Tadeg Quillien

Tadeg Quillien is a cognitive scientist at the University of Edinburgh. He studies causal cognition and theory of mind, using tools and ideas from computational cognitive science, philosophy, linguistics, and evolutionary biology.

Interview

‍

OASIS: Thanks for sitting down with us. What's your favorite ontological entity and why?

TQ: I really like possible worlds, because I work with them, and also from a psychological perspective, they seem crazy. If a linguist explains to a psychologist how necessity and possibility are operationalized, the psychologist will say you are crazy because this idea that we are quantifying over all possible worlds, from a psychological perspective, this can't be the way the brain works, you'd have to run an infinite computation to do that. Yet arguably this is the most elegant theory of modality that we have. But sometimes the best theory is a crazy theory. 

OASIS: So what do you do when a theory is crazy in one domain and elegant in another? 

TQ: In psychology what we like to do is to take the crazy theory, and come up with ways the brain could approximate that computation. It's hard to do, but sometimes you can come up with experimental predictions to test those ideas about how the brain approximates it. 

OASIS: Tell us how your career trajectory turned out in an unexpected way.

TQ: When I started my PhD, I was very interested in moral psychology. So I would read all these papers about the trolley problem, for example. When I started to dive into the topic, I began to get interested in why intentional actions are somehow morally worse. This idea that if someone hurts you accidentally, you might be willing to forgive them. But if someone hurts you intentionally, you'll find that that's worse, maybe it's harder to forgive them.

And so this led me to ask, what do we actually mean by doing something intentionally? I came up with this theory, maybe it has to do with the way we represent the causal relationships between what's in our mind and what's in the world. But then, once I was there, I had to study how the mind represents causation. So I started working on that. But then there are all these theories about how causal reasoning relies on counterfactual thinking. So I also ended up working on counterfactual reasoning.

In sum, the history of a lot of my research is diving deeper and deeper into a rabbit hole, where you get into more and more abstract concepts.

When I started, I thought I would do very traditional moral psychology research. But I got a little bit frustrated by the state of the field, because there was not as much theory as I wanted. This seems to be a general concern in a lot of psychological research. 

OASIS: Hmm. Why is this?

TQ: Somewhat paradoxically, psychologists are not well-trained to think about things in terms of computations. It's very easy to rely on folk psychology and intuitions about it and work like that. 

OASIS: Do you work step by step or follow hunches (or neither, or both)? Explain?

TQ: I very much tend to follow hunches. There is this idea in the research world that you need a research program and you need to work on it consistently. No matter how hard I try to work on one thing, I always seem to have a side project, something that pops up, and it's very hard not to follow it. Very often my best ideas happen when I read something that is completely outside of my current research interests, I get an insight that I feel might be novel or interesting.

OASIS: Tell us about a mentor you'd particularly like to pay tribute to.

TQ: I did my PhD with Leda Cosmides and the late John Tooby. In their lab, people use an evolutionary approach to psychology… 

OASIS: …What does that mean? 

TQ: What this means is that you're basically doing evolutionary biology blended with cognitive psychology. And so in practice, this means thinking a lot about the evolutionary function of cognitive mechanisms. 

In a lot of research labs, the PI works on Topic X and everyone in the lab is more or less expected to work on Topic X. Their lab was not like that at all. There was this expectation that if you follow an evolutionary approach, you can take an evolutionary perspective on any topic in psychology, and you can make progress on that. And so I had previous labmates who worked on things as diverse as logical reasoning, spatial navigation, visual attention, moral psychology, and so on. 

So when I ended up working on causal cognition, which is something no one else in the lab had done before, this was totally ok. 

OASIS: Was it lonely? Did you feel sufficiently supported in that topic? 

TQ: I had to do a lot of reading on my own. But there was a lot of support, both from my mentors, and from other people in the lab, discussing the ideas. So overall I think it's a really great place, a really great atmosphere to be doing research.

OASIS: Tell us about a student you've learned something from.

TQ: I'm working with Can Konuk who is a student at ENS in Paris. We've been working on causal cognition together with Salvador Mascarenhas. Many people in the field of causal cognition are working with symbolic Bayesian models. And we tend to be sceptical of neural networks and connectionist approaches. But Can at the moment is working onwith a new theory that does use neural networks. And I think he makes a good case that they can be useful and they can hold interesting insights, for those of us working on causal cognition. 

OASIS: Thanks for your thoughts! See you in Edinburgh. 

TQ: Thanks!

‍