• 1
You are so beautiful to me right now. ♥

Holy shit the rage began to overwhelm me toward the middle of your post but I kept reading because apparently I don't hate these guys enough at this point.

HOLY. SHIT. I barely knew about the kind of model building they're involved in until now (a bit from the Psych angle, a bit from the AI angle, but very little overall) and I was confused about what it is they are actually hoping to GAIN with this survey for their research. This detailed that for me in beautiful and horrifying colors. JESUS FUCKING CHRIST, THESE PEOPLE.

they stated somewhere that the purpose of their research was to show that not just fMRI results should be used in neuroscience and complaining that currently "it's the only way to get published". I JUST CAN NOT BELIEVE THEY ARE BUILDING BEHAVIORAL MODELS.

SUBCORTICAL REGIONS ARE LESS COMPLICATED THAN THE VISUAL SYSTEM?!?! This is like, sci fi levels of unbelievable coming from people with ANY kind of Academic degree (you'd think they'd been taught to research what they didn't know).

I shall now go and try to stop being angry and fail miserably.

An active brain is one where the neurons are continuously changing in response to each other. If they aren't, you're dead.

Uh-oh...

...

x_x

b)data to program the system; ie, women find THIS sexy which will cause the amygdala to respond'.

Oh, my God. I bet you've got it. It makes SO MUCH SENSE when they say that they're not doing social science now.

It makes SO MUCH SENSE when they say that they're not doing social science now.

My own brain is in conflict. On the one hand, you're right, this *might* actually explain what they thought they were doing.

On the other hand, it's such a *stupid* idea that my brain keeps tossing it back out like a slug in a vending machine.

Oh, read this. Read this now. http://ken-jennings.com/blog/?p=568 It's a description of Dr. Ogas on a game show, and here's the key quote:

The more exposure you get to Ogi, the more you see that he’s just like that. He’s not putting on a little show to mess with your mind–he’s just got a very specific, determined way of doing things. Sure, he’s a bit of an oddball, but in the world of game show uber-contestants, that’s hardly a capital crime. The congenial, funny on-line Ogi is the real Ogi too, but unfortunately for him, you can’t close the barn door once the ass is gone. And Ogi explained to me that the greenroom comas had nothing to do with mind games and everything to do with circadian rhythms (or downloading new data from the mothership, or something).

Me:Basically, he has an area of expertise, and he thinks it explains everything. Everything everywhere.

I think you're right, but it's important to point out that he's not using cognitive neuroscience in an orthodox, extensional way, he also apparently doesn't have any problem *intuiting* what he thinks of as a cog sci *approach* to wildly diverse areas.

I actually think you guys are assuming much too much intellectual honesty on the part of Ogi and co. I doubt his book is a real attempt at cognitive modelling(however flawed) his mention of a repeat with the Right implies to me that he's writing a politicized social science book, with cog sci technobabble laid over it.

Unfortunately, that kind of book where political, moral, or social assumptions are laid out in a nice, self-confirming shallow analysis, with a light glaze of Scientific American-like language is really popular.

I just feel really bad, because things like this end up with me feeling pressure to hide my field, or opinions on it's applicability, because I'm afraid it'll get mixed up with this kind of idiocy.

They say a little bit of knowledge is a dangerous thing, and that goes double if it's knowledge in a really powerful, compelling scientific area. Because people get so excited by what it explains that they then extend that authority as far as it will go. And often further. And then attribute any success they have to the strength of their ideas. And fall further into a black of of self-confirmation, making the rest of us scientists look bad.

Thank you for both the essays.

Very illuminating and clear for people who have little or no background on this subject.

Makes me happy my gut fits with facts.

I haven't really commeted about this on any of the journals because I'm having a hard time just keeping up with reading everyone's fantastic responses. But thank you for this explanation! It does clarify why they say they are not doing social research with their survey.

Thanks for this post! I've been focusing on thinking about the 'datamining' aspects of their work, so it's cool to hear about the modeling side.

You are a much better science teacher than either of those guys. ♥

Thank you--thank you!

Just an additional note which you might have missed: at one point (I think it was in Shaggirl's journal, but I cannot swear to it; maybe it was in a now locked explanation post in Ogi's), he said that some of the questions in the survey were included BY fans, just FOR fans, so they could give something back to the community, even though they were not interested in the questions--he implied (or so I read it as implying) that these were things suggested by their fan correspondents, and in no way that I recall did he identify WHAT ones were fan generated and which ones were real.

Which struck me as bizarre becuase who does that in survey work?

And that also skews any attempt to figure out what they might be doing looking at their survey questions which of course began changing immediately....

This was fascinating to read though, and proof that someone who knows the topic can explain to people who do not. Ogi never managed that, but I don't think he intended to: I think he was obfuscating and bullshitting the whole way through!

Link for info cited above!

http://shaggirl.livejournal.com/186106.html?thread=3047418#t3047418

This is where he talks about his big mistake was including questions FOR fandom (us ungrateful purported anonymouse types rejected his GIFT)--I found it on unfunnybusiness, yay for them!

Thank you for both essays. Even recognizing your suppositions as educated guesses, I think you've provided much needed insight as to what they thought they were (and probably still are) doing. Sadly, I think they'll still get published, and this idiocy won't even be a blip on anyone's radar.

In this context, it worries me that these guys mentioned nifty as well. Another group of othered self-providers.

And, er. A young woman who has had her sexual socialization guided in part by slashgirls *raises hand* will see nothing deviant about slashfic. It is entirely normal.

Thanks so much for this post. Laying things out in a clear, concise manner for the layfolk was helpful and informative and it was a little way in diffusing my own cynical idea that 'yeah, these guys, having used the fandom community will now write a pseudo-science piece of crap book and make a boatload of money off our backs".

I still think they'll go ahead and do this and I still feel like they used the largesse of people who did their best in a variety of ways to educate them. I suppose I see some small difference in the idea that they seem to feel that since they have their newly-minted diplomas in their field of expertise they now think that their extra-special knowledge entitles them to make a buck off the backs of people they never even took the time to understand. Just because they really, truly feel that way doesn't make it okay, though, and that lame-ass 'apology' on ogi_ogas 's now-sealed LJ only reinforces the impression I read from them after things had gone past a certain point, i.e., a pair of condescending, mysogynistic homophobic guys who think way too much of themselves and thier degrees--we don' need no steenkin' ethics!

Again, thanks for the informative post; Aside from the suvey-fail, I learned a thing or two and that's never a bad thing.

Edited at 2009-09-03 10:15 pm (UTC)

We don't know much about the brain, relatively speaking, but we do know this: all human beings have one.

You are awesome--and for both of these posts, not just this particular line.

And for what it's worth, I'd definitely agree with your take on Ogi's motives--I've met a fair number of people like him, in academia and outside it, who think that thanks to the fancy initials after their name, they know everything. Or, to be more specific, that their little corner of expertise gives them the Ultimate and Only Angle on the world.

Edited at 2009-09-04 04:21 am (UTC)

Thank you for this wonderful explanation, and for the other post detailing neuroscience for dummies. Your responses to the comments were things of beauty, as well, and I for one am very glad you needed an LJ to write all this out and continue responding.

I couldn't agree more. The idea that these men have, you know, suddenly discovered the perfect model to analyze desire? sexuality? anything at all? is ludicrous.

To come at this from a CS/AI perspective: what we refer to as 'neural nets' are, in fact, hideously, hideously flawed models of any kind of human behavior. Which is something that most CS research at least has the decency to admit. We've come to the conclusion that what we call neural nets are useful in certain types of programming situations, but that they are in no way predictive of human behavior. Which, as you say, is the whole. frickin'. point.

My best anecdote (as told to me by my college AI professor): a neural net computer was designed to recognize the difference between enemy planes and friendlies. After being 'trained' on some number of 'good' and 'bad' images it was able to identify friend from foe... until someone introduced a new set of images that were taken at a different time of day and it suddenly got confused. At which point the scientists working on the neural net discovered that the whole time they assumed the computer had been looking at the planes, it had been drawing its distinctions from the color of the sky and hadn't been looking at the planes at all.

Seeing as this is the most sophisticated computerized model we have right now for the human brain - and we apparently don't understand what it's doing either... I shudder to think what model these men have concocted from combining two whole academic disciplines that have openly, publicly thrown up their hands and said "WELL WE DON'T KNOW YET OK."

So even if they do use this survey data to train up a neural net into telling them the nature! the very essence! of female desire... how would they even know that women are responding to their (exceptionally ill-founded) assumptions, and not just the scenery?

the scientists working on the neural net discovered that the whole time they assumed the computer had been looking at the planes, it had been drawing its distinctions from the color of the sky and hadn't been looking at the planes at all.

It's not just neural nets that have this problem. In the Second World War, the Russians trained dogs to run underneath tanks, carrying explosive charges that would destroy the tank. When it came to battlefield deployment, the dog units hit a number of problems, not least that the dogs had been trained to run underneath Russian tanks, not German tanks. Specifically, the dogs had learned to go towards the smell of the diesel engines used in Russian tanks, and avoided the unfamiliar smell of the German petol engines.

Just wanting to add my voice to the many thank yous on this page: thank you! I couldn't make heads or tails of the two doctors' explanation, neuroscience not being my field, but you've explained it, and their fail, very eloquently. So thank you for taking the time to do that.

Thank you so much for your explanations, they really helped clarify things for me. And oh, Turns out the brain is actually parsing a near continuous nondifferentiated stream of sound. Oops) this? As one of those frustrated early adopters, SO happy to know it wasn't just me. Sheez.

Thanks for posting this. My science background is chemistry, but I kept looking at their "science" explanations (which had I distinct whiff of "you can't possibly understand this") and going …wuh?

In particular, the argument that slash and male interest in "shemales" is clearly linked to the subcortex because it was not what culture predicted, followed quickly by saying that they weren't looking at culture. You're…using culture when convenient to define the boundaries of your interest and then discarding it entirely in interpreting your data? You can't do that!

Even if you set aside all of the assumptions, all the male privilege, all of that, I was amazed at how useless the questions actually were. As many people pointed out, there would be a question with five options, three of which are about frequency of an activity and two are about whether you like it. How can you get any data from that? You have (in one question!) failed to isolate separate effects. There were also a bunch of questions that were impossible to answer in a meaningful way, like the one asking if we identify with the "dominant" or "submissive" character. Because it was so poorly defined, and even if it had been better defined, was looking at the wrong parameters for identification with a character in a story (and failing to see "identification" as a fluid concept that does not mean "picturing myself as that person"), even if they got a thousand people to choose an option, none of those people mean what the authors are going to assume they mean.

The survey was just…vague and irrelevant in a lot of places, clearly designed to support their assumptions.

And if you make crap assumptions, you're going to get crap science from it. I mean, I could assume the universe is made of pudding, but I'm not going to get any useful predictive models if I do that.

I'd totally read the paper, though.

Hello my friends

(Anonymous)
Hello, I new yours frient on this forum)

  • 1
?

Log in