Content uploaded by Stuart Elden
Author content
All content in this area was uploaded by Stuart Elden on Feb 11, 2021
Content may be subject to copyright.
Reassessing Kant’s geography
Stuart Elden
Geography Department, Durham University, South Road, Durham DH1 3LE, UK
Abstract
This article offers a critical reassessment of Immanuel Kant’s lectures on Physische Geographie and his
contribution to geographical thought more generally. There are a number of reasons why this reassessment
is needed: the lectures are finally about to be published in English translation; careful philological work in
German has exposed how corrupted the standard text of the lectures is; and philosophers are finally begin-
ning to critically integrate an understanding of the Geography into their overall assessment of Kant’s work.
English speaking geographers will therefore soon have access to the lectures in a way that they have not
done before, but they need to be aware both of the problems of the edition being translated and the
work philosophers have undertaken on their situation in Kant’s work and their impact. More broadly,
the reassessment requires us to reconsider the position Kant occupies in the discipline of geography as
a whole. The article examines the history of the lectures and their publication in some detail; discusses
Kant’s purpose in giving them; and looks at the way in which he structured geographical knowledge
and understood its relation to history and philosophy. In terms of the broader focus particular attention
is given to the topics of race and space. While these lectures are undoubtedly of largely historical interest,
it is for precisely that reason that an examination of them and Kant’s thought more generally is of relevance
today to the history of the discipline of geography.
Ó2008 Elsevier Ltd. All rights reserved.
Keywords: Immanuel Kant; History of geography; Race; Space
Introduction
Immanuel Kant’s lectures on Physische Geographie were first published over two hundred years
ago. They were discussed by Richard Hartshorne in his comprehensive study of the discipline of
E-mail address: stuart.elden@durham.ac.uk
0305-7488/$ - see front matter Ó2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jhg.2008.06.001
Journal of Historical Geography 35 (2009) 3–25
www.elsevier.com/locate/jhg
geography almost seventy years ago, a survey that has formed the basis for most accounts of
Kant’s text since.
1
The most comprehensive study in English—J.A. May’s book—is almost forty
years old.
2
So why look again at Kant’s views on geography? There are four main reasons.
The first is that the lectures on geography are far more talked about than read. The first full
English translation of Kant’s Physical Geography is due to appear in the Cambridge Edition of
the Works of Immanuel Kant in the near future, in the volume on ‘Natural Science’.
3
Although
parts of the lectures have been available in English since the late 1960s,
4
this will undoubtedly
make these lectures far more widely available. What are these lectures about, and how might
we assess their claims today?
The second is that even this version of the lectures needs to be treated with caution. It is un-
fortunate that even after such a long wait the English audience is going to have to contend
with a problematic edition. Alongside a contextualisation of the lectures, they need a warning
about their status. The complicated lineage of the lectures will be discussed below, but for the mo-
ment it suffices to say that the text being translated is corrupted, and that it may well obscure parts
of Kant’s intent. What are these complications and how might this mislead?
The other two reasons are perhaps the most important. Third, these lectures and related writ-
ings are now beginning to receive critical attention in English language Kant scholarship. More
detail is provided below, but the basic claim is that just as Kant’s Anthropology from a Pragmatic
Point of View has been reconsidered over the past decade or so, the Geography deserves and is
beginning to receive similar work.
Fourth, just as in recent years scholarship has begun to recognise the complexity of Kant’s
ideas of some key topics, notably race, so too does his view of space need to be rethought. All-
too-often Kant is held up as an instance of views that he did not actually hold or was explicitly
1
R. Hartshorne, The nature of geography: a critical survey of current thought in the light of the past, Annals of the
Association of American Geographers 29 (1939) 173–412; 29 (1939) 413–658. This was reprinted in book form under
the same title (Lancaster, PA, 1939) to which all references are made. See also Hartshorne, The concept of geography
as a science of space, from Kant and Humboldt to Hettner, Annals of the Association of American Geographers 48
(1958) 97–108; Richards, Kant’s geography and mental maps, Transactions of the Institute of British Geographers 61
(1974) 1–16; and M. Bowen, Empiricism and Geographical Thought: from Francis Bacon to Alexander von Humboldt, Cam-
bridge, 1981.
2
J.A. May, Kant’s Concept of Geography and its Relation to Recent Geographical Thought, Toronto, 1970. One of the
rare exceptions among contemporary geographers has been David N. Livingstone, who discusses Kant in the context of
his impact on the discipline of geography, in: The Geographical Tradition: Episodes in the History of a Contested Enter-
prise, Oxford, 1992, 113–117; and in D.N. Livingstone and R.T. Harrison, Immanuel Kant, subjectivism and human
geography: A preliminary investigation, Transactions of the Institute of British Geographers 6 (1981) 359–374. The other
exception is David Harvey, whose reading is more extensively discussed below. Kant’s work only receives the briefest of
mentions in physical geography surveys. See, for example, M.J.S. Rudwick, Bursting the Limits of Time: the Reconsti-
tution of Geohistory in the Age of Revolution, Chicago, 2005; and R. Inkpen, Science, Philosophy and Physical Geogra-
phy, London, 2005. The treatment is little better in G.J. Martin and P.E. James, All Possible Worlds: A History of
Geographical Ideas, New York, 3rd ed, 1993, 109–111.
3
I. Kant, Natural Science, edited by H.B. Nisbet and E. Watkins, Cambridge, forthcoming. The translation of the
Physische Geographie is by Olaf Reinhardt, who has kindly sent me an advance copy of the text.
4
R.L. Bolin, Immanuel Kant’s physical geography, MA Thesis, University of Indiana, 1968. The first six sections of
the introduction also appear in May, Kant’s Concept of Geography (note 2), 255–264.
4S. Elden / Journal of Historical Geography 35 (2009) 3–25
critiquing. Moving from Kant’s Geography to Kant’s geography more generally, we need to con-
sider his position in the tradition more carefully.
Back in 1939, Hartshorne noted Erich Adickes’s assessment of the lectures and agreed that ‘for
geography today.Kant’s work is of little more than historical interest’.
5
This paper does not
seek to suggest that reading Kant today will radicalize the discipline, and this is not an attempt
to rehabilitate Kant. Rather it suggests that reassessing Kant might be a useful moment in
a broader reconsideration of the history of geography, and certainly of the place of Kant within
it. Thus this interpretation takes issue with the ‘little more’ element of the Adickes/Hartshorne
assessment, suggesting that its historical interest is precisely why we should look at it today. As
Livingstone suggests, ‘the history of geography is the history of a contested tradition’.
6
Given
that this is true of the history of ideas more generally, and philosophy especially, their point of
collision in Kant is a significant marker. As philosophers begin the work of integrating these ma-
terials into their overall assessment of Kant’s work, geographers too should reconsider his views
in light of the history of the discipline. Yet rather than do as Hartshorne himself suggests, and
privilege Kant’s other writings over the Physical Geography, which he suggests should be relegated
to a ‘historical footnote’, this reading begins with the lectures.
7
What do these lectures, and fol-
lowing them Kant’s writings on issues of concern to geographers more generally, add to our
understanding of Kant’s geography today?
Immanuel Kant’s geography lectures
Kant lectured on a variety of topics during his career at the University of Ko
¨nigsberg. These
included both philosophical and non-philosophical topics. The lecture courses were often well at-
tended and gained a strong reputation. Many of these courses have come down to us from Kant’s
own manuscripts or from student transcripts. Today these lectures are part of his complete works,
and are invaluable sources of knowledge concerning his work, its substance, coherence and devel-
opment. Standard philosophical subjects such logic, metaphysics and ethics were part of his nor-
mal program, but Kant also gave courses on anthropology and physical geography on a regular
basis. Initially it was geography alone, but from the mid 1770s geography was usually offered in
the summer semester while anthropology was given in the winter. In total geography was offered
forty-nine times over a forty-year period from 1756 to 1796—more frequently than any of his
other topics other than logic and metaphysics.
8
5
Hartshorne, The nature of geography (note 1), 39.
6
Livingstone, The Geographical Tradition (note 2), 101.
7
Hartshorne, The nature of geography (note 1), 44; endorsed by Livingstone, The Geographical Tradition (note 2),
115. C. Glacken, in: Traces on the Rhodian Shore: Nature and Culture in Western Thought from Ancient Times to the End
of the Eighteenth Century, Berkeley, 1967, 530–535, similarly shifts the focus to other texts, in his case largely the be-
ginning of the second part of the Critique of the Power of Judgment, edited and translated by P. Guyer, Cambridge,
2000.
8
Though these figures differ in various accounts, I have taken them from the archival work of Werner Stark (http://
www.staff.uni-marburg.de/wstark/geograph/geo_start.htm) and Steve Naragon (http://users.manchester.edu/Facstaff/
SSNaragon/Kant/Notes/notesGeography.htm).
5S. Elden / Journal of Historical Geography 35 (2009) 3–25
Yet Kant never produced a book from these lectures. The course on anthropology was worked
up into a book by Kant himself, entitled Anthropology from a Pragmatic Point of View.
9
The ini-
tial print-run was for the largest number of copies any book of his produced in his lifetime.
10
In
Kant’s words the Anthropology was ‘the present [i.e. the final] manual for my anthropology
course’.
11
It seems that had Kant more time, he would have produced a volume of the geography
lectures himself. But he remarked in 1798 that a version of them was ‘scarcely possible’ at his own
advanced age, for the manuscript he used to lecture was one he believed only he could read.
12
Shortly after this comment, the first volume of an edition by Gottfried Vollmer was produced,
apparently based on transcripts of lectures from 1778, 1782 and 1793, but this was described
by Kant as unauthorized and illegitimate, of which he did not recognise ‘either the material or
the format’ as his.
13
This is hardly surprising: much of the work is not Kant’s and this is not
made clear in the edition itself; and there is certainly far more material than could ever have
been given in a single semester. The semi-authorized version which was then rushed out in
1802, two years before Kant’s death, was compiled by Friedrich Theodor Rink, utilizing lectures
from 1759, with the introduction from a mid 1770s course, and is much less extensive.
14
Rink
claims that he has based this on three notebooks of Kant’s, but he unquestionably means student
transcripts.
15
Adickes thinks the first part is from 1775, and demonstrates that the text is unreli-
able in that Rink added notes, and altered or dropped passages of which he could not make
sense.
16
Werner Stark has noted that the first half (which he dates to 1774) is much more closely
9
It is important to note that though the anthropology lectures were edited by Kant himself, they too exist in different
forms. Indeed, KGS XXV is devoted to variant drafts; selections from which are forthcoming in the Cambridge Edition
as Lectures on Anthropology, edited by A.W. Wood and R.B. Louden, Cambridge. Kant’s also wrote a large number of
reflections on anthropology, collected in KGS XV (2 parts). For a discussion see W. Stark, Historical notes and inter-
pretative questions, in: B. Jacobs and P. Kain (Eds), Essays on Kant’s Anthropology, Cambridge, 2003, 15–37.
10
M. Kuehn, Introduction, in: I. Kant, Anthropology from a Pragmatic Point of View, edited and translated by R.B.
Louden, Cambridge, 2006, x, citing F.W. Schubert, Immanuel Kants Biographie zum grossen Theil nach handschriftlichen
Nachrichten dargestellt, Leipzig, 1842, 154.
11
Kant, Anthropology from a Pragmatic Point of View (note 10), 6 n.
12
Kant, Anthropology from a Pragmatic Point of View (note 10), 6 n.
13
G. Vollmer (Ed), Immanuel Kants Physische Geographie, Mainz, Four Volumes in Seven Parts, 1801–1805. The
transcripts Vollmer used are lost. For Kant’s repudiation of this edition, see Nachricht, die den Vollmer erschienene
unrechtma
¨ßige Ausgabe der physische Geographie betreffend, 29th May 1801, KGS XII, 372, and E. Adickes, German
Kantian Bibliography, New York, 1970 [1893–1896], 27. There is some dispute about the validity of this refutation, given
Kant’s advanced age and senility when it came out. I owe this point to Robert Bernasconi.
14
F.T. Rink (Ed), Immanuel Kant’s Physische Geographie,Ko
¨nigsberg, 1802, here cited by the KGS edition. On the
provenance of the 1802 edition see Friedrich Theodor Rink, Vorrede des Herausgebers, in KGS IX, 153–155; Richards,
Kant’s geography and mental maps (note 1); and especially W. Stark, Immanuel Kants physische Geographie – eine Her-
ausforderung?, 4th May 2001, http://www.staff.uni-marburg.de/wstark/ws_lese4.htm.
15
Rink, Vorrede des Herausgebers (note 14), 155.
16
E. Adickes, Untersuchungen zu Kants physischer Geographie,Tu
¨bingen, 1911, 278; Zu O. Scho
¨ndo
¨rffesrs Bemerkun-
gen u
¨ber Kants physische Geographie, Altpreußische Monatsschrift 56 (1919) 55–71, 59. P.D. Fenves describes Adick-
es’s Untersuchungen as ‘one of the great monuments of Kant philology’ (Late Kant: Towards Another Law of the Earth,
New York, 2003, 202 n. 10). May suggests the first half dates from 1772 or 73, and claims that it represents Kant’s
mature thought (Kant’s Concept of Geography (note 2), 72, 75).
6S. Elden / Journal of Historical Geography 35 (2009) 3–25
edited, with additions and references, while the second half reproduces Kant’s text almost verba-
tim, even allowing a gap in the text (the omission of India and East Asian islands in the survey) to
remain.
17
Rink alludes to some of the reasons for this uneven treatment in his introduction to the
lectures.
18
His promised revised edition, putting good the problems, never actually appeared.
Vollmer’s edition came out in an eventual four volumes, in seven parts in total, and was com-
pleted in 1805.
It is the Rink edition that is today known as the Physische Geographie, later incorporated into
the Akademie Ausgabe of Kants gesammelte Schriften [the Academy Edition of Kant’s Collected
Writings].
19
This is the one that Hartshorne utilizes, and that almost all accounts in English refer
to, either first or second hand. Yet the dates of the materials used should give us cause to pause.
Kant gave lectures on this topic from 1756 until 1796, and his comment that only he could read
the text in 1798 indicates that he never ceased to revise it over that period. The text we have takes
the introductory materials from a mid-period course (1774 or 1775) and bolts them onto the body
of very early lectures (1759). At the time of the early lectures, anthropology was not yet being of-
fered as a separate course, and some of the materials for that came from the geography lectures.
By way of context, Kant’s mature philosophy is said to begin with the Critique of Pure Reason,
first published in 1781.
20
Nobody seeking to understand Kant’s considered views on morality, on-
tology, logic, physics, or, indeed, anthropology, would content themselves with ‘pre-critical’
works.
21
Yet that is what is usually done with the lectures on geography.
In the early twentieth century Adickes attempted to get the Akademie Ausgabe to produce a new
version of the Geography, instead of relying on Rink’s, but this was declined due to feasibility: the vol-
ume was already typeset. The forthcoming Volume XXVI of Akademie edition will include several
transcriptions of the lectures, prepared by the historian of philosophy Werner Stark;
22
and Volume
17
Stark, Immanuel Kants physische Geographie (note 14). An unpublished analysis by Stark compares the Rink edited
introduction with the manuscript, showing line by line just how much interpretation was involved.
18
Rink, Vorrede des Herausgebers (note 14).
19
It also appeared in other editions, notably I. Kant, Sammtliche Werke Sechster Theil: Schriften zur Physischen Geo-
graphie, edited by K. Rosenkranz and F.W. Schubert, Leipzig, 1839. This volume also includes Kant’s writings on
related topics, and an appendix of variants from Kant’s handwritten materials, 779–805.
20
C. Browne Garnett Jr., The Kantian Philosophy of Space, New York, 1939, 11 (see 119) claims that Kant ‘formu-
lated the main tenets of his philosophy of space’ in 1769, a belief supported by A. Melnick, Space, Time, and Thought in
Kant, Dordrecht, 1989, 25, and J.H. Zammito, Kant, Herder, and the Birth of Anthropology, Chicago, 2002, 264. The
reference is to the developed views found in the Inaugural Dissertation On the Form and Principles of the Sensible and
Intelligible World, in: Theoretical Philosophy 1755–1770, edited and translated by David Walford, Cambridge, 1992,
especially 395–398.
21
F.K. Schaefer, Exceptionalism in geography: a methodological examination, Annals of the Association of American
Geographers 43 (1953) 226–249, 233–234 makes a similar point, although based on the view that the introduction to the
Geography dates from 1756. This is of course not to denigrate valuable work that takes the ‘pre-critical’ philosophy as
its object of study. On this dating in Kant’s work and its relation to the anthropology, see Zammito, Kant, Herder, and
the Birth of Anthropology (note 20), 255–259.
22
This will include the Holstein-Beck manuscript from 1758–1759, several from the 1770–1792 period, as well as in-
troductions from several other years and an appendix of J.G. Herder’s notes from Kant’s courses on logic, moral phi-
losophy, metaphysics and physical geography. See W. Stark, Immanuel Kant, Physische Geographie: Kurze
Information zur Neuausgabe, unpublished paper; and see http://www.staff.uni-marburg.de/wstark/geograph/
geo_edi0.htm.
7S. Elden / Journal of Historical Geography 35 (2009) 3–25
XIV includes notes on geography, along with mathematics, physics and chemistry.
23
Yet unfortu-
nately the forthcoming English translation is of the Rink edition. This is certainly better than nothing
but it is interestingagain to contrast this with the Anthropology, which has been available in English for
several years in a reliable edition, with translations in 1974 and 1978 and an entirely new recent trans-
lation by Robert Louden with extensive notes from Kant’s manuscript and a long introduction;
24
along with forthcoming editions of variant lectures and note materials.
25
There are thus serious philological difficulties relating to reading the geography lectures that go
far beyond their inaccessibility in English. Stark has done invaluable work in ordering and orga-
nizing these lectures, following his work on the lectures on ethics and anthropology, setting them
in their historical context and showing the variation across time.
26
His transcriptions of the hand-
written student transcripts are a major work in themselves. For many purposes the most accessible
edition of the lectures is the French, which although it is a translation of the Rink edition includes
an extensive introduction and a brief appendix of variants from other years.
27
This is the text used
by David Harvey, whose reading is discussed below.
The neglect
Of all of Kant’s work, and of all his wide areas of interest, the neglect of geography is perhaps the
most glaring. It generally merits an entry in dictionaries of Kant’s work, but these tend to be pretty
brief.
28
There is no explicit discussion in the recent A Companion to Kant,
29
and a recent edited book
on Kant and the Sciences makes only a tangential reference to geography.
30
Robert Hanna’s compre-
hensive study, Kant, Science, and Human Nature, makes only two passing references to geography
and offers no sustained engagement.
31
Many other accounts on related topics are similar.
32
The
23
See I. Kant, Notes and Fragments, edited by P. Guyer, translated by C. Bowman, P. Guyer, and F. Rauscher, Cam-
bridge, 2005, 547 n. 34. Unfortunately this volume does not include any of the notes on geography, which appear in
KGS XIV, 541–635. A detailed analysis is provided in the work of Adickes, Untersuchungen zu Kants physischer Geo-
graphie (note 16); and Ein neu aufegefundenes Kollegheft nach Kants Vorlesing u
¨ber physische Geographie,Tu
¨bingen,
1913. See Stark, Immanuel Kants physische Geographie (note 14).
24
The Louden translation of Anthropology from a Pragmatic Point of View is cited above (note 10) (n. 8); the earlier
translations are by M.J. Gregor, The Hague, 1974; and by V.L. Dowdell, Carbondale, 1978. References to this work are
to the Louden translation.
25
I. Kant, Anthropology, History and Education, edited by R.B. Louden and G. Zoller, Cambridge, 2007; Lectures on
Anthropology (note 9).
26
See Stark, Immanuel Kants physische geographie (note 14); and material on his website.
27
I. Kant, Ge
´ographie, translated by M. Cohen-Halimi, M. Marcuzzi and V. Seroussi, Paris, 1999. The appendix is
taken from Kant, Sa
¨mmtliche Werke (note 19). See also M. Marcuzzi, L’e
´criture de l’espace: Re
´cit historique et descrip-
tion ge
´ographique dans la Ge
´ographie Physique de Kant, in: J. Benoist and F. Merlini (Eds), Historicite
´et spatialite
´:Le
proble
`me de l’espace dans la pense
´e contemporaine, Paris, 2001, 118–139.
28
See H. Caygill, A Kant Dictionary, Oxford, 1995, 214–215; H. Holzhey and V. Mudroch, Historical Dictionary of
Kant and Kantianism, Lanham, 2002, 131–132.
29
G. Bird (Ed), A Companion to Kant, Oxford, 2006.
30
E. Watkins (Ed), Kant and the Sciences, Oxford, 2001.
31
R. Hanna, Kant, Science, and Human Nature, Oxford, 2006, 56–57, 268–270.
32
Other examples of the neglect are given by C.W.J. Withers, Kant’s geography in comparative perspective, in:
S. Elden and E. Mendieta (Eds), Reading Kant’s Geography, Albany, forthcoming 2009.
8S. Elden / Journal of Historical Geography 35 (2009) 3–25
same neglect can be found in works concentrating on Kant’s theories of space.
33
And while the work
on anthropology has been discussed by figures of the stature of Martin Heidegger and Michel
Foucault,
34
the Geography has not received anything like the same amount of attention.
Aside from the textual problems, one of the reasons is that philosophers have, by and large, not
known what to make of the works. This is demonstrated in debates about where the anthropology
lectures might be placed in editions of Kant’s work. If it seems unquestionably the case that today
the anthropology and geography belong together, this has not always been the view. Wilhelm
Dilthey and Erdmann argued that their relation was to the cosmology and physical geography;
while Emil Arnoldt and Adickes claimed that they were linked to the metaphysics lectures and
the section on empirical psychology.
35
This confusion might appear surprising, since Erdmann
shows that parts of the Anthropology derived from earlier lectures on geography, and Kant
goes out of his way to demonstrate the linked nature of the inquiries.
36
Yet there is a certainly
a link to the metaphysics lectures. Holly Wilson suggests the following view:
I distinguish between ‘origin’ and ‘arise’: the anthropology lectures arose out of the psychol-
ogy lectures, but had their origin in the physical geography lectures. Kant’s banning of psy-
chology from metaphysics initiated the movement toward an independent series of lectures
on anthropology, but the intent and content of the anthropology lectures finds its origin in
the physical geography lecture, which were initially given fifteen years prior to the start of the
anthropology lectures.
37
Those that do discuss the work on geography do, sometimes, note its relation to wider concerns
in Kant’s work. Howard Caygill, for instance, in his Kant Dictionary says that ‘in addition to this
explicit treatment of geography in his lectures, it is interesting to note Kant’s reliance on geo-
graphical terms and metaphors in his more strictly defined philosophical work’.
38
But this is as
far as his link goes, perhaps understandable given the nature of his project, but shared much
more generally. However, in recent years a number of people have begun to analyse Kant’s
33
See, for example, Browne Garnett Jr., The Kantian Philosophy of Space (note 20); Melnick, Space, Time, and
Thought in Kant (note 20); H. Sidgwick, Kant’s ‘exposition’ of space and time, in: R.F. Chadwick and C. Cazeaux
(Eds), Immanuel Kant: Critical Assessments, London, 1992, Four Volumes, I 89–102; and M.D. Newman, The unity
of time and space, and its role in Kant’s doctrine of a priori synthesis, in: Immanuel Kant: Critical Assessments, Vol.
II, 185–200.
34
See M. Heidegger, Kant and the Problem of Metaphysics, translated by R. Taft, Bloomington, 1997. Foucault trans-
lated the Anthropology as Anthropologie du point de vue pragmatique, Paris, 1964. This edition includes only a brief No-
tice historique, which is all that was published of Foucault’s secondary thesis on the Anthropology. See M. Foucault,
Introduction a
`l’Anthropologie de Kant, The
`se comple
´mentaire pour le doctorat de
`s letters, 1961; now translated by
R. Nigro and K. Briggs as Introduction to Kant’s Anthropology from a Pragmatic Point of View, New York, 2008.
Foucault does not discuss the work on geography.
35
H.L. Wilson, Kant’s Pragmatic Anthropology: its Origin, Meaning and Critical Significance, Albany, 2006, 17; Zam-
mito, Kant, Herder, and the Birth of Anthropology (note 20), 293.
36
B. Erdmann, Zur Geschichte des Textes, in: B. Erdmann (Ed), I. Kant, Reflexionen Kants zur kritischen Philosophie,
Stuttgart-Bad Cannstatt, 1992 [1882], 31–64.
37
Wilson, Kant’s Pragmatic Anthropology (note 35), 3; see R.B. Louden, Kant’s Impure Ethics: from Rational Beings
to Human Beings, Oxford, 2000, 62–63.
38
Caygill, A Kant Dictionary (note 28), 215.
9S. Elden / Journal of Historical Geography 35 (2009) 3–25
work on Anthropology in a new light, which necessarily forces an encounter with the Geography.
These would include Robert Louden’s Kant’s Impure Ethics, John Zammito’s Kant, Herder, and
the Birth of Anthropology,
39
and Holly Wilson’s Kant’s Pragmatic Anthropology. In addition Cam-
bridge University Press has recently published a useful book of essays on the anthropology,
40
and
there is an exceptional recent biography by Manfred Kuehn which, together with Eckart Fo
¨rster
and Peter Fenves’s pioneering work on the late Kant, enables a much more nuanced situation of
the geographical work.
41
It is in the light of these researches, alongside those of Robert Bernas-
coni and David Harvey on race discussed below, that a reassessment can begin.
Kant’s pragmatic purpose
Why did Kant lecture on geography? Although it became one of his most attended courses, this
does not explain things sufficiently. Wilson is valuable in tracking the changing objectives for the
geography lectures, suggesting that initially they were ‘purely scientific, that is, to make a more
certain knowledge of believable travel accounts, and to make this into a legitimate academic
course of study’.
42
But the popularity of the course meant that Kant could begin to suggest
that their aim could be ‘to civilise young students to become ‘citizens of the world’’.
43
Zammito
has similarly shown how the lectures are related to the Anthropology in providing knowledge, but
stresses this is for a philosophical purpose.
44
As Louden notes, therefore, their aim was more than
merely scholastic, but rather:
The anthropology and physical geography lectures are thus not primarily intended as further
contributions to Kant’s critical, transcendental philosophy program.[which] was not his
only concern. A major portion of Kant’s teaching activity was devoted to trying to enlighten
his students more about the people and world around them in order that they might live
(pragmatically as well as morally) better lives.
45
For Louden, anthropology and geography are thus ‘intersecting halves of a larger whole’.
46
The
problematic link between Kant’s views on geography and anthropology, and, especially, on race
and his cosmopolitan ethics will be discussed below, but the point here is somewhat different. This
is that Kant sees these lectures as providing a broad knowledge of the world as a foundation to the
more general studies of his students, and that together the physical geography and pragmatic an-
thropology give an empirical grounding for his thought more generally. In a postscript to his 1775
39
Zammito, Kant, Herder, and the Birth of Anthropology (note 20), especially 57–59, 284–296.
40
Jacobs and Kain (Eds), Essays on Kant’s Anthropology (note 9).
41
M. Kuehn, Kant: a Biography, Cambridge, 2001; E. Fo
¨rster, Kant’s Final Synthesis: an Essay on the Opus Postu-
mum, Cambridge, 2000; Fenves, Late Kant (note 16). This is the case despite the fact that these works do not treat
the Geography in any detail themselves.
42
Wilson, Kant’s Pragmatic Anthropology (note 35), 9.
43
Wilson, Kant’s Pragmatic Anthropology (note 35), 8.
44
Zammito, Kant Herder, and the Birth of Anthropology (note 39), 285–286.
45
Louden, Kant’s Impure Ethics (note 37), 65; see Wilson, Kant’s Pragmatic Anthropology (note 35), 20.
46
Louden, Kant’s Impure Ethics (note 37), 95.
10 S. Elden / Journal of Historical Geography 35 (2009) 3–25
article ‘Von den verschiedenen Racen der Menschen [On the Different Races of Human Beings]’
Kant suggested that the two lecture courses together were Weltkenntnis.
47
This would usually
translate as ‘world-knowledge’, but Wilson has suggested the felicitous ‘cosmopolitan knowledge’,
with ‘cosmopolitan philosophy’ for the related Weltwissenschaften.
48
This knowledge of the
world, for Kant, was integral to the moral and political life of the citizen. Both geography and
anthropology were taught by Kant because of their ‘pragmatic’ dimension, the way in which
this knowledge can guide us in our moral and practical life.
This world-knowledge, this cosmology, is essential to his other writings. Kant suggests that
physical geography is about the world as an ‘object of external sense’; and anthropology as an
‘object of inner sense’.
49
Indeed, in his essay on ‘The Conflict of the Faculties’, Kant divides
the philosophy faculty into two parts—the one that deals with ‘pure rational knowledge’ and
one that deals with ‘historical knowledge’. The former contains metaphysics of nature and morals,
along with pure philosophy and mathematics; the latter includes history, geography, philology,
the humanities and the empirical knowledge of the natural sciences.
50
Of these aspects, physical
geography is ‘the physical description of the earth’, ‘the first part of knowledge of the world’.
51
Wilson is therefore clear that the lectures on anthropology must be seen as philosophy:
Kant explicitly argues that the anthropology is a type of cosmopolitan philosophy. It is not
a scholastic philosophy, and it is not critical philosophy, but it is a type of philosophy.
The twofold field of physical geography and anthropology are viewed cosmologically and
pragmatically. In other words, Kant considered these two disciplines, in the way he taught
them, to be philosophy, and philosophy that was useful for the world.
52
These lectures were to serve as a propaedeutic for ‘practical reason’, and are a ‘history of the con-
temporary condition of the earth or geography, in the widest sense’.
53
This, for Kant, is ‘the pre-
liminary exercise in the knowledge of the world’.
54
Knowledge of the world is thus of both ‘the
human being and nature’.
55
Physical geography studies nature, anthropology the human, but
the latter outweighs the former, since ‘nature exists for the sake of the human being. The human
being is the end of nature’.
56
Nonetheless, the twofold field of Weltkenntis needs to be treated
cosmologically.
57
47
KGS II, 443.
48
See Wilson, Kant’s Pragmatic Anthropology (note 35), 20, 113–115.
49
KGS IX, 156–157; see Anthropology from a Pragmatic Point of View (note 10).
50
I. Kant, The conflict of the faculties, in: Religion and Rational Theology, translated and edited by A.W. Wood and
G. Di Giovanni, Cambridge, 1996, 256.
51
KGS IX, 157. The phrase ‘description of the earth’ is Rink’s interpolation. As B. Jacobs (Kantian character and the
science of humanity, in: Jacobs and Kain (Eds), Essays on Kant’s Anthropology (note 9) 105–134, 132 n. 49) notes, in
Kant’s Groundwork for the Metaphysic of Morals, the same role is played by physics.
52
Wilson, Kant’s Pragmatic Anthropology (note 35) 5, 115.
53
KGS II, 312.
54
KGS II, 443. See also the remarks in x70 of the Education lectures, cited by May, Kant’s Concept of Geography
(note 2), 132.
55
KGS IX, 158.
56
KGS XXV, 470, 733, cited in Louden, Kant’s Impure Ethics (note 37), 95.
57
KGS II, 443.
11S. Elden / Journal of Historical Geography 35 (2009) 3–25
The context and the structure of the geography
Kant was an innovator in geography, if for no other reason that he was one of the very first to
lecture on it as an explicit topic, before it was common to have chairs in geography in Germany.
While others lectured on it in a way that was more akin to travel writing, Kant attempted to system-
atise the subject, synthesising insights from a range of different sources. Indeed, his outline for the
course was unique, and he had to ask for special dispensation from the Minister of Education in or-
der to give a course for which no textbook could be found.
58
The latter was of the opinion that ‘the
worst textbook is certainly better than none, and professors may, if they possess so much wisdom,
improve upon their authors to the extent that they can, but the reading from notes must simply
be stopped. From this we nevertheless make exception of Professor Kant and his course on Physical
Geography, for which no appropriate textbook is yet available’. Kant’s own manuscript was known
as the Diktattext, prepared around 1759 but much amended and now lost. The Holstein-Beck man-
uscript is believed to be the closest to Kant’s Diktattext. After Kant retired his colleague K.L.
Po
¨rschke used the text of George Henry Millar, The New and Universal System of Geography.
59
There is not the space here to outline how Kant’s understandings of specific geographical issues
relate to knowledge at the time and to debates in a range of contexts since. This has been discussed
at some length elsewhere.
60
Nonetheless Kant’s structure of the subject is important in itself and it is
worth focussing on this. Initially his treatment included what was later separated out as the Anthro-
pology. While ‘geography’ in the late 1700s meant something rather different to today, and crucially
depended on where it was being taught or discussed,
61
what is striking is that Kant is interested in
establishing subdivisions within the broad topic, many of which bear similarity to subdisciplines
today, even if others are obsolete. Kant is thus trying to move geography—not always success-
fully—beyond mere ‘earth description’.
62
Anton Friedrich Bu
¨sching’s Neue Erdbeschreibung
(New Description of the Earth) was, as Withers recounts, based on lectures given at Go
¨ttingen
in the philosophy department, and published in eleven volumes between 1754 and 1792.
63
Kant
58
Louden, Kant’s Impure Ethics (note 37), 184–185 n. 6, drawing on K. Vo
¨rlander, Immanuel Kants Leben, Leipzig,
1911, 41–43.
59
The quote and this information come from http://users.manchester.edu/Facstaff/SSNaragon/Kant/Lectures/
lecturesListDiscipline.htm#geography. See G.H. Millar, The New and Universal System of Geography, London, 1782.
60
In relation to geography, see Livingstone and Harrison, Immanuel Kant, subjectivism and human geography (note
2); Withers, Kant’s geography in comparative perspective (note 32); and M. Church, Immanuel Kant and the emer-
gence of modern geography, in: Elden and Mendieta (Eds), Reading Kant’s Geography (note 32). In relation to the nat-
ural sciences more generally, see P.R. Sloan, Preforming the categories: eighteenth-century generation theory and the
biological roots of Kant’s a priori, Journal of the History of Philosophy 40 (2002) 229–253; and his Kant on the history
of nature: the ambiguous heritage of the critical philosophy for natural history, Studies in History and Philosophy of
Biological and Biomedical Sciences 37 (2006) 627–648.
61
On the geographical aspect alongside the temporal in determining geographical knowledge, see Withers, Kant’s ge-
ography in comparative perspective (note 32); and more broadly his Placing the Enlightenment: Thinking Geographically
about the Age of Reason, Chicago, 2007. This is one of the key arguments of D.N. Livingstone, Putting Science in its
Place: Geographies of Scientific Knowledge, Chicago, 2003.
62
See Withers, Placing the Enlightenment (note 61), 178; Church, Immanuel Kant and the emergence of modern
geography (note 60).
63
Withers, Placing the Enlightenment (note 61), 183–184; Livingstone, The Geographical Tradition (note 2), 108–109.
12 S. Elden / Journal of Historical Geography 35 (2009) 3–25
too depends greatly on these kind of empirical examples, but he tries to situate this within a broad
structure of the subject. It is notable that the structure of most of his geography courses is very sim-
ilar, even as the content changes. Kant begins with an introduction which provides a general assess-
ment of the subject and its relation to other topics; moves to a discussion of the mechanics of the
earth and then breaks the substantive part of the analysis into three parts.
The first substantial part of the text is the more accurately ‘physical geography’. It looks at
a range of physical processes concerning earth and water: oceans; land and islands as examples
of the earth and its terrain; earthquakes, volcanoes and electricity; springs and wells; rivers and
water, wind currents; climate, the atmosphere and temperature; transformations of the earth;
and seafaring. The second part concerns the three realms of nature: fauna, flora, and minerals.
Many of the aspects of this second part would not fall within the remit of modern day geography,
with the exception of biogeography. The final part of the book offers a regional geography of the
world, with descriptions of particular regions and places in Asia, Africa, Europe and America.
But Kant’s Physical Geography was also a moral and political account, and included human be-
ings in it, usually just before the section on animals, discussing racial differences. In addition there
are a range of comments in the third part concerning their geographical differentiation. Humans
are thus seen as part of physical geography, both because they are one of the features of the
Erdboden—the earth’s surface—but also because they a causal mechanism for change to the earth
itself, because they build dams, drain swamps and fell forests, thus changing landscape and
climate.
64
After a decade of lecturing, announcing his courses for Winter Semester 1765–66, Kant out-
lined how his vision had developed, and how the more explicitly physical geography could some-
times be underplayed. This is worth quoting at length:
Since then I have gradually expanded this scheme, and now I propose, by condensing that
part of the subject which is concerned with the physical features of the earth, to gain the
time necessary for extending my course of lectures to include the other parts of the subject,
which are of even greater utility. This discipline will therefore be a physical, moral and polit-
ical geography. It will contain, first of all, a specification of the remarkable features of nature
in three realms. The specification will, however, be limited to those features, among the in-
numerably many which could be chosen, which particularly satisfy the general desire for
knowledge, either because of their rarity or the effect which they can exercise on states by
means of trade and industry. This part of the subject, which also contains a treatment of
the natural relationship which holds between all the lands and seas in the world, and the rea-
son for their connection, is the essential foundation of all history. Without this foundation,
history is scarcely distinguishable from fairy-tales.
This is a point worth underlining. As Kant claims in the geography lectures, geography is the
ground or foundation for history, as events necessarily take place somewhere, in some context.
65
This is not to assert a priority for geography over history, but rather to insist on their relation
together, a complementary analysis rather than a separation. Indeed, as Withers has noted, it
64
KGS IX, 298.
65
KGS IX, 163.
13S. Elden / Journal of Historical Geography 35 (2009) 3–25
was a commonplace at the time to see geography as the left ‘eye’ of history, with chronology as the
right.
66
The second part of the subject considers human beings, throughout the world, from the point
of view of the variety of their natural properties and the differences in that feature of the hu-
man which is moral in character. The consideration of these things is at once very important
and also highly stimulating as well. Unless these matters are considered, general judgments
about man would scarcely be possible. The comparison of human beings with each other,
and the comparison of the human today with the moral state of the human in earlier times,
furnishes us with a comprehensive map of the human species. Finally, there will be a consid-
eration of what can be regarded as a product of the reciprocal interaction of the two previ-
ously mentioned forces, namely, the condition of the states and nations throughout the
world. The subject will not be considered so much from the point of view of the way in which
the condition of states depends on accidental causes, such as the deeds and fates of individ-
uals, for example, the sequence of governments, conquests and intrigues between states. The
condition of states will rather be considered in relation to what is more constant and which
contains the more remote ground of those accidental causes, namely, the situation of their
countries, the nature of their products, customs, industry, trade and population.
67
Kant therefore sets out a range of distinctions—the physical, moral and political geography is
alluded to here, but the actual analysis is rather more complicated. By the mid 1770s he offers
a range of possibilities:
-Physical geography: the foundation or ground for other types of geography as well as his-
tory—a general study or outline of nature;
-Mathematical geography: concerned with the measure of the shape, size and motion of the
earth, and its situation in the solar system;
-Moral geography: the relation between moral codes and customs and regions, a kind of spa-
tial differentiation;
-Political geography: the relation of political systems and political laws to physical features of
geography, part of the reason why these are only nominally universal;
-Commercial [Handlungs] geography: concerned with the geographical elements of trade in
surplus products;
-Theological geography: concerned with theological attitudes and principles and their relation
to physical features of the landscape; again a form of spatial differentiation.
68
66
Withers, Placing the Enlightenment (note 61), 178; Kant’s geography in comparative perspective (note 32).
67
Kant, Theoretical Philosophy 1770–1775, 298–299; KGS II, 312–313, translation modified.
68
KGS IX, 160–161, 164–165. Rink changes the order of their first presentation, adding ‘literary geography’ (KGS
IX, 161) which is not in the 1774 transcript or in the fuller elaboration. He also adds detail to the elaboration, usually
based on material later delivered in the lectures themselves. Some of the changes—‘civil society’ for ‘society’ or the re-
placement of ‘Handlungs Geographie’ with ‘merkantilische Geographie’—are more interpretative. May claims that ‘his
concept of the limits and scope of geography is inevitably much broader than any contemporary concept can reasonably
be’ (Kant’s Concept of Geography (note 2), 153).
14 S. Elden / Journal of Historical Geography 35 (2009) 3–25
Kant’s humans thus move from being detached from the world into closer to relation with the
earth, nature and other humans. As the material in the latter part developed this led to the parallel
lectures on anthropology.
69
What the lectures together provide is a grounding of empirical detail.
Crucially though they try to provide a geographical perspective on a range of social, cultural and
physical phenomena. If geography is a study of the difference space makes, Kant can certainly lay
claim to being a geographer. Yet, on the other hand, Kant was writing at a time when there was
the beginning of a noticeable shift away from the compilation of information and reports from
a distance. In the early nineteenth century field-based science was in the ascendance, with the im-
portance of trained eye-witness accounts, following pioneering studies of Alexander von
Humboldt, Georges Cuvier, and somewhat earlier, the scientists involved with Cook’s voyages.
70
The impure physics and the question of race
In his lectures on Logic, Kant suggests that there are four fundamental questions.
1. ‘What can I know?’
2. ‘What ought I to do?’
3. ‘What may I hope?’
4. ‘What is the human being?’
Kant suggests that ‘Metaphysics answers the first question, morals the second, religion the third,
and anthropology the fourth. Fundamentally, however, we could reckon all of this to anthropol-
ogy, because the first three questions refer to the last one’.
71
Just as these other realms of thought
rest on the fundamental question, namely anthropology; so too do Kant’s reflections on the ma-
terial world rest on the understanding of geography. For Kant, knowledge of the world is not
pragmatic merely when it is ‘extensive knowledge of things in the world, for example, animals,
plants and minerals from various lands and climates—but only when it contains knowledge of
the human being as a citizen of the world’.
72
In the Critique of Pure Reason, Kant distinguishes
between nature and freedom, to which accord the natural law and the moral law. The first accords
to everything that is; the second to what should be. He suggests that while both have separate sys-
tems, they can come together in a single philosophical system—his own critical philosophy. In this
critical philosophy there is a pure philosophy of pure reason; but also an empirical element, and
this for both nature and freedom.
73
In the Geography lectures Kant begins by distinguishing be-
tween pure rational knowledge and knowledge from experience, through our senses.
74
In other
69
Wilson, Kant’s Pragmatic Anthropology (note 35) 11, 14.
70
Church, Immanuel Kant and the emergence of modern geography (note 60); Bowen, Empiricism and Geographical
Thought (note 1), 208.
71
I. Kant, Lectures on Logic, translated and edited by J.M. Young, Cambridge, 1992, 538. The first three questions
appear in Critique of Pure Reason, translated by P. Guyer and A.W. Wood, Cambridge, 1998, A805/B853.
72
Kant, Anthropology from a Pragmatic Point of View (note 10), 4.
73
Kant, Critique of Pure Reason (note 71), A840/B868.
74
KGS IX, 156; see Louden, Kant’s Impure Ethics (note 37), 24.
15S. Elden / Journal of Historical Geography 35 (2009) 3–25
words, pragmatic anthropology is the relation of the human to the world, and is thus what Robert
Louden calls ‘impure ethics’.
75
Although much of the material in the geography lectures would
also come under that remit—recall the origin of the anthropology lectures—we might make
a case that much of the rest of the geography is ‘impure physics’, in other words the empirical
detail that inhabits the categories of abstract thought. The use of this kind of detail is pronounced,
for example, in Kant’s third critique, especially in the passages on teleological judgment.
76
And in
addition, while in his Metaphysical Foundations of Natural Science Kant expresses the view that
the empirical sciences require a pure grounding;
77
his last work, the Opus Postumum, bears the
title of ‘Transition from the Metaphysical Foundations of Natural Science to Physics’.
78
Thus while much of the detail of the Geography may be outdated and therefore of merely his-
torical interest, Kant’s way of structuring geographical knowledge and its relation to his thought
as a whole is of enduring importance. This importance lies both in the way he understands geog-
raphy as a counterbalance to history, and in terms of the organization of knowledge. All perceived
things are located in logical classifications such as those of Linnaeus; and in space and time. Logic
deals with the first; physics with space and time, and of these, geography deals with space; history
with time.
79
Kant says that our experience of the world is limited in time—our limited lifespan—
and space—even if we have travelled extensively. So we must rely on others, who can provide us
with either narrative (temporal, history) or description (spatial, geographical).
80
Geography there-
fore allows us access to the ordering and categorizing of the world. Indeed, Kant distinguishes
geography as the description of the whole world from topography as the description of single pla-
ces and chorography as that of regions.
81
Wilson thus suggests that the role is fundamentally dif-
ferent from a merely enumerative account of the world as it is. As she suggests physical geography
in Kant’s terms ‘is not meant to be a description of the world as a scientist would view it, but
rather geography is to be viewed in its purposiveness’.
82
Thus for all their problems and their un-
doubted neglect, these lectures remain a key part of Kant’s work, and in recent years there has
been some important discussion. The remainder of this essay will first discuss race—a theme
treated at length in the various versions of the Geography; and then space, a topic which, while
75
Louden, Kant’s Impure Ethics (note 37). S. Kofman (The economy of respect: Kant and respect for women, trans-
lated by N. Fisher, in: R.M. Schott (Ed) Feminist Interpretations of Immanuel Kant, University Park, 1997, 355–372,
357) has similarly argued that the Anthropology can be seen as an ‘addendum or appendix’ to the work on virtue.
76
Kant, Critique of the Power of Judgment (note 7), 233–255; see Glacken, Traces on the Rhodian Shore (note 7),
530–535.
77
I. Kant, Metaphysical foundations of natural science, in: Theoretical Philosophy After 1781, edited by H. Allison
and P. Heath, translated by M. Friedman, Cambridge, 2002, 181–270. See also May, Kant’s Concept of Geography
(note 2), 148 which schematises Kant’s Classification of Sciences.
78
I. Kant, Opus Postumum, edited by E. Forster, translated by E. Fo
¨rster and M. Rosen, Cambridge, 1993, especially
10–22; Fo
¨rster, Kant’s Final Synthesis (note 41), 1.
79
KGS IX, 159–160, 162. For a criticism of this division, see Schaefer, Exceptionalism in geography (note 21).
80
KGS IX, 159. This is especially the case for someone who travelled as little as Kant, though his location in the port
city of Ko
¨nigsberg meant he had a wider range of contacts than is generally assumed. See Kuehn, Kant: a Biography
(note 41).
81
KGS IX, 159. Orography and hydrography—the description of mountains and water—are also mentioned as
divisions in Rink’s edition, but not in the transcript.
82
Wilson, Kant’s Pragmatic Anthropology (note 35), 10.
16 S. Elden / Journal of Historical Geography 35 (2009) 3–25
not extensively treated in the Geography, is key to any engagement with Kant’s geography in more
general terms.
The question of Kant’s racism has been discussed quite extensively in recent years. More chal-
lenging than his rehearsal of racial stereotypes and entertaining of prejudices is his theorisation
of race. What this means is that Kant cannot simply be excused as a product of his time—trading
on contemporary views about racial superiority and the like—because he went out of his way to ex-
plicitly theorise race, as a crucial category of human life. The charges can be found in a series of es-
says by Robert Bernasconi.
83
For Bernasconi Kant plays a significant role because he can be seen as
the thinker who actually ‘invented the scientific concept of race’, as he was the thinker who ‘gave the
first clear definition of it’.
84
Indeed, Bernasconi contends that in doing so, he did not simply define
race, but ‘played a crucial role in establishing the term ‘race’ as the currency within which discus-
sions of human variety would be conducted in the nineteenth century’.
85
Bernasconi’s phrasing is
telling: ‘Kant opened up a new space for thinking: he took it into new territory’. While Kant may
have stopped there, ‘those who came after him worked in the space he opened up’.
86
Some responses
have contended that this does not affect Kant’s broader thought, and that Kant’s arguments are
stronger than his prejudices.
87
But this is insufficient for Bernasconi. He suggests that there is a dif-
ferent between analytic and Continental philosophy in terms of their treatment of the biography of
thinkers. Analytic is defined as ‘a form of thinking that leaves no room for synthesis, holism or di-
alectic’.
88
While the geographies and histories of this distinction are problematic, as he recognises,
there is some profit in the divide for analysis. For Bernasconi the analytical approach to philosophy,
which tends to be ahistorical in its reading of canonical thinkers, means that:
The racism of a philosopher is easily put to one side. Kant’s racism does not raise a question
for his cosmopolitanism because cosmopolitanism can be defined in such a way as to exclude
racism. Thus, the fact that Kant was a racist has no implications for contemporary Kantians.
However, things are somewhat different for Continental philosophers.
89
83
R. Bernasconi, Who invented the concept of race? Kant’s role in the enlightenment construction of race, in: Ber-
nasconi (Ed), Race, London, 2001, 11–36; Kant as an unfamiliar source of racism, in: J.K. Ward and T.L. Lott
(Eds), Philosophers on Race: Critical Essays, Oxford, 2002, 145–166; and Will the real Kant please stand up: the chal-
lenge of enlightenment racism to the study of the history of philosophy, Radical Philosophy 117 (2003) 13–22. See also
the exchange between Bernasconi and J. McCarney in Radical Philosophy 199 (2003) 32–37, which, while focused on
Hegel, discusses the key issues. See also E.C. Eze, Kant, in: Eze (Ed) Postcolonial African Philosophy: a Critical Reader,
Oxford, 1997, 103–140. Also important is David Farrell Krell, ‘The bodies of black folk: from Kant and Hegel to Du
Bois and Baldwin’, boundary 2 27 No. 3 (2000) 103–314; and more generally his The Tragic Absolute: German Idealism
and the Languishing of God, Bloomington, Indiana University Press, 2005.
84
Bernasconi, Kant as an unfamiliar source of racism (note 83), 147; see ‘Who invented the concept of race?’ (note
83).
85
R. Bernasconi, Introduction, in: R. Bernasconi (Ed) Concepts of Race in the Eighteenth Century, London, Eight
Volumes, Vol. 1, 2001, viii.
86
Bernasconi, Will the real Kant please stand up (note 83), 19.
87
For an account along these lines, see T.E. Hill and B. Boxill, Kant and race, in: Boxill (Ed), Race and Racism,
Oxford, 2001, 448–471.
88
Bernasconi, Will the real Kant please stand up (note 83), 22.
89
R. Bernasconi, Introduction, in: Bernasconi with Sybol Cook (Ed), Race and Racism in Continental Philosophy,
Bloomington, 2003, 1–7, 2; see Will the real Kant please stand up (note 83), 20, 25.
17S. Elden / Journal of Historical Geography 35 (2009) 3–25
Given that the fundamental philosophical influence in contemporary geography is from the
‘Continental’ strand, this is of crucial importance.
One of the few geographers who has recognised the implications of this challenge has been
David Harvey. In a series of publications Harvey has related the geography lectures to the con-
temporary interest in Kant’s cosmopolitanism,
90
suggesting, like Bernasconi, that the claims of
the lectures render this particular concept deeply problematic. Harvey notes that many Kantians
want to dismiss the work on geography as ‘‘irrelevant’, ‘not to be taken seriously’’ or [suggest that]
it ‘lacks interest’‘.
91
In this position they are close to the oft-quoted remark from Benno Erdmann,
who described the Anthropology as the ‘laborious compilation of a seventy-four year old man as
he stood on the threshold of decrepitude’.
92
For Harvey ‘the content of Kant’s Geography is noth-
ing short of an intellectual and political embarrassment’.
93
Both aspects of this assessment might
be encapsulated by Kant’s claim that ‘humanity has its highest perfection in the white race’, with
‘yellow’, ‘Negro’ and ‘red’ races at respectively lower rungs.
94
In his 1755 book Universal Natural History and Theory of the Heavens Kant had displayed the
same penchant for dubious scientific reasoning, uncritical adoption of other material and wild
speculation that was to characterize his geography lectures beginning the next year. In particular,
Kant speculates on the inhabitants of Mercury and Venus, suggesting that their bodily constitu-
tion is such that they could not dwell on earth. In addition those planets like earth and Mars are
midway in the planetary system and therefore their inhabitants are balanced, whereas those of
Mercury and Venus are stupid, weak, unable to submit to justice, and coarse in body.
95
As David
Clark suggests, this ‘functions as a barely displaced allegory of Europe’s close encounters with
Africa and other equatorial regions of the universe’.
96
This text might be considered as an early
work, unimportant and tangential to Kant’s mature philosophical edifice. The same cannot be
said of the lectures on geography, which extended until his retirement, and, from the available
transcripts never ceased to discuss race and racial ideas.
Harvey is right to suggest that there is much material that is offensive, embarrassing and irrel-
evant in the lectures, and makes a compelling case for the difficulties in accepting Kant’s ethical
views in their wake. These views cannot be swept away, disassociated from the thought people
want to salvage. In Harvey’s account, there is a direct linkage between the expression of these
ideas and the philosophical architecture erected on top of the Weltkenntnis. Harvey is right
that the responses of traditional Kantians are inadequate, but over recent years there has been
the beginning of some interest in these issues. While there are a number of Kantians who remain
90
D. Harvey, Cosmopolitanism and the banality of geographical evils, Public Culture 12 (2000) 529–564; and Geo-
graphical knowledges/political powers, Proceedings of the British Academy 122 (2004) 87–115. These themes are also
central to his forthcoming Cosmopolitanism and the Geographies of Freedom, New York 2008.
91
Harvey, Cosmopolitanism and the banality of geographical evils (note 90), 532.
92
Erdmann, Zur Entwicklungsgeschichte von Kants Anthropologie, in: Kant, Reflexionen Kants zur kritischen Philos-
ophie (note 36), 68. I have taken the translation from R.B. Louden, The second part of morals, in: Jacobs and Kain
(Eds), Essays on Kant’s Anthropology (note 9), 60–84, 60.
93
Harvey, Cosmopolitanism and the banality of geographical evils (note 90), 532.
94
KGS IX, 316.
95
I. Kant, Universal Natural History and Theory of the Heavens, translated by S.L. Jaki, Edinburgh, 1981, 195.
96
D.L. Clark, Kant’s aliens: the Anthropology and its others, CR: the New Centennial Review 1 (2001) 201–289, 258.
18 S. Elden / Journal of Historical Geography 35 (2009) 3–25
in denial, both on race and Geography, there are others who have risen to a challenge to think
both in the context of his work as a whole. In part this has been a by-product of important studies
of the Anthropology, but more generally there are now philosophers who have taken Kant’s
geography seriously, which is not to say uncritically.
Louden is a good example of this tendency. He stresses the teleological aspects of Kant’s theory
of race, which implies that ‘racial characteristics are present in the human species because they
help us reach our collective destiny’,
97
that is they come out depending on climatic and landscape
features—a form of geographical determinism. What this means is that Kant is arguing for a the-
ory of race as monogenetic rather than polygenetic, as proposed by contemporaries, such as Lord
Kames in 1774. Kames suggested that different races were different species, from different origins,
whereas monogenism is the belief that humans are all ultimately related, stemming from some
common ancestor. While more traditional biblical accounts stress monogenism—based on the lin-
eage through Adam—accounts at the time began to stress the distinctions of the races. Kant thus
believes in the ultimate unity of the human species, rather than being like those racial thinkers
who thought that different origins implied inequality.
98
But this is not to defend Kant, as he
did say some deeply unpleasant things about race, which in Louden’s balanced judgment means
that earlier views—such as that of Ernst Cassirer that Kant was progressive, ‘humanitarian’ and
‘equalitarian’, and distinct from Gobineau—need to be discarded.
99
For Louden, ‘not all of
Kant’s ideas about race are entirely ‘humanitarian’ and ‘equalitarian’, and the gap between
Gobineau and Kant is unfortunately not always as wide as one would like it to be’.
100
Given the way that race is the link between the Anthropology and the Geography, its minimal
role in the former is perhaps somewhat surprising.
101
As Wilson suggests, ‘the anthropology be-
gins where physical geography ends; the different climates and environments, explored in physical
geography, explain the different kinds of human beings in the world, but the inner germs and nat-
ural predispositions, explored in anthropology, explain why the human being can adapt itself to
the different climates and environments’.
102
Indeed, Kant separated out from the Physical Geog-
raphy course much of what he calls ‘moral geography’, that is that which concerns the ‘customs
and characters’ of different peoples.
103
But in the published version of the Anthropology Kant says
little, and suggests that the observations on the relation of physiognomy to race ‘belong more to
physical geography than pragmatic anthropology’.
104
He does say more in the Geography. This
has led Louden to propose the intriguing possibility that the older Kant may have come to doubt
97
Louden, Kant’s Impure Ethics (note 37), 97.
98
Louden, Kant’s Impure Ethics (note 37) 96–98; Bernasconi, Introduction, in: Concepts of Race (note 85), viii–ix. For
a more detailed discussion, see Bernasconi, Who invented the concept of race? (note 83); Withers, Placing the Enlight-
enment (note 61), Chapter 7; and A.C. Cohen, Kant on epigenesis, monogenesis and human nature: the biological prem-
ises of anthropology, Studies in History and Philosophy of Biological and Biomedical Sciences 36 (2006) 675–693,
especially 681.
99
E. Cassirer, The Myth of the State, New Haven, 1946, 235.
100
Louden, Kant’s Impure Ethics (note 37), 100.
101
Kant, Anthropology from a Pragmatic Point of View (note 10), 223–224.
102
Wilson, Kant’s Pragmatic Anthropology (note 35), 15.
103
KGS IX, 164.
104
Kant, Anthropology from a Pragmatic Point of View (note 10), 199; for a discussion see Louden, Kant’s Impure
Ethics (note 37), 94–95.
19S. Elden / Journal of Historical Geography 35 (2009) 3–25
some of his views on race, which is why they are so brief in the Anthropology, which he edited, and
more extensive in the Geography, based on older materials and not edited until Kant was too old
to proofread the manuscript.
105
But as Louden notes, there is not the evidence to determine this,
and those materials Kant did publish on race (from 1775, 1777, 1785 and 1788) tend to confirm
the worst.
106
Suggesting that it was after the 1788 essay he changed his mind is the last resort, with
Kleingeld arguing that Kant changed his views in the 1790s, and probably after 1792.
107
But we
should note that the 1798 edition of the Anthropology makes reference to Christoph Girtanner’s
U
¨ber das Kantische Prinzip fu
¨r die Naturgeschichte, a work he saw as ‘presented so beautifully and
thoroughly.(in accordance with my principles)’.
108
Thus Kant’s politics are rendered deeply suspect. For geographers this may appear relatively
unimportant in itself—Kant is hardly alone in the tradition for this—but the contemporary res-
onances are hard to escape. Yet while almost anyone today would share his rejection of the lec-
tures’ reactionary attitudes, this does not mean that the lectures are without value. Geographers
have long tried to separate the discipline and its questions from its imperial legacy and the answers
from that time and space. Indeed, it is precisely because we need to trace political ruptures that we
need to think historically.
Space: reassessing Kant’s geography
It is another rupture that may lead to the more important critique: that while we think we know
Kant’s view of space, the account Kant offers of space in the work on geography and, more im-
portantly, elsewhere, is at odds with this received wisdom. Thus this final section moves beyond
the Geography to begin to reassess Kant’s geography more generally. Kant’s position within the
discipline of geography is too often reduced to a caricature, again more often cited than read.
Kant, it is generally supposed, held a view of space that was totalising, based on Cartesian ge-
ometry, absolute in the Newtonian sense. Thus we find a range of adjectival pairings common:
Kantian space is Cartesian, Newtonian, and sometimes even more ahistorically Euclidean. Cer-
tainly in Kant’s Critique of Pure Reason, in the section on the transcendental aesthetic, space
and time are pure intuitions, effectively hardwired into our brains, our very way of perceiving
the world, rather than attributes of it. Space and time are thus the way that we experience, forms
of sensibility, not experienced as such. The relation of this to Descartes’ and Newton’s view of the
material world is more complicated than a simple equation, but this is certainly not an open
system.
105
Louden, Kant’s Impure Ethics (note 37), 207–208 n. 71.
106
Kant’s essays on race are collected in Concepts of Race in the Eighteenth Century (note 85), Vol. 3. As Bernasconi
notes there (Editor’s Note, vii–ix) the 1788 essay U
¨ber den Gebrauch teleologischer Principien in der Philosophie, is an
important step towards Kant’s third critique, the 1790 Critique of the Power of Judgment (note 7).
107
Pauline Kleingeld, Kant’s second thoughts on race, The Philosophical Quarterly 57 (2007) 573–592. See also
Fenves, Late Kant (note 16), 102–104, 111. A more useful, if problematic, account of the possibilities of Kant is found
in S. Muthu, Enlightenment Against Empire, Princeton, 2003.
108
C. Girtanner, U
¨ber das Kantische Prinzip fu
¨r die Naturgeschichte [1796], reprinted in Concepts of Race in the Eigh-
teenth Century (note 85), Vol. 7; Kant, Anthropology from a Pragmatic Point of View (note 10), 223.
20 S. Elden / Journal of Historical Geography 35 (2009) 3–25
Nonetheless Kant’s Critique of Pure Reason and his mature views on space were explicitly di-
rected at challenging a pure Newtonian conception of space. To assert this does not mean that
Kant’s view of space is something that geographers today would want to, or should, adopt.
But they should surely know better what they are arguing against. His earlier writings had
been closer to a relational view, much influenced by Leibniz. But Kant’s critical turn is not entirely
against this perspective. As noted above, Kant’s mature views on space were first presented in his
1770 Inaugural Dissertation, and here he attempts to mediate between a Leibnizian relative view of
space—space is a product of the relations between things—and a Newtonian absolute view—
space is the container for things.
109
Back in 1969, in Explanation in Geography,
110
Harvey discussed Kant in a rather different reg-
ister to his more recent work on the question of race and cosmopolitanism. Then, as now, Harvey
was deeply suspect of Kant’s absolute view of space. Harvey suggests that the absolute view of
space needs to be understood, but alongside it we also need to view and understand space as rel-
ative and relational. Even given the caveat above, Kant might seem to be a long way from that
position. But Harvey in Explanation in Geography offers what might be an intriguing possibility.
This is when he suggests that Kant’s view of space is one in which ‘spatial magnitude is.only
a measure of the intensity of acting forces exerted by the substance’.
111
Harvey then concludes
his analysis:
Such a view of space is contrary to the view on which Kant based his philosophy of geog-
raphy. Thus space is no longer something which can encompass our perceptions of the
world. It is, rather, a collection of measures determined by our perceptions. If space and mat-
ter can no longer be effectively separated and if the properties of space can no longer be re-
garded as given a priori, the logical justification for the particular view of geography adopted
by Kant, Hettner, and Hartshorne, can no longer be sustained.
112
Harvey’s source for the view of Kant on ‘spatial magnitude’ is Max Jammer’s important study
Concepts of Space. In the quotation from Jammer, the analysis is of Kant’s 1755 and 1756 writ-
ings. Jammer actually provides a nuanced account of how Kant moved from a more Leibnizian
view in those earlier works, to a more Newtonian view around 1769, under the influence of Leon-
hard Euler.
113
In 1769 Kant tries to show that ‘absolute space, independent of the existence of all
matter and as itself the ultimate foundation of the compound character of matter, has a reality of
its own’.
114
Kant thus moved, on this account, from a relative view of space to an absolute one.
109
One of the best geography accounts of this is in Richards, Kant’s geography and mental maps (note 1).
110
D. Harvey, Explanation in Geography, London, 1969, 70–78, 207–210.
111
Harvey, Explanation in Geography (note 110), 212; citing M. Jammer, Concepts of Space: the History of Theories of
Space in Physics, Cambridge, 1954, 130. I have used the 3rd ed (New York, 1993). The citation is from page 132 of this
edition.
112
Harvey, Explanation in Geography (note 110), 212. In a more recent piece, Space as a keyword, in: N. Castree and
D. Gregory (Eds), David Harvey: a Critical Reader, Oxford, 2006, 270–293, Harvey does not significantly depart from
this view.
113
See Jammer, Concepts of Space (note 111), 131–132.
114
Kant, concerning the ultimate ground of the differentiation of directions in space, Theoretical Philosophy
(1755–1770) 366.
21S. Elden / Journal of Historical Geography 35 (2009) 3–25
But he does not rest there. As Jammer convincingly claims, Kant further worked on the view of
space that he held, and by the Inaugural Dissertation of 1770 he had reached a different view, at-
tempting to reconcile these positions.
The concepts of absolute space and absolute time are considered to be mere conceptual fic-
tions, a mental scheme of coexistence and sequence among sense particulars. Not itself aris-
ing out of sensations, the concept of space is a pure intuition, neither objective nor real, but
subjective and ideal.
115
This therefore forms the view of space that is well known in Kant, where there are two pure
intuitions—space and time—that structure all that we experience about the world. In Jammer’s
words, space ‘is not an object of perception, it is a mode of perceiving objects’.
116
Briefly, and necessarily crudely, a return to the Critique of Pure Reason is helpful here. Kant
suggests that space is an object of outer sense, and time of inner sense.
117
Space then is as geog-
raphy is, an element of the external world alone, whereas time is both of the external and inter-
nal.
118
Kant then poses the questions he aims to address:
Now what are space and time? Are they actual entities? Are they only determinations or re-
lations of things, yet ones that would pertain to them even if they were not intuited, or are
they relations that only attach to the form of intuition alone, and thus to the subjective
constitution of our mind, without which these predicates could not be ascribed to any thing
at all.
119
Kant then suggests that space must be understood as a representation, thus leading him to talk
of a representation of space. It is a priori, prior to experience, and the necessary condition for it,
‘the ground of all intuitions.the condition of possibility of appearances, not as a determination
dependent on them’.
120
In a marginal note he added that ‘space is not a concept, but an intui-
tion’.
121
It is in this sense, and this sense alone, that time is ‘something real, namely the real
form of inner intuition’.
122
In another note Kant adds a crucial addition: ‘So is space. This proves
that here a reality (consequently also individual intuition) is given, which yet always grounds the
reality as a thing. Space and time do not belong to the reality of things, but only to our
representations’.
123
In the first edition (1781) the point about magnitude is therefore the reverse of the earlier
claim cited by Harvey from Jammer. Kant claims that ‘space is represented as a given infinite
magnitude. A general concept of space can determine nothing in respect to magnitude’.
124
In
115
Jammer, Concepts of Space (note 111), 134–135.
116
Jammer, Concepts of Space (note 111), 138.
117
Kant, Critique of Pure Reason (note 71), A22–A23/B37.
118
Kant, Critique of Pure Reason (note 71), A34/B50.
119
Kant, Critique of Pure Reason (note 71), A23/B37–B38.
120
Kant, Critique of Pure Reason (note 71), A24/B38–B39.
121
Kant, Critique of Pure Reason (note 71), note to A23.
122
Kant, Critique of Pure Reason (note 71), A37.
123
Kant, Critique of Pure Reason (note 71), note to A37.
124
Kant, Critique of Pure Reason (note 71), A25.
22 S. Elden / Journal of Historical Geography 35 (2009) 3–25
the second edition (1787) he slightly alters the formulation: ‘Space is represented as an infinite
given magnitude.Therefore the original representation of space is an a priori intuition, not
a concept’.
125
This leads to the essentially important claim: ‘we can accordingly speak of space,
extended beings, and so on, only from the human standpoint’.
126
A separate 1786 essay on
orientation, that is literally, finding the sunrise in the sky, Kant notes that ‘even with all the ob-
jective data of the sky, I orient myself geographically only through a subjective ground of differ-
entiation’.
127
Space and time are thus not absolute in anything like a Newtonian sense, nor does
Cartesian geometry tell us about the world as it is. Rather they are modes of access to the way
the world appears to us. Kant therefore concludes in a note to his own edition: ‘space and time
of course have objective reality, but not for what pertains to things outside of their relation to
a faculty of cognition, but rather only in relation to it, and thus to the form of sensibility, hence
solely as appearances’.
128
Several points emerge from this. One is that Kant’s view of space (and time) undergoes some
important developments, and that we need to be careful in taking claims in, and about, certain
periods, as indicative of his thought as a whole. The second is that the Geography, given over
several decades, is in part affected by these shifts. Preliminary investigations of the student
transcripts indicate this: the Holstein-Beck from 1758 to 1759 contains no theoretical discussion
of space and time and their relation to geography and history; the courses from the mid 1770s
have the more extensive introductory discussion drawn upon above (such as Ka
¨hler, which is
the basis for the Rink edition introduction); whereas a course from the last years, such as the
Wolter transcript from 1796, underplays the theoretical material but assumes its essential in-
sights. Third, and most important, that Kant’s mature view—the one that emerges in 1770
and is the view of the Critique of Pure Reason, and arguably that of the majority of the geog-
raphy lectures, is not an absolute view of space on the model of Descartes or Newton.
129
The
claim that Kant makes is that space, as this ‘conceptual fiction’, offers a way of insight into
how we perceive the world, and is open to scientific examination, rather than this determina-
tion being necessarily representative of space itself. In many places Kant recognises that di-
rections in space are a product of our body’s encounter with space rather than absolute and
self-evident.
In addition, as the work on space and time in the Geography indicates, Kant is trying to work
the dimensions of space and time together. Kant surely does not advance the kind of historical–
geographical materialism advocated by the likes of Harvey, or the ‘spatial history’ other geogra-
phers have suggested. But his separation of space and time is a product of theorising, when
in practice—either in our experience or in the relation of geographical details to historical
studies—they are continually intertwined.
125
Kant, Critique of Pure Reason (note 71), B39–B40. ‘Given’ and ‘a priori’ are in bold in the text.
126
Kant, Critique of Pure Reason (note 71), A26/B42.
127
Kant, What does it mean to orientate oneself in thinking?, in: Religion and Rational Theology (note 50), 9.
128
Kant, Critique of Pure Reason (note 71), note to A41.
129
Kant’s views on space are also amended in his Opus Postumum. For some instructive indications of where this
might go, see J. Edwards, Substance, Force, and the Possibility of Knowledge, Berkeley, 2000, especially 173–174.
23S. Elden / Journal of Historical Geography 35 (2009) 3–25
Conclusion
Given the renewed interest in the Anthropology and the links this material has to the Geography,
and the forthcoming translation of these lectures in the volume on Natural Science, this seems an
opportune moment to take stock of their historical importance and contemporary relevance. To
read Kant’s work on geography is an inherently interdisciplinary venture, since they cover both
human and physical geography and obviously relate to philosophical concerns. The issues raised
by these texts of Kant’s are textual and linguistic, philological and hermeneutic, philosophical and
political, even as we consider their relation to geography and the wider history of ideas. Only
a multi-disciplinary, and multi-handed, approach can do justice to their complexity.
130
For geog-
raphers these lectures have some fundamental challenges to offer. Most of it is indeed, as Hart-
shorne, following Adickes, suggested, of historical interest. This is the very reason why it is so
interesting. If nothing else, this account offered here is to show just how many difficulties there
are around the texts of Kant’s geographies and their meanings. The link to the anthropology lec-
tures and the question of race is a fundamental concern, precisely because it is so deeply problem-
atic, and because of the challenge Bernasconi and Harvey have thrown down to geographers and
others who wish to appeal to Kant’s cosmopolitan instincts. More positively, and in part going
beyond the Geography itself, there may be something of fundamental interest in the thinking of
the relation between space and time, geography and history. This is not to co-opt Kant into ar-
guments for the primacy of geography over history, nor for claims of their equal status in the
study of social relations, but rather to recognise that the tradition of Western thought has not
always had the imbalance it is often assumed to have had.
The argument made here could easily be extended to suggest that the other important geo-
graphical issues, such as the relation between the human and the environment, and the link or
distinction between the human and nature, might also be profitably be rethought from the per-
spective of a closer engagement with Kant’s critical philosophy as a whole.
131
It seems unques-
tionable, however, that a more nuanced understanding of these lectures and Kant’s work more
generally can only be a good thing for the history of the discipline of geography.
Note on references
Works available in English translation are generally referenced to the Cambridge Edition of the
Works of Immanuel Kant in Translation, which unfortunately does not have volume numbers. Ref-
erences are therefore given to volumes independently. References to the Critique of Pure Reason
are given to the standard A/B pagination of the first and second editions.
130
A forthcoming volume of essays (Elden and Mendieta (Eds), Reading Kant’s Geography (note 32)) attempts to be-
gin this process of interdisciplinary dialogue concerning the lectures. The book seeks to provide a range of essays dis-
cussing, contextualizing and criticizing Immanuel Kant’s work on geography; brings together scholars of geography,
philosophy and related disciplines to allow a broad discussion of the importance of Kant’s text for philosophical
and geographical work, both historically and in the contemporary context; and seeks to deal with some of the complex-
ities of this topic.
131
I owe the specific example to one of the anonymous referees.
24 S. Elden / Journal of Historical Geography 35 (2009) 3–25
For works not in English translation references are to the Akademie Ausgabe of Kants gesam-
melte Schriften, Berlin: Walter de Gruyter, 1900ff, referred to by KGS, and followed by volume
number and page. The Physische Geographie appears in Vol IX: Logik, Physische Geographie,
Pa
¨dagogik.
Student transcripts are quoted from the versions produced by Werner Stark, to whom I am ex-
tremely grateful for his kind access to as yet unpublished materials.
Acknowledgments
This paper was written as part of a project funded by a Leverhulme Research Fellowship and
a British Academy Conference Grant. This support is gratefully acknowledged. Along with
Eduardo Mendieta I organised workshops on Kant’s Geography in New York in November
2007 and Durham in January 2008, and I gave a version of this paper in Uppsala in
May 2008. I am grateful to the workshop participants and audiences for instructive discussions.
I would also like to thank Felix Driver and three anonymous referees for Journal of Historical
Geography for their comments on this paper.
25S. Elden / Journal of Historical Geography 35 (2009) 3–25
