Kant, other minds and intersecting issues...

Interview by Richard Marshall.

'If you deny that Kant thought that perceptual consciousness constitutively depends on the understanding – if you think instead that Kant allowed that we could be perceptually presented with empirical particulars independent of the use of any concepts – then that looks to make possible a rapprochement between Kant and naïve realism.'

'People perceive patients in a persistent vegetative state (PVS) as having less mental capacity than the dead, that we think of them as ‘more dead than dead’. That’s surprising, to say the least. Don’t we ordinarily think of death as bringing to an end a subject’s mental capacities? In which case, how can there be people who have less mental capacity than the dead?'

'Imagine finishing Anna Karenina and being asked: does Anna really care for her son? In answering that question, we think back to the characterisation of Anna’s relationship with her son; we think of the things she does and whether they can be explained as resulting from her love for him. We can disagree about the answer. But it would be odd to answer ‘No’ to the question on the basis that Anna dies at the end of the novel, so is therefore incapable of loving anyone.'

Anil Gomesis Fellow and Tutor in Philosophy at Trinity College, Oxford and an Associate Professor in Philosophy in the Faculty of Philosophy in the University of Oxford. His main research interests are in the philosophy of mind and Immanuel Kant’s Critique of Pure Reasonand, in particular, on issues which arise at their intersection. Here he discusses Kant's uninterested in knowledge of the empirical world, the Transcendental deduction, whether Kant was a naive realist, some issues of other minds, the role of testimony in this, the mental states of people in persistent vegetative states, and whether philosophy - and Kant - has anything to offer in the domain of studying the mind. Start off the new year with a leap into the funky depths...

3:AM:What made you become a philosopher?

Anil Gomes:The idiosyncracies of the British educational system, as much as anything else. When I applied to universities, you were only allowed to make six choices, one of which could be Oxford or Cambridge. I wanted to study politics – I’d been brought up in a left-wing household in a city which prided itself on its radical heritage – so five of my six choices were for politics courses. But at Oxford the only way to study politics was through a degree in Philosophy, Politics and Economics. So I applied for that, and started the degree expecting to concentrate on politics. I soon found that I enjoyed and seemed to be good at philosophy. (It helped that my first teachers were Bill Child, Ian Rumfitt, and Hanna Pickard – I can’t imagine a better introduction to philosophy.) So I concentrated on philosophy, and enjoyed every moment.

Still, it wasn’t clear to me then that I wanted to pursue a career in philosophy. After my undergraduate degree, I went to work in the Houses of Parliament for a Labour MP, which was great fun. I returned to Oxford to do my graduate work, then left again after my doctorate to work in the civil service. After a while I decided I wanted to come back to philosophy, and was lucky enough to get a job at the wonderful Birkbeck College in London. I often hear philosophers say that you should only do philosophy if you can’t imagine doing anything else. This tells you more about the imaginative capacities of philosophers than anything important. (They really can’t imagine doing anything else?) It’s a kind of self-congratulation which drives out from the discipline people who don’t see this as a vocation. And it ignores the way in which some of us find it difficult to imagine ourselves in a discipline which is so overwhelmingly white. I love my job very much, and feel very lucky to be able to spend so much of my time talking with smart, interesting people. But I don’t want to do this forever, and I hope I’ll get to do something else at some point.

3:AM:You’re interested in philosophy of mind and Kant. Can you sketch for us how Kant helps us understand how we come to get knowledge of the empirical world, one which we conceive of as existing unperceived?

AG:There’s a sense in which Kantis very uninterested in knowledge of the empirical world. The Cartesian sceptic only makes a brief appearance in the first Critique, in a small section called ‘The Refutation of Idealism’. There’s even a sense in which Kantis very uninterested in knowledge in the first Critique. His main interest is Erkenntnis – a German term which used to be translated as ‘knowledge’ but which is now usually translated as ‘cognition’. It’s not clear exactly what Erkenntnis is – some kind of object-directed thought, perhaps – but, at the least, we shouldn’t be too quick in identifying it with the kind of propositional knowledge which has been the focus of so much contemporary epistemology. (Karl Schafer, Eric Watkins and Marcus Willascheck, and my friend Andrew Stephenson and I have written on this.) It’s cognition which Kant finds puzzling in the first Critique – and, in particular, a certain kind of cognition which he takes himself to have been the first to identify: cognition of synthetic, a priori truths. Empirical cognition and empirical knowledge just don’t seem as puzzling to Kant, so they’re not the focus of his interest.

But this is not to say that Kant has nothing to say about our ability to think and know about the empirical world, and about the sense in which we can think of it as existing unperceived. On the face of it, it’s kind of odd that we can think of things in this way. ‘It’s still faintly surprising, this rigid fidelity of objects, sometimes resassuring, sometimes sinister’, as Ian McEwan puts it in Saturday. And it’s hard to see what would give us ground for thinking of the world as existing unperceived, particularly if our grounds for doing so are meant to come from experience. Kant seems to offer us an alternative which explains our capacity to think of things as existing unperceived in terms of our possession of certain a priori capacities. That’s a really interesting suggestion, which promises to make clear how we can think about objects existing unperceived without extracting the conception of an unperceived object from experience. It’s a claim which was influential in a certain part of the philosophy of mind. Gareth Evans, for instance, holds that our capacity to think of things as existing unperceived draws on a primitive theory of perception which is not itself derived from experience. That looks similar to the suggestion you find in Kant.

3:AM:Kant’s claim about the Transcendental deduction is that the categories must apply to experience, not that we must apply them. Can you sketch what this is about and what’s at stake in this? Why isn’t it enough to leave it at the more limited claim that we must apply them?

AG:The Transcendental Deduction of the Categories is at the heart of the first Critique. Philosophers like to compare it to a jungle which you have to hack your way across, or a desert which you have to cross without water. (This is the scholar, alone in the library, imagining herself as Indiana Jones.) Kant is aiming in this section to legitimate our use of a certain set of a priori concepts, the categories, a set which includes the concepts of and . One very natural way to think about the Deduction is as responding to a challenge raised by Hume. (Kant calls the Critique of Pure Reasonthe ‘elaboration of the Humean problem in its greatest possible amplification’.) On this way of reading the Deduction, the problem posed by the categories is that they are not given to us in our sensory apprehension of the world: there is just no such thing as an experience of causation, or an experience of substance. This raises a question of how it can ever be legitimate to apply the categories to the objects given to us through sensory experience. If we’re not presented with objects as instantiating the concepts of or , why think that we have grounds for applying those concepts?

Kant’s answer to this question turns on showing that the categories are necessary conditions on experience. But what is that supposed to mean? On one way of cashing this claim out, the categories are necessary conditions on experience because we have to apply the categories to the objects given to us in experience. That would be interesting, if it were true. But it doesn’t seem like it would justify our use of the categories. Because applying a concept is different from that concept being instantiated. And just because we have to apply the categories to the objects of experience, that doesn’t show that they are correctly applied. Application is distinct from instantiation and it seems to me that if we want the Deduction to justify our use of the categories, then it has to aim at the stronger claim and show that the categories are really instantiated. I worry that some readings of the Deduction don’t meet this constraint – that they don’t show anything more than that we must apply the categories. But you might respond to this worry by denying that the Deduction is really taking aim at Hume, or by holding that there’s no real difference between the necessary application of a concept and its instantiation. I don’t think those options are plausible, so I think that the distinction between us merely applying the categories and the categories having to apply is a useful one to wield in assessing reconstructions of the Deduction.

3:AM:Did Kant think we needed concepts to perceive? It seems strange to even think of him as some sort of naïve realist.

AG:It does, doesn’t it! And there’s a history to that strangeness. Naïve realism, as I understand it, is a claim about the phenomenal character of visual perceptions: it claims that visual perceptual consciousness involves primitive relations of apprehension which hold between perceivers and some aspects of their environment. We can trace naïve realist views back to the birth of analytic philosophy. G.E. Moore and Bertrand Russell in Cambridge and John Cook Wilson and H.A. Prichard in Oxford reject the idealism of their predecessors in favour of some form of naïve realism about perceptual consciousness. It’s common for textbook nativity stories to present this revolution as the rejection of neo-Hegelian idealism. But for these philosophers it was primarily a rejection of Kantian idealism – and, in particular, a rejection of Kant’s views about the way in which the mind is active in perceptual consciousness. So from its beginnings, naïve realism was taken to be incompatible with Kant.

Why are they incompatible? Well, Kant seems to think that the understanding – that active, concept-using part of our mind – is involved in perceptual consciousness and one natural way to explain this involvement is by taking Kant to hold that perceptual consciousness constitutively depends on the concept-guided activity of the understanding. At least, that’s the dominant view. And if perceptual consciousness constitutively depends on the concept-guided activity of the understanding, you might think that it can’t be understood as involving primitive relations of sensory apprehension. That was the view of the early twentieth-century realists, and their response was to reject Kant.

But does Kant really hold that the understanding is involved in perceptual consciousness? That’s the traditional reading, I think, but non-conceptualist interpreters of Kant have challenged that view in recent years (Robert Hanna, Lucy Allais, Colin McLear, and others). And if you deny that Kant thought that perceptual consciousness constitutively depends on the understanding – if you think instead that Kant allowed that we could be perceptually presented with empirical particulars independent of the use of any concepts – then that looks to make possible a rapprochement between Kant and naïve realism. That’s one kind of motivation for some of these non-conceptualist interpreters since some of them use naïve realist type views to explicate Kant’s account of perception. The cost is that they sometimes end up denying that Kant allowed the understanding any role in perceptual consciousness at all.

My oxymoronically-titled paper ‘Naïve Realism in Kantian Phrase’ tries to find a middle ground here, by showing how one can think that visual perception involves primitive relations of apprehension, and nevertheless allow that the understanding is necessarily involved in perceptual consciousness. It’s an attempt to show how the insight motivating the early twentieth-century realists – that perception is a primitive form of sensory awareness, whose objects are mind-independent bits of our environment – can be combined with the Kantian recognition that the perceptual consciousness of rational beings is necessarily structured by our rational capacities. Very roughly, reconciliation is possible once we recognise that since modal dependence doesn’t suffice for constitutive dependence, the understanding can be necessarily involved in perceptual consciousness without being constitutive of perceptual consciousness. My naïve realist friends seem to think that this hitches a perfectly good account of perceptual consciousness to an unnecessary Kantian framework, whilst my Kantian friends think that it sullies the insights of Kant’s theories by tying them to a stupid theory of perception. I like to think it offers the best of both worlds, but it probably brings with it the worst of each as well.

3:AM:This is all linked to Phenomenal particularism. Can you sketch out the position and is it the default position of the naïve realist?

AG:The naïve realist says that visual consciousness involves primitive relations of apprehension. This is often glossed as the claim that visual consciousness involves empirical objects as constituents. The thought is that when you perceive an empirical object, you’re related to the object in such a way that the object is literally part of your perceptual consciousness. As I said, naïve realism is a part of early analytic views on the nature of perception. But it disappears from the debate in the middle part of the twentieth-century, arriving back in the latter part of the century through the work of John Campbell, Bill Brewer and, especially, M.G.F. Martin. In a recent paper, Neil Mehta objects to views on which particular external objects are sometimes part of the phenomenal character of perceptual experience. He calls this view phenomenal particularism. Naïve realists endorse this claim for a certain set of experiences – namely, visual perceptions (and perhaps perceptions in general). So naïve realism is one form that phenomenal particularism can take.

3:AM:Are you sympathetic to this position? Are you a defender of naïve realism?

AG:Certainly sympathetic. I’ve heard naïve realism and similar views described as a British disease and given my upbringing in Oxford and London, it’s probably unsurprising that I’m inclined towards the view.

My friend Craig French and I have tried to defend naïve realism against the objections that Mehta raises to phenomenal particularism. One of the things we try to do there is show just how many resources the naïve realist has for responding to various objections. One common kind of objection to naïve realism turns on the fact that there are experiences which seem to involve the same particulars, but which differ in phenomenal character. Mehta gives as an example the taste of wine versus the visual perception of wine. If you thought that perceptual consciousness was just a relation to an object, then it looks like you would have to say that both experiences involve the same phenomenal character (or, at least, some common phenomenal character). But although I glossed naïve realism above as the claim that perception involves relations of apprehension which hold between perceiver and object, in fact many naïve realists take perceptual consciousness to be a three-place relation which links perceivers, objects, and the point of view from which that object is perceived. This is a part of the early twentieth century views as well, and Craig and I try to show how naïve realists can appeal to differences in the point of view in order to explain certain differences in phenomenal character.

A related objection concerns what the naïve realist should say about the phenomenal character of non-perceptual experiences – say, hallucinations or imaginative experiences, the latter of which features in Mehta’s paper. Again, Craig and I are concerned to show the ways in which you can think of there being commonalities between hallucination and perception, or imagination and perception, which don’t turn on the experiences possessing a shared phenomenal character. In the case of imaginative experiences, for instance, we might think of them as representations of the phenomenal relations that one stands in when one is perceiving a certain scene. This can secure a kind of commonality which doesn’t challenge the naïve realist insight about the genuinely perceptual case.

Neil Mehta and Todd Ganson have a really nice response to our paper, which helps clarify what is at stake in the dispute, and we have a further reply. So there are still challenges for naïve realism to overcome. But I do think that there is an insight in naïve realism which makes it worth exploring seriously.