Archive for April, 2009

The Failure of #amazonfail

April 15, 2009

In 1987, a teenage girl in suburban New York was discovered dazed and wrapped in a garbage bag, smeared with feces, with racial epithets scrawled on her torso. She had been attacked by half a dozen white men, then left in that state on the grounds of an apartment building. As the court case against her accused assailants proceeded, it became clear that she’d actually faked the attack, in order not to be punished for running away from home. Though the event initially triggered enormous moral outrage, evidence that it didn’t actually happen didn’t quell that outrage. Moral judgment is harder to reverse than other, less emotional forms; when an event precipitates the cleansing anger of righteousness, admitting you were mistaken feels dirty. As a result, there can be an enormous premium put on finding rationales for continuing to feel aggrieved, should the initial rationale disappear. Call it ‘conservation of outrage.’

A lot of us behaved like that this week, in our fury at Amazon. After an enormous number of books relating to lesbian, gay, bi-sexual, and transgendered (LGBT) themes lost their Amazon sales rank, and therefore their visibility in certain Amazon list and search functions, we participated in a public campaign, largely coordinated via the Twitter keyword #amazonfail (a form of labeling called a hashtag) because of a perceived injustice at the hands of that company, an injustice that didn’t actually occur.

Though the #amazonfail event is important for several reasons, I can’t write about it dispassionately, because I was an enthusiastic participant in its use on Sunday. I was wrong, because I believed things that weren’t true. As bad as that was, though, far worse is the retrofitting of alternate rationales to continue to view Amazon with suspicion, rationales that would not have provoked the outrage we felt had they been all we were asked to react to in the first place.

When trying to explain one’s actions, hindsight is always 20/400. With that caveat, I will say that the emotional pleasure of using the #amazonfail hashtag was intoxicating. There is no civil rights struggle in the US that matters more to me than the extension of equal rights without regard for sexual orientation. Here was a chance to strike a public blow for that cause, and I didn’t even have to write a check or get up from my chair to do it! I went so far as to publicly suggest a link between the Amazon de-listing and the anti-gay backlash following the legalization of gay marriage in Iowa and Vermont. My friend Nelson Minar called bullshit on my completely worthless speculation, which was the beginning of my realizing how much I’d been seduced by righteousness, and how stupid it had made me.

I was easily seduced in part because the actual, undisputed event — the change in status of LGBT-themed work on Amazon, while heterosexual material and anti-gay tracts kept their metadata intact — fit a template I know well, that of the factional use of a system open to public access. Examples are legion; one recent one was the top positions enjoyed by issues related to the legalization of marijuana on the site. (Though I am in favor of the legalization of marijuana, I also recognize that the results were an outcome no representative poll of the American people would have returned.) Seeing the change in status of LGBT books, I believed, vaguely, that Amazon was hosting and therefore complicit in a systemic attempt to remove such material from public discussion.

Here’s how stupid that belief made me. I have been thinking about the internet as hard as I can for the better part of two decades, and for the latter half of that time, I’ve been thinking about the problems of categorization systems, and it never occurred to me that the possible explanation for systemic bias might be something having to do with a technological system instead of a human one, that a changed classification in the Amazon database could trigger the change in status of tens of thousands of books.

I assumed (again, vaguely) that Amazon themselves had not adopted an anti-gay posture, and I recognized the possibility that this might be a trolling attack, but the idea that this was an event of mainly technological propagation, rather than a coordinated bit of anti-gay bias, simply escaped me. This isn’t because I am a generally stupid person; it was because I was, on Sunday, a specifically stupid person. When a lifetime of intellectual labor and study came up against a moment of emotional engagement, emotion won, in a rout.

Many people I love and respect disagree with me on this point; Mary Hodder in particular has written a very thoughtful case for why we should still regard Amazon as culpable and as a target for outrage. I don’t disagree with her interpretations of what Amazon did wrong (and I am using her as a particularly eloquent spokeswoman for a whole class of post-#amazonfail arguments) but I do disagree with her conclusion.

If we wanted to deny Amazon all benefit of the doubt, and to construct the maximum case against them, it would go something like this: it was stupid to have a categorization system that would allow LGBT-themed books to be de-ranked en masse; it was stupid to have a technological system that would allow that to happen easily and globally; it was stupid to remove sales rank from sexually explicit works, rather than adding “Safe Search” options; it was stupid to speak in PR-ese to the public about something that really matters; it was stupid to take as long as they did to dribble an explanation out.

Stupid stupid stupid stupid, yes, all true. If it had been a critique of those stupidities that circulated over the weekend, without the intentional mass de-listing, it would have kicked off a long, thoughtful conversation about metadata, system design, and public relations. Those are good conversations to have, we need to have them, but they are not conversations that would enrage thousands of people in the space of a few hours and kick off calls for boycotts and worse.

Intention is what we were reacting to, and the perception of intention matters, a lot. If you hit me with your car and kill me, the effect on you could be anything from grief counseling to being convicted of murder, and that range of outcomes would rest on a judgment about your intentions, even given the same actual event.

So it is here. Whatever stupidities Amazon is guilty of, none of them are hanging offenses. The problems they have with labeling and handling contested categories is a problem with all categorization systems since the world began. Metadata is worldview; sorting is a political act. Amazon would love to avoid those problems if they could – who needs the tsouris? — but they can’t. No one gets cataloging “right” in any perfect sense, and no algorithm returns the “correct” results. We know that, because we see it every day, in every large-scale system we use. No set of labels or algorithms solves anything once and for all; any working system for showing data to the user is a bag of optimizations and tradeoffs that are a lot worse than some Platonic ideal, but a lot better than nothing.

We know all that, but we’re no longer willing to cut Amazon any slack, because we don’t trust them, and we don’t trust them because we feel like they did something bad, even though we now know, intellectually, that they didn’t actually do the bad thing we’ve come to hate them for. They didn’t intend to silence gay-themed work, and they didn’t provide the means for groups of anti-gay bigots to do so either. Even if the employee currently blamed for the change in the database turned out to be a virulent homophobe, the problem is in not having checks and balances for making changes to the database, not widespread bias.

We’re used to the future turning out differently than we expected; it happens all the time. When the past turns out differently, though, it can get really upsetting, and because people don’t like that kind of upset, we’re at risk of finding new reasons to believe false things, rather than revising our sense of what actually happened.

We shouldn’t let that happen here; conservation of outrage is the wrong answer. We can apologize to Amazon while not losing sight of the fact that homophobic bias is wrong and we have to fight it everywhere it exists. What we can’t do, can’t afford to do if we want to think of ourselves as people who care about injustice, is to fight it in places it doesn’t exist.