Eva Dale 0:00 From the heart of the Ohio State University on the Oval, this is Voices of Excellence from the College of Arts and Sciences, with your host, David Staley. Voices focuses on the innovative work being done by faculty and staff in the College of Arts and Sciences at The Ohio State University. From departments as wide ranging as art, astronomy, chemistry and biochemistry, physics, emergent materials, mathematics and languages, among many others, the college always has something great happening. Join us to find out what's new now. David Staley 0:32 My guest today in the ASC Tech Studios is Vladimir Sloutsky, Professor of Psychology at The Ohio State University College of the Arts and Sciences. His research focuses on conceptual development and on interrelationships between cognition and language. Welcome to Voices, Dr. Sloutsky. Vladimir Sloutsky 0:49 Oh, thank you very much. I'm very pleased to be here. David Staley 0:52 You have recently published an article with your postdoc, Layla Unger, titled "Ready to learn: Incidental exposure fosters category learning", and I want to dive more into this. First of all, tell us your findings in this research. Vladimir Sloutsky 1:07 Well, before delving into the findings, your listeners might want to know what the categories are. David Staley 1:13 Please. Vladimir Sloutsky 1:14 And this is really a fascinating domain, because typically, we don't deal with categories; we deal with individuals. You have a friend, you take a bottle to drink, you drive your car, and yet these things come in more general classes, like there is a class of bottles and there is a class of cars, and even though you love your pet dog, there is a class of dogs. In philosophy, there have been lengthy discussions as to whether categories exist in reality or whether they exist only in language. So, we will set that aside, and we'll just say that they clearly exist in human cognition. And so, one of the questions is, how do people learn these categories, and this is one of the foresights of my research. And so, what we did with Layla in this article, we demonstrated that people can learn some of the categories - and I will talk about which ones in a moment - but people learn some categories without any teaching, without any feedback, just by observing things. And so, incidentally, it happens that these are types of categories that we encounter in real life, like dogs or pots and pans or toys, so things that have much in common among their category members and not the things that we learn in the classroom, like prime numbers or vowels or other kinds of things that are defined by a specific rule. And what we found in this research is that you expose people to these things, and the next question is, how do you know that they learned something? If you start asking them questions whether or not they learned, they're smart people, they will go back in their experience, they realize that this exposure was about something that you were trying to teach them. So, what we did instead is we presented them these category members under a different task, a completely different task that had nothing to do with the learning task, and then we explicitly taught them these categories and compared their learning to a control group where they had exposure to the same features, like if you think about bugs, it will be the same shapes and the same kinds of antenna and so forth, but that didn't have category structure. And what we found was that people who were pre-exposed to these categories, without their knowledge, learned the categories substantially faster than people who were equally familiar with the same stimuli, but those when the stimuli didn't have the structure. So, what it tells us is that when you have categories that are highly structured, you really don't need any teaching signal, you don't need feedback. You can learn them without your awareness that you know them. And what we believe is that this is an evolutionary old system that is present in most vertebrates. That's how pigeons learn to recognize hawks or to recognize appealing food or to recognize people, for that matter. So, we know that pigeons learn these categories, we know that rats learn these categories, we know that primates learn these categories. They live in the wild; there are no classrooms in the wild, so that's the mechanism, how they do that. David Staley 5:25 And you've already sort of moved in this direction - are these categories innate, like Chomsky's ideas, these categories are innate? Vladimir Sloutsky 5:32 No, far from it. If they were innate, you wouldn't need this mechanism, and also, if they were innate, it would be extremely detrimental, because imagine that a pigeon is born in an environment where hawks look differently from the hawks living in the environment where the pigeons parents were born. Well, this pigeon is dead on arrival, and yet pigeons can adapt and learn new dangers. So, the cost of innateness is that it is very difficult for individuals to adapt, so you need generations to adapt. Learning on the fly, in contrast to innateness, is very convenient, because it gives you flexibility to adapt to new things David Staley 6:24 When you say "expose subjects to these different categories or concepts", what does that mean? What was the sort of experimental context? Vladimir Sloutsky 6:33 Well, the experimental context was - and I will give you an example of one experiments, there are five different experiments there - so for example, they would be presented with, imagine that these were artificial stimuli, but suppose, for the sake of the argument, that these are different birds. So they will be presented with a bird, and there will be a sound, and their task will be to report when there are two identical sounds in a row. So, as far as they're concerned, we are testing their memory for sounds. So, birds are simple attention grabber for them, so they don't know that they even need to pay attention to these birds, their task is to pay attention to the sounds. So, all learning happens completely in the background. They are not trying to guess what we want them to do with these items, they just pay attention to the sounds. David Staley 7:35 One of the terms in this article that caught my attention is the idea of latent learning. Tell us what this concept means. Vladimir Sloutsky 7:43 The concept of latent learning is that if you ask people explicitly, they are not aware that they learned something. If you test them immediately after that learning, there is no evidence of learning because we gave them fairly brief exposure; and yet, when you teach them these things explicitly, they learn it substantially faster than people in the control group. David Staley 8:15 Should I take from this research the idea that maybe I need to rethink how I teach my students, how I teach my adult students? Vladimir Sloutsky 8:22 Well, if you're asking for my advice, don't do that. David Staley 8:26 Oh, okay. Vladimir Sloutsky 8:26 Yes, don't do that, because there is a fundamental difference between how we learn things in the wild and how we learn things in the classroom, and there is a place for learning in the wild and there is place for learning in the classroom. In the classroom, we learn, let's call them scientific or scientific-like concepts, concepts that are defined by a single rule. Sometimes the rule could be slightly fuzzy, but there is still a rule. Like, if you teach history, you want people to understand that revolution is not just some unpleasant event, it is a sudden change in political power, and they could be peaceful, they could be bloody, it doesn't really matter. What matters is whether the change in power is gradual, which is not revolution, or sudden, which makes it a revolution. So, real life concepts are very different from that. There are multiple features that make things similar. If you think about cats, there is a similarity in shape and similarity in texture and similarity in size and similarity in diet and similarity in sounds that they produce. So, there is much more commonality among these items, such that you really don't need a defining rule to learn these categories. So, I think that what we do in the classroom is really good for the classroom. We don't want to borrow the classroom ideas to teach about cats or about dangerous things, but equally, we don't want to take real life learning which is slow and robust and move it to the classroom, which is more like discovery learning, you discover a really interesting commonality among things, like you do in physics or in history. David Staley 10:35 You began this conversation by saying that philosophers, for instance, ask the question, do categories exist in reality, and you said, well, let's put this off to the side for a second. To what degree is your work either influenced by or have you influenced philosophical discussions? Vladimir Sloutsky 10:51 Well, I think that I don't want to be naive philosopher, so I will just say that the debate that lasted for hundreds of years was a debate between the so-called realists, who believe that categories exist in the world, and nominalists, who believe that categories are imposed by language. If we think about it, it seems that nominalists propose something really interesting and appealing, but it cannot possibly be true if we agree that non-verbal animals, like pigeons, learn categories. So, it seems that there are some regularities in the world that these organisms pick up. If nominalists were right, and categorical structure is imposed strictly by language, then we wouldn't be talking about category learning in pigeons. David Staley 11:56 So, this research has you working with adults, and I wanted to talk about the work in your lab, the Cognitive Development Lab. And my understanding is that you tend to work with children, small children, in the Cognitive Development Lab. So, I wonder - first of all, tell us about some of the research, some of the findings that you've engaged in this lab. Vladimir Sloutsky 12:16 Well, I would say that the most interesting findings in my mind are interesting difference in cognition at different points of development. So, to put it succinctly, I would say that children, in general, people early in development, not only children, but infants as well, tend to explore their world, tend to sample things broadly, tend to distribute their attention, and as a result, they tend to learn more slowly, but they're well equipped to adapt to sudden changes in the environment - I will explain it in a second. In contrast, adults, they can do everything, but because they assume that all learning in the lab and around them is pretty much like learning in the classroom, they tend to focus their attention, they try to find the defining feature of the categories, and as a result, sometimes they pay the price. If things change rapidly in an unannounced manner, it is somewhat difficult for them to adapt. Imagine that a task - and we have a couple of papers on that - a task is to learn a category, and a category, it is like, let's say squirrels and chipmunks, okay? And so, you can think about a defining feature that distinguishes the two, something like a tail, and the tail would be a defining feature, but there is also an irrelevant feature, something like texture, because texture is very similar in chipmunks and squirrels. And so, you give this task to children and adults. Both learn categories. If you track their eye movements, adults would quickly focus on the tail, whereas children look everywhere. And then you introduce an unannounced shift such that the tail doesn't distinguish so they need to learn new creatures, they just don't know it yet. The tail is not informative any longer, but texture that was uninformative becomes informative. And so, what happens is that children who distribute their attention broadly, so keep sort of plugging along and they learn new categorie and they discover the new feature that reliably distinguish between the two. And for adults, it takes some time, they may struggle. So, in other words, the kinds of classroom learning that we have are very convenient, very good for quick learning of scientific categories. The problem with that is that they require a very stable environment, so scientific concepts do not change, and if they change, you are told that they change, so you have time to rethink, so whereas real life learning is really great for volatile environments, the environments where rules can change. So, this learning is slower, but it also allows you to adapt to these changes. So that's the interesting difference between children's learning, who are slower but more flexible than adult learning, who are fast but less flexible. And I don't want to communicate that adults are somehow dumb. If you tell them that they need to adapt, they can do anything. Human adults are the smartest organisms in existence. It is only if they don't know, and in real life, nobody announces that the environment will change. Eva Dale 16:33 Did you know that 23 programs in the Ohio State University, College of Arts and Sciences are nationally ranked as top 25 programs, with more than ten of them in the top ten? That's why we say the College of Arts and Sciences is the intellectual and academic core of the Ohio State University. Learn more about the college at artsandsciences.osu.edu. David Staley 16:59 As you've been talking, you talk about children, for instance, having this sort of distributed attention versus more focused attention - is that a result of, sort of, formal schooling or formal learning, which strikes me as being about focused attention? Vladimir Sloutsky 17:13 It is interesting that you are mentioning this. To answer your question, let me suggest a hypothetical experiment. Suppose that you have a five month old baby and you put them in the classroom: will they be able to do that? And the answer is clearly no. Now suppose that you have an illiterate adult who never attended school. Will they be able to do that? This is, sort of, more controversial, but research with illiterate people that was conducted in the 20s and 30s would suggest that probably not. So, I think that this is a confluence of brain maturation that allows people, gives them the ability to focus and filter out irrelevant information in the environment where they can practice this ability, which is formal schooling. David Staley 18:19 I saw a brief video from your lab about how you work with children, quite small children, and I'm interested to hear you sort of describe, what's it like? How does a researcher work with, you know, six month old, eight month old, you know, toddlers? Vladimir Sloutsky 18:35 Well, I think that everybody should know that, A., it is hard. It is very hard. And B., it is very rewarding. So, what is hard about it? So, let's start with six to eight month old. So, what happens with six to eight month old? When they get bored, and they get bored quite quickly, they stop cooperating, they start crying. So, the entire period of cooperation could be between ten and fifteen minutes, if you're lucky. So, in general, it is less than that, and you need to pack a rigorous procedure in a very short period of time. So, that is hard. Toddlers are probably the hardest simply because they will not cooperate to begin with. So, it's very hard to do because they are fully mobile, so it's difficult to capture their attention by a small toy, by shaking the toy in front of them. So, they have a tendency to explore their environment, so it's very difficult to capture their attention. So, things become easier as they move towards four years of age, but still, if children don't want to do something, they will tell you that they don't want to do something and you have to stop. But I would say that four and older are easier. Toddlers are the most difficult ones, and infants, six to eight month old infants, are somewhere in between. David Staley 20:21 I'm curious to know how you ended up in psychology, as opposed to my discipline, history, or physics or music, something like that. Why psychology? Vladimir Sloutsky 20:31 Well, there is an intellectual reason and a structural reason. An intellectual reason is that I have been always interested in people. The structural reason is that I was growing up in the Soviet Union, and certain disciplines were not open to people of Jewish origin, so I never wanted to do physics, but I wanted to do mathematics, and mathematics was not open. History, not that I was interested in history, but if I were, it was not open either. So, University of Moscow, which is one of the few places where you can study to become a researcher, had disciplinary departments, and only few departments were open. And so, the ones that I could go and of course, there was no guarantee, because it was extremely selective, was psychology and the other was biology. So, that was my real choice, not among all other disciplines. David Staley 21:42 Give the listeners a sense, why would Jews have been denied entry, say, into the mathematics faculty, or something like that? Vladimir Sloutsky 21:50 Well, the argument was, of course, there was never a written policy, we just knew. Everybody knew. If winners of international Olympiads cannot pass the exam, there is something wrong. But, I think that in the system, the argument was that Jews were prone to immigrating to Israel, so their loyalty is questionable, and so why give somebody an education in the most advanced mathematical field when they would leave. So, of course, it resulted in this self-fulfilling prophecy, but that's a conversation for a different time. David Staley 22:32 And psychology wasn't viewed in the same way as mathematics? Vladimir Sloutsky 22:36 I think that a lot depended on individuals and people who chaired the Department of Mathematics took the task of discrimination much more seriously than people in psychology, so people in psychology probably didn't take their instructions as literally as people in mathematics. So, I don't think that there were different instructions, I think that implementation by individuals was different. David Staley 23:09 I'm curious to know, across your career, have you ever had an "aha" moment? Vladimir Sloutsky 23:13 Have I ever had an "aha" moment? I would say that I have had a number of "aha" moments. David Staley 23:22 You're very fortunate. Vladimir Sloutsky 23:23 But one of them was - and this goes to the beginning of our conversation, about Chomsky - it was a realization that having innate knowledge is not just blessing, as some people would imagine, that makes life easier, it also imposes a tremendous cost, simply because it makes you inflexible. It doesn't give you an ability to change when the environment requires these changes. So, it is probably great for organisms like spiders, who have very short lifespan and who are born with all the skills that they need. But, if you think about humans, or primates in general, the greatest advantage of humans as living organisms is the incredible flexibility with which they can interact with the world. And so, learning gives you this flexibility, and innate things give you sort of rigid external structure from which there is no escape, so harder to adapt. So, realization of that and understanding how smart humans are in that respect, was one of these "aha" moments. David Staley 24:54 In the last, I don't know, say, 40 years or so, what's been the biggest change in your field, big sort of paradigmatic change? Vladimir Sloutsky 25:03 Well, I think that there are probably two sort of converging influences: one came from mathematics and statistics, so it is formal modeling of cognitive processes, and I will explain what it has done, and then another coming from neuroscience, and that's bringing the brain into our thinking about cognition. So, formalization really allows us to test ideas and hypothesis by, A., achieving greater precision in describing the concepts that we would like to study, and B., simulating conditions that we cannot attain in the real world. So, if you have a formal system that can learn an artificial learning system, you can lesion it. You can deprive it of certain capacities and examine how learning will proceed under these conditions. You can change the way it intakes information, so all of that allows you to test multiple hypotheses that would be very difficult to test with real participants. And neuroscience constrains the way we understand cognition. We know that the brain is a real organ, and the more we know about the brain, the cleaner and more precise our thinking about cognition is. For example, we have, right now, we achieved some understanding of brain maturation. It's far from perfect, but we know the key periods in brain maturation, so that gives us very important additional power in our thinking about developing cognition. David Staley 27:11 What's going on in your lab right now, or what's next for your research? Vladimir Sloutsky 27:15 Well, ideally, what we would like to do is to link these two developments together and to understand, to leverage ideas that can come from mathematical and statistical modeling and the ideas that come from our understanding of the brain, to understand a why children tend to explore widely and sample their environment, how this affects their attention precisely, and how their attention supports their learning in the wild and in the classroom. David Staley 28:02 Vladimir Sloutsky, thank you. Vladimir Sloutsky 28:04 Thank you very much. Eva Dale 28:06 Voices from the Arts and Sciences is produced and recorded at The Ohio State University, College of Arts and Sciences Technology Services Studio. Sound engineering by Paul Kotheimer. Produced by Doug Dangler. I'm Ava Dale. Transcribed by https://otter.ai