Log in

No account? Create an account
Trevor Stone's Journal
Those who can, do. The rest hyperlink.
2D Gender Graph 
14th-May-2007 09:20 pm
Vigelandsparken heels over head
In the shower last night, I came up with an interesting idea. Partly it's about using scientific techniques in humorous ways, but there may be something worthwhile. I don't claim it's anything resembling a perfect model of the world, but hopefully it's at least entertaining.

People often talk about a "sexual orientation continuum" where gay is at one end, bisexual is in the middle, and straight is at the other end:

Gay                                Bisexual                                Straight
One can then use fuzzy logic to talk about attraction: "I'm 90% gay," "I like guys about as much as girls," "I'm not as straight as I act." Not a perfect representation of reality, but hopefully more accurate than three words and the instructions "circle one."

We can apply a similar idea to gender and style:

Feminine                                Androgynous                                Masculine
and to physical features and hormones:
Female                                Intersexed                                Male

Suppose that we put the latter two continua in a two-dimensional coordinate system:

(I've put numbers to allow convenient reference and because I'm displaying this with HTML tables instead of gnuplot, but the axes are intended to be continuous, not discreet.)

In this graph, "type 1" people have very female physiology and very masculine behavior. "Type 4" people have moderately male physiology and very masculine behavior. "Type 15" people have very male physiology and androgynous behavior. "Type 18" people have intersexed physiology and moderately female behavior, and so on.

We can use this graph to determine a person's gender empirically. Have them consider a wide variety of people and assign an attraction level to each; let's suppose attraction ranges from -1 for completely repulsed to 0 for no strong feelings to 1 for total infatuation. By plotting each person on the above graph with attraction level in the third dimension we can infer a model of a person's attraction patterns.

Claims like "I'm only attracted to girls" can thus be finessed: does the speaker have a single peak centered in type 6? Does it slope from 0.25 somewhere in type 13 to values near 1 in types 16, 21, and 22? Are there local maxima scattered about the left half of the graph? The claim "I'm not in to guys" could show as a value around 0 on the right half of the graph ("naked men don't turn me on, but they don't weird me out"), the right side might have an average value very close to -1 ("gross! a penis! get it away from me!"), or it might turn out that it's just stereotypical men (e.g. type 5) that turn the speaker off.

Do I have something interesting here? Have I independently discovered a common technique in Gender Studies classes? Does your attraction graph look interesting? I'd like to hear about it. If it's a helpful way of thinking about gender I might be sufficiently motivated to make an interactive version. For now, here's a convenient copy/paste fill-in-the-boxes version for those whose table-fu is not strong:

+0.0 -0.0 +0.0 -0.0 +0.0
-0.0 +0.0 -0.0 +0.0 -0.0
+0.0 -0.0 +0.0 -0.0 +0.0
-0.0 +0.0 -0.0 +0.0 -0.0
+0.0 -0.0 +0.0 -0.0 +0.0

Feel free to suggest graph locations for famous people, fictional characters, or folks you know personally. You can express them as quantized types from the table above or as [-1, 1] valued <sex, gender, attraction> triplets like "I think Arnold Schwarzenegger is a 1, 1, -0.3" (a somewhat repulsive type 5) or "I saw a hot drag queen last night... I'd say Mary was 10%, -90%, 75%" (a rather attractive type 23).

Incidentally, I've oriented the continua and axes such that female and feminine are in the positions traditionally assigned to negative values. My intent is not to imply that female and feminine are "bad" and male and masculine are "good." Like electron and proton charges, the assignment is arbitrary and graphs with axes reversed are just as valid when compared with like-oriented axes. I put female and feminine on the negative side in part because that's the yin-yang association and partly because "F" comes lexically before "M" and English associates "before" with "left." The cultural-linguistic challenge is to disassociate "positive" with "good" and "negative" with "bad." People threatened by flood feel that negative change in the river level is good while people threatened by drought feel that positive change in the river level is good.
15th-May-2007 04:39 am (UTC)
the confusingness of your graph singlehandedly illustrates the absurdity of this thing called sexuality/gender. nice going, bro.
15th-May-2007 05:34 am (UTC)
Wow...this is brilliant. This is interesting because I was talking about the Klein Sexual Orientation Grid today which is useful to a point--and certainly more useful than the Kinsey scale--but really seems inadequate considering all the variation of sex and gender out there. For one, both those scales seem to ignore people who are genderqueer, trans, androgynous, or intersexed in favor of a two-gender/sex model which I guess works for most people but not for everyone. (Bothers me for the same reason I don't like to describe myself as bisexual--I believe in and am attracted to a wider range of genders and sexes than just two.)

Mind if I link this from my blog? I think this really relates to some of the stuff I've written about the subtle distinctions between sex/gender/gender roles and attraction. I know I've talked to people before who ask, "Well, if gender is on a spectrum, what exactly does that look like?" and I've never found a model that described it this well. Of course, there's still the issue of what is "masculine" and "feminine" since I think those words are fairly meaningless and their definitions vary from person to person, but it's also the best descriptor we have, unfortunately. I suppose the only real improvement that could be made here is a distinction between gender identity (gender the person feels they are) and how much they conform to gender roles (socially expected behavior for their sex). That might make things too complicated to easily graph, though.
16th-May-2007 04:34 am (UTC)
If I didn't want it linked, I wouldn't have put it on the World Wide Web :-) That said, it's a rough draft of a musing; I reserve the right to change my mind about anything I said.

Case in point: I deliberately avoided precise definitions of "male," "female," "masculine," and "feminine." The basics of male and female are pretty simple: X or Y chromosome, penis or vagina, chest hair or breasts, and so on. I also have hormone levels and non-sex physical characteristics in mind, which aren't as cut and dry since most are a matter of degree.

I think we can make significant headway in defining "masculine" and "feminine" by correlation to gender. If a lot more men do P we can label that masculine, if Q is a majority-female activity we can call it feminine. This is predicated on the (reasonable, I think) assumption that there are a lot more types 5 and 21 than types 1 and 25. Just as it's produced differences in typical weight and height, evolution has produced tangible differences in typical behavior between males and females, and those differences can be classified masculine and feminine. There are doubtless cultural influences as well, but I think they can be classified (at first pass, anyway) in the same way.

Again, I don't claim this is a perfect model nor is my goal to conclusively define gender terms. The graph may even be useful to people with different perspective on what constitutes gender. Someone raised in a matriarchal society would consider leadership a feminine attribute and thus might place, say, Hillary Clinton as a type 22 while someone raised in patriarchy would label her a type 2. A composite plot of a few hundred people done by many participants in two cultures could be a way to study differences in gender expectations.
15th-May-2007 06:02 am (UTC)
Hmm... it's an interesting idea, but I think there are enough other dimensions of attraction to confuse things.

I like masculine guys, but it's a particular flavor of masculinity. Most of what I like is type 5/10, basically, but there are plenty of type 5/10s that I don't like. The women I find attractive are usually relatively butch -- but there are individual exceptions. And so on.
16th-May-2007 04:46 am (UTC)
Hmm. Let's say height and strength dimensions both project positively to the male dimension and thus tall and buff mean a more male value than short and weak. The projection to one dimension loses the distinction between tall and weak (goths) and short and buff (wrestlers). I can definitely understand attraction to one but not the other, but they might both project to the same "male" value.

Perhaps smoothness of the graph indicates how well the model fits someone. If a person assigns wildly different attractiveness ratings to near neighbors on a regular basis, it's not a very good model of their patterns. But if others have a rather smooth graph the model may be useful for them.

Perhaps it's mostly useful as an introductory tool... "Single type 9 seeks type 21 or 22 for gender discussions at coffee shops..."
15th-May-2007 05:35 pm (UTC) - ::chuckles::
That is fairly brilliant.A simple system that covers the most commonly addressed types. While usefull for generating stats, I don't think it fits most of the people we know...my own place/number would change daily!
16th-May-2007 04:57 am (UTC) - Re: ::chuckles::
Time can certainly be a fourth dimension in the model. One could periodically make a 3D attraction plot and then show a Time Lapse of Sexiness. "Wow... back then I was totally into type X but I've learned how awesome type Y can be."

Ratings of a particular person also vary over time. After a few dates, someone might say "I met this type 18... I'd say at least a 0.8!" Three months later the comment might me "That type 18 hasn't called in weeks. Downgraded to a 0.3!"

And a person can certainly change position, particularly along the gender spectrum. Changes to sex characteristics are tougher but certainly not impossible.
15th-May-2007 11:47 pm (UTC)
Okay, first I kept reading it as "Masculine, Hydrogynous, Feminine" so that took me a little while to figure out. Sometimes my brain plays practical jokes on me.

Second, in response to the question "Who's to say what's masculine and what is feminine?", the Qabalah gives a pretty exhaustive account. Once I got to thinking about that, it reminded me of an interpretation I've heard of the word Yahweh/Jehovah (alternate pronunciations of the Hebrew characters YHwh, where w for some reason is the equivalent of small Y). This is actually a name that cannot be pronounced - not which may not be pronounced - because it's not a name; it's what the characters symbolize. Specifically, Y means the spiritual feminine and H means the spiritual masculine (or possibly the other way around; my Hebrew is weak), and w and h symbolize the physical masculine and feminine. So the simple "name" for God is "the union of spiritual and physical, masculine and feminine".

Well that's almost too obvious, isn't it? I mean, what else is creation but mingling the physical and/or spiritual masculine and feminine? By which I do not mean to imply one male and one female. Ideally, a single meat-suit-dweller could be prefectly balanced in all four aspects and theoretically this would make hem a profoundly and prolifically creative person.

So kinda what you've got there is a god graph.
16th-May-2007 05:02 am (UTC)
Maybe God is a type 13. Maybe Jerry Falwell is flipping out right now as he learns that God is a bi switch with an ambiguously long clitoris/short penis.

Or maybe God is all types at once. A superposition from -1 to 1 in all dimensions. Simultaneously the most attractive and most repulsive being you can think of.
This page was loaded Apr 23rd 2018, 9:22 am GMT.