OAC Press Working Paper 12: Thomas Sturm on Kant's cosmopolitan point of view in history.

This forum post is to advertise an upcoming OAC online seminar by Thomas Sturm:


What Did Kant Mean by and Why Did He Adopt

a Cosmopolitan Point of View in History?

The paper can be viewed here (as html or in a downloadable pdf ) and the seminar itself will begin on the 19th of March.



Currently in Anthropology, cosmopolitanism is a topic whose 'star continues to rise', as one commentator has recently put it. A number of international conferences, including the 2006 ASA meeting at Keele and a range of recent volumes, including a special issue of Social Anthropology in 2010, have highlighted the overlap between aspects of anthropological practice and central ideas in the philosophy of cosmopolitanism. At the OAC we have discussed Kant and cosmopolitanism in anthropological terms from the very inception of this web community; for example here.

Thomas Sturm's article contextualises elements of Kant's pragmatic and cosmopolitan viewpoint on history, placing it within discussions taking place in the late 18th Century. He points to tensions that existed during the enlightenment concerning the possibilities for a cosmopolitan history. Amongst these were central questions for historical inquiry such as: 'what is human nature?' and 'how malleable is it?' 'What constitutes a cause in history?' 'What part do human motives play in historical change?' Figures such as Herder came to the fore at this time to decry an enlightenment tendency to project current values onto the history of other peoples and epochs. Sturm proposes that Kant argued for a cosmopolitan viewpoint that simultaneously accepted human plasticity without relinquishing the claim for a universal human nature.

Our hope for this seminar is that it will create a target for fruitful cross-disciplinary dialogue around the idea of cosmopolitanism, help clarify Kant's role, and provide areas for future debate.


Thomas Sturm has recently published two articles closely related to the theme of this seminar:

(2008) Why did Kant reject physiological explanations within his pragmatic anthropology? Studies in History and Philosophy of Science, 39, 495-505.


(2011) Freedom and the human sciences: Hume’s science of man versus Kant’s pragmatic anthropology. Kant Yearbook, 3, 23-42.


(preprint version of this one: http://uab.academia.edu/ThomasSturm/Papers/844478/Freedom_and_the_Human_Sciences_Humes_Science_of_Man_versus_Kants_Pragmatic_Anthropology)

Views: 1791

Replies are closed for this discussion.

Replies to This Discussion

Thomas Sturm said:

 I'm unsure about the humanism/antihumanism divide in anthropology today, so I can't help here. I need more explanation what this has to do with the issue of the relation between "pragmatic" and "cosmopolitan" in Kant.

Thomas, Thank you for your scholarly and sympathetic answers to our questions. I don't want to exhaust you and would remind you that you are not obliged to answer every question in such detail. But since you end with a questionmark concerning our disciplinary discourses, I will just leave you with a pointer before I go to bed, so that we can resume the issue of the relationship between pragmatic and cosmopolitan history later.

Albert Camus wrote The Plague (La peste) when he had given up on active engagement in the anti-colonial revolution because he doubted if personal "activism" was an appropriate answer to a war that eventually took a million lives. In the novel a plague hits a town and people try to explain it in personal and moral terms, but its causes are much wider than that. This is what I mean by anti-humanism. In the next decade, French structuralism (influenced by American systems theory) took out the subject, dialectic and even history from Marxism as so much outmoded German philosophy. Deleuze and Guattari developed this strand and their influence on anglophone anthropology, along with various strands of post-structuralism, has been considerable.

John was referring to this aspect of what I wrote. How is it possible to write a universal history which operates in terms that are personally meaningful to human agents? I don't say it can't be done or that Kant doesn't have some interesting anwers, but this is the first question that would occur to most contemporary anthropologists and it touches on widespread suspicion of Enlightenment "grand narratives" (including Kant's) which are taken by postcolonial theorists, for example, to have underpinned western imperialism.

The founders of modern ethnography referred to the tradition that they overthrew as "conjectural history" and that critique is never far below the surface. This is a good reason to take up the example you cite of Hume's History of England. In his touching brief autobiography, Hume tells how disappointed he was that his first two major philosophical works bombed, while his History was a best seller, making him financially comfortable. It is clear that his ability to tell the story of English kings (whose individual names head each chapter) as a vehicle for universal history was obviously prepared for by his philosophical work. This makes the source of the grand narrative something other than observation of human action and motives. Shakepeare perhaps did a better job of presenting the dialectic of person and structure in ways that a live audience can readily appreciate.

If Hegel mounted, as I would argue, a counter-revolution against Kant's liberal revolution, it took the form of disparaging the limitations of Kant's notion of Moralität, a species of bourgeois individualism that could never grasp the movement of societies in history. My usage may be idiosyncratic, but that is what I mean by humanism. In any case, western social science has taken Hegel's line ever since and that is why Kant's work is routinely ignored or, worse, put down superficially. Like you and Huon, I feel that this is a bum rap, if you will pardon the expression, and I look to you and your colleagues to show me more clearly why.

Dear Thomas, you make some good points here. But to clarify further discussion, it may be useful to distinguish between two perspectives on Kant.

To the Kant scholar, what Kant did or did not say is the heart of the matter. To this anthropologist the more interesting question is what in Kant may be useful to understanding the empirical data my research reveals. Here I follow the lead of the anthropologist Victor Turner, with whom I once had the privilege of studying. Vic remarks in an essay to which I have often returned that a theorist's ideas are useful in the field only when they illuminate something about the lives of the people among whom the anthropologist is working — which often occurs only after the ideas have been disentangled from the "logical sludge" in which we find them embedded. It is often the chance remark or unexpected angle that proves more useful than the system from which we extract them.

There is also the question raised by Clifford Geertz, whether there is, in fact, a primary human nature revealed when the "second nature" of culture is stripped away. This onion model of human nature in which deeper layers are seen as truer or more fundamental than the external layers stripped away to reach the core is, he suggests, misguided. I would agree.

I would not, however, on that account, reject the possibility of a general, natural history of humanity in which cultural differences, considered as data, are of no greater or less significance than, for example, details of human anatomy are to the great evolutionary story that begins with the solitary tunicates and continues through Chordata to Vertebrata to Mammalia to Primates, Hominids and, to date, Homo Sapiens. What may be useful here is the observation that biologists have long since transcended sentimental and simplistic theories of unlineal progress converging on a Man created in the image of his God and made considerable progress in understanding the genetic, ecological and other mechanisms that shape the immense variety of the world we inhabit. That is an area in which cultural anthropologists, forever trapped in debates over nineteenth century antinomies, have singularly failed to make progress.

Thomas, I am very interested in the normative side of Kant's cosmopolitan project partly because anthropologists often say 'well it is not my business to make grand meta-narrativistic historical or sociological claims - If I am in favour of anything it is local ontological self-determination' (or similar).

Again, one of the aspects I see in Kant is that there is a general principle of self-determination. So, for instance, he complains about students who know exactly what Wolf says but don't have any idea of how those ideas connect to their own experience or how to use them judiciously and so on. The same could be said of some anthropology students who can talk endlessly about Derrida or Bourdieu or Foucault but have no idea how the ideas involved connect to their own experience or what the implications of actually using those concepts in 'real' life would imply. A main theme is self-awareness and self-determination - using concepts in a way that makes sense for the individual life and so on rather than mimicking whatever is fashionable or available. In addition there is the attempt to enlarge one's thinking - to live with an 'enlarged' mind. 

As regards world history there is arguably an expectation that anyone who wants to enlarge their thinking should be able to see the scope of their own life in terms of the interconnectedness of human life in space and time, though this sense could take many forms. So, each person should morally be engaged in creating a 'world' history-geography for their own life; not simply imbibing someone else's pre-prepared history or geography. To me, there are some links here with some more complex aspects of what anthropologists' call reflexivity (generally the idea that the anthropologist is a participant in what she observes and theorises). 

Qua world history in the standard sense, for example, Adam Ferguson is arguably a more interesting and in certain ways more perceptive historian than Kant because he tends not to fall back on Nature and Providence as much as Kant does and he has an almost Darwinian awareness of chance, spontaneity and random change in social arrangements. What Kant is adding is a demand that everyone should take responsibility for their own standpoint on the world. (there is of course the huge dilemma that everything apart from basic human subjectivity is laid out unequally in time and space).

Again, I see in Kant a view that human freedom and cosmopolitan awareness are linked to the contingency of the individual life, though I am quite influenced by Hannah Arendt in this. Because individuals are born at a contingent moment in time and their life span is also contingent their thoughts and decisions are ultimately their own in this rather specific way - whether they like it or not the judgements they make are 'spontaneous' and self-guided. This is a reading that we could find in some late 19th neo-Kantians like Simmel who see philosophical anthropology as a highly personal task where one is leaving the door open to other people to create their own anthropologies.

But of course this is not the positivistic model of social science that we have been used to - and I admit it is not necessarily what Kant would have intended either - perhaps it is more 'post-modern'. Either way, the positivistic model has hit a dead end in all the social sciences, where, despite claiming some kind of value free basis for what they do, all the social sciences have been involved in importing various metaphysical ideas and moralities into their descriptions. So, finally, what strikes me about Kant's normative cosmopolitanism is the very drastic claim it is making that each individual whether they like it or not is in some way responsible for everyone else and that the job is to work out, on their own behalf, for themselves, how. This idea, whatever else it does, certainly cuts through the intellectual clutter.

Either way, the positivistic model has hit a dead end in all the social sciences, where, despite claiming some kind of value free basis for what they do, all the social sciences have been involved in importing various metaphysical ideas and moralities into their descriptions. 

For what it's worth, this strikes me as one of those pious platitudes that needs closer examination. That social science imports metaphysical ideas and moralities into its descriptions is true. That this implies that the social sciences have hit a dead end is false. 

I say this having just returned to Japan from Los Angeles, where I participated in the 32nd annual Sunbelt conference of the International Network Analysis where I met a number of very smart people who are happily advancing the state of the art in their branch of the social sciences—both in terms of computational tools and methods and in terms of the richness of the data on which those tools and methods operate. Are they the end all and be all? Of course not. But of progress there is plenty.

There is, in addition to data mining, which reveals demonstrable patterns in large data, the construction of agent-based models of the sort to which an introduction is provided in the free course by Scott Page to which Jacob Lee and I have pointed elsewhere. Yes, it is impossible to construct an agent-based model without making assumptions about how the agents in question see the world and make judgments about how to act in it. Like the human beings they mimic, they, too, require models-of their worlds and models-for behavior — a metaphysics and morality, if these terms be preferred. Yes, current examples are still simplistic. But those who construct them are aware of that and gradually making progress on finding ways to improve them. There may be a limit to that progress, but to claim that it has already been reached is palpably absurd. 

My own presentation was a case study using some of my Winners' Circles data. In it my goal was to explore the relationship between social network analysis (SNA) and historical and ethnographic research (HER). It began with something that seemed to me utterly miraculous. Applying a technique called p-Core analysis to a network of 7,018 creators and extracting the top 45 in terms of their p-Core vector scores generated a diagram of relations among top creators in which I instantly recognized the history of Japanese advertising during the period the data covers. I realized that I knew or knew about most of these individuals. I didn't, however, know all of them. The analysis revealed a highly central individual who had been entirely off my radar — which required me to think about why that was so. In this respect, SNA surprised the ethnographer. The ethnographer, however, also had a surprise for the network analyst. An interview with one of the two most central figures in this subnetwork pointed to one of its peripheral members as a critical contributor to that most central figure's career. The p-Core scores did not capture the importance of this relationship.

What the exercise had demonstrated clearly was that SNA and HER have much to say to each other. One provides a way to quickly and efficiently parse a mass of data unintelligible to the unaided eye, providing, in effect, a social science microscope that highlights structures of interest embedded in what otherwise are unreadable hairballs of information. The other enriches our understanding of the relationships that constitute those structures and the critical paths by which they were formed. And each surprising the other with an unanticipated finding? That, not some metaphysical truthiness, is what science is all about. That we don't know all the answers yet and probably never will. That's what keeps it fun.

This is fun, John, but what has this got to do with the seminar paper?

Sorry, one leap too long. The connection for me was in my previous comment, the snarky remark that cultural anthropology, trapped in debating 19th century antinomies, has failed to make progress in the way that biology, for example, has. The biggest of the traps in question appears to me to be setting up a binary opposition between concrete, local ethnography and abstract theory, instead of pursuing the question whether our theories can explain the range and variety of ethnographic particulars. Perhaps Thomas can tell us whether Kant remains caught in that opposition or points to a useful way out of it. Serendipitously, I have in my Kindle collection Michael Fox's The Accessible Hegel, which suggests that Hegel might be a better choice than Kant as far as I am concerned.

Dear Thomas,

Of course I didn't mean to say that Kant's references to the shape of our home planet in the Metaphysik exhausts his writings on natural history. My point is twofold: (1) we can move beyond Kant's own intentions as an author and note that his Enlightenment interest in the natural order leads him to assigning a certain place to human beings in this order; (2) we can even infer that his natural philosophy not only precedes his anthropology and ethics, but remains a fundament on which he built the rest of his philosophy.

As for Kant v. Hegel (or why not Aristotle v. Plato): it is possible to trace versions of anti-humanism in both. (By the way, Keith, do you take 'structuralism' and 'structural analysis' to be synonymous with anti-humanism?) In contrast to Kant, Hegel begins by saying that everything that is must be true because it is (Latour now claims that Tarde came up with this idea). After twisting and turning for decades, Hegel returns to this point, now with the added insight that Sein is ordered by Fürsein. Human history is the history of realization (in both senses). Is this anti-humanism? I'm not sure.

Kant, on the other hand, defines the Enlightenment as 'man's emergence from his self-incurred immaturity’. Already Horkheimer and Adorno pointed out that, on this view, the most liberated self, which uses its mental faculties without dictation by someone else, may be found in the writings of the Marquis de Sade. Isn't Juliette, the cruel and criminal libertine of Sade's novel with the same name, the epitome of the Enlightenment? Famously, Lacan and Žižek have pointed to other continuities between Kant and Sade: we do not only find sadism in Kant (‘the Kantian Law is a superego agency that sadistically enjoys the subject's deadlock, his inability to meet its inexorable demands’) but Kantian ethics in Sade (evil accomplished out of principle). Is this anti-humanism? Again, I'm not sure.

Nonetheless, I agree with Keith that anti-humanism is (or should be) the number one concern in contemporary anthropology.

Dear Ronald,

Could you say a bit more about what you mean by "anti-humanism" when you say that it should be the number one concern in contemporary anthropology? Are you talking about theories that devalue and disempower the human individual by reducing social life to rigid laws or mechanical forces, Calvinist predestination reconstructed as physics? 


Sorry to have wandered so far away from the focus of your paper. The truth is I know of Kant, not at all the same thing as actually knowing Kant. My impressions are along the following lines.

1. Kant responded to Hume's empiricism by insisting that inferences (causal, moral, aesthetic) require premises not given in sense data themselves. This seems entirely valid. Computers require programs to process incoming data; brains require neurological structures to process sensory input. Should human minds be any different? I shouldn't think so.

2. Kant found those necessary premises in transcendental categories, presumed to be inherent in the structure of mind. Since, however, they were not inherent in sense data, they could not be inferred by the usual forms of induction and deduction. They could only be discovered by transcendental deductions—which I take to be logically equivalent to what mathematicians call proof by contradiction. 

3. It would appear, however, that Kant mistook proofs that transcendental categories are necessary for proofs that particular forms of such categories, the axioms of Euclidean geometry, for example, were inherent in all minds. We now know that this is not true. There are, continuing the example, a variety of non-Euclidean geometries. 

4. Given the multiplicity of possible transcendental categories, it is possible to conceive of a history in which human beings shift from one set to another. Here time is a good example, with the gradual evolution of time from definition in terms of rudimentary repetitive motions (sun rises, sun sets, for instance) to the abstract linear time of Newtonian mechanics, to the flexible time of General Relativity. Self, moral horizon and the nature of the good and the beautiful may also be good examples, or so it would appear in such works as Charles Taylor's Sources of the Self. And, then, of course, there is

5. Durkheim, who argued that both the categories and the differences we find in them from one people and place to another are rooted in the organization of society—a third set of social facts distinct from both mind and sense data. The virtue of Durkheim's proposal, when seen as a springboard for ethnographic investigation and anthropological theory, is that it offers a possible explanation for the variations documented by Taylor and others.

This, then, is the sum and substance of what I think I know about Kant, which I'm sure to a proper Kant scholar is nothing short of pitiful. I look forward to learning where I have erred and what I am missing.

Many thanks for the contributions of yesterday. I was busy and so can reply only today; but this has an advantage, because the discussion has developed its own dynamics, and that helps me to assemble and structure my replies.

Before I begin, please allow me one methodological remark. There was some name-dropping in the contributions of yesterday, sometimes with a bit overly sketchy hints at their views and arguments. I would like to resist that as far as possible. I doubt that we can assume that we already know what the authors mentioned have meant, let alone that we agree on their views granted that we know what they were. I have been trained as an analytic philosopher. Even when I study the history of philosophy and science, I look for arguments and possible counter-arguments. I reconstruct texts in as argumentative ways as possible, in order to learn what their precise meaning was - at least part of what one means when making a claim follows from one’s reasons for the claim), and in order to make it possible that we relate them to current debates without anachronism. This allows for an integration of serious historical work with a serious interest in what to learn (if anything) from thinkers of the past for our current problems inside and outside of philosophy (thus, John’s “two perspective” remark does apply to many Kant scholars but I hope not to me). So, here is my request: Perhaps we can at least sketch the argument – stating the conclusion and telling us what premises are supposed to support it – of the author introduced, be it Kant or Hegel or Tarde and Deleuze, Foucault and Latour, Lacan and Sade, and make explicit what to think of the argument, and otherwise avoid mentioning them. I am aware that in many academic corners conversation follows different rules, but I can only provide reasonable answers of reasonable length (thanks for the reminder, Keith!) if I know what’s really meant and on what grounds it is. Every additional name-dropping increases the complexity of our already very rich debates. (Please accept that I my remarks are not meant as engaging in familiar polemics between analytic and continental philosophy or the like. I sometimes have learned from Heidegger, full stop.)

The next reply will contain my attempt to structure our debates, and reply to them. Many thanks for your patience!

Thanks for the call to order, Thomas, and it is of course you who has been patient not the other way around. 

Yesterday's contributions concerned a current predicament of anthropology, which was described in different ways. (1) Some hope that Kant can help current anthropologists to overcome a “Hegelian” fixation that is said to dominate social science (Keith). (2) The normative side of Kant’s cosmopolitanism is seen as an antidote to a positivistic model in the social sciences which is said to have hit a “dead end” (Huon) – a characterization of the state of the art the adequacy of which was called into doubt, however (by John). (3) Biology is expected to help anthropology to get out of a predicament, namely the conflict between telling a grand narrative or abstract theory and pursuing its research in the standard way, studying local customs. (John) - To some extent, these views were connected to the topic of humanism over anti-humanism in anthropology, but I shall resist from engaging in this additional complexity until I read a clear statement about what the doctrines are.

(1) As to the Hegelian influence upon social science, it seems that the charge against Kant is that the latter erected an ideology of bourgeois individualism under the name of a universalist ethical theory (as Keith reports – I understand, however, that Keith doesn’t subcribe to this view). There is the additional remark that Kant even erected impossible things – citing Lacan or Zizek, Ronald writes that “Kantian Law is a superego agency that sadistically enjoys the subject's deadlock, his inability to meet its inexorable demands”.

First, the last point is pretty unacceptable. Kant’s ethics contains the important and reasonable principle of “Ought implies Can” – a moral theory should be livable. No doubt moral demands can be challenging, but that’s precisely their point. (A point, by the way, which in the Critique of Practical Reason is illustrated by means of a thought experiment, alluding to an actual case described perhaps in Hume’s History or other historical accounts (I have to check): When Anne Boleyn was charged of adultery against Henry VIII, the courtier Henry Norris denied he had an affair with her, and was found guilty of treason and executed - and he knew that this would be the most likely outcome of his decision.)

Second, Kant was perfectly happy to accept that the social sciences have to start with the undeniable fact of the diversity of human morals or customs; Hegel’s opposition of “Sittlichkeit” and “Moralität” is overstated. Among other things, Kant’s ethical theory does not completely determine all aspects of the moral life; it mostly rules out certain actions and maxims. Therefore, it is compatible with many (though not all) different directions the development of individual and society may take over history. 

(2) I agree with Huon’s remarks that Kant’s normative cosmopolitanism, like his ethical theory, is closely connected to the ideas of self-regulation but also with the demand to broaden one’s horizon. I’m not sure this is incompatible with a “positivistic” model of social science – I would first need to know what’s meant. But I have some sympathy for John’s remarks here.

(3) I think that’s a nice suggestion. While anthropology (like many other social sciences) has shown a trend towards “localism” or microstudies, what may have been missing is a complementary account of why certain customs or behaviors did not only fit into their historical or geographical environments, but also how all such “relativism” is perfectly compatible with accounts of human nature that are more universal.

That's for starters. Please hit me with further questions and objections, and I'll be happy to explain further.




OAC Press



© 2018   Created by Keith Hart.   Powered by

Badges  |  Report an Issue  |  Terms of Service