Enlightenment 2.0 (Interview with Thomas Metzinger)

COLLAPSE V Enlightenment 2.0 Interview with Thomas Metzinger Thomas Metzinger’s work is at the forefront of interdisciplinary research between the philosophy of mind and cognitive neuroscience. Marrying an encyclopedic grasp of the philosophical literature on consciousness with a superlative mastery of the latest neurobiological research, his Being No One is a groundbreaking work that recasts the terms in which the problem of consciousness is formulated. Although advances in cognitive neuroscience over the past twenty years have sparked a notable resurgence of interest in this problem among philosophers, many have argued that consciousness cannot be reductively explained by cognitive neuroscience, while others have gone so far as to insist that consciousness is a mystery that cannot be explained tout court, and that cognitive neuroscience lacks the resources to tackle ‘the hard problem’ of explaining how first-person subjective consciousness could ever arise from un-conscious neurophysiological processes. In our interview, Metzinger discusses not only how he confronts this philosophical challenge head-on by forging new conceptual resources capable of bridging this allegedly irreducible ‘explanatory gap’, but also how innovations such as the concept at the heart of his new theoretical framework – the ‘phenomenal self-model’ – might impact upon the personal and social experience of being human. 187 COLLAPSE V COLLAPSE: At the beginning of Being No One1 you criticize the tendency towards ‘arrogant armchair theorizing’ in what you call ‘analytical scholasticism’ in the philosophy of mind, a tradition which you charge with ‘ignoring firstperson phenomenological as well as third-person empirical constraints in the formation of one’s basic conceptual tools’. Rather than taking into account empirical constraints on concept formation, such neo-scholastic ‘armchair philosophers’ will often explicitly stipulate that it is our commonsense or ‘pre-philosophical’ intuitions which must ‘serve as the benchmark for a satisfactory account’.2 But while you agree that such intuitions ‘have to be taken seriously’, you are scornful of the idea that they should serve as the explanans rather than the explanandum in the philosophy of mind, pointing out that it may well be the case that ‘our best theories about our own minds will turn out to be radically counterintuitive’, presenting us with ‘a new kind of self-knowledge that most of us just cannot believe’. Though you suggest that ‘nobody ever said that a fundamental expansion of our knowledge of ourselves has to be intuitively plausible’, it is unfortunately the case that many philosophers of mind continue to act as if that were precisely the case. As Daniel Dennett has recently characterised the situation, philosophers of mind have typically proceeded ‘as if the deliverances of their brute intuitions were not just axiomatic-for-the-sake-of-the-project but true, and, moreover, somehow inviolable’. Whereas in most sciences 1. Thomas Metzinger, Being No One: The Self-Model Theory of Subjectivity (Cambridge, Mass.: MIT Press, 2004). The quotations in this paragraph are taken from pages 1-3 of the book. 2. Tim Thornton, Wittgenstein on Language and Thought (Edinburgh: Edinburgh University Press, 1998), 1. 188 Metzinger – Enlightenment 2.0 ‘few things are more prized than a counterintuitive result’, given that it ‘shows us something surprising and forces us to reconsider our often tacit assumptions’, in philosophy of mind a counterintuitive result is ‘typically taken as tantamount to a refutation’.3 In your own work, you are not content with simply reversing this order of priorities by taking the commonsensical intuitions and concepts of the folk and the philosophers as the explanandum rather than the explanans; you also aim to provide an account of precisely how such deeply ingrained intuitions, and the philosophical conservatism which follows from them, are ‘ultimately rooted in the deeper representational structure of conscious minds’. So for you, explaining consciousness and subjectivity is not simply a matter of ‘changing the way we talk’, but of forging conceptually convincing links between sub-personal and personal levels of description which are empirically rather than intuitively plausible. Can you say more about what this project of ‘interdisciplinary philosophy’ involves, and the kinds of ‘empirical constraints’ which have guided the formation of your own conceptual tool-kit? THOMAS METZINGER: Daniel Dennett is of course right – there is a certain deflationary Gestus in a lot of good, professional analytical philosophy of mind, and it sometimes blocks progress by simply being conservative without independent argument. There may be deeper evolutionary reasons for this attitude (quite obviously, our brains 3. Daniel Dennett, Sweet Dreams: Philosophical Obstacles to a Science of Consciousness (Cambridge, Mass.: MIT Press, 2006), 34. 189 COLLAPSE V are ‘coherence engines’, maximizing the coherence of our internal model of reality at any point, often sacrificing veridicality for short-term functional adequacy, with the ‘goodness of fit’ then being what determines the degree of intuitiveness of a given thought), as well as reasons based in the history and culture of analytical philosophy – which was itself a revolutionary enterprise in the past, a beautiful rebellion against academic pretentiousness and narcissistic obscurantism. I do not want to speculate about these reasons any further, but obviously, to talk in an unnecessarily ‘revolutionary’ tone and to cultivate counterintuitive interpretations of empirical results in order to achieve media attention (a strategy sometimes found in neuroscience) is not the solution either. It could be just another kind of corniness. Our resources are limited. My intuition (!) is that, in these exciting times, we should not waste too much time and energy investigating what exactly the relationship between philosophy of mind and, say, neuroscience, cognitive science or AI ideally should be. Rather, in this particular historical epoch, I think further epistemic progress can be generated by philosophy becoming partly embedded into interdisciplinary research programs, following a strategy which one of my best critics, Josh Weisberg, has called the ‘method of interdisciplinary constraint satisfaction’ (MICS).4 Part of the idea is to first confine ourselves to homing in on the target phenomenon in the actual world, using all sources of information available from every 4. See J. Weisberg, ‘Consciousness Constrained: A Commentary on Being No One’, Psyche 11 (5), 2005, and Metzinger’s ‘Reply to Weisberg: No Direction Home – Searching for Neutral Ground’ in Psyche 12 (4), 2006. These papers, as well as a Symposium on Being No One can be found online at http://psyche.cs.monash.edu.au/. 190 Metzinger – Enlightenment 2.0 single discipline researching the human mind, in order to compress the space of possible solutions dramatically. Then we can see whether, in the course of this historical phase, empirical results have actually changed some of our deeper, theoretical intuitions. Maybe then we might have a better idea of how intuitions evolved in the first place; what concepts actually are; why exactly certain forms of self-deception were adaptive and became superbly robust, spilling over into the enterprise of philosophy and science itself – or perhaps even what actually rendered unargued conservatism and corniness successful strategies in primate societies and the world of our ancestors. Let’s say we are interested in consciousness, or in a specific issue like non-conceptual self-consciousness: instead of kicking the problem upstairs into formal semantics and modal logic (i.e., by making modal intuitions explicit; intuitions which themselves have an evolutionary history), and instead of analyzing the sociohistorical accident of how we happen to talk about our own minds today, we should perhaps strive to invent a new form of philosophizing, which would be partly data-driven and conducted by philosophers who do not just talk about ‘first-person methods’, but actually are phenomenologically well-travelled. C: Cultivating counterintuitive interpretations of the results of the empirical sciences in order to grab media attention would be ‘corny’ indeed. Yet the fact remains that much of modern natural science is really in no need of any such deliberate ‘cultivation’ in order to bring out its ‘counterintuitive’ nature, quite simply because it explores aspects of reality to which our evolutionary ancestors 191 COLLAPSE V simply had no access, and in regard to which it would therefore not have been reproductively advantageous for them to develop any ‘intuitions’. This is a point, of course, which you develop at length in Being No One with regard to the likely evolutionary provenance of what you call our ‘naive-realistic self-misunderstanding’. Indeed, the ‘self-model theory of subjectivity’ which you formulate is not only something which is ‘radically counterintuitive’ for the time being, as it were – the kind of theory which is deeply unsettling at first but is destined ultimately to be absorbed into the culture at large – but is something which you insist will necessarily always remain counterintuitive. ‘Even if you are intellectually convinced by the current theory’, you write, ‘it will never be something you can believe in’.5 Thus, you insist that we can never come to intuitively believe, properly digest or internalize the fact that the ‘selves’ which we take ourselves to be do not in fact exist. But in this regard you have sometimes been accused of having cultivated an unnecessarily ‘radical’ or ‘counterintuitive’ interpretation of empirical results yourself. Of course, science frequently discovers that our commonsensical or folk conceptions of things are inadequate and so radically revises them. Our folk conceptions of space and time, for example, have undergone extremely profound revolutions, yet it is rarely claimed that space and time do not exist (even if there are certain avant-garde physicists willing to make precisely such a claim).6 Rather, the claim is more often that they are not what we, and previous science, took them to be. Thus a predictable and somewhat inevitable objection to your thesis would be that, in claiming that ‘no such 5. Being No One, op. cit., 567. 6. See ‘The View from Nowhen: An Interview with Julian Barbour’, this volume. 192 Metzinger – Enlightenment 2.0 things as selves exist in the world: Nobody ever was or had a self’,7 you are overstating your case, since your arguments only motivate a much weaker, revisionist position regarding the self, rather than the strong eliminativist one indicated by the title of your book. Why, then, do you feel it is necessary to insist that selves do not exist rather than that they are not what we thought they were? Why can’t we think of the project as one of revision and redefinition; of replacing our folk-psychological intuitions with the richer conceptual vocabulary provided by cognitive neuroscience; or of developing new and better-informed intuitions which are grounded in science? Can you say more about the ontological constraints that necessitate eliminating rather than revising the idea of ‘the self’ and replacing it with the theoretical entity you call the ‘phenomenal self-model’? TM: A lot of very good questions, for many of which I do not have a ready-made, official answer. First, if it makes you feel better (it will), to be a revisionist and come up with some weaker notion of selfhood, fine. The title of the book Being No One attempted to allude to many different things at the same time, and certainly wasn’t about the rather trivial and unoriginal point that the self is not a substance – Hume said that, Kant said that, Buddhist philosophy of mind had made the point long before, and few people uphold this ontological claim today. Yet, for the purposes of scientific psychology and neuroscience we can certainly do without the ontological commitment to ‘a’ self. 7. Being No One, op. cit., 1. 193 COLLAPSE V I do claim that everything we want to understand can be understood at the representational, functional, and dynamicist levels of description – and this really is not too breathtaking a point either, at least in my own view. The concept of ‘a’ self just is not a necessary ingredient of explanatory theories. We can predict what we want to predict and achieve the growth of knowledge we want to achieve with a more parsimonious conceptual framework. For example, relative to our interests and a given context, ‘self’ can be substituted by simpler notions like ‘PSM’ (phenomenal self-model), ‘SMT-system’ (self-model theory), or ‘MPS’ (minimal phenomenal selfhood).8 In other contexts, such as literature, art, or everyday communication, we may continue to use the notion of ‘self’ in order to drive and shape each other’s mental models of reality in a way that we find attractive. But since selfhood ultimately is a form of phenomenal content, we must take the phenomenology seriously – and the phenomenology is Cartesian; it is expressed by what I have called the ‘substantiality intuition’: We actually do experience ourselves as ontologically autonomous, indivisible entities, as something that can ‘hold itself in existence’ all by itself, and as something that knows about its own existence, with certainty, because in some hard to understand way it is ‘infinitely close to itself’. If we simply take the deflationary attitude again, or try to be good citizens and not exaggerate things too much by saying that, for example, this is just a straw man fallacy and then proposing some weaker notion of the phenomenal 8. See O. Blanke and T. Metzinger, ‘Full-Body Illusions and Minimal Phenomenal Selfhood’, forthcoming in the January 2009 issue of Trends in Cognitive Sciences. 194 Metzinger – Enlightenment 2.0 self, then we overlook the fact that the phenomenology itself is deeply substantialist. Of course, the phenomenology of substantial selfhood is not an epistemically justified claim, it is just a naturally evolved form of phenomenal content. But it must be taken seriously, as I have tried to do. Deflationary accounts cannot do that: Phenomenal selves are not particulars, simples, or individual entities, but dynamical processes, creating locally supervening forms of conscious content. C: Clearly, you are not simply reiterating the now familiar claim that the self is not a substance: as you point out, Hume, Kant, Nietzsche, and even Buddhism had already taught us that. Many philosophers, including critics of your views like Dan Zahavi,9 are perfectly willing to concede this, but only in order to conclude that the self is a process rather than a substance. And it precisely by redefining the self as a process that one avoids the stronger (and rather more interesting) thesis that there are no such things as selves. Yet it seems clear that you cannot be content with such a redefinition. Were Being No One merely proposing to redefine phenomenal selves as dynamical processes, this would be just a redescription, not grounds for elimination. Rather, your argument would seem to be that the self experienced at the level of phenomenal content is an effect generated by non-phenomenal, sub-personal representational processes. Consequently, the self cannot be identified with these processes themselves, since that would be to conflate content at the level of the representandum with its representatum. This would be why it is no more legitimate to redefine the self 9. See D. Zahavi, ‘Being Someone’, in Psyche 11 (5), 2005. 195 COLLAPSE V as a ‘process’ than it was to call it a ‘substance’. Would you agree that this is what distinguishes your thesis from the more familiar ‘self-as-process’ thesis favored by many? TM: My perception is that some who now jump onto the ‘self-as-process’ reading (which I offered myself in the very first paragraph of Being No One) perhaps only do so because this wording has a trendy ‘narrative’ ring to it – and this, in turn, would set the stage for more of what my excellent British PhD student (who is a bit chaotic, but from whom I learn a lot) has recently termed ‘Continental Jazz’. (I now use this new technical term all the time!) But there is no narrator, just selfless dynamical self-organization, and if the self is a process, then the least we must admit is that it is a very intermittent, patchy one – it comes to an end every night, in dreamless deep sleep. And waking up is obviously not performed by some transcendental technician of subjectivity pushing a magical reboot button in our brain – there is no subpersonal ‘narrator’ or ‘author’ piloting the conscious self. But it is also true that we do need good, modern neurophenomenology providing us with new conceptual instruments; a phenomenology that is not driven solely by ideological anti-reductionist resentment, but one that is at once historically and empirically well-informed. This is an extremely demanding challenge. But I am guilty of having indulged in ‘process’ talk as well. A process is a chain of events, and, roughly, events are property instantiations in some given domain, at a specific location in space and time. As long as we stay with the representational level of description, we may differentiate between the representational content of such a process 196 Metzinger – Enlightenment 2.0 (an abstract property), its carrier (concrete neural dynamics), and the phenomenal properties co-instantiated along with it (Shoemaker and others would say phenomenal ‘character’ here, I mostly say ‘phenomenal content’ in order to distinguish it from intentional content per se). ‘Selfhood’ is precisely such a phenomenal property. But please note that even the vehicle/content distinction is not a necessary one, because it depends on your chosen theory of mental representation. More empirically-plausible ones, like the connectionist model of representation, demonstrate how the content can be directly realized in the physics of the carrier – for example in the sheer synaptic connection strength between individual units. It is the connectivity itself that does the job, and we have a host of formal tools to pick out the abstract properties we are interested in – partitions of activation vector space, points in weight space, trajectories, and so on. My own position is that introspectively available phenomenal content is just a very small subset of the overall intentional or representational content systems like ourselves generate and use, namely that portion which is locally determined, because it depends on microfunctional properties realized by the NCC (neural correlate of consciousness). Consciousness is in the brain, but a lot of cognition certainly isn’t – and if you don’t believe that, you should read Andy Clark’s new book Supersizing the Mind.10 In the new TICS paper, Olaf Blanke and myself,11 based on our own experimental work attempting to create out-of-body experiences in healthy individuals, have 10. A. Clark, Supersizing the Mind: Embodiment, Action and Cognitive Extension (Oxford: Oxford University Press, 2008). 11. See note 8 above. 197 COLLAPSE V presented a framework for the simplest form of conscious selfhood. Self-consciousness is in the brain, a lot of selfrelated cognition certainly isn’t. In particular, agency is not a necessary condition for conscious selfhood, but self-location in time (presence), self-location in space (embodiment), and identification (with the transparent content of a body-image) seem to be the three decisive dimensions. So if I speak of a ‘process’ I envision an intermittent neural dynamics that can sometimes instantiate and functionally integrate phenomenal properties of this kind. The existence of content does not, of course, imply that there is reference as well. The subjective aspect of substantiality in this representational process could be entirely misrepresentational, a form whose character is entirely phenomenal. C: The Cartesian ‘substantiality intuition’ regarding selfhood which you have mentioned – rooted as it is, as you say, deep in the evolutionary past of our species, and forming the very content of our understanding of who we are and our place in the world – is certainly not one which is likely to be abandoned any time soon. Indeed, as we’ve already mentioned, it is extremely difficult to imagine how the idea that what people pre-reflectively take to be their innermost ‘self’ is in fact nothing other than a kind of ongoing phenomenal hallucination generated by their brains could ever take root at an intuitive level. Broadly similar ideas which were afloat in the psychophysiology and philosophy of the mid-to-late nineteenth century were compared by Friedrich Albert Lange to the impact of the Copernican revolution in astronomy: Just as it was once amazingly difficult for people to think of this fixed earth 198 Metzinger – Enlightenment 2.0 on which we stand, ‘the prototype of rest and stability’, as moving, so it would be all the more difficult for them to recognise their own phenomenal selves, ‘the prototype of all reality for them’, as ‘a mere scheme of representation’ generated by what Lange called their ‘psycho-physiological organisation’.12 In this regard, you have written that neuroscience may instigate a revolution in our self-conception more radical than any previous scientific revolution, a revolution which might have ‘greater social and cultural ramifications than any previous theoretical upheaval’.13 Could you explain what kind of upheaval you have in mind here? How do you envisage such a cultural and social upheaval given that, as you also insist, your self-model theory of subjectivity is constitutively indigestible for us at any intuitive level of lived experience, and so necessarily confined to the level of the philosopher’s conceptual understanding? What conditions would need to be in place for this ‘cultural integration’14 of your thesis to take place? TM: Again, a lot of very important questions, and I have tried to address some of them in a forthcoming non-academic book, The Ego-Tunnel: The Science of the Mind and the Myth of the Self.15 To cut a very long story short, neuroscience turns into neurotechnology, neurotechnology turns into phenotechnology, and in the course of the next fifty to 12. F. A. Lange, The History of Materialism (London: Routledge & Kegan Paul, 1925), Book 2, Section 3, Chapter 4, 206. 13. T. Metzinger (ed.), Conscious Experience (Paderborn: Ferdinand Schöningh, 1995), 13. 14.See Being No One, op. cit., 2. 15. T. Metzinger, The Ego Tunnel: The Science of the Mind and the Myth of the Self (New York: Basic Books, forthcoming 2009). 199 COLLAPSE V two hundred years such phenotechnologies will gradually invade our Lebenswelt, or ‘life-world’. Even given a culture of denial this may then actually have a lot of psychosocial consequences – which, however, are very hard to foresee today. We see the very first, early developments in neurotechnology today – cognitive enhancers such as modafinil,16 forensic neurotechnologies like brain-fingerprinting, and new types of lie detectors, or the development of new brainmachine interfaces, as in the case of the monkeys at Duke University controlling humanoid robots in Japan via the internet,17 and so on. A small subset of these new technologies is already used to directly control information flow within the NCC itself, for example by alleviating severe, treatment-resistant depression through direct deep brain stimulation with electrodes and so on.18 My point is that a small fraction of neurotechnology will slowly turn into consciousness technology and that this may have cultural consequences, even if we haven’t got the proper, comprehensive theory of the human mind yet. The ethically relevant class of actions could be defined as all those forms of practical intentionality in which the content-specifier consists of satisfaction conditions centrally involving phenomenal contents. Now, more and more actions become technologically available, which directly influence information flow in the NCC and directly determine phenomenal content. I’m sure many people will have heard about the full-body illusion 16. See Henry Greely et al., ‘Towards responsible use of cognitive-enhancing drugs by the healthy’, published online in Nature, 7 Dec. 2008: 702-5. 17. See http://www.sciam.com/article.cfm?id=monkey-think-robot-do. 18. See Helen Mayberg et al., ‘Deep Brain Stimulation for Treatment-Resistant Depression’ in Neuron, Vol. 45, 2005: 651–660. 200 Metzinger – Enlightenment 2.0 experiments carried out in Lausanne by Bigna Lenggenhager, Tej Tadi, Olaf Blanke and myself,19 and perhaps some have heard about Henrik Ehrsson’s brand-new variation of his own earlier studies in which you actually have the conscious experience of shaking your own hand in cyberspace.20 What if it actually became feasible to directly control an artificial body-model in VR with your brain and to fully identify with the avatar on the phenomenological level (a possibility the self-model theory predicts)? Many might spontaneously say this would be terribly ‘cool’ or ‘awesome’ – but for my part, I am not so sure. If one really tries to think through, in a more sober and critical fashion, some of the consciousness technologies that are now on the horizon, one finds that the challenges they present are not only ethical, but perhaps also challenges to our mental health. C: It is tempting to say that once such ‘consciousness technologies’ become possible, it will be very difficult for any ethical agency to prevent or control their deployment. But perhaps part of the motivation for your publishing a book aimed at a wider audience is precisely to pre-emptively open up a public discussion of these possibilities before they become reality. What did prompt you to write a nonacademic work like The Ego Tunnel, and could you summarize its contents for us? 19. O. Blanke, B. Lenggenhager, T. Metzinger and T. Tadi, ‘Video Ergo Sum: Manipulating Bodily Self-Consciousness’, Science, Vol. 317, no. 5841, 2007: 1096-9. 20. See V. Petkova & H. H. Ehrsson, ‘If I Were You: Perceptual Illusion of Body Swapping’ in PLoS ONE, Vol. 3, Issue 12, 2008, 1-9. See http://www.plosone.org/ article/info:doi/10.1371/journal.pone.0003832. 201 COLLAPSE V TM: I just wanted to try something new. Being No One was my third monograph. I had also done service for the scientific community by editing two or three anthologies focused on consciousness and by helping to construct an academic society (the Association for the Scientific Study of Consciousness),21 setting up an infrastructure for consciousness research. So I thought it could perhaps be interesting to move on from research to teaching and the public understanding of science, at least on the level of books. So I began working on a three-volume text-book called Grundkurs Philosophie des Geistes (a basic course in philosophy of mind covering mostly the second half of the twentieth century, which will eventually appear in an English version as well). I am now working on Volume 3, which is on intentionality and mental representation and scheduled for publication in summer 2009. Volume One, on phenomenal consciousness, and Volume Two, on the mind-body problem, already appeared in 2006 and 2007 respectively. The Ego Tunnel has the subtitle ‘The Science of the Mind and the Myth of the Self’. It is a popular book which will appear in March 2009. It addresses a wider audience, has a playful and experimental character to it, and attempts to explain an entirely idiosyncratic selection of those issues I personally judge to be most important in what I call the ongoing ‘consciousness revolution’, to the educated layperson. Of course, there is a quite a bit on self-models in there, with many concrete examples, and three interviews with prominent neuroscientists as well. The two final chapters address some of the consequences of these new scientific insights into the nature 21. See http://assc.caltech.edu. 202 Metzinger – Enlightenment 2.0 of the conscious mind-brain: the ethical challenges they pose and the social and cultural changes they may produce, given the naturalistic turn in the image of humankind. I close by arguing that ultimately we will need a new ‘ethics of consciousness’. If we arrive at a comprehensive theory of consciousness, and if we develop ever more sophisticated tools to alter the contents of subjective experience, we will have to think hard about what a good state of consciousness is. So yes, as you say, one of the functions of this project could be to draw the attention of a wider public to the relevance of neuroethics and normative issues that have been neglected. (Another ethical question with which I am deeply concerned these days is whether or not I should stop writing books now!) C: In his ‘Empiricism and the Philosophy of Mind’ Wilfrid Sellars suggests that eliminativist claims such as the claim that ‘physical objects are not coloured’ ought to be construed not as statements within the framework of common sense or the manifest image, but rather as challenges to the framework itself.22 Accordingly, is your claim that selves do not exist meant to be construed as part and parcel of a broader challenge to the manifest image, and if so, what else might need to be eliminated from it? After all, it would seem strange to say that it is only our folk conception of human beings as persons that is false; that beliefs and desires and selves do not exist but that the things about which we say we have beliefs and desires are adequately captured by common sense. So if we envision a future in 22. Wilfred Sellars, Science, Perception and Reality (Atascadero, California: Ridgeview, 1963), 172-73. 203 COLLAPSE V which we employ the language of cognitive neuroscience rather than folk psychology, wouldn’t this have to be part and parcel of a much broader overhauling of all our other folk conceptions such as our intuitive or commonsensical folk physics, folk biology, etc.? TM: Let me first give a ‘Continental Jazz’ type of answer to this question: Folk psychology is not only a theory, but also a practice. A lot of things will happen in science and philosophy, but the question of course is how this will actually change the practice, what it will do to everyday social interactions for example. As I pointed out above, although the way we speak about ourselves may only change slowly and gradually, the more likely prediction is that new technologies will invade our life-world and have an impact on the way we look at ourselves – because they directly change social interactions. The ‘manifest image’ is anchored in the phenomenal model of reality that evolved in our brains over millions of years, and it will certainly prove to be quite robust, at least in its transparent sensory and motor partitions. Social emotions and the opaque cognitive layers in our self-model, however, may prove to be more sensitive to a new cultural context. Actually folk psychology – understood as a pre-theoretical set of conceptual assumptions about reality – has already been completely demolished, just like folk physics. What will prove to be more robust is the pre-linguistic phenomenal model of reality anchored in our brains. And, functionally speaking, this is predominantly what drives our behaviour. There is a deeper aspect to your question, however: it pertains to the ‘self-as-epistemic-subject’. What many do 204 Metzinger – Enlightenment 2.0 not see is that there is also a sort of ‘folk epistemology’ – a deeply engrained, naturally evolved set of assumptions about what it means to know, what certainty is, how we mentally represent the world. If cognitive neuroscience also dissolves our traditional ideas about what a knowing self – an ‘epistemic agent’ – is, then this might also eventually have interesting consequences for the theory of science and philosophical epistemology. C: Do you foresee an increasing resistance to your attempt to integrate philosophy and neuroscience, given philosophy’s institutional complicity with servicing socio-cultural interests (morality, religion, liberal-democratic consensus, etc.)? What scope might there be for the kind of radically revisionary philosophical naturalism to which you are committed given how energetically some professional philosophers insist on providing alibis for folk superstition? Is the human race doomed to eternal resistance to such difficult and uncomfortable scientific truths? Or is such ‘difficulty’ simply based upon a misconception as to what theses like yours entail? What sort of cultural effects might we expect were they to be integrated into our everyday understanding of the world? TM: A sincere sensitivity to ethical issues, a commitment to liberalism and democracy – if taken seriously – are very radical affairs, not something old-fashioned, and if philosophical conservatism (you may be disappointed to hear) means keeping such sociocultural interest alive then we certainly need good, professional, analytical philosophy to 205 COLLAPSE V do this. The overall situation on our planet reminds me of one big ocean of irrationality, suffering and confusion – if more conservative philosophy’s ‘complicity’ with certain social institutions can help at least a little bit to improve our situation, then this is a good thing. I think, taken as a whole, humankind will prove to be absolutely robust in its resistance against the growth of knowledge, against what I sometimes like to call the ‘Enlightenment 2.0’ introduced by modern philosophy of mind and cognitive neuroscience, AI and AL, etc. There may be some rare individuals or even groups who find an interesting new synthesis of intellectual honesty and spirituality. But on a larger scale I believe two types of reactions are most likely to spread: A primitive form of hedonism, based on vulgar forms of materialism and a cynical, simplistic version of normative neuroanthropology; and the rise of irrationalism and fundamentalism, even in secular societies, supported by all those who desperately seek emotional security and espouse closed worldviews, simply because they cannot bear the naturalistic turn in the image of humankind. Both types of reaction are deeply human, and obviously, both are a danger to the very few stable, open societies we have established on this planet so far. C: We certainly don’t intend to denigrate the value of hard-won civil liberties or the democratic ethos: a sincere sensitivity to ethical issues and a commitment to democracy are indeed very grave and important matters, and certainly in no way ‘old-fashioned’. But the question remains whether social institutions shaped by a conception of human beings as maximally rational, self-determining agents – 206 Metzinger – Enlightenment 2.0 a conception with deep ties to a philosophical tradition that construes ‘selfhood’ as the defining feature of properly human being or ‘personhood’ – will prove adequate to the challenge of improving a human condition characterized by the undeniable surfeit of what you yourself describe as ‘irrationality, suffering, and confusion’. Might there not be a fundamental difficulty in reconciling the naturalistic understanding of selfhood as an inevitable ‘user-illusion’ generated by the human organism with a commitment to the irreducible worth of persons understood as the ultimate loci for the freedom and dignity attributed to the individual? Does the notion of ‘selfhood’, given its close ties to the concept of ‘personhood’, not serve to anchor the ethical, political, and juridical norms upheld by those who espouse liberalism and democracy? Can one really retain these norms while denying the existence of the ‘selves’ that were supposed to feature among the necessary conditions for human personhood and serve as the bearers of freedom and dignity? Can there be persons without selves? TM: Persons are not something we find out there, in objective reality. Persons are constituted in societies – through mutual acts of, say, acknowledging each other as rational and morally sensitive individuals. There can be selfless persons. Selfless individuals could certainly acknowledge each other as rational, and agree on a basic political and moral consensus. The problem is rather that we may turn out to be less rational than we thought, and less self-determining (whatever this might mean exactly) than our own previous anthropological self-idealizations have encouraged us to believe. My point is that (if this unargued empirical 207 COLLAPSE V speculation turns out to be true) it would be irrational as well as unethical in itself if we didn’t take such new empirical insights into account. But will we be rational and ethical enough to let new data about the evolutionary history and the neural underpinnings of rational thought and moral cognition themselves influence decision-making, say, in pedagogy or politics? I think you make a very good, and relevant, observation in describing why many of our best political theorists assume they must resist a naturalization of selfhood: They see the danger of bulldozing the normative dimension of personhood, of losing the ‘critical subject’; the dangers associated with an erosion of precisely what you call the ‘anchor’ of our ethical, political, and juridical norms. You cannot imagine how well I know these worries and antireductionist resentments! When I studied philosophy at Frankfurt in the late 1970s, it was not only trendy, but a frequently repeated and politically-correct commonplace move among we students to call the kind of stuff I and others do today ‘proto-fascist’. I was part of the more radical wings of the alternative movement of the late 70s, and when I sat in the seminars of Jürgen Habermas and so on all of this was certainly too superficial and not radical enough for me – at that time I thought that if you were really serious about getting in touch with political reality then you could do this by smelling some tear-gas, or by actually walking through shanty-towns in India. I certainly do respect Habermas and old-school political philosophy, but I have always thought that one has to dig much deeper in order to understand not only how the German catastrophe could happen, but why we seem unable to end all this violence and oppression, the 208 Metzinger – Enlightenment 2.0 constant turmoil on our planet that we call our ‘history’. I think it is a kind of intellectual tragedy that most of the best political philosophy is systematically shielding itself off from neuroscience, evolutionary psychology, and so on. It does damage to our discipline. C: This obdurate refusal on the part of most political philosophers to acknowledge any possible connection between the normative and the neurophysiological is indeed a kind of intellectual tragedy. But tragedy also veers into farce, as when the veritable phobia towards ‘scientism’ and ‘reductionism’ exhibited by Frankfurt School Critical Theory leads its adherents to issue moralistic denunciations of science and technology which are virtually indistinguishable from those of institutions like the Catholic Church. But critical theorists are certainly not alone in this regard. Returning to your earlier comments about analytic philosophy: We share your admiration for the analytic tradition’s commitment to intellectual sobriety, responsibility, and rigour. However, despite these virtues, it cannot be wholly absolved from responsibility for the kind of neo-scholastic ‘intuition-mongering’ critically invoked at the beginning of this interview. Thus we are less convinced that this tradition is entirely innocent when it comes to promulgating the kind of ‘closed worldviews’ which actively obstruct the growth of naturalscientific knowledge. While twentieth-century analytic philosophy may have started out in part, as you put it, as ‘a beautiful rebellion against academic pretentiousness and narcissist obscurantism’, once should not forget that it was also in many cases an avowedly anti-naturalist rebellion bent on preserving the autonomy and ‘purity’ of philosophy 209 COLLAPSE V by actively preventing its hybridization with empirical disciplines such as experimental psychology (a story minutely documented in Martin Kusch’s Psychologism: A Case Study in the Sociology of Philosophical Knowledge).23 Indeed, one might argue that the active resistance on the part of many academic philosophers to any miscegenation of philosophy with the empirical sciences ultimately has the unfortunate effect of helping to make the world safe for the irrationalism and fundamentalism of those who, as you put it, ‘cannot bear the naturalistic turn in the image of humankind’. In this regard, your use of the expression ‘Enlightenment 2.0’ is particularly suggestive: could you elaborate? Are you referring to the way in which the techniques and methods of the natural sciences are now being used to explore those phenomena which we take to constitute the core of our humanity, i.e. mind and consciousness? The emergence of a ‘science of cognition’ would seem to represent science’s attempt to understand itself by investigating the structure of the mind engaged in investigating reality. Is this the ultimate twist in the trajectory of Enlightenment: the twist whereby the explanatory strategies hitherto used to disenchant ‘outer nature’ are now being deployed to disenchant our own ‘inner nature’? TM: Well, first, anti-naturalism could be true. We have to take one philosopher and one argument at a time. I certainly do understand your point – academic disciplines are self-modeling entities too, and they tend to preserve their boundaries. There will always be individuals who 23. New York and London: Routledge, 1995. 210 Metzinger – Enlightenment 2.0 profit from isolationism, from keeping the boundaries hard and the methods ‘pure’. There is this conservative social movement within analytical philosophy, yes. But let us not commit a psychologistic fallacy or indulge in paranoia – I do not know any anti-naturalists who intend to make the world safe for the irrationalism and fundamentalism. The question is simply who is right here. Enlightenment 2.0 will tell us more about what the conditions of possibility for knowledge are. We will use our intelligence not only in order to understand its own evolution, but also to begin to optimize it. We will use scientific rationality to gradually understand the finegrained, microfunctional properties realized by our brains that enabled scientific rationality in the first place. If one takes a closer look, it is unclear what ‘Enlightenment 1.0’ actually was, but at least there were certain general attitudes increasingly shared and cultivated in Western society – a critical questioning of traditional institutions, customs, morals, and so on. I think that in the decades to come we may be able to elevate this old philosophical project to a new level of precision and generate a sustainable growth of knowledge by connecting it to our new partner disciplines like cognitive neuroscience. In particular, we could free the Enlightenment project of the accusatory, moralizing tone it has tended to exhibit in the past – we may obtain a better understanding of the ways in which in-group behaviour, tribalism, traditionalism, unargued conservativism, higherorder forms of self-deception and cultures of denial actually evolved, what their neural underpinnings are, why they proved to be adaptive in the world of our ancestors and how they now spill over into the process of doing science. 211 COLLAPSE V We may also come to understand how this effect could be minimized. In the course of this process we will also see that actual scientific practice is not at all something carried out by philosophical saints pursuing the ideal of self-knowledge, but is often driven by jealousy, ill-will, ostentatious behavior and brute career interest. This could indirectly bring about a great improvement in the process itself; it could serve to foster intellectual honesty. But Enlightenment 2.0 may also have a sobering quality about it, extending Max Weber’s ‘disenchantment of the world’ to the conscious self. We will have to pay an emotional price for it – and what I am interested in is what this may do to the forces of social cohesion, to what Habermas would have called soziale Bindekräfte. C: In Being No One, you describe the object which I consciously experience as holding in my hand as ‘a dynamic, low-dimensional shadow of the actual physical object in your hand, a dancing shadow in your nervous system’.24 Does this not also imply that the hands which are holding the book, and indeed, the entire manifest image of the world, turns out to be ‘a dancing shadow in my nervous system’? There are several other places in your book where you say that the content of our phenomenal image of the world is ‘solely’ and ‘exclusively’ due to our neurophysiology. For example, you write: ‘Our conscious model of reality is subterranean in that it is determined exclusively by the internal properties of our nervous system [...] phenomenal experience as such unfolds in an internal space, in a space quite distinct from the world 24. Being No One, op. cit., 549 212 Metzinger – Enlightenment 2.0 described by ordinary physics [...] [and] evolves within an individual model of reality, in an individual organism’s brain, and its experiential properties are determined exclusively by properties within this brain’.25 Similarly, the cognitive scientist Donald Hoffman has argued that our perceptions of the world no more resemble the world itself than a computer icon resembles the inside of a computer.26 Would it be fair to say that, for you, representation is more akin to simulation, understood as the process whereby the mind-brain generates its own virtual phenomenal reality, than to replication, understood as the process whereby a mirror reflects a pre-existing reality? But then what is the ontological status of the reality that is being perceived or simulated? Moreover, how do we get to know about the non-phenomenal objects to which the phenomenal ‘shadow-objects’ of our experience supposedly correspond in some way, and what is the nature of this correspondence? Does your conception of representation allow for the possibility that our perceptions might resemble aspects of the unperceived world in some sense, at least at a structural level? Or does the world generated by representational processes, the world of commonsense phenomenal experience, screen us from the world in itself? We might be tempted to insist that the only real objects are those quantified over by fundamental physics, but the problem with this is that fundamental physics nowadays doesn’t quantify over any objects (a point made by James 25. Ibid., 547-8; emphasis added. 26. Donald Hoffman, Visual Intelligence: How We Create What We See (New York: W. W. Norton and Co., 1998); cf. ‘Conscious Realism and the Mind-Body Problem’ in Mind & Matter, 6:1, 2008: 87-121. 213 COLLAPSE V Ladyman in another interview for this volume),27 and certainly doesn’t provide us with anything like the ‘actual physical objects’ experienced in phenomenal consciousness. How, then, is science, which would seem to be rooted in the domain of phenomenal experience at some level, able to access reality? Can science dispense with appeals to perceptual consciousness altogether? Where do you stand on these basic epistemological issues? TM: Well, we should never conflate intentional and phenomenal content. Conscious experience per se is how the world appears to you – that is all. Epistemology only comes in with intentionality, with semantic properties of representations, such as reference or truth. Unfortunately I don’t know Ladyman’s work yet, but it seems obvious to me that he has a very good point – we have to assume those entities as existing that our best current theories with the highest predictive success etc. need to postulate, and classical objects will not be among them. I like to look at our phenomenal model of reality – including objects, properties, relations, naïve physics and all – as a multimodal interface to navigate our behavioural space, a tool enabling the organism to successfully close certain sensorimotor causal loops on a very coarse-grained level, shielding it from the underlying causal complexity. Consciousness is autoregulation in a system that has become much too complex to understand itself. The phenomenal model of reality is a naturally evolved virtual organ, and segmented scenes, sets 27. See ‘Who’s Afraid of Scientism? Interview with James Ladyman’ (this volume), and James Ladyman and Don Ross with David Spurrett and John Collier, Every Thing Must Go: Metaphysics Naturalized (Oxford: Oxford University Press, 2007). 214 Metzinger – Enlightenment 2.0 of properties bound into robust objects, and so on, were functionally adequate fictions – just as selves were. But the majority of these fictions must have picked out at least some of the relevant causal regularities governing ‘middlesized’ physical phenomena in our space of interaction, or else they would not have helped us to copy genes more efficiently. Other virtual models may have been adaptive precisely because they created functionally adequate forms of self-deception. But given the evolutionary history of nervous systems as a background assumption, it is hard to see how the majority of relational/structural assumptions about our ecological niche could have been outright false – our ancestors would not have survived. You may remember that in Being No One I explicitly and deliberately decided not to develop a theory of mental representation, and to ignore epistemological issues in favor of fine-grained neurophenomenology. But I did – very vaguely and perhaps inconsistently – endorse connectionism, embodied dynamicist cognitive science and JohnsonLaird’s theory of mental models as working background assumptions. For a mental model, the key phrase you are looking for could be ‘partial relational homomorphy’ – perhaps it is like a child’s colouring-book: an extremely selective set of outlines is given, but all the colours are self-made. The extremely small and selective set of spatiotemporal relations and causal regularities faithfully carried over into our perceptual model of reality might then be something that is strongly relative to our bodies and the actions beings like us can perform in the world – a world without objects or selves, but one that can be known through the right combination of science and philosophy. 215