« If Everybody Had An Ocean | Main | simulating women »

October 26, 2003

social construction of technology

I have a tendency to point out when interfaces have gender biases in them. Lately, i keep getting asked what it would mean to create a separate interface for women. This question astonishes me, and i keep chewing on how to properly respond.

<broad generalizations> There are a lot of things that men and women do differently. For example, men are much better at spatial rotation, but their skills at differentiating shades are atrocious. Yet, there are certain tasks in the real world that men are better suited for and certain tasks that women are better suited for.</broad generalizations> The real world doesn't have a gendered interface; it simply allows for different readings, different levels of access. For example, we all manage to negotiate three dimensions, but the cues we use to do so differ.

Technology is socially constructed. In other words, technology does not exist devoid of its creators' prejudices, biases, cultural assumptions, etc. When men design and build toys and then have other men test them, it should not be surprising that the common experiences of those men get imbibed in the technology. (If this seems surprising, imagine what happens when you assume American roads, habits when designing a car for Amsterdam or Japan. Cultural dependence is not that different.)

The trick is not to design a separate interface (remember, separate but equal never really worked). The goal is to incorporate a wide variety of perspectives into the design and creation of a system, to create a system that people can repurpose to meet their needs. The goal is to encourage flexibility of expression, to not project a limited perspective into the technology. Designers must take into consideration the vast array of potential users, experiences, expectations, not simply their own. This is why things get tricky.

Posted by zephoria at 11:51 AM in Academia | Permalink


Exactly! Separate interfaces is a terrible idea! But adaptable interfaces is wonderful - at dinner last night a bloke kept going on and on and on about the differences between men and women, you know, all those annoying dualisms where you have to be this or that and if you're this, you can't be that, and my god it drives me crazy.

I'd love to hear some examples of gendered interfaces, too, Danah - I've not explicitly thought along those lines that I can remember.

Posted by: Jill at Oct 26, 2003 12:11:07 PM

Jill - the latest bout of it stemmed from my irratation that sites presume the majority of desired users to be male and thus elect male/female even when they know it's not alphabetical. (i.e. there are those who don't realize and thus make cultural blunders and there are those who explicitly feel as though men are the target audience because they are the early adopters).

Of course, this topic has a larger history for me. I spent two years looking at how prioritization of depth cues is dependent the level of sex hormones in the brain. This in turn impacts our ability to negotiate virtual reality and 3D interfaces because the cue conflicts are so great. My favorite quotes during that research came from the military who told us that women could adapt to the new cues, but that they got violently sick returning to reality.

The language that sites tend to choose has a lot of biases in it. [Of course, then there are the sexual inuendos generated by the UNIX world.] The pictures tend to express who the target audience is. Things like "shopping carts" instead of "baskets" convey an American-centric perspective. The types of ads that are targeted express a specific notion of feminity/masculinity and when they aren't targeted, they're often male-centric.

Even tech advertisements tend to have these interesting connotations... wait... i'll post that one up top.

Posted by: danah boyd at Oct 26, 2003 12:19:30 PM

you seem to have missed something; interface design is a historic weakness of i.s./i.t. in general; is it at all surprising that it fails in terms of being loaded with the same sorts of stupid gender-insensitive issues as technical documentation? no doubt the situation is easily improved by using more talent -- particularly female talent -- and better revision practices to take care of interface issues.
it also won't happen because that costs time and money outside open source, which is still not acceptable to many developers. i am wary of open source though because in my experience it is often a place where the worst of gender insensitivity flows due to there being no corporate human resources people to complain to. there isn't much "homophobia" since there are plenty of male gay programmers, but i've seen "chs" and "red ones" used in the discussions there.
--random noise

Posted by: anonymous at Oct 26, 2003 1:24:45 PM

One thing to avoid in discussions of this nature is broad generalizations which don't help women or men. I've read discussions about how generally women favor games that emphasize building and long-term strategy over destruction and quick sensory overload, which might make some sense, except acting on a generalization like that leaves some people out in the cold, such as some of my female friends who are rabid FPS players.

Nobody is the mean -- who do you know that has 2.6 kids? -- so perhaps another way to express Danah's point is that designers should use easy categorizations (gender, race, etc.) as a way to push themselves outside of their own specific design needs. A male designer who considers the average spatial skills of women isn't just helping out women users; he's also helping out the men whose spatial skills might be off-mean as well.

There was, IIRC, an analogous movement occurring in the world of consumer product design with respect to handicapped users -- designers realized that you could just design for both normal users and the handicapped. They called it "universal design". A telephone with large number buttons, for example, is helpful to the vision impaired or those with motor skills deficiencies. It also looks pretty good in your average dorm room.

Posted by: Francis Hwang at Oct 26, 2003 1:55:52 PM

If you had given an example, I would understand this post, but it looks like another biased female view on the issues and an attack on male employees. So according to the author, it is ok to kick male employees and recruit female employees just to get this perspective. This is just sexist. The author is openly sexist, but probably thinks she can get away with it because she is a female!

Posted by: Tim at Oct 27, 2003 1:39:19 AM

Tim, i don't understand where you're getting your impressions. I am not saying that it is OK to kick male employees and i don't understand what example you are referrring to or why you are calling me openly sexist. What i'm arguing for is a variety of perspective, of conscious consideration of the various ways in which people perceive and interact with the world. People can understand experiences different than theirs when they open their eyes. The point is to encourage everyone to realize that technology is socially constructed and that we need to move outside of our comfort zones to include a greater variety in order to be more encompassing.

Posted by: danah boyd at Oct 27, 2003 1:48:08 AM

Francis - i actually disagree with your devaluation of broad generalizations. Our brains constantly assess and negotiate stereotypes (i.e. broad generalizations). The problem is not in processing those generalizations, but in relying on them to be always valid and in expecting everyone under the supposed umbrella to fit there.

Broad generalizations often give us concrete positions with which to engage dialectically. Through the comparison, contrast and questioning of these broad generalizations, we can develop a richer picture that is more likely to encompass a wider diversity of whatever the broad generalization is addressing. The key is to acknowlege that you are working with broad generalizations in order to draw a larger picture.

Given that, you're dead on. While i can tell you about cue conflict processing differences between men and women, there is a far wider range amongst men and amonst women than between the two. Yet, by discussing it dichotomously, we not only make room for a wide range of in-between behavior, but we provide a structure that allows us to address observational claims about the differences between men and women (such as those by the Army). If we simply talk about the large range amongst humans, we have no foundation on which to address claims that women appear more susceptible to simulator sickness.

Thus, my general belief is to tease out the different axes on which you can analyze or generalize, but always be aware of doing so. [Thus, my HTML indicators that i was making broad generalizations for the sake of the discussion.]

Posted by: danah boyd at Oct 27, 2003 1:58:36 AM

Hmm, interesting. Perhaps this explains why my mother can't seem to navigate through the computer interface but she sure knows what she wants to write once she gets to the WP. She is also 77, and I've started to wonder if there is an age factor, most discussions on Indigo Children talk about how people are being born with different wiring than they used to, and so maybe youth of both genders can easily relate to and navigate a virtual interface.
As a not-so-great example, my dad told me that back in horse and buggy days, if you got drunk and had to drive home, the horses knew the way, which is a lot different than pressing buttons and levers to drive a car, whether impaired or not.

danah wrote:
There are a lot of things that men and women do differently. For example, men are much better at spatial rotation, but their skills at differentiating shades are atrocious. Yet, there are certain tasks in the real world that men are better suited for and certain tasks that women are better suited for.

Posted by: elf at Oct 27, 2003 2:21:14 AM

As I railed about in a blog entry a while back, one of the main issues with designing technology either objectively (ha!) or inclusively is that the "well-designed technology" approach is in direct opposition to the "educate the user" approach. At the moment our biases are not only built into the technology, but taught as lore to those users who must use the technology, by people who have those same biases plus some biases of their own.

More to the point, there are huge numbers of people who make their paychecks by providing the educational service of reconciling how people think things should work (user expectation) with how things actually work (user experience). Well-designed technology is a big step toward putting the technological priesthood out of work. It's not just a gender issue, or a competence issue, or an accessibility issue. Poorly-designed technology serves a very significant function, namely, to provide a social role to those whose job it is to make up the gap between expectation and experience.

Posted by: Michael at Oct 27, 2003 3:39:10 AM

As far as I know girls are better at literature classes. They can write better and they seem to be keen on reading too. I think the teachers should take this fact into account when they grade. My literature grades were always lower than girls' grades, and my girl friend seem to be better at writing too. So, I say male students should get some adjustement, or more male teachers should be employed to correct this unjust situtation. See the point? This is sexist as much as your comments are.

The post doesn't say any specific, it doesn't give a concrete example. It tries to show that there is a problem, without saying what exactly the problem is or whether the real cause of the problem is the fact that males design only for themselves. You suddenly divide normal people as male and female and then you argue that, male designers are selfish and that they only design for themselves, you don't even say what is it that they only design for themselves. It sort of tries to provoke normal females to believe that they are victims, and I think this is outrageously sexist. I am not calling you sexist at all, cause every normal person can see that already.

Posted by: Tim at Oct 27, 2003 5:59:03 PM

what a great contribution this site is... thanks for posting the link... brings to the consciousness again, a realisation of the gendered nature of technology.
it reminds me of an interesting international anthology publ. by spinifex press australia, called "cyberfeminism" (ed. susan hawthorne and renate klein). presents the thinking of feminists working in the fields of electronic publishing, data delivery, multimedia, virtual campus, program development etc.
grapples with the dominant discourses and pitfalls of the medium.
i am also reminded that we presume much about both the intention of ... and the manner in which technology is developed
my interest lies in examining which voices are privileged as we advance (?) our society.

Posted by: elena at Oct 28, 2003 6:29:45 AM

I have to say that any differences between men and women that might affect computer interfaces are probably so minimal that, if a typical computer interface causes problems to one or other group, it is very poorly designed.

For example, how much spatial rotation do you need to do to operate the average checkbox? Which precise shades need differentiating to select a menu item?

Perhaps the spatial rotation issue is one more reason why the '3D virtual interface' (to software such as the operating system, not games) was an unbelievably cretinous conception, but did we really need another reason? Those interface attempts died years ago.

Differences enforced by disabilities are far more significant, particularly complete loss of vision. General practice in the accessibility world is that blind users should use the same software as everyone else; there are many arguments in favour of this approach, but it does usually result in poor software for blind users.

That doesn't at all apply for any male/female distinctions: excellently designed software for male users will be excellently designed for females too. All the arguments in favour of common software, however, do still apply: version parity, ability to use software together/share a computer, etc. etc.

What's more, I'm fairly certain that 'women's software' would translate into 'pink'... ;)

The ordering and/or default selection of male/female radio buttons is a completely different issue, really, and I think a rather silly one: you have to pick an order and there's no particular reason why either ordering should apply, beyond the societal norm of 'male/female' rather than 'female/male'. Using the societal norm is a better idea, just as, in software aimed at techies, it's better to order colour options 'red/green/blue' rather than 'blue/green/red' (even though the latter is alphabetical).

You may not like the societal norm but it does provide a convenient, arbitrary interface default. (Similarly, I don't think the typical order of titles, which is probably Mr / Mrs / Ms / Dr, discriminates against PhDs.)

--sam (who likes pink, actually, but hey)

Posted by: sam at Oct 28, 2003 11:40:43 AM

Very interesting post (actually, whole site enlightening). Just a minor gripe: calling things that men design "toys" will only encourage people like Tim above. The rest of the post felt-calm, even-handed - hell, a grown-up discussion of an interesting point. No need to get lazy.

Posted by: rich at Oct 28, 2003 11:44:39 AM

Rich - thanks for catching me on my hiccup. I was trying to remember to be more serious with this post, but my general language does come out...

What you are running into is my tendency to refer to all tech creations as toys (instead of tools since i think that they are far more for play than usefulness in most consumer uses). This doesn't connotate 'bad' so much as trying to show that what we all create is not simply for the productivity that we pretend to focus on in the States. I realize that there are some productive tools, but i tend to default to the inverse.

Posted by: danah boyd at Oct 28, 2003 2:42:00 PM

It would be interesting to study this through the lens of Air Traffic Control. Though I don't have exact data, it appears that at least 25-30% of all controllers are women. Except for tower and ground controllers, whose work is mostly visual, controllers interact with the world in a computer-mediated fashion. I know that the FAA and its contractors have done extensive human-factors work; it would be fascinating to know if they have picked up on any gender differences in interface design.

That said, I believe there are tremendous *individual* differences in cognitive style (and hence in optimum interface design), and that these probably outweigh any gender differences.

Posted by: David Foster at Oct 29, 2003 1:29:56 PM

People! you are forgetting genetics and biology. Sexual Dimorphysm means that the two sexes are halves or parts of a whole. Not separate at all. Any attempt to treat men or women as separate entities in any way is doomed to fail consensus.
Look at things from their origin, outward. If humans were hermaphrodite as a majority, this conversation would not exist, period.
Some ppl in this thread are thinking flexibilty, That's a great start. Now Plan for the organism, not for the individual.
People, WAKE UP. And PLLLEEAZZZE DON'T VOTE FOR BUSH unless you make more than $150,000 a year ; )

Posted by: Adolfo at Jan 27, 2004 2:50:56 PM

"girl vs boy..." Very interesting post!

Posted by: perekop at Mar 27, 2004 6:14:34 PM

Very interesting post .. , a grown-up discussion of an interesting point.

Posted by: Crimea at Dec 12, 2004 11:55:01 PM

Very interesting post -Yes
girl vs boy- No

Posted by: speculator at Dec 12, 2004 11:58:31 PM

The discussions on this site are really very interesting. I am a design studnet and currently involved in design of a mobile phone for Working WOMEN.
The most innovative idea i got is that we should stop thinking as women are different, but have an outward view.

Posted by: Pradeep at Dec 15, 2004 2:02:25 AM