Categories We Live By: How We Classify Everyone and Everything

I Basics of Categories

Our investigation into categories is divided into two parts. Part I, comprising chapters 1–4, considers the general character of categories: where they come from, how they’re related to the world and society, and their connection to language. Part II, comprising chapters 5–12, considers case studies, showing how those general issues play out in specific examples, which are often surprisingly complex.

I have kept references in part I to a minimum. The material presented consists mostly of well-established facts and arguments that are found widely in the philosophical and scientific literature. If readers would like to know more about the scientific evidence, a good source is (ahem) the first three chapters of Murphy (2002), but they are much longer and can get technical. I do cite specific experiments and theories, and readers can find full references at the end of this book.

1 Introduction

If you were born before 1960, it is very likely that the first words you ever heard put you into a category: It’s a boy! or You have a healthy girl, Mrs. Smith. (In those days, parents didn’t know the sex of a baby before it made its first appearance.) Sad to say, the first thing said about you wasn’t about your bright prospects, your sparkling personality, your unique heritage, or even your wonderful ability to kick and cry. Instead, you were categorized and from then on treated as a member of that category.

For some children born in the twenty-first century, that categorization happens even earlier. Because of the wonders of ultrasound, most American parents learn their child’s sex during pregnancy, perhaps via an explosion of blue or pink confetti during a (misnamed) gender reveal party. Yes, now they don’t even wait for your birth to categorize you—they’re categorizing you six months before you’re even born!

Of course, people are categorized all the time. Walking down the street, we see and think of people as boys, women, police officers, elderly Black men, hipsters, Chinese people, Orthodox Jews, street musicians, or high school students. We categorize people we know as alcoholics, introverts, rich people, obsessives, grandparents, doctors, and so on.

This is not simply a testament to the human desire to classify other people. We do the same thing with objects we encounter every day, categorizing them as BMWs, sippy cups, fireplugs, lawns, poodles, or forks. Like people and objects, events are classified into groups too, such as parties, production meetings, arguments, going to the movies, wars, and giving birth.

This book is not about parties or sippy cups or boys; it is about the human need to classify. Although every object, person, and event is unique, with its own history, composition, and set of properties, we nonetheless group things into categories that ignore those unique aspects, and then we think about them in terms of properties that they share—or that we think they share. Every party is different, but we use the word party to describe all of them. It is a bit strange that a crowded gathering in a loud, dark hall filled with people dancing to thunderous music gets the same name as a small group of people sitting in an all-white decorated apartment, clutching cocktails and awkwardly trying to make conversation, with no music and long silences. Yet both are parties.

Why do we do this? And where do these categories come from? The categories that we use over and over seem to have a reality that cannot be questioned. There are really doctors and lawyers, of course. Some trees are really oaks, and others are really maples.

Attending a party is not the same as attending a movie. But we could have used other categories, such as classifying people by their age, by their height, or by their income, rather than dividing them into doctors and lawyers. Parties and movies both involve groups of people engaged in activities that they hope will be enjoyable. If we had wanted, we could have made a category including both of them instead of separating them. And although botanists certainly distinguish maples and oaks, many people group them into a single category like tall, leafed trees. Others divide up trees according to their functions in landscaping, like specimen trees or ornamentals.

Such examples inevitably raise the question of whether categories are real or made up, or perhaps make us ask which are real and which are made up. Where do categories come from? If they come from us—our heads or our societies—then they do not seem entirely real; rather, they are constructions. If they come from the world itself, then they must be real. It seems pretty obvious that dividing trees into categories like ornamental, shade tree, fast-growing tree, or specimen tree is a human construction. But shade trees are alike in some ways, and they serve similar purposes in landscaping. Is this really not a real category? And if we doubt that, then perhaps we should doubt other categories, like oaks, which play a role in our biological theories. After all, those theories are also human constructions.

Does someone decide which categories are real and which are not? If you call a palm tree a tree, is there an expert who can tell you that it isn’t actually a tree (as some people think)? If so, who gave that expert the authority to decide our categories? If I keep calling a palm a tree, what harm am I doing?

This book tries to answer these questions. They are important because, to paraphrase Sigmund Freud, categories are destiny. If you are categorized as a boy in our society, you will likely do a number of things, and a number of things will likely happen to you that are different than if you are categorized as a girl. If something is a cat, it might be allowed to reside in your apartment, but if it is a dog, it might not be. If you are convicted of a felony, you are going to have a lot more problems than if you are convicted of a misdemeanor. Of course, felonies differ in many respects, from their details to their seriousness (e.g., passing bad checks versus mass murder), but we nonetheless group them together to decide your punishment, which prison you’ll go to, and whether you can vote afterwards (in many states). Your insurance company makes categories of accidents and disasters and determines whether it will reimburse you depending on whether your situation fits its categories. A burst pipe that ruins your basement might fall under a plumbing problem that is covered, while the same amount of water from rainfall is a flood that is not covered—even though the damage is identical in the two cases. If you have a virus that is causing cardiac problems, your health insurance might provide full coverage. But if the cardiac problems are coming from an inborn heart defect, they are a pre-existing condition that might not be covered. As a result of such consequences, people who get to make up these categories wield an enormous influence in our society.

If there are categories that exist independently in the world, then they may determine much of what happens to their members. An apple tree will grow apples, not oranges; a cat will meow rather than bark; a piece of granite is likely to be harder than a chunk of shale. For these natural things, categories may tell us a lot about the nature of the objects in them, and as such they are extremely useful.

Nowadays we are suspicious of categories, especially categories of people. Gender, ethnic, racial, and class categories are seen as a source of discrimination and inequity. If we saw each person as a unique individual, perhaps we wouldn’t have all the problems caused by racism, sexism, and all the other -isms that afflict society.

In spite of these problems, I will argue that we cannot get rid of categories in general. They simplify and distract us from individual identities, which can be bad, yet if we didn’t have them, we would find it impossible to navigate the world and deal with its incredible diversity. It is because we classify a unique object as a tomato that we know we can eat it. We don’t have to investigate the object and its unique properties or wait for the results from a lab test to tell us whether it is OK. Tomatoes are edible; knowing such things saves us hours of trouble in deciding, object by object, what we can eat. That said, categorical predictions like this cannot be absolute. A rotten tomato or one that has been in contact with ptomaine should be avoided. Relying on categories can save our lives (starving? eat the tomatoes!) or can lead to mortal errors (no, not that tomato!). Even person categories are extremely useful. When introduced to a friend of my nieces who is a finance major, I engaged her in a conversation about the stock market, which she knew all about. If they had introduced her as a dance major or a biology major, I would not have wasted both our time asking about stock valuations. Categories of people can help us adapt our behavior appropriately to that type of person and thereby smooth social encounters. Of course, such categories are not perfect and can lead to gaffes; we might be better off without some of them. But without any person categories, we would have a much harder time engaging with new people we meet.

Categories are so essential that we can’t count or calculate statistics without them. For example, how many are in your home right now? You are no doubt going to ask, "How many what? Exactly! The only way that we can count things is to categorize them into countable units: how many dogs, books, people, parties, donors, salads? You could try to count the number of things," but good luck with that. Is each page of each book a thing? Each thread of your clothing? Each pile of stuff? Will you count the families and nationalities and professions of all the people you encounter? They are things. There is no way to know what to count without first stating the categories that you want to tally. Statistics that we read in the news are inevitably based on an initial categorization scheme: Americans between the ages of nineteen and twenty-five; ash trees in New England forests; accidents on interstate highways; people who registered as independent but who voted for Joe Biden for president; and so on.

Thus, categories are essential to many aspects of our lives, so it is important to understand their nature, where they come from, how we learn about them, and how to negotiate their complexities. This book is an introduction to these topics.

The Psychological Power of Categories

Categories sometimes have effects that they shouldn’t. One obvious case is prejudice—when you decide to like or dislike someone because you found out that she is Asian or White, or a Yalie or a member of the Chamber of Commerce. But even in dealing with simple objects, categories can have an effect beyond what they should. For example, I have a friend who dislikes cheese. I put cheese in quotes to indicate that he dislikes anything that is called cheese. The problem with this is that cheeses are incredibly diverse. There are hard, nutty ones like Emmental; sharp ones like aged cheddar; tasteless ones like American cheese; soft, sweet ones like ricotta or mascarpone; ultrarich ones like triple-cream Camembert; ones that taste like milk, such as mozzarella; and so on. Some cheeses come from cows, but others come from goats or sheep (or even soy), and those have different textures and flavors. People who say, I don’t like cheese for some reason hate all these things: hard and soft, sharp and mild, creamy and nutty and funky—they hate them all. I don’t believe it.

If you told me that you don’t like sharp cheeses, I would believe you. If you said that you don’t like the texture of very soft cheeses, I would pity you, but I would believe you. But if you say that you don’t like any cheese, I don’t believe you because cheeses simply don’t taste, look, or feel the same. What my friend dislikes is something being called "cheese. I’m sure that he had a bad cheese experience in childhood. He ate something he didn’t like and learned that cheese is bad. With that negative attitude, he quickly rejected anything else called cheese (and very likely did not have a wide experience with cheeses anyway), reinforcing his belief that cheese is bad. As a result, this friend, who is in other respects normal, cannot bear to eat anything that he believes is called cheese." He actually thinks that there is some flavor or property of cheese that makes it taste bad to him. But that is not the case: cheeses don’t share a particular texture or flavor or source. He hates the category, not the actual stuff.

I think we know other people who have such dislikes. Once they’ve identified something as country music or the color yellow or a dog or a chick flick or whatever, they immediately reject it. Others have categorical likes: all French food, every single Bruce Springsteen song, or Ford vehicles. Of course, some Fords are good, but some are gas-guzzlers that are unsafe or unfun to drive. This is not to say that these categories are always unhelpful. If you liked Springsteen’s last few albums, it may make sense to buy his new album without hearing it first. However, if the album has a few losers on it, the true fan may refuse to recognize this. In contrast, the Springsteen-hater may dislike every song on that album, even if she would have liked a couple if she didn’t know that Springsteen had written them. Categories are helpful in allowing us to select and attend to things that we generally like or dislike. They help us decide which movie to go to, which person to date, which car to buy, and so on. On the other hand, faith in your feelings about a category can be overblown to the point that it obscures the values of individual members.

The other side of categories’ ability to influence our attitudes towards an object is their ability to explain things. Suppose that Stephanie tells Anna that Fred keeps chasing her cat around. Why? Anna asks. Fred is my new dog, Stephanie replies. This seems to explain it. Stephanie doesn’t have to go into any more detail about Fred, his personality, his early experiences and training, and so on. Merely by categorizing him as a dog, she has provided an explanation of Fred’s behavior: Dogs chase cats, after all. Fred chases this cat because he’s a dog. Now, this may well be true, but in other cases, the explanatory power of categories may be illusory.

Imagine the following exchange:

Anna: Why does Todd go on and on about taxing millionaires?

Stephanie: He’s a real liberal, you know.

Anna: Oh, I didn’t know.

The conversation could end here because Anna may well feel that by categorizing Todd as a liberal, she has now accounted for his behavior. But the parallel with the cat-chasing dog is not so clear. Does Todd talk on and on about this because he’s a liberal? Or do we call him a liberal because of his feelings about taxation? Is being liberal a real category that can make you do things, the way being a dog makes Fred do things, or is it just our own description of someone’s beliefs? Political liberals are a meaningful category (in my opinion), but it’s not clear that they have the force of categories like those from biology or physics. Granite is hard because it’s granite—it’s an inherent property—whereas it’s difficult to say that Todd is a liberal first, and as a consequence, he wants to tax millionaires. Rather, we call him a liberal because of his beliefs.

This doesn’t mean that there’s a problem with categories like liberal, but it does point out that different categories have different powers, although people may not discriminate among them very well.

Definition

If you’re going to write a book about categories, people expect you to define exactly what you’re writing about. As the next chapters will make clear, this may not be doable, but I will give a rough answer. One popular definition is that a category is something that groups together different entities that are treated as being the same thing, for at least some purposes. In most categories (and the most interesting ones), you can tell that the members are not identical, and yet you treat them as identical. You can clearly tell that a dachshund is not the same as a greyhound, but you call them both dog, feed them both dog food, take them both to the vet rather than to your doctor, and so on. You also probably think that they have the same kind of liver and the same blood chemistry. If your dachshund were sickened by eating some food, you would keep your greyhound away from that food too. The fact that you can see that objects are different and yet you treat them as similar is what makes categories interesting, and powerful.

Why do you treat category members as the same thing (for some purposes)? Generally, it’s because they have some properties in common. It would be absurd to take a random group of entities and treat them equivalently. Dogs can be treated as the same thing because of their similar body shapes and parts, behaviors, and biology. A dog, two trees, a wedding, a door hinge, and a song would not be treated as equivalent because they have nothing in common. There is nothing to stop you from claiming that they are all in the same category and calling them snookles, say. But snookles is not going to catch on as a category because its members do not share important properties, and therefore it would not be of use.

Now there is a problem with this definition of category, which is that it is too broad. It allows us to put in a category any items that share any commonality. So, my sweater and my hair might both be brown, so we can say that they are both in the category of brown things. They are equivalent in the sense that I call both of them brown. However, the category of brown things is not very interesting. When you know that something is in that category, you know exactly one thing about it—its color. There’s nothing else to be known. Maybe the brown thing is alive or maybe it’s inanimate; maybe it’s microscopic or maybe it’s as big as a planet; maybe it moos or maybe it sings or maybe it is silent. The category doesn’t tell you any of those things. These single-criterion categories are kind of degenerate. Yes, items that share the criterion (brownness, being three inches long, taking at least half an hour, or whatever) are equivalent, but only in the one property that defines the category. Indeed, you can define trivial categories that only have one ridiculous feature in common, like things that you have touched in the last forty-nine seconds, or objects that are exactly fifteen miles from Cincinnati. Those are not categories that people form, because they are not useful.

Real categories that we use are usually rich and informative. When you know that something is a dog, you are pretty sure that it has four legs, that it barks, that it lives with or around humans, that it has a four-chambered heart, and on and on. That is a category. It is this informativeness that makes categories useful to us. So, over the course of this book, when I talk about categories, I’m only talking about rich categories that carry a lot of information, like the examples at the beginning of the chapter: party, sippy cup, boy, and so on. Single-criterion categories may technically be categories, but they are not particularly interesting, and there is little to be said about them.

Overview

The questions facing us, then, are: What is the nature of categories? Are they real? Do they exist in