Tim's journal

Saying "I don't censor myself. I just say what I think" is popular. I used to say it a lot myself, and I probably still sometimes say something that amounts to that.

My preferred way of saying it now looks more like "no fucks given" -- which is, I think, a little bit more accurate in that it's a statement about my assessment of the risks and benefits of saying something in a particular situation. Which is to do with how much power I have in that situation.

So somebody who says "I never censor myself" is either extremely powerful (and if that person is Donald Trump, he might just be making a completely straightforward statement of truth); is foolish (somewhat more common than the Donald Trump scenario); or isn't being totally honest. (Ironically.)

It's the last case -- the "not totally honest" case -- that I want to look at more carefully. I think a lot of people take pride in their putative lack of self-censorship because they like TV shows like "South Park" or admire some particular comedian. But they're not as funny as the comedians they admire, or even as funny as "South Park" can occasionally be.

More to the point, I think "I don't censor myself" often comes with an implied moral judgment: that there's something dishonest about not saying what you really think, in every possible situation. Tell your friend that his haircut looks nice, when you think he looks like someone put a bowl on his head and cut around it? YOU ARE A TERRIBLE PERSON, because somehow honesty (about something unimportant) gets weighted much higher than the value of maintaining a relationship and making someone else feel nice. Why is that? We know there's no single moral principle that trumps everything -- most decisions are some form of balancing test or another.

Interlude

What does the expression x + y mean in a program? Pick whatever programming language you like (except Lisp, I guess -- sorry) for the purpose of answering; at least, any one where x and y denote variable references (so, not Erlang or Prolog either).

You don't know, right? It depends on what x and y refer to in the lexically (or dynamically, depending what language you picked) enclosing environment when this expression gets evaluated at runtime. If you are a programmer, you understand that context doesn't only affect meaning. It is meaning. Or at least, you understand that when you're reasoning about programs.

Context

So why would I choose to not say exactly what I think in a given situation? If the same person with the haircut was a total stranger, and my job was to do quality assurance for a haircutting place, then probably I would say that his haircut looked bad. So that suggests that context matters.

Not only does context affect the meaning of what you say, context is meaning in and of itself. For example, if I was at a bar with a very close friend and we were 3 drinks in, I might tell a fantastically filthy joke. (I mention "3 drinks in" because shared intoxication is a legible indicator of intimacy in my culture, rather than because drinking makes people behave badly.) I wouldn't tell the same joke at 10:00 AM on a Monday in a meeting at work. Why is this? Am I a hypocrite because I'd tell the joke in one situation but not the other? If the joke is somehow bad if I tell it at work, isn't it also bad if I tell it to my friend?


If you came here via a Dreamwidth friends page, you saw the following content warning: 'some discussion of sexualized presentations, trigger/content warning debates, and racism'.

Well, it depends on the joke, of course, but for the sake of argument let's suppose it's just gross and/or lewd and isn't a joke that depends on putting anybody down. I still wouldn't tell it during the work meeting, and the reason is because telling it in that context has an entirely different meaning than telling it to my friend. Telling it in a meeting, more so than whatever the content of the joke is, is a way to communicate "I have power in this situation, so the usual social norms of propriety don't apply to me. You better keep that in mind." It's a threat.

I like amusing my friends and don't generally like to threaten my co-workers, so that's why I would tell the joke in the bar but not the boardroom. It's not that I somehow have to bite my tongue or hold myself back from threatening my co-workers! It's that I don't want to.

Is this a silly example that wouldn't happen in practice? I don't think it is, actually. I worked at a place where a recently-hired executive made a joke in the office -- apropos of pretty much nothing -- that fell into the popular genre of jokes that basically don't make sense unless you assume all women cheat (and are heterosexual, and tend to be in monogamous partnerships with men). He, like everyone else in the room at that moment (but not like everyone else who worked at the company), was a guy. He also outranked everybody else in the room, as far as the org chart went. I don't think either of these things was a coincidence. On its own, the joke seemed silly and out of place. But I think as a new executive, he must have had this tool in his toolbox of ways to assert power -- to let people know that he (and not necessarily we) were allowed to violate social norms. Had he been a more misogynist person than he seemed to be, he would have waited for some of our female colleagues to be in the room. But because he wasn't a complete fool, he didn't make the joke with the CTO or CEO in the room.

Another real-life example is that of sexualized presentations at technical conferences. The Geek Feminism wiki lists 13 examples of talks during programming conferences that contained pornographic slides or similarly inappropriate sexually objectifying remarks.

Something about the debates about sexualized presentations that I've never seen remarked on is: often, these debates turn into debates about the virtues or lack thereof of erotica or pornography. To which I say: please.

Again, consider two different situations: one in which my (hypothetical) partners and I are at home watching some porn on the couch together. And a second one (not hypothetical) in which a man is giving a talk about how to use CouchDB, an open-source database system, in front of a large audience at a Ruby conference, and uses an image of a woman's nearly-naked buttocks to illustrate his talk (which is subtitled "Perform like a Porn Star").

The debate over the second incident, which I followed at the time, seemed largely focused on whether or not porn (and sometimes, by extension, sex) is good, and whether or not women in the audience were being "prudes" for objecting to it, and whether or not men in the audience were being "white knights" for pointing out it would make some women feel unwelcome.

All of that missed the point completely. I can't know the true reasons why Matt Aimonetti chose to include that slide in his talk (and, I suspect, neither does he), but based on my understanding of politics and sociology, I can make some educated guesses. As a well-respected member of the Ruby community, he felt comfortable transgressing a social norm (though not too comfortable, I'm guessing, since if he wanted to be really transgressive, I could have suggested a number of other images that didn't cater as much to the male gaze). He assumed that most or even all of the audience was going to be men -- what's more, heterosexual men, who he presumed to enjoy a particular kind of erotic or pornographic image. So he felt that the way to simultaneously assert and reinforce his status in the community, and win over the people he wanted to win over, was to break a taboo. But not a big taboo -- a medium-sized one that, I suspect, many of the men in the audience would also have liked to break, but lacked Aimonetti's status.

The debate over Aimonetti's talk (and others liked it) seemed to focus on whether an image of a naked butt was good or bad. The assumption, only inconsistently questioned, was that if you don't believe a naked butt belongs on a slide in a database talk, then you think there's something inherently wrong about naked butts.

But on its own, a naked butt doesn't mean anything. As many of the pro-sexualized-presentation crowd pointed out, we're all human, are all naked at least sometimes, and all have butts. Implicitly, the sexualized-presentation-critical crowd recognized that it was the context for the butt that communicated the harmful -- perhaps even abusive -- message. But, at least when I wrote about this at the time, I didn't spell that out completely.

In debating sexualized presentations, many of the people who support sexualized presentation want us to treat the context as a free variable. They want to treat it as if we don't know anything about what it could be, and argue accordingly. But in fact, in any situation that actually happened, context is at least a partially-bound variable. We have some information about it; it couldn't be just anything. A claim that's true about what would be acceptable for me to do with my partners together in one of our apartments would not also be true about what would be acceptable for me to do in a public talk. If you're familiar with propositional logic or with type theory, you can make up your own analogy here about the difference between existential and universal quantification.

I'm making an analogy with generalization (or lambda-abstraction, if you prefer) not because I think that analogy is going to persuade anybody who doesn't already get it. (I don't.) But rather, because it's clear that the pro-sexualization crowd was, mostly, perfectly capable of understanding logic. They were programmers. They just chose not to apply their logical reasoning abilities to a social situation where being logical wouldn't further what they saw as their own interests. Even programmers who aren't very good at writing code understand that the meaning of x + y depends on context, and yet, when talking about anything that wasn't code, they acted as if context didn't and couldn't determine meaning

Triggers and content warnings

Some people see the issue of trigger warnings or content warnings to be more ambiguous. They may agree that it's wrong to make people feel unwelcome or unsafe by asserting your social power to violate norms. But what about -- they say -- if you're a professor teaching a novel that contains a rape scene? We can agree that the professor (probably) isn't trying to make students feel unwelcome, but only to teach a work of literature that they think is important. [Edited to add in light of comments:] This is a totally separate example, in my mind, from the examples of rape jokes at work or porn images in a tech talk. I don't think there is any kind of warning that could make that content acceptable in a professional setting -- because in those settings, rape jokes or porn images are not there to educate in the first place! They're there to further the domination of other people; there is no "safe" way to do that in a professional environment. I am going to assume that professors, on the other hand, choose curricula because they hope to educate students, and that causing a strong and lasting emotional reaction in some students who have PTSD is truly an unintended side effect. [End of edits.]

"So," I would say, "no instructor should have to self-censor and decline to teach what they think is a good work of literature because of content that may trigger some people's traumatic memories. Rather, the humane thing to do is to provide a warning beforehand so that students can mentally prepare themselves."

But some people seem to think that this, too, is a form of self-censorship, even though it means giving more information; not less. "And anyway," they say, "couldn't anything potentially be a trigger to somebody? If we warned about everything that someone could find triggering, wouldn't we have no time left to do anything?"

I find this disingenuous, again because it's an argument that denies context. In a culture where rape, sexual assault, and physical abuse of children are widespread, it's a reasonable assumption that a group of (say) 30 college-aged students will include some people who have experienced them. Yes, somebody might also have had a traumatic experience involving a yellow bus, but there is no shared context that would alert all of us to traumatic experiences involving yellow buses. That we might sometimes cause accidental harm isn't a good reason to give up on trying to prevent harm.

So I think the argument about triggers is ultimately the same as the argument about sex jokes in the workplace or about sexualized conference presentation: it's an argument predicated on context not existing or at least not mattering.

Who gets context?

I think it's useful to notice, when discussing the virtues or lack thereof of any kind of speech, which speakers are being afforded context, and which ones are being denied it.

To give a more specific example: In the 1990s in the US, it was popular for middle-class white people (of which I was one, at least culturally) to participate in a group ritual involving wringing their hands about gangsta rap, and how violent and misogynistic it was. This probably wouldn't have happened at all if the same music hadn't been very popular among white suburban teenagers, but I digress.

When Ice-T sang "Cop killer, I know your family's grieving (fuck 'em!) / Cop killer, but tonight we get even", it seemed that overwhelmingly, adult white listeners interpreted this line in the absence of any context. I know, since I was one -- well, I was twelve, but I was more interested in being like the grown-ups than in liking what other kids my age liked. At the time, I thought it was obvious that this line literally was an imperative to the audience to go kill cops (or, as I would have said at the time, "police officers"), and that was it.

Did some of those middle-aged white adults in the '90s have fond memories of Johnny Cash singing "I shot a man in Reno, just to watch him die"? I bet some of them did. Because Johnny Cash (being white) was someone they could empathize with, they were willing to afford him more context: first of all, recognizing that the songwriter is not the character in the song. Even that tiny bit of understanding seemed pretty absent from my memories of what white people said about "gangsta rap" in the '90s.

When I read newspaper articles as a young teenager about the deleterious effect of rap music on the youth (that is, I thought at the time, youths who weren't smart enough to read the newspaper, unlike me), I was unaware of the context for songs like "Cop Killer". I didn't know that police systematically targeted Black people, especially (but not only) poor Black people, for brutality. I believed that the adults around me said (or implied) -- that if Black people were targeted more by police, it was because they committed more crimes. So when I read quotes from rap songs in newspaper articles decrying them (usually the only way I ran into them, since I only listened to classical or folk at the time and lord help me if I ever turned on the radio), I digested them the way they were being fed to me: as senseless, context-less violence, rather than as a reasonable emotional response to oppression and powerlessness.

But even at twelve, if I'd heard "Folsom Prison Blues", I don't think I would have concluded Johnny Cash was a guy who advocated shooting a man just to watch him die. I would have understood that Cash was just telling a story. This is because I was a white kid in America and thus had been brought up to be racist. Part of how white supremacy works is the belief that only we are bright enough to tell stories, to use metaphor, to use nuance -- all of which involve context. We limit our own understanding by refusing to see that people we place as below us in the hierarchy might be doing all of those things, too; might be using them to undermine our power.

So if you're skeptical, I encourage you to notice patterns. Who makes excuses for whose speech or action based on context, and when are they believed? It's a warning sign that someone is being denied context when you hear something like "If a man said 'I hate women', people would be enraged, whereas it's okay for a woman to say 'I hate men', and that's unfair!" (Never mind that people do get enraged at women for saying "I hate men".) Again, if we are programmers (or just people who do arithmetic), we know that x - y doesn't have to be the same as y - x unless y and x are the same. In this statement, women and men are not the same, because they occupy different social positions (moreover, different women occupy different social positions based on their intersecting oppressions, and the same for men).

Context is not just a flimsy excuse for doing or saying something bad or wrong (when you don't want to defend the person doing it), or a nervous justification for something you're uncomfortable with (when you do want to defend the person doing it). It's a part of the meaning of everything that we as language-users say. Even when we are writing equations that rigorously follow from a set of explicit axioms based on logical rules, we still share the context of what our chosen notation means. In ordinary life, whether you are talking to your friend in the bar, to your partner at home, to your co-workers in a meeting, or to an audience at a professional conference, you are using your knowledge of the context you're in to communicate meaning, consciously or not. So that also means that you are accountable for the meaning you convey.

Why do I censor myself?

So why do I censor myself? Not to spare others' feelings, though sometimes that enters into it. Not to be polite, though that may enter into it too. More broadly, I censor myself because I often believe I know the entire meaning of everything that comes into my head (though often, as when I was 12 and reading white editorial writers' critiques of rap music, I didn't), but I don't expect everyone else to share all of the same context. Moreover, I'm aware of context that is shared that would cause the listener to extract a different meaning than the meaning I intend inside my head -- as with the dirty joke example. And when I censor myself, the meaning isn't a meaning I actually want to convey. I want to share jokes I think are funny; I don't want to present myself as a dominant male; unfortunately, in some situations, I can't do both. If I told my hypothetical friend the truth about his terrible haircut, his understanding (unless we had a very specific kind of relationship) would be not that I was sharing objective and unfiltered truth, but that because I was violating a social norm about politeness over trivial matters between friends, I didn't like him. That wouldn't be what I want to communicate.

Sometimes, the easiest way to say what you mean is to shut the heck up.