Survey says: gays & lesbians preferred over homosexuals in military | Hacker News Reader

Survey says: gays & lesbians preferred over homosexuals in military

languagelog.ldc.upenn.edu

33 pointsanateus21 comments

> The largest observed difference in support for spending was between the traditional category "welfare" and the two variant forms "assistance for the poor" and "caring for the poor."

It's the same reason unpleasant jobs have the name of the job changed every 10 to 20 years. Janitorial services, custodial services, sanitation technian, etc... in the beginning, it avoids the stigma of what it is - cleaning up dirt/filth/etc. Over time, people catch up, and you have to change the name of it again. Arguably "Secretary" to "Executive Assistant" is the same thing.

Welfare works similarly - when it's rebranded, it does well for a while, then sinks back down into unpopularity. When someone says "social safety net" these days it tends to evoke a more favorable response, but they usually don't mean anything significantly different from general welfare, unemployment, and other state-provided assistance. But a social safety net doesn't have the same associations with welfare, so it's more favorable right now. If social safety net acquires the same connotations as welfare, the name will be changed again in 10-20 years.

They should have also tested "gay men" and "lesbians" separately; I have a feeling that people were only more positive to the "gay men and lesbians" aggregate category because people haven't actually internalized the idea that women serve in the military—so when you say "homosexuals", the image triggered is purely of gay men, and that gets neither the feminist nor lewd-fantasy vote.

Those interested in the (extensive) science of framing biases in human reasoning will want to consult Kahneman and Tversky's classic "Choices, Values, and Frames": http://amzn.com/0521627494

Yes, it's dry and scholarly and will require that you relearn basic statistics. You want to read it anyway.

Thanks! I commented on the article mentioning that since Mark Liberman was asking for empirical studies.

Read "Judgment Under Uncertainty" first.

  "A cynical commenter at TPM suggested the control experiment
  of asking people if they favor "heterosexuals" serving
  openly in the military."

I think that would actually be a very enlightening experiment.

Reminds me of an old story of a politician "slandering" his opponent in the campaign by telling the mostly poorly educated locals things like "Do you know what that man did at college? He matriculated!" in a tone that suggests that's something sleazy. There was no means to rebut it. It was true and if you then gave a vocabulary lesson to your voters, you would just be calling them stupid.

Citation if available?

I don't have a citation. My understanding is that it is a true story. If you want to look for it, you might check for mudslinging political campaigns in the Deep South of the US prior to, say, 1950 (possibly even 1800's).

EDIT: Found a reference to this:

Reynolds told the backcountry crowds that his opponent had once sunk so low as go up to Harvard (pronounced HAW-vud). What did the man do there? Why, he "matriculated"! And, worse, he became "a thespian"! Imagine.

Naturally, Reynolds won the race.

http://archives.cnn.com/2000/ALLPOLITICS/stories/02/21/morro...

Arigatou.

More proof that humans are fucking crazy.

Why is this downvoted? This is exactly what the experimental results coming out of the field of heuristics and biases say.

Because it's not particularly insightful? Also, humans aren't "fucking crazy" so much as they make near-instant judgments sometimes in flawed ways, with associations, context, language, and thought getting jumbled up and being hijackable sometimes.

(That said, wow, it's so cool to be responding to a comment by Eliezer Yudkowsky, you're like a hero to me and brilliant and thanks for all your work on Bayes, rationality, and thinking, and tsuyoki naritai, and that three worlds collide story, and all the rest - seriously, huge admirer of yours here)

"... Because it's not particularly insightful? Also, humans aren't "fucking crazy" so much as they make near-instant judgments sometimes in flawed ways, with associations, context, language, and thought getting jumbled up and being hijackable sometimes. ..."

"fucking crazy" approximately equates to "make near-instant judgments sometimes in flawed ways, with associations, context, language, and thought getting jumbled up and being hijackable sometimes" and it's shorter. As a rule I don't downvote these kinds of remarks, just ignore them.

> Also, humans aren't "fucking crazy" so much as they make near-instant judgments sometimes in flawed ways, with associations, context, language, and thought getting jumbled up and being hijackable sometimes.

With respect, and with gratitude for your compliments, I ask: Is there really that much of a difference?

> With respect, and with gratitude for your compliments, I ask: Is there really that much of a difference?

I think so. "Fucking crazy" implies a madness/insanity/etc that can't be overcome. It's fine if you're trying to make a point and need quick words ("Wow, humans can be really fucking crazy sometimes" in response to a story), but pointing out how humans are crazy also shines a light on potential solutions.

Saying, "People are prone to making associations with certain words instead of concepts" is helpful - you can think, "Okay, then how can I use descriptive words that aren't loaded with unnecessary baggage?" Or if you're trying to persuade, you can look for terms that are favorable. Or you can point out the error in word association. Or you can try to "humanize" the concept - that's my guess at what happened in this particular story. "Homosexuals" implies someone else - a third party, an enemy, "not one of us". "Gay men and lesbians" implies people, humans.

Just writing off humanity with, "People are fucking crazy" makes it easier for people to shrug their shoulders and say, "Well, that's how it is." Instead, I think it's something to be overcome. The more we point out exactly what our errors tend to be, the more we can work specifically on them.

Fair enough, but I hope you won't overlook the points that:

1) People are, in fact, fucking crazy.

2) People often have difficulty in coming to terms with the above fact in its simple form, to the detriment of their development as rationalists; it's why I use "People are crazy, the world is mad" as a slogan. They pass up opportunities like a valid startup idea or signing up for cryonics because they think that other people will laugh, and they haven't gotten on a deep gut level that the people laughing at them are crazy. Vide: http://wiki.lesswrong.com/wiki/Conformity_bias

This is why I objected to the downvoting of the original comment (I'm pleased to see it's been voted up now). The referenced poll is, in fact, more proof that humans are fucking crazy, and people need proof of that.

Coding monkey, programmer, software engineer. What's the difference?

PHP, Python and J2EE respectively ;-)

"... PHP, Python and J2EE ..."

Hack and get the job done by yourself, hack get the job done and others can help, engineer and get the job done with lots of overheads.

Well, if using Python over J2EE is wrong, I don't want to be right.

Reply on news.ycombinator.com