‘It is about time to get to a more unified and coherent narrative in the cognitive sciences’

An interview with Jelle Bruineberg

15 July 2016

Jelle Bruineberg is a Ph.D. candidate at the Institute for Language, Logic and Computation (ILLC) and the department of Philosophy of the Faculty of Humanities, at the University of Amsterdam. His supervisors are Martin Stokhof and Erik Rietveld; he is part of Erik Rietveld’s VIDI-project ‘Landscape of Affordances: Situating the Embodied Mind’. His current research interests lie on the intersection between theoretical neuroscience, embodied cognitive science and phenomenology. He is interested in how brain, body and world contribute jointly to everyday skillful behavior.

 

“I started studying Physics and Astronomy at the UvA in 2007. When I was half-way, I took some extra courses, and after I finished I could do a one-year extra bachelor in Philosophy. Then I did the master ‘Brain & Cognitive Sciences’, and the one-year philosophy master at the UvA. After that I started my PhD with Erik Rietveld. I met him at the end of my Philosophy bachelor; I was about to start my thesis in logic, but I switched to doing Embodied Cognitive Science and wrote my bachelor thesis with him. He introduced me to many of the texts and the concepts in Embodied Cognition, and some of the people. So when he had a PhD position available, it was a nice match I think.”

 

“The title of my project is ‘Naturalizing skilled intentionality’. The idea is that skilled intentionality, or skilled action, consists of pretty much unreflective ways of going on with the world. I always give the example of someone biking through Amsterdam: what are the aspects of the environment, the practices and the dynamics of the environment that are important there. That is, how can someone gain his or her balance in the interactions with the environment. My question is how you can naturalize that: how can you build a cognitive (neuro-)science, or any kind of science actually, based on these concepts. What I mean by naturalizing is finding some kind of isomorphism between descriptions at a phenomenological level and descriptions at a systems level, perhaps a dynamical systems level, or a neuroscience level. My work is mainly focused on work in theoretical neuroscience, such as predictive coding and more broadly the Free Energy principle, and how that could fit in with this account of skilled action. Related to this is the question of how brain dynamics stand in a relationship with the environment. Even more broadly my project engages with debates about representation, computation and content; that is more the philosophy of mind side. Next to that I also do more applied or practical things like modeling, where I am interested in how we can model certain behavior using particular kinds of paradigms from the cognitive science. The project is pretty interdisciplinary. To model neurodynamics we use ideas from theoretical physics, from computational neuroscience, and from phenomenology and ecological psychology. That is part of the challenge, or the weakness of the project, that it spans the whole range of the sciences. But that is what makes it really exiting. Understanding all aspects of skilled intentionality interests me so much because it combines neurodynamics with dynamics of the environment, with the social practices that are needed to enable a particular kind of behavior, with the skills needed to do that, with the intentionality. This intentionality is not goal-directedness in the sense that you have a goal in mind, but more being directed towards some kind of grip. In that sense the phenomenon of skilled action combined so many aspects of the whole width of sciences that I think that understanding those is an important key in having a more unified understanding of these sciences and how they relate to each other.”

 

“Part of my job is also to teach. I have been teaching in theoretical philosophy the last few years: a first year course covering metaphysics, epistemology, philosophy of language, and phenomenology. That is a lot of fun. It is also pretty challenging, in the sense that you teach lots of things that you are not completely familiar with yourself. Because it is so broad, no one really has the background to be at home in all those fields. For me it might have been sometimes even more difficult, because I do not have a traditional philosophy background, because I always took the shortcut in philosophy: I did a shortened bachelor’s and the one-year master’s. So in a sense, I did not read much of the classics. Then teaching on Kant for three sessions is quite challenging. But then again, after these three times you have really learned a lot about Kant. I think it gave me a bit more of a broad background, which is good as well for the rest of my work.”

 

“The idea for the modeling project is that I would do a ‘proof of principle’ kind of modeling, using some relatively easy studies. Of course, the example of biking through Amsterdam is very under-constrained: it is a crazily complex problem. An easier case to model is sports. Sports are typically cases where behavior is very complex, but it is at least bound by particular kinds of rules, such that possibilities are not infinite. My current idea is to look at boxer-boxer interactions, and how two boxers together get to some kind of grip on the situation. We model the idea of skilled intentionality using Merleau-Ponty’s concept of ‘tendency towards an optimal grip’. That is the example of two people in an elevator standing too close together, taking a step backwards to reduce some kind of self-perceived tension. One thing we came upon in our reading of the ecological psychology literature and the sport science literature is something we conceptualize as a ‘hypergrip on the field of affordances’. The idea is that experts boxers compared to novices do not punch the thing that is most readily available, but move to a situation where a multiplicity of action possibilities is available. This has been shown already for ice climbers: novices only look one step ahead, whereas experts have a grip on the whole field of affordances. Experts can thus prospect: they do not always take the next best thing, but they can make a more difficult move that in the end allows them to get to the top more easily. Thus, you do not tend toward the closest minimization of tension, but a more prospective anticipation of your situation. The idea is that in these boxer-boxer interactions, is that you would find similar kind of things; the attempt is to see if you can model that.”

 

“I use mainly concepts from phenomenology, ecological psychology and the Free Energy principle. In phenomenologists’ work, mostly in the work of Hans Jonas, and also in the ecological psychology literature, there is a strong ‘continuity between life and mind’. This is the idea that the structures of life are continuous with the structures of the mind: what it takes for an organism to self-organize, or to self-maintain or self-produce itself, is related to the structures of intentional directiveness, or the structures of consciousness. That is a pretty bold claim, but you find it both in ecological psychology, in phenomenology, and in the work of the Free Energy principle. The Free Energy principle tries to give a unifying account of the workings of the brain and how the brain is situated in the animal and in the brain-animal in the environment. The basic idea behind the Free Energy principle is that in order to flourish, organisms need to minimize the prediction error with respect their interactions with the environment. It is a theory with a crazy amount of formalisms, but it is a bit harder to actually compute with it. Therefore, most computations are done with a derived theory of the Free Energy principle: Predictive Coding, which is now very hip in neuroscience. There the idea is that there is anticipation of the interaction with the environment, which you can model by something like prediction error minimization. The Free Energy Principle is more difficult to compute with because it wants to answer the why-question of everything. In the case of Predictive Coding you can just build a model for instance for a random-dot-motion task, and then it does what it is supposed to do, and that is really cool. But the Free Energy principle states that you can see perception-action, development, evolution, etcetera all as different ways of minimizing free energy, where some kind of species is just kind of settling on the transient towards its most ecologically optimal state. I think this ambition of giving a more unifying account of cognition within biology is really important. There is so much pet theories lying around; I think it is about time to get to a more unified and coherent narrative. And I think that potentially the Free Energy principle could provide this.”

 

“In a sense my project is about front-loading specific kinds of questions for cognitive science and cognitive neuroscience. For example the question of relevance: why is one aspect of the environment more relevant than another aspect? There is something about that aspect of the environment in a particular kind of context, given a particular kind of agent. And in so many cognitive science experiments and modeling studies, the goals and the tasks of an agent are already predefined, and the experiment is only about the performance within this particular kind of setting. I think it is really important to make the switch to more open-ended kinds of tasks, where the task is actually to find out what the aim of the task is, while interacting with the environment, that is way more natural. There is so little known about what determines relevance, and I think that will be a key question to cognitive neuroscience, hopefully within the next few decades.”

 

“I hope my work is pretty accessible for cognitive scientists; I try to write it with a specialized but non-philosophical audience in mind. But I think so far, I have pretty much failed in that. There are a few people who read my work well, but I think there are also lots of people who read it for whom it is too esoteric. It is still pretty technical, and there is lots of vocabulary from lots of different fields. Making that cross-connection is really challenging. I hope to make my articles more readable and slightly less technical in the future, because it is of course also a sign of really understanding your own work very well if you are able to explain it to a lay audience. The first paper I wrote in my PhD has actually been cited quite a bit, by people doing sport science, by ecological psychologists, by some computational neuroscientists, and by philosophers. Cognitive neuroscience seems to be missing a bit from this, but at least it seems to be read by a wide audience, and that I find really nice. Hopefully these cognitive neuroscientists will follow as well.”

 

“When I was very young I wanted to be a fireman, until I was like 7 or 8 or so. And then, strangely enough, at some point I told my grandma that I was going to be a professor, also when I was still very young. Which was a bit weird, because it is not in my family or so, my parents did not go to university or anything. I do not think I am so serious anymore about being a professor, but being an academic has been something that I have been wanting to so for a really long time. I think it was just a matter of finding the right topic to work on. I had a lot of interests, and I did not really know what was the right one. This seems to be a theme where I can really combine lots of my interests, so that is nice. What I also like a lot about this job is that you get to travel so much. Traveling is just really nice; everywhere you come you meet very interesting people. And of course in the end I find it also important that what you do has at least a little bit of impact. Going somewhere and someone having read your paper, that is actually pretty rewarding. Surely there is an intrinsic content-wise motivation, but those kind of moments show you that you are not just doing it for yourself, but there are at least people that get some kind of inspiration or ideas from what you do. And that is really worthy as well.”

 

“I am starting to think about what I would like to do after my PhD as well now; the first year is just working and slowly getting your way into it. I think I want to continue working on these kind of ideas; they are very rich and I think there is a lot more to get out of it. Perhaps I would also like to work a bit more on the empirical side, the modeling side. Otherwise it tends to get very abstract. That is nice for a philosopher, but sometimes you want a bit more concrete work to be done with it.”

 

“When I started my master’s in philosophy and cognitive science, it was the first year of Julian Kiverstein’s appointment, and I have been extremely lucky to have both him and Erik Rietveld close to me during my master’s. I think that allowed me to have this rich array of ideas, and to do this interdisciplinary work that I do now. It is really a shame that this interdisciplinary environment that was there at the time is not available to current students. In that sense I am concerned with myself if I would have studied at the UvA five years later, whether I would have been able to do the same kind of things that I am doing now. I think currently, the answer is no, and that is really a pity. As I see it, there is continuity between philosophy and cognitive science, and the way in which the university is structured does not facilitate this continuity: some people are at the faculty of science, some are at the faculty of humanities; I think this is very detrimental. SMART Cognitive Science is very important to bring these researchers closer together, but I think it is also important to achieve this at the level of education; to have joined courses, study tracks and programs.”

by Gisela Govaart, July 2016

Published by  SMART Cognitive Science