Bryan Caplan  

Against Argumentative Definitions: The Case of Feminism

Italian elections update... There's no success like failur...
Suppose I define socialism as, "a system of totalitarian control over the economy, leading inevitably to mass poverty and death."  As a detractor of socialism, this is superficially tempting.  But it's sheer folly, for two distinct reasons.

First, this plainly isn't what most socialists mean by "socialism."  When socialists call for socialism, they're rarely requesting totalitarianism, poverty, and death.  And when non-socialists listen to socialists, that's rarely what they hear, either. 

Second, if you buy this definition, there's no point studying actual socialist regimes to see if they in fact are "totalitarian" or "inevitably lead to mass poverty and death."  Mere words tell you what you need to know.

What's the problem?  The problem is that I've provided an argumentative definition of socialism.  Instead of rigorously distinguishing between what we're talking about and what we're saying about it, an argumentative definition deliberately interweaves the two. 

The hidden hope, presumably, is that if we control the way people use words, we'll also control what people think about the world.  And it is plainly possible to trick the naive using these semantic tactics.  But the epistemic cost is high: You preemptively end conversation with anyone who substantively disagrees with you - and cloud your own thinking in the process.  It's far better to neutrally define socialism as, say, "Government ownership of most of the means of production," or maybe, "The view that each nation's wealth is justly owned collectively by its citizens."  You can quibble with these definitions, but people can accept either definition regardless of their position on socialism itself.

Modern discussions are riddled with argumentative definitions, but the most prominent instance, lately, is feminism.  Google "feminism," and what do you get?  The top hit: "the advocacy of women's rights on the basis of the equality of the sexes."  I've heard many variants on this: "the theory that men and women should be treated equally," or even "the radical notion that women are people."

What's argumentative about these definitions?  Well, in this 2016 Washington Post/Kaiser Family Foundation survey, 40% of women and 67% of men did not consider themselves "feminists."  But over 90% of both genders agreed that "men and women should be social, political, and economic equals."  If Google's definition of feminism conformed to standard English usage, these patterns would make very little sense.  Imagine a world where 90% of men say they're "bachelors," but only 40% say they're "unmarried."

What would a non-argumentative definition of feminism look like?  Ideally, feminists, non-feminists, and anti-feminists could all endorse it.  If that's asking too much, all these groups should at least be able to accept the proposed definition as a rough approximation of the position they affirm or deny.  My preferred candidate:
feminism: the view that society generally treats men more fairly than women
What's good about my definition? 

, the definition doesn't include everyone who thinks that our society treats women unfairly to some degree.  In the real world, of course, every member of every group experiences unfairness on occasion.

a large majority of self-identified feminists hold the view I ascribe to them.  Indeed, if someone said, "I'm a feminist, but I think society generally treats women more fairly than men," most listeners would simply be confused.

Third, a large majority of self-identified non-feminists disbelieve the view I ascribe to feminists.  If you think, "Society treats both genders equally well," or "Society treats women more fairly than men," you're highly unlikely to see yourself as a feminist.

At this point, you could declare, "Given all the #MeToo revelations, it's obvious that society does treat men more fairly than women."  Or, "Men are vastly more likely to be violently killed than women, so it's obvious that society treats women more fairly than men."  Similarly, you could declare, "Since women earn x% less than men, society treats men more fairly than women" or "Since men are jailed nine times more often than women, society treats women more fairly than men."  (In both cases, naturally, someone else could respond, "After basic statistical corrections, these gaps go away.") 

And you know what?  Despite their overconfidence and impatience, all of these statements are on point.  They're real arguments, not semantic trickery.  If you calmly collect and carefully quantify a few hundred such arguments, you won't just know whether feminism is true.  You'll know how close the other side is to being right.

Comments and Sharing

COMMENTS (20 to date)
Greg writes:

I read your Against Argumentative Definitions article, and while I agree that argumentative definitions are a barrier on the way to the truth, I don't think your feminism example is as clear cut as presented.

My objection is that I think "the theory that men and women should be treated equally", "the advocacy of women's rights on the basis of the equality of the sexes", and "the view that society generally treats men more fairly than women" are all substantially the same argument. The point is to define what we're talking about, and what we're talking about is whether society treats men and women equally. They just differ as to where on the fair/non-fair continuum to set the feminist/non-feminist fulcrum.

The ideological approach would be to set it at equal fairness or some other ideologically determined point. Your approach is to make the definition match the perception as evidenced by the survey results i.e. make the definition feminists match the survey feminists.

I find this problematic. First, definitions should probably not be a popularity contest because popularity changes. So what was feminist in the year 2000 is now not feminist because the fulcrum has shifted in time. Or in space! What is feminist in Malaysia is not in Canada because of different opinions, or different realities for men and women.

There are benefits to having fixed definitions.

Ben Kennedy writes:

You'll just end up wrangling over the definition of "fairness". I don't think self-described feminists are somehow unaware that controlling for 12 different things shrinks the pay gap. They just don't care. Unless your arguments somehow bridge the is-ought gap (good luck!) a different definition doesn't really resolve the conflict.

Lawrence D'Anna writes:

The one nice thing about argumentative definitions (or "semantic infiltration" or whatever you want to call it) is that it's a really reliable tell that a person isn't worth talking to or reading. If they use an argumentative definition, especially if they insist on it after you point it out, then they're either too stupid to spot it, or they view debate as subordinate to their political aims and only of instrumental value.

Either way, they don't deserve to be in your bubble. Especially if they agree with you on whatever the underlying issue is. If they disagree with you then at least they might know something you don't. If they agree with you they're just propagandists for your priors.

Steve writes:

Not sure a new definition is needed.

The real distinction is between the concept and the political movement. One definition doesn't necessarily need to encompass both.

Greg G writes:

Nice post Bryan.

No one can possibly make the best argument for something without knowing how to also make the best argument against it.

Thomas writes:

Why "more fairly"?

I think feminists believe that society generally treats men better than women, rather than more fairly. After all, women could be treated less fairly and also better than men. In fact, I think quite a few men actually believe that is the case in important areas such as criminal law, family law, flexibility regarding social roles, military conscription, importance of victims of violence, credibility, etc. But, I don't believe any feminists think women are treated better than men because of unfairness -
they invariably believe women are generally treated worse than men, and thus that men are treated better than women.

So why not just say that?

Adam writes:

Interesting, but tend to agree with Greg. "Isms" are evaluative, not descriptive. Isms desire a movement toward a desired ideal from an inadequate present.

Words ending in "ism" are evaluative and make evaluative claims. They are not descriptive. Here's more:

Anon writes:

My guess is that the definition of feminism has evolved. Your definition would be fine in the past, but right now I think it feminism is the view that power dynamics define just about every interaction between genders, and women are virtually always oppressed. Because of this oppressed status, they understand gender relations better than men.

Dmitry writes:

Thomas is exactly right. You are using "fairly" when you should be using "better."

Mark writes:

Part of the problem here is that ideologies typically have both analytical and normative positions that often get confused, and they also have their own idiosyncratic internal terminologies.

Both feminists and anti-feminists generally agree that men and women should be treated equally (that is, they share the same central normative position). They disagree, however, on there analytical assessment of the world: feminists think 'society' is stacked against women; anti-feminists believe it is either generally fair, with respect to gender, or stacked in women's favor; and/or they have disagree about the definition of equality (for feminists, equality usually means population-level variables like average income are equal for men and women; for antifeminists it typically means mere equal treatment before the law).

So, defining movements or ideologies ("black lives matters", "family values", etc.) in terms of vague, often nearly universally applauded oral sentiments is really meaningless; it's done for propaganda reasons, not to to make a clear statement of a position.

Complicating things even more, anyone can define any 'ism' any way they please. Most feminists may agree with Bryan's definition 'society treats men better than women,' but some feminists may argue that feminism merely means society *should* treat men and women equally, and may even believe that it does so today, and that their feminism is therefore generally merely support of the status quo (comparable to a 'liberal democrat' who supports the current liberal democracy against potential alternatives).

Since 'feminism' and 'socialism' and other political ideologies have no essential definitions, we can only define them however the preponderance of people identifying with the term define it. But since the definition itself, as employed by most of the 'members', typically assumes a host of disputed premises. Those are what actually matter. Most ideologies are in fact not simple enough to be encapsulated by a single sentence, so there's really no point in arguing over (or using) the definitions or the labels themselves. We should, therefore, approach these words in a 'non-essentialist' way, and regard them as useful only as good predictors of what people who identify with them believe about particular political or social questions.

If someone says they're a feminist, it's useful because it tells you they probably think discrimination against women is widespread; they probably support laws to redress said putative discrimination; they probably support legal abortion. It does not tell you whether they support "equality", because 1) if you're not a feminist, you likely define equality different from how they define it, and 2) there's considerable disagreement and confusion (whether acknowledged or not) among them over how they themselves define the word. The same can more or less be said of libertarians and 'liberty', democrats (small 'd') and 'democracy', and so on.

Weir writes:

Kate Millet, Cordelia Fine and Judith Butler are feminists, but so is Janet Radcliffe Richards.

There's Andrea Dworkin, Catharine MacKinnon and Mary Daly, but there's also Laura Kipnis, Ellen Willis and Catherine Deneuve.

There's Cathy Newman, but there's also Cathy Young. There's Margaret Atwood, Lionel Shriver, and Angela Carter. There's Lara Prendergast, Emily Yoffe, and Wendy Kaminer. For every Emily Lindin, there's a Caitlin Flanagan.

Feminism is a massive tent. There are the fainting-couch feminists, but there are feminists who treat women as grown-ups. There are the Twitter feminists, but there are also feminists who are interested in the outside world, beyond Manhattan and Brooklyn.

There's Ayaan Hirsi-Ali, Shirin Ebadi, and Wafa Sultan. There's Martha Nussbaum. She knows that Indian society is not the same as Norwegian society. That's the place to start, with the definition of "society" that we're working with. If "society" just means the Hudson River, then calling yourself a "feminist" doesn't amount to much.

HH writes:

Everyone who wants to continue with a really interesting expansion of these ideas, try searching Slate Star Codex for "motte and bailey." Very insightful, and in line with Bryan's thinking.

Miguel Madeira writes:

«My objection is that I think "the theory that men and women should be treated equally", "the advocacy of women's rights on the basis of the equality of the sexes", and "the view that society generally treats men more fairly than women" are all substantially the same argument. »

No they are not - the first and the second are purely normative position ("men and wome should be treated equally"), while the third is a combination of normative and positive positons ("men and wome should be treated equally" but "in present society, men are treted better than women")

John Thacker writes:
I don't think self-described feminists are somehow unaware that controlling for 12 different things shrinks the pay gap. They just don't care.

But that's a worthwhile thing to argue about, indeed *the* thing to argue about. You want a shorthand for the position "the current situation is unfair" and the counter position "the current situation is fair" (even if not equal, as fair and equal are not the same), such that everyone can agree on the definitions. Then the proper thing to debate is whether the current situation is fair or not, and why. It at least reduces the silly argument about who is a feminist or not, or attempts to win the arguments through chicanery about definitions.

Disagree with Thomas and Dmitry. You definitely will find self-described feminists (even on the left) who argue that in certain situations the government is unfair and treats women better and that it should stop. They may not be a majority, but certainly the more hardcore feminists think that women should be as eligible for Selective Service as men.

Christophe Biocca writes:

You can find even more eggregious cases, where it's not the definition, but the name itself, that is argumentative.

"Foreign policy realism" is probably the best example. It has a specific model of the world (self-interested states as the relevant actors, anarchic world stage in which they interact), which may or may not be a good fit, but the very name makes sentences such as "Foreign policy realism poorly models the real world" kind of nonsensical.

Greg writes:

@Miguel Madeira

I agree they're different in that he third makes an assertion about the current state, but I maintain they're the same in terms of Bryan's central thesis: none of them are argumentative and none get in the way of discussing the declarations he provides later.

Thomas writes:


I don't think feminists are more focused on equality of population-level metrics. They are not focused on life expectancy equality, death-on-the-job equality, imprisonment equality, violent victimhood equality, etc.

Often, when related topics come up, feminists focus on areas where women are differentially affected (thus, domestic violence, where women are about 50% more likely than men to be victims, rather than all violent crime, where men are about nine times as likely to be victims) or, rather than compare to men, use some ideal outcome as the norm to compare to.

In this sense, some feminists are not about gender equality at all, but rather are about advancing the welfare of women, without regard to men (nothing wrong with that, aside from being rather sexist).

NS writes:

Why use the word "society" when people usually just mean "men." By feminisms own canon, women can't treat anyone unfairly because they don't have the power. So just tell the truth. Feminism is a claim that men treat women unfairly and feminists don't want that to continue. But to say "society" treats anyone in any way is to reify something beyond its proper realm.

Edward writes:

Great point!

I'm very curious to know how "basic statistical corrections" could eliminate the sexual abuse #metoo gap that women face. Someone please educate me.

Andy writes:

I don't like the active voice here. I don't really believe that "society" really does anything. How about "women experience more unfairness than men".


Return to top