The History and Future of Consciousness Studies | Anil Seth

Anil, to get a sense of where we are today in consciousness studies, it's good to know the historical development, certainly in recent times, about how we got to where we are. So from your perspective, what's a good time to start? I mean, people have always been talking about the mind-body problem in some sense or another, but in terms of the modern consciousness studies community, what do you look towards as the initial period of time and what have been some of the major developments until now? I do think it's important to reach back a little bit because it really was Rene Descartes hundreds of years ago that gave us the problem of consciousness in the form that people have been wrestling with it ever since, which is this dualism, this dualistic perspective between conscious experiences and the brain, the body, the material world. And then there were lots of sort of ups and downs over the hundreds of years. And a lot of people talk about consciousness research being born or reborn in the 1990s. Now, it was there at the very beginning of psychology and neuroscience. People like William James talked about it a lot towards the end of the 19th century. But because of perceived problems with how you can study consciousness, the reliability of introspection, then consciousness research kind of went to the fringes and then completely beyond out into the wilderness for many decades. Yeah, I did my doctorate in neuroscience and neurophysiology in 1964 and 1968. And my motivation was to think about how can I understand consciousness? That was really my motivation. And I don't remember once in four years using the term in terms of working with my professors or anything else. That's interesting. I mean, even when I was, I did my undergraduate degree 1991 to 1994. And that was in psychology, ended up in psychology in Cambridge. And I was taught by some of the leading behaviorist psychologists who were very against consciousness being within the remit of psychology, even in the early 1990s. From the perspective as a student, it was still not really a legitimate topic of study. But by then, unbeknownst to me at the time, it was becoming rehabilitated in other places. And this rehabilitation I think happens for a couple of reasons. On the one hand, there were some serious philosophers who started talking about consciousness, not only from a philosophical perspective, but also bringing in some experiments, evidence from psychology and neuroscience. So people like Daniel Dennett, who we sadly lost earlier this year. That changed the game. Then you had work in science that was also beginning to say sensible things about consciousness. Bernard Baars, a psychologist, formulated the first version of what became, or still is, the global workspace theory in the late 1980s. And then in the early 1990s, we had the beginnings of the Neural Correlates of Consciousness research program, inaugurated, or at least officially inaugurated, by Francis Crick and Christoph Koch. The idea there was, okay, look, there are these intractable philosophical problems about how consciousness relates to the physical world, if indeed it does. But we know as a matter of empirical fact that the brain is intimately related to consciousness. So let's just focus on these empirical correlations, the Neural Correlates of Consciousness, and let's look at something that is relatively well understood from other branches of neuroscience, the visual system. So let's focus on visual awareness and look for the Neural Correlates of Visual Perception. That was also game-changing because at the same time, brain imaging methods are becoming more widely available. So suddenly you can do experiments. You can have people in brain scanners where their subjective experience might change. In binocular rivalry, for instance, the stimulus is the same, but conscious experience flips from one thing to the other. So you can begin to ask questions like, well, what happens in the brain when perception changes but the stimulus doesn't? I think all of these things contributed to this first wave of excitement in the 1990s. You had the formation of new journals, the Journal of Consciousness Studies, Consciousness and Cognition, a series of conferences that are still going on now. So from the 1990s on, that's really the path that we've been following. What's been interesting to see is the progressive integration of philosophy with neuroscience and psychology. I think when I started, again back in the early 1990s, still being interested in it, even though my behaviorist supervisors were not so much, there were still largely distinct communities. That's been changing over the last two, three decades for sure. I think that's been an essential part of it. There's also been this growth in theoretical work and the availability of new methods. The power of computers to sustain large, detailed models of neural systems has also helped. So now I think at least in some areas, there's a very productive interaction between theory and experiment. Some theories have really become among the more prominent. There's probably a group of four, global workspace theory, integrated information theories, higher order theories, and broadly predictive processing theories, of which my ideas are part of that family. Those are not necessarily totally mutually exclusive. You could have some nestings from one to the other, I would assume. Yeah, you can. There's been work trying to figure out where are the overlaps. That can be a useful thing to do because you can leverage different ways of thinking. If you can make connections between theories, you're better placed to understand where they differ and where they say the same thing. They're also not mutually exclusive in a more problematic way, which is that they are often theories of different things. They all purport to be theories of consciousness, but then when you drill down into what are they actually theories of, if they're theories of different things, then of course they're not mutually exclusive. I'm known to take the opposite view. That says that there's a danger in narrowing down too quickly because we have uncertainty to narrow down to 4 or 22 or whatever. That gives the impression that that's the only game in town. Some of that is because those communities have similar kinds of, shall we say, fundamental assumptions, physicalism or whatever it is. There's really a very broad range of theories or ways of thinking about consciousness. I'd be concerned that we don't squeeze those off too soon in order to really understand what consciousness is. Yeah, I think I broadly agree with that. I think all the current theories that we have are wrong, but hopefully they're not entirely wrong, or hopefully at least some of them are not entirely wrong. I think the key is the difference between necessity and sufficiency, that some of the theories that are known in physicalism may be necessary but may not be sufficient. Right. This brings us to what would be the criterion for a successful theory. We want it on the one hand to provide explanatory insight, predictive power, so we can tell what is conscious, what isn't, what kind of consciousness might be happening. We also want it to provide both necessary and sufficient conditions. I don't think many theories are like that. There is one integrated information theory that does propose sufficient conditions, but many of the other theories that we have, like global workspace theory, higher order theory, to me they seem more descriptive of how consciousness unfolds in human beings and perhaps other animals. They're not laying down, okay, if you have these conditions, then you have a conscious system. That's where we need to go. There's also, I think, this distinction between scientific theories and philosophical theories. This speaks to assumptions that may be common to all. Most of the prominent theories are in some way or another physicalist, materialist theories. Many are computational theories too. I think, for me, that's a more problematic assumption. Why should that be the case? Of course, we have other philosophical theories too, besides materialism, panpsychism, monism, idealism. These may turn out to be ultimately more useful frameworks, but they are different. Philosophical frameworks, in this sense, as far as I can see, behave according to different rules and should be evaluated by different criteria. You don't test physicalism. You ask whether it's useful in the game of explaining the phenomenon in question. You can't refute it in an experiment.