https://youtubetranscript.com/?v=tEuGpc16fB8
One of the courses I teach in the cognitive science program is Cognitive 250, which is the introduction to cognitive science. And I’m going to try and compress the first sort of six weeks of lecture into ten minutes. So we’ll see how well that goes. So what is cognitive science? And very often you get sort of this thing from the catalog. Well, it’s the interdisciplinary study of the nature of cognition. But why? Why interdisciplinary? Well, that has to do with a little bit of the history. So we’ve been trying to understand the mind and cognition for a long time. And philosophy did quite a bit of work at it. But we realized that in addition to that conceptual work, we had to do some experimental work where we were studying more carefully behavior. And so psychology also starts to study the nature of the mind. But we start to realize that we can’t study that just on its own. We need to understand the brain level. How is the brain causing and contributing to behavior? And that’s where, of course, neuroscience comes in. But it also matters how brains link up to each other. And so we need linguistics and we need anthropology to also tell us how minds and brains work. So what we need is these various disciplines to work together in order to get a complete picture. Because what matters is no one of these levels is sufficient for understanding the mind. Therefore, no one of these disciplines. Because all of these levels are very causally interdependent with each other. They affect each other. The brain level affects behavior, behavior affects the brain. The behavioral level of one brain is affected by the behavior of other brains. So that anthropological, cultural level or that linguistic communicative level matters. And so what we need is we need a discipline that isn’t doing each one of these levels, but the discipline about how they’re causally interdependent and therefore theoretically relevant to each other. That’s what cognitive science is. And cognitive science is therefore pursuing a project of synoptic integration. It is trying to create a conceptual vocabulary, a theoretical grammar for how these various disciplines can talk to each other so as to afford this very valuable synoptic integration. So that means that cognitive science is this really creative project. We’re trying to come up with these theoretical constructs that will serve this important purpose. So, what do you do when you construct a theoretical construct? Well, what you do is you’re trying to make the most plausible construct you can. See, what do I mean by plausibility? I want a theoretical construct that holds good promise that it’s going to predict where I’m going to find good empirical evidence. I want a construct that’s going to help guide and constrain future empirical research. So I need plausible constructs. Well, how do we come up with plausibility in general? Well, I’m going to indicate to you a model of plausibility that I’m working on with Leo Ferrara. And this plausibility model looks something like this. There’s a couple things we’re doing to try and increase the plausibility of our constructs. The first thing we’re doing is we’re trying to get convergence. We’re trying to get how the various disciplines, various theoretical lines of argument, various lines of research all converge on a shared conclusion. Why do we do that? Why does that matter? Because when you get convergence, you increase trustworthiness. Trustworthiness is the conclusion that your construct has not been produced by bias. Look, I can look at the book, but in order to make sure that the book is really there, that I can trust my perception, I also reach out and touch this book on neurocomputing. Because if both my sight and my touch converge on this, it’s less likely that I’m just hallucinating the book, that it’s just being produced by bias. It’s much more likely that there’s something really there. Now, is it perfect? No. But it increases the probability that the conclusion here that is resulting is not due from any single line of bias, but from something really there. But I want something more than convergence. What I want is I want to help afford insight. I want to indicate new ways of thinking, help people in the various disciplines to solve problems, to open up new lines of research. I need to promise insight. One of the things my constructs have to do is they not only have to have convergence for trustworthiness, my constructs have to be elegant in the sense that I can use the same construct in many different situations. This is a highly powerful or insightful construct because I can take it here and frame and solve problems I couldn’t solve before. Here, here, here. If the various disciplines can share this construct, this is one powerful way of affording synoptic integration. I want a balance between these. The more I’m trying to transform the field with my insight, the more I need to guarantee or at least assure myself that my construct has high trustworthiness. Because if I do this without ruling bias, I get weirdness, I get conspiracy theories, I get fantasy, I get craziness. But if I have a lot of convergence, a lot of trustworthiness without a lot of insight, then I get triviality. I get what is obvious, what everybody already knows. And that’s also not scientifically useful. But if I can get something that’s highly convergent and highly elegant in this manner, then has a high degree of plausibility, then I have something powerful, potentially profoundly powerful. So, cognitive science is about creating these constructs.