February 27, 2014 | Issue#2:mind
Interview conducted by Signe Cane in Riga, Latvia, during the 9th International Symposium of Logic, Communication and Cognition, hosted by the University of Latvia.
Andy Clark is a Chair in Logic and Metaphysics at the University of Edinburgh. He was Director of the Philosophy/Neuroscience/Psychology Program at Washington University in St. Louis. Andy’s interests span not just philosophy of mind, but also artificial intelligence, including robotics, artificial life and embodied cognition.
David Chalmers is Distinguished Professor of Philosophy and director of the Centre for Consciousness at Australian National University. He is also Professor of Philosophy and co-director of the Center for Mind, Brain, and Consciousness at New York University. David is best known for his work on consciousness, but is also interested in metaphysics and epistemology, and the foundations of cognitive science.
While you were colleagues at Washington university in the mid-nineties, you came together to publish “The Extended Mind” (1988), in which you proposed to look at the mind in a different way — as “not just in the head”. What exactly does that mean?
DC The idea of the extended mind was to look for a closer integration between the mind and the environment, getting it right into the loop, so to speak.
AC I think the real attraction of the extended mind story is that the activity upon which mindfulness depends is much more spread out than we thought. Maybe our ongoing use of things like [smartphones] and other sorts of external structures is really part of creating a web of activity, where mind is what happens when that web happens. The paper itself is about the possible role of external structures in establishing that you can hold what philosophers call a ‘dispositional belief’. It’s the kind of stuff that you don’t have to be rehearsing in your mental life. For example, my belief that it’s May doesn’t require me to be forever rehearsing that it’s May — it’s a dispositional belief of mine.
DC We actually had two central examples. In the paper, there is the case of playing Tetris and rotating things in the environment and making the analogy between that and mental rotation. The idea is that it doesn’t really matter either way whether it’s physical rotation or mental rotation. If it’s a way of figuring out whether the Tetris tile fits into the rack, it doesn’t matter whether it’s outside or inside the brain. Now that, I guess, is actually a current activity rather than dispositional state.
AC That’s right, yes.
DC And then we had the case of belief. Inga believes that the museum is on 53rd St and therefore walks there. Otto, on the other hand, has Alzheimer’s and keeps a notebook where he writes everything down. He’s previously written down that the museum is on 53rd St, looks it up, and walks there. We argued that the case of Otto is completely analogous to the case of Inga. Ordinarily we’d say that Inga believes the museum is on 53rd St because she had the disposition to do this. Otto is exactly the same, only the memory, rather than being stored in the biology of his brain, is in his notebook. The notebook is serving as part of Otto’s memory constituting his cognitive states. As Andy said, that’s a case of a dispositional belief — once he reads it and it comes into his consciousness, then it’s also affecting his brain. The special point here is, even before that happens, it’s part of the broad infrastructure of his mind.
How did you come to collaborate on the idea?
DC This has been a perpetual theme of Andy’s for years. It was more of a side interest for me. Andy came to me with an article called something like “The Mind and World Breaching the Plastic Frontier, First Draft”.
AC Yes, I had a little first draft. I think it was called “Mind and World: The Plastic Frontier”. I think part of the genealogy is like this — when the paper only had the Tetris case, you might very well think, “we’ll be excited by that.” But philosophers may not be that excited, because they could re-describe that case as just input and output to the real system upon which “mental stuff” depends. And that’s all going on in the head. The idea that Dave had with Otto and Inga case was to create something philosophers couldn’t ignore. Where something that they already care about — these states of dispositional believing — now seemed to depend on a more widespread structure. Of course, it wasn’t even called extended mind at that stage.
DC I was a big fan of the book by Richard Dawkins called The Extended Phenotype, and thought, ah, we’re kind of trying to do the same thing here for the mind, so let’s call it “The Extended Mind”! Someone might have said about the Tetris case that it is just an extended process, or extended cognition in some sense — the processing goes on outside, the leap from state to state is outside, but all the real mental stuff is inside. We thought, well, let’s try and do this for a paradigmatic mental state, the state of belief.
Would you say the basic concept of the extended mind has changed since?
AC I think in the extended mind debate since then there have been lots of questions raised about to what extent it depends on taking the stuff that goes on internally as already somehow privileged.
DC We don’t want to fall into the trap of brain chauvinism, where the brain defines the properties of cognition and the stuff out there has to be just like it in order to be mental. Maybe this could be broadly analogous to it in the way that a Martian with a very different cognitive system could still be analogous enough to a brain to count as having cognition.
AC It still leaves the question, could it be totally different from the thing that goes on in the head, but just because of the way that it gets together, it’s analogous? I actually think this question doesn’t have an answer. I think that the concept of the mind we started with doesn’t deliver an answer to whether it’s a case of extended mind or a case of something else that’s interesting and important.
DC I think dialectically and rhetorically it makes sense to start with the easiest. For the purposes of the paper we tried to take a case that was as closely analogous as possible, this case of Otto with the notebook and Inga. We said that we don’t propose to offer necessary and sufficient conditions for being a case of extended cognition, just pointed out — here are some relevant factors. Ever since people have been saying “but what’s the mark of the cognitive, what are the necessary and sufficient conditions,” and I think that in general that’s pretty hopeless.
AC We did get lots of that. I don’t think the internalist needs a “mark of the cognitive” to recognise some states as being cognitive, so why should the extended mind theory?
DC My colleague Ned Block liked to say: “When you wrote the article, the thesis was false, but since then it’s become true!” I think what he probably means is the advent of things like the [smartphone, search engines]… people using this kind of stuff all the time. Although I’m not sure whether he says that because now there are cases of doing extended cognition, or whether he thinks that the concept of the mind has actually changed in our technological world.
When it comes to the biggest objections that have come up over the past 15 years or so, how do you think the thesis has fared?
DC When we published the article, I saw the attractions of the thesis but had some reservations. We had this footnote saying “Authors are listed in order of the degree of their belief in the central thesis”. Andy was very confident, I was on the fence. It’s probably still true that I go back and forth on it a bit more than Andy. But one thing that has made me much more confident over the years has been looking at all the objections that people have come up with. I have to say, they’re really not that strong. You go through the best of them and the most well-known ones, and they all strike me as quite easy to requite. If that’s the best they can come up with to reject the thesis, it’s starting to look quite good to me.
AC I agree. I haven’t actually seen anything that strikes me as a convincing objection yet. In fact, I think, most of the things that are put forward as objections, we more or less anticipated in the paper. There are tricky cases, you know. Some of the things that might give you pause, is for example, the role of perception in the cases of the extended mind. It always seems like in all of these cases there’s something perceptual going on.
DC To say it’s got to be bound by the skin or the skull, it seems unprincipled to me. If you want to look for something principled, argue that perception and action are the boundaries of the mind. When Otto reads the notebook he’s doing it via perception, and the mind is what happens on the other side. And sometimes I can come to feel moved by that, but then why can’t we redescribe what’s going on in the Otto case? Viewed through one lens, it’s perception, viewed through another, it’s memory retrieval.
AC Here’s an example that I quite like for this one. It’s the ‘memory glasses’ experiment at MIT. These glasses were developed to help people with Alzheimer’s or visual form agnosia (where people can’t recognise objects in front of them). The glasses have a little camera that’s taking input from the surroundings. In the case of someone with mild Alzheimer’s, the system that they’re wearing “knows stuff,” if you like it. As it recognises a face, it will give you a very quick flash of information — that’s your wife, uncle, dog. Interestingly, that information can be flashed so quickly that you don’t experience it, but nonetheless it helps these patients to do better.
DC Charles Stross’s novel Accelerando has one illustration of the extended mind — glasses that provide you with all kinds of information. People become totally reliant on them, so when this one guy’s glasses are stolen, he becomes a “gibbering wreck” in the same way one would if a part of their frontal cortex was removed. Now, [a new product like this] has just been released, so maybe we’re moving into the era where this thesis becomes truer for all of us in everyday life.
How has your work diverged since you were colleagues?
AC Quite a bit, I think. It had diverged already, and it just so happened that we came together on this one issue.
DC It was around the same time when I was doing work on consciousness. That’s continued to be my central theme among many other issues, I guess. For Andy, maybe extended cognition’s been his central thesis?
AC Yeah, it’s likely.
DC Every now and then I try to get Andy to write something on consciousness. In general, I fall closer to the Cartesian side of things than Andy does. A lot of my intuitions and approaches in philosophy are fairly internalist. Still, I think there’s room for certain kinds of externalism. The nice thing about this project was to try and accommodate some externalist insights about the role of the environment.
What about extended consciousness, then?
AC I have been quite sceptical about it, but I’m slightly less sceptical today. There’s only one argument I’ve seen that’s pushed me a little bit to wonder about that. Suppose a conscious experience depended on moving through a set of neural states in a certain way, and it’s only because of a certain influence of the external world that you can move through those states in that way. That’s where one might say that the inner stuff isn’t sufficient. That might actually be true, but I don’t think it gives you anything much like the extended mind, because it’s simply what it would take to perceive the external world at all.
DC I think both of us would certainly want to allow that, in principle, there can be extended consciousness that arises through interactions with the environment. I’m still dubious that connections to the environment like a notebook or calculator will extend consciousness in that way. The environmental stuff certainly affects one’s consciousness, though. It does so by affecting one’s brain.
AC I suppose, if there were group conscious states, that might be the case of something like extended consciousness, but you don’t see the evidence for that. If a group of people could somehow have, depending upon their interaction at a certain time time, some sort of communal state that is only available because they’re there in that group…
DC That’s not, obviously, quite like the extended mind. I think that’s more like a “collective mind.” In extended mind, roughly the same person is somehow extended, whereas in collective mind, who knows — is it a whole new person?
AC Yeah, it is different.
People have taken to the extended mind idea and have tried to apply it to different things. What has been the most surprising development to you?
AC I was surprised and interested in some developments in cognitive archaeology. Some archaeologists were arguing that the extended mind provided a framework in which to reconceptualise archaeology as involved in dealing with things that have been parts of people’s minds, almost like a fossil trace.
DC In traditional philosophy it’s beginning to have some impact. Andy’s colleagues at Edinburgh in the epistemology department proposed the extended knowledge project, where you start thinking of knowledge as this extended process that involves interaction with the environment. I think that reconfigures some issues on “how we know” about the world. A few years ago I went to a talk by Joel Anderson on the extended will. It was about the role of things like environmental cues in increasing your willpower. When you get up from your desk, there’s a sign saying “go back to work, you idiot”. That is a mechanism of the extended will — it constitutes your willpower. Here the extended mind actually feeds into self-help, and has some practical use.
AC A real-life application! I think that’s a good case.
Do you think you might collaborate on something again?
AC I wouldn’t rule it out.
DC To be fair, we have pretty different interests. I’ve been going off the last few years doing stuff on metaphysics, epistemology, philosophy of language, which Andy doesn’t have a strong interest in. Although he is a philosopher, so he’s interested in everything, and so am I. There are a few issues where we meet in the middle in philosophy of mind and have some overlaps. Who knows.
David, you published a book last year, titled Constructing the World. Can you tell us a little about it?
DC The basic idea is building blocks. If you want to build a model of the world, what are the basic building blocks going to be? Rudolf Carnap wrote a book back in 1928 called Logical construction of the world where he basically said that from a few simple ‘primitives’ you can define every expression in a language, like English, and build up every concept that we need to describe the world. He did it with a very limited vocabulary, and tried to build everything up from there. And people said that it’s a wonderful project, but it doesn’t work. It’s seen as a “noble failure”. What I’ve tried to do is see if you can take a version of Carnap’s project, tweak it a little bit and see if something like it can be made to work. Ultimately we get a kind of blueprint for the world in terms of five or ten basic concepts — such that all truths about the world can be derived from there. I’ve tried to argue that has consequences for thinking about a lot of issues. Not just philosophy of mind, but philosophy of language, thinking about meaning, epistemology, knowledge, metaphysics, about the character of reality.
AC I haven’t gotten to grips with it yet…
DC It’s a long book!
AC The request that I would have is that I would want to know what these ‘conceptual primitives’ have to do with processes that actually do construct mental spaces in humans and other animals, because it doesn’t seem likely to me that the primitives that they start from will look like the primitives that you’re talking about. Instead, I think the primitives you’re talking about are actually quite advanced accomplishments.
DC It’s not intended as a contribution to cognitive psychology or to the empirical theory of the mind in the first instance. There are a whole bunch of interesting questions on how exactly one connects this to the question of whether there are primitives of the mind and primitive concepts. There is a lot of debate about that in psychology and elsewhere.
What are you working on right now, Andy?
AC I’m working on something that is quite internalist in one way and definitely hugely empirical and computational. It’s an interesting question where these things might meet. I’m working on predictive processing as I’ll call it, which is just a particular model of how you build up models to get to grips with the world as a living creature. The basic idea is that we try to predict our own sensory flow using a multi-layered structure. What I think is interesting about it is that internal processes deliver perception and action as being computationally built in just the same way. They sort of squeeze cognition out because you don’t need it any more, it’s all done by the stuff that is doing perception and action. And that stuff, precisely because it’s doing perception and action, rolls the world in at any point in which that’s going to be a useful strategy.
DC It’s a wonderful framework, ridiculously ambitious. And like all of these ridiculously ambitious ideas, my suspicion is that at the end of the day it’s only going to be a part of the story. Usually the advocates of these ideas want it to be the whole story or at least a huge part of it. I’m suspicious that the process of predicting the next sensory signal will explain everything about the mind or everything about cognition. But I do think it’s a useful and interesting mechanism, and it’s a bold hypothesis that by building world models that are good at predicting the next sensory signal, it will also be good at doing all these other things that we need the mind to do. My suspicion is that at the end of the day that’s a very strong constraint. If evolution is resourceful, it’s going to help itself to all kinds of other mechanisms that you don’t get out of predictive coding which will go a little bit further.
You both have opinions on each other’s philosophical views, perhaps there is even disagreement?
AC I suppose, one of the views that Dave has pursued is the fairly hard line with respect to conscious experience. He says that even when we finish telling the full story functionally, using contemporary or near future science, explaining all the responses of the organism, we wouldn’t actually have done the work to understand how conscious experience is possible. I’ve always thought that actually we would.
DC We do disagree about the case of consciousness. I’m inclined to think that the modern computational paradigm, although it’s explained many wonderful things about the mind, has significant limitations when it comes to explaining conscious experience, how things feel from the inside of having a mind and being a subject. I’ve argued that the standard methods of neuroscience and computational science at least have to be augmented to bring consciousness into the picture. I think Andy is much more deflationary about consciousness than I am. I’ve always been encouraging him to write more about it. I think he’s got some interesting views on the matter. I think, one of these days — Andy writes a book every few years — one of these days it’s got to be a book on consciousness. Maybe connecting it to the extended mind?
AC More likely I’d end up connecting it to predictive processing.