Log in

Previous Entry | Next Entry

On Heuristic Tradeoffs

My friend Jenna is quitting smoking, on which I congratulate her unreservedly. Nicotine is a devilishly hard drug to kick; I managed, after countless false starts, to quit smoking, but I will probably leach nicotine into the soil long after I am dead. Note that I said I quit smoking but not nicotine; quitting smoking was, in the end, a matter of finding a nicotine administration route that was less of a hassle than going outside, lighting a small paper tube on fire, and inhaling the smoke from it. Patches weren't it. (I get rashes.) Gum wasn't it either. (The release rate is all wrong.) No, I am one of those e-cigarette dweebs, microdosing flavoured nicotine like a pain patient on a PCA, all day erryday. I admire everyone who manages to disentangle the routines of addiction from their other everyday routines, because I'm probably never going to get there myself: I like my drug and it's a stupid hobby. Yes, I suppose I'd be a better person if I could learn to rise above most of life's minor inconveniences through my own poise and willpower, but let's face it, nicotine is faster.

I bring this up to establish a standard of charity that I'm going to try to keep to in what follows.

I was catching up recently with an old friend going through something of an identity crisis. He's always thought of himself as a feminist: "the radical notion that women are people," check, all one happy motte here. Recently, though, he dated a third-wave feminist, and ran into a lot of cognitive dissonance, in that they were basically unable to talk about the kinds of problems Ozy Frantz brings up in this post. Ozy's opening paragraph is worth quoting in full, the sort of thing that ought to be cross-stitched into samplers and framed in bathrooms as a reminder that the map is not the territory:
Memo To The Social Justice Community At Large: the privilege/intersectionality model of how oppression works? Is a model. It’s an oversimplification that people use because the actual reality of how oppression works is way too complicated to talk about. It is not the Ultimate Truth Of How Oppression Works Forever and Ever.

There are lots of reasons why a model might not adequately explain real-world outcomes in its domain, but most of them boil down to "you took a shortcut somewhere." This usually happens because it's necessary. In machine pattern recognition, we grade the performance of systems on two metrics: precision, or "of all the times the system predicted outcome FOO, how often was it right?", and recall, or "of all the times the system should have predicted outcome FOO, how often did it?" Tuning a system that you plan to use for decision-making on real data often involves making tradeoffs between precision, recall, model size, and engineering constraints. For example, most of the bulk of any voice recognition application on your phone or tablet will be a few megabytes of precomputed matrices that the app uses to transliterate your speech sounds from your microphone into words on your screen -- that is, when your device can't reach the cloud servers where much more nuanced and accurate models that can run to the tens of gigabytes or more sit. Such is life in a world where device storage is finite. The hybrid system uses the remote, more accurate method when it can, and the local, does-the-best-it-can method when it has to.

It's hazardous to get too quantitative when comparing silicon to brains (the architecture's just so different), but I think the qualitative analogy -- that our mental models necessarily incorporate heuristic shortcuts -- should be pretty uncontroversial. It's which shortcuts people choose to take (or refuse to consider not taking), I think, that creates the sort of cognitive dissonance which my friend experienced on a micro level, and which on a macro level creates and perpetuates outgroups.

My brain is pretty meta sometimes, and occasionally goes about installing heuristics without warning me about it, which in combination is probably how I ended up with a heuristic aversion to people who cling to sloppy heuristics. (I am also kind of a jerk, though, don't forget.) This is going to be an interesting one to deal with, because look, sometimes sloppy heuristics are all somebody can afford. And, unfortunately, we're not silicon-based pattern-matching systems, so replacing a cheap-but-sloppy heuristic with a ballpark-as-cheap-but-less-sloppy one isn't as simple as a matter of pushing an app update. Just like my stress, the decision problems never go away; just like nicotine, the sloppy heuristic is a quick and cheap solution with long-tail side effects that we cheerily hyperbolically discount away, because the opportunity cost of switching to the better solution is higher than we're willing to pay up front.

Thus do we trap ourselves in coordination problems of our own making. So it goes.


( 22 comments — Leave a comment )
Apr. 23rd, 2015 12:08 am (UTC)
¿Third-wave feminist?
A definition of this term for readers not familiar with the jargon of this particular discourse would be helpful.
Apr. 23rd, 2015 12:20 am (UTC)
Re: ¿Third-wave feminist?
Wikipedia's right over there, dude.
Apr. 23rd, 2015 01:51 am (UTC)
Re: ¿Third-wave feminist?
Ummm, yes. Looked it up, and ...

"This article has multiple issues. Please help improve it or discuss these issues on the talk page.
The examples and perspective in this article deal primarily with the United States and do not represent a worldwide view of the subject. (February 2013)
This article's lead section may not adequately summarize key points of its contents. (August 2013)
The neutrality of this article is disputed. (March 2014)
This article lends undue weight to certain ideas, incidents, or controversies. (May 2014)"
John Edwards Cummings
Apr. 24th, 2015 09:07 am (UTC)
Re: ¿Third-wave feminist?
The dude actually raises an important question.

The line between second and third wave is pretty damn vague and weird (trick question: which wave is Camille Paglia partial to and why? )

This is actually pretty important to discourse, because when you say that a friend dated a "third wave feminist", that's supposed to convey some kind of useful information about the "third wave feminist".

Given current state of definitions available, at most I may conclude that your friend dated someone who is somewhat more likely to be supportive of individual autonomy, somewhat more likely to be sex-positive and sex-work positive, significantly less likely to harbor ill-justified animosity towards transsexuals, somewhat less likely to be ignorant of queer issues, somewhat more likely to be friendlier towards homosexual people, and somewhat more likely to value science-based, empirically grounded arguments (and thus eschew ideologically driven, empirically failed programs).

In a few words, a "third-wave feminist" is more likely to be like feminists at http://feministire.com/ and less likely to be like feminists at feministcurrent.

But none of those are definitive, well-established traits, and frankly, I know plenty of self-described "third wavers" who do not fit some (or even most) facets of the above ad-hoc "profile".

Oh, and not to mention that there are classifications that appear to run in "parallel" to the wave construct (can one be "radical feminist" and "third wave feminist" at the same time? Why?)

So while the phrase "he dated a third-wave feminist" does tell me a little bit about what the person your friend met might be like (and the fact that the relationship developed feminism-related "issues" does tell me a little bit about what your friend might be like), it tells me very little, and in very uncertain terms.

And that is a problem.

When someone tells me they are Marxist, I can be fairly certain what kind of views the interlocutors espouse (and act accordingly).

When someone tells me they're "X-wave" feminists, most I can have is a rather vague guess (which pretty much ensures that there will be surprises, and not all of them pleasant)
John Edwards Cummings
Apr. 24th, 2015 09:17 am (UTC)
Re: ¿Third-wave feminist?
Oi, looks like my previous one got filtred :(

I can assure you that my comments are wordy, but not, in fact, suspicious
Apr. 24th, 2015 09:45 am (UTC)
Re: ¿Third-wave feminist?
LJ's spam filter is a bit brain-damaged, I think. Should be fixed.

You raise a valid point which I can't really address in the instant example, because I don't have enough familiarity with the beliefs of the person in question. My friend and I talked mostly about his ex's tendency to disregard empirical evidence that didn't fit her model, which is largely what prompted this post. I agree that the various subgenres aren't well-typed, and what one person means when they call themselves a third-wave feminist and what another person means when they say the same thing are often entirely unrelated.

OTOH, "what kind of feminist is she?" is pretty incidental to the phenomenon I'm writing about, because it's hardly confined to any particular current of feminism, or indeed feminism at all. Bounded cognition is a human universal.

My curt response to the previous commenter has some basis in historical interaction, which I'm only going to allude to because I'm not interested in embarrassing him.
Apr. 25th, 2015 12:21 pm (UTC)
Re: ¿Third-wave feminist?
Sloppy definitions beget sloppy heuristics based on implicit assumptions. E>I.
John Edwards Cummings
Apr. 30th, 2015 10:18 am (UTC)
Re: ¿Third-wave feminist?

My experience has been approximately that way with people whom I tentatively consider *second*(-ish) wave feminists and many radical ones, that is, said people systematically disregarded empirical evidence or downplayed the need for well-structured, evidence-based arguments (one radical feminist I happened to interact with even claimed that empirical science should be eschewed in feminist discourse because it is "a patriarchal construct and antithetical to critical theory". The scarequoted part is a verbatim quote).

I wonder if there is some kind of symmetric distortion of outgroup perception going on here (basically, "second wave" people thinking that it's the third wave who disregard good science, while "third wave" people think that it's the "second wave" folks who disregard good science).

While I do agree on bounded cognition in general, I think that terminological issues and intra-group classification issues in feminism are especially pronounced, far in excess of what one can find in other venues of human thought.

That might be of course just toxic influence of critical theory (which is itself, at this point, a rather poorly organized and empirically impoverished field), or a mere coincidence - the reason behind the phenomenon is a discussion unto itself, but I think it's a pretty apparent trait of feminist discourse.

And it is, at the very least, counterproductive, and will prove outright disruptive if wave-count (and/or number of "parallel" feminism-classifications) keeps rising.

I mean, I kind of want to identify as a feminist (I do consider gender equality to be more efficient - and better justified by existing evidence - than the "traditional" layout of gender roles and gender-driven "power" imbalances) but I do not want to be associated with certain specific currents in feminism (for instance, anything sex-negative and/or "nordic" model supporting is an instant "no go") and certain specific thinkers (For instance, Farley) or support sweeping claims about gender essentialism (I think the data on biological predisposition to specific "gendered" behaviors is very poor and contaminated with dubious, unreplicated studies and outright "salmon necromancy MRI"-type research, and thus I prefer to withhold any judgement until better data arrives).

It's pretty damn clumsy and hard to convey this particular set of beliefs and attitudes in an efficient manner.

I do think my beliefs are closer to liberal third-wave individualistic feminism than to anything else (or are they ?) but that's still a goddamn mouthful.
May. 11th, 2015 06:31 am (UTC)
Re: ¿Third-wave feminist?
Apologies for the slow reply; I am traveling and visiting relatives and juggling that with dayjob. Hooray for weekends. (LJ used to be a good place for extended, thoughtful discourse once, though; let's see whether federation helps that recover.)

FWIW, I tend to think they both disregard good science, but agreed, it would be quite interesting to study the mechanisms of in-group/out-group formation (particularly on the left, where it seems to happen awfully goddamn fast these days) in general and this might well be an angle.

My position is roughly as follows: for the internet-connected world, at least, it is impossible not to encounter outliers along just about any biologically based axis you care to name. This deprecates out the traditional layout of gender roles &c, in the sense of code being deprecated out of a later revision.

There's a reductio ad absurdum argument about intersectionality that I've been meaning to make for a while, and it would probably make a good follow-on to this post. I will try to get that written up soon.
Apr. 23rd, 2015 02:32 am (UTC)
Very interesting article. We often take mental shortcuts to simplify multi-factor situations.
May. 11th, 2015 06:38 am (UTC)
I actively think of this as principal component analysis over an N-dimensional graph, when I'm thinking about it (i.e., using System 2). When I'm doing it, I just do it. (System 1.)
Apr. 23rd, 2015 12:46 pm (UTC)
idk if I've mentioned it to you, but just so's you know - vaporizers are v. hazardous to my ability to breathe. I doubt you'd be rude about it, as you were always polite in the past with cigarette smoke, but I know some people don't realize that even the vaporizers can cause problems for asthmatics. And I'd probably eventually have issues with the nicotine if I was exposed to too much of that, but I haven't ever been in that situation (only been stuck in clouds of cigarette smoke in clubs and pubs and bars and whatnot, which made me really sick, which I believe you and Len opined was probably due to hypersensitivity to nicotine, but quite understandably none of us thought testing it was a good idea, with how badly it mucked with me! I really hate blacking out.)

On to other stuff. Yeah, the models we use for intersectional issues are decidedly imperfect. I think of them as something like a trend line on a chart - it doesn't show the lie of every data point, but it shows the general trend. (ugh, I think that's the right terminology, but I should have been asleep hours ago. Bad insomnia! Stupid autoimmune crap!)
Apr. 23rd, 2015 01:16 pm (UTC)
You're not the only one, actually! A few other people I know -- one an ex-smoker, one who's never smoked -- get a buzz off secondhand vapor, so if someone has a bad reaction, I go outside.
Apr. 24th, 2015 06:33 am (UTC)
Yeah, I have trouble even with the non-nicotine ones. Something in the vapor triggers me. But I do have stupidly sensitive lungs. Fortunately, they seem to have gotten a bit better since I left Philly.

I survived about an hour at Christmas inside the house of my aunt and uncle, who both smoke (though they've laid off smoking inside it seems like except in their bedroom). Thank heavens they've backed off, because I had to leave early a couple years ago due to the smoke coming out of the soft furnishings, and it causes A Thing with that side of the family if someone who is physically in a location where they can come to a thing doesn't. I only went to Christmas because I was told we'd be eating outside...which we eventually did, but the food wasn't actually ready when it was supposed to be, and they only had one heater for outside, which wasn't actually sufficient for the group. We of course sat my grandma and my mother (who both get cold very easily) right under the heaters. I wished I'd brought a heavier jacket, as I hadn't been thinking about eating outside and only had my thin raincoat. If I'd been clever, I could have borrowed something warmer from my mom, but...I didn't think of it until we got to my uncle's.

Anyhow, before we left Philly, that would have had me outside and huffing off my inhaler and coughing, where this time I did need the inhaler but was just...uncomfortable, a bit tight...instead of having a more severe attack.
Apr. 24th, 2015 09:59 am (UTC)
Propylene glycol can be very drying to the throat, which I imagine could set off a chain of responses leading to an asthma attack. (OTOH it's also a carrier liquid for nebulizers and some atomizers, which I bet makes finding the right asthma medication a special hell for some people.)
Apr. 25th, 2015 03:32 am (UTC)
It's possible. It could be any one of the dozens of ingredients, though. I'm somewhat infamous for how sensitive my lungs can be. I wouldn't be surprised if there is a common flavor/scent agent that I react to, either, as some scents (especially artificial ones) have been a trigger for me since I was in my teens. The worst asthma attacks I had back then were because of a new shampoo.
Apr. 23rd, 2015 01:58 pm (UTC)
And yeah, "statistical models predict trends, for which there will always be outliers" is one of those basic numeracy skills that has somehow managed to escape a lot of people. A few months ago I was reading the blog of a high school math teacher who was pretty horrified to find out that many of his colleagues interpreted "students in certain socioeconomic groups are at risk of worse educational outcomes" to mean "all students in those socioeconomic groups should be put in remedial classes, regardless of their previous outcomes." Literally, people were arguing to take black honors students out of honors classes because the teachers didn't understand what statistics are for.
Apr. 24th, 2015 06:26 am (UTC)
oh FFS...that's...that's...AGH.

I don't even have words for it. The level of stupid, it hurts. It hurts.

By that kind of thinking, my mother - who grew up in a poor Irish-Catholic household - should never have gotten into any honors classes. She was a runner up for the National Merit Scholarships.
Apr. 23rd, 2015 04:26 pm (UTC)
I find myself implementing intellectual kludges. Like, on the one hand, I am pretty hardcore feminist. On the other hand, I am a spirited pro-sex kind of person. In many ways, these two positions are beautiful compatible. In fewer cases, but not a tiny number, each viewpoint "clearly" espouses a position opposed by the other.

It's been fashionable lately to draw male superheroes in poses similar to those often given to female superheroes. That's beautiful. It's subversive, funny, thought-provoking, and points to glaring inequities. But at it's core, the concept devolves partly on the concept that we're supposed to say, "Okay, that's not right," and then think, "So female superheroes probably shouldn't be depicted in that way, either." Sometimes, that is obviously true. There is some terrible so-called art of female superheroes out there.

But sometimes, when I see a picture of a male superhero in the genderswapped costume, I'm like, "Whoa, hot."

Sexual objectification disproportionately affects women. But de-sexualizing men erases the sexual being of women, as well as gay and bisexual men and others. So I end up saying, "I am pretty much fine with sexy male or female superheroes, as long as it's not stupid or overtly demeaning, BUT aside from that, male and female characters still need more balanced presentation in general." It's a definite compromise. Because I can't articulate a particularly pure principle it applies, it doesn't help much for people who don't share my separate views on feminism and sex.
Apr. 24th, 2015 10:11 am (UTC)
In a classical game theory model, this probably looks like you playing a mixed strategy, and in the annoyingly naive interpretation (where the pure strategies are either "all sexy pictures are fine" or "all sexy pictures are objectification"), a mixed strategy doesn't really compute to someone playing one of the pure strategies ("why are they okay with some sexy pictures but not others?")

The classical model unfortunately flattens out the nuance in "why", which makes it a bad model. I can mentally model the preferences you've expressed just fine, though, to the point where I think you could show me pictures of superheroes and I could probably guess whether you like the pictures or not. I'm going to have to think about how I represent that internally, because your position doesn't strike me as at all self-contradictory.
Apr. 24th, 2015 05:32 am (UTC)
Do those two errors roughly translate to false positives and false negatives, as in the old "imprison one innocent person or let ten guilty people go free" tradeoff?
Apr. 24th, 2015 09:48 am (UTC)
False positives and false negatives are involved, but precision and recall calibrate false positive count and false negative count against the true positive count. Precision is (true positive / (true positive + false positive)), and recall is (true positive / (true positive + false negative)).
( 22 comments — Leave a comment )

Latest Month

July 2015


Powered by LiveJournal.com
Designed by Tiffany Chow