Dietmar H. Heidemann (Editor) - Kant Yearbook 3-2011 - Anthropology-Walter de Gruyter (2011)

Kant Yearbook
3/2011
Anthropology
Kant Yearbook
3/2011
Anthropology
Edited by
Dietmar H. Heidemann (University of Luxembourg)
Editorial Assistant:
Katja Stoppenbrink (University of Luxembourg)
Editorial Board:
Henry E. Allison (University of California at Davis), Karl Ameriks
(Notre Dame), Gordon Brittan (Montana State University), Klaus
Düsing (Universität zu Köln), Daniel O. Dahlstrom (Boston Univer-
sity), Kristina Engelhard (Universität zu Köln), Brigitte Falkenburg
(Universität Dortmund), Hannah Ginsborg (University of California
at Berkeley), Michelle Grier (University of San Diego), Thomas
Grundmann (Universität zu Köln), Paul Guyer (University of Penn-
sylvania), Robert Hanna (University of Colorado at Boulder), Georg
Mohr (Universität Bremen), Angelica Nuzzo (Brooklyn College/
CUNY), Robert Stern (Sheffield University), Dieter Sturma (Univer-
sität Bonn), Ken Westphal (University of East Anglia), Marcus Willa-
schek (Universität Frankfurt)
De Gruyter
The Kant Yearbook is an international journal that publishes articles on the philosophy of Imma-
nuel Kant. Each issue is dedicated to a specific topic. Each annual topic will be announced by
way of a call for papers. The Editorial Board of the Kant Yearbook is composed of renowned
international experts, and selects papers for publication through a double blind peer review
process.
Online access for subscribers: http://dx.doi.org/10.1515/kantyb
ISBN 978-3-11-023653-8 (Print)

ISBN 978-3-11-023654-5 (Online)
ISBN 978-3-11-023655-2 (Print+Online)
ISSN 1868-4599 (Print)
ISSN 1868-4602 (Online)
Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.
” 2011 Walter de Gruyter GmbH & Co. KG, Berlin/Boston

Cover image: Martin Zech, Bremen
Printing and binding: Hubert & Co. GmbH & Co. KG Göttingen
⬁ Printed on acid-free paper
Printed in Germany
www.degruyter.com
Contents
Andrew Stephenson
Kant on Non Veridical Experience . . . . . . . . . . . . . . . . . . . . . 1
Thomas Sturm
Freedom and the Human Sciences:
Hume’s Science of Man versus Kant’s Pragmatic Anthropology 23
Liesbet Vanhaute
Systematic Classification or Purposive Moralization?
On why Teleology is not the (only) Key to Kant’s Anthropology
from a Pragmatic Point of View . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Matthias Wunsch
The Activity of Sensibility in Kant’s Anthropology.
A Developmental History of the Concept of the Formative
Faculty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Thomas Wyrwich
From Gratification to Justice.
The Tension between Anthropology and Pure Practical Reason
in Kant’s Conception(s) of the Highest Good . . . . . . . . . . . . . 91
Job Zinkstok
Anthropology, Empirical Psychology, and Applied Logic . . . . 107
Gnter Zçller
Kant’s Political Anthropology . . . . . . . . . . . . . . . . . . . . . . . . . . 131
List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Note to the Studi Kantiani . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Kant on Non Veridical Experience
Andrew Stephenson
Abstract
In this paper I offer an interpretation of Kant’s theory of perceptual error based
on his remarks in the Anthropology. Both hallucination and illusion, I argue, are
for Kant species of experience and therefore require the standard co operation
of sensibility and understanding. I develop my account in a conceptualist frame
work according to which the two canonical classes of non veridical experience
involve error in the basic sense that how they represent the world as being is not
how the world is. In hallucination this is due to the misapplication of categories
and in illusion to the misapplication of empirical concepts. Yet there is also
room in this framework for a distinction in terms of cognitive functionality be
tween the level of experience, which is merely judgementally structured, and
that of judgement proper, which involves the free action of a conscious
agent. This distinction enables Kant to allow for the otherwise problematic
phenomenon of self aware non veridicality.
Introduction
Non veridical experience has been a central topic in epistemology and

philosophy of mind since at least the time of Descartes. Even disregard
ing worries about radical sceptical scenarios in which experience is glob
ally delusive, the residual but trenchant issue of how to deal with local
delusion remains the dividing question in analytic philosophy of percep
tion.1 Yet little has been written that explicitly aims to expound Kant’s
views on the matter.2 My aim in this article is to begin to remedy this
fact. Doing so will not only shed new light on the on going debate con
cerning the fundamental nature of Kant’s theory of human experience;
1 See Crane (2006).

2 One notable exception is Beck (1978), whose account differs markedly from
mine. A recent two part article by Frierson (2009) discusses Kant’s account
of mental disorder at length and also focuses on the Anthropology, but his con
cern is primarily with psychiatric and practical issues. My thanks to Dan Rob
inson for bringing this paper to my attention.
2 Andrew Stephenson
it will also expose certain Kantian tools that have so far gone unexploit
ed in contemporary discussion.
I will not be directly concerned with Kant’s relation to any form of
scepticism, but rather with his account of the particular mechanisms of
non veridical experience. Kant’s most extended remarks concerning
these mechanisms occur in his Anthropology From a Pragmatic Point of
View. Like the topic itself, this is a text that has been almost entirely ne
glected in the literature on the theoretical philosophy,3 and it is another
key aim of this article to begin to remedy this fact as well.
First I outline how sensibility (§1) and the understanding (§2) func
tion in cases of hallucinatory experience. Then, partly by way of shading
in this outline, I explain how Kant’s model also has a place for illusion
(§3), before further addressing the general issue of how it can cope with
cases in which the subject is aware that her experience is in some way
non veridical (§4). Finally I offer some very brief, largely promissory re
marks about the potential consequences of my account (§5).
1.
Hallucination is the experience as of an object when no object is pres

ent. A canonical example is Macbeth’s seeing a dagger before him, the
handle toward his hand. There are plenty of complications if one goes
into the details of the notion but this much will suffice here.
Hallucination, as a species of experience, must involve sensibility.
This lower mental faculty is commonly characterized in terms of passive
receptivity. What it is for us to possess a passive faculty for receptivity is
for us to be able to receive information about the surrounding environ
ment merely through being causally affected by it. And what it is for us
to have representations that belong to this faculty is for such information
to get encoded in mental states that have their sufficient causes in are
nothing but a causal effect of sensory stimulation light being reflect
ed to strike the retina and dissipate molecules of fatty acid, for example.4
3 One notable exception is Brandt (1999).

4 The notion of information encoding might sound rather anachronistic but it is
simply supposed to latch on to the idea that the different representations at this
very low level of mental function have different features in virtue of the differ
ent circumstances in which they are brought about, and that it is these features
that are exploited in various ways by the higher faculties in whatever subsequent
Kant on Non Veridical Experience 3
Kant calls these representations sensations or intuitions, which for pres

ent purposes I will take to be roughly equivalent (though I am far more
neutral on this controversial matter than such a stipulation might sug
gest, which fact I return to in the last section).
However, what is crucial for hallucination is that there is a sense in
which sensibility is not entirely passive. Kant’s model of the human
mind is essentially that of a data processor, with input, function, and
output.5 Sensibility’s role is to provide the input, without which
“there would be no material that could be processed” (Anthr., 144).6
In itself this role dictates nothing about the source of that data. The
bare fact that a data processor needs informational input creates logical
space for a certain restricted degree of underdetermination in this regard.
I say ‘restricted’ because I do not mean to introduce sceptical worries
about malign supernatural implantation or brains in vats. Let us suppose
for the sake of argument that we fix the distal origin of our data as gen
uine causal affection by external physical objects. It nevertheless remains
a possibility that once such data has been originally received once gen
uine causal affection has occurred it could somehow be regurgitated
after the fact and then presented for processing in the normal way.
Such would remain the work of sensibility, understood as data provider,
and yet would require more than sheer passivity. This active element to
sensibility, I propose, is at least one of the roles Kant intends for the
imagination (though it certainly plays others):
Sensibility in the cognitive faculty (the faculty of representations in intu
ition) contains two parts: sense and the power of imagination.—The first is
the faculty of intuition in the presence of the object, the second even with
out the presence of it. (Anthr., 153)
cognitive role such representations are to play. Sellars (1968) talks about this in
terms of counterpart properties.
5 For discussion of this general interpretive approach see Hanna (2001, 14 – 66)
and Longuenesse (1998, 35 – 58).
6 All in text page references are to volume VII of the Akademieausgabe (AA),
which contains Kant’s Anthropologie in pragmatischer Hinsicht (Anthr.). References
to other of Kant’s works are restricted to footnotes and, with the exception of
the Kritik der reinen Vernunft (CPR), are given by volume and page number of
the Akademieausgabe along with a short English title. References to the Kritik
take the standard A/B format. I follow the English translations of The Cambridge
Edition of the Works of Immanuel Kant but have made modifications where
deemed appropriate. The details of the particular volumes I have used from
the Cambridge collection are contained in the bibliography.
4 Andrew Stephenson
The power of imagination (facultas imaginandi), as a faculty of intuition even

without the presence of the object, is either productive, that is, a faculty of
the original presentation of the object (exhibitio originaria), which thus pre
cedes experience; or reproductive, a faculty of the derivative presentation of
the object (exhibitio derivativa), which brings back to the mind an empirical
intuition had previously. (Anthr., 167) 7
And Kant even goes on to point out the ultimately parasitic nature of
the imagination in this naturalistic guise:
The power of imagination (in other words) is either inventive (productive),
or merely recollective (reproductive). But the productive power of imagina
tion is nevertheless not exactly creative, for it is not capable of producing a
sense representation that was never given to our faculty of sense; one can
always furnish evidence of the material of its representations. To one
who has never seen red among the seven colours, we can never make
this sensation comprehensible… the sensations produced by the five senses
in their composition cannot be made by means of the power of imagina
tion, but must be drawn originally from the faculty of sense. (Anthr.,
167 – 8) 8
The imagination normally plays such a role in cognitive processes that
are entirely epistemically legitimate, like memory (Anthr., 182 5),9
but there is room here for an altogether more pernicious function
in hallucination.
Consider the following picture. In veridical experiences there is an
object that is causally affecting us at the time of the experience and in
virtue of which act of sensing we acquire a manifold of intuitions. In
hallucinatory experience, however, there is no such object, no such causal
affection, and no such act of sensing. Instead there is the imaginative re
call of a previously acquired manifold (perhaps even one that has been
somehow carved up and combined with other parts of previously ac
quired manifolds). It is this (perhaps gerrymandered) manifold that is
then presented for intellectual processing. What results is an experience
7 Cf. Metaphysics Mrongovius (AA 29:881), Metaphysics Volckmann (AA 28:449),

Metaphysics L2 (AA 28:585), Metaphysics Dohna (AA 28:672).
8 Cf. CPR A 770 – 1/B 798 – 9, and the Vienna Logic (AA 24:904).
9 It would be entirely appropriate to be reminded of Hume (1978) in observing
this connection to memory, as there is much in the following account that par
allels the role he attributes to the imagination. I cannot go into this topic here
but I should note that it differs significantly from the aspect of the relation be
tween Kant and Hume’s theories of imagination discussed, for example, in
Strawson (1971).
as of whatever object originally caused the relevant intuitions, but with

out the actual and current presence of that object.10
Furthermore it seems entirely possible that things might go awry in
this way without us noticing or even being able to notice. On the data
processor model there might be nothing intrinsic to the data we process
that tells us whether it is the result of current affection or rather of some
more surreptitious procedure. Again we need not go as far as entertain
ing radical sceptical hypotheses. Indeed, we might even concede that
there is always something intrinsic to the data that could tell us this
much, at least when we take a large enough collection of it. It would
still seem entirely possible that this feature is not always accessible to
us, and this is sufficient to motivate the present worry. Kant himself ges
tures towards this issue of the apparent possibility of subjective indistin
guishability:
[…] the power of the imagination, which puts material under the under
standing in order to provide content for its concepts (for cognition),
seems to provide a reality to its (invented) intuitions because of the analogy
between them and real perceptions. (Anthr., 169) 11
At the level at which representations are processed, there might, at least
under certain circumstances, be no way to tell the proximal origin of
those representations. Or to put it another way, the higher faculties can
not as it were reach down into sensibility in order to check which part
of it, sense or imagination, is currently activated.
So far I have talked as though the imagination provides information
for processing directly, bypassing passivity altogether. In fact we need to
complicate this picture slightly because of certain things Kant says in this
context about a division in the passive part of sensibility:
10 Something very similar is implied at CPR B 278. Allais (2010, 59) points out
that Kant talks about intuitive representation there, rather than intuition as
such, but in light of the passages we have just seen I cannot agree that this is
particularly significant. Rather we should look more carefully at the oft cited
claim in the Prolegomena (AA 4:281 – 2) that “An intuition is a representation
of the sort which would depend immediately on the presence of an ob
ject”—note the subjunctive tone and the fact that Kant goes on in the (rarely
cited) next sentence to talk about an object’s presence “either previously or
now”.
11 He is more explicit in the Prolegomena (AA 4:290): “The difference between
truth and dream, however, is not decided through the quality of the represen
tations that are referred to objects, for they are the same in both”. Cf. also CPR
B 279.
6 Andrew Stephenson
The senses, however, are in turn divided into outer and inner sense (sensus
internus). The first is where the human body is affected by physical things;
the second, where it is affected by the mind. (Anthr., 153)
Suppose, then, that some passive part of sensibility is always involved in
experience outer sense for veridical cases and inner sense for halluci
nation. After all, it seems right to say that hallucination involves actually
undergoing sensuous modifications of the mind, rather than merely a
punctiform episode of information recall. On this approach, the imag
ination fully replaces the object and is attributed similar causal powers.
What distinguishes hallucination from veridical experience at this level
of mental function is not the bypassing of the senses. It is simply that
it is the imagination rather than the object that fulfils the role of causal
instigator in cases of hallucination.
The addition of this intermediary step has the benefit of emphasiz
ing the important fact that the kind of imaginational activity currently
under discussion is very different to intellectual activity. And it also sug
gests an intuitive alternative role for the otherwise difficult notion of
inner sense. In other works Kant focuses on distinguishing inner sense
from apperception and relating this distinction to the epistemic humility
of transcendental idealism.12 But this matter “does not really belong to
anthropology” (Anthr., 142), so in his lectures on this subject he talks
instead about
[…] taking the appearances of inner sense for external appearances, that is,
taking imaginings for sensations… it is mental illness: the tendency to ac
cept the play of representations of inner sense as experiential cognition, al
though it is only fiction […] and accordingly to trick oneself with the in
tuitions thus formed (dreaming when awake). (Anthr., 161)
Indeed I think it is generally fair to say that the account I have outlined
brings us closer than most to the idea of imagination many of us would
have prior to reading the Transcendental Deductions.
2.
We have seen what has to happen at the level of sensibility in order for
hallucination to occur, but this is not the whole story. Hallucination is
one kind of output of our cognitive system, and while we can suppose
12 See e. g. CPR B 153 – 6.

that there is a perfectly robust sense in which the input of sensibility con-
strains this output, it does not alone determine it. Different input can pro
duce different output, but this can also depend on the nature of the
functions that map one onto the other.13 In Kantian terms, as a species
of experience, hallucination must also involve our higher mental faculty,
the understanding.
More specifically, hallucination is a normative phenomenon at least
insofar as it involves some kind of mistake. Yet sensibility, both in its
passive and active part, functions entirely within the natural realm. Its
representations are the immediate, unprocessed result of causal affection,
and whether this occurs because of the object or because of the imag
ination, it is a natural event and in no sense assessable for correctness.
Kant makes this absolutely crucial point in the following striking pas
sage:
The senses do not deceive […] not because they always judge correctly, but
rather because they do not judge at all. Error is thus a burden only to
the understanding.—Still, sensory appearances (species, apparentia) serve to ex
culpate, if not exactly to justify, understanding. (Anthr., 146) 14
It is the intellectual activity of the understanding its act of information
decoding that first brings us into a normative realm at all. Kant’s ex
plicit ought talk is largely restricted to the practical sphere,15 but his
core machinery makes a certain structural connection to a certain
kind of normativity clear: concepts can be correctly or incorrectly ap
plied and judgements can be true or false. It is true that so far this is
an extremely bare kind of normativity, in effect exhausted by truth con
ditions. Indeed one might even question the wisdom of calling it nor
mativity in the first place and later we will have cause to turn briefly to a
far richer and more uniquely Kantian notion. But it is significant
enough for now that it is only once the understanding and its cognitive
machinery becomes involved that representation, and thereby misrepre
sentation, is possible. The product of sensibility functioning alone does
13 This is one way of stating the problem that McDowell (1996) and (1998) has
addressed under the Sellarsian rubric of the Myth of the Given. More specifi
cally that functions could produce output, if not without input, then in some
sense regardless of it.
14 Cf. CPR A 293/B 350, Prolegomena (AA 4:290 – 1), Metaphysics Mrongovius (AA
29:759, 833).
15 Though see CPR A 135 – 6/B 174 – 5. My thanks to Ken Westphal for the
pointer, who cites this passage in his (2004, 168).
8 Andrew Stephenson
not represent the world as being a certain way at all, and so cannot be
mistaken in even this very basic sense.
So exactly how does hallucination involve a mistake? Well in a nut
shell the Kantian intellectual machinery works as follows: the concept of
an object provides the rule in accordance with which the understanding
processes the data it receives from sensibility; what it is for the under
standing to carry out this process is for it to interpret that data as collec
tively indicative of the perceptual presence of an object; and the expe
rience that is the result of this process has what I will call a judgemental
counterpart something like ‘that there is an object here and now’
which is in effect the experience’s content. Hallucination involves a
mistake, then, in the sense that the concept of an object is misapplied,
the data provided for intellectual processing in some sense should not
be interpreted as indicative of the perceptual presence of an object,
and the resultant judgemental counterpart is false.
Note that it is crucial for the application to hallucinatory cases that
this judgemental counterpart to experience does not depend for its sense
on the actual existence of its object. If there is in fact not an object here
and now, then the claim that there is an object here and now is false, but
at least it is meaningful. This application proscribes singularity of refer
ence in virtue of demonstrative type reference to a particular object; as
suming, that is, that judgements that refer in such a way judgements to
the effect that that object is here and now are not merely false but fail
to have any determinate truth value when they fail to refer because the
object does not exist. But it is consistent with singularity of reference in
virtue of spatiotemporal indexicality. So there is something analogous to
demonstration about the judgemental counterpart to experience. It is
just that it is not the object part that sustains the analogy. After all,
the concept of an object is as general and descriptive as any other con
cept (as we will see in the next section). Rather what sustains the anal
ogy are the spatiotemporal indexicals, and these are immune to refer
ence failure regardless of what objects happen to exist at the time
and place of an experience, veridical or no, there is always a time and
place to that experience.16
16 My thanks to Arthur Melnick for prompting me to clarify this point and sug
gesting the particular locution I have adopted. A similar position, though with
differences concerning the Kantian jargon, is developed in depth by Howell
(1973). It contrasts with that of Thompson (1972).
Moreover, I take it that this reading of the judgemental counterpart

to experience as the claim that there is an object here and now need
not necessarily fly in the face of any facts about ordinary linguistic prac
tice. Even if we do appear to use demonstratives in reporting on what
we experience ‘this object exhibits such and such features’ it is not at
all obvious that these reports really express any object dependent con
tent. And it would be thoroughly in keeping with the current proposal
that the demonstrative locution in such cases simply reflects a general
default assumption that experience is veridical (as we will see in §4).
The thought is that we tend to talk about our experiences as though
they are object dependent simply because we tend to believe that
there is an object in the world that we are experiencing; that we tend
to act as though it would not be a problem if our experience were ob
ject dependent because we tend to presume we are experiencing an ac
tual object. But that this practice and this assumption exist entails neither
that the content of our experience is in fact object dependent nor even
that it is on reflection somehow more intuitive to say that it is. Nor, of
course, was ordinary linguistic practice of particular concern to Kant.
This account of hallucination remains fairly skeletal. One way to
flesh it out further would be to consider some salient variations on
the theme. For example, how does Kant’s model accommodate cases
of partial hallucination? If only the dagger is unreal, and the courtyard
in which Macbeth sees it situated is real, then it seems that his under
standing cannot be entirely wrong in all its conceptualizing activity.
And what about coincident hallucination? Suppose that what I halluci
nate just so happens to coincide with the actual state of my environ
ment, that the data my imagination provides just so happens to be ex
actly the same as the data my outer sense would have provided had I
been experiencing veridically. Something has clearly gone wrong, and
my experience will likely not be a suitable basis for knowledge, but con
cepts are instantiated and judgemental counterparts come out true. Fi
nally, how can a subject hallucinate something she has never come
across? Presumably it is possible to hallucinate a unicorn without ever
having sensed one, but then the imagination even in this naturalistic
manifestation must do something a little more than mimic whole senso
ry manifolds. I am confident that Kant has as sophisticated resources as
any to deal with standard complications like these. Some have already
been mooted and some will come out in what follows. What I want
to do now, however, is move on to the second core class of non verid
ical experience, illusion.
10 Andrew Stephenson
3.
For present purposes I will treat illusions as cases in which an object is

experienced as possessing properties it does not as a matter of fact pos
sess. A canonical example is seeing a straight stick half submerged in
water and its looking bent. To understand Kant’s account of illusion
we need to turn in more detail to the role of the concept of an object.
There is a crucial ambiguity in this notion between a representation
that involves no empirical elements and a representation that involves
some empirical elements, albeit alongside pure elements. It is specifically
the concept of an object in general that involves no empirical elements.
This is the concept articulated by the categories. There are both sche
matized and unschematized versions of it, depending on whether it is
separated from our particular modes of sensibility, space and time.
The unschematized concept of an object in general is not my concern
here. It is something like the concept of a whole with parts that enjoys
some more or less limited reality and which can stand in some kind of
ground and consequence relation to other such things. Many things
qualify as objects in this sense, including noumena and Platonic abstrac
ta. The schematized concept of an object, on the other hand, is more
restrictive. It is something like the concept of a spatiotemporally extend
ed and located, sensible property possessing, fully causally functioning,
really existing particular, event, or state of affairs. So the categories are
mobilized wholesale, as collectively articulating this concept of an object
in general.
Paradigmatic of such objects are medium sized dry goods, like tables
and chairs. Yet what makes it the concept of an object in general is that
what further empirical concepts are instantiated is left undetermined. If
something is cognized as an object in general then it could but is yet to
undergo further descriptive specification according to what particular
sensible properties it happens to possess, such as that of being rectangular
or that of being wooden or that of being a table. Once one starts to
specify in this way then one is no longer working solely with the
pure, categorial concept of an object in general. It becomes, for example,
the concept of an object that is rectangular and wooden. The categories are
still involved since whatever instantiates the species instantiates the
genus but there is now an admixture of the empirical as well.17
17 Note, however, that even prior to empirical specification the pure concept of
an object, in both its schematized and unschematized versions, is descriptive.
This point also has its judgemental counterpart. The judgement that
there is an object here and now does not involve any concepts other
than the categories (assuming that whatever spatiotemporal determina
tion it involves is non conceptual). But it could be empirically specified
without altering its basic structure simply by adding predicates. Our ex
perience might be as of a rectangular wooden object here and now, to
which the corresponding judgemental counterpart would be that
there is an object that is rectangular and wooden here and now.
The relevance of this distinction for hallucination is that things go
wrong at the categorial level, before any empirical specification takes
place, for there is nothing perceptually present that is spatially located
and extended, fully causally functioning, or really existing.
But what has this got to do with illusion? Like hallucination, illusion
is at a normative (=representative) level of cognitive output and in
volves a mistake of some kind. To continue a passage from which I
quoted earlier:
Thus the human being often mistakes what is subjective in his way of rep
resentation for objective (the distant tower, on which he sees no corners,
seems to be round; the sea, whose distant part strikes his eyes through high
er light rays, seems to be higher than the shore (altum mare); the full moon,
which he sees ascending near the horizon through a hazy air, seems to be
further away, and also larger, than when it is high in the heavens, although
he catches sight of it from the same visual angle). And so one takes appear
ance for experience; thereby falling into error, but it is an error of the under
standing, not of the senses. (Anthr., 146) 18
But unlike hallucination, this mistake does not occur at the level of cat
egorization. In cases of illusion, the application of the categories, bare as
they are, is entirely correct. Unlike hallucination, there is in cases of il
lusion an occurrent act of outer sensing, so there is indeed in the subject’s
perceptual presence something that instantiates the concept of an object
in general, that of a spatiotemporally extended and located, sensible prop
erty possessing, fully causally functioning, really existing particular,
Indeed I take this to be a definitional characteristic of concepts per se. It there

fore stands in place of the predicate rather than the variable in sentences like
‘there is an x, such that Fx’. If anything plays an analogous role to that of
the variable here—and we have to be very careful in translating into an anach
ronistic idiom—it is intuition. (See the comment below on how to articulate
the principled distinction between hallucination and illusion.)
18 Kant gives further examples of illusions at p. 137, which include the observa
tion that “white stockings present fuller calves than do black ones”!
event, or state of affairs. Illusion, then, goes awry at the level of empirical
specification, when a subject ascribes properties like roundness, high
ness, farness and largeness.
Note, however, that it is not empirical specification per se that causes
the problem. When I experience a straight stick as bent, I am correct to
apply the concept of an object that is a stick. This concept remains ge
neric enough to be instantiated by the object instigating the illusory ex
perience. What I ought not do is apply the still more specific concept of
an object that is a straight stick.
And in the other direction, note that I have not expressed the dis
tinction between hallucination and illusion schematically. For it would
not be correct to do so in terms of the difference between ‘there exists an
x’ (for hallucination) and ‘there exists an x such that x is F’ (for illusion),
since for all this dictates ‘F’ could be a pure categorial predicate like
‘causally functioning’ or ‘property possessing’ rather than an empirical
one like ‘high’.
Sticking with the simple formulation in terms of error at the catego
rial level and error at the empirical level, then, on the face of it this looks
to be a plausible and principled way to draw the distinction. This is pri
marily because it entails that actual mental malfunction only occurs with
hallucination, which is important due to the utter ubiquity of illusion in
comparison to hallucination. Despite the fact that the notion of a mis
take is a notionally normative one and we are in some basic sense at a
normative level of processing here, in both classifications of non verid
ical experience it is ultimately the natural realm that determines whether
or not a mistake has occurred; the question of whether or not we are
undergoing non veridical experience, be it of the hallucinatory or illu
sory variety, ultimately depends on how the world is. Indeed this simply
indicates the fact that at the current level of cognitive function the
norms in play can be fully articulated by a list of truth conditions.
But in hallucination, and crucially only in hallucination, there is a
cause that we can locate more specifically in sensibility. In illusion, sen
sibility is functioning as usual and we are sensing in exactly the same way
as we do when we experience veridically it is the object and outer
sense that are involved rather than the imagination and inner sense. Un
like Macbeth seeing a dagger, there is nothing wrong with the subject
who sees a stick looking bent when half submerged in water.
Still, we need to elaborate a little, for going on what we have so far
one might worry that cases of radical misrepresentation will get incor
rectly classified. If a subject is looking at a chair but seeing an elephant,
intuitively it would be correct to say that she is hallucinating. This kind

of case seems radically different to that in which someone sees a white
chair as red due to nonstandard light conditions. But the categories
might appear to remain applicable. Should it then be classified as illu
sion? In fact the current model very naturally suggests how we are to
deal with such cases.
The categories are applied uniformly across all manifolds irrespective
of contingencies of content, for all empirical intuitions are as of objects
in general. Whether I perceive a bird or a tree or a bird in a tree or a
bird flying out of a tree, for example, I perceive a spatiotemporally lo
cated and extended, sensible property possessing, fully causally function
ing, particular, event, or state of affairs. With empirical concepts how
ever, just which ones are mobilized in the production of any given ex
perience is determined by the particularities of the relevant intuitions.
Not all intuitions are caused in us by red things, for example, so not
every experience will require us to mobilize the empirical concept of
red or be as of an object that is red. In this way the activity of the under
standing counts as radically spontaneous in its categorization of the sen
sible manifold, and yet at the same time can remain constrained from
below insofar as it also consists in the application of empirical concepts.19
In the kind of case proffered above, the bottom up constraint is
clearly missing, and this indicates that there has indeed been some men
tal malfunction, that there has been a break in the mechanistic chain. A
good explanation of such a case would require positing the surreptitious
working of the imagination, for the fact that there is a chair present is
explanatorily irrelevant. The categories do not remain applicable because
they have not been applied to a manifold caused by that chair at all. In
this way such cases would qualify (correctly) as hallucinations.20
19 See Watkins (2008) for an extended treatment of this anti McDowellian point.
Note also that I take this characterization of spontaneity to be entirely compat
ible with the more standard one (which does not distinguish between the em
pirical and pure manifestations of the understanding): deploying spatiotemporal
concepts—i. e. both empirical ones, which are automatically spatiotemporal,
and pure ones, which are spatiotemporal when schematized—does not carve
the world of things in themselves at its joints.
20 Further complications include Müller Lyer cases, in which it is also far from
clear that there is ‘nothing wrong’ with susceptible subjects.
4.
Having addressed the mechanisms involved in both canonical classes of

non veridical experience and utilized distinctive Kantian tools to sug
gest a principled distinction between them, we need to confront an
issue that applies generally in this context. Any vaguely credible account
needs to allow room for cases in which the subject is aware that her ex
perience is non veridical. The motivation is not merely that non verid
ical experience is rarely if ever actually subjectively indistinguishable
from veridical experience. Even if this were common, the fact would
remain that our ulterior knowledge of the way the world works
could well inform how we decide to take the evidence of our senses.
Yet at first it might seem as though Kant cannot cope with this simple
and ubiquitous phenomenon, for he ties experience too closely to
judgement. If I know that my experience is in some way non veridical
when, for example, I see a stick half submerged in water looking bent,
then surely I ought not to judge that there is a bent stick here and now.
Indeed it seems plausible that I ought to judge that it is not the case that
there is a bent stick here and now. The worry is that something like a
Moorean paradox arises for subjects who do what they ought in such
cases; that in cases of self aware non veridicality rational subjects
judge that the stick is bent and that it is not bent. And this worry is
not specific to illusion. The same would apply if Macbeth decided
that what he saw before him was in truth a dagger of the mind, a
false creation, proceeding from the heat oppressed brain. Fortunately,
this objection underestimates the complexity of Kant’s model.
I have talked about judgemental counterparts to distinguish the kind
of thing that accompanies experience from what we might call full
blown judgement. The basic thought is that these are significantly dif
ferent kinds of representation, occurring at different levels of cognitive
engagement, indeed with different attendant kinds of normativity, and
that this allows room for a subject to be able, coherently, to make full
blown judgements that contradict her experiential judgemental counter
parts. This needs some explanation.
By full blown judgement I mean the kind of unification of concepts
that Kant so famously divides into analytic and synthetic. Both full
blown judgement and the judgemental counterpart to experience qual
ify as judgement because they constitutively involve concepts and are
objectively valid in the sense that they are truth evaluable. However,
full blown judgement constitutively involves only concepts. Experience
and its judgemental counterpart, on the other hand, constitutively in

volve concepts and intuitions in experience we unify intuitions with
concepts, not just concepts with concepts. The involvement of concepts
at all is enough to secure a kind of judgemental form, but the constit
utive involvement of intuitions ensures that it is of quite a unique
kind. The most important and general consequence of this has to do
with the relative primitiveness of experience in comparison to full
blown judgement. There is a definite hierarchy here. Experience and
its judgemental counterpart justify or provide rational ground for full
blown judgement. They are not equivalent to it. For example, the
full blown synthetic a posteriori judgement that some sticks are bent
would normally garner justification from the experience as of a bent
stick and the judgemental counterpart that there is a bent stick here
and now. But this relation only holds firm in epistemically suitable sit
uations. It can readily be overridden by a variety of countervailing fac
tors, as when a subject is fully aware that putting straight sticks in water
does not bend them.
Another way to approach this issue, and one that will bring us back
to the text of the Anthropology, is via the notion of consciousness. Kant
defends the explicitly anti Lockean doctrine that we can ‘have represen-
tations and still not be conscious of them’. In particular:
The field of sensuous intuitions and sensations of which we are not con
scious, even though we can undoubtedly conclude that we have them;
that is, obscure representations in the human being (and thus also in animals),
is immense. Clear representations, on the other hand, contain only infin
itely few points of this field which lie open to consciousness; so that as
it were only a few places on the vast map of our mind are illuminated.
(Anthr., 135)
It seems as though Kant virtually defines intuitions and sensations as ob
scure in this passage, and experience as clear, for remember that whole
manifolds of information encoding natural mental states must be con
ceptually unified by the understanding in order to produce a single rep
resentational experience as output. If so, we are not given manifolds in
any sense that allows us to consciously choose how to conceptualize
them we do not consciously decide how to process our data set,
and although the activity of the understanding is more intellectually so
phisticated than the activity of the imagination, it is not as sophisticated
as voluntary action.21
21 Bennett (1974, 19) makes this mistake, presumably unaware of Kant’s discus
One could go further and say that we are not normally conscious at
the level of sensibility at all and that the understanding is itself partly
constitutive of normal human consciousness. But this is a difficult and
controversial issue.22 Fortunately, all that is essential here is the much
weaker claim that it is only on the basis of consciousness as of objects
that we do things like make decisions. This is enough to produce the
required hierarchy between full blown judgement on the one hand
and experience and its judgemental counterpart on the other. For one
especially salient way to make a decision is to make a full blown judge
ment about how the world is. Now that is not to say that experience is
entirely neutral in this regard. After all, it presents us with the world as
being thus and so, not merely as possibly being thus and so. In modern
parlance we might say that experience does not simply provide us
with bare propositional content but also prompts us to adopt an attitude
of belief towards it; or that experience both supplies accuracy conditions
and suggests that they are fulfilled. But what is crucial is that experience
has this manner even if we know that the world is not thus and so. As
Kant says, “Illusion is that delusion which persists even though one
knows that the supposed object is not real” (Anthr., 149).23 And it is pre-
cisely because we have no choice in the matter that we fall short at this
level of making the kind of commitment typical of full blown judge
ment. Under epistemically optimum conditions we can readily consent
to asserting that the way the world is presented as being in experience is
indeed the way the world is. Yet we could also resist this if there were
reason to do so. The conscious strengthening of propositional attitude,
in either direction, represents a decision and constitutes a cognitive shift
up from the level of experience and its judgemental counterpart to the
level of full blown judgement.
Note that the claim is not that every judgement actually involves
something that could be correctly described as a choice. This would
be implausible, for judgements can surely be as unbidden and uncon
sion in the Anthropology of ‘the involuntary course of one’s thoughts and feel
ings’ (Anthr., 133).
22 Kant certainly acknowledges some form of consciousness at the level of bare sen
sation (see e. g. CPR B 207, A 320/B 376), but it is not at all clear what role it
plays or how we should characterize it. Sellars (1975) and George (1981) offer
deflationary adverbial accounts.
23 Cf. CPR A 293/B 349 A 298/B 355, where Kant discusses the fact that non
veridical experience shares this feature with the transcendental illusion against
which his critical method is designed guard.
trolled as experience. Rather the claim is that every judgement possibly

involves this kind of thing. But this is enough to bring with it a much
richer and more thoroughly Kantian kind of normativity. The mere
possibility of exercising a certain freedom makes judgement normatively
binding in a way that experience is not. Ultimately, in fact, it means
nothing less than that judgement is governed by a categorical impera
tive. But all that is crucial here is that, according to Kant’s model of
the human mind, the case of self aware non veridicality is not analogous
to a Moorean paradox at all. It is far more like the case of deciding to do
one thing whilst nevertheless being tempted to do another, and in this
there is no hint of irrationality, merely humanity.
5.
If the account I have given in the preceding four sections were com
plete, there would be immediate repercussions for several central dis
putes in contemporary Kant scholarship. In particular I am thinking
of two issues that draw heavily on recent work in analytic philosophy
of perception.
First of all, the account I have sketched is clearly not naive realist.24 I
mean by naive realism the particular form of disjunctivism about hallu
cinatory experience that maintains it is of a fundamentally different kind
to veridical experience because veridical experiences have as metaphysical
constituents their real world objects. Neither intuition nor experience,
as I have understood these notions, are relational in the sense that they
require the actual and current existence of their objects. For we have
seen that one species of experience is hallucination, and that the repre
sentations produced by the imagination’s affect upon inner sense therein
are a species of intuition. Though we have also seen that hallucination is
essentially reproductive, that it is ultimately parasitic upon genuine caus
al interaction with the external environment, so there does seem to be at
work some weaker, externalist sense in which intuition and experience
depend on that environment. And there might even be room for weak
er forms of disjunctivism, though this is not least because of the broad
ness of disjunctivism’s criterion of fundamental difference.25 Despite the
similarity of structure and constituent on the current picture, it is not
24 Contra e. g. Hanna (2010) and Allais (2010).

25 See Soteriou (2010).
obvious that it should not count as a fundamental difference between

veridical and hallucinatory experiences that only in the former are we
in direct causal contact with actual objects, especially if we consider
that this issue might well partially determine whether or not the episode
in question can be theoretically and practically assimilated into a single,
law governed picture of the world.
Second, as I have explained it here Kant’s model is not one that ac
tively makes room for non conceptual world oriented content. Very
roughly, the content of a mental state is what is conveyed to the subject
who is in that state; this content is world oriented if it is intentional in
some suitable sense and puts the subject in cognitive touch with the
world in some suitable way; and it is non conceptual if it is not itself
conceptually structured, for example in a way analogous to the semantic
structure of a proposition, and does not itself constitutively involve con
cepts.26 Going solely on what I have said above, Kant thinks that mental
content is either non conceptual but not world oriented sensation/in
tuition/information encoding natural representation or world orient
ed and conceptual experience. For all we have seen here, there is no
middle ground.27
However, the account I have given is clearly not complete there is
clearly much more to say. And so in light of this I in fact make no claim
to have established either that Kant was not non conceptualist about
some world oriented content or that none of the Kantian classes of rep
resentation are object dependent in the naive realist, relationalist sense.
To positively rule these out would require much fuller treatments of all
the key notions of ‘imagination’, ‘intuition’, ‘concept’ and ‘experience’,
not to mention ‘synthesis’ (which I have talked about in terms of proc
essing and function). But I do think anyone would be forced to admit
that Kant uses each and every one of these absolutely pivotal terms in
different ways at different points, and the research questions would in
clude: Is there a significant usage of ‘intuition’ where it denotes some
26 In fact there is a much weaker sense in which one might think content could
qualify as non conceptual, namely if it does not require the subject to possess
concepts which would enable her to interact with it in certain ways, such as
describe it. I cannot discuss this influential alternative in these brief remarks, ex
cept simply to say that it does not look especially Kantian (see Hanna (2010) for
quasi independent reasons to suspect its robustness as a form of non conceptu
alism).
27 Contra e. g. Hanna (2008), Hanna (2005) and Allais (2009), and in line with
Ginsborg (2008).
thing in between a spatiotemporally ordered manifold of sensations and

full cognitive experience? Or is there a significant usage of ‘experience’
according to which it remains objective but does not require the under
standing? Ought we to impose a principled distinction between intu
itions and imaginings, even though Kant simply does not make one ter
minologically explicit and regular? And is there a much richer role for
the imagination that yet does not go as far as rendering it subordinate to
the understanding? I do not claim to have answered or even addressed
any of these questions.
Nevertheless, there are several points at which my account, such as
it is, goes against the explanatory grain of the positions under consider
ation. For example, I have shown that Kant’s model has the resources to
provide quite a rich bottom up explanation of both canonical classes of
non veridical experience as well as the general potential for subjective
indistinguishability. Modern day naive realists, on the contrary, tend
to use subjective indistinguishability as a primitive, and they do so in
order to offer a much more austere definition of non veridical experi
ence, as merely that which is subjectively indistinguishable from verid
ical experience. And concerning the issue of non conceptualism, one of
its key motivations is the idea, which is very natural indeed, that expe
rience is more basic than judgement, that seeing is not believing. Yet by
considering how Kant could deal with self aware non veridicality we
have seen that he makes room for this distinction within varieties of con
ceptually structured representation.
Moreover, the way I have explained the Kantian concept of an ob
ject in general as collectively articulated by the categories can be put to
work in a similar regard. A key argument in favour of the relational
view of experience is that it is necessary to explain how we can derive
concepts from experience, and in particular the concept of mind inde
pendence. Now it is a complex matter as to just what this concept
amounts to in the context of transcendental idealism, but suppose that
mind independence is entailed by having a spatiotemporal extension
and location, being fully causally functioning, and being really existing,
which seems highly plausible. Then we have seen that for Kant this is a
concept that is precisely not derived from experience. Rather it is mo
bilized as a necessary and constitutive condition of it. Kant’s innatism in
this regard discharges much, though admittedly not all, of that essentially
empiricist motivation for the relationalist view.28 And this tool might
also be used to undercut a significant class of arguments in favour of
non conceptualism. For the view I have attributed to Kant is only mod-
erately conceptualist in the sense that, strictly speaking, only the pure cat
egories and not empirical concepts are necessary for experience. Savages
can cognitively experience houses as objects in general without having
any conception of what a dwelling established for men would be, and
although they could not thereby describe the objects of their experien
ces as such, they could certainly track those objects through spacetime
and even causally reason about them in some limited way.29
Still, these are really nothing more than promissory notes. I leave a
fuller discussion for another occasion.
Conclusion
In the preface to his Anthropology From a Pragmatic Point of View, Kant

defines the discipline as “the investigation of what he [the human
being] as a free acting being makes of himself, or can and should
make of himself” (Anthr., 119). So formulated, it sounds very broad.
The references to autonomy and normative injunction imply that it is
to include moral philosophy, and this relationship has been much dis
cussed. Less attention has been paid to the fact that such marks are
equally apt to pick out key aspects of Kant’s theoretical philosophy.
The spontaneity exhibited by the human mind in its application of con
cepts (in particular the pure ones), the freedom the conscious agent can
express in deciding what to make of her experiences, and the rich and
complex normative structure of the realm we thereby enter, are all
foundational issues in Kant’s mature philosophy of mind and epistemol
ogy. And though we have seen that the spectre of non veridicality is
built right in to the nature of the human mind on Kant’s model,30 it re
mains the case that we ought not be seduced into making false judge
28 See Gomes (2010) for a much more in depth treatment of this topic. His target
is Campbell (2002).
29 Cf. the Jsche Logic (AA 9:33) (and the Dohna Wundlacken Logic (AA 24:702)).
30 In fact this suggests another structural connection to anthropology (even more
broadly construed), for it is Kant’s anthropic turn that leads him to his data pro
cessor model in the first place. Contrapositively put, God could not undergo
non veridical experience. There is an intriguing passage in the Critique of Judg
ment (AA 5:401 – 3) that can be read as elaborating on this connection.
ments on the basis of it. We should make of ourselves beings who agree
with one another, and are right, about the world.31
Bibliography
Allais, Lucy (2009): Kant, Non Conceptual Content and the Representation of
Space, in: Journal of the History of Philosophy 47, pp. 383 – 413.
Allais, Lucy (2010): Kant’s Argument for Transcendental Idealism in the Tran
scendental Aesthetic, in: Proceedings of the Aristotelian Society 110,
pp. 47 – 75.
Arens, Patrick E. (2010): Kant and the Understanding’s Role in Imaginative
Synthesis, in: Kant Yearbook 2, pp. 33 – 52.
Beck, Louis W. (1978): Did the Sage of Konigsberg Have No Dreams?, in: L.
W. Beck (ed.): Essays on Kant and Hume, London, pp. 38 – 60.
Bennett, Jonathan (1974): Kant’s Dialectic, Cambridge.
Brandt, Reinhard (1999): Kritischer Kommentar zu Kants Anthropologie in
pragmatischer Hinsicht, Hamburg.
Campbell, John (2002): Berkeley’s Puzzle, in: J. Hawthorne and T. S. Gendler
(eds.): Conceivability and Possibility, Oxford, pp. 127 – 143.
Crane, Tim (2006): Is There a Perceptual Relation?, in: T. S. Gendler and J.
Hawthorne (eds.): Perceptual Experience, Oxford, pp. 126 – 146.
Frierson, Patrick (2009): Kant on Mental Disorder. Part 1: An Overview;
Part 2: Philosophical Implications of Kant’s Account, in: History of Psy
chiatry 20, pp. 267 – 310.
George, Rolf (1981): Kant’s Sensationism, in: Synthese 47, pp. 229 – 255.
Ginsborg, Hannah (2008): Was Kant a nonconceptualist?, in: Philosophical
Studies 137, pp. 65 – 77.
Gomes, Anil (2010): Kant and the Explanatory Role of Experience, unpublish
ed MS.
Hanna, Robert (2001): Kant and the Foundations of Analytic Philosophy, Ox
ford.
Hanna, Robert (2005): Kant and Non Conceptual Content, in: European
Journal of Philosophy 13, pp. 247 – 290.
Hanna, Robert (2008): Kantian Non Conceptualism, in: Philosophical Studies
137, pp. 41 – 64.
Hanna, Robert (2010): The Rational Human Condition, unpublished MS.
Howell, Robert (1973): Intuition, Synthesis, and Individuation in the Critique
of Pure Reason, in: Nous 7, pp. 207 – 232.
Hume, David (1978): A Treatise of Human Nature, eds. L. A. Selby Bigge and
P. H. Nidditch, Oxford.
31 In addition to those already mentioned, I would like to thank Keith Wilson and
the other attendees of the Warwick work in progress seminar, an anonymous
reviewer from the Kant Yearbook, Bob Hanna, Anil Gomes, and above all
Ralph Walker for such useful comments on earlier drafts of this article.
Kant, Immanuel (1998): Critique of Pure Reason, ed. and trans. P. Guyer and
A. W. Wood, Cambridge.
Kant, Immanuel (2000): Critique of the Power of Judgment, ed. P. Guyer,
trans. P. Guyer and E. Matthews, Cambridge.
Kant, Immanuel (2001): Lectures on Metaphysics, ed. and trans. K. Ameriks
and S. Naragon, Cambridge.
Kant, Immanuel (2002): Theoretical Philosophy after 1781, eds. H. Allison and
P. Heath, trans. G. Hatfield, M. Friedman, H. Allison and P. Heath, Cam
bridge.
Kant, Immanuel (2004): Lectures on Logic, ed. and trans. J. M. Young, Cam
bridge.
Kant, Immanuel (2007): Anthropology, History, and Education, ed. G. Zöller
and R. B. Louden, trans. A. Gregor, P. Guyer, R. B. Louden, H. Wilson,
A. W. Wood, G. Zöller, A. Zweig, Cambridge.
Longuenesse, Béatrice (1998): Kant and the Capacity to Judge, Princeton.
McDowell, John (1998): Having the World in View: Sellars, Kant, and Inten
tionality, in: Journal of Philosophy 95, pp. 431 – 491.
McDowell, John (1996): Mind and World, Cambridge.
Sellars, Wilfrid (1968): Science and Metaphysics: Variations on Kantian
Themes, London.
Sellars, Wilfrid (1975): The Adverbial Theory of the Objects of Perception, in:
Metaphilosophy, pp. 144 – 160.
Soteriou, Matthew (2010): The Disjunctive Theory of Perception, in: The
Stanford Encyclopedia of Philosophy (Winter 2010 Edition), ed. Edward N.
Zalta, URL = <http://plato.stanford.edu/archives/win2010/entries/per
ception disjunctive/>.
Strawson, Peter F. (1971): Imagination and Perception, in: Foster and Swanson
(eds.): Experience and Theory, London, pp. 31 – 54.
Thompson, Manley (1972): Singular Terms and Intuitions in Kant’s Epistemol
ogy, in: Review of Metaphysics 26, pp. 314 – 343.
Watkins, Eric (2008): Kant and the Myth of the Given, in: Inquiry 51,
pp. 512 – 531.
Westphal, Kenneth R. (2004): Kant’s Transcendental Proof of Realism, Cam
bridge.
Freedom and the Human Sciences:
Hume’s Science of Man versus Kant’s
Pragmatic Anthropology1
Thomas Sturm
Abstract
In his Anthropology from a Pragmatic Point of View, Kant formulates the idea of the
empirical investigation of the human being as a free agent. The notion is puz
zling: Does Kant not often claim that, from an empirical point of view, human
beings cannot be considered as free? What sense would it make anyway to in
clude the notion of freedom in science? The answer to these questions lies in
Kant’s notion of character. While probably all concepts of character are in
volved in the description and explanation of human action, Kant develops a
specific notion of character by distinguishing character as a “mode of thought”
(Denkungsart) from character as a “mode of sensing” (Sinnesart). The former no
tion is distinctively Kantian. Only mode of thought reveals itself in human ac
tion such that actions can be seen as linked to an agent’s first person perspective
and the capacity to rationally reflect one’s own intentions and desires. By ref
erence to this concept human actions can be empirically explained qua free ac
tions. The point of this paper is not only to rule out the interpretation that Kant
is an incompatibilist concerning the dilemma of freedom and causal determin
ism. It is also argued that Kant defends a version of soft determinism which is
more sophisticated and more adequate for the human sciences than Hume’s.
1 This essay is a strongly updated and extended version of a conference paper that
first appeared in German (Sturm 2001a). It also extracts materials of (Sturm
2009, esp. ch. VII §§14 – 18 and ch. VIII §§ 5 – 6). Work on this article was sup
ported by the Max Planck Institute for the History of Science, Berlin, and the
Spanish Ministry for Science and Innovation (MICINN), reference number FFI
2008 – 01559/FISO.—I am grateful to Christopher Green, Paul Erikson, and to
an anonymous reviewer from the Kant Yearbook for several suggestions that
helped to improve my English. All translations are my own.
24 Thomas Sturm
Introduction
The project of a science of human nature was of great concern to eight

eenth century philosophers and scientists. A wide range of conceptions
for it were available at the time, all competing for the distinction of the
most adequate and fruitful program for how to investigate the human
being. There were serious controversies over the methods, aims, and
the characterization of the subject matter of the human sciences. In Ger
man speaking countries the debate raged in the 1770s, when in partic
ular the advocates of what was called “empirical psychology” and “med
ical” or “physiological anthropology” happened to meet (see, for in
stance, Bonnet (1770 71); Platner (1772); Herz (1773); Schütz
(1771); Tetens (1777, Introduction)). They discussed, among other
things, the legitimacy of introspection (is it a reliable and useful tool
for justifying empirical knowledge about the mind?), the nature and
the architecture of the human mind (what are its basic and derived fac
ulties?), and the possibility of physiological explanations of the mental
(can they be given? are they necessary at all?).
Kant observed these debates. He had serious problems with both of
these approaches and their answers to various methodological and met
aphysical questions (see Sturm 2009, chap. II, IV V). I cannot discuss
these topics here, but I shall assume the following, for which closer ar
guments can indeed be given (ibid., chap. III IV). Despite his well
known skepticism about the then existing empirical psychology and
its aspiration to become a true natural science (see AA 4:471) 2, Kant
did not deny the possibility of a systematic empirical investigation
into the human mind and behavior at all. From 1772/73 until 1795/
96, he taught in every winter semester what he soon came to call “prag
matic anthropology”. In AA 25 we have an excellent critical edition of
manuscripts from these lectures. Kant’s Anthropology from a Pragmatic
Point of View of 1798 stems from these lectures. He even often described
this anthropology as at least potentially a systematic science with its
2 Kant citations are taken from the Akademie Ausgabe (Kant 1900 ff.), indicating
volume and page numbers. For the Critique of Pure Reason (CPR) I depart
from this practice and, as is common, indicate the page numbers of the first
(A) and second (B) edition (1781/1787). Also, a few of the passages quoted
here have been taken from lectures on anthropology not contained in AA 25
and are cited here by their title and page number (e. g., Reichel 109); these lec
tures are accessible on the internet (http://web.uni marburg.de/kant//web
seitn/gt ho304.htm).
Freedom and the Human Sciences 25
own architectonic, or as built from a definite idea (AA 7:119, 121 f.; AA
10:146 f., 242; AA 25:470, 551, 782). It would also incorporate what is
valuable in empirical psychology (CPR A 849/B 877; AA 28:223, 541,
584 and 876; AA 19:756 f.). Both the lectures and the book reveal his
continuing reflection on and reaction to then ongoing debates, and how
this led to some of his puzzling ideas concerning the human sciences.
One of these ideas arguably the most puzzling, but also most im
portant one is the following. Kant determines the subject matter of his
anthropology as an empirical inquiry into what “the human being, as a
free agent, makes, or can and should make, of himself” (AA 7:119). In
other words, human freedom is not only incorporated into his anthro
pology, but even plays a central role in it. Interpreters of Kant’s Anthro-
pology have often ignored his programmatic statement;3 and insofar as
they recognize it they often wonder how the program could be realiz
ed.4 What is on Kant’s mind here? Clearly, this part of his conception
relates closely to a challenge for the human sciences up until today,
namely: When we look at actions from a scientific point of view,
what sense should we make of the ordinary idea that we are free agents,
capable of initiating our own actions, and able to act responsibly? Is this
assumption perhaps only to be used when we consider actions from a
practical perspective and thus, theoretically or scientifically speaking,
an illusion? Or could it be integrated into science in a serious way?
The problem cannot be foreign to Kant. After all, he also says that if ac
tions are viewed empirically, they are as causally determined as any
other natural event (CPR A 549/B 577; AA 5:99). Does he not strictly
hold apart viewing ourselves from an empirical point of view, as mere
natural objects, and viewing ourselves as moral agents? These questions
are connected to the familiar metaphysical problem of freedom and de
terminism. However, I shall not deal with this problem here, but rather
consider how different stances towards it can lead to different concep
tions of the human sciences.
In order to discuss Kant’s approach, it is useful to reflect on a com
parable project for the investigation of human action a project that
3 Many scholars have for long followed Beck’s view on Kant’s Anthropology, who
thought that it simply examined “how men should conduct themselves in or
dinary affairs of life”, resulting only in an “episodic elaboration of practical
rules” (Beck (1960, 7 and 54)). After the appearance of AA 25, this has clearly
changed.
4 Thus Hinske (1966, 425). About a few recent attempts to make sense of Kant’s
statement, more in part 2 of the present essay.
26 Thomas Sturm
also struggled with the problem of freedom and determinism in an era

that began to scientifically investigate human thoughts, feeling, inclina
tions and action. I mean Hume’s project of a “science of human na
ture.” As will become clear, Kant and Hume agree in that if the concept
of freedom is to be integrated into science in a serious way, this integra
tion has to use the notion of character. However, they do so in strikingly
different ways, related to their different conceptions of the “science of
human nature” and “pragmatic anthropology.” In part 1, I will explain
Hume’s approach, showing how he tried to defend a version of compa
tibilism concerning freedom and causal determinism. I shall also note
two serious problems with his view. In part 2, I turn to Kant’s different
approach and will show how it could avoid the problems Hume’s ap
proach was faced with.
1. Freedom in Hume’s project of a “science of human nature”

1.1. Hume’s compatibilism concerning freedom and causal
determinism
Hume argues that there does exist a meaningful concept of freedom that
is perfectly compatible with the causal determination or necessity of our
actions. He characterizes the concept as follows: “By liberty […] we can
only mean a power of acting or not acting, according to the determinations of the
will; that is, if we choose to remain at rest, we may; if we choose to
move, we also may” (Hume 1748 51, 95). A human being acts freely
if (and only if) she acts according to her desires or passions, where this
course of action is neither undermined nor produced by (external) co
ercion.
The point of this claim, and its consequences for a scientific inves
tigation of human nature, becomes clear when we bring into play one of
Hume’s distinctions. Like many others, he distinguishes two different
concepts of freedom: the “liberty of spontaneity” on the one hand,
and the “liberty of indifference” on the other. Hume understands the
notion of indifference as the total absence of causes or necessity. He
maintains that it is the view of many philosophers that an action can
only be free if it is not caused, since causes bring about their effects nec-
essarily. On the view of advocates of the liberty of indifference, howev
er, it cannot be necessary that we act in the ways we do, if we act freely
(Hume 1739 40, 407; 1748 51, 94 n.). We must always have a real
possibility of doing otherwise, and of doing so under the very same

causal circumstances. Similarly, our past actions have only been free if
we could have done otherwise, again under the same causal circumstan
ces. Hume opposes this notion of freedom on various grounds. He
thinks it implies the existence of chance, that it is incompatible with
the general causal determinism, and that is a threat to the project of a
science of human nature.5 The idea that human action is subject to caus
al laws is a well known and crucial tenet of his project of an empirical
science of human nature. Indeed, what sense would it make to first cau
sally explain an action and then to claim that those causes need not have
that action as their consequence? The notion of freedom as spontaneity,
fortunately, does not undermine this project. Spontaneity is what Hume
has in mind when he explains freedom by saying that we act freely if and
only if we act according to the “determinations of the will” according
to our desires or intentions. This, he thinks, is the only meaningful no
tion of freedom that there is. Spontaneity is enjoyed by “everyone, who
is not a prisoner and in chains” (Hume 1748 51, 95). The freedom of
spontaneity is, therefore, only opposed to (external) coercion or vio
lence. It does not require the absence of causes or the possibility to
act otherwise even under identical causal conditions. To be true, spon
taneity allows for a certain meaning of the common phrase that one
could have acted otherwise: I could have acted otherwise if I also had de-
sired otherwise if, in other words, it hadn’t been the case that the ante
cedent causal conditions had been the same. Since we can, in Hume’s
view, explain and predict our actions by reference to such determina
tions of the will according to the laws of human nature, freedom of
spontaneity is compatible with the causal determination of actions and
with the project of the empirical science of human nature.
1.2. Two problems of Hume’s compatibilism
Hume’s argument is widely taken as a paradigmatic defense of compa

tibilism, and indeed a successful one, at least in part.6 But we should not
overlook the problems with his views. To begin, it is certainly true that
Hume’s concept of freedom as spontaneity is compatible with the ideas
of the causal determination and the empirical investigation of actions.
5 See Stroud (1977, 141 – 150); Penelhum (1993, 130 f.).

6 Stroud (1977, 144); Penelhum (1993, 129).
28 Thomas Sturm
But, at the same time, the concept is totally irrelevant for this investiga
tion: Since all actions can be explained completely by reference to pas
sions, intentions, and other “determinations of the will,” the concept of
freedom itself can be eliminated from the “science of human nature.” In
order to give causal explanations of human actions, one cites the rele
vant “determinations of the will,” and that’s it. That the actions are
also free in the sense of spontaneity might be mentioned, but does
not add anything to the explanations. One need not complain about
this. However, it is important to note this point since, as we will see,
this marks a critical difference between Hume’s and Kant’s approaches.
There are also separate problems with Hume’s theory, two of which
I shall point out here.7 The first is the problem of explaining our impres
sion of being free in the sense of indifference. Hume fully admits that
we do have the impression that we are free in the sense of being able
to act otherwise under identical antecedent causal conditions. He
notes that from a first person point of view it often appears to us that
no causes for our actions are given, and that we therefore appear to our
selves as being able to choose arbitrarily between different alternatives.
But this impression is wrong, he maintains, and it is so for the following
reason:
We may imagine we feel a liberty within ourselves; but a spectator can
commonly infer our actions from our motives and character; and even
where he cannot, he concludes in general, that he might, were he perfectly
acquainted with every circumstance of our situation and temper, and the
most secret springs of our complexion and disposition. (Hume 1739 – 40,
408 f.; 1748 – 51, 94 n.)
Such an external spectator is, of course, the “scientist of human nature.”
But even if we grant that an external spectator may be capable of ex
plaining and predicting our actions causally, or that the impression of
indifference or absence of causes is misleading why does that impres
sion arise specifically from the first person point of view? It can hardly
be the general lack of knowledge of the causal relations between deter
minations of the will and actions, since Hume assumes that we all have
such knowledge and apply it when observing other agents. So, we
should expect that Hume takes it for granted that we apply such knowl
edge to ourselves as well. But he does not. Moreover, as he explicitly
asserts, we do not only in fact have such causal knowledge, we also
need to have it. Only beings having such knowledge of causes can in
7 See also Stroud (1977, 146 – 149).

teract with other agents, ascribe responsibility to them, and so on.

Hume does not explain why such knowledge, if it must be ascribed
to every agent, coheres with the idea that every agent, under certain cir
cumstances, has the firm impression that no causes necessitate his own
course of action. Neither does he explain why this impression or idea
goes hand in hand with an asymmetry between the first person and
third person point of view if indeed any such asymmetry exists,
which of course may be doubted. In any case, Hume’s treatment of
the liberty of indifference is not sufficiently thorough.
A second problem for Hume’s theory of freedom concerns the con
nection between that theory and his conception of the science of man.
As he admits, human beings act in ways which seem to be quite irregular
and, therefore, not accessible to an empirical analysis of action based on
universal causal laws. He proposes the following: Human beings or
groups of human beings may differ in their courses of action, but one
can, nevertheless, observe regularities within particular agents or partic
ular groups of agents. Here Hume brings into play the concept of char
acter. Traits of character fulfill an explanatory role, since they are more
or less persistent causal properties of human beings leading to constant
behavior. Most importantly and typically (though perhaps not always),
character traits are long term desires, intentions and other determina
tions of the will.8 There are, Hume says, a “variety of maxims”
among mankind. But it is also true, he says, that each person has “max
ims” or a “fixed and established character,” and according to this “fixed
and established character” the agent acts in typical ways (Hume 1748
51, 85 88). That does not mean that people cannot change, nor that
they cannot occasionally act out of character. But they do so in accord
ance with the causal laws of human nature. That it is not always easy for
us to discern a person’s character shows merely that the scientific study
of mankind is not yet mature. In any case, for Hume “character” is a
technical term.
Examples of the explanatory function and the complexity of this
concept can be found in Hume’s History of England (Hume 1754
62). Almost every chapter on the rule of a particular monarch ends
8 See Bricke (1995); McIntyre (1990). I shall not discuss here how tight the re
lation between character and action is in Hume (see Johnson (1990)), nor the
qualifications and exceptions one would have to give concerning the idea that,
by and large, character traits are composed of long term desires or passions (see
the beautiful discussion of this topic by Baier (2008, ch. 1).
30 Thomas Sturm
with a section on that person’s “Death and Character,” listing the com
bination of traits observed in that person that led to him living and act
ing as he did. The character of Charles I was mixed “as that of most
men, if not of all” (Hume 1754 62, V 542). On Henry VIII, Hume
writes:
It is difficult to give a just summary of this prince’s qualities: He was so dif
ferent from himself in different parts of his reign, that […] his history is his
best character and description. The absolute, uncontrouled authority which
he maintained at home, and the regard which he acquired among foreign
nations, are circumstances, which entitle him, in some degree, the appella
tion of a great prince; while his tyranny and barbarity exclude him from the
character of a good one. He possessed, indeed, great vigour of mind, which
qualified him for exercising dominion over men; courage, intrepidity, vig
ilance, inflexibility […]. (Hume 1754 – 62, III 321 f.)
Now, an important problem that arises with Hume’s notion of character
relates to the following point. Traits of character are what make sense of
the idea of an ascription of responsibility that, for Hume, is such a cru
cial part of his theory of freedom. Actions are, as he says, “temporary
and perishing,” and we do not hold a person responsible if she does a
certain thing only once, because it is always unclear whether it was real
ly she who performed the action. Henry VIII had a complex and partly
evolving character, but it is pretty clear that through most of his life he
was a dominating, strong person. This trait also supported his immoral
traits (his “barbarity”, say), and they are the reason why we think that he
had no good character, and why we blame him for many of his actions.
Only a reference to traits of character permits an ascription of responsi
bility, since those traits are “durable and constant” features of one’s
mind (Hume 1748 51, 98). However, at a closer look this only
means that the reference to traits of character is a reference to causal reg
ularities. Hume does not say why it is precisely the concept of character
that provides the link needed for practically evaluating a person, ascrib
ing responsibility, and so on. Put differently, he does not explain how
the agent himself can cause an action. After all, a fixed attitude or a con
stant motive for a type of action can be determined by external coercion
in a way that does not reveal this coercive force, thereby undermining
Hume’s own criterion for free action. An empirical investigation of
human action, at least one with Hume’s intentions, should introduce
and reflect relevant conceptual and empirical differences here. But
Hume’s “science of human nature” never achieves this. In this regard,
his project neglects a task crucial to any moral psychology. So much for
Hume.
2. Kant’s pragmatic anthropology as a science of free agency
That leaves us with Kant. As mentioned, he specifies the subject matter

of his Anthropology as an inquiry into what “man, as a free agent, makes,
or can and should make, of himself” (AA 7:119). The first question we
must ask is what notion of freedom is on his mind here.
2.1. Transcendental and practical freedom
Let me begin by noting what Kant does not maintain about freedom.
Though many influential commentators assert the contrary, he was al
ways a vigorous opponent of the notion of freedom as indifference.
In other words, Kant is not an incompatibilist about freedom and deter
minism.9 In the Critique of Pure Reason he introduces a distinction that
might appear to be identical with the Humean distinction between
9 Allison speaks of Kant’s “contra causal” or “incompatibilist” notion of freedom

(Allison (1990, 1, 25 f.)). This suggests—and Allison does not offset the impres
sion—that Kant thought indifference is a necessary condition for free action.
Hudson explicitly claims that between Nova Dilucidatio and the critical phase
Kant changed his view accordingly: He made the capacity to do otherwise a
necessary condition for free action (Hudson (1994, 7 and particularly 72 f.)).
This assumption shapes essential considerations in Hudson’s discussion (espe
cially Hudson (1994, 72 – 98)). But Hudson is wrong. Kant consistently rejects
the notion of freedom as indifference, from Nova Dilucidatio (1755) to Metaphy
sics of Morals (1797) (AA 1:403; AA 6:226 f.). Neither does he change his mind
in between (see, for instance, R 4226=AA 27:465 (1770s) or similar statements
in lectures on metaphysics from the 1780s and 1790s, AA 29:901 f.; AA
29:1022 f.). Kant mentions, among other things, Buridan’s ass: a total indiffer
ence towards motives can never explain how and why we act freely (AA
28:677 f.). Hudson’s treatment of passages from the first Critique and other writ
ings (CPR A 554 – 556/B 582 – 584) is also unsatisfactory: for Kant the fact that
we can do otherwise means that we can reflect whether our action is guided by
the right practical norm, and that this reflection can have causal relevance for
what we do. This varies the causal prerequisites, instead of keeping them strictly
identical. For Kant the possibility stressed by Hudson, namely, that we may de
viate from the norm, is not proof of indifference freedom; not all possibilities
are capacities (Wood 1984, 81 f.).
32 Thomas Sturm
the liberties of indifference and of spontaneity, namely the distinction

between “transcendental” and “practical” freedom. Transcendental
freedom refers, as he says, to the capacity of bringing “about a state of
affairs by oneself” or of “beginning a series of events completely by one
self” (CPR A 533 f./B 561 f.), that is wholly free of any natural causes.
However, his defense of the possibility of transcendental freedom in the
third Antinomy by no means advocates the idea that we might act in
differently, or that we could have done otherwise under completely
identical causal circumstances. One consequence of this for his anthro
pological project should be clear. His intention is not to realize the proj
ect of explaining human action causally and simultaneously maintaining
that the action thus explained might also have been done otherwise.
That would undermine the idea of causal explanation itself, and it
would make Kant an incompatibilist. A lengthy defense of both the lib
erty of indifference and a program for a causal explanation for human
action had been developed by Johann Nicolas Tetens (1777, vol. II,
1 148). Kant justifiably dismisses Tetens’s discussion of freedom and
determinism as confused and useless (AA 10:232).
This rejection of incompatibilism both on the metaphysical and the
scientific level seems to imply that all Kant has left is the Humean no
tion of freedom as spontaneity, or that “practical freedom” has to be un
derstood as the ability to act according to one’s given desires or inclina
tions. Henry Allison maintains that Kant’s anthropology has no room for
the concept of transcendental freedom. Many passages in Kant’s writings
seem to support that reading. Transcendentally free actions are said to
not be something we ever encounter in experience. Conversely, Kant
also claims that if we consider actions from an empirical point of
view, they are as causally determined as any other natural event. In con
trast, he characterizes practical freedom as the capacity of human beings
to develop a certain “independence from sensuous impulses”, and he
explicitly states that this freedom can be “proven from experience”
(CPR A 802/B 830). One might thus believe that practical freedom,
and it alone, is the concept appropriate to Kant’s anthropological inves
tigations. That is what Alix Cohen (2009) has recently proposed. Un
fortunately, that won’t work either. Kant maintains that practical free
dom presupposes transcendental freedom. In other words, we get no
“independence from sensuous impulses” (no practical freedom) without
a capacity to “bring about a state of affairs by oneself” (A533 f./B561 f.).
More positively, this shows that Kant is interested in an aspect of the
notion of freedom that has been overlooked by many interpreters, but
which, in his view, is crucial both for the empirical explanation of

human agency and for practical purposes such as ascribing responsibility.
It is the idea that we are active subjects, or that we can, in a certain sense
yet to be explained, be the originators (Urheber) of our actions. This is a
different assumption about freedom than the popular assumption of
being able to do otherwise even under identical causal circumstances.
According to the intuition of self determination, it may very well be
that our actions are caused in a certain fashion, namely by the agent.
In this view, the very concept of freedom becomes predicated not pri
marily of actions, but of agents. This does not lead Kant to adopt an oth
erwise unexplained notion of agent causation, as had been maintained
in his time by, say, Thomas Reid. There are certain capacities a human
being must possess to be such an agent and, moreover, these are ca
pacities that one does not possess from the beginning but which we
have to develop. This developmental aspect is another essential difference
between Hume and Kant. It also leads to the next step we have to take:
Kant explains his view that we can be, and can become, originators of
our actions by one of his anthropological concepts of character.
2.2. Character in Kantian anthropology
In his anthropology lectures, Kant uses as textbook the chapter on em

pirical psychology from Alexander Gottlieb Baumgarten’s Metaphysica
(1757). This text was basically a faculty psychology: a doctrine of the
basic mental powers such as cognition or volition and their sub powers
such as the senses, memory, the intellect, and so on. Kant’s strongest de
viation from this structure comes in the latter half of the 1770s. From
then on, he models only the first part of his anthropology, which he
also describes as the “general part” or a “doctrine of elements,” after
the faculty psychologies. To this, he adds a second part, the so called
“Anthropological Characteristics”.10 This is no accident. These “Char
10 Kant’s Characteristics are either dealt with skimpily or not at all in Marquard
(1965); Hinske (1966); Van de Pitte (1971); Firla (1981); Kim (1994); Wilson
(2006); not even by those interested in Kant’s theory of action from a modern
perspective (Meerbote (1984); Allison (1990); Hudson (1994)). Brandt (1999)
addresses the “Characteristics” at length, but does not tackle the guiding ques
tion of the present essay; neither does Frierson (2005) in his otherwise useful
account of Kant’s empirical theory of action.
34 Thomas Sturm
acteristics” provide the key to the problem of how we can be studied as

originators of our actions from an empirical point of view.
In the opening section of the “Characteristics” Kant begins with a
distinction that resembles Hume’s distinction of properties that are
shared by all human beings alike from properties which Hume thinks
force us to develop a differentiated analysis of human individuals or
groups. But it would be false to view Kant’s “Characteristics” as another
early instance of what we would nowadays call a differential psychology.
From at least the mid 1770s11 Kant additionally introduces a distinction
between human beings as natural beings and as free beings. A typical
statement goes as follows:
The character of a human being is the distinguishing mark of one human
being from another, or of mankind in general from other beings. It is [first]
the character of a person, then that of sex, that of a nation and, finally, of
the species. One may, then, consider the human being [both] as a natural
and as a free being. As a natural being one considers him in terms of the
predisposition to be found in him, and that is the character of human beings
as animals. […] The character of the human being as a free being is being
made part of his will. (AA 25:1530; cf. AA 25:1384; Reichel 109 and 122)
Kant’s reflections do not always take the same route, but all aim in the
same direction: In order to express what is special about man’s character
as a free being, one must distinguish one kind of character that consti
tutes a “mode of sense” (Sinnesart) from a kind of character that consti
tutes a “mode of thought” (Denkungsart; alluded to in AA 25:649 and
clearly in AA 25:821 f.; cf. AA 7:285, 292). They are both kinds of
character for the following two reasons. First, to have a character gen
erally means for Kant (as for Hume) that one’s own behavior can be de
scribed as subject to certain regularities. If one knows the regularity
under which the behavior of a person falls, one knows what one “can
expect of a man” (AA 7:292). This holds both for the concept of char
acter as a person’s “mode of sense” and her “mode of thought.” There
fore, both concepts can fulfill the explanatory role that the concept of
character is traditionally expected to fulfill. Second, Kant assumes that
having a character does not mean having another mental capacity. Rath
er, character is always the result of the development and exercise of certain
powers (AA 25:227, 437, 624). The next natural question is what these
11 In 1772/73 Kant still says that a person “cannot give himself a character other
than the one nature has given him” (AA 25:228). But perhaps in 1775/76 and
clearly as of 1777/78 he does claim the opposite.
powers are, and how they are developed and exercised and how that
establishes a difference between mode of sense and mode of thought.
That there are differences between the two forms of character Kant
supports by pointing out that we not only speak of character as some
thing that describes the type of behavior a person normally exhibits.
We also speak of a person as having or lacking character at all:
If we can say of a man simply: “He has character”, we are not merely say
ing a lot about him but also paying him a great compliment; for this is a rare
thing that inspires respect and admiration. (AA 7:291 f.)
Thus we do not merely use the character vocabulary for descriptive pur
poses. We also use it for evaluating a person’s deeds. This, of course, is a
familiar phenomenon. However, although the descriptive and the eval
uative concepts do differ, one should not think that Kant would separate
them such that the evaluative notion of character only finds use in prac
tical life and the descriptive one only in the scientific investigation of
human action.12 That our notion of character enables us to evaluate de
meanor is also an empirical fact. Any empirical investigation that ne
glects that fact would neglect crucial features of human conduct.
2.3. Character as “mode of thought” (Denkungsart)
Now, for the practice of evaluating other agents Kant uses the concept
of character as a “mode of thought.” This he does because he relates
only this notion to the concept of freedom. As he puts it at one
point, “mode of thought” is a disposition by which we “fixate our free
dom” (R 1517 = AA 15:867 f.). But how can we do that?
Two essential features of the concept of a mode of thought need to
be pointed out here. They are not surprising to those familiar with
Kant’s other works, but it is important that he developed them within
his anthropology and never omitted them there. First, he assumes that
the fixation of one’s freedom involves the use of one’s practical reason
(AA 25:227, 437, 624). This capacity consists in the ability to formulate
practical principles or maxims, or to evaluate given practical rules (AA
7:199; cf. CPR A 547/B 575). Just as central as this feature is the second
one, which is peculiar to the Kantian conception, but also linked to the
first feature: To have or employ a mode of thought involves having a
12 This goes against, say, Beck (1960, 29 – 32; 1975) or O’Neill (1989, 67).
36 Thomas Sturm
certain first-person perspective on one’s own actions and the intentions and
rules that guide them. Describing a person’s deeds by reference to her
mode of thought requires that any explanation of her acts will involve
a reference to her own perspective. Not all behavior must be viewed in
that way, as Kant’s distinction between a mode of thought and a mode
of sense makes clear. We may often view human beings as not acting
according to rules that they have evaluated and deliberately adopted.
For instance, we may act according to the traditional rules of a certain
time or society, which may become, as Kant says, “another nature”
(AA 7:121). But something that may be important for the explanation
of significant areas of our lives would be missing if we ignored our
own perspective as agents.13
The topic of the first person perspective or the self consciousness
in thought and action is indeed central for Kant’s approach. His lectures
begin with it and he claims that it arises over and over again in the an
thropological investigation of human cognition, feeling and action:
The Self involves what distinguishes man from all other animals. If my
horse could grasp the notion of Self, I would have to dismount and view
him as my companion. The Self makes man a person, and this notion
gives him the power over all else, it makes him the object of his own re
flection. The Self is involved in all our thought and action, and is our great
est concern. (AA 25:859; cf. AA 25:1215 f., 1438; AA 7:127)
The general claim of the relevance of the notion of the self can best be
illustrated by the issue of human “egoism” and the complex dynamics of
human social interaction it leads to. Every human being craves recogni
tion and enjoys drawing attention to his or her “beloved Self”. Kant ob
13 It is therefore difficult to see what Allison means by claiming “that there is no

room even for the thought of such freedom [namely transcendental freedom] in
connection with the empirical character of rational agents” (Allison 1990, 46).
Allison sees the “incorporation thesis” as the crucial thesis in Kant’s theory of
freedom. According to this thesis motives can only enter explanations of our
actions if they are “taken up” in a maxim; and maxims are expressions of
the spontaneity of our selves and their rational capacities (Allison 1990, 5 f.;
cf. CPR A 547 f./B 575 f.). These claims are by no means excluded from An
thropology, however. In a another way Hudson is mistaken by saying that the
Kantian idea of an empirical explanation of action can only be an explanation
in terms of natural causality, and by equating the view of the same actions using
the vocabulary of freedom and a strictly non empirical point of view. Kant is a
compatibilist in a far reaching sense: it is possible to empirically describe human
actions as being free because we can empirically use concepts of the self as an
agent and concepts of that self’s practical maxims.
serves that people don’t overly appreciate egoism in others, even though
they themselves are tempted to act similarly (AA 7:127 130). If I wish
to act prudently, I must learn to take into account that others share traits
of egoism; I must take their point of view into account. More impor
tantly, I must grant others opportunities to speak favorably of them
selves. By sensing my esteem for them, they will in turn form a better
opinion of me a fact I can then exploit in order to gain their support
or recognition. Throughout the lectures Kant discusses similar needs to
understand actions as being conditioned by one’s view of oneself and
the related view that others have opinions about me. One may cite
Kant’s analysis of the omnipresent practice of role playing in society,
of “concealing” one’s own intentions and beliefs and “dissembling” be
fore others and so on (AA 7:151 f.). In other words, we acquire a mode
of thought if we do acquire it at all only through rather complicated
social dynamics and learning processes.
Kant’s views naturally involve difficulties. To mention but one
point: He embeds the idea of an agent being the originator of his or
her action in his anthropology, thereby giving sense to his claim that
practical freedom presupposes transcendental freedom, but he also con
nects the concept of transcendental freedom in Critique to more de
manding and problematic assumptions as well. In the Critique, he links
the concept of transcendental freedom to the notorious distinction be
tween “things in themselves” and “appearances” (it is a matter of serious
dispute whether all of his statements can be read as involving a so called
two aspect instead of a metaphysically problematic two world view).
This leads him to a number of obscure claims for example, that acting
transcendentally free or from “absolute spontaneity,” means acting from
an atemporal cause. One cannot save the “whole Kant”: Certain deci
sions have to be made. Mine is, as should be clear, to side with the more
mundane view.
2.4. Comparison with Hume’s views
It would be wrong to think that I have argued that Kant’s views about
the notion of freedom in relation to the human sciences constitute his
considered reaction to Hume’s views on the same topic. There is, of
course, considerable circumstantial evidence for such a historical con
nection: Kant was familiar with both Hume’s Enquiries and with his His-
tory of England as well. Moreover, Kant was following the general call of
38 Thomas Sturm
a comprehensive human science, for which Hume’s work was often

seen as paradigmatic. Finally, the close conceptual, even terminological
connections between the notions of freedom and character in both au
thors are striking, too. But my point is more systematic than historical.
As mentioned, Hume’s views about the relation between the scientific
investigation of the human mind and behavior and the assumption of
freedom are nowadays often seen as a major inspiration. This cannot
be said about Kant. We may wonder whether that is justified.
So, how do Kant’s solutions compare to Hume’s? To begin, some of
the aforementioned topics in which the concept of the self seems to play
an essential role are familiar to Hume, of course. The ideas of self love,
pride, and humility are often involved in his description of the charac
ters of human beings (cf. Hume 1748 51, 233 f.; 1739 40, 277 ff.).
However, Kant’s notion of the self in connection with the concept of
character has an additional aspect. He remarks:
[…] when we simply say that [a person] has character, we mean [he pos
sesses] that property of the will by which he binds himself to explicit prac
tical principles which he has prescribed himself irrevocably through his
own reason. (AA 7:292; cf. 285 and 294; AA 25: 649 f., 823, 1385)
The rules the Kantian anthropologist uses in explaining human action
when he refers to a person’s mode of thought are not merely rules
under which those actions either do or do not fall. Rather, they are
the result of the capacity to view those rules as being one’s own in
the sense that one is able to reflect and evaluate them rationally, to
find them correct or incorrect according to certain criteria, and so on.
If in a given case such a rational reflection has a positive outcome,
one identifies actions as one’s owns in a way that one doesn’t for other
behavior. Hume’s views cannot cover these sorts of cases because he
thinks of the regularities governing human conduct along the lines of
mere laws of nature.14 In this sense, to speak of a person’s character is
14 Against Frierson, who maintains (2005, 3 and 8) that Kant’s empirical account
of human action is in terms of strict natural laws and has in itself no role for the
notion of freedom. Later on, Frierson (2005, 26 ff.) first notes that the Kantian
explanation of the origin of character seems not to be naturalistic in this strong
sense, but then continues to insist that it has to be. In my view, Frierson over
looks the conceptual and developmental centrality of the first person perspec
tive in his otherwise complex interpretation of Kant’s account of character. The
converse mistake is made by Cohen, although she (2009, 117) claims to be fol
lowing Frierson’s account. Cohen frequently refers to Kantian statements that
bring into play the self determination of free actions (interpreting these in
not to ascribe to her the character or rule but to say that she has devel
oped her own character as opposed to being merely formed by nature or
tradition. In this sense and in no other Kant views us as possible
originators of actions or, more precisely: as originators of principles of
action. Moreover, by using this notion of character, Kant introduces a
developmental perspective to the investigation of human conduct,
and a specific one one in which we are not mere products of our
own development, but in which we are also viewed as potential (and
partly actual) producers of our development.
This first person point of view can surely be tied to Kant’s concept
of the self or ‘I’ as found in his theoretical and practical philosophy
(without being identical with these notions). What we have here is a
formal conception of the self: a conception of a hierarchy of possibly
reflected rules of action of a person. Such rules are intentional states
of a higher order, but not merely this (versus Frankfurt (1982)). I can
observe myself remembering something, or I can remember that I
once had a certain belief. Such second order intentions are merely no
ticings. Other second order intentions, in contrast, might form and direct
first order intentions, and Kant identifies the former as principles of rea
son. Which principles have priority, and why they do, are different
questions which he answers elsewhere. In his anthropology, he is
quite careful not to overstep the boundaries of an empirical investiga
tion: what anthropology does is to study the empirical conditions
which further or hinder the development of character as a mode of
thought.
Finally, remember that there were two main problems with Hume’s
views. First, his unexplained assumption of an asymmetry between the
view of one’s own actions and those of other human beings: only in the
former case we are said to experience a liberty of indifference or a non
existence of causes. Hume has no explanation for this claim, which is
even incompatible with his claim that we must assume for various rea
sons the connection between causes of actions and actions themselves to
be necessary. Second, he admits that actions appear oftentimes irregular,
terms of practical freedom in a problematic way, as noted above) but does not
connect these statements with Kant’s account of character. Moreover, her claim
(2009, 116), that bringing into play the notion of freedom implies “no dimin
ution of the necessity of natural laws” is fine. One never violates any physical
law when acting freely, of course. But we should not infer from this that self
legislated rules of reasoning or Denkungsart character are themselves mere laws
of nature.
40 Thomas Sturm
but claims that we can nevertheless explain them by using the concept
of character, which he thinks also explains the practice of ascribing
agency and responsibility to persons instead of the fleeting actions. How
ever, it is unclear why he thinks so, since his notion of character in
volves only the idea of agent specific regularities; it remains open
whether we should take the practice of ascribing responsibility also seri
ously.
In Kant’s approach, in contrast to Hume’s, agent and spectator do
have a symmetrical explanation for our belief that we could have
done otherwise without making use of the dubious notion of the liberty
of indifference: We know what it means to apply the principle of one’s
action to oneself, to justify such a principle, to follow it correctly, to vi
olate it consciously, and so on. We can ascribe the same abilities to oth
ers just as well. We do not mean thereby that their or our actions are
uncaused, or that we could act otherwise under identical causal circum
stances. Character as a mode of thought is, as Kant’s freely admits, some
thing to be acquired through a learning process in which other human
beings are our teachers. This process may never be perfect, but that does
not mean that actions done on the basis of merely following tradition
and actions performed by following one’s own reflected maxims are
the same. The anthropological point of view is not to be equated
with the first person point of view, but a Kantian anthropologist
takes into consideration that such a first person point of view does
exist, and that it can figure in our lives, and how it is restricted by cer
tain circumstances and furthered by others.
Admittedly, that is at best a framework for an empirical investigation
of free agency. But it is clear that, however one would have to carry out
such a project, it would connect Kant’s anthropology to his views on
the relevance of understanding human education, history, and society,
and it would particularly involve the actual empirical investigation of
all educational, sociological, psychological, and historical conditions
which further or hinder the development of character (cf. AA 4:289;
AA 6:417). In other words, Kant anticipates work currently pursued
where disciplines like personality and developmental psychology, and
social and moral psychology intersect. Historically speaking, Hume’s
views about the “science of human nature” and its relation to the ordi
nary assumption of human freedom have prevailed, at least in certain
camps, and Kant’s conception of anthropology has, as far as I know,
never been viewed as a precursor to later developments in, say, person
ality or developmental psychology. That is an interesting historical ex
planandum. I suspect part of the explanation has to do with the heavy

weight of later perceptions of Kant as being the author of the Critiques
rather than also as a serious philosopher of the human sciences.
Bibliography
Allison, Henry E. (1990): Kant’s Theory of Freedom, Cambridge.
Baier, Annette (2008): Death and Character, Cambridge, MA, London.
Baumgarten, Alexander Gottlieb (1739/41757): Metaphysica, Halle.
Beck, Lewis White (1960): A Commentary on Kant’s Critique of Practical
Reason, Chicago.
Beck, Lewis White (1975): The Actor and the Spectator, New Haven.
Bonnet, Charles (1770 – 71): Analytischer Versuch über die Seelenkräfte, transl.
Christian Gottfried Schütz, Göttingen.
Brandt, Reinhard (1999): Kritischer Kommentar zu Kants Anthropologie in
pragmatischer Hinsicht, Hamburg.
Bricke, John (1995): Hume’s Conception of Character, in: S. Tweyman (ed.):
David Hume: Critical Assessments, 6 vols., London, vol. VI, pp. 248 – 254.
Cohen, Alix (2009): Kant’s Concept of Freedom and the Human Sciences, in:
Canadian Journal of Philosophy 39, pp.113 – 136.
Firla, Monika (1981): Untersuchungen zum Verhältnis von Anthropologie und
Moralphilosophie bei Kant, Frankfurt, Bern.
Frankfurt, Harry G. (1988): Freedom of the Will and the Concept of a Person,
in: Frankfurt, Harry: The Importance of What We Care About, Cam
bridge, pp. 11 – 25.
Frierson, Patrick (2005): Kant’s Empirical Account of Human Action, in: Phi
losophers’ Imprint 5, No. 7, pp. 1 – 34.
Hinske, Norbert (1966): Kants Idee der Anthropologie, in: H. Rombach (ed.):
Die Frage nach dem Menschen, Tübingen, pp. 410 – 427.
Hudson, Hud (1994): Kant’s Compatibilism, Ithaca, London.
Hume, David (1748 – 51): Enquiries Concerning Human Understanding and
Concerning the Principles of Morals, ed. L. A. Selby Bigge (3rd ed. P.
H. Nidditch), Oxford 31975.
Hume, David (1754 – 1762): The History of England, London (rev. edition
London 1778, reprint 1983).
Johnson, Clarence Shole (1990). Hume on Character, Action, and Causal Ne
cessity, in: Auslegung 16, pp. 149 – 164.
Kant, Immanuel (1781/1787): Kritik der reinen Vernunft, ed. Jens Timmer
mann, Hamburg 1996.
Kant, Immanuel (1900 ff.): Gesammelte Schriften, Akademie Ausgabe, Berlin.
Kim, Soo Bae (1994): Die Entstehung der Kantischen Anthropologie, Frankfurt
am Main.
Marquard, Odo (1965): Zur Geschichte des Begriffs ‘Anthropologie’ seit dem
Ende des achtzehnten Jahrhunderts, in: Collegium Philosophicum, Basel,
Stuttgart, pp. 209 – 239.
42 Thomas Sturm
McIntyre, Jane L. (1990): Character: A Humean Account, in: History of Phi

losophy Quarterly 7, pp. 193 – 206.
Meerbote, Ralf (1984): Kant on the Nondeterminate Character of Human Ac
tions, in: W. A. Harper, R. Meerbote (eds.): Kant on Causality, Freedom,
and Objectivity, Minnesota, pp. 138 – 163.
Penelhum, Terence (1993): Hume’s Moral Psychology, in: D. F. Norton (ed.):
The Cambridge Companion to Hume, Cambridge, pp. 117 – 147.
Platner, Ernst (1772): Anthropologie für Aerzte und Weltweise, Leipzig.
Schütz, Christian Gottfried (1771): Betrachtungen über die verschiednen
Methoden der Psychologie; nebst einem kritischen Auszug aus des Hrn.
Abt von Condillac Traité des sensations, in: C. Bonnet: Analytischer Ver
such über die Seelenkräfte, Göttingen 1770 – 71, vol. II, pp. 187 – 273.
Stroud, Barry (1977): Hume, London.
Sturm, Thomas (2001): Eine Frage des Charakters: Kant über die empirische
Erklärung freier Handlungen, in V. Gerhardt, R. P. Horstmann, R. Schu
macher (eds.): Kant und die Berliner Aufklärung, Akten des IX. Internatio
nalen Kant Kongresses, Berlin, vol. 4, pp. 440 – 449.
Sturm, Thomas (2009): Kant und die Wissenschaften vom Menschen, Pader
born.
Tetens, Johann Nicolaus (1777): Philosophische Versuche über die menschli
che Natur und ihre Entwickelung, 2 vols., Leipzig (reprint Hildesheim,
New York 1979).
Van de Pitte, Frederick P. (1971): Kant as a Philosophical Anthropologist, The
Hague.
Wilson, Holly L. (2006): Kant’s Pragmatic Anthropology: Its Origin, Meaning,
and Critical Significance, Albany, NY.
Wood, Allen W. (1984): Kant’s Compatibilism, in: A. W. Wood (ed.): Kant on
Self and Nature, Ithaca, London, pp. 73 – 101.
Systematic Classification or Purposive Moralization?
On why Teleology is not the (only) Key to Kant’s
Anthropology from a Pragmatic Point of View
Liesbet Vanhaute
Abstract
This paper is intended to place the suggestion that Kant’s Anthropology from a
Pragmatic Point of View is an essentially teleological work, in the right perspec
tive. Kant does indeed make use of teleological judgments in his anthropology,
but this form of judgment does not provide an exhaustive characterization of
the work. The assumption that it does unduly stretches Kant’s concept of tel
eological judgment. I investigate the merits and deficiencies of the teleological
reading. Though to a certain extent the reading is confirmed by Kant’s critical
theory of teleological judgment, it leads to two problematic implications. The
teleological reading implies first, that non teleological disciplines like empirical
psychology are too unsystematic to be included in pragmatic anthropology. But
though Kant doubted empirical psychology could ever become a full fledged
science, he did believe it could bring about systematic, non teleological infor
mation that is valuable for anthropology. The teleological reading secondly im
plies that pragmatic anthropology can be classified as ‘moral anthropology’.
Contrary to this, it can be argued that pragmatic anthropology as a whole (so
not only insofar as it contains empirical psychology) can do without teleological
presuppositions. This means that the focus on human morality is not predom
inant and that Kant’s Anthropology from a Pragmatic Point of View cannot be re
duced to moral anthropology.
Introduction
Several Kant scholars1 have recently suggested that Kant’s Anthropology

from a Pragmatic Point of View should be understood as an essentially tel-
eological work. This paper is intended to place this suggestion in the right
1 Alix Cohen, Robert B. Louden and Allen Wood have all, in different ways,
made the claim that pragmatic anthropology is teleological. Cohen (2008,
511): “The guiding principles and procedures of Kant’s anthropological meth
od are, I believe, essentially teleological”; Louden (2003, 72): “The strong tel
44 Liesbet Vanhaute
perspective by showing that Kant does indeed make use of teleological

judgments in his anthropology, but that this form of judgment does not
provide an exhaustive characterization of the work. The assumption
that it does unduly stretches Kant’s concept of teleological judgment
and overlooks another form of judgment that also plays a role in Anthro-
pology from a Pragmatic Point of View.
The first section of this paper investigates the merits and deficiencies
of this teleological reading. After showing that it derives some confir
mation from Kant’s critical theory of teleological judgment, I will indi
cate two implications that result from it. The teleological reading im
plies first that non teleological disciplines like empirical psychology
are too unsystematic to be included in pragmatic anthropology, and sec
ond, that pragmatic anthropology can be classified as ‘moral anthropol
ogy.’ The second and third sections of the paper will argue that these
implications are problematic: pragmatic anthropology does incorporate
some empirical psychology, and it cannot be reduced to moral anthro
pology. This will allow me to conclude that Kant’s pragmatic anthro
pology is more than a teleological study of human nature.
1. Kant’s pragmatic anthropology and teleological judgment

Kant’s Critique of Judgment certainly suggests a strong connection be
tween teleological judgment and human nature. Insofar as commenta
tors emphasize this, they are faithful to Kant’s writings. However,
when they further argue that Anthropology from a Pragmatic Point of
View is primarily a teleological work, they overburden Kant’s theory
of teleological judgment. This first, introductory section will show
that Kant believes teleological judgment is best suited to understand a
specific range of phenomena, including some features of human nature
(1.1). After that, it will suggest that it is nevertheless problematic to in
terpret pragmatic anthropology as consisting exclusively out of teleolog
ical judgments (1.2).
eological thrust of these descriptions […] within the anthropology lectures also
serves as a correction to the view that Kantian anthropology is simply empirical
science”; Wood (2003, 45): “Insofar as Kant has a conception of its methods at
all, he thinks of anthropology as following the looser method of biology, based
on regulative principles of teleological judgment”.
Systematic Classification or Purposive Moralization? 45
1.1. Kant’s theory of teleological judgment
Kant understands teleological judgment as a form of reflective judg

ment. Reflective judgments do not simply apply a priori principles to ex
perience: they search for rules and concepts that are not (yet) given.2 As
such, they are useful to understand realms of experience that cannot im
mediately be understood as constituted by the categories of the under
standing or the principles of practical reason.3 The activity of reflective
judgment consists in striving for systematic comprehension of these
realms. In the course of this activity, specific assumptions prove to be
heuristically useful.4 The most basic, transcendental presupposition
which all forms of reflective judgment share is that nature is systematic
enough for judgment’s activity to be successful. Reflecting judgment as
sumes that general rules and principles can be found, even in seemingly
contingent, mere empirical phenomena. Kant terms this assumption
“the subjective purposiveness of nature.”5
Teleological judgment is one type of reflective judgment. Human
reason resorts to it whenever it encounters specific phenomena that
can only be understood by conceptualizing them in terms of purposes.
In addition to transcendentally assuming subjective purposiveness, teleo
logical judgment also holds that it is heuristically necessary to assume
real, objective ends. In Critique of Judgment, Kant indicates that the phe
nomenon most in need of such conceptualization is that of organic na
ture.6 But because reason successfully judges organisms as (internally)
purposive, it necessarily7 starts to look for (external) purposive relations
between individual phenomena in nature.8
2 First Introduction to the Critique of Judgment (FI), AA 20:209.

3 In Critique of the Power of Judgment (CJ), AA 5:175, Kant argues that there is a
form of cognition which, unlike the understanding or practical reason, does not
legislate through a priori principles. This cognition finds empirical rules and
concepts that are different from the laws of practical reason or the categories
of the understanding.
4 See Guyer (2003) and FI, AA 20:214.
5 FI, AA 20:202
6 CJ, AA 5:369. For an in depth analysis of the internal teleological judgment on
organisms, see Beishart (2009).
7 CJ, AA 5:379 – 380: “[…] only matter insofar as it is organized […] necessarily
carries with it the concept of itself as a natural end […]. However, this concept
necessarily leads to the idea of the whole of nature as a system in accordance with
the rule of ends” (my italics).
46 Liesbet Vanhaute
Kant makes it expressly clear that external teleology is more prob

lematic than internal teleology. Whereas he describes the judgment
on organic organization as immediately and necessarily asserting that a
thing in nature is internally structured in a purposive way, he argues
that judgments on external purposiveness contain an element of contin
gency: one could say that a thing A is the purpose of a thing B, but also
that this thing B is the purpose of thing A.9 The only way to overcome
the triviality of judgments on external purposiveness, according to Kant,
is to assume that events in nature have an absolute end. But the only co
herent way to think of such an absolute end (that is no longer a means to
anything) is by understanding it as human morality.10 Hence, Kant ar
gues that if we want to acknowledge purposive relations between nat
ural phenomena, we are to understand them as having one final pur
pose. This purpose can only be human morality. To see progression
in nature, we must assume human morality as the end of natural evolu
tion.
With this theory of external purposiveness, Kant critically legiti
mates the possibility of progressive history in Critique of Judgment. If
morality is taken as its ultimate goal, history can be understood reflec
tively as a teleological process guided by the interaction between
human nature and the challenges posed by hostile, external nature.11
From this perspective, human nature can be seen as fundamentally dis
posed to develop reasonably and morally throughout every generation.
Humanity’s evolution can be captured in external teleological judg
ments; its dynamic can be understood as having a final purpose. In
this way, external teleological judgment allows us to grasp the otherwise
unintelligible realm of historical events.12
8 Its purpose here is to conceive nature in accordance with the idea of “the whole
of nature as a system in accordance with the rule of ends.” CJ, AA 5:379 – 380.
9 CJ, AA 5:426 – 427.
10 CJ, AA 5:448: “it is a fundamental principle, to which even the most common
human reason is compelled to give immediate assent, that if reason is to provide
a final end a priori at all, this can be nothing other than the human being (each
rational being in the world) under moral laws.”
11 See, for example, CJ, AA 5:432.
12 On the unintelligible character of non teleological history, see Idea for a Univer
sal History with a Cosmopolitan Aim (IUH), AA 8:17.
The concept of a human predisposition is an essential part of this ex

ternal teleological view on human history.13 History can be understood
as a coherent whole by focusing on the development of the human pre
disposition to reason.14 The teleological perspective thus provides us
with a coherent, systematic theory of human nature insofar as it allows
us to understand humanity as developing towards a moral end. The in
dividual human being is part of this evolution and carries a predisposi
tion that will unfold throughout history.
1.2. Teleology, empirical psychology and moral anthropology
Kant’s firm conviction that the teleological perspective is necessary to

conceive of certain aspects of human nature has led some commentators
to presume that every anthropology must adopt a teleological view. The
preceding elaboration on teleological judgment has shown that this
makes sense to a certain extent: human nature in its historical develop
ment is a phenomenon that must be understood teleologically. But does
this mean that pragmatic anthropology consists entirely of teleological
judgments; that all its “observations and reflections are subordinated
to the ethical final ends of human existence”? 15 I will argue that this
is not the case by showing that the teleological reading is deeply inter
woven with the following two problematic assumptions.
1) A teleological understanding of human beings can bring about a co
herent, systematic interpretation of human development. The teleological
reading of Anthropology from a Pragmatic Point of View thus has the advantage
that it can explain in which manner the work is systematic.16 This is some
thing Kant is deeply concerned about: he stresses that pragmatic anthropol
13 Kant introduces the concept of a ‘predisposition’ in the context of biological

problems—the concept thus seems to have its proper place in internal teleolog
ical judgments about biological (organic) nature. Nevertheless, in CJ, AA
5:378, Kant explicitly states that the organic model of ‘natural ends’ (i. e. organ
isms) cannot be applied to human behavior. Insofar as history takes up the talk
of ‘predispositions’ to understand the history of human behavior, the concept
there takes on another, non biological meaning (see Vanhaute (2011)). This
contradicts Wood’s suggestion, quoted in footnote one, that pragmatic anthro
pology would follow biology’s methods.
14 IUH, AA 8:17.
15 Wilson (2006, 24).
16 Wilson (2006, 96): “judgments of purpose allow us to judge in a systematic way
instead of merely enumerating an aggregate of experience.”
48 Liesbet Vanhaute
ogy cannot be an unsystematic collection of random facts.17 But can we

conclude from this concern with systematicity that the perspective of prag
matic anthropology is identical to that of teleological judgment? This is
only possible if we assume that Kant can only concieve of one kind of sys
tematic theory of human nature, namely a teleology of human nature. He
would have to be of the opinion that non teleological ways of studying the
human being, like empirical psychology, are fundamentally deficient.
This is not Kant’s view. He believes human phenomena can be sys
tematically understood without teleological presuppositions. In fact, it is
exactly his view on empirical psychology that shows most clearly that
this is the case. This view will be examined extensively in the next sec
tion of this paper.
2) If the teleological reading were correct, Anthropology from a Pragmatic
Point of View would assume a moral end of human evolution. Its perspec
tive would require a focus on the morally predisposed nature of human be
ings. This focus is unavoidable: a non trivial teleology of human nature re
quires the assumption of morality as an end (cf. supra, 1.1). As such, the
teleological reading suggests that pragmatic anthropology is primarily con
cerned with those features of human beings that facilitate morality. It
would amount to the discipline that Kant calls ‘moral anthropology’:
The counterpart of a metaphysics of morals […] would be moral anthro
pology, which […] would deal […] with the subjective conditions in
human nature that hinder people or help them in fulfilling the laws of a met
aphysics of morals.18
Such a pragmatic anthropology would ignore more accidental features
of human behavior, features that are not related to the (lack of) unfold
ing of moral predispositions. The third section of this paper will show
that this conflicts with Kant’s intentions in the pragmatic anthropology.
Though human morality plays a role in Anthropology from a Pragmatic
Point of View, we cannot just conclude that Kant saw the work as a
‘moral anthropology’ or as concerned with the historical development
of human predispositions. The perspective of pragmatic anthropology
may include a teleological perspective, but it cannot be reduced to it.
Pragmatic anthropology has its own methodological singularities and
advantages, and these are overlooked if the work is read as primarily
concerned with teleological judgment.
17 Anthropology from a Pragmatic Point of View (AP), AA 7:119.

18 Metaphysics of Morals, AA 6:217.
2. Pragmatic anthropology contains empirical psychology
Kant defines empirical psychology as the study of the experiences of

inner sense. It is a discipline that strives to be a science, but that will
probably never become one. This section considers whether these
doubts led Kant to believe that empirical psychology is deficient and un
systematic. I argue that there are reasons to believe that Kant wanted
empirical psychology to play a role in pragmatic anthropology (2.1). I
then show that he nevertheless perceived a number of problems
which empirical psychology must avoid (2.2). If it succeeds in avoiding
its inherent problems, empirical psychology can generate useful infor
mation. Kant insists that this information can be systematic (2.3).
2.1. Empirical psychology is included in pragmatic anthropology
In the published version of the Anthropology from a Pragmatic Point of

View, Kant does not explicitly comment on the place of empirical psy
chology. He did make several statements on the issue, however, in the
decade before publishing this work. These remarks often assert that em
pirical psychology can provide only a very limited amount of knowl
edge.19 Nevertheless, Kant is concerned about the place this knowledge
should take in the system of sciences. He observes that many have treat
ed it as a part of metaphysics, but that this is not the appropriate place for
it. It would be better if it were to become a part of anthropology:
Empirical psychology must thus be entirely banned from metaphysics […]
it can establish its own domicile in a complete anthropology.20
Contrary to Holly Wilson,21 and in agreement with Thomas Sturm,22 I
maintain that Kant never really changed his mind on this point. Al
19 Vorlesungen ber Metaphysik 1790 – 1791, AA 28:735, Vorlesungen ber Metaphysik

Volckmann 1784 – 1785, AA 28:367.
20 Critique of Pure Reason A 848 – 849/B 876 – 877.
21 Wilson (2006, 21 – 24) seems to suggest that Kant may have included empirical
psychological data in anthropology, but always within the context of teleolog
ical judgments. Without teleological additions, empirical psychology is too un
systematic to be part of pragmatic anthropology.
22 Sturm’s book, Kant und die Wissenschaften vom Menschen, presents a comprehen
sive analysis of Kant’s thoughts on the relation between empirical psychology
and anthropology. I believe that he makes a very strong point, and I make fre
quent reference to his work throughout this section.
50 Liesbet Vanhaute
though he doubts that empirical psychology will ever become a science

(these doubts are the topic of the next section), he never explicitly ar
gues against including the small amount of knowledge that it can pro
duce in anthropology. I present three arguments to support this claim.
The first refers to the relation of pragmatic anthropology to what
Kant calls ‘knowledge of the schools.’ A second argument concerns
the structure of anthropology, and a brief final argument refers to the
specific content of anthropology. It is important to note that these argu
ments are not intended to show that Kant’s anthropology can be re
duced to empirical psychology, or that it relies heavily on empirical psy
chology. I present them only because they suggest that Kant does not
explicitly reject the inclusion of empirical psychological data in prag
matic anthropology.
1) At first sight, my initial argument might appear to be a rejection
of empirical psychology. In an anthropology lecture from 1791, Kant
remarks that empirical psychology should be classified as “school
knowledge” (Schulkenntniss).23 He contrasts this form of knowledge
with “world knowledge” (Weltkenntniss), as found in pragmatic anthro
pology. Understood correctly, however, this remark appears to be an ar
gument for rather than against the interrelatedness of anthropology and
empirical psychology. Note that Kant’s intention in asserting that Welt-
kenntniss is different from Schulkenntniss is to say that neither of the two
forms of knowledge can be reduced to the other. He certainly does not
want to argue that they should be separated completely. On the contra
ry, he believes that they are complementary. School knowledge necessa-
rily precedes world knowledge.24 Hence, having school knowledge (of
which psychology is a part) is a prerequisite for finding world knowl
edge (like anthropology). As such, the distinction between school
knowledge and world knowledge certainly does not rule out the possi
bility that some empirical psychological knowledge could find its way
into the worldy pragmatic anthropology.25 On the contrary, it would
actually make sense that anthropology occasionally uses empirical psy
chological knowledge, which Kant defines as knowledge of the appear
ances of the soul. Cognition of the human psyche can be helpful in the
23 Anthropologie Dohna Wundlacken 1791, Ausgabe Kowalewski s.72

24 AP, AA 7:120, see also Sturm (2009, 293 – 296).
25 Sturm (2009, 189 – 191) argues that anthropology probably incorporates psy
chological knowledge, but that Kant’s conception of anthropology differs
from the conception of empirical psychology.
quest for useful world knowledge about human beings.26 Without some
insight in the human soul, we cannot acquire useful knowledge of
human behavior (see also infra, 2.2, Objection 3).
2) The structure of the Anthropology from a Pragmatic Point of View
confirms the inclusion of some elements of empirical psychology. The
tripartite division of the opening section (“On the cognitive faculty”,
“The feeling of pleasure and displeasure” and “On the faculty of desire”)
is taken from empirical psychology; it corresponds to what Kant consid
ers to be the structure of the human soul.27 We experience this division
by means of our inner sense, which is the source of empirical psycho
logical data. The threefold division of human mental powers is thus
based on a datum from empirical psychology. Pragmatic anthropology
uses this structure as a guideline for research on human behavior.
Hence, Kant does not seem to object to the use of empirical psychology
in anthropology. This is also clear from his use of Baumgarten’s Psycho-
logia Empirica as a handbook for his lectures. Kant notes that he uses it as
a guiding thread for structuring his lessons. The structure of anthropol
ogy is thus fundamentally influenced by empirical psychology. Kant
maintained this structure throughout all of his anthropological lectures,
as well as in the published Anthropology from a Pragmatic Point of View.
This demonstrates that empirical psychology has always played a certain
role in the development of pragmatic anthropology.
3) That Anthropology from a Pragmatic Point of View can be character
ized as partially based on empirical psychology is also clear on a very
specific, textual level. When legitimating his use of Baumgarten’s
work, Kant not only claims that he uses it to structure his lectures, he
also notes that it provides useful material. 28 Some of the content of
Baumgarten’s empirical psychology is included in pragmatic anthropol
26 Empirical psychology proceeds from self perception. Although self perception

is certainly not the only way in which we can arrive at anthropological knowl
edge, Kant does not rule it out altogether as a source of anthropological knowl
edge. He remarks that self perception is often the basis of our judgment of oth
ers. AP, AA 7:143: “… knowledge of the human being through inner experi
ence, because to a large extent one also judges others according to it, is more
important than correct judgment of others.”
27 Metaphysik Mrongovius 1782 – 1783, AA 29:877.
28 Menschenkunde 1781 – 1782, AA 25:859: “Da es kein anderes Buch über die
Anthropologie gibt, so werden wir die metaphysische Psychologie Baumgart
ens, eines Mannes, der sehr reich in der Materie; und sehr kurz in der Ausfüh
rung ist, zum Leitfaden wählen.”
52 Liesbet Vanhaute
ogy. Moreover, Kant states several times29 that he is referring to matters

of psychology. For example, when discussing suicide, he mentions that
the question of whether suicide presupposes courage is a psychological
one. He then immediately starts answering it, and we can only assume
that the answer is composed of psychological ideas.30 He also introduces
his discussion of the temperaments by stating his intention to analyze
them psychologically.31
Although pragmatic anthropology certainly cannot be reduced to
empirical psychology, I believe that data from the latter discipline are
indisputably included in the former discipline. But in which form? It
is certainly not in the form of a full fledged science: Kant denies that
empirical psychology can be scientific. He formulates a number of ob
jections against scientific empirical psychology, and these objections in
fluence the way in which empirical psychology is used in pragmatic an
thropology. I present these objections in the following section.
2.2. Problems with empirical psychological knowledge
Kant does not consider the empirical psychological data that are includ
ed in anthropology to be results of proper scientific research. Empirical
psychology can provide information that is correct and relevant enough
to become part of pragmatic anthropology, but that does not mean that
it is a proper science. This is enigmatic: how can empirical psychology
provide some (but not very many32) valuable facts without being a sci
ence?
Kant discusses the problems of empirical psychology in a short and
dense fragment of Metaphysical Foundations of Natural Science. In his study
Kant und die Wissenschaften des Menschen, Thomas Sturm distills from this
(and from some other passages in Kant’s work) four problems with em
pirical psychology. One problem is that its method of introspection is
29 AP, AA 7:214, 7:238.

30 AP, AA 7:258.
31 AP, AA 7:286. Kant writes that we can understand temperaments as conse
quences of physiological phenomena, or that we can use physiological termi
nology as a mere metaphor for psychological phenomena. In the next section,
he states that he will use the physiological terminology only for classification.
Apart from this his approach is thus psychological.
32 Vorlesungen ber Metaphysik 1790 – 1791, AA 28:735, Vorlesungen ber Metaphysik
Volckmann 1784 – 1785, AA 28:367.
psychologically unhealthy. This is not, however, an argument against

empirical psychology as a science (and Sturm does not present it as
one either): many sciences can be psychologically unhealthy but still
be proper sciences. I shall ignore this objection to empirical psychology
and discuss the three remaining objections by referring to both Kant’s
original text and Sturm’s interpretation of it. The objections I present
are that (1) empirical psychological data can be mathematized only to
a very limited extent; (2) no psychological experiments are possible
and (3) the objective observation of psychological phenomena is impos
sible.
1) Sturm offers the following interpretation of Kant’s belief that
“mathematics is not applicable to the phenomena of inner sense:”33
for a discipline to be a science, Kant requires that it studies data that
can be mathematized. Only when a priori categories can be applied to
such data can scientific laws be derived from them. Intuitions can be
fully mathematized if they are constituted in space and time, the a priori
structures of experience. External experience is constituted in both
space and time. Inner experience, which is the basis of empirical psy
chology, is only temporal and not spatial. This does not mean that psy
chological phenomena cannot be mathematized at all, but that this is
possible only to a limited extent. Psychological phenomena can be
mathematized only in so far as their temporal extension can be meas
ured. But how is this possible? It makes sense to suggest that, in princi
ple, psychological phenomena could be timed: we could quantify them
in terms of minutes and seconds. What form these measurements should
take, however, is completely unclear.34 Internal experience takes the
form of an entirely temporal permanent flux in which one mental
state constantly evolves into a next one. There is no ‘space’ in between
psychological phenomena; although they are somehow distinct, it is not
clear where one ends and the next begins. For this reason, it is impos
sible to know when to start measuring and when to stop. Hence, al
though it is in principle possible to mathematize psychological phenom
ena to a certain extent (by measuring temporal phenomena), it is not
clear how this would appear in practice. Because of this, Kant doubts
that empirical psychology will ever become a science.
It is important to note that the preceding characterization of empir
ical psychological phenomena does not imply that they are marked only
33 Metaphysical Foundations of Natural Science (MFNS), AA 4:471.

34 Sturm (2009, 238 – 239).
54 Liesbet Vanhaute
by their temporal dimension. Sturm remarks that Kant sometimes sug

gests that, in addition to duration, psychological phenomena also have a
dimension of intensity.35 Kant believes that we can state with quite some
certainty that one psychological experience is either more or less intense
than another; this allows for a certain classification of psychological phe
nomena. Such classicification, however, is based only on a comparison
of the intensity of psychological states. An exact and complete mathe
matization of the intensity of these phenomena is just as problematic
as is the adequate measurement of their temporal dimension.
2) The lack of secluded phenomena, which is the root of the prob
lem of mathematization, is also the reason for Kant’s second objection to
scientific empirical psychology: because there are no separate occurenc
es, we cannot ‘recombine them at will’:
[T]he manifold of inner observation can be separated only by mere division
in thought, and cannot then be held separate and recombined at will (but
still less does another thinking subject suffer himself to be experimented
upon to suit our purpose).36
According to Sturm, Kant here refers to the problematic fact that no ex
periments are possible in empirical psychology. Neither my own, nor
someone else’s psychological states can be adequately manipulated,
and it is therefore impossible to experiment on them. In a proper sci
ence, it is possible to manipulate phenomena: situations are created in
order to determine whether a phenomenon occurs as a predictable result
of them. In psychology, this is impossible. Because there are no secluded
phenomena, it is impossible to state with certainty that a certain event
re occurs regularly.
I want to emphasize the fact that Kant does not say that the psycho
logical flux cannot be divided at all. He writes that inner observation can
be “separated by mere division in thought:”37 we can assume a certain
division in our thinking about these phenomena. Such a division can
serve as a ‘working hypothesis’ to which we can resort when speaking
and thinking about these phenomena. Even if it is impossible to identify
exactly when a certain psychological state (e. g. a moment of sadness)
started and ended, it is still possible to refer to it as distinct from the
states occurring before and after it.
35 Sturm (2009, 242 – 250).

36 MFNS, AA 4:471.
37 Ibid.
3) Even if we are not setting up an experiment, but simply observing

our own psyche, we can never be sure if we get to know our object as it
is: “observation by itself already changes and displaces the state of the
observed object.”38 In this assertion, Kant is alluding to the familiar phe
nomenon that, by directing our attention towards our inner experien
ces, these experiences can change: self consciousness has an influence
on our mental states.
It is not clear, however, whether Kant believes that this makes ob
jective self observation only very difficult or completely impossible. In
his metaphysics lectures of 1792 93, he asserts the impossibility of psy
chological experiments. 39 He attributes this not to the lack of secluded
phenomena but to the fact that experimentation changes the phenomenon
we try to observe. This is contrasted with self observation, which is very
hard but not impossible. This suggests that only the awareness of being
involved in psychological experimentation fundamentally alters the psy
chological state ‘mere’ self observation does not necessarily do this.
The published Anthropology also appears to reflect the belief that self ob
servation is difficult, but not impossible:
[…] knowledge of the human being through inner experience, because to a
large extent one also judges others according to it, is more important than
correct judgment of others, but nevertheless at the same time perhaps more
difficult. For he who investigates his interior easily carries many things into
self consciousness instead of merely observing.40
Kant thus believes that self observation is problematic and an inappro
priate basis for a science. However, in so far as it is necessary in order
to make correct judgments about other human beings, it appears to
be unavoidable (which is probably one of Kant’s reasons for including
empirical psychological knowledge in pragmatic anthropology, cf.
supra 2.2., Argument 1).
Though the three difficulties listed above make it impossible for em
pirical psychology to become a full fledged science, they certainly do
not rule out the possibility of interesting and relevant empirical psycho
logical data. Empirical psychology can avoid its inherent problems: it
38 Ibid.
39 Metaphysik Dohna 1792 – 93, AA 28:679: “wir verfahren überhaupt method
isch, durch observiren, oder experimentiren; das erste ist schwer, und das letz
tere unmöglich; denn das Experiment was wir machen, ändert schon unsern
Gemüthszustand.”
40 AP, AA 7:143.
56 Liesbet Vanhaute
can divide psychological phenomena as a ‘working hypothesis’ and

compare the length and intensity of the phenomena that result from
this division. It should avoid psychological experimentation, which is
deeply problematic, but it can base itself to some extent on self obser
vation. If empirical psychology takes these guidelines into account
(avoid experiments, be careful with self observation, …) it can navigate
around its inherent problems. Although it will probably never become a
science in the narrow, Kantian sense of the term, it can provide us with
some insight into our own psyche. Without this self insight, it would be
very difficult to understand others. Therefore, psychological insights
play a role in pragmatic anthropology: individuals cannot make sense
of the behavior of others without having some understanding of their
own mental lives.
2.3. How systematic can empirical psychological data be?
One might offer the objection that the findings of empirical psychology
are fragmented and diffuse. If empirical psychology must indeed slalom
around the various difficulties listed above, it would be logical to expect
it to generate only disparate and incoherent results. If this is the case, it
would make sense to argue (cf. supra, 1.2.(1)) that Kant thought that
empirical psychological findings should be included in teleological
judgments in order to ensure that they would be a proper part of sys
tematic pragmatic anthropology.
But Kant never thought of empirical psychology as entirely unsyste
matic. Even in the Metaphysical Foundations of Natural Science, in which
he presents the three above listed objections to scientific empirical psy
chology, he describes it as a discipline that arrives at a certain level of
systematicity:
Therefore, the empirical doctrine of the soul can never become anything
more than […] a natural doctrine of inner sense which is as systematic as
possible, that is, a natural description of the soul, but never a science of
the soul […].41
In Critique of Pure Reason, Kant states that empirical psychology seeks to
explain phenomena,42 hence not just to enumerate disparate facts. He
41 MFNS, AA 4:471.
42 Critique of Pure Reason A 347/B 405.
makes a similar suggestion in a lecture on logic from 1789 90, in which

he states that empirical psychology studies the human soul’s phenomena
by placing them under general rules.43 If empirical psychology searches
for such general rules, then it must be more than a mere collection of
disparate facts. It must be able to achieve systematic results.
The key to understanding Kant’s conception of empirical psychol
ogy as a search for systematic rules in empirical data can be found in
the the First Introduction to the Critique of Judgment, where Kant notes
the following:
If what is to be found is merely the ground for the explanation of that
which happens, then this can be either an empirical principle, or an a priori
principle, or even a composite of the two, as one can see in physical me
chanical explanations of events in the corporeal world, which find their
principles in part in the general (rational) science of nature, and partly in
those sciences which contain the empirical laws of motion. Something sim
ilar takes place when one seeks for psychological grounds of explanation for
what goes on in our mind, only with this difference that, as far as I am
aware, the principles for this are all empirical, with only one exception,
namely the law of the continuity of all changes (since time, which has
only one dimension, is the formal condition of inner intuitions), which
is the a priori ground of these perceptions, but which is virtually useless
for the sake of explanation.44
The discipline of empirical psychology has thus only one a priori princi
ple, and that principle is useless for explanation. It therefore resorts to
‘empirical laws’: general rules that are separate from this a priori princi
ple. Empirical psychology uses these empirical laws to classify and ex
plain facts; because of this, it can provide systematic information.
Elsewhere in the First Introduction and the Critique of Judgment, these
empirical laws are understood as the result of an activity of the reflecting
power of judgment. This, however, is the general heading under which
Kant treats teleological and aesthetic judgment, which suggests that
the empirical laws of psychology are teleological judgments (that they
would be aesthetic judgments is rather unplausible). Is this the case?
As I explained in the first section of this paper, Kant states that we resort
to teleological judgments if we are dealing with phenomena like organ
isms and human history: realms of facts that cannot be reasonably under-
stood without introducing the concept of an objective end. Empirical psychol
43 Logik Busolt 1789 – 90, AA 14:611.

44 FI, AA 20:237.
58 Liesbet Vanhaute
ogy studies neither organisms nor human history.45 It searches for em

pirical generalities within the realm of inner experience. Kant’s argu
ment in Critique of Judgment (cf. supra, 1.1.) suggests that it would be
superfluous to take the teleological perspective in order to find such em
pirical laws: it suffices to assume that nature as a whole is subjectively pur-
posive with regard to our cognitive capacities, that it lends itself to find
ing systems of empirical laws, even in seemingly contingent facts like
those provided by inner experience. Empirical psychology does not
need to resort to objective purposiveness, hence it does not proclaim
teleological judgments. Asserting that it does comes down to unduly
stretching Kant’s concept of teleological judgment.
This means that empirical psychology proclaims judgments that are
neither teleological nor aesthetic, but which Kant nonetheless considers
reflective. Kant thus seems to assume that there is a form of reflective
judgment that cannot be reduced to either teleological or aesthetic judg
ment. Paul Guyer has noticed this (not specifically in relation to Kant’s
thoughts about empirical psychology) and has argued that Kant assumed
a third46 form of reflective judgment: judgment that searches for empir
ical laws and concepts.47
This third type of reflective judgment exists in the capacity to find
empirical laws and principles that are able to systematize facts underde
termined by a priori principles. By drawing upon this capacity, empirical
psychology can find an order in the experiences it studies, even if it will
45 Empirical psychology does not deal with human beings as either organisms (as
biology does) or as part of a historically developing species (it studies the data of
inner sense rather than the historical development of mankind).
46 In fact, Guyer (2003, 1 – 62) maintains that there are five forms: aesthetic judg
ment and teleological judgment are general terms, each of which incorporates
two forms of reflecting judgment. For the sake of clarity, however, I will ab
stract from this part of his argument, even though I believe it is correct. In
the remainder of this article, I will thus speak of a third form of reflecting judg
ment.
47 Guyer (ibid.) presents this form of reflective judgment as playing an important
role in the emergence of systems of natural science. It finds regularities that con
nect the a priori categories of a science to a posteriori facts. One could object that
my interpretation of empirical psychology as reflecting judgment goes against
Guyer’s point because empirical psychology is not a natural science. However,
it is possible to understand empirical psychology as analogous to a natural sci
ence in a very early stage. It has only one a priori principle and seeks to add em
pirical regularities to this meager basis. For this reason, it searches for empirical
laws and concepts. It will not find as many of these as a real science would, but
that does not mean it cannot assume the same method.
never become a full fledged science. As a search for coherence in inner

experience, empirical psychology does not stand in any direct connec
tion to the final end of human beings, morality (cf. infra). In spite of
this, it plays a role in pragmatic anthropology. I admit that this role
might be limited, but it cannot be ignored. On the contrary, it deserves
to be emphasized: it shows that pragmatic anthropology is not fully
composed of teleological judgments. In so far as it takes up data from
empirical psychology, pragmatic anthropology is systematic in a non
teleological way. This suggests that a systematic study of human beings
is not completely dependent on assumptions about objective ends or
(moral) predispositions. At least one other strategy is possible: that of
careful, systematic self observation.
3. Pragmatic anthropology is more than moral anthropology
As we saw in the first section, Kant writes that moral anthropology fo

cuses specifically on those aspects of human nature that promote or hin
der morality. The teleological reading of Anthropology from a Pragmatic
Point of View seems to imply that it is the equivalent of moral anthropol
ogy: it studies the human being from the standpoint of a priori ethics.
But is this correct?
3.1. Empirical psychology’s blindness for morality
Kant’s writings make it quite obvious that whenever pragmatic anthro

pology resorts to the findings of empirical psychology, it will not con
tain any reference to morality. The free ego that feels moral motivation
does not appear empirically to inner sense; and for this reason it cannot
be grasped by empirical psychology.
In the self cognition of the human being through inner experience […] he
can only recognize himself as an object through his representation in expe
rience as he appears to himself, not as he, the observed, is in himself.—If he
wished to cognize in the latter way, he would have to rely on a conscious
ness of pure spontaneity (the concept of freedom), (which is also possible),
but it would still not be perception of inner sense and the empirical cog
nition of his inner self (inner experience) which is based on it. Rather, it
can only be consciousness of the rule of his actions and omissions, without
thereby acquiring a theoretical (physiological) cognition of his nature,
which is what psychology actually aims at.—Empirical self cognition
60 Liesbet Vanhaute
therefore presents to inner sense the human being as he appears to it, not as
he is in himself […].48
3.2. Pragmatic anthropology versus teleological history
It is not only in view of the empirical psychological elements in anthro

pology that the teleological interpretation of the work is untenable. The
minor interest in teleology is characteristic for the work as a whole. This
can be noticed by comparing the pragmatic anthropology to Kant’s de
scriptions of history.49 The latter discuss the development of human pre
dispositions as they stand in interaction with nature. They are able to
teleologically judge history as a progressive evolution towards mankind’s
final, moral end and as such they answer questions that arise from a spec
ulative, theoretical interest.50 To get a grasp of their subject matter, they
take the perspective of a distant observer of mankind.51 As such, they
regard the human species as a whole.
Pragmatic anthropology has a different interest and takes a different
perspective:
Therefore, even knowledge of the races of human beings as products be
longing to the play of nature is not yet counted as pragmatic knowledge
of the world, but only as theoretical knowledge of the world.52
In addition, the expression ‘to know the world’ and ‘to have the world’
are rather far from each other in their meaning, since one only understands
the play (Spiel) that one has watched, while the other has participated (mit
gespielt) in it.53
Contrary to the philosopher of history, who rises above the ‘play’ he
studies, the anthropologist searches for knowledge that is not merely
theoretical. Anthropological knowledge is based on participation rather
48 AP, AA 7:397.
49 I focus on IUH and §§82 – 84 of CJ.
50 In IUH, AA 8:17, Kant suggests that the desire to find order in human history
arises from “observing [human] activities as enacted in the great world drama.”
It is the philosopher who looks for a “definite plan of nature.” In CJ, AA 5:24,
the interest in finding such a plan derives from speculation on why things are
there.
51 IUH, AA 8:17: “History […] allows us to hope […] that if it considers the play
of the freedom of the human will (Spiel der Freiheit des menschlichen Willens) in the
large, it can discover within it a regular course.”
52 AP, AA 7:120.
53 AP, AA 7:120.
than distant observation. The anthropologist studies human behavior

and interactions from an insider’s perspective. He understands what
it’s like to live in a human society and interact with human individuals
and he takes this as the basis for his theory. This perspective is difficult to
harmonize with that of the historian, who looks on to the development
of the human species as a whole.
In the course of his speculative attempts to understand the develop
ment of humanity, the historian judges teleologically. But in the shift in
perspective from history to anthropology, his subject matter disappears
out of sight. The historical evolution of humanity as a whole is no im
mediate topic for the pragmatic anthropologist, who bases his theory on
the concrete behavior of human individuals. This means that resorting
to teleological judgment is no longer immediately necessary.54 To co
herently understand human phenomena from the anthropologist’s ‘in
sider’s perspective’, no teleological assumptions are required. The an
thropologist does not need to accept morality as the final end of nature,
and he can drop the strong focus on humanity’s process of moralization.
As such, his perspective is broader than that of the historian.
3.3. Pragmatic anthropology and reflective classification
If describing the human process of moralization is not the pragmatic an

thropology’s priority, what is the main purpose of the work? Textual
evidence suggests that the work is primarily concerned with classifica
tion of human phenomena. In a letter to Herz from 1773, for example,
Kant remarks that his intent in the anthropology lectures is to arrive at
general knowledge of the human being by “seeking the phenomena and
their laws” rather than to bring about a modification of human nature.55
The intention that is expressed in this letter is preserved in Kant’s ma
ture conception of anthropology. In the published version of the work,
54 Nevertheless, teleological concepts like that of a ‘predisposition’ are used regu

larly in anthropology. This does not contradict my interpretation. I argue that
the teleological perspective is not predominant, but I concede that anthropol
ogy can occasionally take it up, for example when the anthropologist, from his
‘insider’s perspective’ ponders over what the future development of humanity
will be. Kant is obviously doing this in AP, AA 7:321 – 333.
55 Correspondence, AA 10:145: “The intention that I have is to […] seek the phe
nomena and their laws, rather than the ultimate conditions of the possibility of
the modification of human nature in general.”
62 Liesbet Vanhaute
he notes that one of its greatest advantages is that it provides a system of

headings for a classification of human phenomena.56 The content of the
work confirms this: Kant treats very diverse topics, including non moral
aspects of the human being like the quest for happiness,57 the extent to
which humans are determined by natural temperaments, contingent
facts as nationality or sex,58 and trivial things about why humans laugh.59
If this is pragmatic anthropology’s focus, it makes sense that Kant did
not choose to compose it out of teleological judgments. Teleological
judgment’s unavoidable assumption of morality as a final end of nature
would interfere with the anthropology’s attentiveness to the diversity of
phenomena that are to be classified. A teleological perspective requires a
focus on those (predisposed) features of human nature that promote
morality. Such a focus would be much too restrictive for a discipline at
tempting a classification of human phenomena.60 Pragmatic anthropol
ogy needs a broader perspective.
Nevertheless, though the general perspective of teleological judgment
is not called for in pragmatic anthropology, Kant’s anthropological proj
ect will most probably have to resort to reflecting judgment. The realm of
human actions is one that seems underdetermined by a priori principles.61
Hence, if pragmatic anthropology wants to systematically classify seem
ingly disparate human phenomena, it will have to search for general
principles that are not yet given. In view of this, it will need to assume
subjective purposiveness: it needs to presuppose that human behaviour and
interaction constitutes a realm of facts that can be systematically under
stood by human thinking. Only if it presupposes this will its reflective
activity of searching for general principles to classify the manifold of
human phenomena make sense.
56 AP, AA 7:121 – 122, 7:270 and 7:312.

57 For an interesting discussion of this quest, see Kain (2003, 242)
58 AP, AA 7:303 – 308, 7:311 – 321.
59 AP, AA 7:255.
60 That the focus on morality is not predominant is also clear from AP, AA 7:119:
pragmatic anthropology studies what the human being “as a free acting being
makes of himself, or can and should make of himself.” The ‘should’ thus con
stitutes only a third of the subject matter.
61 Kant calls it a manifold. Anthropology’s task is to find order in this manifold.
Menschenkunde 1781 – 1782, AA 25:856: “aus dem Mannigfaltigen, was wir
am Menschen wahrnehmen, Regeln zu ziehen.” See also IUH, AA 8:17.
3.4. The usefulness of reflective classification for morality
The systematic classification of human phenomena will never be com

plete if it overlooks phenomena indicating human morality. According
to Kant, every human being has the capacity to respect the moral law.62
And even though humans rarely (perhaps never) 63 act from mere respect
for the law, their moral capacities and sensitivities will shine through and
have impact on their actual behavior.64 In systematically classifying this
actual behavior, pragmatic anthropology will thus encounter morally
promising aspects of human nature, as well as immoral behavior.
These are inventoried, along with many other significant aspects of
human nature that do not stand in direct relation to human morality.
As such, anthropology is not possible without some reference to mor
ality. An experience based study of human beings can and must also
concieve of moral responsibility. This implies that pragmatic anthropol
ogy, as a systematic study of human beings and their social world, does
contain some moral anthropology. Pragmatic anthropology, in being
knowledge “for the world’s use”65 is thus useful for the moral philoso
pher who is curious about “the subjective conditions in human nature
that […] help [people] in fulfilling the laws of a metaphysics of morals.”66
Also, the non teleological interpretation of the work allows us to
acknowledge that the moral relevance of pragmatic anthropology is
not limited to this. Apart from providing the moral philosopher with
a moral anthropology, an empirical counterpart to his theory, pragmatic
anthropology is able to help every reader to better understand herself
62 See for example Religion within the Boundaries of mere Reason, AA 6:27.
63 Groundwork for the Metaphysics of Morals, AA 4:406.
64 This ‘shining through’ often takes interesting forms. For example, Kant de
scribes the capacity to feign virtue as somehow indicating the capacity for
true virtue. In AP, AA 7:152 – 153, he calls it ‘permissible moral illusion’, because
there is some, though very little, virtue involved. For an interesting discussion
of such ‘empirical markers of morality,’ see Frierson (2008). See also Brandt
(2003, 30), who notes: “The relationship between [anthropology and ethics]
mirrors […] the relationship between [empirical and intelligible character] in
and/or of human beings. If one compares the relation between these two char
acters with the two sides of a coin, then this ‘human coin’ has a peculiar char
acteristic: the images on the two sides resemble one another. Neither side is in
dependent of the other, rather, they stand in a fixed relation to each other—at
least a similarity relation.”
65 AP, AA 7:119.
66 Metaphysics of Morals, AA 6:217.
64 Liesbet Vanhaute
and her world. Such an understanding is morally promising in several

ways: insofar as it presents the human being as endowed with moral ca
pacities, it might cultivate confidence in one’s own moral talents,67 or in
the case of a less virtous reader, knowledge of one’s moral deficiencies.68
As an inventory of the different kinds of empirical characters and tem
peraments, it can help to understand how to adapt one’s good intentions
to the specific person one is dealing with.69 To the extent that it explains
how humans act from principle, it helps educators and politicians to
promote genuine rather than superficial morality.70 So interestingly, as
a discipline that is mainly interested in systematically describing all
kinds of anthropological phenomena, pragmatic anthropology’s useful
ness for morality is more diverse than that of a teleology describing mor
alization, or a moral anthropology confirming the possiblity of a priori
ethics.
4. Conclusion
My goal in this essay was to demonstrate that Kant’s pragmatic anthro

pology is primarily interested in the classification of human experiences,
whether they stem from inner sense or from other sources. I have shown
that the anthropologist needs to assume subjective purposiveness, not
only when he uses insights from empirical psychology, but also when
he studies human society from his characteristic ‘insider’s perspective’.
The additional assumption of objective purposiveness is not as unavoid
able. We can thus conclude that anthropology always judges reflective
ly, and occasionally teleologically. Readings that present the pragmatic
67 AP, AA 7:153 deals with the importance of believing in the possibility of

human virtue. Human characteristics that indicate man’s moral capacities are
politeness or sociability (AP, AA 2:277).
68 Characteristics that promote viciousness rather than virtue are laziness, coward
ice and duplicity. (AP 7:276) Also a tendency to imitate others might lead to
wickedness. (AP, AA 7:293).
69 Kant’s account of the temperaments in AP, AA 7:287 – 291 could be useful in
the following manner. We can imagine a sympathetic person with pragmatic
knowledge of human beings taking the anger of a phlegmatic man very serious,
because she knows the anger will not pass away quickly, as the anger of a chol
eric man will. She will also realize that she must help a choleric person in a dif
ferent manner, for example by not paying too much attention to his angriness,
or by helping him deal with his anger.
70 Menschenkunde 1781 – 1782, AA 25:858.
anthropology’s perspective as predominantly teleological, ignore the

richness and multiplicity of Kant’s anthropological project.
Bibliography
Page references to Kant’s writings refer to Kants Gesammelte Werke (1900, Kö
niglich Preussische Akademie der Wissenschaften (ed.), Berlin) (Akademie Aus
gabe, AA) with the exception of the reference to the Dohna Wundlacken lec
ture notes, which is to the electronic edition of Kant’s works on the Kant im
Kontext CD ROM (2009, Karsten Worm Verlag, Berlin). English citations
are taken from the Cambridge Edition of the Works of Immanuel Kant (see refer
ences below).
Beishart, Claus (2009): Kant’s Characterization of Natural Ends, in: Kant Year
book 1, pp.1 – 30.
Brandt, Reinhard (2003): The Guiding Idea of Kant’s Anthropology and the
Vocation of the Human Being, in: B. Jacobs and P. Kain (eds.): Essays
on Kant’s Anthropology, Cambridge, pp. 85 – 104.
Cohen, Alix (2008): Kant’s answer to the question ‘what is man?’ and its im
plications for anthropology, in: Studies in History and Philosophy of Sci
ence 39, 4, pp. 506 – 514.
Frierson, Patrick (2008): Empirical psychology, common sense, and Kant’s em
pirical markers for moral responsibility, in: Studies in History and Philos
ophy of Science 39, 4, pp. 473 – 482.
Guyer, Paul (2003): Kant’s Principles of Reflecting Judgment, in: Kant’s Cri
tique of the Power of Judgment, Lanham, pp.1 – 62.
Kain, Patrick (2003): Prudential Reason in Kant’s Anthropology, in: B. Jacobs
and P. Kain (eds.): Essays on Kant’s Anthropology, Cambridge, pp. 230 –
265.
Kant, Immanuel (2007): Anthropology from a Pragmatic Point of View, in:
Kant, Immanuel: Anthropology, History, and Education, trans. R. B.
Louden and G. Zöller, Cambridge.
Kant, Immanuel (1999): Correspondence, trans. A. Zweig, Cambridge.
Kant, Immanuel (2000): Critique of the Power of Judgment, trans. P. Guyer,
Cambridge.
Kant, Immanuel (1997): Critique of Pure Reason, trans. P. Guyer and A.
Wood, Cambridge.
Kant, Immanuel (2000): First Introduction to the Critique of Judgment, in:
Kant, Immanuel: Critique of the Power of Judgment, trans. P. Guyer,
Cambridge.
Kant, Immanuel (1996): Groundwork of The Metaphysics of Morals, in: Kant,
Immanuel: Practical Philosophy, trans. M. Gregor, Cambridge.
Kant, Immanuel (2007): Idea for a Universal History with a Cosmopolitan
Aim, in: Kant, Immanuel: Anthropology, History, and Education, trans.
R. B. Louden and G. Zöller, Cambridge.
66 Liesbet Vanhaute
Kant, Immanuel (2002): Metaphysical Foundations of Natural Science, trans.

Michael Friedman, in: Kant, Immanuel: Theoretical Philosophy after
1781, eds. H. Allison and P. Heath, Cambridge.
Kant, Immanuel (1996): Religion within the Boundaries of mere Reason, in:
Kant, Immanuel: Religion and Rational Theology, trans. A. W. Wood
and G. Di Giovanni, Cambridge.
Kant, Immanuel (1996): The Metaphysics of Morals, in: Kant, Immanuel:
Practical Philosophy, trans. M. Gregor, Cambridge.
Louden, Robert B. (2003): The second part of morals, in: B. Jacobs and P. Kain
(eds.): Essays on Kant’s anthropology, Cambridge, pp. 60 – 84.
Sturm, Thomas (2009): Kant und die Wissenschaften vom Menschen, Pader
born.
Vanhaute, Liesbet (2011): How biological is human history?, in: Logical Anal
ysis and History of Philosophy 14. [Forthcoming]
Wilson, Holly L. (2006): Kant’s Pragmatic Anthropology: its Origin, Meaning
and Significance, Albany.
Wood, Allen (2003): Kant and the problem of human nature, in: B. Jacobs and
P. Kain (eds.): Essays on Kant’s anthropology, Cambridge, pp. 38 – 59.
The Activity of Sensibility in Kant’s Anthropology.
A Developmental History of the Concept of the
Formative Faculty
Matthias Wunsch
Abstract
Kant usually characterizes sensibility as receptivity. Hence it can seem paradox
ical to speak of the “activity of sensibility” in his philosophy. Yet that sensible
representations are receptive in origin does not necessarily mean that their con
tent is due to our receptivity alone. In fact, as early as his 1770 inaugural dis
sertation Kant assumes acts of coordinating the sensible as conditions of sensible
knowledge. In the context of his anthropology he then attributes these acts to
the so called “formative faculty” which he conceives as part of sensibility. With
the concept of the formative faculty Kant unifies Baumgarten’s conception of
the lower cognitive faculty. Moreover he outlines his own theory of the activity
of sensibility by means of the formative faculty and its various facets. Further
more, a closer look at the various transcriptions of Kant’s lectures on anthropol
ogy shows that, in the late 1770s, the concept of the imagination supplants that
of the formative faculty as the foundation of his conception of an active sensi
bility, and shows also how the distinction between productive and reproductive
imagination is able to stand in for the various facets of the formative faculty.
The paper concludes with a brief look at the prospects beyond the field of an
thropology.
Introduction
It can seem paradoxical to speak of the “activity of sensibility” in Kant’s

philosophy. Kant’s standard characterizations of sensibility, beginning
with his 1770 inaugural dissertation On the form and principles of the sen-
sible and the intelligible world (De mundi sensibilis atque intelligibilis forma et
principiis; hereafter Dissertatio), leave no doubt that sensibility is passive.
“Sensibility (sensualitas)”, as he writes in the Dissertatio, “is the receptivity
(receptivitas) of a subject in virtue of which it is possible for the subject’s
own representative state to be affected in a definite way by the presence
68 Matthias Wunsch
of some object.”1 Although this initial characterization has a decisive

and systematic significance for Kant, nonetheless it only describes one as
pect of sensibility: its transcendental aspect, or the aspect of the origin of
sensible representations. Kant turns this against the Leibniz Wolff
school, which he criticizes for only having defined sensible representa
tions according to their epistemic type (i. e. logically), as representations
of something represented in an obscure or confused way. On Kant’s
conception, such logical classifications absolutely fail to capture the dif
ference between sensible and intellectual objects,2 which depends in
stead on the difference in origin between the various representations.
The aspect of the origin of sensible representations is to be distin
guished from that of their content. Their being receptive in origin
does not necessarily mean that their content is due to our receptivity
alone. In terms of their content, sensible representations are appearances,
for which Kant distinguishes between matter and form. Whereas the
matter consists in passively received sensations, the form of the appear
ance is “the aspect namely of sensible things [sensibilium species] which
arises according as the various things which affect the senses are co or
dinated by a certain natural law of the mind.”3 Thus the form of the ap
pearance derives from acts of coordinating the sensible, which are deter
mined by a law of coordination that is binding and identical for every
individual item of sensible knowledge.4
However, in the Dissertatio Kant is not concerned with examining
the ways that the sensible gets coordinated in any more detail, since
he focuses here on the principles of the form of appearances. He identifies
these principles as space and time principles that guarantee the general
connectability of the parts of the sensible world and are the subjective
conditions of coordination.5 The Dissertatio does not explain which
faculty is responsible for this work of coordination anymore than it clari
fies the ways of coordinating. Essentially two faculties or capacities come
1 MSI, § 3, AA 2:392. All quotations in English from the Dissertatio are taken
from Kant, Immanuel, Theoretical Philosophy 1755 – 1770, trans. and eds.,
David Walford and Ralf Meerbote, Cambridge 1992.
2 MSI, § 7, AA 2:394.
3 MSI, § 4, AA 2:392.
4 As this law holds for the form of all sensible knowledge, Kant subsequently
identifies it with the form of sensible representations: the form of sensible rep
resentations “est […] non nisi lex quaedam menti insita, sensa ab obiecti prae
sentia orta sibimet coordinandi” (MSI, § 4, AA 2:393).
5 MSI, § 14, 5; § 15, D.
The Activity of Sensibility in Kant’s Anthropology 69
into question for Kant at this time: sensibility and understanding. What
could be more obvious than to argue that sensibility is ruled out, since it
is characterized from the beginning as receptivity? And yet nowhere in
the Dissertatio does Kant state that coordination is the work of the un
derstanding. In my view it can be shown that the question of where to
situate coordination in the theory of faculties was still open, for Kant,
around 1770; however, I cannot deal with this point here.6
Only in the course of the 1770s, in the context of his anthropolog
ical investigations, does Kant arrive at a closer examination of the ways
in which the sensible gets coordinated and by which faculty.7 As we
know, starting in the winter semester of 1772/1773 Kant regularly
held lectures on anthropology based on the “Psychologia empirica”
chapter of A. G. Baumgarten’s Metaphysica. 8 It is a central thesis of
this paper that in his anthropology, following Baumgarten’s doctrine
of the lower cognitive faculty, Kant developed a differentiated concep
tion of specifically sensible activity under the heading of “formative
powers” and the “formative faculty.” The activity of coordination
finds a place in this conception as well: “The formative faculty concerns
the form of the entire lower cognition, namely the coordination, since
one can join representations to one another in different ways.”9 Thus
the activities of the formative faculty that Kant investigates in his an
thropology are the conceptual heirs and further specification of the co
ordination of the sensible as conceived in the Dissertatio. Moreover, it is
worth mentioning three other reasons (which have been hinted at al
ready to some extent) why Kant’s conception of the formative faculty
deserves our attention:
With the concept of the formative faculty Kant succeeds in unifying
Baumgarten’s conception of the lower cognitive faculty.
6 I plan to make up for this deficit in a future publication on Kant’s conception of

the coordination of the sensible.
7 The generally decisive texts for Kant’s anthropology are the Reflexionen zur An
thropologie (AA 15:55 – 654), the Entwrfe zu dem Colleg ber Anthropologie aus den
70er und 80er Jahren (AA 15:655 – 899), the transcripts of his Vorlesungen ber
Anthropologie (AA 25) and the Anthropologie in pragmatischer Hinsicht (AA 7).
8 The Metaphysica, Baumgarten’s main theoretical work, first appeared in 1739.
Kant used the fourth edition (1757), which was the first to contain Baumgar
ten’s German translation of several important terms. The chapter “Psychologia
empirica” comprises §§ 501 – 739 of the Metaphysica and is reprinted in AA
15:5 – 54 (§§ 504 – 699) and AA 17:130 – 40 (§§ 501 – 503 and 700 – 739).
9 Refl 331 (1776/8?), AA 15:130 f.
70 Matthias Wunsch
Kant outlines a conception of the activity of sensibility by means of

the formative faculty.
Historically, Kant’s conception of the formative faculty has to be
seen as the precursor to his later conception of productive and re
productive imagination.
In this paper, following a brief sketch of Baumgarten’s “doctrine of
lower knowledge” (gnoseologia inferior) (1) I will clarify how Kant unifies
Baumgarten’s conception of the lower cognitive faculty in his anthro
pology under the aspect of the formation of sensible representations
(2). From here I will show that the formative faculty as Kant conceived
it is the active part of sensibility, and will outline the basic features of his
theory of the formative faculty (3). It will then be possible, with the help
of the various transcriptions of Kant’s lectures on anthropology, to re
construct how, in the 1770s, the concept of the imagination supplants
that of the formative faculty as the foundation of his conception of an
active sensibility (4), and to show how the distinction between produc
tive and reproductive imagination is able to stand in for the various fac
ets of the formative faculty (5). The paper then concludes with a brief
look of the prospects beyond the field of anthropology (6).
1. Baumgarten’s doctrine of lower knowledge

(gnoseologia inferior)
Kant had several reasons for choosing the chapter “Psychologia empiri
ca” from Baumgarten’s Metaphysica as the textual basis for his anthropol
ogy lectures. Firstly, he had enormous respect for Baumgarten, despite
his increasing theoretical distance.10 Secondly, he was already quite
well versed in Baumgarten’s Metaphysica, which he regularly used as
the basis for his lectures on metaphysics. Thirdly, the “Psychologia em
pirica” chapter is very rich in material and more sophisticated than other
texts in the tradition of empirical psychology. Finally and this is espe
10 In his Neuen Anmerkungen zur Erluterung der Theorie der Winde (1756) Kant
mentions the Metaphysica as “the most useful and exhaustive of all such com
pendia” (AA 1:503). According to the Nachricht von der Einrichtung seiner Vorle
sungen from the winter semester of 1765/6 Kant appreciates Baumgarten for
“the wealth and precision of his teaching” (NEV, AA 2:308). And in the Cri
tique of Pure Reason Kant praises the “excellent analyst Baumgarten” (A 21/B 35
Anm.).
cially significant here Baumgarten’s philosophy, and particularly his

empirical psychology, lends itself to an important motif in Kant’s
thought since 1769/70: the philosophical reappraisal of sensibility.
Baumgarten brought an innovative reorientation to the tradition of
German philosophy going back to Leibniz and Wolff, taking as his start
ing point the inadequacy of the traditional logic for the problems of
philosophical poetics.11 He formulated the need to broaden philosophy
to include a science that would do proper justice to the lower cognitive
faculty a science which he would later become the founder of, and
which he entered into the philosophical vocabulary under the name
of ‘Aesthetica.’12 For him, ‘aesthetics’ did not have today’s narrow
sense of a theory of art, but rather referred to a comprehensive science
of sensible knowledge (scientia cognitionis sensitivae). Baumgarten divided it
into four aspects: “theory of liberal arts” (theoria liberalium artium), “doc
trine of lower knowledge” (gnoseologia inferior), “art of beautiful think
ing” (ars pulchre cogitandi) and “art of the analogue of the reason” (ars an-
alogi rationis).13 For our purposes here, only the second aspect is signifi
cant, the doctrine of lower knowledge, which plays a very instructive
role for Kant’s anthropological theory of sensibility. We can see it as a
science of (a) the indistinct representations and (b) the lower cognitive
faculty.
(a) Baumgarten turns to the Leibnizian hierarchy of ideas, but reas
sesses the theoretical value of obscure and confused ideas by introducing
or emphasizing more positive characterizations of them. Rather than
going into this in detail,14 we should just note Baumgarten’s termino
logical innovation relative to the Leibniz/Wolff tradition of designating
those ideas that are acquired through the lower part of the cognitive fac
ulty and are “indistinct” (non distincta) as “sensible representation” (reprae-
sentatio sensitiva),15 a term that Kant then took up in his Dissertatio.
11 Baumgarten, Meditationes philosophicae de nonnulis ad poema pertinentibus (1735),

§ 115.
12 Ibid., § 116. In the Aesthetica (1750/8) Baumgarten introduces his conception
of aesthetics.
13 Baumgarten, Aesthetica, § 1.—On the significance, content and context of
Baumgarten’s aesthetics see Poppe (1907), Baeumler (1967) (11923) and
above all Franke (1972), where the various aspects of the aesthetics are compre
hensively discussed.
14 Cf. Franke (1972), esp. 44 and 46 – 8.
15 Meditationes, § 3; Metaphysica, § 521.
72 Matthias Wunsch
(b) Baumgarten calls the faculty of obscure and confused or indis

tinct cognition the “lower cognitive faculty.”16 Thus his terminology al
ready implies that he emancipates that which Wolff called the ‘lower
part’ of the cognitive faculty from the ‘higher part,’17 since he speaks
of an independent cognitive faculty rather than a part of the cognitive
faculty. Accordingly, the discussion of the lower cognitive faculty
takes on a much larger role relative to the higher cognitive faculty in
Baumgarten’s empirical psychology than it does in the corresponding
discussion by Wolff. Moreover, Baumgarten’s investigation of the
lower cognitive faculty is not just more comprehensive, relatively, but
also more fine grained. In his Psychologia empirica, under the heading
“Of the lower part of the cognitive faculty” (De facultatis cognoscendi
parte inferiori), Wolff only distinguishes between Sensus, Imaginatio, Fac-
ultas fingendi (faculty of fiction) and Memoria (Oblivio, Reminiscentia),18
whereas Baumgarten draws a much finer set of distinctions. When he
comes to the corresponding point of his theory, he discusses a number
of faculties: Sensus, the ability to feel the state of my own soul or body;
Phantasia, the ability to have representations of the past (phantasmata,
imaginationes); Perspicacia, the ability to recognize the congruences (ingen-
ium or wit) and differences (acumen or acutenes) of things; Memoria, the
ability to recognize reproduced representations; Facultas fingendi (faculty
of fiction), the ability to isolate phantasmata to some extent and recom
bine them; Praevisio (foresight), the ability to be aware of one’s future
state and the future state of the world; Iudicium ( judgment), the profi
ciency in judging things, that is, to recognize their perfection or imper
fection; Praesagitio (expectation), the ability to represent an anticipatory
representation as the same representation one will have in the future;
and Facultas characteristica (faculty of signs), the ability to correlate signs
with the things signified.19
16 Baumgarten, Metaphysica, § 520: “Facultas obscure confuseque seu indistincte

aliquid cognoscendi cognoscitiva inferior est.”
17 Wolff, Psychologia empirica, §§ 54 f.
18 Ibid., §§ 56 – 233.
19 Baumgarten discusses these faculties in the order indicated in his Metaphysica,
§§ 534 – 623. In the Aesthetica, §§ 30 – 7, he discusses these faculties in a less
gnoseological and more genuinely aesthetic context as the faculties of the ‘in
genium venustum.’
2. Kant’s unification of Baumgarten’s ‘lower cognitive faculty’

under the aspect of formation
In the course of his anthropology, Kant takes up the activities of several

of the lower cognitive faculties discussed by Baumgarten under one and
the same aspect: the formation of sensible representations. Thus with his
conception of the formative faculty he seeks to unify Baumgarten’s
“doctrine of lower knowledge”. To make this clear, we will first
have to specify the concept of the formative faculty that is discussed
both in Kant’s Reflections on anthropology and in the various transcrip
tions of his anthropological lectures. The notes from his early lectures
distinguish between a whole series of various kinds of faculties of forma
tion:20
1. the faculty of direct image formation (Abbildungsvermçgen);
2. the faculty of reproductive image formation (Nachbildungsvermçgen);
3. the faculty of anticipatory image formation (Vorbildungsvermçgen);
4. the faculty of imaginative formation (Einbildungsvermçgen);
5. the faculty of completing formation (Ausbildungsvermçgen).
Another formative faculty can be added to these:
6. the faculty of counterimage formation (das Vermçgen der Gegenbil-
dung, the faculty to represent something by means of signs).21
Since every one of these can be described as a formative faculty, we can
take ‘formative faculty’ as the generic term for them. The formative fac
ulty in its broader sense is to be distinguished from the narrower sense of
the expression according to which the ‘formative faculty’ could just
20 See V Anth/Collins/Philippi (1772/3), AA 25:45, 76; V Anth/Parow (1772/3),

AA 25:303; V Anth/Friedlnder (1775/6), AA 25:511 f. At one point this list is
preceded by the heading “On the formative faculties” (V Anth/Philippi, AA
25:76.4).
21 This faculty is not mentioned in any listing of the individual formative faculties
in the transcriptions of Kant’s anthropology lectures. However, the expression
‘counterimage (formation)’ is found in the Reflexionen zur Anthropologie. See
Kant’s remarks on “Gegenbild, symbolum” and “Bezeichnung: signatrix
(s Gegenbild)” within the discussion of the formative faculties; Refl 314
(1769?), AA 15:124.17 and Refl 326 (1769/70?, s addendum: 1770s), AA
15:129.6. In Refl 313a (1769?), AA 15:123, Kant explicitly adds “The counter
image: symbolum” to the series of the five formative faculties initially men
tioned. The lecture transcription V MP L1/Pçlitz, AA 28:237 also mentions
the “the faculty of counterimage formation” as a “faculty of formative power.”
74 Matthias Wunsch
refer to the first of the six faculties (the faculty of direct image forma
tion).22
Rudolf A. Makkreel has already argued for the thesis that Kant uni
fies Baumgarten’s lower cognitive faculties with his conception of the
formative faculty.23 However, in my view a few more things need to
be added to his characterization. Makkreel claims that the Abbildungs-
vermçgen, the faculty of “direct image formation” from sensations,
“roughly” corresponds to Baumgarten’s faculty of exact perception,
the ‘acute sense’ (sensus acutus).24 But for Kant and Baumgarten both,
acute senses are senses that already register fine distinctions and can be
sharpened through practice.25 Thus they are only needed for an especial
ly accurate formation of images and not for the ability to represent im
ages of present objects per se. Thus Kant’s ‘Abbildungsvermçgen’ just cor
responds to Baumgarten’s ‘sensus’, insofar as it is already associated with a
formative activity. This is precisely what Baumgarten seems to assume,
as suggested by the line: “Every sensation is a sensible representation
that has to be formed by the lower cognitive faculty” (omnis sensatio est
sensitiva formanda per facultatem cognoscitivam inferiorem).26 Kant’s Nach-,
Vor- and Einbildungsvermçgen (the faculties of reproductive, anticipatory,
and imaginative image formation), as Makkreel notes, correspond to
Baumgarten’s phantasia, praevisio (foresight) and facultas fingendi (faculty
of fiction). The Ausbildungsvermçgen, the faculty of completing forma
tion, does not correspond to anything in Baumgarten. Kant probably
adopted it into his theory of formative faculties following A. F. Hoff
mann and C. A. Crusius.27 Crusius discusses a “power to finish incom
plete ideas” and emphasizes that Hoffmann had “first dealt with this in
22 See V Anth/Collins, AA 25:76.5 – 7, and V Anth/Brauer/Parow, AA 25:303.19 –

23.
23 Makkreel (1990, 11 – 14, here: 12). On this thesis see already Schmidt (1924,
7 f.).
24 Ibid., 13 f.—On Baumgarten’s ‘sensus acutus’ see Metaphysica, § 540, and Aes
thetica, § 30.
25 V Anth/Collins, AA 25:57.17 – 20.
26 Baumgarten, Metaphysica, § 544, my emphasis.—This passage also clearly shows
that Makkreel’s (1990, 2) claim that Baumgarten only emphasizes the formative
ability in connection with the “ars formandi gustum” (§ 607) is inadequate.
Rather, references to this ability are also found in the empirical psychology
of the Metaphysica in §§ 522, 544, 557, 570, 571, 632 and 637.
27 This is overlooked by both Makkreel (1997) and Schmidt (1924, 8).
particular.”28 The Vermçgen der Gegenbildung (faculty of counterimage

formation), finally, corresponds to Baumgarten’s facultas characteristica
(faculty of signs). Makkreel does not mention that Kant’s choice of
the generic term ‘formative faculty’ has also a concise textual basis in
Baumgarten, who notes that the sensitive share of a representation ba
sically “is formed by the lower cognitive faculty” (formatur per facultatem
cognoscitivam inferiorem).29
Although Kant’s conception of the formative faculty and Baumgart
en’s conception of the lower cognitive faculty are closer than Makkreel
presents them to be, it should not be forgotten that Kant’s concept of
the formative faculty is developed against the background of a different
theory of representation from Baumgarten’s doctrine of lower knowl
edge. As mentioned at the outset, starting with the Dissertatio at the lat
est, Kant no longer primarily classifies representations as obscure, con
fused or distinct; the differences in their origin are more fundamental
than this logical distinction. In establishing this difference, Kant is led
to a theory that assumes a pure sensibility and laws founded on this
pure sensibility. The various formative faculties have to be seen as a
part of this theory; according to the Collins transcription, in 1772/
1773, Kant dealt with them in his anthropology lecture under the head
ing “Of Sensiblity.”30 But now this leads us straight back to the thesis
mentioned at the beginning, namely that Kant develops the conception
of an active sensibility in his anthropology.
28 Crusius, Weg zur Gewissheit und Zuverlssigkeit der menschlichen Erkenntnis, § 101,
italicization removed by author, and § 104. This text by Crusius is a reworking
of the Vernunftlehre of his teacher Hoffmann; see Wundt (1945, 262 f., 248 –
54).
29 Baumgarten, Metaphysica, § 522, my emphasis. In § 570 Baumgarten stresses this
for the ‘phantasia’: “[O]mnis imaginatio est sensitiva […], formanda per facul
tatem cognoscitivam inferiorem”. Of Baumgarten’s other lower cognitive fac
ulties not mentioned so far, perspicacia (i. e. ingenium and acumen), iudicium and
praesagitio do not correspond to anything in Kant’s formative faculties (even
though they are often discussed in the transcriptions of the anthropology lec
tures), whereas memoria, as I will explain at the end of the 3rd section, is a
mode of the faculty of reproductive image formation.
30 V Anth/Collins, AA 25:44 f. See also the early approach taken in Refl 316
(1769?), AA 15:125.19 f.
76 Matthias Wunsch
3. The formative faculty as the active part of sensibility
Kant calls representations that do not rest on mere affectivity, as impres

sions and sensations do, ‘images.’ The primary epistemological charac
teristic of images is that we generate them ourselves through the action
of the formative faculty. This distinguishes the formative faculties in
general from the senses as capacities for receiving sensation. By the
end of the 1760s Kant already distinguished the production of sensible
representations according to whether it was “either passive or active,”
i. e. whether they were generated passively through affection or related
to an activity of the mind.31 By 1769/70 Kant was considering the idea
of allocating this activity to sensibility: sensibility was to be responsible
for the material aspect of sensible representations through its receptive
sense, but was also to include a “faculty of intuition” that would account
for the sensible form of these representations.32 A Reflection that accord
ing to Adickes probably dates from 1773/5 captures this dichotomy and
relates it to the formative faculty:
Our sensible faculties are either senses or formative powers. The latter are
produced by us ourselves, not through the impression of the senses but none
theless under the conditions in which the objects would affect or have af
fected our senses.33
In the Metaphysik L1 lecture transcription (from the second half of the
1770s) Kant distinguishes accordingly between the “faculty of the senses
themselves” and the faculty of the “imitated knowledge of the senses,”
which “can quite properly be called the formative power.”34 Thus here as
well we find Kant drawing a two fold distinction within the “sensible
cognitive faculty,” i. e. “for sensibility,”35 based on the origin of the rep
resentation. In sense the representation arises “entirely through the im
31 Refl 314 (1769?), AA 15:124.4.

32 Refl 680 (1769/70), AA 15:302: “Sensibility can be considered according to its
matter or its form. The matter of sensibility is sensation, and its faculty is sense;
the form of sensibility is appearance, and its faculty the intuition.”– Refl 650
(1769/70), AA 15:287: “Sensible representations are either sensations and re
quire sense, or appearances and are founded on the faculty of intuition.”
33 Refl 287 (1773/5?), AA 15:107 f.; my emphasis.—According to the quotation,
we produce the formative powers themselves; but what Kant means is their
products.
34 V MP L1/Pçlitz, AA 28:230.8 – 10 (the first two quotations), AA 28:235.17 – 8
(third quotation).
35 Ibid., p. 230.8 and p. 235.11.
pression of the object,” and in the case of the formative power it arises
“from the mind, but under the condition in which the mind is affected
by the objects of the senses.”36 Sensible representations arising “from the
mind” are thus representations subject to the formal principles of intu
ition and generated by the formative activities of the mind. Thus Kant
counts them among the representations that are ‘made’ rather than
‘given.’37 He sees them as representations of the formative power that
“originate from the spontaneity of the mind,” and at the same time
he believes that the formative power “belongs to sensibility.”38
The conclusion from all this is obvious: in conceiving the formative
faculty as a part of sensibility, Kant ascribes a dimension of spontaneity
and activity to sensibility. Thus he goes beyond the simple equation
‘sensibility = receptivitas’ not just in supposing that sensibility is the
source of its own laws which all sensible content is subject to, but
also in seeing this content as fashioned by an active faculty of knowl
edge, the formative faculty. In what follows I will sketch the basic fea
tures of Kant’s anthropological theory of an active and spontaneous sen
sibility by describing the activities of the various formative faculties and
examining how they relate to each other and depend on one another.
The first three formative faculties mentioned above distinguish
themselves by the fact that their representations come with a temporal
index and a reference to reality: the faculty of direct image formation
represents a present object as actually present, while the faculties of re
productive and anticipatory image formation represent an absent object
as actual in past or future. The anthropology prepares the way for Kant’s
renowned thesis from the Critique of Pure Reason that “imagination is a
necessary ingredient of perception itself” by ascribing the active mo
ment of formation to the Abbildungsvermçgen, the faculty of direct
image formation.39 This faculty “is active,” since it “goes through the
acts of making images from impressions;” its product, the “direct
image,” is a sensible representation put together into an intuitive
whole from impressions.40
36 Ibid., p. 230.10 – 14; cf. p. 235.19 – 21.

37 Ibid., p. 230.26.
38 Ibid., p. 230.21 f.; p. 231.5.; cf. p. 235.21 f. On Kant’s conception of an active
sensibility according to V MP L1/Pçlitz; see also Caygill (2003, 180 f.).
39 CPR, A 120 note.
40 V Anth/Parow, AA 25:303.33 (first quotation); Refl 327 (1772?), AA
15:129.25 f. (second quotation); Refl 314 (1769?), AA 15:124.16 (third quota
tion).—See also V Anth/Collins/Philippi, AA 25:45.3 – 9, as well as the follow
78 Matthias Wunsch
In regards to Baumgarten, Kant identifies the faculty of reproductive

image formation with both “Phantasie” and “Imagination”41 and the fac-
ulty of anticipatory image formation with “Praevision.” Not just reproduc
tive images but also anticipatory images depend on the associations of
temporally linked types of representations.42 If an associative link exists
between two representations that occurred successively in the past, then
the occurrence of a representation of the first type can allow one to an
ticipate a representation of the second type. The anticipated representa
tion is a preliminary formation generated on the basis of associative con
nections. The faculty of anticipatory image formation clearly depends
on the faculty of reproductive image formation, since the ‘terminus an
tecedentus’ of the associative connection first has to be reproduced from
the current representation before its associate correlate can be formed in
anticipation.43
What makes Kant’s conception of the relation between the faculty
of direct image formation and the faculties of reproductive and anticipa
tory image formation particularly interesting is that it seems to anticipate
several of the phenomenological insights of the 20th century. For one
thing, Kant sees the faculty of direct image formation as the “founda
ing passages: “The formative faculty is the compilation of impressions, from

which a whole then emerges.” (V Anth/Philippi, AA 25:76 note 3)—“For
every sensibility there is at the same time an act of direct image formation, in
which we gather the images of the impressions that occur to our senses and rep
resent them to ourselves at once” (V Anth/Parow, AA 25:269.3 – 6).—“When
the eyes are opened many sensible impressions occur, my mind puts them to
gether and makes out of them a whole, and this is the formative faculty.” (V
Anth/Parow, AA 25:303.26)
41 V Anth/Collins, AA 25:45.18. On the identification of the faculty of reproduc
tive image formation with the first see also V Anth/Parow, AA 25:304.2, and
with the second see V Anth/Collins, AA 25:78.8 – 9 and 95.11.
42 V Anth/Friedlnder, AA 25:512.23 – 5: “The faculty of reproductive image for
mation has its law, it follows the law of association of the representations.”—
Refl 315 (1769?), AA 15:125.6 – 7: “The subjective ground of the reproductive
image formations is also the ground of the anticipatory image formation.” See
also V Anth/Parow, AA 25:304.10: “Anticipatory image formation happens just
like reproductive image formation.”
43 See Refl 316 (1769?), AA 15:125.22 – 7. “The Phantasie [reproductive image
formation] consists in this: the present reproduces the past, and one of these
representations another; the former is the cause of the reproduction, the latter
[is the cause of] the continuation, both according to the law of association. The
Praevision [anticipatory image formation] is only an effect of this, when we set
the terminum antecedentem of Phantasie in the present moment.”
tion” of the other modes of formation,44 since reproductive and antici

patory images are produced on the occasions of current direct images
and in light of past direct images. For another, the faculty of direct
image formation also presupposes the work of reproduction and antici
pation, as can be seen in several of the examples from Kant’s anthropol
ogy. For instance, when we are reciting something, it is a precondition
for the coherence of the recital that we ‘protend,’ i. e. form in anticipa
tion, that which immediately follows.45 Kant also mentions that an im
provising pianist has to “in part preview the future notes and in part re
view the notes produced if no dissonance is to occur.”46 We can con
clude from this that a ‘direct image’ of complete temporal wholes re
quires reproductive and anticipatory activities. While we are ‘directly’
forming the melody played by the pianist, we also reproduce sequences
played earlier and anticipate this or that continuation or rule it out.47 Yet
the direct formation of spatial wholes and not just temporal wholes also
involves reproduction and anticipation: “In intuiting what is present we
continuously look to the past and the future. In this way we bring it into
a connection and become aware of it.”48 The intuition of that which is
present does not consist in the impressions filling some possible Now; it
is equally constituted by the activity of reproduction and anticipation.49
Even if Kant’s anthropology is not designed to aim at any fundamental
epistemological analysis, it nonetheless implies the general idea that it is
a condition of all our acts of speaking and thinking as well as action and
perception that we incorporate that which is past, anticipated, or possi
ble.
The three formative faculties discussed so far are characterized by the
temporal reference and context of their representations and by repre
senting their objects as actual. In contrast, the “Einbildung is an image
44 Refl 315 (1769?), AA 15:125.3.

45 V Anth/Collins, AA 25:119.4 f.; see also V Anth/Parow, AA 25:335.5 f.
46 V Anth/Parow, AA 25:249.29 f. In Refl 390 (1776/8), AA 15:156, Kant formu
lates this thought as follows: “That in speaking we always look a distance back
and a distance ahead, without which there would be no connection.”
47 We do not explicitly rule out certain conceivable continuations in listening; but
that we have ruled them out de facto can be shown by the fact that we are sur
prised when they in fact occur.
48 V Anth/Philippi, AA 25:87.26 – 8.
49 Cf. V Anth/Philippi, AA 25:119 note 2: “If we understand the present to be a
whole stretch of time, then we do not see something like a point before us that
expresses the present, but rather a whole district in time.”
80 Matthias Wunsch
of fabrication, facultas fingendi, it is an image of an object that is neither

present nor past or future, but rather is a fiction.”50 The work of the fac-
ulty of imaginative formation enables us to “create a whole new object,”
i. e. “produce self created objects.”51 However, the possibilities of imag
inative formation are constrained by what is sensibly given, since we
cannot create sensations,52 but only combine them.53
The faculty of completing formation relates to this point. When a given
representation is incomplete or imperfect, we add the missing element
in our representation by interpolating, supplementing or improving
it actions which characterize the faculty of completing formation,
the “faculty of perfection” (facultas perficiendi).54
The faculty of formation has a tendency to complete everything in our
mind. Thus when we become aware of something, we make ourselves a
concept of it. If the object does not accord with the concept, the mind in
cessantly strives to complete it.55
Thus the activity of completing relies on a concept that functions as a
standard. In the case of a half hidden spatial object, for example, this
standard consists in an empirical conception of the thing. But the con
cepts of numbers or aesthetic ideals can also serve as standards, such as
when we ‘round up to a dozen’ or improve “an imperfect piece in
the comedy” in our thoughts.56
The faculty of counterimage formation is the faculty to represent things
by means of signs. The signs (counterimages) in question can be divided
50 V Anth/Friedlnder, AA 25:511.21 – 4; my emphasis.

51 V Anth/Parow, AA 25:269.13 (first quotation); V Anth/Philippi, AA
25:87.24 – 5 (second quotation).
52 See V Anth/Parow, AA 25:321.33, u. V Anth/Collins, AA 25:95.23 – 4.
53 V Anth/Collins, AA 25:46.1 – 2: The “Einbildung only takes the materials from
the senses, but it creates the form itself.” Cf. V Anth/Philippi, AA 25:76 f.: “We
can never entirely fabricate something, rather we have always so to speak cop
ied the materials and thus we can only change the form.”—An additional limit
to our freedom in forming new representations is that we cannot imagine any
thing logically impossible (V Anth/Parow, AA 25:326.6 – 7; V Anth/Collins, AA
25:100.1 – 2).
54 V Anth/Friedlnder, AA 25:512.22.
55 V Anth/Friedlnder, AA 25:512.1 – 5. The “tendency […] to perfect (perficien
di)” is already mentioned in the Refl 313a (1769?), AA 15:123.
56 See the examples in V Anth/Friedlnder, AA 25:512.6 – 17.
into linguistic signs (words) and non linguistic signs (symbols).57 We

cannot go into the understanding of symbols here, since for Kant it pre
supposes an analogical use of reason58 and thus can’t be ascribed to the
active part of sensibility. In the case of words, however, Kant sees the
formation of counterimages primarily as an associative reproductive
image formation. “For each word one reproduces (nachbilden) the repre
sentation that one tends to associates with the word.”59
Kant’s theory of the formative faculty is made even more complex
by his distinction between two different ways of using each faculty: an
involuntary and a voluntary use. Whereas the former is described as un
conscious, unrestrained and without rules, the latter is a conscious use in
accordance with the understanding. For Kant, all humans face the chal
lenge of bringing all of their faculties under their own free will and thus
attaining self control. This demand of pragmatic anthropology is not pri
marily ethical in nature, but rather a law of prudence. Prudential behav
ior towards oneself and in the world entails acting with self determina
tion. For this purpose, according to Kant, all of our faculties have to be
disciplined, cultivated and instructed by the understanding. Thus Kant
continually warns us against the involuntary use of the formative faculty,
whereas he attributes great significance to its voluntary use, since it se
cures the proper goal directed functioning of the formative faculties
(epistemically, for example).
That Kant allows for a voluntary and involuntary use of the individ
ual formative faculties leads to a sort of doubling of the entire theory of
formative faculties. I would like to briefly illustrate this using the exam
ples of the faculties of reproductive image formation and imaginative
formation. a) The involuntary formation of reproductive images causes
some images to reappear all of a sudden, and is to be distinguished from
memory, the activity of remembering, i. e. the faculty of reproducing
images from the past at will. ‘Memory’ refers to the conscious reference
to the past. b) Kant also distinguishes between voluntary and involuntary
imaginative formation. He discusses involuntary imaginative formation
in conjunction with dreams, hypochondriac behavior, and various men
57 V MP L1/Pçlitz, AA 28:237.16; Refl 313a (1769?), AA 15:123.21; cf. Refl 314

(1769?), AA 15:124.17. Sometimes Kant calls symbols ‘sense images’; see for
example V Anth/Collins, AA 25:126.23; V Anth/Friedlnder, AA 25:536.2 – 3.
58 V Anth/Friedlnder, AA 25:552.2 – 6; V Anth/Pillau, AA 25:778.33 – 35; cf.
Anth, AA 7:191.
59 V Anth/Parow, AA 25:305.24 f.
82 Matthias Wunsch
tal disorders. Voluntary imaginative formation, in contrast, is typical of

all acts of discovery, invention and creativity, such as in art or in the for
mulation of hypotheses.
4. Transformation: From ‘formative faculty’ to

‘Einbildungskraft (imagination)’
For the purposes of getting a clear outline of Kant’s theory of the for
mative faculties, it would be better to view the theory as a stable struc
ture and to ignore questions of the shifts, terminological and otherwise,
that the theory undergoes. Nonetheless, in what follows I would like to
turn to the historical development of the theory in order to reconstruct
the prehistory of Kant’s later theory of the Einbildungskraft (at least in
sofar as it concerns the anthropology). The transcripts from Kant’s an
thropology lectures of the 1770s are particularly indispensable to
this.60 Together with Kant’s Reflections on anthropology they give us a
clear history of the various stages in the development of the theory of
the formative faculties and allow us to date when Kant started to take
the concept of the Einbildungskraft as the new foundation of his concep
tion of active sensibility.
Kant had already used the concept ‘formative faculty/power’ before
his first anthropology lecture in the winter semester of 1772/3.61 His
early conception of the formative faculty first emerges in 1769/1770,
thus at the same time that he was developing his new theory of sensibil
ity. The relevant passages are the Reflections 313a 326, where Kant de
scribes the modes of formation that he later sets out synoptically in his
60 H. Mörchen, who has produced the most detailed and informative investigation
of “Kant’s anthropological discussions of the imagination” thus far (Mörchen
(1930, 319 – 52)), was not yet able to make use of these transcriptions, and
thus treats Kant’s statements on the formative faculty and the imagination,
which he characterizes (ibid., 334 – 50) based on the V MP L1/Pçlitz, as a static
structure. The same holds for V. Satura (1971, 113 ff.).—Yet no sufficient clari
ty has been achieved on the anthropological prehistory of Kant’s theory of Ein
bildungskraft since the publication of Kant’s anthropological lectures in the Aka
demie Ausgabe (1997).
61 If Adickes’ dating is correct, then the Refl 352 (1762/3?), AA 15: 138 is to my
knowledge the first mention of the formative power. However, the question
Kant poses in that passage (“Whether the stream of Phantasie, as well as the di
rection of its formative power, stems from the mind?”) is not sufficient to re
construct his earlier concept of the formative power.
first anthropology lecture.62 Thus according to the published material

Kant’s theory of the formative faculties emerges in the time from
1769 to 1773. As it concerns the conception of sensibility, the central
thought which, judging by Adickes’ dating, Kant probably first set
down in 1773/1775 is that sensibility includes not just the receptive ca
pacities, the senses, but also the active faculties of the formation of sen
sible representations.63
The theory of the formative faculty remains essentially constant
from 1769 to 1776; and within this context, ‘Einbildungskraft’ refers
solely to the faculty of producing new representations. This changes
in the Pillau transcription of the anthropology lecture from the winter
semester of 1777/8, where Einbildungskraft takes on the functions that
had previously been reserved for other formative faculties. Among other things,
Kant now says of the Einbildungskraft that it is “the representation of
things past.”64 Since ‘Einbildungen (imaginings)’ thus take on a possible
reference to the past, one could expect that this would simultaneously
push into the background the terms previously used for this, ‘Nachbil
dung (reproductive image formation)’ and ‘Phantasie.’
And this is in fact the case for terms like ‘Nachbildung(svermögen)’
and ‘nachbilden (reproductive image forming).’ They appear in the
Friedlnder transcription (1775/6),65 but not in the Pillau transcription,66
which instead talks about the ‘Reproduktionsvermögen (faculty of re
production)’67 or, as just quoted, the ‘Einbildungskraft.’ The term
‘Phantasie,’ in contrast, continues to be used throughout all transcrip
tions, though it takes on a narrower meaning over the course of time.
While in the transcriptions of the first lecture in 1772/3 it is used (a)
synonymously with ‘reproductive image formation’68 and (b) to refer
62 Refl 313a 326, AA 15:123 – 9. V Anth/Collins/Philippi, AA 25:45 and 76 f.; V

Anth/Parow, AA 25:303.
63 Refl 287 (1773/5?), AA 15:107 f.; see as well Kant’s earlier Reflections quoted in
note 32.
64 V Anth/Pillau, AA 25:751.10 f.
65 V Anth/Friedlnder, AA 25:511.19 – 21 as well as p. 512.22 – 4.
66 A look at the guiding texts of the edition of Kant’s anthropology lectures shows
that such expressions occur most frequently in Collins and Parow (around ten
times each), less often in Friedlnder (four times), not at all in Pillau, only occa
sionally (twice) in the longest guiding text (Menschenkunde) and then later not at
all ( just as in Anth).
67 V Anth/Pillau, AA 25:752.6 f.
68 V Anth/Collins, AA 25:45.15 – 18, 78.8 f. and 95.11; V Anth/Parow, AA
25:305.22 f.
84 Matthias Wunsch
to the uncontrolled production of images,69 for example as is character

istic of the so called ‘fantasists,’70 the first meaning (a) ‘Phantasie’ as
the reproduction of the past has disappeared by 1775/6. The latter
meaning is now dominant and thus with it the ‘fantasticalness’ and
the perverse, unruly, arbitrary, rhapsodical, dissipated, depraced and
pathological fantasy.71 Thus Kant increasingly distances himself from
Baumgarten’s concept of ‘phantasia,’ which referred primarily to the fac
ulty of becoming conscious of the past states of the world and only mar-
ginally included the “vana phantasmata” and unrestrained fantasy (phan-
tasia effrenis).72 Kant restricts the meaning of ‘Phantasie’ to this marginal
area and from this point on gives the expression a primarily negative
connotation. Thus ‘Phantasie’ falls under the purview of the involuntary
formative faculty and is essentially identified with it.
In the Pillau transcription (1777/8) this restriction of the meaning of
‘Phantasie’ to a merely involuntary activity which means: the episte
mological devaluation of the concept goes hand in hand with an ex
tension and higher valution of the concept of ‘Einbildungskraft’, of
which the following passage is typical:
In a more narrow sense the Einbildungs Kraft is the representation of things
past. But typically people see it as the whole field of the formative faculty, in
dependently of the presence of the objects. […] Einbildung is sometimes
distinguished from Phantasie. The Einbildung is the formative faculty insofar
as it is in some measure subject to the will [Willkr]: Phantasie is without
will [Willkr], even contrary to the will [Wille]. A fantasist is not someone
who has no many images, but someone for whom the images impose
themselves contrary to his will. Einbildung also does not act in accordance
with our will, but we can direct it towards objects.73
Two points concerning the relation of the Einbildungskraft to Phantasie
and the formative faculty are worth emphasizing.
Firstly, Kant distinguishes Einbildungskraft from Phantasie. His crite
rion is that the latter is in essence involuntarily active, whereas we can at
least point the former in a certain direction. Since, in contrast to Phan-
69 For example: V Anth/Collins, AA 25:87.21; V Anth/Parow, AA 25:310.16 –

20.
70 V Anth/Collins, AA 25:105 – 7; V Anth/Parow, AA 25:326.25 – 7 and p. 330 f.
71 V Anth/Friedlnder, AA 25:514 f., p. 528 – 31; V Anth/Pillau, AA 25:753.8 –
30, p. 758.21 f., p. 763 – 5; V Menschenkunde, AA 25:945.22 – 946.12, p.
948.19 – 949.20, p. 955.9 – 959.14, p. 1006, p. 1012.14 – 8, p. 1033.19 – 22.
72 Metaphysica, § 557 and § 571.
73 V Anth/Pillau, AA 25:751.10 – 22, my emphasis.
tasie, we are able to ‘steer’ Einbildungskraft, Kant plays this off against the
negatively connotated Phantasie and gives it a higher relative epistemo
logical value. As Kant sees involuntary activity in the Reflection 338, we
are in a sense not the subject of the act of Phantasie, and thus Phantasie is
passive in this regard: “Phantasie is passive. [/] It plays with us;” whereas
in contrast: “Einbildungskraft is active. [/] We play with it.”74
Secondly, and this is the decisive point, the voluntary use of the Einbil-
dungskraft is identified with almost the entirety of the formative faculty. As will
shortly become clear, there is only one exception. The Einbildungskraft,
Kant writes, is “the whole field of the formative faculty, independently
of the presence of the objects.” This should be taken to mean that all the
representations that represent things “that are not present” can be called
‘Einbildungen (imaginings).’75 However, we can relate to such things
not just through new, self created representations, but also through rep
resentations of the past or the future and by means of signs. Completions
of images (Ausbildungen) are also to be understood as representations of
things that are not present; since it is precisely because an object is not
entirely present (in the sense that it is partially hidden, for example) that
we fill in the complete representation of it. This means that the Einbil-
dungskraft so conceived covers all formative activities with the exception
of direct image formation: the formation of reproductive, anticipatory,
imaginative, completing, and counter images. Thus Kant quite consis
tently modifies the distinction of sensibility described above, between
senses and formative powers, and now writes: “Sensibility. Sense and
Einbildungskraft.”76 Thus the Einbildungskraft has taken the place of the
formative faculty.77 This terminological shift shows that it has now
taken over the function of almost the entire epistemically relevant activity
of sensibility.
74 Refl 338 (1776/9?), AA 15:133.20 f. Around this time Kant also emphasizes the
difference between Phantasie and Einbildungskraft in Refl 337 (1776/9?) and Refl
369 (1776/9?), AA 15:133 and 144.
75 V Anth/Pillau, AA 25:750.27.
76 Refl 223 (1776/8?), AA 15:85. Cf. Refl 225 (1783/4), AA 15:86.4ff; Refl 342
(1780/9), AA 15:134; Refl 1482 (s addendum 1775/89), AA 15:675.18 f.; Refl
1503 (s addendum 1780/1804), AA 15:801 and Anth, AA 7:153.
77 Both are identified accordingly in Refl 339 (1780/9), AA 15:134.5 f.
86 Matthias Wunsch
5. The distinction between ‘productive’ and

‘reproductive’ Einbildungskraft
Since as of 1777/8 ‘Einbildungskraft’ serves as the general term in

Kant’s anthropology for all activities of the formative faculty (aside
from direct image formation), Kant now requires a terminological appa
ratus to register the distinctions within the concept of the formative fac
ulty within his new and comprehensive conception of the Einbildung-
skraft. He makes use of the following distinction: the “faculty of the
Einbildungskraft is twofold, productive and reproductive.”78 This distinc
tion, which is the foundation of Kant’s conception of the Einbildung-
skraft, first emerges in the surviving transcripts from his anthropology
lectures in 1781/2, i. e. in the Petersburg and Menschenkunde transcripts.79
The reproductive Einbildungskraft or the “faculty of reproduction is the
faculty of producing again the images of things that were formerly pres
ent.”80 This matches the substance of Kant’s earlier characterization of
the faculty of reproductive image formation. Accordingly the passage
goes on immediately to say: “This faculty underlies […] all memory,
where our Einbildung only reproduces.”81 It is characteristic of memory
that it recalls representations of the past “with consciousness” and “vol
untarily.”82 The substance of this characterization is also familiar material
from the earlier transcriptions of the anthropology lectures. Termino
78 V Menschenkunde, AA 25:945.1 f.
79 The Pillau transcription from 1777/8 does not yet explicitly make this distinc
tion, even if we find there the contrast between a ‘faculty of reproduction’ and
a ‘faculty of creation’ (V Anth/Pillau, AA 25:752.5 – 9). The passages on pro
ductive and reproductive Einbildungskraft in V Menschenkunde are: AA
25:945, 974, 981 and 1062.—The Petersburg transcription does not belong to
the guiding texts of the edition of the anthropology transcriptions in AA 25.
However, like all other transcriptions, it is available on the internet in the elec
tronic documentation of Kant’s lectures on anthropology (http://www.uni
marburg.de/kant/webseitn/gt ind30.htm). The passages on reproductive and
productive Einbildungskraft in the Petersburg transcription are found on pages
62, 83, 90 and 172.—It is worth noting several other anthropological remarks
by Kant that do not rule out putting the date of the distinction before 1781.
According to Adickes’ dating, their terminus post quem is 1780 (Refl 340 and
341, AA 15:134) or even 1776 (Refl 1485, AA 15:699.12); yet the late terminus
ante quem, 1789, does not force the earlier dating.
80 V Menschenkunde, AA 25 945.2 – 4.
81 Ibid., p. 945.4 – 6.
82 Ibid., AA 25:974.8 – 14.
logically, however, memory is now associated with the generic term

‘Einbildungskraft’ or more precisely with the voluntary mode of that
Einbildungskraft which is called ‘reproductive.’83 Its counterpart in Kant’s
anthropological theory of Einbildungskraft is the involuntary activity of
reproductive Einbildungskraft, which is called “Imagination” and “Phan
tasie” in the transcript of the 1781/2 lecture.84
The ‘faculty of imaginative formation,’ i. e. the ‘facultas fingendi’ of
the earlier transcripts of the anthropological lectures is now absorbed
into the productive aspect of the more broadly conceived Einbildungskraft.
The faculty of production is creative and produces things that were not thus
in our senses previously.—This creative faculty is also called the productive
Einbildungskraft or the poetic faculty.85
Unsurprisingly, Kant also distinguishes two modes of exercising the pro
ductive Einbildungskraft as well. “This faculty of production is divided
into the voluntary and involuntary imagination. […] The involuntary
is called Phantasie.”86 Poetic creation, in contrast, is “actually the inten
tional creation of new representations, […] thus the act whereby I de
liberately use given materials to make for myself new ones.”87 Thus
whereas of ‘poetizing’ and ‘memory’ refers to the voluntary use of
the productive and reproductive Einbildungskraft, ‘Phantasie’ refers to
the entirety of the involuntary use of the Einbildungskraft.
This distinction between reproductive and productive Einbildung-
skraft gives Kant the conceptual apparatus he needs in his anthropology
in order to trace the individual formative faculties (with the exception of
direct image formation) back to the Einbildungskraft. The connections
that we found in our discussion of the formative faculties can help to
83 In the earlier Pillau transcript, memory, “as the power of will over the products
of the Einbildungskraft,” is first brought together with the Einbildungskraft, (V
Anth/Pillau, AA 25:756.17 f.), although Kant does not yet explicitly distinguish
between reproductive and productive Einbildungskraft. This offers further sup
port for the view defended in the previous section that the Pillau transcription
already features a decisive expansion of the meaning of the term ‘Einbildungs
kraft’.
84 V Menschenkunde, AA 25:974.14 – 22.
85 Ibid., AA 25:945.6 f.—Ibid., 981.4 – 6.
86 Ibid., AA 25:945.22 – 6. Elsewhere Kant also names the involuntary productive
Einbildungskraft simply “Imagination,” thus calling it by the name used in the
above quotation for the general term for both types of productive Einbildungs
kraft. Ibid., AA 25:981.9 – 12.
87 Ibid., AA 25:981.7 – 9.
88 Matthias Wunsch
clarify how this works. Anticipatory image formation is based on repro

ductive image formation, which in the new terminology has to be clas
sified as an activity of the reproductive Einbildungskraft. Completing for
mation is founded essentially on adding imaginatively to the representa
tion, i. e. imaginative formation, which within the new terminology has
to be considered an activity of the productive Einbildungskraft. Counter
image formation, if the signs in question are words, is an associative re
productive image formation in Kant’s conception and thus is based on
reproductive Einbildungskraft. However, nowhere does Kant explicitly
set out the reduction sketched here; yet the fact that the terms ‘forma
tive faculty/power’ cannot be found in the Critique of Pure Reason and
any of his later publications, and that the activities previously discussed
under this heading are now attributed to the Einbildungskraft, shows that
Kant had made this reduction de facto.
Conclusion
From the end of the 1760s to the second half of the 1770s Kant had de
veloped a conception of the formative faculty that is active and sponta
neous and yet at the same time belongs to sensibility. Kant’s develop
ment of this conception began with the Dissertatio concept of the coor
dination of the sensible, and in the context of his anthropology it was
motivated by his goal of unifying the diverse activities that Baumgarten
attributed to the ‘lower cognitive faculty.’ It is precisely for the purposes
of achieving a unified theory that the concept of Einbildungskraft then
proves to be more stringent, since it avoids the redundancies that had
marked the concept of formative faculty and yet due to the internal
‘productive/reproductive’ distinction it can almost completely absorb
the diverse jobs the formative faculty had been responsible for.
Only direct image formation cannot be reduced to the Einbildungs
kraft. As the examples above have shown, it presupposes reproductive
and anticipatory work and thus presupposes the work of the reproductive
Einbildungskraft. Yet in the Critique of Pure Reason it is part of Kant’s core
conception of direct image formation, i. e. the perception of empirical
objects, that it also involves the activity of the productive Einbildungskraft.
In his anthropology Kant lacks both the interest and the conceptual
means to get a better view of this activity. For this reason the question
of the epistemological consequences of the anthropological conception
of active sensibility cannot be answered here. What would it take to
bring these consequences to light? The anthropological concept of pro-

ductive Einbildungskraft would have to be grounded in a transcendental
concept.88 On Kant’s view we can only understand this transcendental
concept of productive Einbildungskraft in working through the problem
of the transcendental deduction of the categories. Here we would also
have to determine precisely how autonomous the imaginative activity of
sensibility is, that is, to what extent it should be seen as independent of
the understanding. Thus in terms of historical development we would
have to examine Kant’s outlines of such a deduction preceding the Cri-
tique of Pure Reason with a view to the Einbildungskraft. 89 In this light it is
quite significant that, so far as all present source materials indicate, Kant
first explicitly distinguishes between reproductive and productive Einbil-
dungskraft in an outline for such a deduction stemming from the spring
of 1780 the ‘loose page’ B 12 in which Kant also introduces the
transcendental Einbildungskraft. 90 This text would have to play a central
role if we wished to expand this examination of the historical develop
ment of the activity of sensibility in Kant’s anthropology in connection
with Kant’s epistemology.
Bibliography
Baeumler, Alfred (1967): Das Irrationalitätsproblem in der Ästhetik und Logik
des 18. Jahrhunderts bis zur Kritik der Urteilskraft. Mit einem Nachwort
zum Neudruck [reprint of the 1st edition, Halle 1923], Darmstadt.
Baumgarten, Alexander Gottlieb (1954): Reflections on Poetry: A. G. Baum
garten’s Meditationes philosophicae de nonnullis ad poema pertinentibus,
Halle 1735, trans. K. Aschenbrenner and W. B. Holther, Berkeley.
Baumgarten, Alexander Gottlieb (41757) (11739): Metaphysica, Halle, reprinted
in: AA 15:5 – 54 and AA 17:5 – 226.
Carl, Wolfgang (1989): Der schweigende Kant. Die Entwürfe zu einer Deduk
tion der Kategorien vor 1781, Göttingen.
Caygill, Howard (2003): Kant’s Apology for Sensibility, in: B. Jacobs (ed.): Es
says on Kant’s Anthropology, Cambridge, pp. 164 – 193.
88 Then the ‘productive’ activities of Einbildungskraft would include not just those
by means of which we can creatively see a dog in a cloud, for example, but
those that we need to see anything as a dog at all. Cf. Strawson (1974).
89 These outlines are found in Kant’s Reflexionen zur Metaphysik (AA 17); see also
the text mentioned in the following note.
90 The ‘loose page’ B 12 is found in AA 23:18 – 20. For a detailed analysis of the
text see Carl (1989).
90 Matthias Wunsch
Crusius, Christian August (1965): Weg zur Gewissheit und Zuverlässigkeit der
menschlichen Erkenntnis, Leipzig 1747, reprint: Hildesheim.
Franke, Ursula (1972): Kunst als Erkenntnis. Die Rolle der Sinnlichkeit in der
Ästhetik des Alexander Gottlieb Baumgarten, Wiesbaden.
Makkreel, Rudolf A. (1990): Imagination and Interpretation. The Hermeneut
ical Import of the Critique of Judgment, Chicago.
Mörchen, Herrmann (1930): Die Einbildungskraft bei Kant, in: Jahrbuch für
Philosophie und phänomenologische Forschung 11, pp. 311 – 495.
Poppe, Bernhard (1907): Alexander Gottlieb Baumgarten. Seine Bedeutung
und Stellung in der Leibniz Wolffischen Philosophie und seine Beziehun
gen zu Kant. Nebst Veröffentlichung einer bisher unbekannten Hands
chrift der Ästhetik Baumgartens, Leipzig.
Satura, Vladimir (1971): Kants Erkenntnispsychologie in den Nachschriften
seiner Vorlesungen über empirische Psychologie, Bonn.
Schmidt, Raymund (1924): Kants Lehre von der Einbildungskraft. Mit beson
derer Rücksicht auf die Kritik der Urteilskraft, in: Annalen der Philosophie
und philosophischen Kritik, H. Vaihinger and R. Schmidt (eds.), 4, 1/2,
Kant Festschrift, pp. 1 – 41.
Strawson, Peter F. (1974): Imagination and Perception, in: P. F. Strawson:
Freedom and Resentment and other essays, London, pp. 45 – 65.
Wolff, Christian (1968): Psychologia empirica, Frankfurt, Leipzig 21738
(11732), in: Gesammelte Werke, vol. II.5, reprint: Hildesheim.
Wundt, Max (1945): Die Schulphilosophie im Zeitalter der Aufklärung, Tü
bingen.
From Gratification to Justice. The Tension between
Anthropology and Pure Practical Reason in Kant’s
Conception(s) of the Highest Good
Thomas Wyrwich
Abstract
The aim of this paper is to demonstrate that there is a tension between anthro
pological and solely ‘pure’ rational elements in Kant’s conceptions of the ‘high
est good.’ Whereas ‘happiness’ serves in the second Critique still as a humanly
conceptualized form of gratification commensurate to virtue, Kant is searching
for a purely moral form of ‘happiness’—as the objectification of virtue—in his
later works. Distancing himself from certain aspects of the postulates, Kant lo
cates this purely moral form in the concept of a (basically punitive) justice. Ac
cording to the main idea of this paper, moral justice is not exclusively or not
even in the first instance an ‘all too human’ concept but a demand of pure, di
vine practical reason itself.
Introduction1
It is a well known fact that Kant considers it necessary to expose a ten

sion between specific anthropological conditions and ‘self sufficient’
pure practical reason in order to establish a categorical metaphysics of
morals. In the Groundwork, Kant points out that a true moral law
“must hold not only for human beings but for all rational beings as
such, not merely under contingent conditions and with exceptions but
with absolute necessity” (AA 4:408; see also 4:411 f.). If there are also un
conditional laws in the realm of practical philosophy, they cannot essen
1 Translations have been taken, as far as possible, from the Cambridge Edition of the
Works of Immanuel Kant. In some cases I have given my own translation or I
have modified a translation. References to Kant’s works are given in the
text, by volume and page number of the Akademie Ausgabe.—I would like
to thank Silvia Jonas (Berlin) and an anonymous referee for the Kant Yearbook
for helpful comments.
92 Thomas Wyrwich
tially depend on empirical conditions, not even if they are “necessary

human” conditions.2
Although this claim is widely accepted as a core of what one might
call the “analytic part” of pure moral philosophy (including at least
Groundwork I and II and the Analytic of the second Critique), it is notice
able that this systematic difference does not attract the same interest in
the analysis of the Dialectic of the second Critique at least not in recent
time. In this paper, I will try to show that the consideration of this dif
ference is important in order to reconstruct Kant’s theory of the “high
est good” in the second Critique and classify it adequately, namely, as a
primarily anthropological conception (1). After that, I will argue for the
claim that in later works, Kant transforms his theory of the highest good
into a “purely” moral form by focusing on the concept of justice (2). Fi
nally, I will provide reasons for the thought that such a transformation
must also involve a new ascertainment of the relation between God and
pure practical reason (3).
1.
In the Introduction to the Critique of Practical Reason (CPrR), Kant claims

that
[…] we shall not have to do a critique of pure practical reason but only of
practical reason as such. For, pure reason, once it is shown to exist, needs
no critique [!]. […] It is therefore incumbent upon the Critique of Practical
Reason as such to prevent empirically conditioned reason from presuming
that it, alone and exclusively, furnishes the determining ground of the
will (CPrR, AA 5:15 f.; see already the very beginning: AA 5:3).
On the other hand, he asserts outright at the beginning of the Dialectic
that pure reason in its “practical use” also has “its dialectic” (CPrR, AA
5:107). The rising “unavoidable illusion” could only be removed
“through a complete critical examination of the whole pure faculty of rea
son” (ibid., my italics). An obvious question to ask is whether Kant is
simply contradicting himself here, or whether it is legitimate to speak
about pure practical reason from two different standpoints: the as it
were ‘analytical’ standpoint of pure practical reason (that might need
2 Therefore, “anthropological” might also be translated into “not purely rational”

in the critical philosophy. Two instructive German essays deal with that prob
lem/topic mainly with reference to the Groundwork and the Analytic of the sec
ond Critique: Forschner (1983) and Cramer (1991).
From Gratification to Justice 93
no critique) and the (as we will see) ‘dialectical’ standpoint of pure prac
tical reason (that needs a critical investigation and an additional stabili
zation). If there is such a dialectical standpoint, the assumption will sug
gest that this position is based on a certain mixture of pure and empiri
cally conditioned practical reason (as already becomes clear in the Intro-
duction).
To clarify this question, it seems helpful, first of all, to take a closer
look at the emergence conditions of the dialectic of pure practical rea
son in the second Critique. This dialectic rests upon the following cir
cumstance:
As pure practical reason it likewise seeks the unconditioned for the practi
cally conditioned (which rests on inclinations and natural needs), not in
deed as the determining ground of the will, but even when this is given
(in the moral law), it seeks the unconditioned totality of the object of
pure practical reason, under the name of the highest good (CPrR, AA 5:108).
It is eye catching here that Kant is speaks about a form of pure practical
reason that is from the outset structurally and “synthetically” related to
given empirical conditions like “inclinations and natural wants.”3 Like pure
theoretical reason, this “pure” practical reason tries to include all finite
“natural wants” in an all embracing way.4 At the outset, the Dialectic
3 Kant emphasizes this explicitly in the Religion within the Boundaries of Mere Rea
son: “The proposition, ‘Make the highest possible good in this world your own
ultimate end’, is a synthetic proposition a priori which is introduced by the moral
law itself, and yet through it practical reason reaches beyond the law [!]. And this
is possible because the moral law is taken with reference to the characteristic, natural to
the human being, of having to consider in every action, besides the law, also an end” (AA
6:7, n. 2, my italics).—This “a priori” might be called an ‘a posteriori condi
tioned a priori.’ In contrast to the law of pure practical reason (that could be
formulated by a complete non sensual, holy being as well), this “synthetic prop
osition a priori” essentially depends on human nature.
4 This already implies a tension with the Analytic: The demand of pure practical
reason is adapted here to the natural circumstances, whereas the categorical im
perative itself requires an adoption of all natural circumstances to its command.
Whereas pure practical reason seeks the unconditioned “for the practically con
ditioned (which rests on inclinations and natural needs)”, here in the Dialectic,
pure practical reason rather seeks the unconditioned against or through the subju
gation of the “practically conditioned” in the Analytic (see, e. g., CPrR, AA 5:44,
5:80 f. etc.).—One might notice furthermore that already the “pure” theoret
ical reason is not so “absolute pure” in case of the four antinomies: the a pos
teriori given, contingent conditioned resp. the contingent string of conditions is needed
here in order to ‘conclude’ the unconditioned as well (in contrast to the onto
logical proof of the existence of God, which is actually based on a pure notion).
94 Thomas Wyrwich
is therefore about “the” pure practical reason of a finite being and its
supposed perspective (which is by no means necessary, as the Ground-
work see once again AA 4:408 and 4:411 f. has brought to mind).
This becomes even clearer through the following dissection of the in
tended highest good:
That virtue (as worthiness to be happy) is the supreme condition of what
ever can ever seem to us desirable and hence of all our pursuit of happiness
and that is therefore the supreme good has been proved in the Analytic. But
it is not yet, on that account, the whole and complete good as the object of
the faculty of desire of rational finite beings [!]; for this, happiness is also re
quired, and that not merely in the partial eyes of a person who makes him
self an end but even in the judgement of an impartial reason, which regards
a person in the world generally as an end in itself. For, to need happiness, to
be also worthy of it, and yet not to participate in it cannot be consistent
with the perfect volition of a rational being that would at the same time
have all power, even if we think of such a being only for the sake of the
experiment (CPrR, AA 5:110; partly my italics, Kant’s italics partly delet
ed).
On the one hand, Kant is reactivating his old claim (inter alia spelled out
in the first Critique (CPR), see CPR A 806 ff./B 834 ff.) that the moral
worthiness to be happy is the supreme and irreducible condition of
every rational hope for happiness. On the other hand, the extending
(and maybe new) claim is added that even the pure practical reason of
finally “rational finite beings” itself should regard it as unjust and unaccept-
able if a being which naturally requires happiness and even deserves it,
might never achieve it.5 The “perfect volition of a rational being” is ar-
ranged here in such a way that this volition would intend in the case of a
being, whose “reason certainly has a commission from the side of its
sensibility which it cannot refuse, to attend to its interest” (CPrR, AA
5:61; my italics), the gratification of this natural, morally neutral com
mission as well of course if the being proves itself worthy.6
5 Milz (2002, 109 f.) draws attention to the possibility that Kant is referring here
to the motive of the “disinterested observer,” taken from the tradition of Eng
lish political philosophy. Milz points out that the function of this reference is to
indicate that this fair minded disinterested observer is coincidentally judging
from the point of view of pure practical reason, as the finite being, as I
would add, insinuates here in the Dialectic.—Before Milz also Albrecht (1978,
61 (n. 208), 70 (n. 229)) has taken the view that this “highest good” is an un
ambiguous idea of the “pure” practical reason.
6 Kant defines “happiness” in the Dialectic as “the state of a rational being in the
world in the whole of whose existence everything goes according to his wish
However, there are other passages in the Kantian work that are ap
propriate to indicate that such attributions and conclusions have not al
ways been as strict as they appear here. In Reflection 7059 (1776 78),
Kant notes more carefully:
The worthiness to be happy isn’t our first wish indeed but it is the first and
indispensable condition under which reason approves it. However, it seems
that reason also promises us something in this demand: namely, that some
one can hope to be happy if one behaves in a way that doesn’t make him or
her unworthy to be happy (AA 19:238; my italics and translation).
And consequently, it is no complete ‘pure moral reason’ in the “Canon”
of the first Critique (1781) that legitimates the assumption of a highest
good, but a reason of a gracious being “whose business it is to dispense
all happiness to others” (CPR A 813/B 814; my italics and translation).
Putting together these passages, the basic anthropological presuppo
sition of the Dialectic and the “old” idea of the gratification of the virtuous
man, that Kant has already discussed in lectures (see n. 6), we can draw
the following conclusion: Kant presupposes in the Dialectic a form of
pure practical reason that shall additionally want the fulfilled happiness
of a finite human being, insofar as the being acts on purely moral mo
tives. A being that is “sentenced to happiness” must have the real pos
sibility to become happy that is the extending postulation of this rea
son.
But once again the question arises if this is actually a demand of (ab
solute) pure practical reason or rather of the “empirically conditioned” prac
tical reason. The second possibility appears much more plausible once
we recall what Kant accentuates in the Analytic:
and will, and rests, therefore, on the harmony of nature with his whole end” (CPrR,
AA 5:124; my italics). What I personally wish and will, should—as long as it is
based on merely moral motives!—come true. Here, Kant obviously refers back
to his “old” idea (e. g. mentioned in different lectures) that virtue can claim
gratification or reward in the form of happiness, see: “The natural, moral belief
is with all virtuous man, as one believes that good actions will be rewarded as
such” (AA 24:243; my italics and translation).—“The man who lives morally
can hope to be rewarded for it” (AA 27:284; my italics).—“Every upright
man […] cannot be possibly upright, without hoping at the same time, on
the analogy of the physical world, that such righteousness must also be reward
ed” (AA 27:285).—In contrast to that, as Kant puts it 1788, the moral law itself
“demands of us disinterested respect”, and that “without [!] promising or
threatening anything with certainty” (CPrR, AA 5:147).
96 Thomas Wyrwich
But it is requisite to reason’s lawgiving that it should need to presuppose

only itself, because a rule is objectively and universally valid only when it
holds without the contingent, subjective conditions that distinguish one rational
being from another (CPrR, AA 5:20; my italics).
If that is true, it seems justified to assume that the finite, practically pure
rational being tries to constitute a “synthetic” harmony between the
necessary ought and its subjective, natural will under the name of the
“highest good.” But this “more” appears as a mixture of moral and an
thropological needs. Therefore, the Dialectic seems to rest upon the hy
postatization of a contingent human need (sensible happiness), that is a
“problem imposed upon him by his finite nature itself” (CPrR, AA 5:25;
my italics) 7, to an objective and therefore necessary demand of pure prac
tical reason.
It has been mentioned in the literature as well that such a hyposta
tization cannot work without a shift of perspective. Beck convincingly
points out:
The need of reason to believe in the existence of a highest good and to pos
tulate the existence of its conditions does indeed arise from ‘an objective
determining ground of the will’, but only because of ‘inescapable human
limitations’. It is therefore not a ‘need of pure reason’ but a need of all
too human reason (1960, 254).
And Hegel already notes in his Faith and Knowledge that the speculative
aspect of the ‘highest good’ has “certainly been transfused by Kant into
the human form that virtue and happiness harmonize” (TWA, 2:330; my
translation).
However, it does not seem necessary to presume that Kant had no
awareness of these relations.8 On the contrary, in the following passages
he is exhibiting “our reason” that isn’t able to conceive the highest good
“only on the presupposition of a supreme intelligence” (CPrR, AA
5:126). And it does not need Jacobi or Hegel to conclude that this
7 See also just before: “To be happy is necessarily the desire of every rational but
finite being and therefore an unavoidable determining ground of its faculty of
desire” (CPrR, AA 5:25). It shouldn’t be ignored that this “problem” isn’t im
posed upon us by reason but by “our nature”. When Denis, for example, ex
plicates (in accordance with Engstrom): “Our own happiness becomes an object
of pure practical reason when it is pursued on a maxim that gives priority to
virtue” (2005, 35), she is eluding the question why a naturally grounded (and
not only “moral”) happiness should be finally a real object of “pure” practical
reason.
8 Schwarz (2004, 258) argues in a similar way.
“our reason” cannot be fully identical with the “pure rational reason”
that sets from itself moral “bounds […] to the mankind [!]” (CPrR,
AA 5:85), as Kant underlines in the Analytic. 9 In fact, it is therefore
not the ‘pure’ practical reason itself that produces a dialectic subsequent
to the presupposition of the “highest good” but a finite, “empirically
supported” practical reason that apparently cannot avoid to grade up a
need “imposed upon us by our own finite nature” to a finally purely
moral demand.10 And, as we will see, also the aspired solution has to
rest upon a similar ‘upgrade’.
There is no room here for a detailed analysis of the complex antin
omy of practical reason. For the purpose of my topic it should be suf
ficient to mention that reason can only resolve the antinomy if it can
be demonstrated how the postulated highest good as a synthetic propo
sition a priori is possible. 11 Analogous to the first Critique, the first step
relies on the introduction of the critical difference between the sensible
and the intelligible world. This difference (that has already been opened
ex negativo by theoretical reason) shall principally reveal that the possibil
ity of the highest good is cogitable at all. Natural happiness as propor
tion to pure virtue could be an “effect” (see CPrR, AA 5:115) of an in
telligible cause; and even if this happiness would never be completely re
alized in the sensible world, it could be the case that this realization oc
curs in the intelligible world. However, this conception seems not suf
ficient to explain how the highest good could be realized. For such a task
a more specified theory about such proportion as causality is needed.
And since this highest good requires a certain form of “distribution”
(of happiness), it appears legitimate for finite reason to postulate a
“wise and all powerful distributor” (CPrR, AA 5:128). Nevertheless,
in a later retrospection, Kant consequently also limits that idea in the fol
lowing way:
9 See also: “Pure reason is practical of itself alone and gives (to the human being)
[!] a universal law which we call the moral law” (CPrR, AA 5:31); cf. AA
6:406.
10 It is quite difficult to decide whether the antinomies of pure theoretical reason
rest upon such an ‘anthropological absolutization’ as well. My conclusion here
refers only to the constellation of the second Critique. Nevertheless, it could be
quite interesting to take a closer look at Groundwork, AA 4:411 f. (“we must not
make its principles [pure practical reason] dependent on the particular nature of
human reason, though in speculative philosophy this may be permitted”) for
further investigations in this context.
11 Among others Milz (2002) has suggested such a convincing interpretation.
98 Thomas Wyrwich
I said above that in accordance with a mere course of nature in the world
happiness in exact conformity with moral worth is not to be expected and
is to be held impossible, and that therefore the possibility of the highest
good on this side can be granted only on the presupposition of a moral au
thor of the world. I deliberately postponed the restriction of this judgement
to the subjective conditions of our reason […]. In fact, the impossibility re
ferred to is merely subjective, that is, our reason finds it impossible for it to con
ceive, in the mere course of nature, a connection so exactly proportioned
and so thoroughly purposive between events occurring in the world in ac
cordance with such different laws (CPrR, AA 5:145).
The same restriction can be assigned to the other postulates, whose as
sumption is obviously altogether “subjective, that is, a need, and not ob-
jective, that is, itself a duty” (CPrR, AA 5:125). The postulate of the im
mortality of the soul, that should provide a possibility to make the “per-
fect accordance of the mind with the moral law” (AA 5:122) thinkable,
turns out to be a prolongation of anthropological structures into the in
telligible world. Because “all the moral perfection that a human being
can attain is still only virtue” (AA 5:128), and never holiness,12 there
is nothing left for “the creature but endless progress” (ibid.). And
whereas the Analytic was able to provide a “deduction” of freedom
(AA 5:47 f.) and its “reality” (AA 5:49 and 56) based on the moral
law, freedom becomes finally only a “postulate” in the Dialectic as
well (see AA 5:132 f.). The reason for that seems to be clear: The finite
human being can only hope and postulate that its freedom can actually be
naturally realized in the sensible world, in order to establish a condition
for the existence of a good the creature is hoping for on his part.
2.
Although Kant does not explicitly speak about “gratification” in the Di-
alectic, it seems obvious that this “old” conception has left its mark in the
idea of “happiness as proportion (to virtue)”. However, it is quite inter
esting to see that Kant refrains from certain elements he has exposed in
the Dialectic in subsequent works. In his essay about The End of All
Things from 1794, he takes a stand for the so called “dualists” from a
practical point of view (see AA 8:328 330). In contrast to the “unitar
ians”, who want to allow the possibility of an enduring process of ex
piating even after death (in order to enable a potential “eternal beati
12 See also CPrR, AA 5:82 – 84.

tude” for all humans), the “dualists” accept the mode of moral probation
only for the finite life (in order to let some take part in final “beatitude”
but all others in “eternal perdition” (AA 8:329) 13). So the ‘anthropolog
ical’ way of moral execution is strictly reserved for the sensible world by
the “dualists,” whose system should be, as Kant henceforth clarifies, the
“assumable” (AA 8:330) in practical philosophy, whereas the previous
postulate of a continuous immortal soul reminds much more of the sys
tem of the “unitarians.”
But such positional clarifying can already be found in the third Cri-
tique (CJ) from 1790. Kant’s restriction of what Hegel would call a “bad
infinity” (see TWA, 5:150 155, 288) to the sensible world is also re
flected in his reconstruction of the experience of the “sublime:” Our
frustrating feeling of an endless and always insufficient progress of our
imagination coincides for Kant with an elevating feeling of a “supersensible
faculty in us” (CJ, AA 5:250).14 The negative feeling has to be encom
passed by something positive and complete, which also accords with the
idea of a “true infinity” of the moral world (CPrR, AA 5:162; my italics)
that has already been exposed at the very end of the second Critique.
And once again, such a complete “true infinity” can obviously not be
identified with an “endless” moral pursuit of happiness. Corresponding
ly, such a progress of the soul has lost its basic function in the third Cri-
tique. Instead of that, we find a quite different argument for the necessity
of the “highest good.” The problem of the “virtuous atheist Spinoza”
(who just wants to act morally, does not believe in God and immortality
and does not care about the consequences of his actions at all) is not that
he ignores his own anthropological happiness. Rather his problem rests
upon the fact that
deceit, violence, and envy will always surround him, even though he is himself
honest, peaceable, and benevolent; and the righteous ones besides himself
that he will still encounter will, in spite of all their worthiness to be happy,
nevertheless be subjected by nature, which pays no attention to that, to all
the evils of poverty, illnesses, and untimely death, just like all the other an
imals on earth, and will always remain thus until one wide grave engulfs
them all together (whether honest or dishonest, it makes no difference). (CJ,
AA 5:452, my italics)
13 For Kant, this notion is also strongly connected with the idea of a “last judge
ment” (AA 8:328) and therefore, as we can conclude, connected with a kind
of ‘intelligible accountability’.
14 See also Wundt (1924, 431).
100 Thomas Wyrwich
“Honest or dishonest, it makes no difference” this seems to be decisive

here. An absolutized neutral nature (as well as every “unitarian”) has to
level this crucial difference. Therefore, the problem of the “highest
good” transforms into a merely moral problem of justice in the third Cri-
tique. This “moral” transformation can also be detected in Reflection
6454, written between 1790 and 1804, where Kant specifies the pro
portionality of virtue and happiness as a “purely moral wish” (AA
18:725). But “happiness” has to do very little here with our natural
“wish and will” or with gratification. Instead of that, our need to assume
a “highest moral good […] is a moral need to assume a just judge, who is
not [!] a being whose benevolence we hope for but whose holiness we
fear” (ibid.; my translation). “Happiness” (here as correlate to “punish
ment”) only means moral justice in this Reflection.
The justification for this moral transformation might be found in the
fact that Kant tries to exhibit a kind of analytical relationship between
pure morality and justice in his later works, last but not least in his Met-
aphysics of Morals from 1796/97. In the first part, in the Doctrine of Right,
he concludes:
The law of punishment is a categorical imperative, and woe to him who crawls
through the snake windings of the doctrine of happiness in order to dis
cover something that releases the criminal from punishment or even re
duces its amount by the advantage it promises, in accordance with the
pharisaical saying, ‘It is better for one man to die than for an entire people
to perish.’ For if justice goes, there is no longer any value in human beings’
living on the earth. (AA 6:331 f., my italics)
So the categorical imperative can also be reformulated in the following
way: Be just and try to donate justice! And since for Kant, “ought im
plies can,”15 such justice must be possible as well. But whereas in the ex
ternal realm of right and law such justice seems to be up to rational hu
mans, at least to some extent, such justice can never be adequately es
tablished in the realm of morality by human beings, since they have
no access to their own “real” moral characters16 and since they have
no full physical control of the external word. Therefore, Kant also
draws the conclusion at the very end of the Metaphysics of Morals that
only a divine and punishing being is able to guarantee such exhibited
moral justice that should not be mixed up with love or happiness:
15 Cf. CPrR, AA 5:30, AA 6:62, AA 8:287 f., AA 6:49 f., AA 22:507, AA

23:245.
16 See for example: CPR A 551/B 579 (n.), AA 6:392 and AA 8:284 f.
The divine end with regard to the human race (in creating and guiding it)
can be thought only as proceeding from love, that is, as the happiness of
human beings. But the principle of God’s will with regard to the respect
(awe) due him, which limits the effects of love, that is, the principle of
God’s right, can be none other than that of justice. We might, speaking as
we must do after the way of human beings [!], express it that way, that God
has created rational beings from the need, as it were, to have something
outside himself which he can love and by whom he can be loved in return.
[break] But not only as extensive, but even more extensive (for the principle is
restrictive) is the demand, which, even our own reason judges, divine justice, as
punitive, raises compared to us. Because a reward (praemium, remuneration
gratuita) has no reference to a justice against beings, that have only several
duties and no right against the other [divine being], but only to love and
benefaction (benignitas); even to a lesser extent such a being can have an en
titlement to wages (merces), and a remunerate justice (iustitia brabeutica) in
the relation of God to human beings is a contradiction [!] (AA 6:488 f.;
partly my translation, italics added).
“After the way of human beings” happiness is the last aim of a loving, di
vine will. But even our own (pure) reason tells us that a divine, in case
of doubt punitive justice is an “even more extensive,” all encompassing
principle.17 So the anthropological concept of reward, that is only con
nected with the ideas of “love and benefaction” and that has sustained
the concept of the highest good in the second Critique, is now clearly
contrasted with the concept of a purely moral justice.
3.
However, irrespective of the transformation of the basic conception of

the highest good, isn’t the idea of a judging and just being a very similar
construction to the one of the Dialectic of the second Critique? Isn’t it
again a very anthropomorphic idea to seemingly “postulate” such a
17 See also AA 27:284 ff., AA 28:1085, AA 28:1292 – 1294.—The idea of a moral,

punitive justice should not be mixed up with the contrarian concept of an iustitia
distributiva (see for example: AA 6:297). It is the latter concept of a rewarding
justice that seems to be associated in the following passage of the second Critique
as well: “And the holiness that his command inflexibly requires in order to be
commensurable to his justice in the share he determines for each in the highest good
is to be found whole in a single intellectual intuition of the existence of rational
beings” (CPrR, AA 5:123; my italics). Nevertheless Kant already exposes a
close conceptual connection between pure morality and the idea of a punitive
justice in the second Critique (AA 5:37 f.).
102 Thomas Wyrwich
(personal) divine judge? One might think so, but again, we find some
evidence in Kant’s late works that such an interpretation would not
be really appropriate. First of all, in the Religion within the Boundaries of
Mere Reason, Kant establishes an instructive connection between the
moral law and the idea of a divine lawgiver:
Agreement with the mere idea of a moral lawgiver for all human beings is
indeed identical with the moral concept of duty in general, and to this ex
tent the proposition commanding the agreement would be analytic (AA
6:6, n.).18
So here we can identify an analytical connection between the moral law
and a divine lawgiver, whereas in the second Critique, this connection
could only be established synthetically on the basis of a very human
“highest good.” Kant develops this connection further in the Opus post-
umum:
There must also, however, be—or at least be thought—a legislative force
(potestas legislatoria) which gives the laws emphasis (effect) although only
in idea; and this is none other than that of the highest being, morally and
physically superior to all and omnipotent, and his holy will—which justifies
the statement: There is a God. (AA 22:126)
Putting together the notion of an analytical relation between the moral
law and a correlative divine lawgiver on the one hand, and the notion of
a “legislative force which gives the laws emphasis (effect),” that can be
connected or even identified with the idea of justice, on the other hand,
we can draw the two following conclusions:
1. There must be a certain form of identity between God and the moral
law, and that is exactly what Kant says in the Opus postumum: “God
is [!] the moral practical autonomous reason” (AA 21:145; my ital
ics). “The concept of God is the idea of a moral being, which, as
such, is judging [!] and universally commanding. The latter is not a
hypothetical thing but pure practical reason itself” (AA 22:118; my ital
ics).19
2. If a judging God, or the moral law respectively, is connected with the
idea of a “legislative force,” then it seems legitimate to claim that
pure practical reason itself demands justice and since ought implies
18 See also Förster (2000, 136).

19 See very similar: AA 4:408 f., AA 8:264, AA 8:350 (n.), AA 22:104 f., AA
22:116, AA 28:1076.
can also has to “guarantee” the possibility of this justice (at least in
an intelligible world).
But such assumable and believable guarantee can only work if an “ob
ject” or “material” is demanded that is directly connected with pure prac
tical reason. And this cannot be, all in all, a naturalized happiness, but
only a (moral) justice.
Conclusion
In this paper I have tried to show that Kant’s dialectical theory of the
“highest good” and the three postulates that are required in order to
demonstrate the possibility of this highest good, intentionally rest
upon an assumption that someone might call an absolutization of an an
thropological standpoint the necessary moral “ought” and the very
natural “will” shall harmonize in this object as finite reason insinuates.
I have argued that it is also necessary (entirely analogous to the analytic
part of Kant’s pure moral philosophy) to restrict an overwhelming an
thropological perspective in order to establish a critical (meaning “pure
ly” moral here) conception of the highest good as well. I have tried to
show that Kant himself takes this route, starting in the third Critique and
ending in the Opus postumum. Ideas like “gratification” and (even di
vine) “love” are restricted to a human perspective in the Metaphysics
of Morals. All in all, in the critical philosophy it can be shown that con
cepts like a reward of virtue, an endless prolongation of the probation of
the soul (see once again Kant’s criticism in The End of All Things and his
theory of the “sublime”), a merely “postulated” freedom and last but
not least a benevolent distributor are conceptually taken from the sen
sible world of the finite human being. Kant finds an instructive classifi
cation for such conceptions in the third Critique: They turn out to be
just “jat’ %m¢qypom”, and not “jat’ !k¶¢eiam” (CJ, AA 5:463).
In contrast to that, the “true” idea of justice could be exposed as an
analytical, “material” implication of the moral law that in turn could it
self be identified with God from a certain perspective.20 It is important
20 It should be mentioned that even after the second Critique, Kant nevertheless
sometimes refers to the “anthropological” conception of the highest good as
a synthetic extension. Concerning the history of its development, it might be
more accurate to say that Kant is working with both conceptions in his late
works.
104 Thomas Wyrwich
to mention that this concept of moral justice should not be mixed up

with the “happiness” of the stoics in the second Critique, who pretend
to feel happy while acting morally. Justice (also as punitive) does not
mean a heteronomous “motivating” object, because it is equivalent to
say someone is motivated by the moral law itself or by the purely
moral “object,” which means that she is motivated to donate justice
(as pure practical, divine reason commands).21 So, if man is searching
for an “object” of pure practical reason, she should also search for some
thing that is not only “human.” This is something we might conclude in
accordance with Kant. It is, of course, Fichte who will take the same
direction.
Bibliography
Albrecht, Michael (1978): Kants Antinomie der praktischen Vernunft, Hilde
sheim, New York.
Beck, Lewis White (1960): A Commentary on Kant’s Critique of Practical
Reason, Chicago, London.
Cramer, Konrad (1991): Metaphysik und Erfahrung in Kants Grundlegung der
Ethik, in: Neue Hefte für Philosophie 30/31, pp. 15 – 68.
Denis, Lara (2005): Autonomy and the Highest Good, in: Kantian Review 10,
pp. 33 – 59.
Förster, Eckart (2000): Kant’s final synthesis: an essay on the Opus postumum,
Cambridge, London.
Forschner, Maximilian (1983): Reine Morallehre und Anthropologie. Kritische
Überlegungen zum Begriff eines a priori gültigen allgemeinen praktischen
Gesetzes bei Kant, in: Neue Hefte für Philosophie 22, pp. 25 – 44.
Hegel, Georg Wilhelm Friedrich (TWA 1969 ff.): Theorie Werkausgabe
(Werke in 20 Bänden), eds. E. Moldenhauer and K. M. Michel, Frankfurt
am Main.
Kant, Immanuel (1993): Opus Postumum, ed. E. Förster, Cambridge.
Kant, Immanuel (1996): The Metaphysics of Morals, ed. M. Gregor, Cam
bridge.
Kant, Immanuel (1997): Critique of Practical Reason, ed. M. Gregor, Cam
bridge.
Kant, Immanuel (1997): Groundwork of the Metaphysics of Morals, ed. M.
Gregor, Cambridge.
Kant, Immanuel (2000): Critique of the Power of Judgement, ed. P. Guyer,
Cambridge.
Kant, Immanuel (2005): Notes and Fragments, ed. P. Guyer, Cambridge.
21 In later passages this moral justice de facto substitutes the conception of a (hu
manly desired) gracious allocation that has sustained the concept of the highest
good in the second Critique.
Kant, Immanuel (2005): Religion within the Boundaries of Mere Reason, ed.
A. Wood and G. Di Giovanni, Cambridge.
Milz, Bernhard (2002): Der gesuchte Widerstreit. Die Antinomie in Kants Kri
tik der praktischen Vernunft, Berlin, New York.
Schwarz, Gerhard (2004): Est Deus in nobis. Die Identität von Gott und reiner
praktischer Vernunft in Immanuel Kants „Kritik der praktischen Ver
nunft“, Berlin.
Wundt, Max (1924): Kant als Metaphysiker. Ein Beitrag zur Geschichte der
deutschen Philosophie im 18. Jahrhundert, Stuttgart.
Anthropology, Empirical Psychology,
and Applied Logic
Job Zinkstok
Abstract
Kant’s anthropology has always been taken as a practical discipline (either moral
or pragmatic). Such readings neglect the fact that Kant also envisaged a use of
anthropology in logic. In this paper I explore this logical relevance of Kant’s an
thropology. I do so by first arguing that Kant’s anthropology is for an important
part concerned with empirical psychology. I then show that this empirical psy
chological part of anthropology is highly relevant for the branch of logic that
Kant calls applied logic, viz., the kind of logic that is “directed to the rules
of the use of the understanding under the subjective empirical conditions
that psychology teaches us”. I illustrate the relevance of anthropology for
logic by analyzing Kant’s conception of prejudice and showing how empirical
psychological knowledge is used in applied logic to prevent the errors arising
from prejudice.
Anthropological logica is applicata.1
Introduction
In the past decade, interest in Kant’s anthropology has risen steadily.

Thanks to this new engagement, anthropology has come to be recog
nized as an important supplement to Kant’s philosophy, as the empirical
and less formal side to a system otherwise (in)famous for its strong a priori
and formal nature.2 In these interpretations, Kant’s anthropology is al
1 Reflection (henceforth abbreviated as Refl.) 3332, AA 16:738 (dated 1780s, per

haps late 1770s; translation mine). References to Kant’s works use to volume
and page number of the Akademie Ausgabe (I. Kant, Gesammelte Werke, ed.
königlich preußische (später deutsche) Akademie der Wissenschaften, Berlin
1900 f.), abbreviated AA. Translations of Kant’s works are from the Cambridge
Edition of the Works of Immanuel Kant, unless otherwise stated.
2 See especially Louden (2000).
108 Job Zinkstok
ways taken as a practical discipline, i. e., as a discipline dealing with

human action and comportment within society, be this action moral
or merely pragmatic.3 This is not strange, as Kant explicitly advertises
his anthropology as an anthropology from a pragmatic point of view (cf.
the title of his 1798 book). Moreover, he also explicitly envisioned a
moral anthropology, for example in the Metaphysics of Morals (1797):
The counterpart of a metaphysics of morals, the other member of the di
vision of practical philosophy as a whole, would be moral anthropology,
which, however, would deal only with the subjective conditions in
human nature that hinder people or help them in fulfilling the laws of a
metaphysics of morals. It would deal with the development, spreading,
and strengthening of moral principles (in education in schools and in pop
ular instruction), and with similar teachings and precepts based on experi
ence.4
As I will argue, however, this specific interest has given rise to a rather
one sided evaluation of Kant’s anthropology as a merely practical disci
pline. Although it is certainly not wrong to read Kant’s anthropology as
such, I will argue that the contents of this discipline are of broader ap
plicability than just the action of human beings in society. This is due to
the fact that an important part of anthropology consists in empirical psy-
chology. As I will show, Kant actually considered at least one other area
of application for this psychological content namely that of logic, more
specifically, applied logic. Applied logic, for Kant, is a branch of logic in
the broad sense, and it is described in the Critique of Pure Reason as “di
rected to the rules of the use of the understanding under the subjective
empirical conditions that psychology teaches us”.5
In this paper I will explore this connection between anthropology
and logic. In section 1, I will start by briefly reviewing the common,
moral interpretation of Kant’s anthropology. After that, I will turn to
the relation between empirical psychology and Kant’s anthropology, ar
guing that Kant indeed envisioned an empirical psychological discipline
as part of anthropology. Moreover, I will argue that even in the anthro
3 Note that I use the term ‘discipline’ not in the Kantian sense, but rather as a
general term indiscriminately referring to any more or less systematic or scien
tific body of knowledge. I use the term ‘practical’ here and below in a wide
sense, i. e., as encompassing both moral and non moral action. This is in con
trast to Kant’s use of the term in (for example) the Critique of the Power of Judg
ment (CPJ), where it is explicitly tied to morality (e. g. on AA 5:171).
4 Metaphysics of Morals (MM), AA 6:217.
5 Critique of Pure Reason (CPR), A 53/B 77.
Anthropology, Empirical Psychology, and Applied Logic 109
pology lectures, although they have a decidedly pragmatic (non theo

retical) aim, empirical psychology nonetheless constitutes an important
layer.6 The third section will then turn to logic, in order to investigate
Kant’s view on the connection between logic and psychology. We will
see that Kant came to reject any psychological foundation for formal
logic (or pure general logic), but that he explicitly allowed for another
kind of logic, applied logic, which draws on psychology in order to
teach the way human beings should reason in concrete situations. In
the fourth section, I will flesh out Kant’s conception of applied logic
by offering an example and showing how applied logic is informed
by empirical psychology as it is offered in anthropology.
1. Kant’s anthropology and a common interpretation of it
When speaking of Kant’s anthropology, one generally refers to the book

Anthropology from a Pragmatic Point of View, which Kant published in
1798. This book is the outcome of the lectures on anthropology that
he offered every year from 1772 until 1796, as a companion course to
his lecture on physical geography. Together these courses formed a
kind of popular introduction to knowledge of the world, and they
were indeed Kant’s most popular lectures.7
In the Anthropology Kant defines this discipline as a “doctrine [Lehre]
of the knowledge of the human being [Kenntniß des Menschen], system
atically formulated”.8 He distinguishes two points of view from which
anthropology can be pursued: a physiological and a pragmatic one.
Physiological anthropology concerns “what nature makes of the
human being”, whereas pragmatic anthropology concerns “what he as
6 I use the term ‘theoretical’ in this paper as referring to cognition that pertains to
the domain of nature as opposed to the (practical) domain of freedom (cf. CPJ,
AA 5:171). Note that when I speak of theoretical cognition in this paper,
which deals with the empirical cognition of anthropology and psychology, I
do not wish to include transcendental philosophy.
7 For more on Kant’s anthropology lectures, see Stark (2003), as well as Brandt
and Stark (1997).
8 Anthropology from a Pragmatic Point of View (Anthr.), AA 7:119. Note that a doc
trine [Lehre] is not taken here as Doktrin, which is opposed to Kritik (cf. CPR
B 25). It rather refers to the way Kant also uses the term ‘Lehre’ in the preface
to the Metaphysical Foundations of Natural Science (MFNS, AA 4:467). Moreover,
keep in mind that the Lehre of anthropology concerns an empirical discipline.
110 Job Zinkstok
a free acting being makes of himself, or can and should make of him
self”.9 As follows from the fact that Kant, with his anthropology lecture,
wanted to prepare his students for the world outside of the university,
his lecture provides a pragmatic anthropology. Taking the faculty of
memory as an example, Kant explains that theoretical reasoning on
the physical causes of memory in the brain is a futile enterprise. More
successful is the following approach:
But if he uses perceptions concerning what has been found to hinder or
stimulate memory in order to enlarge it or make it agile, and if he requires
knowledge of the human being for this, then this would be a part of an
thropology with a pragmatic purpose, and this is precisely what concerns
us here.10
The Anthropology is divided into two main parts. The first, the “Anthro
pological Didactic”, concerns the “manner of cognizing the interior as
well as the exterior of the human being”,11 and it treats the three main
faculties of the human mind: the faculty of cognition, the feeling of
pleasure and displeasure, and the faculty of desire.12 In all three cases,
Kant treats both the higher and the lower part of the faculty (e. g., in
the case of the faculty of cognition both the understanding and sensibil
ity). The contents of this part thus present the human mind as it is given
in all human beings. The second part, the “Anthropological Character
istic”, concerns the “manner of cognizing the interior of the human
being from the exterior”.13 The “Characteristic” treats the notion of
character and especially the differences in character among human beings,
more specifically, differences that depend on for example sex, nation
and race.
As I indicated above, interpretations of Kant’s anthropology have
focused on its practical orientation. Reinhard Brandt (partly together
with Werner Stark), for example, has argued that although Kant started
his anthropology lectures as a theoretical course on empirical psycholo
gy, by the Winter Semester of 1773 1774 the lecture had undergone a
“pragmatic turn”.14 After this turn, Brandt argues, Kant’s anthropology
is a “doctrine of prudence [Klugheitslehre]”, in which students are taught
9 Anthr., AA 7:119.
10 Anthr., AA 7:119.
11 Anthr., AA 7:125 (translation modified).
12 This division is of course also to be found in the CPJ, AA 5:178 ff.
13 Anthr., AA 7:283 (translation modified).
14 Brandt (1994); for the term ‘pragmatic turn’, see, Brandt and Stark (1997, xvii).
the rules of the game that is human society. Now although Brandt con
cedes that in this new science of pragmatic anthropology there are still
traces of the older psychological content, he is quite adamant that “Kant
frees himself from the idea of the natural cognition of man as an empir
ical psychology qua theoretical discipline”,15 and that he starts to “in
creasingly refrain from presenting his new discipline [i.e., anthropology]
as psychology, and even emphasize that his anthropology is not psychol
ogy”.16 Moreover, Brandt claims that “anthropology has, as a practical
discipline, changed its epistemological place” in that it “is excised not
only from metaphysics, but also from the academic disciplines in the
narrower sense, to which physics and empirical psychology belong”.17
For confirmation of his reading, Brandt draws on the passages in the An-
thropology and the lectures in which Kant explicitly says that he is dealing
with pragmatic anthropology, and he also points to a letter Kant sent to
Marcus Herz toward the end of 1773, in which he distances his ap
proach to anthropology from the psychosomatic approach of Ernst Plat
ner, who published his Anthropologie fr rzte und Weltweise in 1772.18
Brandt further argues that the anthropology lectures are not identical
to the moral anthropology Kant calls for in some of his writings.19 Al
though Kant touches on moral issues here and there in the lectures, he
never uses the terms ‘categorical’, ‘imperative’ or ‘autonomy’.20
Robert Louden has a somewhat different approach. Although he
shares Brandt’s view that Kant’s anthropology is a practical discipline,
he explicitly hails Kant’s anthropology as “the second part of morals”,
thus interpreting it as moral anthropology. 21 Louden admits that Kant no
where in his lectures nor in the Anthropology explicitly says he is treating
this moral anthropology, and he also concedes that the anthropology
lectures do not offer a systematic and straightforward account of it.
Still, disagreeing with Brandt, he is convinced that moral anthropology
is an important ingredient of Kant’s anthropology: “[a]lthough Kant
nowhere (i. e., neither in the anthropology lectures nor anywhere
else) hands over to readers a single, complete, tidy package of moral an
thropology, I aim to show that a bit of careful detective work neverthe
15 Brandt (1994, 21).

16 Brandt and Stark (1997, xi).
17 Brandt and Stark (1997, xiv).
18 AA 10:145 – 146. I come back to this issue in the next section.
19 For example in MM, AA 6:217.
20 Brandt and Stark (1997, xlvi–xlvii).
21 Louden (2003); cf. Louden (2000), esp. Chapter 3.
112 Job Zinkstok
less can lead us to some fulfilled hopes regarding Kant’s philosophia mo

ralis applicata.”22 Louden goes on to elaborate on the moral dimension of
anthropology, dealing with Kant’s treatment of particular, subjective
hindrances to carrying out the moral law, the relevance of knowledge
of the world for making morality effective in the lives of human beings,
and the way in which anthropology specifies the “moral map” that ori
ents our life at the “normative destination” of humanity.23
Although it is clear that Brandt and Louden disagree concerning the
moral dimension of Kant’s anthropology, they do agree that it is a prac-
tical discipline: anthropology deals with human action, be it moral or
merely prudential. In what follows I will not challenge this line of in
terpretation by arguing that Kant’s anthropology does not concern ac
tion at all rather, I will show that the use of some of its contents
does not limit itself to practical matters (whether moral or pragmatic),
but that it is relevant as well for applied logic.
2. Empirical psychology and anthropology
As I indicated in the introduction above, applied logic is a branch of

logic that “is directed to the rules of the use of the understanding
under the subjective empirical conditions that psychology teaches
us”.24 It is thus a branch of logic that is informed by empirical psycho
logical knowledge. In order to prepare our investigation of applied
logic, in the present section I will argue that Kant’s anthropology indeed
contains empirical psychology; and that for him empirical psychology,
although a problematic discipline and not a proper science, is still a via
ble enterprise. In order to see this, we have to take another look at the
anthropology lectures, more specifically at its first and biggest part, the
part which is called “Anthropological Didactic” in the Anthropology.
As said before, this first part of Kant’s anthropology deals with the
three main faculties of the mind. It derives mainly from Kant’s concern
with the empirical psychology as it was taught in the Wolffian tradition.
More specifically, it is based on the empirical psychology part of
Alexander Baumgarten’s Metaphysica, the textbook Kant also used for
22 Louden (2003, 64).

23 Louden (2003, 67 – 75; the quotations are on p. 72). Firla (1981) and Frierson
(2003) emphasize the moral importance of anthropology, too.
24 CPR A 53/B 77.
his metaphysics lectures. In fact, even before Kant started lecturing on

anthropology in 1772, he already treated the empirical psychology
part in those metaphysics lectures, and quite extensively at that.
Throughout his lecturing career it remained a substantial part of his
metaphysics lecture, even though he scaled it down because he incorpo
rated the material in his anthropology lectures.25 However, although
Kant retained empirical psychology in metaphysics, from 1770 onwards
he was very clear that it does not properly belong there: being an em
pirical discipline, it has no place in metaphysics, which is by definition a
priori. 26 In the Critique of Pure Reason Kant explains this, as well as the
reason for retaining it in his lectures on metaphysics:
[Empirical psychology] comes in where the proper (empirical) doctrine of
nature [Naturlehre] must be put, namely on the side of applied philosophy,
for which pure philosophy contains the a priori principles, which must
therefore be combined but never confused with the former. Empirical psy
chology must thus be entirely banned from metaphysics, and is already ex
cluded by the idea of it. Nevertheless, in accord with the customary scho
lastic usage [Schulgebrauch, i. e., academic usage] one must still concede it a
little place (although only as an episode) in metaphysics, and indeed from
economic motives, since it is not yet rich enough to comprise a subject
on its own and yet it is too important for one to expel it entirely or attach
it somewhere else where it may well have even less affinity than in meta
physics. It is thus merely a long accepted foreigner, to whom one grants
refuge for a while until it can establish its own domicile in a complete [aus
fhrlichen] anthropology (the pendant to the empirical doctrine of nature).27
It would thus seem to be the case that Kant came to consider empirical
psychology as a part of the empirical discipline of anthropology. Indeed,
in the Critique of the Power of Judgment Kant repeatedly characterizes em
pirical psychology as an anthropological discipline, for example where
he claims that psychology “is really merely an anthropology of the
inner sense, i. e., knowledge of our thinking self in life, and as theoretical
25 See the letter to Marcus Herz of 20 October 1778 (AA 10:242): “My discussion
of empirical psychology is now briefer, since I lecture on anthropology.” In the
metaphysics lectures empirical psychology can be found in the following tran
scripts. For the 1760s: Metaphysik Herder, AA 28:143 – 144, 850 – 886 and 924 –
931. For the 1770s: Metaphysik anon Korff (K1), AA 28:1519 – 1520 and Meta
physik anon L1 (Pçlitz), AA 28:228 – 262. For the 1780s: Metaphysik Mrongovius,
AA 29:877 – 904. For the 1790s: Metaphysik anon L2 (Pçlitz), AA 28:584 – 590;
Metaphysik Dohna, AA 28:670 – 679; Metaphysik anon K2, AA 28:815 – 816; and
Metaphysik Vigilantius (K3, Arnoldt), AA 29:1009 – 1024.
26 See the Inaugural Dissertation (ID), AA 2:397. Cf. Brandt (1994, 16).
27 CPR A 848 – 849/B 876 – 877.
114 Job Zinkstok
cognition it also remains merely empirical”.28 But what is important in

this quotation is not only the fact that empirical psychology is said to be
part of anthropology, but also that Kant characterizes it as a theoretical
discipline. Indeed, in other works it is confirmed that empirical psychol
ogy is part of the doctrine of nature (as opposed to the doctrine of
freedom). This can be made clearer by drawing on the Metaphysical
Foundations of Natural Science. There Kant divides the doctrine of nature
in general into that of outer sense (the doctrine of body) and that of
inner sense (the doctrine of the soul). Both, we learn further, have a ra
tional and a historical (empirical) part.29 He thus explicitly allows for an
empirical theoretical discipline treating the soul. It is this doctrine of na
ture of inner sense that is explicitly called the anthropology of inner sense
in the third Critique. 30 I would suggest, therefore, that for Kant a “com
plete anthropology” not only comprises the empirical counterpart to
practical philosophy (moral anthropology), but also part of the empirical
counterpart to theoretical philosophy (next to empirical physics): the
empirical doctrine of man as an object of outer sense (anthropology
of outer sense, i. e., doctrine of the body, or medicine) and of inner
sense (anthropology of inner sense, i. e., empirical psychology).31
One may raise several objections to this account. Whereas Kant in
the passages just cited relegates empirical psychology to anthropology
as a theoretical discipline, in other places he claims that a theoretical em
pirical psychology is not viable. He does so most conspicuously in the
Anthropology. With regard to inquiring into the natural causes of
human mental faculties, he says that “all theoretical speculation about
this is a pure waste of time”.32 And already in a letter to Marcus Herz
of 1773, Kant raises such criticism, in contrasting his plan for the an
28 CPJ, AA 5:461; cf. CPJ, AA 5:277.

29 MFNS, AA 4:467 – 468.
30 Cf. MM, AA 6:385; and What Real Progress Has Metaphysics Made in Germany
since the Time of Leibniz and Wolff ? (RP), AA 20:308.
31 An early reader of Kant, G.S.A. Mellin, is quite clear in dividing Kant’s view of
anthropology thus: see the lemma “Anthropologie” in Mellin (1797, 277 –
282). Also Kim (1994, 147 – 153), in his careful investigation of Kant’s concep
tion of anthropology, proposes this as a reconstruction of Kant’s views. It must
be noted, of course, that medicine does not belong to the (physical) doctrine of
body as it is actually treated in the MFNS; rather, it would be part of biology, or
the doctrine of organisms. This is outside of the scope of this paper, however.
32 Anthr., AA 7:119.
thropology lecture with Ernst Platner’s book Anthropologie fr Aerzte und
Weltweise, which was published in 1772:
This winter I am giving, for the second time, a lecture course on Anthro
pology, a subject that I now intend to make into a proper academic disci
pline. But my plan is quite unique [gantz anders; i. e., different from Plat
ner’s]. I intend to use it to disclose the sources of all the sciences, the sci
ence of morality, of skill, of human intercourse, of the way to educate and
govern human beings, and thus of everything that pertains to the practical. I
shall seek to discuss phenomena and their laws rather than the first grounds
of the possibility of the modification of human nature in general [erste
Grnde der Mçglichkeit der modification der menschlichen Natur berhaupt].
Hence the subtle and, to my view, eternally futile inquiries as to the man
ner in which bodily organs are connected with thought I omit entirely.33
This objection, however, is not justified. One must realize that in the
eighteenth century there were different approaches to empirical psy
chology (and to anthropology in the broader sense), and that Kant’s
criticism in the passages cited is directed against a specific kind of psycho
logical investigation rather than against empirical psychology in general.
Kant criticizes an empirical psychology that searches for the physical or
biological causes of mental processes an approach that can be designat
ed as physiological psychology.34 What is problematic is the attempt to
trace back mental phenomena to their “first grounds”. Hence Kant does
not think that empirical psychology can be a science in which phenom
ena are explained by grounding them. This kind of psychology would
involve tracing mental phenomena back to their physical, corporeal
grounds something Kant is not very optimistic about. In short: Kant
criticizes a naturalistic empirical psychology, an approach that was
33 Letter to Marcus Herz, end of 1773, AA 10:145 (translation modified).

34 See Sturm (2008) for more on eighteenth century physiological approaches to
psychology. He mentions Johann Theodor Eller (1689 – 1760), Ernst Platner
(1744 – 1818), Johann Gottlob Krüger (1715 – 1759), Charles Bonnet (1720 –
1793) and David Hartley (1705 – 1757) as representatives of this approach. It
must be noted, by the way, that Sturm uses the term ‘empirical psychology’
to designate the (among others) Wolffian approach to the science of man; he
also, alternatively, refers to physiological approaches as ‘anthropological’, as
the latter term was commonly understood to refer to man as a whole, i. e., com
prising both mind and body. If I use these terms in this paper, however, I do
not follow Sturm. I use the term ‘empirical psychology’ as a general term for the
science of the mind, encompassing both the Wolffian and the physiological ap
proach.
116 Job Zinkstok
known as physiological psychology.35 As it is, Kant’s criticism of a spe

cific approach to empirical psychology (viz., physiological psychology)
does not imply that other approaches to empirical psychology are equal
ly problematic. Indeed, it would be very strange if Kant would take a
discipline he considers entirely futile to be an important source for his
anthropology.
Still, this is not to say that Kant is highly optimistic about the pros
pects of empirical psychology as a science. As is well known from the
famous passage in the preface to the Metaphysical Foundations of Natural
Science, Kant denies psychology the status of a proper science: “Yet the
empirical doctrine of the soul [empirische Seelenlehre] must remain even
further from the rank of a properly so called natural science than
chemistry.”36 He gives two reasons for this. First, psychology does not
easily lend itself to mathematization; and second, in psychology one
cannot properly isolate phenomena and conduct experiments. This
claim and the reasons Kant gives are difficult to interpret, and I will
not go into this too deeply. Suffice it to say with regard to the first rea
son that psychology, dealing with phenomena of inner sense, does not
allow the a priori construction of its objects in space, only in time its
mathematical treatment is thus more difficult than that of physics.37
The second reason Kant gives concerns a number of problems encoun
tered in trying to investigate mental phenomena: it is difficult to get
hold of such fleeting phenomena, and they do not lend themselves to
systematic, repeated observation, but only to thought experiments.
Moreover, observation of one’s own mental phenomena tends to alter
these phenomena.
In the Anthropology Kant addresses this second, methodological rea
son for the lower scientific status of empirical psychology as well. There
he does not only draw attention to the disturbing effect that observing
someone has on his or her behaviour, but also to the fact that observing
35 I thus do not agree with Sturm when he states that Kant “never claims that
these anthropologies [viz., physiological ones] cannot be scientific” (Sturm
2008, 496). As Kant says physiological psychology is “eternally futile”, I do
not see how Sturm can conclude that Kant “never argues against the possibility
of a physiological anthropology” (ibid, 499). In my view, Kant’s remarks do not
seem to leave open much room for the possibility of a scientific physiological
psychology.
36 MFNS, AA 4:471.
37 See Sturm (2001) for a convincing argument for this specific “restricted impos
sibility claim”.
oneself is difficult: that “when the incentives are active, he does not ob
serve himself, and when he observes himself, the incentives are at
rest”.38 Moreover, Kant expresses serious reservations concerning the
value of introspection: it easily leads to “enthusiasm [Schwrmerei] and
madness”, especially when one tries to observe the inadvertent course
of one’s mind.39 Because of these methodological problems with
which empirical psychology is confronted, Kant came to advocate a dif
ferent methodological approach. Empirical psychology should rely on
the observation of actual behaviour rather than on introspection.40 Indeed,
in the Anthropologie Mrongovius (1784/1785) we find the remark that
“This [the state of one’s mind] I can experience just as well by means
of attention to my actions”.41 Similar remarks can be found in the Men-
schenkunde (1781/1782),42 and also in the Anthropology Kant states that
one can obtain knowledge of human beings through social intercourse,
and that the main features of characters in novels and plays “have been
taken from the observation of the real actions of human beings”, which
is why they are useful to anthropology.43 Consequently, the methodo
logical emphasis in Kant’s study of the human mind shifts from intro
spection to observation of behaviour.
In sum, the problems with empirical psychology for Kant come
down to two things. In the first place there are methodological reasons
on account of which Kant argues that no kind of empirical psychology
can be a proper science: mathematization is impossible, and experimen
tation and introspection are unviable methods of research. Although
these reasons imply that empirical psychology cannot attain the highest
scientific status, they do not relegate empirical psychology to fiction it
can still be a systematically ordered science. Indeed, Kant proposes that
instead of relying on introspection, (any kind of) empirical psychology
should rely rather on observation of external behaviour in order to
draw conclusions about internal psychological mechanisms. This
38 Anthr., AA 7:121. The Cambridge Edition of the Works of Immanuel Kant has “and
when he does not observe himself” (my emphasis) in the second part of this sen
tence, which is a mistake. The original has “und wenn er sich beobachtet”.
39 Anthr., AA 7:132 – 134.
40 See Sturm (2001), esp. 174 – 178, for more details on this. Sturm argues that
Kant, with this methodological stance, argues especially against Baumgarten,
whom he considers to be a strong proponent of introspectionist psychology.
41 AA 25:1219 – 1220 (translation mine). Cf. Sturm (2001, 174 – 175).
42 AA 25:856 – 857.
43 AA 7:120 – 121.
118 Job Zinkstok
comes down to a methodological restriction. Secondly, there is Kant’s

criticism of physiological psychology on account of the impossibility
(or at least difficulty) of getting to know the material causes of mental
phenomena. With this, Kant limits the scope of empirical psychological
knowledge. As a result, he considers a specific kind of empirical psy
chology (viz., physiological psychology) futile, because it aims at obtain
ing knowledge of precisely those material causes. The upshot is thus that
empirical psychology is restricted: both in its method and in its scope.
What he does not argue for, however, is the impossibility of empirical
psychology.
Still, even if empirical psychology is a viable theoretical discipline
that is part of a complete anthropology, how does this square with
Kant’s repeated claims that anthropology is a pragmatic discipline? Cer
tainly it cannot be denied that Kant’s anthropology has a pragmatic aim,
and that its main intention was to provide knowledge of the world in
order to teach what man “as a free acting being makes of himself, or
can and should make of himself”.44 For that purpose, anthropology is
to provide students with knowledge that helps them to actually go
about in the world, and to govern and influence both themselves and
others. It thus aims at knowledge of man as a citizen of the world:
Such an anthropology, considered as knowledge of the world, which must
come after our schooling, is actually not yet called pragmatic when it contains
an extensive knowledge of things in the world, for example, animals, plants,
and minerals from various lands and cultures, but only when it contains
knowledge of the human being as a citizen of the world.45
There is thus no doubt concerning the aim of Kant’s anthropology: it is
pragmatic, and therefore practical. That does not mean, however, that
everything it contains is pragmatic or practical in nature as well. The
passage just cited indeed confirms this, as it presents theoretical knowl
edge of the world (here specifically of physical and biological objects) as
a first step in the direction of a pragmatic discipline: just in virtue of
containing this theoretical knowledge it is “not yet” pragmatic. Rather,
it has to be applied to, implemented in a pragmatic context. Hence Kant
does not have any problems with presupposing theoretical knowledge as
a source for anthropology and it is obvious that an important part of
the knowledge you need in order to govern yourself and deal with
other people concerns psychology.
44 Anthr., AA 7:119.
45 Anthr., AA 7:120.
In order to clarify this further, it is useful to distinguish here be

tween anthropology as the lecture course Kant offered, and the com
plete anthropology as it is part of Kant’s encyclopaedia of the sciences.
As I argued, in the latter Kant distinguished between a theoretical and a
practical branch of anthropology, empirical psychology being part of the
theoretical branch. The lecture course can conversely be thought of as a
mixture of both branches: it has a strong pragmatic slant (and may justly
be called pragmatic anthropology because of that), but incorporates the
parts of theoretical anthropology (empirical psychology) where it is nec
essary to grasp the pragmatic knowledge that is the purpose of the
course.
It will be clear that the results we have obtained so far deviate from
the common interpretations of Kant’s anthropology I discussed in the
previous section. Although Brandt is certainly right that the pragmatic
anthropology as Kant taught it is a practical discipline, his claims do
not hold for the broader conception of anthropology, which, as I
have argued, encompasses at least some theoretical cognition. More spe
cifically, according to Kant this broader conception of anthropology
contains empirical psychology, as the quotations from the Critique of
Pure Reason and the Critique of the Power of Judgment prove. Any argu
ments that can be found in Kant’s works against the viability of empiri
cal psychology, I have argued, are to be understood as directed at spe
cific approaches to or methods applied in empirical psychology, not at
empirical psychology in general. It is thus clear that even after the
“pragmatic turn” in the lectures Kant thought that a theoretical, empir
ical psychological part within anthropology was viable. It is therefore, I
would suggest, more plausible that this pragmatic turn was confined to
anthropology as the specific lecture course, and did not apply to the more
encompassing discipline of a complete anthropology.
Further, it is perfectly all right if Louden singles out the moral rele
vance of Kant’s anthropology. At the same time, however, it will by
now be clear that this point of view is one sided, and should be com
plemented by a broader view which encompasses not only the non
moral side to human action in society, but also the theoretical parts
that have relevance outside of the merely practical.
120 Job Zinkstok
3. The use of anthropology in logic
Our next step is to turn to logic, and to see where empirical psycholog
ical considerations come in. Let me first introduce Kant’s division of
logic into several branches, as he offers it in the introduction to the
Transcendental Logic in the first Critique. There Kant defines logic as
“the science of the rules of understanding in general”.46 He then pro
ceeds to divide logic according to the way it can be undertaken: either
with regard to the general use of the understanding (general logic) or
with regard to its particular use (particular logic). General logic “contains
the absolutely necessary rules of thinking, without which no use of the
understanding takes place, and it therefore concerns this [understanding]
without regard to the difference of the objects to which it may be di
rected” a condition he later specifies by stating that general logic ab
stracts “from all content of cognition, i. e. from any relation of it to the
object”. 47 Particular logic, on the contrary, “contains the rules for correctly
thinking about a certain kind of objects”.48 General logic, in turn, is sub
divided into pure logic and applied logic. Applied general logic concerns
the “rules of the use of the understanding under the subjective empirical
conditions that psychology teaches us”, and is called a “cathartic of the
common understanding”.49 Pure general logic, on the other hand, abstracts
from these psychological conditions of thought. It is the “pure doctrine
of reason [Vernunftlehre]”, and it is said to be “properly scientific, al
though brief and dry, as the scholastically correct presentation of a doc
trine of the elements [Elementarlehre] of the understanding requires”.50
With this division of logic, Kant first of all takes a stance against the
Wolffian conception of logic. Wolff famously argued that logic derives
principles from both ontology and psychology: from the former because
ontology must teach “what to look for in order to know things”, the
latter because psychology must teach “how the operations of the intel
46 CPR A 52/B 76.

47 CPR A 52/B76 and A 55/B79; translation modified, and emphasis in latter
quotation mine. For more on the difference between both characterizations
of the relation of logic to objects, and a defence of the second as the more ac
curate one, see Tolley (2007, esp. 127 – 132).
48 CPR A 52/B 76.
49 CPR A 53/B 77 – 78.
50 CPR A 53 – 54/B 78.
lect are used in knowing truth”.51 Kant’s view of logic is diametrically

opposed to this, as he thinks that there is a formal core of logic, i. e.,
pure general logic, which has recourse neither to any information con
cerning the object of thought, nor to any information concerning the
subject of thought (i. e., the mind). Because of this radically new concep
tion of logic, Kant has been hailed as the first to speak of “formal
logic”.52
This strict conception of the core of logic, however, should not
make us forget that Kant, as his division of logic shows, also has a broad
er conception of logic, in which various other logical, though less for
mal disciplines find their place. Important among those are the particular
logics, logics that refer to specific objects to which logic may be applied,
and which form methodologies of specific sciences. An important ex
ample would be the particular logic of mathematics.53
Our concern, however, is not with the logics that specifically refer
to the object to which logic may be applied, but rather with the logic that
specifically refers to the human mind that is to carry out the thinking, ap-
plied general logic. Kant’s use of the term ‘applied logic’ is somewhat
idiosyncratic here, as he himself admits: it does not comprise “certain
exercises to which pure logic gives the rule”, as was common in his
time.54 Rather, it concerns the conditions under which thought is exer
cised in the case of human beings, as can be inferred from the fact that its
counterpart, pure logic, is said to abstract
from all empirical conditions under which our understanding is exercised,
e. g., from the influence of the senses, from the play of imagination, the
laws of memory, the power of custom, inclination, etc., hence also from
the sources of prejudice, indeed in general from all causes from which cer
tain cognitions arise or may be supposed to arise, because these merely con
51 See § 89 of the “Discursus praeliminaris” in Wolff’s Latin Logic, Wolff

(1983[1740]), 39 – 40: “quaenam sint ea, ad quae in rerum cognitione attendere
tenemur”; “usum […] operationum ipsius [i.e., intellectus] in veritate cogno
scenda”. Translation from Wolff (1963 [1728]).
52 See for example MacFarlane (2000, 20 – 22 and esp. 95 – 113). Michael Wolff
(1995, 207, esp. n. 44), however, claims earlier ancestry for this term (most im
portantly Joachim Jungius in his Logica Hamburgensis of 1638), but Longuenesse
(2006, 163 n. 8), argues that formality nevertheless plays a ground breaking role
for the first time in Kant’s work.
53 See Michael Wolff (1995, 205 – 221), for an explication of what this particular
logic of mathematics would be for Kant; he carefully argues that the rule of
mathematical induction would belong to this particular logic of mathematics.
54 CPR A 54/B 78.
122 Job Zinkstok
cern the understanding under certain circumstances of its application, and

experience is required in order to know these.55
A few pages later Kant adds to this that applied logic “deals with atten
tion, its hindrance and consequences, the cause of error, the condition
of doubt, of scruple, of conviction, etc.”.56
Now what do these characterizations of applied logic tell us about
this discipline? It seems quite clear that applied logic provides a descrip
tion of the way thought proceeds in concreto, as carried out by actual
human beings. Moreover, as especially the last description quoted re
veals, applied logic is in effect an empirical doctrine of error: it deals
with the obstacles to correct reasoning, the ways in which concrete
thinkers (may) go astray. Therefore, by looking at what Kant says
about error and its causes, we can learn more about applied logic.
The basics of Kant’s view of error can be found (among other pla
ces) in the Critique of Pure Reason. At the start of the Transcendental
Dialectic Kant explains that error can arise when the understanding in
teracts with other faculties (and especially with those of sensibility). Such
interaction may cause the understanding to deviate from its own laws:
In a cognition that thoroughly agrees with the laws of the understanding
there is no error. In a representation of sense (because it contains no judg
ment at all) there is also no error. No force of nature can of itself depart
from its own laws. Hence neither the understanding by itself (without
the influence of another cause), nor the senses by themselves, can err;
the first cannot, because while it acts merely according to its own laws,
its effect (the judgment) must necessarily agree with these laws. But the for
mal aspect of all truth consists in agreement with the laws of the under
standing. In the senses there is no judgment at all, neither a true one nor
a false one. Now because we have no other sources of cognition besides
these two, it follows that error is effected only through the unnoticed in
fluence of sensibility on understanding […].57
Hence, as applied logic deals with error, and as error is caused by the
illegitimate interaction of the understanding with other faculties of
the mind, applied logic has to deal with this interaction among mental
55 CPR A 53/B 77 (translation modified).

56 CPR A 54/B 79 (translation modified).
57 CPR A 293 – 294/B 350. Cf. Jsche Logic ( JL), AA 9:53. See also Refl. 2244, AA
16:283 – 284 (dated 1760s); Refl. 2250, AA 16:286 (dated 1770s); and
Refl. 2259, AA 16:288 (dated 1780s–1790s).
faculties.58 The distinction between pure and applied logic, therefore, is

a distinction between studying the laws of the understanding as it acts in
isolation and studying the laws of the understanding as it interacts with
sensibility, for that will teach us what kind of misleading influence sen
sibility may have on the understanding. This state of affairs explains as
well why applied logic needs recourse to empirical psychology: under
standing the way in which human beings can go wrong in concrete in
stances of reasoning must be based on empirical knowledge of the var
ious faculties of the mind and the way in which they interact that is, it
must be based on empirical psychology.
Indeed, in the Jsche Logic applied logic is characterized in exactly
these terms. In the first section of the introduction, which expounds
the concept of (pure general) logic, it is made clear that pure general
logic does not take recourse to psychological, contingent principles, be
cause “[i]n logic we do not want to know how the understanding is and
does think and how it has previously proceeded in thought, but rather
how it ought to proceed in thought”59 that is, pure general logic is not
a descriptive discipline, but one offering prescriptive rules. The second
section then turns to the divisions of logic (in the broad sense), and there
we find applied logic defined as follows:
In pure logic we separate the understanding from the other powers of the
mind and consider what it does by itself alone. Applied logic considers the
understanding insofar as it is mixed with the other powers of the mind,
which influence its actions and misdirect it, so that it does not proceed
in accordance with the laws which it quite well sees to be correct. Applied
logic really ought not to be called logic. It is a psychology in which we
consider how things customarily go on in our thought, not how they
ought to go on. In the end it admittedly says what one ought to do in
order to make correct use of the understanding under various subjective
obstacles and restrictions; and we can also learn from it what furthers the
correct use of the understanding, the means of aiding it, or the cures for
logical mistakes and errors. But propaedeutic it simply is not. For psychol
ogy, from which everything in applied logic must be taken, is part of the
philosophical sciences, to which logic ought to be the propaedeutic.60
This passage confirms my interpretation of applied logic and its relation
to psychology. In order to explain errors as made by actual human be
58 Of course, Kant distinguishes transcendental error as well, caused by transcenden

tal illusion. Applied logic does not deal with that, only with what one could call
empirical error.
59 JL, AA 9:14.
60 JL, AA 9:18.
124 Job Zinkstok
ings and to enable one to deal with them, we need a discipline that tells
us how the understanding interacts with other faculties of the human
mind, and how these interactions can make the understanding deviate
from its own laws. Such consideration presumably gives us rules, addi
tional to those of pure general logic, that teach us how to avoid or to
overcome the errors we are prone to or at least it shows what kind
of processes need care and attention in order not to run off course.
Given this characterization of applied logic, it is not surprising that
the passage in which Kant introduces applied logic in the Critique of Pure
Reason, indeed mentions a number of mental faculties that might inter
fere with the understanding: “the influence of the senses, […] the play
of imagination, the laws of memory”.61 Neither should it be a surprise
that these three faculties are all treated in Kant’s anthropology.62 There
fore, connecting this to the results of the previous section, it will be
clear that there is a close connection between applied logic and anthro
pology, at least to the part of anthropology that consists of empirical psy
chology. Because Kant relegates empirical psychology to anthropology,
it is in anthropology that we must seek the principles that applied logic
requires in order to fulfil its task. In order to substantiate this claim, I
will now turn to a concrete example of what applied logic would be.
4. An example of applied logic: prejudice
Although Kant thus had a rather clear view of applied logic, and repeat
edly indicates its place in the broader science of logic, he never actually
offered an explicit applied logic. However, we can reconstruct an exam
ple from his works, especially from his lectures on logic and his anthro
pology. I will reconstruct the case of prejudice. In the eighteenth century,
prejudice was commonly dealt with in logic,63 and Kant, in his lectures
on logic, is no exception. According to the Critique of Pure Reason treat
ment of “the influence of the senses, […] the play of imagination, the
laws of memory, the power of habit, inclination, etc., hence also from
61 CPR, A 53/B 77.

62 Anthr., AA 7:153 – 167 (the senses); 7:167 – 196 (the imagination); and 7:182 –
185 (memory, being a sub faculty of the imagination).
63 See for example §§ 1011 – 1016 of Wolff (1983 [1740], 729 – 733); §§ 168 – 172
of Meier’s Auszug aus der Vernunftlehre, AA 16:396 – 429.
the sources of prejudice” pertains to applied logic.64 Consequently, an

analysis of what prejudice is and how it is caused should provide us
with a concrete example of what applied logic is about and how it is
connected to anthropology.
Prejudices, according to the Jsche Logic, are “provisional judgments
insofar as they are accepted as principles”. 65 A provisional judgment, in turn,
is a judgment “in which I represent that while there are more grounds
for the truth of a thing than against it, these grounds still do not suffice
for a determining or definitive judgment, through which I simply decide
for the truth”.66 Hence the main problem of prejudices is that they
are taken as objective and determining instead of merely provisional;
they are caused by the mistake that “subjective grounds are falsely
held to be objective, due to a lack of reflection, which must precede all
judging”.67
It is thus quite clear that for Kant prejudice is a source of error: any
cognition built on principles that are provisional judgments, and that
thus rests on merely subjective grounds, is liable to error. To see
what Kant’s account of prejudice has to do with empirical psychology,
and thus why it is part of applied logic, we have to look into the sources
of prejudice after all it is these sources that Kant says are treated in ap
plied logic. Now the principal sources of prejudice, as the Jsche Logic
tells us, are “imitation, custom, and inclination [Nachahmung, Gewohnheit,
und Neigung]”.68 This is confirmed by the description of the contents
of applied logic in the Critique of Pure Reason, where Kant mentions
both custom and inclination and connects them with the sources of
prejudice (note the ‘hence’ in the quotation given above). Even though
prejudice itself is not treated explicitly in the Anthropology, we do find a
treatment of its sources. I will briefly go over them.
64 CPR A 53/B 77.

65 JL, AA 9:75.
66 JL, AA 9:74.
67 JL, AA 9:76. For similar accounts of prejudice, see the Logik an Wien (Wiener),
AA 24:863ff; Logik Philippi 3 (Philippi), AA 24:424 – 425; Refl. 2530 (dated
1775 – 1778), AA 16:406 – 407.
68 JL, AA 9:76 (translation modified; the Cambridge translation has “habit” for
“Gewohnheit”. Cf. Logik an Blomberg (Blomberg), AA 24:166; Logik Philippi 3
(Philippi), AA 24:425; Logik Pçlitz 3.1 (Pçlitz), AA 24:547 – 548; Refl. 2519
(dated 1760s), AA 16:403; Refl. 2550 (dated 1790s or early 1800s), AA 16:412.
126 Job Zinkstok
Inclination is defined in the Anthropology as “habitual sensible de

sire”.69 From this definition it is immediately clear that inclination ex
plicitly pertains to sensibility, more specifically to the sensible part of
the faculty of desire. If therefore inclination serves as a basis for a judg
ment, this judgment is subjective and cannot serve as a principle for ob
jective knowledge. If it is taken thus, however, it is a prejudice. This
source of prejudice, therefore, can be understood only by inquiring
into the way the faculty of desire interferes with the understanding.
Custom is also treated in the Anthropology, and it is defined as “sub
jective practical necessity […], and so designates a certain degree of will,
acquired through the frequently repeated use of one’s faculty”.70 In a
note from the 1780s Kant alternatively describes it as “the ease of exe
cution through frequent repetition.”71 He connects custom to the law of
association as it belongs to the reproductive imagination: “empirical ideas
[Vorstellungen] that have frequently followed one another produce a cus
tom in the mind such that when one idea is produced, the other also
comes into being”.72 This interpretation of the basis for the subjective
necessity of custom is confirmed by a note in Kant’s logic textbook,
where he writes that concepts belong to a consciousness “partly in ac
cordance with laws of the imagination, thus subjectively, or of the un
derstanding, i. e., objectively valid for every being that has understand
ing”.73 Consequently, custom is a subjective necessity, not an objective
one, based upon the workings of the reproductive imagination.74 Hence
if a judgment based on custom, i. e., based on subjective necessity deriv
ing from the laws of association, are taken for sound judgments express
ing objective necessity on which one may base other judgments, a pre
judice arises. In order to understand and explain the possibility of this
kind of prejudice, we need to understand the way the reproductive
imagination works and how it interferes with the understanding.
Imitation, finally, is the least clear cut case. In his anthropology Kant
treats it most conspicuously in the context of explaining genius, i. e., the
69 Anthr., AA 7:251; cf. AA 7:265.

70 Anthr., AA 7:147.
71 Refl. 1510 (dated 1780s), AA 15:829 (translation mine).
72 Anthr., AA 7:176 (translation modified). For the distinction between produc
tive and reproductive imagination, see CPR B 152.
73 Refl. 3051 (dated late 1770s to 1780s), AA 16:633.
74 Cf. Kant’s treatment of David Hume’s conception of custom, CPR B 127 –
128, A 760/B 788, and A 765/B 793.
“originality of the cognitive faculty”.75 Imitation is contrasted with gen

ius, as the latter is supposed to produce original things. Still, not all imi
tation is considered harmful for genius: it is indispensable in order to
learn “certain mechanical basic rules, namely rules concerning the ap
propriateness of the product to the underlying idea; that is, truth in
the presentation of the object that one is thinking of”.76 Here imitation
is connected to a treatment of imagination, something which is explicit in
a note where Kant writes: “Imagination […] is either productive or re
productive. The first belongs to genius, the second to imitation and
memory”.77 In this specific context, however, it is not clear what imi
tation has to do with prejudice. The most likely connection can be
found in Kant’s treatment of character. For Kant, the character of a per
son is closely related to authenticity it refers to traits that are character
istic or peculiar. In the 1770s Kant notes: “the spirit of imitation aims at
the singular, and is also not characteristic [eigenthmlich]”,78 and “Imita
tor: no character”.79 He is more specific on what character is in another
note dating from the 1780s:
For character in general it is required, first, that man has a will of his own,
that is neither imitated nor guided by someone else. Therefore he has to
decide on the basis of his own consideration what he decides, not on the
basis of fashionable rules of life. At the same time he has to be not as
weak as to want to please and comply with everybody […]. Second, he
must practice not to act through instinct and tempers or whims, but ac
cording to principles; in the same vein not according to custom. Third,
he has to keep word to himself, even if only in order to know that his in
tention will not be futile, that is to say perseverance or firmness of inten
tion: tenax propositi vir, not out of stubbornness, but out of principles.
Fourth, always keep word to others, so that he has character also in the
eyes of others. Hence choice of maxims.80
From these notes, I suggest, we can gather that for Kant imitation is a
form of distortion of one’s own character, leading one to act in ways
that are not based on proper, objective principles. Obviously, these
notes have a pragmatic slant they concern our behaviour towards oth
ers. Nonetheless, analogous considerations apply to the context of logic:
75 Anthr., AA 7:224. Cf. CPJ, AA 5:309 and 318 – 319.

76 Anthr., AA 7:225 (translation mine).
77 Refl. 1504 (dated 1780 – 1784), AA 15:805.
78 Refl. 761 (dated 1772 – 1773), AA 15:332 (translation mine).
79 Refl. 1495 (dated 1773 – 1778), AA 15:759 (translation mine).
80 Refl. 1517 (dated 1780s), AA 15:865 – 866 (translation mine).
128 Job Zinkstok
blindly adopting other people’s judgments and taking them as objective

grounds for other cognitions makes one liable to mistakes. Hence it is
not difficult to see that imitation can be a source of prejudice. However,
it remains unclear what psychological knowledge is needed in order to
explain imitation. It may be the case that the (reproductive) imagination
plays a role, as is hinted at in the passages on imitation and genius the
imagination may be (partly) responsible for the process of copying other
people’s judgments, or that of getting so used to those judgments that
one starts trusting them as if they are based on sound principles. Perhaps
it is character, or rather the lack thereof, which explains one’s suscept
ibility to imitation. Reading through the anthropology lectures and the
Reflections pertaining to them, I cannot find more conclusive leads. In
this case, therefore, the link between anthropology and empirical psy
chology remains weak.
Taking stock, the example of prejudice makes clearer what applied
logic is, and how it is connected to anthropology. As we saw, the sour
ces of prejudice all have to do with the interaction of the understanding
with other faculties, which lead one to take subjective grounds as if they
are objective. Understanding such errors, and being able to avoid them,
therefore requires knowledge concerning the various faculties of the
human mind. For this kind of knowledge applied logic has to draw
on the empirical psychology as it can be found in anthropology.
Conclusion
In this paper I have argued that interpretations of Kant’s anthropology
that solely stress its practical (moral or pragmatic) nature fail to take into
account another interesting side of this discipline, namely its use in
logic. I have argued, first, that a complete anthropology comprises em
pirical psychology, i. e., the empirical, descriptive study of the faculties
of the human mind. Although Kant is critical of a number of methodo
logical tools used in empirical psychology, and although he denies the
possibility of certain kinds of empirical psychology altogether, he none
theless does not want to get rid of it completely. In Kant’s view, empir
ical psychology, even though it is not a proper science, still enables us to
obtain systematically ordered knowledge of the mind that can be used in
various other disciplines. Among these disciplines are not only moral an
thropology (the application of moral principles in the case of human be
ings, including the specific obstacles and hindrances to moral behaviour
that derive from the specific mental constitution of human beings) and
pragmatic anthropology (the application of knowledge of human beings
in the service of prudential comportment in society), but also applied
logic: the logical discipline that deals with the interaction of the under
standing with the other faculties of the mind and the errors that might
originate from this interaction. Because of this connection to anthropol
ogy, one might, in analogy to moral anthropology, call applied logic as
well logical anthropology, or, as Kant calls it himself in a note in his text
book for his lectures on logic, anthropological logic: “anthropological logica
is applicata”. 81
Bibliography
Brandt, Reinhard (1994): Ausgewählte Probleme der Kantischen Anthropolo
gie, in: H. J. Schings (ed.): Der ganze Mensch. Anthropologie und Litera
tur im 18. Jahrhundert. DFG Symposion 1992, Stuttgart, Weimar, pp. 14 –
32.
Brandt, Reinhard, and Werner Stark (1997): Einleitung, in: R. Brandt and
W. Stark (eds.): Kants gesammelte Schriften, Vol. XXV, Göttingen, pp.
vii–cli.
Firla, Monika (1981): Untersuchungen zum Verhältnis von Anthropologie und
Moralphilosophie bei Kant, Frankfurt am Main.
Frierson, Patrick (2003): Freedom and Anthropology, in: Kant’s Moral Philos
ophy, Cambridge.
Kim, Soo Bae (1994): Die Entstehung der kantischen Anthropologie und ihre
Beziehung zur empirischen Psychologie der Wolffschen Schule, Frankfurt
am Main.
Longuenesse, Béatrice (2006): Kant on A Priori Concepts. The Metaphysical
Deduction of the Categories, in: P. Guyer (ed.): The Cambridge Compan
ion to Kant and Modern Philosophy, Cambridge, pp. 129 – 168.
Louden, Robert B. (2000): Kant’s Impure Ethics. From Rational Beings to
Human Beings, New York, Oxford.
Louden, Robert B. (2003): The Second Part of Morals, in: B. W. Jacobs and
P. P. Kain (eds.): Essays on Kant’s Anthropology, Cambridge, pp. 60 – 84.
MacFarlane, John (2000): What Does It Mean to Say that Logic Is Formal?,
Pittsburgh.
Mellin, Georg Samuel Albert (1797): Encyclopädisches Wörterbuch der kriti
schen Philosophie, Band 1, Züllichau, Leipzig.
Stark, Werner (2003): Historical Notes and Interpretative Questions about
Kant’s Lectures on Anthropology, in: B. W. Jacobs and P. P. Kain
(eds.): Essays on Kant’s Anthropology, Cambridge, pp. 15 – 37.
81 Refl. 3332, AA 16:783 (dated 1780s, perhaps late 1770s).

130 Job Zinkstok
Sturm, Thomas (2001): Kant on Empirical Psychology. How Not to Investigate

the Human Mind, in: E. Watkins (ed.): Kant and the Sciences, Oxford,
pp. 163 – 184.
Sturm, Thomas (2008): Why Did Kant Reject Physiological Explanations in
His Anthropology?, in: Studies in History and Philosophy of Science 39,
4, pp. 495 – 505.
Tolley, Clinton (2007): Kant’s Conception of Logic, Chicago.
Wolff, Christian (1963 [1728]): Preliminary Discourse on Philosophy in Gen
eral, transl. R. J. Blackwell, Indianapolis, New York.
Wolff, Christian (1983 [1728]): Philosophia rationalis sive logica, Gesammelte
Werke, Vol. II:1, ed. J. École, Hildesheim.
Wolff, Michael (1995): Die Vollständigkeit der kantischen Urteilstafel. Mit
einem Essay über Freges Begriffsschrift, Frankfurt am Main.
Kant’s Political Anthropology
Gnter Zçller
Abstract
The essay investigates the anthropological foundations of Kant’s political
thought. Section 1 argues for the mutually supplementary relation between
the critical theory of reason and the natural history of reason in Kant. Section
2 deals with the implied politics of Kant’s anthropology focusing on the relation
between nature and culture. Section 3 addresses the human social character, in
particular the dual process of the civilization and the moralization of human be
ings, in Kant. Section 4 presents the political vocation of the human being elu
cidating the paradoxical relation between good and evil and the role of civic
republicanism in Kant’s political anthropology.
“For from an animal that engages in

reasoning everything is be expected.”1
Introduction
Immanuel Kant’s stature as a political philosopher depends chiefly on his

late work on the moral conditions of political action, contained in On
the Common Saying: That May be Correct in Theory, But It Is of No Use
in Practice (1793), Toward Perpetual Peace (1795), “Metaphysical First
Principles of the Doctrine of Right” from The Metaphysics of Morals
(1797) and “Second Part. The Conflict of the Philosophy Faculty
with the Faculty of Law” from The Conflict of the Faculties (1798). In ad
dition, Kant’s foundational work in moral philosophy in general, pre
sented in Groundwork of the Metaphysics of Morals (1785) and Critique of
Practical Reason (1788), often has been appropriated for purposes of po
litical philosophy, recently so in the works of John Rawls and Jürgen
Habermas. Generally then, Kant’s contribution to political philosophy
has been taken to consist in a normative account of the ethical and
1 Reflexion 1521, AA 15/2:891.

132 Günter Zöller
legal principles that are to underlie moral politics. By contrast, little at

tention has been paid to the place and the role of political thought in
Kant’s extensive anthropological writings, which span almost his entire
career as an academic philosopher and which supplement the critical ac
count of the principles of theoretical and practical reason with an ac
count of the natural and cultural conditions of the development of rea
son in human practices during pre historical and historical times.
The following essay aims at redressing the one sided reception and
effective history of Kant’s political philosophy by gathering and present
ing the anthropological foundations of Kant’s political thought. The
term, “political anthropology,” which is not be found in Kant himself,
here serves to convey both the political dimension of Kant’s anthropo
logical thought in general and the anthropological basis of his political
thought in particular. Section 1 addresses the relation between the crit
ical Kant and the anthropological Kant and argues for the mutually sup
plementary relation between the critical theory of reason and the natural
history of reason in Kant’s philosophy. Section 2 deals with the implied
politics of Kant’s anthropology and focuses on the relation between na
ture and culture in Kant’s bio politically based anthropology. Section 3
turns to Kant’s explicit anthropology of politics concentrating on the
human social character in general and the dual process of the civilization
and the moralization of human beings in particular. Finally section 4
presents the political vocation of the human being according to Kant
by elucidating the paradoxical relation between good and evil and the
role of civic republicanism in his political anthropology.
The focal point of the essay throughout is Kant’s detailed analysis of
the human species in terms of naturally based and culturally developed
characteristics, which set human beings apart from other animals, while
maintaining their integration into the realm of nature. Politically perti
nent aspects of Kant’s naturalist account of cultural anthropogenesis ad
dressed in the essay include an anti essentialist understanding of the
human species character, the conception of a natural history of
human reason, a non theological understanding of good and evil, a
strictly secular perspective on the origin, development and end of
human history and the political goal of negative freedom in matters of
education, legislation and religion. Given the wealth of anthropological
material in Kant, the essay has to confine itself to the main outlines of
Kant’s natural and cultural anthropology in general and of his political
anthropology in particular.
Kant’s Political Anthropology 133
1. Geo-anthropology
On the standard account, Kant’s contribution to modern philosophy

rests on his project of a self critique of reason involving the assessment
of the origin, the extent and the boundaries of reason in its various, spe
cifically different modes of employment, chiefly in its theoretical use for
the determination of objects (knowing) and its practical use for the de
termination of the will (doing). In disciplinary terms, Kant’s work
therefore is located mainly in epistemology and moral philosophy,
more specifically in the elucidation of the non empirical (“a priori”)
principles of reason based knowing and doing. Given Kant’s particular
concern with “our,” human potential for rationality in theory and in
practice and his espousal of a mind based (“transcendentally idealist”)
account of knowing and a will based (“autonomous”) account of
doing, Kant’s twin theory of a priori knowledge and moral volition
draws on the constitutive role of the human mind in knowing and act
ing, thus lending his critical philosophy the character of a philosophical
psychology or a philosophy of mind.2
But Kant’s philosophical œuvre is not exhausted by his groundbreak
ing and multi disciplinary critical philosophy. Already before he under
went his critical turn in the 1770s, and from then on contemporaneous
with the development and elaboration of the critical philosophy, Kant
pursued wide ranging and far reaching philosophical interests that led
to numerous publications and several lecture series the popularity of
which stands in a marked contrast to the enormous intellectual demands
and the highly scholastic nature of the critical philosophy.3 While Kant
did not believe that the critical philosophy could ever become popular,
he always aimed a substantial part of his teaching and writing at the non
professional philosophical public. In fact, much of his philosophy in the
popular vein can be seen as an exoteric extension of his esoteric core
project of the critique of reason. Moreover, for Kant, popular philo
sophical work constituted an essential part of his philosophy in that it
afforded to expand philosophy from its narrow “scholastic concept”
2 On Kant’s transcendental theory of mind and its subsequent reception, see Zöl
ler (1993).
3 On the peculiar relation of Kant’s critical philosophy to the Enlightenment
concern with intensional and extensional popularity, see Zöller (2009).
134 Günter Zöller
(Schulbegriff) to its wider “cosmic concept” (Weltbegriff) 4 and to put the

results and achievements of the critical system of reason to use for prac
tical purposes that would be of interest to a wider audience or reader
ship.
In essence, Kant’s popular philosophical project is a sustained study
of the human being in his natural and cultural variety as well as unity.5
The main forum for the development and presentation of his popular
philosophy were Kant’s lectures in anthropology, which he held regu
larly every winter semester starting in 1772, after having obtained his
chaired professorship in logic and metaphysics in 1770, until retiring
from teaching at the University of Königsberg in 1796.6 In fact, Kant
was the first to devote a lecture course to this subject matter drawing
on the then still comparatively obscure term, “anthropology” (Anthropo-
logie), to designate a comprehensive treatment of the human being tran
scending existing or emerging disciplinary divisions and combining psy
chological, ethnological and socio cultural perspectives on the patterns
and courses of human existence.7 After finishing his teaching career
Kant issued a textbook that presented the cumulative results of his dec
ades long research and teaching on anthropological matters. The title of
the work neatly stressed the practical, worldly purpose of the enterprise:
Anthropology from a Pragmatic Point of View (Anthropologie in pragmatischer
Hinsicht, 1798).
In Kant’s own development as a teacher of philosophy the lectures
on anthropology were preceded by another long term teaching project,
which, in part, prepared the independent course in anthropology, viz.,
his regular lecture course on physical geography, which he had offered
since 1756 and continued to offer after starting the anthropology lec
tures, alternating between the course in anthropology during the winter
semester and the course in physical geography during the summer se
4 On the distinction between the two conceptions of philosophy, see Critique of

Pure Reason (CPR) A 838/B 866.
5 As a matter of historical accuracy and in order not to obscure Kant’s biased
usage of language, which addresses the natural and cultural conditions of
human beings primarily with reference to the male representatives of the spe
cies, the English word, “human being,” used to translate the German masculine
noun, “Mensch,” will be associated with the masculine gender in this essay.
6 For a representative selection of student transcripts from Kant’s anthropology
lectures, see AA: 15/1 and 15/2.
7 On the disciplinary origins of anthropology in Kant’s time, see Eckardt/John/
van Zantwijk/Ziche (2001).
mester of each of his years of academic service.8 While the main focus of
the lectures on physical geography lay on the natural properties and
conditions of the earth as a whole and of its geographically diverse
parts (continents, oceans, mountains, rivers, etc.), already the earlier of
Kant’s popular twin lecture set included sections on the human being
in general and on specific human populations considered in terms of
the geo physical conditions of their generic and characteristic proper
ties. The later lectures on anthropology continued the geographical in
terest in the role of natural ambient factors in the development and dif
ferentiation of the human species but concentrated on the individual
characters of the European peoples, leaving the treatment of the geo
graphically conditioned character of non Europeans to the lectures on
physical geography.9 In addition, the newly developed lecture course
in anthropology included ample material on the cognitive and appetitive
powers and the feeling ability of the human mind, for which Kant drew
on contemporary German academic philosophy (the discipline termed,
“empirical psychology,” in the metaphysical systems of the Wolff
school) and his own, emerging or established, critical philosophy.
Further evidence of Kant’s long term, in breadth as well as in depth
academic occupation with anthropological matters include: an early es
sayistic work on the anthropology of aesthetics, Observations on the Feel-
ing of the Beautiful and Sublime (1764); a tract on the taxonomy of mental
illnesses, Essay on the Maladies of the Head (1764); an article on the ana
tomical distinction between animals and humans, Review of Moscati’s
“Of the Corporeal Essential Difference Between Animals and Humans”
(1771); a trilogy of writings on the geographically influenced differen
tiation of the human species into relatively stable subspecies (“races”),
Of the Different Races of Human Beings (1775), Determination of the Concept
8 For a documentation of Kant’s lecture notes on physical geography, see AA 26/

1:1 – 320 (Manuscript Holstein); see also AA 9:151 – 436 (a compilation of
Kant’s lecture material on physical geography published under Kant’s name
in two volumes by D. F. Th. Rink in 1802). The publication of a representative
selection of student transcripts of Kant’s lectures on physical geography, to ap
pear in AA 26/2, is in advanced preparation (ed. Werner Stark).
9 Kant’s former student and eventual critic as well as rival, Johann Gottfried
Herder, retained the joint treatment of geographical and anthropological mat
ters in his multi volume unfinished main work, Ideas for the Philosophy of the His
tory of Humankind (1784 – 1791). On the historical and systematic relation be
tween Kant’s and Herder’s geo anthropological projects, see Zöller (2011)
and Zöller (2011a). For an alternative assessment of the relation between
Kant and Herder, see Zammito (2002).
136 Günter Zöller
of a Human Race (1785) and On the Use of Teleological Principles in Philos-

ophy (1788); as well as occasional reflections on the bodily influence of
the mind, On the Philosopher’s Medicine of Body (1786), and on the rela
tion between mind and brain, From Soemmerring’s “On the Organ of the
Soul” (1796). In addition, Kant’s anthropological thought has entered
into his contributions to the philosophy of history and pedagogy, in par
ticular Essays Regarding the Philanthropinum (1776/1777), Idea for a Uni-
versal History with a Cosmopolitan Aim (1784), Review of J. G. Herder’s
“Ideas for the Philosophy of the History of Humanity”, Parts 1 and 2
(1785), Conjectural Beginning of Human History (1786) and Lectures on
Pedagogy (1803).10
The nature and extent of Kant’s philosophical concern with anthro
pological matters is all the more surprising given the strictly non empir
ical, a priori character of the principles established by the critique of rea
son and the methodological distinctions drawn by Kant between the ob
jectively valid, apodictically certain rules of knowing and willing and
their contingent, fallible instantiations. Compared to the standards and
strictures of his “pure philosophy” (reine Philosophie) 11 Kant’s anthropo
logical interest may seem a sideline to his main work in critical philos
ophy and to have the function of a temporary and contingent relief from
the latter’s rigorous requirements of universality and necessity.
But rather than amounting to a lighter and more approachable ver
sion of Kant’s core philosophy, or even a mere diversion from it, his an-
thropological corpus provides an important and much needed supplemen
tation of the critical account of reason’s a priori principles. While the
critical philosophy purports to identify and legitimate the sum total of
a priori principles that govern knowing and willing as well as feeling,
the system of the critical philosophy largely leaves unaddressed the fac
tual origin, subsequent development and eventual actualization of the
potential for rationality in the human species. In focusing on the “ob
jective validity” (objektive Gltigkeit) of the rational principles governing
10 For a comprehensive edition of Kant’s published anthropological works, in

cluding the overtly historical and pedagogical publications, in modern English
translations with detailed introductions, notes and bibliographical information,
see Kant (2007). In what follows references to this edition that do not specify
the title of the cited work by Kant are to the Anthropology from a Pragmatic Point
of View.
11 CPR A 840/B 868 (contrast to “empirical philosophy” [empirische Philosophie])
and A 848/B 876 (contrast to “applied philosophy” [angewandte Philosophie]).
knowing, willing and feeling,12 Kant’s critical philosophy systematically

brackets the contingent, concrete circumstances and conditions of the
practice of reason. To be sure, this methodological restriction is not
to be counted as an intrinsic shortcoming of Kant’s critical project,
which could only succeed in establishing a finite body of universal prin
ciples governing the employment of reason at the expense of pursuing a
high level of abstractness and generality. Neither is it to be expected
from Kant’s anthropological works that they achieve the complete inte
gration of concrete, naturally determined and culturally diverse human
existence into the critical theory of reason.
Instead of correcting or improving upon actual or alleged shortcom
ings of the critical project, Kant’s work in anthropology counterbalances
the intentionally narrow methodological and systematical orientation of
the critique of reason, geared at a critical theory of a priori principles,
with a broadly conceived, empirically informed and practically geared
account of the human condition. The technically philosophical, critical
Kant and the popular, anthropological Kant share an interest in the
human being as rational, or at least: as capable of reason. In the critical
perspective, reason is taken objectively, with an eye toward the set of
principles that govern its various modes of employment. From the
standpoint of anthropology, reason is considered subjectively, in view
of the conditions that favor and shape the first emergence, subsequent
development and actual employment of reason. The former project ad
dresses the context of justification with regard to reason its normative
status independent of factual matters; the latter aims at the context of
discovery with regard to reason at its emergence and appearance
among actual human beings located in place and time, or rather: in ge
ography and history.13
In extensional terms, the difference between Kant’s critical project
and his anthropological project turns on the distinction between the
human as an individual being or as many, ideally all individuals of hu
mankind and the human being as a species (or a subset thereof) and as
a subject of natural and cultural development in its own right, concep
tually independent of the single or collective comportment of individual
12 On the central role of the axiological concept of validity in Kant’s theoretical

philosophy, see Zöller (1984).
13 On the analogy and disanalogy between space and time as factors in human de
velopment, see Reflexion 1404, AA 15/2:612 f.
138 Günter Zöller
human beings.14 The difference between the critique of reason’s essen

tially individualist conception of the human being and the species based
approach of Kant’s anthropology is especially striking when it comes to
the contrast between a critico rationally geared account of interhuman
relations centered around non empirical “laws of freedom” and an an
thropologically oriented account of human social life in terms of natural
regularities.15
But the supplementary relation between the critical and the anthro
pological project in Kant goes in both directions. Not only does the
concern with the natural and cultural particulars of human existence
materially enlarge the formal focus on the pure rules of reason. A min
imal normative conception of rationality in turn provides orientation in
the wealth of dimensions and details that characterize naturally and cul
turally particularized humanity. The guiding role of reason is all the
more needed given that in Kant the portrayal of the human condition,
while seemingly descriptive and empirical, is laden with normative or
normatively charged preconceptions and premises regarding human
pre historical and historical developments and therefore fraught with
prior notions and prejudices that, retrospectively, can be unmasked as
racism, sexism, classism and many other ways of turning distinction
into discrimination and difference into deficiency. Kant the anthropol
ogist certainly is not free of such prejudice. But the inner affinity and
structural convergence between his bottom up account of reason in
the anthropological project and his top down account of reason in
the critical project helps him and helps his readers differentiate be
tween alleged anthropological facts and their normative interpretation
and significance for an account of the flourishing of reason under nat
ural conditions.
14 See Reflexion 1467, AA 25/2:645.

15 The methodological and systematic distinction between the critical project and
the anthropological project in Kant maintained in this essay is intended as an
exegetical alternative to recent Anglo American work on Kant that seeks to re
dress the perceived shortcomings, one sidedness or otherwise defective struc
ture of his ethics (formalism, rigorism, universalism) by drawing on his anthro
pology for a materially specific, historically concrete and culturally inflected re
interpretation of Kant’s ethics. Rather than integrating Kant’s anthropology
into his ethics, the present essay maintains the distinction between critical phi
losophy and anthropology in Kant as two closely related but differently organ
ized and oriented philosophical projects. Not the merging of the two projects
but their joint consideration is able to provide the entire Kant.
2. The bio-politics of anthropology
At the conceptual and doctrinal center of Kant’s work in anthropology

stands the relation between nature and culture in the development of
human life. Yet Kant’s consistent focus is not on the opposition between
nature and culture but on their dynamic interrelation. More specifically,
Kant considers the human being as an animal thereby likening him to
other animals which operate under the guidance of instinct and yet as
an animal with the capacity to reason, a capacity that sets him apart from
the other animals by weakening the grip of the instincts on the human
being and thereby making him both more vulnerable and more danger
ous than other animals. Moreover, the animal endowed with reasoning
ability that the human being is makes him dangerous not only to other
animals, which he may hunt, tame and devour, but also to his own kind.
A chief concern of Kant’s anthropology throughout are the relations be
tween human beings in their twofold dependence on natural and cultur
al factors. Kant’s social anthropology (again a term not employed by
Kant himself) extends from the natural basis of societal life through var
ious processes of social cultivation to the overtly political organization of
human society under the guise of the state.
While the shapes and stages of the process of culturation are the
main focus of Kant’s anthropology, he devotes considerable attention
to the purely natural preconditions and manifestations of human life.
The concern with the natural development of the human being reflects
an interest on Kant’s part in the human species as an animal species sub
ject to naturally induced differentiation and resulting in natural diversity
even prior to the cultural processes of specification and individualiza
tion. Kant’s particular concern is with the natural unity of the human
animal species amidst its geographically induced differentiation into
populations that are marked and distinguished from each other by un
failingly hereditary traits, specifically epidermal pigmentation (“human
races”).16 But Kant’s detailed and repeated discussions of these matter,
which involved him in a protracted argument with the naturalist and
explorer, Georg Forster, over the original unity or plurality of the
human species monophyleticism being defended by Kant, polyphyle
ticism by Forster ,17 also served the methodological function of provid
ing Kant with a model from nature for understanding and assessing cul
16 On Kant’s work on the human races, see Lagier (2004).

17 On the dispute between Kant and Forster, see Riedel (1981).
140 Günter Zöller
tural development. When charting the cultural development of the

human species, Kant retains the naturalist terminology and conceptual
ity of pre given “predispositions” (Anlagen) and “germs” (Keime) and
that of “drives” (Triebe) along with their “development” (Entwicklung)
and selective actualization in response to ambient factors. This allows
him to account for the radical openness and extreme malleability of
the human being, while at the same time acknowledging the limited
range of actualizations afforded in each case by a given predisposition
that enables as much as pre structures and thereby channels human de
velopment.
But the natural and the cultural are not only conceptually opposed
and analogically connected in Kant’s account of anthropogenesis. For
Kant human cultural development is, to a large extent, itself natural,
part of an encompassing nature that includes the shaping influence of
natural means on the development of human culture. In a move that
continues older teleological thinking about the operations of nature,
Kant tends to personify nature, addressing “her” as an agent with prac
tical intelligence and in pursuit of her own designs. In English transla
tions of the German noun for “nature” (Natur) Kant’s personification
of nature is often rendered by capitalizing the word’s initial letter (“Na
ture”). While the personification of nature in Kant is more than a mere
figure of speech, it should not be taken literally, as if Kant were actually
attributing to nature intelligent design. Kant was weary of dogmatic tel
eology even before elaborating his critical account of the teleology of
nature in the Critique of the Power of Judgment. In Kant’s anthropological
thinking reference to nature’s ends serves the methodological purpose of
identifying developmental structures in nature, and specifically in living
nature, that span the modal difference between potentiality and actuality
and that present themselves with the directionality and orientedness
characteristic of human purposive activity. In addition, the teleological
perspective on nature and the human being in it allows Kant to address
the natural and cultural development in non descriptive or normative
terms that do not simply state or describe what is the case but that an
nounce and prescribe what is to be or ought to be the case, according to
some standard further to be specified.
The critically mitigated teleological perspective on living nature also
allows Kant to draw together the more narrowly natural, “animal,” as
pects of human life and those cultural processes that, while being exclu
sively characteristic of human life, operate prior to or independent of
conscious planning by individual human beings. To be sure, the pres
ence or absence of complete instinctual control radically distinguishes

mere animals or brutes from human animals or intelligent beasts. But
the loosened grip of the instincts on the human being is not per se tan
tamount to completely rational self control. Rather, on Kant’s anthro
pological account, the human freedom from instinctual determination
first and for the longest time, and for the most part even all the time,
goes together with alternative, not directly instinctual forms of natural
control and natural guidance. For Kant the anthropologist, the factual
freedom from the instincts opens the human being up to other, more
subtle ways of natural direction that typically elude the awareness of
the human being creating the impression or the illusion of free choice
in human conduct.
The key concept by means of which Kant marks the essential differ
ence between instinctually driven animals (brutes) and instinctually free
but, for the most part, alternatively driven animals (humans) is the no
tion of character. In Kant’s anthropological thinking “character” does
not have the narrow meaning of the basic mental constitution of an in
dividual human being. Rather the term serves to designate the ingrained
marks and traits of entire populations that reach in size and scope from
families through ethnic groups to all of humanity. Moreover, the term is
also employed by Kant to identify the basic behavioral set up of non ra
tional animals. Since the character of a given population of living beings
is an inner constitution that eludes direct inspection by others, a given
character has to be ascertained by external means, chiefly, especially in
the case of human beings, by observing outward traits and comport
ment. Kant here draws on the ancient epistemic practice of physiogno
my or the cognition of the inner character from external features, espe
cially facial features.18
Kant devotes the entire second half his published anthropology to
the “anthropological characteristic” (anthropologische Charakteristik) or
“the way of cognizing the interior of the human being from the exte
rior.”19 After first discussing in quite some detail the characteristic prop
erties of persons in terms of “natural aptitude” (Naturell) and “tempera
ment” (Temperament), Kant turns to the characteristics of the two sexes,
of different peoples and of the various races. In the final section of the
second part of his published anthropology, at the very end of the entire
work, he addresses the character of the species (Gattung), presenting
18 See Reflexion 1498, AA 15/2:774 – 777.

19 AA 7:283; Kant (2007, 383).
142 Günter Zöller
“basic traits of the depiction of the character of the human species.”20

The culmination of the published anthropology in a characteristic as
sessment of the entire human species is prepared by the prominent
final position of the characteristic of the human species in the student
transcripts of Kant’s anthropology lectures since the mid 1770s.21 In ad
dition, Kant’s literary remains (Nachlaß) in anthropology contain numer
ous entries on the species character of the human being, with further
related entries to be found in his literary remains in logic and the phi
losophy of right.22
Kant’s extensive reflections on the character of the human species
do not lead to the identification of a fixed and finite set of properties
that would uniquely characterize the human being. To begin with,
Kant voices skepticism about the very possibility of finding and defining
the human character as such, when no comparison group of other, non
earthly rational animal beings is available for ascertaining what might be
characteristic of finite rational beings in general and of human such be
ings in particular.23 In this situation Kant resorts to the contrastive com
parison of the human being, as the only rational animal being on earth,
with non rational animals. While every other animal species has its own
specific character, the human animal, on Kant’s account, does not pos
sess a species character that is set from the beginning and removed from
change and development. Rather the sole species character that Kant
sees fit to attribute to the human being is what could be termed a
meta character, viz., to acquire a character in the first place and to do
so by his own means and measures, although not without nature’s guid
ance.24
Kant’s minimalist and developmentalist conception of the human
species character makes his anthropology intellectually inhospitable to
20 AA 7:330; Kant (2007, 425) (translation modified).

21 See AA 25/1:675 – 728 (“On the character of humanity in general;” Anthroplo
gy Friedlnder; winter semester 1775/75); AA 25/2:838 – 847 (“Character of the
human species;” Anthropology Pillau; winter semester 1777/78); AA 25/
2:1194 – 1203 (“On the character of the entire human species;” Menschenkunde;
winter semester 1781/82); AA 25/2:1415) (“3rd chapter on the character of
the human species;” Anthropology Mrongovius; winter semester 1784/85).
22 See Reflexionen in anthropology, AA 15:602 – 654. See also Reflexionen in logic,
AA 16:170 – 813, and Reflexionen in the philosophy of law, AA 19:603 – 612.
For a compilation of the pertinent material from the Nachlaß, see Kant
(1985, 201 – 255).
23 See AA 7:321; Kant (2007, 416).
24 See AA 7:329; Kant (2007, 424).
the traditional question, “What is the human being?,” which Kant him
self avoids asking and answering in his anthropological writings. Also
outside of Kant’s anthropological works the question rarely occurs,
and if it does, then only in marginal texts and occasional contexts25
that would not justify attributing to him any particular interest in the
question as such or in answering it. Kant’s anthropological concern is
not with a fixed being or an essence of humans but with their open
ended developmental potential. Even the traditional locutions of the
“nature of the human being” or of “human nature” and the appeal to
it when judging what is possible and what is not possible for the
human species to achieve attract Kant’s criticism. For Kant, one cannot
know in advance what such nature might be and encompass with regard
to future developments.26
By effectively denying the human being a species character akin to
that of the other animals, Kant has opened up the human being for an
existence in history capable of undergoing long term change and espe
cially long term development by natural and cultural means. He also
has, in the exceptional case of the human species, broken with the tradi
tional, religiously based conception of the constancy of the animal spe
cies. But unlike Darwin’s theory of evolution over half a century later,
Kant’s revolutionary anthropology does not envision an “origin of spe
cies” and specifically the “descent of man,” but concerns the far reach
ing development of the human species from purely natural origins,
which are considered given and not subject to further derivation, to
the eventual world wide expansion of human cultural achievements.
Moreover, unlike Darwin, Kant does not countenance a change in
the physical constitution of human beings but a vast, various and volatile
cultural anthropogenesis.
Given the radical openness of the human species for future develop
ment and its complete lack of an initially fixed and subsequently con
stant character, the possible directions and the eventual outcome of
the development of the human species cannot be ascertained empirically
by referring to something given in past or present experience. Rather
the philosophical reconstruction of the overall course of human devel
opment has to take recourse to modes of thinking that exceed what is
empirically given in light of conceptions of reason (“ideas”), chiefly
25 See AA 11:429 (letter to C. F. Stäudlin from 4 May 1793) and AA 9:25 (Logic,
ed. G. B. Jäsche).
144 Günter Zöller
those of the ends or purposes attributable to the development of the

human species. The chief device drawn on by Kant to address the
non empirical purposive structure of human species development is
the contemporary, theologically based discourse of the “destination,”
“calling” or “vocation” of the human being (Bestimmung des Menschen),
which Kant transposes from the religious to the secular and from the es
chatological to the historical sphere.27 Morever, Kant regards the factual
or intentional pursuit of the human vocation as a process of perfection
ing, and of self perfectioning at that, during the course of which the
human species progresses toward a goal that is elevated and remote
without being elusive and excluded.
In a remarkable move that has deep repercussions beyond his an
thropological project Kant excludes the happiness of the human being
from the scope of his vocation, thereby relegating the natural and to
that extent unavoidable pursuit of happiness to a mere means for achiev
ing the genuine end of human development attributable to nature, viz.,
the maximal unfolding of his abilities or “talents.”28 The philosophical
point behind Kant’s anti eudaemonist anthropological vision is the ex
clusive linkage of the human vocation with activity, spontaneity or
“work” (Arbeit) rather than passivity, receptivity or “enjoyment”
(Genuß). To be sure, the vocation of the human being to the develop
ment of talents of all kinds does not remove happiness altogether from
the course of human development. Yet the achievement of happiness is
not itself the naturally pregiven goal of human history but an effect,
however welcome and even to be wished for, of pursuits that aim at
something else. In Kant’s natural teleology of the human species happi
ness is considered not as a factually pursued and eventually obtained goal
but normatively as an ideal state tied to desert and merit. A politically
sensitive consequence of Kant’s anhedonic natural teleology of human
existence is the exclusion of modes of human life that he considers to
be merely “passive” and oriented solely to sensory gratification
(“ease” [Gemchlichkeit], “good living” [Wohlleben], “happiness” [Glck-
seligkeit]).29 For Kant all such primitive forms of life lie at the margins of
cultural development and fall outside of the proper scope of human his
tory. In geo political terms, Kant’s discriminatory ethnology leads him
27 On the contemporary discourse of the “vocation of the human being,” see Zöl
ler (2001).
28 Reflexion 1521, AA 15/2:887.
29 See AA 7:325; Kant (2007, 420).
to restrict the effective cultural anthropogenesis to Europe and its cul

tural development from classical Greece onward.30
3. The anthropology of politics
Resorting to the teleological discourse of the vocation or destination of

the developmental life of animate beings, Kant presents a detailed ac
count of the structural difference between mere animals and humans be
ings. The common premise in describing the vocational course of mere
animals and of human animals is the teleological principle that all pre
dispositions in a living being are destined to be actualized. Nothing in
nature is in vain, and no predisposition goes to waste.31 Mere animals
achieve their vocation over the course of their individual lives, each sin
gle one for itself. Typically, they grow, mature and reproduce before
dying, thereby fulfilling their natural life cycle. Considered as animals,
human beings, too, typically reach their vocation over the course of
their life time. However, when considered as rational beings, human
beings, on Kant’s account, do not reach their vocation during the
span of their individual lives.32
The main reason for the asymmetry in the vocational success of
mere animals and rational animals is the latter’s lack of a fixed and finite
set of characteristic predispositions, which delineate the life course of
the mere animal. The freedom of human beings from complete instinc
tual control renders their vocation open and infinite. To reach the in
finite human vocation exceeds the possible accomplishment of any in
dividual human being and even that of any joined number of human
beings. It is only the human species, so Kant, that may perdure long
enough to eventually reach its vocation. Hence the proper subject of
an anthropology geared at determining the vocation of the human
being as such has to be an anthropology not of individual human beings
but of the human species. To be sure, the human species is not an entity
of its own, really distinct from any and all human beings and subject to a
different course of life than the human beings themselves. Rather it is
30 See Reflexion 1501, AA 15/2:788 – 790.

31 See AA 8:18; Kant (2007, 109) (Idea for a Universal History, First Proposition).
32 For the distinction between individual vocation and species vocation, see Re
flexion 1454, AA 15/2:635; Reflexion 1467, AA 15/2:645; AA 8:18; Kant
(2007, 109) (Idea for a Universal History, Second Proposition).
146 Günter Zöller
the culturally aggregated and historically accumulated result of the de

velopment of infinitely many individual human beings, none of
whom, except the very last ones,33 get to experience the eventual fru
ition of their fragmentary contributions to the development of the
human species. Moreover, Kant’s systematic references to the human
species as the secret subject of historical change and development do
not imply a guarantee that the complete human vocation in fact will
be achieved in a future however remote. Kant is not predicting the fu
ture course of human history but exposing the natural and cultural dy
namics underlying it and presenting the rational prospects for the suc
cessful approximation of the naturally given vocational end of the
human species.34
Yet while the members of the human species are endowed with un
characteristically multiple predispositions capable of many, indeed infin
itely many modes and manners of development, they also lack the im
mediate propulsion that motivates and orients instinctually driven mere
animals or brutes. The freedom from the instincts therefore threatens to
render human beings inert and indolent. In order to compensate for the
lack of immediately effective specific drives, nature, on Kant’s account,
has provided the human being with a surrogate of instinctual compul
sion that functions as a general predisposition for activating further pre
dispositions and germs that might otherwise remain dormant and, per
haps forever, inactive. The compensatory meta disposition of human
beings consists in their peculiar social character.
As animals that are not longer ruled by instinct alone, human beings
rely on their own volition to govern their conduct and seek to live ac
cording to their own will. This exercise of the will represents a minimal,
anthropological sense of freedom that may well go together with the
human being’s will being in turn determined by factors not subject to
volitional control. With respect to other human beings, the predisposi
tion to follow one’s own will leads, positively, to the imposition of one’s
will onto that of others and, negatively, to seeking to avoid having one’s
own will governed by that of others. The primary trait of the human
social character reveals itself to be asocial, viz., seeking to dominate oth
ers while simultaneously fleeing their dominion.

34 On the distinction between a factual description of humankind and its norma
tively conceived history, see Reflexion 1405, AA 25/2:613.
But this controlling evasive streak is only one half of the human so
cial character, as analyzed by Kant. As a result of the many needs that a
single human individual is not able to fulfill entirely on his own, human
beings not only flee each other, for fear of being dominated by another
human being, but also seek each other in the hope of gaining support,
gathering influence and increasing their own dominion. Famously Kant
terms the antagonistic social constitution of the human being that makes
him flee the company of his kind as much as seek it, his “unsocial soci
ability” (ungesellige Geselligkeit).35 For Kant the countervailing tendencies
in the human social character do not result easily in a stable equilibrium.
Rather the development of the human species is marked for Kant by the
conflicts arising from the contrary aspirations to asociality and sociality,
to an anti social self sufficiency that the human being cannot really af
ford and a social integration that the human being cannot really bear.
For Kant the diverse and multiple attempts at resolving the basic an
tagonism in the human social character shape the course of human de
velopment and especially of human history, which is to be regarded as
the arena for the experimental reconciliation of human asociality and
sociality. In the process of negotiating a precarious balance between
their conflicting social asocial orientations, human beings, on Kant’s ac
count, are driven into actualizing many a previously dormant predispo
sition and developing abilities they otherwise never would have known
to possess, much less brought into usage. According to Kant, the socially
provoked abilities are not restricted to technical skills in dealing with the
material world but essentially involve social skills for dealing with other
human beings, all of which operate under the secret desire to follow
only their own will, while seemingly cooperating with each other. In
Kant’s bleak picture of the origin and development of human social
life social mores are essentially marked by dissimulation, conceit and hy
pocrisy, not to mention the violent alternatives to the failures of these
measures at assuring domination and superiority.
But Kant’s sober and disillusioning portrayal of the genealogy of
morals in the human species is counterbalanced by his firm belief in
the human possibility and even in the human potential (“vocation”)
for societal peace on a large scale, however remote that end of history
may be and however arduous the path toward it may prove to be. Con
sidered in its entirety, the interplay between cunning, cheating and con
flict at the level of individual human beings serves the slow overall prog
35 AA 8:20; Kant (2007, 111) (Idea for a Universal History, Fourth Proposition).
148 Günter Zöller
ress at the level of the human species, so that the faults and failures of the
human species prove to be or may prove to be or will prove to be so
many indirect but indispensable naturally prepared means for achieving
the best possible human world.
Kant distinguishes three main stages in the naturally prepared and
culturally executed progressive development of the human species,
each based on a natural predisposition that finds it gradual and incre
mental actualization over the long and lingering course of human histo
ry. First comes the “technical predisposition” (technische Anlage), fol
lowed by the “pragmatic predisposition” (pragmatische Anlage) and finally
the “moral predisposition” (moralische Anlage).36 Kant correlates the three
predispositions, which he takes to be present in human beings at all
times and in all places, with three types of human development that
occur successively in human history and actualize the three simultane
ously present predispositions in serial form, beginning with the unfold
ing of the technical predisposition, then moving on to that of the prag
matic predisposition and finally turning to the realization of the moral
predisposition. In each case, the actualization of the respective predispo
sition is not a temporally and spatially fixed event but takes the form of a
lengthy and open historical process.
All three predispositions of the human species distinguished by Kant
concern ways in which the human being acts and does so in a manner
radically different from the acting abilities of non rational animals. The
technical predisposition of the human being consists in his ability of act
ing upon things by intentionally employing mechanical means. The re
maining two specifically human predispositions concern the ability of
the human being to interact with other human beings. The pragmatic
predisposition consists in his ability to employ other human beings for
his own purposes. The moral predisposition consists in his ability to
act upon himself as well as others according to non natural laws involv
ing freedom as a principle of acting.37 All three predispositions are
geared toward the successful and expansive use of reason, with the tech
nical predisposition providing mechanical skills for the efficacy of reason
and the pragmatic and moral predispositions furnishing social skills for
reason’s prudential and moral efficacy.38
36 AA 7:322 – 324; Kant (2007, 417 – 419).

37 See AA 7:322; Kant (2007, 417).
38 See AA 7:323 – 325; Kant (2007, 418 – 420).
Kant considers all three basic predispositions of the human being to

be natural predispositions and hence ways in which nature in him rather
than the individual human being for himself, by his own willing and
choosing, enables the engagement of his reason with things and with
other human beings. That the basis for the history of the use of reason
is entirely natural also holds for the moral predisposition. In an anthro
pological perspective the path to morality goes from inculcated practices
to moral principles and not vice versa, as in moral philosophy, from
moral principles to practiced morality.39
Kant distinguishes the specifically different gradual and incremental
processes through which the three basic human predispositions unfold as
the “cultivating” (cultiviren), the “civilizing” (civilisiren) and the “moral
izing” (moralisiren) that the human being undergoes as much as, uninten
tionally, brings upon himself.40 The notion of cultivation and particular
ly the cultivation of talents of all kinds derived from the old Latin
word the working of the soil (cultura) here does not have the generic
meaning of a contrast to nature but the specific sense of indicating the
development of technical skills that reach from the artisanal to the artis
tic and that involve the able operation of mechanical means for intelli
gently chosen ends of all kinds. Like all post natural development (“cul
ture” in the broadest sense) culture in the specific, technical sense in
volves activity, even “work” (Arbeit), while for Kant pre cultural,
merely natural human existence remains passive and geared toward “en
joyment” (Genuß).41
By contrast, the notion of the human being becoming civilized of
effectively, if not intentionally, civilizing himself involves not techni
cal, essentially mechanical processes but the social, essentially interhu
man transformation the nature guided self transformation of the
human being from the “natural state” (Naturzustand) to the “civil
state” (Civilzustand).42 This mode of culture (the concept being taken
in the widest sense) consists in substituting the crudity of mere “personal
force” (Selbstgewalt) through a “well mannered” (gesittet) conduct, even
if the latter is not yet ruled by principles that are genuinely “ethical” (sit-
39 See AA 7:328 f.; Kant (2007, 423).

40 See AA 7:324 f.; Kant (2007, 420).
41 On culture and enjoyment as alternative basic modes of human life, see Reflex
ion 1521, AA 15/2:889. On the significance of work, rather than intellectually
lazy intuitive vision, in philosophy, see AA 8:387 – 406 (Of a Recently Adopted
Elevated Tone in Philosophy).
42 Reflexion 1521, AA 15/2:889.
150 Günter Zöller
tlich).43 The transition to the civil state and the ability to maintain it re
quires “education” (Erziehung) under the twin shape of “instruction”
(Belehrung) and “discipline” (Zucht, Disciplin).44 In becoming a “citizen”
(Brger) or entering into a “civil constitution” (brgerliche Verfassung),45
the human being has entered into a social life that is essentially a shared
or common life, even though it is deeply shaped by the asocial procliv
ities of the human beings that enter into it and that makes them inclined
to distrust each other as much as rely on each other.
Considering that in Greek the linguistic equivalent of the Latin
based terms, “citizen” and “civil” (from civis and civilis, respectively),
are “politikos” and “polites,” the basic character of human development
under Kant’s second, pragmatic predisposition can be viewed as that
of his political development. More precisely, human development
under the pragmatic predisposition is political in a twofold sense: on
the basis of the pragmatic predisposition the political dimension of
human existence first arises, and the further unfolding of the pragmatic
predisposition involves the progressive development of the political
mode of human existence, from fairly elementary forms of communal
life to abstractly organized and efficiently administered statehood in its
various modes of governance and further on to the international associ
ation of individual states. In essence the unfolding of the pragmatic,
socio political predisposition of the human species makes up the entire
course of human history, including a long distance future that may, or
rather is to, include the political perfection of the human species. In the
meantime, though, human beings, on Kant’s account, can be considered
refined and polished but not really “civically minded” or “civilized”
(brgerlich gesinnet, civilisirt).46
Even less successful than the political progress toward true civility is,
in Kant’s eyes, the progress made so far toward the perfect actualization
of the third, moral predisposition. Kant diagnoses in history so far und
up to the present “morals” (Sitten) without “virtue” (Tugend), “sociable
ness” (Geselligkeit) instead of “righteousness” (Rechtschaffenheit) and
“vanity” (Eitelkeit) rather than “love of honor” (Ehrliebe), so that
human beings “on the whole” (im Ganzen), i. e., as a species, are “al
43 AA 7:323; Kant (2007, 418).

45 AA 7:327; Kant (2007, 422).
46 Reflexion 1524, AA 15/2:897.
most not all all […] moralized” (beynah gar nicht […] moralisirt).47 From
an anthropological viewpoint Kant is not concerned with moral prog
ress in the lives of individual human beings, which may be achieved
at any moment in time und under all circumstance due to the radical
freedom that the human being possesses as a consciously free agent en
dowed with the faculty of practical reason (“person”).48 Anthropologi
cally considered, morality derives from moralization or the lengthy for
mative process by which the socially camouflaged pursuit of one’s own
will gradually is superseded by genuine concern for the common good
and the latter’s pursuit for its own sake. For Kant the moral anthropo
genesis concerns not so much a novel set of ends to be set by a morally
predisposed agent as a motivational reorientation in the social life of
human beings from practical “solipsism”49 to the felt (“moral feeling”)
distinction between “right” (recht) and “wrong” (unrecht) in actions con
cerning the agent himself as well as others.50 For Kant the anthropologist
morality as a condition as well as an end of human practice is a socio
political matter belonging to the naturally based and artificially devel
oped culture of human coexistence.
4. The political vocation of the human being
The eminently political character of the anthropological account of

human developmental history, as presented by Kant in terms of the
three successively realized basic predispositions of the human species,
becomes apparent in his critical engagement with Rousseau’s political
philosophy of culture. Kant correlates each of the three specifically
human predispositions with a particular work and a systematic aspect
in Rousseau’s thinking about nature, culture and politics. In each case
Kant pairs his own progressist assessment of the development of the pre
disposition in question with Rousseau’s pessimist picture of human cul
tural development. In particular, Kant correlates his discussion of the
technical predisposition with Rousseau’s analysis of the physical and
mental “weakening” that the human species undergoes through cultural
progress, especially in the development of the arts and sciences, as de
47 Reflexion 1524, AA 15/2:897.

48 See AA 7:324; Kant (2007, 419).
49 Reflexion 1471, AA 15/2:649.
50 See AA 7:324; Kant (2007, 419) (translation modified).
152 Günter Zöller
tailed in the Discourse on the Arts and Sciences (1750). Kant’s analysis of the
pragmatic predisposition of the human species for the process of civili
zation is said to have its correlate in Rousseau’s discussion of the cultur
al political origin of inequality and mutual suppression among human
beings, as detailed in the Discourse on the Origin of Inequality (1755). Fi
nally, Kant confronts his treatment of the moral predisposition and the
progressive moral education of the human species with Rousseau’s por
trayal of “education contrary to nature and deformation of the mind
set,” as illustrated in Rousseau’s novel, Julie, or the New Heloise (1761).51
The point of Kant’s sustained parallelism between his own anthro
pology of human development and that of Rousseau is not simply to
contrast a positive and a negative account of the transition from nature
to culture. Rather Kant takes over substantial aspects of Rousseau’s cul
tural pessimism into his own account of human progress, just as he in
corporates elements of his own optimist general outlook on human his
tory into his interpretation of Rousseau. In particular, Kant expands on
the threefold pairing of the specifically human predispositions in his
own cultural anthropology and Rousseau’s three works in the critique
of culture with a second triad of writings by Rousseau which, according
to Kant, supplement the negative assessment of culture in the first triad
with Rousseau’s founding and envisioning of a counter culture destined
to overcome the essential shortcomings of failed arts and sciences, failed
politics and failed pedagogy. Kant mentions specifically Rousseau’s So-
cial Contract (1762), his Emile (1762) and the Profession of Faith of a Savo-
yard Vicar (from Book IV of the Emile) as works for which the three cor
related works in the negative critique of culture were meant to provide
the “guiding thread” (Leitfaden) for an alternative form of political,
pedagogical and moral culture.52
On Kant’s reading of the architectonic of Rousseau’s philosophy,
the latter’s overall strategy is not to advocate the return to the state of
nature but the regard for nature as a measuring stick for human cultural
development. By advocating “looking back” (zurck sehen) onto the state
of nature, rather than “going back” (zurck gehen) 53 to the state of nature,
Rousseau, on Kant’s revisionist reading, orients the human species to a
51 See AA 7:326; Kant (2007, 422). On the identification of the works alluded to
by Kant, see Kant (2007, 542 note 145). For a more detailed discussion, see
Brandt (1999, 326 f.).
52 See AA 7:327 f.; Kant (2007, 422).
53 AA 7:326); Kant (2007, 422) (translation modified; emphasis in the original).
possible future culture free from the ills of current cultural corruption
a final state of culture that retrieves under cultural conditions the pre
historical state of nature that was lost by the advent of culture.
The reintroduction of Rousseau as interpreted by Kant into Kant’s
own anthropology results in a threefold scheme of human development
according to which the state of nature is followed by the state of culture,
the long term development of which ultimately is to lead to a state in
which “perfect art again becomes nature” (vollkommene Kunst wird wieder
zur Natur).54 Drawing on Rousseau’s “three paradoxical propositions”
(drey paradoxe Stze) 55 about the harms caused by the apparent benefits
of scientific progress, a civil constitution and unnatural pedagogical
means and reverting Rousseau’s negative criticism of cultural develop
ment into the latter’s defense, Kant formulates what could be termed the
paradox of culture, according to which the inventions of culture prove
both objectionable, even reprehensible, when compared to the lost state
of nature, and functional, even beneficial, when considered in their in
direct preparatory role for the eventual restitution of nature under the
terms of culture. What in Rousseau might have seemed an attack on
culture citing its constitutive ills, is turned by Kant into an apology of
culture citing the benefits that come or are to come out of those very
ills. For Kant, under Rousseau’s influence, culture is both anti nature
and ante nature, the opposite of nature and the condition for its return.
The very evidence that leads Rousseau or rather, Rousseau as inter
preted by Kant to the indictment of culture makes Kant mount its de
fense.
The Kantian reading of Rousseau and the concomitant Rousseauian
inspiration of Kant’s anthropology also manifest themselves in moral
terms, when it comes to ascertaining the predispositional presence of
good or evil in the human species. Kant acknowledges the dual presence
of good and evil in the predisposition arguing that the “inborn propen
sity” (angeborener Hang) to the good constitutes the “intelligible character
of humanity in general” (intelligibeler Charakter der Menschheit berhaupt),
while the equally “inborn propensity to the evil” (angeborener Hang […]
zum Bçsen) constitutes the human being’s “sensible character” (sensibeler
Charakter).56 Kant argues that any contradiction between the opposed
basic inclinations falls away upon considering that the “natural voca
54 Reflexion 1454, AA 15/2:635, and Reflexion 1523, AA 15/2:896.

55 Reflexion 1521, AA 15/2:889.
154 Günter Zöller
tion” (Naturbestimmung) of the human being is to progress continuously

toward (moral) improvement.57
But Kant’s anthropological analysis of the human natural predispo
sition toward evil is not limited to his discussion of the moral predispo
sition. Kant widens the scope of the initially specifically moral treatment
of good and evil to an outright anthropology of good and evil that draws
on Rousseau’s association of naturalness with goodness and his condem
nation of the evils of culture’s break with nature, while preserving
Kant’s own overall assessment of human cultural development as pro
gressive and ameliorative. In particular, Kant’s maintains with regard
to the general course of human history the “production” (Hervorbrin-
gung) of the good from the evil, more precisely, of a good that is not
intended by the human being himself but that, once developed, pre
serves itself and that arises from the evil due to the fact that the latter
is “always internally at odds with itself” (innerlich mit sich selbst immer
sich veruneinigendes Bçses).58
In Kant’s analysis of culture the conception of the “origin of the
good from the evil” is tied closely to the reverse conception of the “ori
gin of the evil […] from the good.”59 In leaving the state of nature and
with it nature’s instinctual tutelage, the human being employs his newly
discovered ability to reason for freely pursuing his own well being in
ways that infringe upon other human beings and that give rise to all
kinds of “vice” (Laster) and “misery” (Elend). Yet due to the essential
instability of selfishly governed social life, the lapsarian passage from
the good to the evil at the beginning of human history, on Kant’s assess
ment, is to find its eventual inner worldly redemption in the inverse
origin of the good from the evil. By turning evil into an “incentive
for the good” (Triebfeder zum Guten) 60 and vindicating the Rousseauist
vilification of culture into culture’s paradoxical self overcoming, Kant
undertakes an anthropodicy, or a justification of the evils of human cul
ture in view of the good they secretly serve. While this move, inspired
57 On the distinction between the “natural vocation” (Naturbestimmung) of the

human being, which is culturo political, to be promoted by natural means
and to be fulfilled in the natural order, and the “rational vocation” (Vernunft
bestimmung) of the human being, which is ethico religious, to be promoted
by non natural means and to be fulfilled in the moral order, see Reflexion
1521, AA 25/2:885, 888.
58 AA 7:328; Kant (2007, 423).
59 Reflexion 1521, AA 15/2:891.
60 Reflexion 1501, AA 15/2:790.
by the earlier religious project of the justification of God in the face of

the evils of the world (theodicy), does not take away from the horrors
and the suffering in human history, it furnishes the philosophical reflec
tion on the nature and the course of history with a point of view that
integrates historical processes into a comprehensive structure of signifi
cance on a spatially and temporally comprehensive, anthropological
scale. Moreover, the anthropological prospect of inner worldly self re
demption of the human being, with the aid of nature, lacks the other
worldly perspective characteristic of Kant’s moral philosophy61 and its
extension into ethico theology and moral religion.62
In an anthropological perspective the radical reality of evil in the
human being is not a matter of a sinful fall and its long term moral con
sequences; nor is the restitution of the good an affair involving an indi
vidual human being’s act of inner moral revolution under divine assis
tance. Rather good and evil are features of the naturally induced and
naturally governed development of human culture. They are normative
predicates used to classify, at the anthropological level, the various forms
of interaction between nature, freedom and reason that determine the
course of human existence. In particular, animality combined with in
stinct, as characteristic of the state of nature under a Rousseauist descrip
tion, amounts to the good; so does freedom combined with reason, as
characteristic of the eventual perfectly civilized state of the human
being. By contrast, animality combined with freedom, as characteristic
of the imperfectly civilized state of the human being, amounts to the
evil, yet is ultimately productive of the good due to the eventual matu
ration of reason beyond its merely instrumental beginnings.63
For Kant the precarious position of the human being between the
good and the evil is a result of his complex constitution as, at once,
an “animal human being” (Tiermensch) and a “moral human being” (mo
ralischer Mensch).64 In the former regard as an animal human being ,
the human being is an “animal capable of reason (animal rationabile)” (ver-
nnftiges Thier) or a being capable of employing its reasoning ability in
the service of its animality and of the latter’s naturally selfish needs of
61 See AA 5:122 – 132; Kant (1999, 238 – 246) (Critique of Practical Reason).
62 See AA 6:18 – 53; Kant (1996, 69 – 97) (Religion Within the Boundaries of Mere
Reason, Part One. Concerning the Dwelling of the Evil Principle Alongside
the Good or Of the Radical Evil in Human Nature). See also AA 5:434 –
474 (Critique of the Power of Judgment, §§ 84 – 91).
64 Reflexion 1521, AA 15/2:888.
156 Günter Zöller
self preservation, self propagation and self enjoyment. On Kant’s assess

ment, the instinctually free but not yet rationally self determined human
animal instrumentalizes reason for its animal ends, thereby deforming
instinctually regulated natural pursuits into unnaturally liberated, vicious
practices.
In the latter regard as a moral human being , the human being is a
“rational being (animal rationale)” (Vernunftwesen) 65 capable of, conscious
of, called upon and conscious of being called upon to act on purely ra
tional grounds. Brought together the two halves of human existence do
not actually fit together until the human being himself, with nature’s
aid, has turned his freed animality enhanced by instrumental reason
into a freed rationality enhanced by good will. But unlike in his
moral philosophy, Kant’s concern in his anthropology is not with the
norms and form of such moral willing and acting but with the arduous
path of the human species through rationally enhanced animality toward
morally conditioned rationality. That path leads not through the heart of
the individual moral agent but through history and employs not the
inner constraint of conscience, moral respect or moral feeling but the
outer means of socially organized constraint. Accordingly, the domain
of socially mediated and historically manifest human self perfection is
the political.66
Under conditions of rationally enhanced animality, the political
project of the self perfection of the human species takes on the form
of human self governance. Kant’s political point of departure is the
twofold recognition that the human animal is in need of a “lord”
(Herr) able to curb the natural human tendency to socially harmful prac
tical egoism and that, in the absence of other earthly rational beings,
only the human being himself is available for the indispensable task of
social control.67 In Kant’s natural history of human social life, the begin
ning of human self governance is the tribal monarchical reign of one
human being over many others with the ensuing rivalries and outright
warfare both within the tribe and between several such tribal entities.
The political instability of tribal and intertribal strife leads, on Kant’s ac
count, to the transition from the “natural state” (Naturzustand) to the
65 AA 7:413; Kant (2007, 416 note); see also AA 7:321; Kant (2007, 416).
66 For an analysis that extends the specifically political character of Kant’s philos
ophy into his moral philosophy, in particular his late ethics in The Metaphysics of
Morals, see Zöller (2010) and Zöller (2010a).
67 See Reflexion 1500, AA 15/2:785 f.
“civil state” (brgerlicher Zustand), in which the governance is said to re

volve around “freedom” and “law” (Freiheit, Gesetz), supplemented by
“power” (Gewalt) to ensure the rule of law under conditions of free
dom.68
But first and foremost human self governance under “civil legisla
tion” (brgerliche Gesetzgebung) is imperfect and lacking the proper bal
ance of freedom, law and power. Kant presents a fourfold combinatorics
of the three constitutive elements of political governance, three of
which represent deficient political constellations with a different one
of the three constitutive elements missing in each of the three cases.
First, there is law and freedom without power, resulting in “anarchy”
(Anarchie). Then there is law and power without freedom, constituting
“despotism” (Despotism). Finally there is power and freedom without
law, amounting to “barbarism” (Barbarei). The complete combination
of all three constitutive political elements and with it the only “true”
civil constitution power together with freedom and law yields
what Kant terms a “republic” (Republik). The term here serves not to
indicate a particular organization of the body politic but “a state in gen
eral” (ein Staat berhaupt),69 or the core conception of the state as such,
in which the well being of the commonwealth is the highest political
principle. Kant stresses that the well being constituting the objective
of the supreme law of a republic is not the “sensual well being” (Sinnen-
wohl) or the “happiness” (Glckseligkeit) of its citizens but the “intellec
tual well being” (Verstandeswohl) or the preservation of the constitution
of the state.70
Kant’s political conception of a republican constitution of the state
draws on the Latin origin of the term, “republic” (res publica, meaning
“common cause”), in pre imperial, free Rome and places his political
anthropology into the context of modern republicanism, as advocated
by Machiavelli in the Discourses on Livy, Spinoza in the Tractatus politicus
and Rousseau in the Social Contract. 71 In Kant’s anthropological perspec
tive the freedom required for and realized in a true republic, where it is
conditioned by law and safeguarded by power, is essentially republican
68 Kant (1900, 7:330); Kant (2007, 425) (translation modified; in the original em
phasis).
69 Kant (1900, 7:330 f.); Kant (2007, 426) (translation modified).
70 Kant (1900, 7:331); Kant (2007, 426) (translation modified; in the original em
phasis).
71 On the Greek roots of modern republicanism in Plato and Aristotle, different
from its Roman roots in Cicero and Livy, see Nelson (2004).
158 Günter Zöller
freedom, not the minimal, formal and permissive freedom of individual

preference and choice but the substantial, contentually specific and po
litically obligating freedom of a civil society that is a society of citizens.
In addition, the normative identification of state and republic in the an
thropology serves as a politically significant supplement to the primarily
juridical conception of the state in Kant’s late philosophy of right.72
Given the radically asocial social propensities of human beings, the
republican political order has to exercise constraint for its instauration
and continued maintenance. But it is to do this only in such a way
that the constraint is mutual and according to law. Most importantly
yet, the laws limiting the constraint are to be laws that the human beings
have given themselves. Republican freedom involves “mutual con
straint under laws issuing from the human beings themselves” (wechsel-
seitiger Zwang unter von ihnen selbst ausgehenden Gesetzen).73 To judge
from the extant student transcripts, Kant was even more outspoken in
his lectures on anthropology than in the printed version on the egalitar
ian and even revolutionary implications of his political anthropology of
the perfect civil constitution envisioning a “society of equal beings”
(Gesellschaft gleicher Wesen) to be hoped for as a result of “many revolu
tions” (viele Revolutionen) that yet would have to occur.74 In the print
version the former final vision of the perfect polity is rendered as one
of a “cosmopolitan society” (weltbrgerliche Gesellschaft) encompassing
all human beings, but only with the status of a “regulative principle”
for the infinite approximation of an unreachable end.75
But the political freedom hat Kant envisions under a republican civil
constitution is not limited to the narrowly political freedom of self leg
islation and self governance. For Kant the improvement of the human
species in all three areas of social life under the basic predispositions to
self cultivation through (public) education, self civilization through
(state) legislation and self moralization through (moral) religion is
geared toward freedom from extraneous authority and the establishment
of self rule. In particular, Kant’s criticizes the predominant politico cul
72 See AA 6:229 – 372; Kant (1999, 386 – 506) (The Metaphysics of Morals, The
Doctrine of Right). On Kant’s republican conception of the state, see also
AA 8:349 – 353; Kant (1999, 322 – 325) (Toward Perpetual Peace). See also
Kant’s discussion of Plato’s republic (platonische Republik) as a practical idea in
the CPR A 316/B 372.
74 AA 25/1:690 f. (Anthropologie Friedlnder).
75 AA 7:331); Kant (2007, 427) (in the original emphasis).
tural condition of human beings in past and present under a comprehen

sive system of threefold “tutelage” (Unmndigkeit): the “domestic”
(huslich) tutelage in matters of education, which prevents human beings
from thinking for themselves even after their formal education is finish
ed; the “civil” (brgerlich) tutelage in political matters, which prevents
human beings from governing their own affairs; and the “pious”
(fromm) tutelage in religious matters, which keeps human beings from
following their own conscience and moral judgment.76
On Kant’s analysis, removing the extraneous constraints from the
three domains of possible human self improvement will result in an al
ternative cultural politics of “negative education”, “negative legislation”
and “negative religion.” In negative education the child’s talents, incli
nations, choices and morals are encouraged to develop free from the
strictures of the mere copying of conduct, artificial constraint, fixed ex
amples and corrupting enticements. In negative legislation the citizens
stand under laws issued from simple reason free of indoctrination and
obfuscation. In negative religion the concept of God is reduced to its
moral core freeing religion from the intercession of priests and an arcane
theology.77 Kant concedes that the called for “freedom of education,
civil freedom and religious freedom,” which in turn is to enable the
“general improvement” (allgemeine Verbesserung) of the human species,
is still a remote goal for a humankind not yet “susceptible” (susceptibel)
of such radical measures. But in regard to the natural history of the
human species from nature, with nature, through cultivation and civ
ilization to moralization , he is equally certain that in the long run
“nothing brought about forcefully [rather than freely] will last” (nichts
erzwungenes Bestand hat).78 Kant’s eminently political anthropology con
cludes and culminates with the vision of the human world this world
in the future rather than a future world characterized by the very no
tion that is also at the core of his moral philosophy, viz., freedom.
76 Reflexion 1524, AA 25/2:898 f.

77 See Reflexion 1524, AA 25/2:898 f.
78 Reflexion 1524, AA 15/2:898.
160 Günter Zöller
Bibliography
Brandt, Reinhard (1999): Kommentar zu Kants Anthropologie (1798), Ham
burg.
Eckardt, Georg, John, Matthias, van Zantwijk, Temilo, Ziche, Paul (eds.)
(2001): Anthropologie und empirische Psychologie um 1800. Ansätze
einer Entwicklung zur Wissenschaft, Cologne, Weimar, Vienna.
Kant, Immanuel (1900 f.): Gesammelte Werke, ed. Königlich Preußische (spät
er Deutsche) Akademie der Wissenschaften, Berlin, New York (Akade
mie Ausgabe; AA).
Kant, Immanuel (1985): Schriften zur Geschichtsphilosophie, ed. M. Riedel,
Stuttgart.
Kant, Immanuel (1996): Religion and Rational Theology, ed. A. W. Wood
and G. Di Giovanni, Cambridge.
Kant, Immanuel (1999): Practical Philosophy, ed. M. J. Gregor, Cambridge.
Kant, Immanuel (2007): Anthropology, History, and Education, ed. G. Zöller
and R. Louden, Cambridge.
Lagier, Raphaël (2004): Les races humaines selon Kant, Paris.
Nelson, Eric (2004): The Greek Tradition in Republican Thought, Cam
bridge.
Riedel, Manfred (1981): Historizismus und Kritizismus. Kants Streit mit G.
Forster und J. G. Herder, in: Kant Studien 72, pp. 41 – 57.
Zammito, John H. (2002): Kant, Herder, and the Birth of Anthropology, Chi
cago, London.
Zöller, Günter (1984): Theoretische Gegenstandsbeziehung bei Kant. Zur sys
tematischen Bedeutung der Termini “objektive Realität” und “objektive
Gültigkeit” in der “Kritik der reinen Vernunft”, Berlin, New York.
Zöller, Günter (1993): Main Developments in Recent Scholarship on the Cri
tique of Pure Reason, in: Philosophy and Phenomenological Research 52,
pp. 445 – 466.
Zöller, Günter (2001): Die Bestimmung der Bestimmung des Menschen bei
Mendelssohn und Kant, in: V. Gerhardt, R. P. Horstmann and R. Schu
macher (eds.): Kant und die Berliner Auflärung. Akten des 9. Internationa
len Kant Kongresses (26. bis 31. März 2000 in Berlin), Berlin, New York,
vol. 4, pp. 476 – 489.
Zöller, Günter (2009): Aufklärung über Aufklärung. Kants Konzeption des
selbständigen, öffentlichen und gemeinschaftlichen Gebrauchs der Ver
nunft, in: H. F. Klemme (ed.), Kant und die Zukunft der europäischen
Aufklärung, Berlin, New York, pp. 82 – 99.
Zöller, Günter (2010): Autocracy. Kant on the Psycho Politics of Self Rule,
in: S. R. Palmquist (ed.): Cultivating Personhood. Kant and Asian Philos
ophy, Berlin, New York, pp. 401 – 414.
Zöller, Günter (2010a): Autocracy. The Psycho Politics of Self Rule in Plato
and Kant, in: E. B. Pires, B. Nonnenmacher and S. Büttner von Stülpnagel
(eds.): Relations of the Self, Coimbra, pp. 385 – 404.
Zöller, Günter (2011): Genesis und Klima. Geo Anthropologie bei Herder und
Kant, in: S. Bacin, A. Ferrarin, C. La Rocca and M. Ruffing (eds.): Kant
und die Philosophie in weltbürgerlicher Absicht. Akten des XI. Internatio

nalen Kant Kongresses Pisa 2010/Proceedings of the XIth International
Kant Congress Pisa 2010, Berlin, New York.
Zöller, Günter (2011a): Mensch und Erde. Die geo anthropologische Parallel
aktion von Herder und Kant, in: M. Heinz and A. Nuzzo (eds.): “Meta
kritik.” Transformationen vorkritischer—Figurationen nachkantischer Phi
losophie, Stuttgart Bad Cannstatt.
List of Contributors
Andrew Stephenson, Laming Junior Research Fellow, Queen’s College,
University of Oxford
Thomas Sturm, Dr., Department of Philosophy, Autonomous Univer

sity of Barcelona
Liesbet Vanhaute, PhD Fellow of the Research Foundation Flanders,

Department of Philosophy, University of Antwerp
Matthias Wunsch, Dr., Department of Philosophy, University of Wup

pertal
Thomas Wyrwich, Dr., Department of Philosophy, University of Mu

nich
Job Zinkstok, Department of Philosophy, University of Groningen
Günter Zöller, Professor of Philosophy, Department of Philosophy,

University of Munich

Dietmar H. Heidemann (Editor) - Kant Yearbook 3-2011 - Anthropology-Walter de Gruyter (2011)

Uploaded by

Document Information

Original Title

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

Dietmar H. Heidemann (Editor) - Kant Yearbook 3-2011 - Anthropology-Walter de Gruyter (2011)

Uploaded by

Copyright:

Available Formats

Kant Yearbook

Online access for subscribers: http://dx.doi.org/10.1515/kantyb

ISBN 978-3-11-023653-8 (Print)

Bibliographic information published by the Deutsche Nationalbibliothek

” 2011 Walter de Gruyter GmbH & Co. KG, Berlin/Boston

List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Note to the Studi Kantiani . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Non veridical experience has been a central topic in epistemology and

1 See Crane (2006).

Hallucination is the experience as of an object when no object is pres

3 One notable exception is Brandt (1999).

Kant calls these representations sensations or intuitions, which for pres

The power of imagination (facultas imaginandi), as a faculty of intuition even

7 Cf. Metaphysics Mrongovius (AA 29:881), Metaphysics Volckmann (AA 28:449),

as of whatever object originally caused the relevant intuitions, but with

12 See e. g. CPR B 153 – 6.

Moreover, I take it that this reading of the judgemental counterpart

For present purposes I will treat illusions as cases in which an object is

Indeed I take this to be a definitional characteristic of concepts per se. It there

intuitively it would be correct to say that she is hallucinating. This kind

Having addressed the mechanisms involved in both canonical classes of

and its judgemental counterpart, on the other hand, constitutively in

trolled as experience. Rather the claim is that every judgement possibly

24 Contra e. g. Hanna (2010) and Allais (2010).

obvious that it should not count as a fundamental difference between

thing in between a spatiotemporally ordered manifold of sensations and

In the preface to his Anthropology From a Pragmatic Point of View, Kant

The project of a science of human nature was of great concern to eight

also struggled with the problem of freedom and determinism in an era

1. Freedom in Hume’s project of a “science of human nature”

possibility of doing otherwise, and of doing so under the very same

1.2. Two problems of Hume’s compatibilism

Hume’s argument is widely taken as a paradigmatic defense of compa

5 See Stroud (1977, 141 – 150); Penelhum (1993, 130 f.).

7 See also Stroud (1977, 146 – 149).

teract with other agents, ascribe responsibility to them, and so on.

2. Kant’s pragmatic anthropology as a science of free agency

That leaves us with Kant. As mentioned, he specifies the subject matter

2.1. Transcendental and practical freedom

9 Allison speaks of Kant’s “contra causal” or “incompatibilist” notion of freedom

the liberties of indifference and of spontaneity, namely the distinction

which, in his view, is crucial both for the empirical explanation of

2.2. Character in Kantian anthropology

In his anthropology lectures, Kant uses as textbook the chapter on em

acteristics” provide the key to the problem of how we can be studied as

2.3. Character as “mode of thought” (Denkungsart)

13 It is therefore difficult to see what Allison means by claiming “that there is no

2.4. Comparison with Hume’s views

a comprehensive human science, for which Hume’s work was often

planandum. I suspect part of the explanation has to do with the heavy

McIntyre, Jane L. (1990): Character: A Humean Account, in: History of Phi

Several Kant scholars1 have recently suggested that Kant’s Anthropology

perspective by showing that Kant does indeed make use of teleological

1. Kant’s pragmatic anthropology and teleological judgment

1.1. Kant’s theory of teleological judgment

Kant understands teleological judgment as a form of reflective judg

2 First Introduction to the Critique of Judgment (FI), AA 20:209.

Kant makes it expressly clear that external teleology is more prob

The concept of a human predisposition is an essential part of this ex

1.2. Teleology, empirical psychology and moral anthropology

Kant’s firm conviction that the teleological perspective is necessary to

13 Kant introduces the concept of a ‘predisposition’ in the context of biological

ogy cannot be an unsystematic collection of random facts.17 But can we