Articles / Categorization

Describing gender categories: clusters and radii, Rosch and Lakoff

A while ago on the My Husband Betty message boards I posted an analysis of the category “woman” in contemporary American culture, and where transgender people fit into it. A couple of months ago there was another discussion about this issue, and since then I’ve wanted to rework my original post and make it available here. In my view, a lot of discrimination against transgender people has its origins in overly rigid views of gender. I have no illusion that posting my analysis here will suddenly enlighten bigots around the world, but I hope it will be helpful to some people. On the other hand, it’s not quite as helpful to the transgender movement as some might like.

It’s important to note here that this is a descriptive analysis. I think it’s a waste of time to present the way you want things to be before you figure out how they are. I am trying to describe the way that people understand gender, and after that I will talk about how it could be different. Please don’t take my description of an attitude or belief as an endorsement of it.

My analysis is based on the categorization theories of Eleanor Rosch, as presented in George Lakoff’s excellent book Women, Fire and Dangerous Things. In Chapter 4, Lakoff shows how it is more useful to describe the category of mother with a “cluster model” than with the classical categorization model that uses necessary and sufficient conditions. For example, Dr. Johnson defined a mother as a woman that has borne a child; Lakoff calls this the “birth model” of motherhood. But Lakoff identifies four other models that are in wide use: a genetic model, a nurturance model, a marital model and a genealogical model. He invents a series of more or less plausible sentences with the phrase “real mother” in them, each one affirming one of the five models and rejecting the others.

Critically, mother includes mutually exclusive subcategories like surrogate mother and foster mother. Some people may try to be fundamentalist and dogmatic about the birth criterion, but most agree that both of these kinds of mothers are still mothers in the end.

In chapter 5, Lakoff observes that mother is not just a cluster category, but a “radial category” based around the sterotype of a housewife who nurtures the children that she has genetically conceived with her husband and then given birth to. Other kinds of mothers, such as working mothers are defined in contrast to the stereotype. Lakoff writes (page 80):

Consider an unwed mother who gives up her child for adoption and then goes out and gets a job. She is still a mother, by virtue of the birth model, and she is working–but she is not a working mother!

Similarly, it seems strange to think of egg donors as unwed mothers, because the category of unwed mother is based on the birth model. To identify normal (I’d say normative) expectations, he gives a “but” test:

Normal: She is a mother, but she isn’t a housewife.
Strange: She is a mother, but she’s a housewife.

Lakoff goes on to list other kinds of mothers: stepmothers, adoptive mothers, birth mothers, foster mothers, biological mothers, surrogate mothers, unwed mothers and genetic mothers. Mother is a radial category because people judge these kinds of mothers as being good examples of motherhood based on their distance from the prototype of the housewife.

If you’re at all familiar with the endless transgender terminology wars, you’re probably already thinking of ways that this could be applied to gender. Here’s my take on it.

Like mother, woman in contemporary America is also a radial category, based on a stereotype of a human being with XX chromosomes, a female reproductive system and secondary sex characteristics. The stereotypical woman is shorter and less muscular than average, with a higher-pitched voice. She was born and raised a girl and went through a number of female rites of passage. She is submissive and indirect, in sex as in the rest of her life, and she puts family before career. She is attracted to men, and plans to marry one if she hasn’t already. She wears skirts, high heels, jewelry, make-up and long hair and manicured nails. And yes, the stereotypical woman is less intelligent than average.

Anyone who’s been involved in a gender-based flamewar or two can probably remember a number of sentences with “real woman” in them, and can point to non-prototypical kinds of women: tall women, strong women, deep-voiced women, assertive women, women who top, career women, lesbians, unmarried women, women who don’t wear skirts/high heels/etc. Here are a few “but test” examples:

Normal: She’s a woman, but she’s a tall woman.
Strange: She’s a woman, but she’s a short woman.

Normal: She’s a woman, but she’s assertive.
Strange: She’s a woman, but she’s submissive.

Normal: She’s a woman, but she’s a lesbian.
Strange: She’s a woman, but she’s heterosexual.

Normal: She’s a woman, but she doesn’t wear make-up.
Strange: She’s a woman, but she wears make-up.

The category of man is similar, and I’ll just skip to the “but test” here:

Normal: He’s a man, but he’s got a high-pitched voice.
Strange: He’s a man, but he’s got a deep voice.

Normal: He’s a man, but he takes care of the kids all day.
Strange: He’s a man, but he works all day.

Normal: He’s a man, but he’s not very good at sports.
Strange: He’s a man, but he’s very good at sports.

Normal: He’s a man, but he has long hair.
Strange: He’s a man, but he has short hair.

Man, these stereotypes suck, don’t they? Fortunately, a lot of people out there realize that there’s a wide range of gender traits and celebrate this diversity.

So the categories of man and woman are radial and prototype-based, but are they cluster categories? Can they contain members with non-overlapping traits? There is certainly room in some people’s gender categorization systems. I know plenty of people who would accept both a butch lesbian and a femme non-body-modifying transwoman as members of the category woman, but I don’t know how typical they are.

My feeling (and again I’m being descriptive here) is that the fundamentalist interpretations of gender categories are stronger than with mother. A lot of people insist that genitals are the essential criterion, and are unaware of the wide range of intersex conditions that complicate that picture – or uninterested in them.  Many people will either ignore exceptions or put them into a catch-all category of “freaks.”

A lot of this seems to be due to the fact that there are two categories, woman and man, and they’re generally considered to be all-inclusive (everyone is a man or a woman) and mutually exclusive (no one is both a man and a woman). To allow people with penises into the category of woman is to open up the possibility that someone could be both man and woman. There is no corresponding category of non-mother that corresponds to mother in this way.

It would probably be possible to devise a clever test to determine whether gender is a cluster category for a given person, and test a sample of Americans to figure out how widespread it is. In the absence of this, we’ve just got speculation.

That’s as far as I think I can go being purely descriptive. Stay tuned for my prescriptive ideas about how gender should be.

1 Comment

Comments are closed.