[Noisebridge-discuss] Gender & Technology followups from 5Mof

[maymay]

On Sep 18, 2009, at 10:11 PM, Sai Emrys wrote:

>> "Do you want to get busy with a person who has a penis?"
>
> Not really, IMO. For instance, pre-op MTF trannies (aka "chicks with
> dicks") are almost exclusively a straight male thing.

Sorry, can you clarify? A straight male thing in being that pre-op  
"MTF trannies" are straight men or that straight men want to  
experience sexual activity with these people? If the latter, how do  
you know this is true?

> So I think it is more accurate to say that people are looking for
> specific gender(s), broadly.

I think it's difficult to quantify the components that make up  
someone's feelings of desire, and while a significant portion of that  
might be described as desiring specific genders, an equally  
significant portion is not. One of the problems with the Yay!  
Genderform example is that it overloads the term gender to describe  
things that, IMHO, the word gender cannot meaningfully describe; using  
specific language to dissect the context is really important, and  
we'll lose that if we always call anything that has to do with  
sexuality someone's "gender."

>> This algorithm has a ton more assumptions in it from a technical  
>> standpoint
>> than I think necessary,
>
> "A ton"? Enh... not sure about that. I think it's fair to say that the
> majority of people are into one standard gender or the other.

I *think* I disagree with you here. Being "into" something is as vague  
as desiring something in an unspecified way (see above paragraph), and  
I think that "the majority of people" really aren't aware of the  
nuanced distinctions that make individuals different when it comes to  
their expressions of femininity or masculinity or something else. The  
fact that the majority of people *appear* to be "into" a "standard  
gender" looks more like a mirror onto the current state of most  
people's (mis)understanding of gender than anything else to me. In my  
experience, very "standard"/"normal"/"straight"/"vanilla"/whatever-you- 
want-to-call-it people are quite a bit more nuanced than even they  
realize at first.

> Those
> who aren't probably list themselves as bi, and that moots that issue.
> ('cause bi actually means 'no preference' in this context; insert
> nonduality spiel here.)

As a self-identified bisexual, I'd dispute that, too, but that's a  
totally different argument. :) More on that subject to be found on an  
old blog post of mine.[0]

>> self-identify as "male" have penises as well as assuming that this  
>> is the
>> way every person will interpret the question "Are you male or  
>> female?" that
>> they were explicit about what they were asking.
>
> What about the above trannies, who identify as female and are mostly
> sought by straight males, despite their having cocks?

Okay, so, there are a few issues all wrapped into one question here.  
One is the way in which people identify themselves independent from  
the physical bodies they have. Another is the distinction between  
desiring something and wanting to be something—sometimes these are  
similar and sometimes they are not (e.g., "Wow, that person is really  
cool. Do I have a crush on them or do I just want to mimic their  
behavior because I admire it?"). I'm not sure which you're addressing,  
if either…?

> For that matter: the more questions you ask, the more people go away.
> You simply cannot expect to be able to ask all your users a long
> inventory.
>
> More especially, you cannot ask all your users questions that
> presuppose that they have a genderqueer identity. Most do not. Shoving
> something like that on them inherently makes a sociopolitical
> statement, and my apps at least try to be studiously neutral on
> anything not directly related to the functionality.

Yeah; from a technical standpoint, it's all about functionality in the  
end. That's why it frustrates me to see male/female drop down menus in  
places where the answer to that question means nothing to the  
functionality being offered (i.e., hotspot example in presentation).

>> but I think you and I both know that "good enough" code is often
>> pretty damn lousy when examined more closely.
>
> Of course. But why would you?
>
> Especially when dealing with something when the reality is
> impossible-to-categorize-ly complex, all you want to even think about
> is 'good enough'.

Therein lies why it was someone like *me* who came up with that  
presentation in the first place. I totally examine "good enough"  
solutions to problems like gender assumptions in application code that  
were not handled well enough for me, probably because I am an edge  
case. So I don't want to think about what "good enough" for other  
people is, I care about making things better for myself, which—being  
an edge case—is tough. :) The personal is totally political.

>> I think the Yay! Genderform web site is a perfect example of why  
>> trying to
>> provide every single possible option in an application that doesn't  
>> need to
>> know too much about sexuality information is a bad idea. :) The UI  
>> is simply
>> unusable and there's an ontology problem here, too. (And not to  
>> complicate
>> things further but note that Facebook and FetLife's examples  
>> implement the
>> same values ["male","female"…] in a form field with a different  
>> label;
>> Facebook calls "male" a "sex" while FetLife calls "male" a  
>> "gender". Who's
>> correct? Not sure….)
>
> Indeed. And even those categories are insufficient; at some point you
> have to make it a generic text entry, and then you know you've lost
> any categorization utility. :-P

Not necessarily. I'd prefer a generic text entry over a drop-down menu  
in the case of, say, FetLife.com's "Gender" field since that field  
(AFAIK) is only being used as a display string and not for any  
operations that require metadata like profile searching. Using a text  
entry encourages the kind of free text-entry that emulates the  
behavior of gender variant people—who only loosely have any sort of  
standardized vocabulary today—than does any of the other interaction  
paradigms available in the modern web site interface.

Of course, a free text entry is not without its problems. The biggest  
of which is grouping, or categorization. This might range from the  
simple case of normalizing "Male" and "male" to the more complex of  
parsing individual words or phrases, but it's certainly not an  
impossible task. Moreover, the extensible vocabulary infrastructure  
widely used in the semantic web gives us some insight into how you can  
create formal taxonomies using these kinds of diverse terms. Creating  
a formal ontology and/or vocabulary for sexuality and gender has been  
on my mind as a side project for some time.

>> So, I'm not really arguing against categorization. I've worked as an
>> information architect before; I love categories. :) What I'm  
>> arguing for is
>> more attention to a very fundamental categorization scheme
>
> How about this: propose a better system. What ought a site like
> OKCupid do that's practical and better - both for questions of users'
> gender and what they're looking for?

OkCupid does a number of interesting things that they *don't* relate  
to gender but that I think could be interesting to experiment with. I  
say "interesting to experiment with" because I have no formal  
recommendations of what a better system might be, since this isn't  
even a question I've been looking at for a very long time and I've  
heard of no one else asking similar things.

One of the things they do is have this "describe yourself" bubble at  
the top of a profile page in which they give you three free text entry  
boxes to write what amounts to self-tags in the system. That's neat  
for a lot of reasons because it pools the user base into categories  
created by the user base, much like what's going on with gender in  
gender variant communities today. I don't know how or if that  
interface is tied into the user-matching algorithms that OkCupid uses,  
though.

It will always come back to functionality, and that's highly  
application-specific, of course. For a site like OkCupid, which shows  
little reservation about making users answer a whole lot of questions  
at some point or another, I think this kind of expanded self- 
identification scheme is likely to be a pretty good fit.

But, bluntly, I'm actually not as personally interested in dating  
sites as I am in the other places I see gender biases in society or  
technical implementations. Dating sites are just where the most  
obvious examples are. :)

>> Now we have
>> "web developer" versus "web designer" and even within there we have
>> specialities like "back end developer" and "front end developer,"  
>> if you'll
>> pardon my crude terms here.
>
> FWIW, I do full stack work - everything from making the (simple)
> graphics to optimizing database calls to ... well, everything.
>
> So I actually use those terms pretty interchangeably; I'm used to
> being me, as it were.

I've been a freelancer for years so, like you, I know how wearing many  
hats feels. :) That being said, that's no longer how the capitol-I  
Industry is most commonly structured anymore, and that's kind of what  
I was getting at. As the field has evolved, it has developed  
specialties *because we paid attention to the evolution of the field*.  
People who do web development work of any stripe at least know about  
the existence of the specialties they don't do, and I think the same  
thing can happen to an awareness of sexuality and gender in a  
significant way, just like it happened when web designers started  
paying attention to accessibility for colorblind, deaf, legally blind,  
or mobility impaired people.

>> I think it's unfortunate that a site like LinkedIn can, for example,
>> subdivide their "Industry" dropdown menu with a plethora of diverse
>> options—completely sensical on a site ostensibly trying to improve  
>> the
>> nature of your business's presentation—but doing something similar  
>> with a
>> sex or gender drop down menu on a dating site like OkCupid might  
>> actually
>> alienate the non-genderqueer-friendly population because somehow  
>> looking
>> beyond a binary is "wrong."
>
> I don't know whether it's unfortunate or not. But certainly it's true.
>
> IME: never expect normal people to be introspective, or to be aware of
> (let alone care about) tricky issues of identity. It just ain't true
> often enough.

Totally agreed. I know a lot of people who don't care how the web  
sites they visit work, and as a web developer I make a very concerted  
effort that they can remain blissfully ignorant and don't have to  
know. But I also try to make sure that the web site works really well  
and addresses their goals (on the site). My argument is centered  
around the belief that more sites—regardless of what their exact  
purpose is—can meet the goals of more people more effectively by  
becoming more aware of gender bias at every level, from interface to  
business logic, and if executed well, users who prefer not to  
introspect about the details won't have to know anything's changed.

> And bluntly, not enough people give a damn what your gender "really"
> is; they just want to know what categories they can use that will
> more-or-less accurately predict how they should interact with you.

Some people, myself included, are constantly confronted by the reality  
that by using the categories most known we are interacted with in ways  
we don't like and that, at their worst, don't feel respectful. I  
acknowledge that in this way, I'm a social edge case, but from a  
technical standpoint, making one's systems robust enough to handle the  
edge cases is what iterative improvement is all about. I see an  
analogy on a political level here, but I also see a practical,  
technical principle at work here that can actually benefit every case,  
not just the edge cases. (I.e., "be liberal in what you accept" RFC  
1122 slide from the presentation I gave.)

> Like I said, I see that sort of thing as something that comes up in
> one-on-one nonsuperficial relationships. Not that one can handle even
> en masse of one's acquaintances, let alone programmatically across
> millions of users.
>
> - Sai

Yet. :) I don't actually think many technical minds have been aimed in  
this direction yet. My Googling certainly shows a dearth of material  
about this particular subject. Or maybe I'm just not well educated  
enough to know what my keywords should be. But we also didn't know how  
to do public key cryptography before a whole ton of math geniuses  
focused on the problem, either. I hope we'll get there eventually.

Cheers,
-maymay
Blog: http://maybemaimed.com
Community: http://KinkForAll.org
Volunteering: http://ConversioVirium.org/author/maymay

EXTERNAL REFERENCES:

[0] http://maybemaimed.com/2007/08/30/what-almost-everybody-else-doesnt-get-about-bisexuality/