Greg Murphy studies the psychology of concepts. How do we use language to understand things, and how do we sort the world into categories? In our conversation, we consider what makes a category, why we love them, and where they steer us wrong.
Dr. Murphy released a book on this topic a few years ago: Categories We Live By
How We Classify Everyone and Everything
Transcript
Please note that the transcript is for the interview portion of the episode only, and it was automatically generated with AI assistance. It has not been checked for accuracy.
Andy Luttrell: When I was a freshman in college, I stumbled onto what turned out to be a very handy conversation starter. I would ask people whether cereal, like breakfast cereal in a bowl with milk, whether that was a type of soup or not. And everyone’s first reaction is to just be confused, like what? Obviously cereal is not a soup. But then they would start to reckon with the question. They’d say, well, soup has like meat and stuff in it. Cereal doesn’t have meat. Ah, but there are plenty of meatless soups. I would say, well, sure, they’d say, but the milk in cereal has no place in a soup. Ah, but have you heard of cream soups, my friend? They would say, but a soup is hot, cereal is cold. And I would say, oh, but there are plenty of cold soups. Gazpacho was always the wild card that I was ready to play, and I always did, because everybody wanted to call soups hot. And on and on and on. Try as you might, you’d either have to come to admit to yourself in a moment of desperation and nihilism, that cereal was in fact a type of soup, or acknowledge to yourself that you didn’t know what the heck soup actually is.
And that really is the insight. What is soup? What is any category of things? We walk around pretending we have a clear set of rules for sorting the world into groups of things, but we don’t actually have a real rulebook, no specific logic behind the categories. It’s very walks like a duck, talks like a duck world out there. You don’t have a clear set of parameters for calling something soup. You’ve just been introduced to all sorts of things in your life that different people have referred to as soup. And then your confused little brain has to build some fuzzy idea of what they all have in common, hoping and praying that some self important college freshman never found you and asked you the annoying question, is is cereal soup?
Little did I know at the time that these sorts of questions were about to have a mini renaissance of their own. The most famous version is the ever frustrating dilemma of whether a hot dog is a sandwich. Same gotcha tone as the cereal conundrum. A hot dog, you may be forced to conclude, is a bit of meat presented between bits of bread, and therefore it is certainly by definition a sandwich. But it’s the same experience. You take for granted what a sandwich is, even though nobody ever sat you down and explained to you the rules behind what is and isn’t a sandwich. You just know. At least you think you do. Like if I asked you directly do you know what a sandwich is? You’d tell me, rightly, to get lost. Of course you know what a sandwich is. But then lob hot dogs into the machine and suddenly you have to reckon with the fact that you’ve never really known what a sandwich is. Nothing is a sandwich. Everything is a sandwich. Oh, God. What are we doing here?
Okay, but what does any of this have to do with opinions? Okay, well, opinions often live in the world of categories. When we think about what we like and what we don’t like, a lot of the time, those things exist at the level of categories. I would say I like sandwiches, but that assumes I even know what that label means. So understanding how people form and use categories is actually essential to understanding how they go about grouping up things in the world to form a sense of what they would approach and what they would avoid. Tell me you’re serving soup for dinner and I will accept the party invitation because I love soup. But if I show up and it’s bowls of Cheerios everywhere, I mean, I’ll stay and I’ll eat the Cheerios, but just know that’s not what I expected, and you’ll just look at me with a smug look and ask me the question I’ve been running from for the last 18 years. What is cereal? Not soup?
You are listening to Opinion Science, the show about our opinions, where they come from, and how we talk about them. I’m Andy Luttrell, and this week I’m excited to share a conversation I got to have with Dr. Greg Murphy. He’s an emeritus professor at New York University, and he’s spent his career looking into the cognitive psychology of concepts, like how do we know about what stuff is? This includes our use of language to refer to things, and, you guessed it, how people categorize the world. A few years ago, he released this great book called Categories. We live by how we classify everyone and everything. I’ve been interested in the idea of categorical thinking for a long time. Check out my SciCom summer episode with Lulu Miller for a little more on that. But I saw this book and was immediately excited to read it, and it’s great. He shares a lot of the basic psychology of categorical thinking and also gets into what that all means for everyday decisions, like legal categories, racial categories, clinical diagnoses, natural species, regulatory categories, all sorts of stuff.
Anyhow, I loved the book and it stuck with me. So I reached out to Greg after a while to see if he’d be up for talking with me for the podcast, and I’m glad he was willing. It was really a great conversation and I think it’ll get you to really question the categories you use to think about things, where they come from and what good they’re doing you or not. So let’s get into my chat with Dr. Greg Murphy.
Andy Luttrell: I have been super interested in the sort of cognitive psychology of categories for a while as a non expert. It’s just like a curiosity of being a person in the world and thinking about. In my world of attitudes, we talk about the attitude object, right? The topic about which you have a positive or negative association. And we always wave our magic wand when we talk about that attitude object and just say, just trust us, it’s a thing. But there are so many ways you could think about the level at which you would attach a sense of positivity or negativity to some concept in the world. And then a lot of the applications to things like stereotypes and prejudice really lean on this notion of social categories. And there again they’ve given a little more attention to the psychology of the category itself. But even then it gets pretty fuzzy pretty quickly, right? It’s very clear that like that’s not the thing they care about, it’s the next part. But let’s get down to like what the thing is.
Greg Murphy: The whole, the whole nature of prejudice is that you find out some group that somebody belongs to a category and then you use that to judge them and interpret their behavior and perhaps greatly, you know, biasing your own interpretations of what they’re doing and of who they are. But it only works if you have a category. If you only, you know, realize that you’re. And of course this is not just prejudice, positive negative prejudices, but also positive prejudices. So you know, I find out that you’re a Republican or a Democrat and that I already hate you without knowing anything about you. I. Because I have feelings about those categories, right, of people. And if I admire you because you’re a doctor the first time and trust you the first time I meet you, that’s because I have some positive beliefs and maybe emotions about that category. So yeah, prejudice certainly is a case where the categories of gone wrong. You need a category to explain it, but it hasn’t been to our benefit arguably.
Andy Luttrell: And the quirky thing, I guess there are two that’ll be themes that we’ll talk about is one, people are multiply categorizable and sort of flexible in the moment, right. I can categorize you any which way, any of those categories. I have different kinds of associations and so our relationship really hinges on this decision I make as to what you are. And then also that, like, we are imposing these neat categorical structures onto the messiness of nature, and we’re inventing so much of the nomenclature and the psychology behind what counts as this category. And so, yeah, prejudice just, it’s a. It’s a nice kind of intersection of sort of basic cognitive psychology of category and then the complicated social world that we find ourselves in, right?
Greg Murphy: And people have very often have an inflated idea of the reality of a category, of and of the similarity of the objects or people inside the category. So, you know, again, you may have some feeling about, say, Republicans, right, and have some ideas, positive or negative, and kind of think, you know, Republicans are. They’re all alike. They’re all, you know, and then start spouting things. But the governor of Vermont, Governor Scott, is a Republican, the only Republican elected to statewide office in Vermont. And he’s apparently the most popular governor in the United States. So in a state that is overwhelmingly Democrat, except for the progressives and the socialists, the Democratic socialists, we’ve elected a Republican governor. And I used to describe him to people, this doesn’t work so well anymore. I used to say, well, he’s not a bad guy. He’s just slightly to the left of Andrew Cuomo. Now he’s a Republican. And that just shows he would not fly as a Republican candidate in Ohio. And even within the Republicans in Ohio, you know, you’ve got your. J.D. vance’s, but you’ve also got, you know, other people who are like traditional Midwest Republicans. When I lived in Illinois, I was exposed to this idea. People who are kind of fiscally conservative and kind of ahead towards social things, like, well, why don’t, why isn’t everybody just minding their own business, you know, and they. And that’s very different from what we see now from someone like J.D. vance or whatever. So we have a tendency to think that for people, certainly the category is a little more all encompassing than they really are. And we have that same tendency to think about categories of objects like chairs or events like parties or things like that, that, you know, I hate parties. I want to go to parties. You have to talk. You’ll meet strangers. What do you say to the strangers? What’s going. You know, it sounds horrible, but then, you know, occasionally you have to go to a party. It’s like, well, it wasn’t that bad. But you go back to thinking about parties are terrible, you know, a little bit, a little bit later. So that the kind of simplification of categories, such as thinking that they’re more similar than they really are, is very common, I think. And it’s not just true with person categories, it’s true with all of ours.
Andy Luttrell: To cut to the fundamentals, I have been wondering sort of how much is everything categories in terms of knowledge, like things we know, our concept of the world, how much is it almost fully represented by categories? And it’s, it’s really rare that we’re thinking of like a very specific exemplar and the knowledge we have about it. But most of our understanding of the world is at the level of categories.
Greg Murphy: You know, certainly like the school taught, knowledge that we have is virtually all about categories and you know, knowledge, what we would call knowledge is often knowledge about categories. So the fact that say dogs are a mammal and that mammals have a four chambered heart and things like that, so it’s at a level of generalization applying to, you know, millions of creatures in the world that have existed and currently exist and will exist. So yeah, we’re thinking about a whole category in this case, But we also have a lot of knowledge about individuals. If you have a dog, you know a lot about your dog. And you know, that knowledge can then come and trump whatever it is that you believe about dogs in general. So, you know, you might think that dogs will chase cars, but you know, your dog never chases cars. So you don’t worry about your dog chasing the car, but for some other dog that you happen to be dog sitting or walking or whatever, you know, you’re worried about the cars. And it’s hard to say in real life how much we rely on those two sources. You know, right now I’m meeting you for the first time. I’m thinking of you as a social psychologist and an academic, which I have a lot of knowledge of, you know, for my own academic career. But as I get to know you more and more and more, you know, I might end up thinking like, well, you know, he doesn’t really like to argue, it’s very strange, you know, and other things that then preempt that categorical knowledge and then finally there’s some idea that things that are really similar, new things that you meet in a category. So if you happen to be really similar to some social psychologist, I knew I might draw on that specific example. So my knowledge of this other person, I would say, well, you know, Andy’s probably interested in this too, or he doesn’t want to be bothered with these things or whatever, because that’s what my Friend is like, so the categorical knowledge is really important for our general knowledge of the world. And it’s also really important when we’re encountering new things, because when you encounter new things, you don’t have any specific knowledge about it. You don’t know anything about that individual. And this is just, you know, I like to use the example of cooking. You go shop and you’re going to make a recipe. You know, they go tell you tomatoes, they don’t tell you which tomato to look for, and you don’t know anything. As you go to the supermarket, you know there’s a sign saying tomatoes, but you don’t know anything about those objects and you’ve never seen them before. And you’ve got to assume that they’ve got the same properties of tomatoes that you’ve encountered before and that are going to be good in this recipe and you don’t have to go and investigate them. In particular, you’re not running chemical tests on the tomatoes you buy from the supermarket because you think you know about tomatoes in general. Now, it turns out, like I this last summer, I encountered black tomatoes. That was a little bit of a shock.
Andy Luttrell: We grew purple tomatoes in our garden this year that were purple all the way through to the middle.
Greg Murphy: Wow. There you go. So, yeah, so the day before, I would have said, tomatoes are red, you know, and in fact, if you ask me now, I would still say tomatoes are red because I only saw that one black tomato plant, you know, and so it wouldn’t take too long for me to revert to my stereotype as it is about tomatoes, which is, you know, 99% of the time, perfectly useful. So it’s a simplification, but it’s a simplification that works for us in everyday life in many cases. And it’s really, I think that we started to question how well it works in understanding people, that the simplifications we apply can be very damaging and wrong and unhelpful to us, as well. As we don’t know something about a person, maybe we should know because we’ve been blinded by what category they’re in and what we think about that category.
Andy Luttrell: We can run with tomato, too, as a recurring example. It’ll come up. You have a great way to get at people who think very differently, categorically about tomatoes than maybe they ought to. So I’m going to put a pin on it for that for a moment. But I think this is a good opportunity to be clear about what a category means, and at the risk of asking you to define category in the same way that it’s risky to ask people to define categories themselves that exist. How do you generally think about what it means for there to be a category that exists in. In people’s conception of the world?
Greg Murphy: So a category is a group of objects or entities, we should say, really, because it includes events and all kinds of other things as well. It’s a group of entities that have some significant properties in common and that you can treat different members in that category the same way to some degree. So tomatoes are generally red. They’re generally round. They’ve got a thin skin, they have a certain flavor. They’ve got seeds inside all these things we know. And their uses are similar, their tastes are similar. So tomato is a good category. Most of the categories we talk about are kind of like that. They’re relatively useful categories that when, you know, if I. If I tell you, like, I’ve got a tomato in the car, you know kind of a lot about what it is that I have in the car. Or if I say I have a dog or anything else, you have some idea of what to expect when we get there. There’s other categories that are much more minimal, that are kind of technically categories, but that are not that interesting. And I don’t, you know, it’s the kind of thing that philosophers will try to engage you in, but I don’t really like to talk about them. So, like, categories of brown things. So sing. That’s kind of based on one property. So, yeah, all brown things are in the category of brown things. But the problem is, if I tell you I’ve got a brown thing in the car, you don’t know anything more about it except that it’s brown. It’s not really very useful. Right. On the other hand, if I told you, oh, I’ve got a red brown thing that I bought at the supermarket, you know, and you decided that’s probably a tomato, then you would know just by making that leap tomato. Therefore, he’s going to make a salad out of it, or he’s going to throw it at somebody’s window or whatever it is that you can use tomatoes for. So you can have these kind of minimal categories of brown things or even silly categories like things I touched yesterday, but they don’t really do you any good. So this is not really the kinds of categories the way we think about categories. They’re just, again, kind of philosopher’s delight to talk about those things empirically.
Andy Luttrell: So I’m curious about this. So it seems like those are Technically categories, but just not ones that people engage with very much empirically. Do we know much about the categories that people spontaneously turn to? How could we document that in ordinary life? I’m not actually engaging my brown things category scheme as often as I might be engaging other category schemes.
Greg Murphy: Well, I mean, we rely a lot on language to tell us that. Again, you come, let’s say, I’ll give you a ride home. You get to the car and you say, oh, there’s a dog in the backseat, or, what are you doing with all those tomatoes? You don’t say, there’s a brown thing in the backseat, or what are you doing with all those round things in the backseat? Right. You overtly classify them into one of these richer categories, typically. And the richer categories are the ones that help us reason about the world. So again, if you just classify something as being a brown thing, it doesn’t lead to any more information. So if you want to draw conclusions about some new object that you’re seeing, it only really does you good to put it into one of these richer categories. And so if I see something and say, okay, it’s a dog, say, running around your yard, I might decide, I don’t want to go into the yard because, you know, dogs sometimes bite strangers, and I don’t care to go in there. On the other hand, if I look and I see it’s a brown thing and don’t go any farther, like, well, there’s a lot of brown things. It doesn’t really tell me what to do or what not to do. So I think language is the key to the, to our use of the rich categories. When you are going to use these other categories is when they have some specific purpose. So you’re doing set design, and you need a lot of brown things. You start looking around. If you’ve ever been in that situation where you’re looking for something that that is, meets, like, one specific criterion, you know, you’ll really start to annoy your friends as you’re walking down the, you know, walking down the street going like, hey, look at that. It’s really round. What are you saying that, like, no, I could really use some round things, you know, and maybe you could really use some round things, and that’s when you start, you know, thinking of things that way. But again, it doesn’t. It doesn’t help you reason about the world any any further.
Andy Luttrell: Generally, I think we’ve. We’ve generally hit on this. But, like, what the value of these categories are, like, it seems like, um, and maybe this is also the time to talk about, like, how real they are. And to the extent that are they. Is the world truly presented to us categorically versus we are completely inventing lines to draw around stuff and obviously whatever, somewhere in between. And then if, if we’re doing that, like, why are we doing that? How is it serving us to engage with the world this way?
Greg Murphy: Yeah, the realness of categories is very hard to give a definite answer to because we can’t get outside our own heads and see the world in some other objective way. We’re already seeing it with the categories. So it’s a little bit like saying, can I. When I see something that’s green, is it really green? Is it the same green that you see? Like, well, probably. But there’s no way to kind of get outside and see like, what the perception of green really is. Maybe that’s not a great example because you can in fact measure, say, the wavelength of light and measure whether it’s green or not. My feeling is that the human categories that we use are some mixture of things that are real, things that are not so real, that are like human. Human inventions and conventions. So. And the way to tell about their being real is simply that if the categories are not real, if they’re just completely arbitrary things, like things that I happened to touch yesterday, they don’t do you any good. They won’t stand up to repeated use. They won’t. So, you know, I touched my car handle and I touched the door at the supermarket and I touched. I touched my daughter, say. Right. But knowing that these things are somehow together in some sense doesn’t really tell me more about them or allow me to extend it to new items. Which is really the main use of categories is that when you encounter something new, you can apply the knowledge you have about its category to know how to interact with it and make predictions about it and know how to respond to it. And the categories that are not based in reality at all are just not going to be that way and they’re just going to fall away. You don’t even have to decide they’re not real. People will not be using them because they’re not. That they’re not. They’re not useful. On the other hand, the world has a lot of different possible ways of being grouped together, which. So you already mentioned about, you know, the people person categories are extremely numerous for any one person. And there’s lots of ways that we could divide up the biological world or the physical or the world of artifacts that we all Live in lots and lots of ways, you know, and there’s some degree of arbitrariness or at least human interest that tells us which one we want to do. So there’s often kind of a conflict between, say, biologists who are going around, you know, deciding what are the categories and the taxonomies of biological things in the world and ordinary people who are interacting with the world. And then, you know, you read an article that says something like, you know, rabbits are no longer rodents. And then you’re like, okay, great, what do I do now? You know, so what’s the word for those animals that come and, like, cut, you know, eat up my plants or eat the bird seed or whatever, like the rats, the mice, the rabbits, like. And, you know, you don’t really pay that much attention to them. And it’s not that the category you’re using isn’t real. It’s just a different category with other criteria that’s useful to you. So there’s an example of, boy, I’m not sure whose it is anymore. It’s in the book. But somebody who talks about the category of shellfish, and it turns out that shellfish, you know, and this is. If you go to the fishmongers, is the idea, like, to get shellfish, or you’re a cook and you’re going to get shellfish. Like, I need all the different shellfish for my dish. You’re going to get a lot of things that are not related to each other biologically at all. You know, you’re going to get clams and oysters, but also maybe, you know, crayfish and crabs and some lobster tails or whatever else. These things are not in the same biological category, not near to each other. But it’s a very useful category. It’s a real category because they all share certain properties that we’re particularly interested in when we’re eating them. And therefore, for us, it’s a perfectly good category that doesn’t do the biologist any good, and vice versa. Some of the biological categories are very good, for example, from some biological perspectives, like putting things together that share an evolutionary history, but not good even for other biological perspectives, like putting things together that share an ecosystem, which is something that other biologists will do and that some people are very interested in.
Andy Luttrell: Speaking of tomatoes, the part of your book that just changed my life was dismantling the superiority I got to feel by pointing out that tomatoes are not really a vegetable, they’re fruit. And it totally. I just. There was something about it that just really gripped me because you know, who really cares? But so the delights that people take in correcting each other over this inane category distinction, do that. And I think you not only just get to like, allow us to roll our eyes at that, but give us a new way of thinking about what, why we shouldn’t care if botanically a tomato is one thing, if it conflicts with our ordinary, everyday experience of it categorized in a different way.
Greg Murphy: And part of the problem is due to, and we haven’t really touched on this yet, it’s, you know, the misleading aspect of language, the way that language. So I said before, when we talk about things, we use these categorical terms all the time. You know, like, I eat a lot of fruit. The problem is that the language arose spontaneously, you know, to fulfill different needs. And it was not made a. It was not designed in advance to be unambiguous and clear about everything. And the result is that we have a word like fruit, which is been taken by botanists to refer to the product of a fertilized flower on a plant, right, that contains seeds, something like that. And that’s great for botanists. That’s a really good category because, you know, you need to know how plants reproduce. And things that have fruits have different properties and different ways of living the environment than things that don’t have fruits. And that’s great. But it turns out it includes all these things that we don’t call fruit, like corn and almonds and beans, I think. And the, the more non technical use of the word fruit is a culinary use that came about through, you know, people picking things off trees or off bushes and eating them and saying, oh, look at these things, they’re really good. They’re fruit. And anything that you could eat that was picked off a tree or, or a bush was then called fruit. But, you know, things of that sort. And it’s because those things tend to be sweet and kind of soft and good, eaten cold, they don’t need to be cooked and so forth. You know, those properties began to sort of adhere around that category of fruit. Almonds, you know, don’t really look like those things. They don’t taste like those things. And tomatoes they grow on a plant, right, but not a tree, not really a bush, and they’re not sweet and you don’t eat them for dessert, and they don’t, you know, you can’t eat them cold. But mostly you cook with them. And so somehow they didn’t make it into our informal category of fruit. Now, you know, both those categories are perfectly good. If You’re a botanist and you start saying that a tomato is not a fruit, you’re gonna, you know, you’re gonna be in trouble if. But on the other hand, if you’re talking about making a fruit salad and you’re putting a lot of beans and corn and, you know, and almonds, the people are gonna think that, you know, you mislabeled it. It’s not. There’s, where’s the fruit? So that really leads to unnecessary confusion because both uses are perfectly normal and right. And it’s just a shame that they have the same name, really. And I think maybe the botanists are to blame on this, that they called the apples and the pears and the blueberries, they called them fruit and then later discovered, this is my guess, discovered, like, oh, beans are basically the same thing. We have to call them fruit as well.
Andy Luttrell: I have this same dilemma just in our world of psychological science of, like, we study things and we have constructs and concepts, but we’re kind of beholden to language that already exists for them. And sometimes it sure would be nice if we could just like, start over and say, like, there’s this experience that has these qualities and forget what we’ve called it before. That word is not helpful for us to categorize. I tell students all the time that, like, the real dilemma of being a psychologist is that we study a lot of times the words people use for the things that they do.
Greg Murphy: Right.
Andy Luttrell: Like, when I design a survey, the only way I can understand my construct is using the words that people use to describe it, but it’s not always a match. Right. Like, that kind of lay definition of a concept may not match my scientific definition of the same concept. And then here we’re in trouble.
Greg Murphy: Right? Yeah. And when psychologists make up technical words that have non technical meanings, it can lead to problems. You’re just like, okay, well, we don’t. I know why you’re thinking you’re saying that, but the word fright doesn’t mean the same thing to Freud that it does to you. And then you’ve got to go, okay, well, now I’ve got a new meaning for fright I’ve got to figure out. Yeah, it’s a problem. Language is a big problem, I would say. And, you know, I’ve studied this phenomenon of polysemy, which is multiple meanings of words. And one of the interesting things is that it’s the oldest words in the language that are the most common, that are also the most polysimus. So if you look up a dictionary and you just Count how many numbered definitions there are like under a single entry, right? So the really common words are loaded with them. You know, a word like stand, a verb like stand or bear, to bear, B, E, A, R. You know, it’s, it’s like a whole, it’s like half a page of a dictionary to list all their different meanings. No, really, it’s an, it’s a very large number. And even some of them, you’re saying like, well, this could actually be divided into a couple of different things as well. So in terms of communication, it can be very difficult to know exactly what sense of a word the person has in mind. And there are some classic categorization cases like this, like fruit. There are some others that are not. They’re a little more systematic, like some of these kind of value laden terms. So a word like art, which, you know, you can use to refer to, say, decorations you put on the wall or, you know, on tables which are primarily there for their decorative purpose or something. But then there’s another meaning of art that has to do with embodying true, what we think of as true artistic values, you know, so as being inspirational and creative and somewhat novel and that, showing, showing skill. And so I use this example from my poor brother who does his paint by numbers painting of, you know, sailboats on the bay. And like, you know, I have to tell him this is not art. You know, it’s not, you know, you, like you. It’s, it’s a cliched pattern. Like the colors are simplistic, there’s no skill in the execution. There’s nothing inspiring about it. Like, this is not art. And, you know, and he can just as easily say, look, it’s a painting. You put it on the wall. It’s art. What else could it be? And it’s because we sort of have these two ideas about some of these value laden terms like art, one being sort of a normative value and the other one being a descriptive term. And you know, that leads to confusion as well because there’s a number of terms like this and a lot of them referring to professions or human activities that we, you know, ascribe both of these kinds of meanings to. And that leads to difficulty in categorization. I think sometimes that you’re, you’re really arguing over two different, two different things rather than, you know, even he will admit, okay, it’s not that creative. This is. But this is a made up example. My brother does not.
Andy Luttrell: Okay, he’s off the hook.
Greg Murphy: No, yeah, he doesn’t have that Much talent.
Andy Luttrell: Similarly, though, I mean, I think that culturally we’re having this reckoning with AI generated products in the same sort of way of, like, what does this. It’s sort of forcing us to contend with, like, oh, there was this implicit assumption in what art was, that if exactly the same image was generated through a different process, how do we feel about it now? And so, like, as. As the times change, so too must our categories and the way we think about them.
Greg Murphy: Or maybe getting off topic here, but I think a lot of our appreciation of art and my enjoyment of art is trying to figure out what does the artist have in mind here? What is he or she trying to, you know, accomplish? And once it’s the AI that’s out the window, you know, it’s generating a new pattern based on previous patterns it’s made. And there’s no. There’s no. In fact, perhaps it does have some deeper significance to it, but it’s not intended. You know, it’s not. It’s not there in the program. And I. You know, maybe we are over overestimating, you know, human. Human creativity and value by attributing those things to the artists. But. Yeah, I mean, how much time do you want to spend thinking about the inner meanings of some computer program, you know, that you have never met and don’t really know anything about?
Andy Luttrell: Yeah, it seems, though, that, like, you know, until there, like, that just wasn’t maybe not a consideration before because anytime you’d see something, you could easily assume there was some degree of intentionality there. And now that that’s not a given, you go, well, wait, is that essential to my category of arts, or is it a feature that distinguishes some art from other arts? And now we’re playing the category game.
Greg Murphy: Yeah, well, I mean, there was a time in which very cheap Russian paintings were being imported into the country. Have you seen them? You would go to, you know, there would be some kind of art sale at a local hotel, and you go to the lobby and there would be a hundred paintings on the wall. And they were all very kind of scenic things, you know, with a few. With a few adorable children thrown in. And it turns out they’re made literally in a factory, you know, painted by hand with very cheap Russian labor back in the time when we were then trading with Russia. And they would just put these out, you know, they would. There’d be somebody like, this week, you’re doing the sparkling lights at moonlight, you know, with the. Over the snow. And they would do that same painting over and over and over again and then send it, you know, off to be exported. And you know, some people thought this is a beautiful scene, sparkling moonlight, you know, and on the snow and put it in their living room. But I think the same question arises. Is that really art? It’s art in a descriptive sense. But once it becomes formulaic like that and done to instruction and repeated over and over, you just think that in the kind of normative sense, this is not why we’re interested. Most people are interested in art. This is not why we, we value. So it’s not really art.
Andy Luttrell: Similar dilemma. This past weekend we went to all these, you know, Christmas time markets and one of the ones we went to was an art fair and the other was a craft market. And I had this sense of like, well, which of them am I in right now? Like, is this a meaningful distinction or is it not? And there again, I mean, coming to language, right. It seems like there was like an arts and crafts movement historically, you know, late 19th century arts and crafts movement that was sort of of like repurpose these terms that we had and sort of put them in new lights. And so maybe this is an opportunity also to get back to that language question, which is like, it seems like language evolves independently of the evolution of like category meanings. It seems like. Or cross culturally. And what I’m trying to do is tee up this study that I love that you talk about in the book.
Greg Murphy: That I used to try to explain.
Andy Luttrell: Which is all these many words that have sort of independently emerged, but that don’t necessarily match the intuitive categories that we bring to bear.
Greg Murphy: Yeah.
Andy Luttrell: So I wonder if you could sort of take the reins and fill us in on that.
Greg Murphy: Right, so you’re referring to this very interesting study by Barbara Malt and her colleagues who had the very peculiar idea of going to the supermarket and getting like 60 containers of things when they presumably ate all the contents of the containers. That’s not really explained. But so. And they’re things like juice boxes, mustard. Okay, well, Andy, what do you, what do you put mustard in? You buy it at the supermarket.
Andy Luttrell: I would probably call it a mustard bottle.
Greg Murphy: Would you?
Andy Luttrell: Yeah.
Greg Murphy: Okay. What about, what about pickles?
Andy Luttrell: Pickle jar.
Greg Murphy: Okay, thank you. All right, so yeah, so they got the pickles, they got the mustard, they got juice boxes of various shapes, they got shampoo, they got all these things that are liquid or semi liquid things that need some kind of container to hold them. And they went to three different places to test what people call these containers. So they wanted to know what do they call them not the ingredients, but what do they call the container itself? And so they went to Argentina and got Spanish speakers. They went to, I think, Shanghai and got Chinese speakers. And then Moltz Lab is in Pennsylvania. So they had English speaking Americans there. And so they had them label all these objects and they looked at what was the similarity of the labeling across countries. Now, actually wrote down some numbers here, so I wouldn’t make it up entirely. So for English, there were three main names of jar, bottle, and container, and that named like 50 of the objects. And then there were just a few extras that got some of the leftovers. Spanish has one main name, frasco, which names 28 objects, which is I think we would call a bottle. Then it had a lot more specific names for one or two objects. So Spanish had more names than English. And The Chinese had one word that named 40 objects. So that’s far bigger than any of the other categories of the other languages. The second one that named 10, and they just had a couple of others. So these are really, you know, very different names to very different ways to name objects. So English has kind of a medium number of names that are used moderately often. Chinese has one name that’s used most of the time, and Spanish had one somewhat popular name and then a load of more specific names. So they have a way of doing a correlation among the naming, like which objects have the same names, and then comparing that across languages. And the correlations were pretty moderate. They were like around a 0.5. So you might be thinking that, okay, based on this, the speakers of these different languages must think of these objects somewhat differently as well. So when Chinese speakers look at most of these objects, they just say they’re like, okay, they’re all containers, right? They’re all the same container. Container, container. Okay. There’s an actual Chinese word which I will not attempt to pronounce, but anyway. And Spanish speakers must have seen them as being very different. You know, this one is a bottle, this one is a tiny jar kind of thing. And then they had different names for all of them, and therefore they must think they’re different kinds of things. So then Malta l did this measure of conceptual structure, which is very common in the field, which is you get people to rate the similarity of the objects. So you show them the tomato, the ketchup bottle, and you show them the pickle jar, and you say, how similar are these two containers.
Andy Luttrell: Words?
Greg Murphy: Right? No, you show them the pictures and again, explain that. We just want to know about the containers, not asking how similar Ketchup is to pickles, right? And so, yeah, they do a lot of ratings and then you can basically put them into an analysis program that can tell you, like, how similar are the whole domain of objects. What is the structure, the similarity, the whole thing. So it tells you, like, what is the likely categorical structure? Like, some things are clustered together over here, other objects are clustered together over here. Then there’s another object that’s off by itself, that’s the tube or something, that doesn’t get the same name as anything else and so forth. What they found out was that basically the structure of the categories was identical for the three populations. The correlations were all over greater than 0.9. And the fact that the Chinese called two things by the same name didn’t necessarily mean that they thought, yeah, these are the same kind of container. And the fact that the Spanish speaker has called them by different names didn’t necessarily make them think that these are really different. The influence of language on the categories that people seem to have was less than you would less, to put it mildly, less than you would think. On the other hand, I think that we do often rely on language to tell us about categorical structure that we may not know about. So an example is this is from a work of fiction, but like guinea pigs. Guinea pigs are, you know, they’re a small pet in the us, right? Somewhat cute, furry things. And I think most people would say, you know, they’ve got some relationship to pigs because they’re called pigs. And if they don’t have any relationship, why are they called guinea pigs? Right. Well, they’re called guinea pigs because some dumb English speaker went to South America and found them and decided to call them guinea pigs. Not a biologist or anybody who knew anything about how related different mammals are. But whoever it was had to come up with a name and without much time to think about it, apparently called them guinea pigs and they don’t have anything to do with pigs. But that seems surprising. It seems surprising to me even now. I might have thought, well, they’ve got. There’s something a little bit there. And, you know, that’s not that uncommon that things, especially newly discovered by some language members, objects are given a name that’s from the categories you already know. So the English speakers go down to Australia and they start, you know, calling things, you know, Tasmanian pine. Now, is it really a pine? Is it similar to pines of Europe or that. Well, they weren’t biologists, but, you know, what else, what else were they going to call it? And then Years later, when people are thinking, like, what is a pine? And then you find out that Tasmanian pine is not even a pine, you know, then you don’t, you know, you’re kind of thrown for. You feel like the languages let you down, which it kind of has. But just as you were saying with, you know, in psychology, you can’t just go and rename things. Right? No, we, you know, have you said we’re not going to call guinea pigs guinea pigs anymore? We’re going to give them their Peruvian name, Something else like, well, good luck. I mean, it’s not a bad idea, but, you know, how are you going to make it happen? Because guinea pigs out of the bag. Yeah, yeah, guinea pigs are out of the bag. I mean, we can’t even change over to the metric system. Like, how are we, how are we going to change a thousand names for all these animals and plants that should not be called that, you know, and the house sparrow isn’t a sparrow. Like, well, it’s not. It’s not a sparrow like we have in the U.S. so, you know, we’re just, we’re kind of stuck with that and having to remember the house sparrow isn’t really a sparrow.
Andy Luttrell: You know, the, the jar study, I think why I love it so much is it’s just such a potent reminder that we can’t rely on the words as like a proxy for the categories that people use.
Greg Murphy: Yeah, right.
Andy Luttrell: And so, like, they, they’re, they’re not completely divorced from the categories that people use, but they’re not reliable indicators that you can have totally different sets of words for things that have emerged for any which reason. But it doesn’t mean that that is a good cue to, like, oh, people think of these things similarly. Right? Like these, these jars are the same and these tubes are the same, and these bottles are the same. That it’s just not that, that, like, we need to understand and sort of. It helps keep our debate sensibilities in check when we want to start saying like, okay, I’m going to anchor on the truth that jar is a real category. And now I’m going to fight you over whether this thing is a jar or not. No, it was. Jar was never a thing to begin with.
Greg Murphy: Yeah, well, part of the reason for this, it seems that the accepted explanation for the problem with jars is that sometimes these names made more sense a long time ago. So your shampoo bottle was actually a glass bottle that looked a lot more like a soda bottle, you know, than the ones are now, which are plastic. Flip top Containers usually that don’t look anything like a wine bottle or whatever else, but they, you know, they did then. And juice boxes. So, okay, they have a picture. What if there’s a juice box in the shape of a Smurf? Well, whenever, whenever the English language was invented that did not exist, you know, and then when they made it, they gotta, they’ve gotta come up with the name of it. And in some cases there’s commercial reasons, like companies are making the object and they want it to be connected to something else, maybe something it doesn’t really deserve to be connected to, but which for marketing reasons would be good to do it. So that kind of reasoning, which if, then it gets adopted into the world at large, the language at large, you’re stuck with a bad motivation for this name, which, you know, really is not accurately describing the object. So, so it means that we kind of have to keep these different meanings of the word separate in our heads. You know that some bottles are plastic and have flip tops, some are made of glass and have cork in the, in the top, you know, and all the different things that they hold, you’ve just gotta, you’ve just got to learn it all, you know, and there’s no way around it. And, but it doesn’t stop you from making, you know, categories of say, all glass containers, which may not have a simple name in English, but it could be a way that you, you know, think about things when you come across them. And the same way you may have categories of, say, people that don’t have a simple one word or short phrase description, but is a category that you made up. And you go like, oh yeah, she’s one of those, you know, positive or negative. And that affects your interaction with the person as well. So because even though there’s no name, you’ve kind of noticed some similarity out there in the world among certain kinds of people.
Andy Luttrell: I have a 4 year old daughter who’s very sharp and picking up all sorts of words. And we, we’re now like facing this all the time of like, okay, you, you did catch me, like I used the same word to talk about two totally different things. And now I have to explain to you that like, there’s no good reason for that. You just kind of have to put both of them in your bag.
Greg Murphy: Yeah, it’s embarrassing. Why is, why is the thing you swing at the baseball have the same name as the animal? It’s, I mean, it’s, it’s awkward. And, and there’s a reason for each actually, which actually uses Example in the book, that they’re derived from two different sources. Like the word bat derived from two different sources, one from French and one from, they think, Scandinavian or Northern German of some kind. And originally they weren’t both bats, but through phonological change in historical English, it turned from baka or something to bata to bat to bat in American English. Anyway, so there it is. We got two bats, you know, and they. It wasn’t that anybody ever thought they were the same kind of thing, but that’s the way the language generated, you know, and yeah, if only we. If only we’d listened to people and kept with Esperanto.
Andy Luttrell: Guinea pigs out of the bag.
Greg Murphy: Yeah, exactly. But even that, you know, it also saves us making up new words. So, okay, English is a little bit awkward in this regard. But you think of a horse as the animal, right? If you want to go eat part of that animal, you would say, yeah, I had a nice horse stick last night. We ate some horse, we made a horse stew. You would use the same word to describe the meat as the animal. And that’s a productive kind of meaning ambiguity, which is used for lots of things. The word for the plant is used when you eat the plant. You don’t make up a new name for rhubarb when you put it in a pie. Now, English has this weird thing like pork and beef and a couple of other things which turn out to be due to the influence of French. When the French invaded England. So in the old days, actually, I don’t know what they said. I assume they just said, let’s eat some pig, right? I don’t know, let’s eat some boar. But the French came over and they talked about pork, you know, which is a French word, and buf. Which is the word for beef, Right. And what’s usually said about this as well, given the difference between English and French cooking, it seems wise to adopt the French words for the food. That kind of makes sense, but that’s rather atypical and in many languages. So in French, it’s not that way. The pork is the pig and the boeuf is the animal and the meat, right? So. So the fact that we have these different meanings is a useful. Can be useful for generating words for things even when they’re not in the same category. But the good news is there’s hardly any confusion over that. So if somebody says, like, I ate some, you know, in French, I ate. I ate pork last night. They don’t. You don’t think that maybe it was, you know, grunting and had hooves and, you know, it was a whole living pig. And if somebody says, I’ve got five porks on my farm, you know, you don’t think they’re chops, you think that they’re actual pigs that live in the Barn, you know, etc. So that doesn’t lead to that much confusion. But in other cases, like the art case or the tomato, the fruit case or whatever, it can leave in. And technical definitions, like scientific definitions, are often kind of the. The problem.
Andy Luttrell: I like your perspective, too, because it seems to just empower the people in that, like, you know, it sort of reminds me of kind of pedantic grammar. People who are like, no, there’s a certain way things are said and spoken and spelled and put together, but language is also, like, there’s no reason to think that we should. There’s no inherent meaning to those rules, only to the extent that they’re useful for communication. Same with categories. I’m going to read you a quote. I think this quote from your book really captures all of these things. Things which is there’s no mechanical rule to tell us whether Pringles really are or are not potato chips. So we have to decide why we want this category and how Pringles fit into that picture.
Greg Murphy: Right.
Andy Luttrell: Which is just this great example of a modern object that we need to force into a category. But that whole undertaking makes us wonder, like, well, what do we need this category for and why do we want it to fit into it?
Greg Murphy: Right?
Andy Luttrell: Which is just the inherent subjectivity of categories.
Greg Murphy: Right? And just, you know, for a little context, there was a big lawsuit in which the Pringles company said that they were not potato chips because they didn’t want to be subject to the VAT tax in England and Great Britain. And therefore they were arguing, you know, look, we’re not potato chips and we’re not. Part of it was like, we’re not at all natural. We don’t have that much potato. All these terrible things about their products in order to try to get out of this classification and these legal decisions. I mean, this is an important use of categories. The law makes categories all the time. And, you know, tax law is really obvious because we tax some categories of objects differently from other categories, immediately raising all kinds of questions like, well, which category is it in? So if Pringles are not potato chips, then they’re not subject to this particular snack tax that other potato crisps in England were subject to. But there’s no law to tell us what the right answer is. Here we have to kind of decide what we want the answer to be, especially when it comes to legal things. And the legal answer is, may be different from what’s useful to us in everyday life. So if people decide, like if the courts decide that Santa Claus costumes are not clothing, you know, you can still wear one as clothing. You can put it up in your closet, you can wash it, you know, in the wash. You can do whatever it is you do. The fact that the courts have decided they’re not going to tax it or exempt it from taxation, you know, whatever it is because it’s a costume and not clothing, doesn’t mean that we have to accept it that way. Just as the law that says that some of that white gooey stuff on the shelves near the mayonnaise is not really mayonnaise. People buy it and use it as mayonnaise, and they refer to as mayonnaise. They may not even notice that it doesn’t say mayonnaise on the jar. So because, you know, it’s all the white gooey stuff you put on your bread or make into a salad dressing, you know, use as part of a salad or whatever, that’s a useful category to refer to things that way, even if the FDA has told them they can’t label it that way. So neither one of these ways of categorizing things is necessarily wrong. What I think is wrong is to say that, well, some government agency told us this is the category, and therefore you have to use it this way or you’re making. You can be told off. You can be ridiculed if you don’t, you know. And by the way, apparently the Supreme Court, in spite of what some people say, did decide that tomatoes could be taxed as a vegetable. They said, well, botanically it’s a fruit, but, you know, in every other respect is treated like vegetables. And the intent of the legislation, when they called it vegetable, you know, they wanted to do it was some 1920s protectionist case. They actually decided that it was a vetrillo. So it’s a rare Supreme Court decision that we can all agree with.
Andy Luttrell: Well, at least we have that.
Greg Murphy: I want to say it would be a 4, 3 decision now or something. It would be very close. 5, 4 decision. Yeah. Oh, well.
Andy Luttrell: Well, out of respect for your time, I want to thank you for talking about categories with me. This has been super fun and I appreciate you taking the time.
Greg Murphy: This has been enjoyable. Thank you very much.
Andy Luttrell: Alrighty. That’ll do it for another episode of Opinion Science. Thank you so much to Greg Murphy for taking the time to talk about categories and what they do for us psychologically. As always, you can check out the episode webpage for links to more about him and his book again is Categories. We live by how we classify everyone and everything. Be sure to follow this show Opinion Science wherever you listen to podcasts. Happy New year to you 2026. We’re here, we’re doing it. Let’s get to business and make things happen this year. I know you can do it. I’ll be here as always, still on the show’s monthly release schedule. I tell you, it’s been really helpful that I still get to share all this cool stuff with you, dear Listener, without burning out on the very regular release schedule that I used to have. Who knows, maybe one day I’ll be able to hire an editor, a social media maven, and a production assistant and then we can do more. But for now it’s a one man band and that man also has a full time job. But how do you support such an operation? Tell the world. Please write reviews. Suggest the show to people who would be into it. Do it. Do whatever comes to mind. Also, I don’t say this enough, but I’m very open to listener feedback. Get in touch with me if there’s something that you’d want to hear more of, or if you have any thoughts about making the show even more accessible to folks or, I don’t know, even. Just let me know that you listen and that you like it. It still astounds me when I get emails from people from all over the world who happen to hear about this show and listen to it. So let me know you’re out there. I’m otherwise literally alone right now in my office talking into a microphone, having an imagined conversation with a person that I can’t see. But that person is you. And I’m grateful you’re there. Okay, that’s it from me. I’ve got important 2026 business to attend to, so see you next month for more Opinion Science.
