Sexing the Machine

Three digital women debate gender, technology and the Net. An e-mail roundtable with authors Ellen Ullman and Sadie Plant.

Published September 11, 1997 4:55PM (EDT)

this three-way e-mail conversation about technology, the changes it's working on our lives and how those changes affect women in particular was held at the request of a national magazine. Later, the magazine decided the discussion would "go over the heads" of its readers. Perhaps its editors were also confused when so little of the conversation amounted to what we dubbed "whining at the gates" -- that is, the familiar complaint that women have been excluded from the world of high technology. To us, the debate about the role of computers in our lives has moved on from those early days.

Instead, we felt at home enough to question the more fundamental ways high technology is reshaping our world. If more and more women are practicing "multi-tasking" as a way of life, is that liberating or maddening? Do computers concentrate or decentralize authority? Does the World Wide Web give users more power or less? Can machines ever be considered "intelligent"? Our conversation might indeed go over the heads of the shrinking ranks of the resolutely unwired -- but for everyone else, we think it goes right to the heart of the matter.


Laura Miller, a senior editor at the Internet magazine Salon

Sadie Plant, a lecturer in cultural studies at the University of Birmingham and author of "Zeroes and Ones: The Matrix of Women and Machines" (Doubleday, 224 pages, $23.95)

Ellen Ullman, a software engineer and consultant since 1978 and the author of "Close to the Machine: Technophilia and Its Discontents" (City Lights Books, 189 pages, $21.95 -- October 1997)

1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1

Laura Miller: We've been asked why so few women seem to be among the movers, shakers and philosophers of high-tech industry and culture. I'll make the usual, obvious observations about how girls are socially and culturally steered away from mathematics at a fairly early age, and this leads to fewer women in computer science and engineering schools, and therefore fewer women in the community of people that high-tech industries draw their leaders and pioneers from. It's intellectually fashionable to see girls' lack of interest in math as innate, but the same thing was once said about athletics, and the new surge in women's and girls' sports shows how much that can change with a little encouragement and reforms like Title IX. We may see more women in these industries within the next 10 years.

I see another problem, though. The work culture of high tech is obsessive and single-minded, on both the creative and the business sides. When you're on a project, you work every waking hour, sometimes losing track of the time of day or the day of the week. Everything else falls away. It's a work style that's carried over from the computer science programs in universities. If we assume that women are generally more invested in having balanced lives, and are often responsible for taking care of kids, then they may be unwilling to drop everything else for work and as a result don't find high-tech careers tenable.

The culture of the Net is a slightly different matter, since participating in it doesn't make that kind of demand. But from its early years, the Net was populated with people -- mostly men, but some women -- from tech companies and the tech-related programs in universities, and it reflected their social inexperience. There was a lot of talk about Net culture being hostile to women early on, but that seems to have died out. There are a lot more women online now, and it's now clear that loutish behavior -- and the reaction of being hurt or offended by it -- are not the exclusive property of either gender.

2. Doing many things at once

1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1

Sadie Plant: We shouldn't assume, even now, that women turn their backs on high-tech careers because of the obstacles in their way, and we certainly shouldn't assume that programming a computer or even managing a corporation is the most desirable or socially significant thing to do.

As for getting wrapped up in work, I would have thought that this trait far predates computer science programs, and historically has been as closely associated with women as with men -- all those hours of knitting, for example. And while computing can demand this kind of engagement from individuals at particular times, the broader cultural effect of new technologies is to allow people to do several things at once.

It's also important not to start with the downside. Given the prejudice about women's ability with math and machines, the number -- and certainly the quality -- of women working in both the creative and business sides of new technologies might be said to be pretty high.

Ellen Ullman: Rather than asking why women are not more involved in computing, we might just as well ask why anybody enjoys watching a Net browser tell you how many bytes remain to be downloaded. I have been a programmer and software engineer for nearly 20 years, and the deformations of character that go into the making of a technical person are not necessarily salutary.

Programming (and using computers in general) involves a narrow-well horizon -- the world in a small frame -- and a tight, nervous, cycling energy. This strange energy can be alternately exhilarating or unpleasant, but the long-term effect is an abiding impatience. Once you get used to the machine cycling back at you, responding to you and only you, you come to hate it when it's slow. And the diffuse, ambling ways of normal humans can become unbearable. Even after hours and days of programming (or using the system all day at work), returning to the terminal after dinner is seductive. Here is where something, if not someone, can keep up with the strange, nervous energy you've gotten used to. If you are a reasonably sane woman with several friends and acquaintances -- someone who enjoys the ambling ways of normal humans -- it's not hard to understand why the PC in the study is not so alluring.

But I have to question Sadie's looking fondly on the Sherry Turkle-ish idea that "the broader cultural effect of new technologies is to allow people to do several things at once." I don't believe computers let us do "several things at once" quite in the way that, say, a mother can bounce a baby, talk on the phone, write a note and watch someone come up the walk. These motherly several things are done at various gradations of attention -- a human way to multi-task. But life around computers reproduces the operating-system version of multi-tasking: a time-slice model. One task is the foreground, and all others are rolled out. That the operating system seems to be doing several things at once is only an illusion, an adjustment of time-slices, a round-robin look at all the possible things to do, with each thing getting a nanosecond or two, and the rest swapped into the background. (Humans have six senses. A computer has one: the current instruction.)

Think about talking on the phone with someone who is sitting by a computer. One moment he or she is there, talking to you, present in a regular way. The next you sense a vagueness. Gone. The tack-tack of the keyboard is the dead giveaway: You have been swapped out.

My point is that the "culture" that has grown up around computers has tended to re-create the deeper technical solutions embedded in the systems we use. That this culture may be less than pleasing to women (and men) is no surprise. The technical solutions eventually used were not really planned by anybody. Engineering proceeds by tweaking and tinkering: One version follows another; features and errors accumulate; the next version fixes mistakes, adds new features, introduces new errors. It's really no use analyzing all this as a male cultural plot or a bane to females. In the end, from the standpoint of the engineer, a computer program has one and only one meaning: It works, more or less.

Laura Miller: To go back to Sadie's point about knitting: That's a highly interruptable activity, which programming (or, for that matter, writing) isn't. One of the things we're talking about here is concentration, and I tend to agree with Ellen that the celebration of multitasking (switching back and forth among functions or activities -- among applications really) among some academics strikes me as misguided. It has to do with deciding that it's wonderful for people to "realize" that their identities are not unified or singular, which is considered a Good Thing in the realm of theory. But in my experience, as a computer user and simply as a worker in a technically sophisticated environment, constantly switching back and forth between activities is exhausting and stressful. Everyone complains about it.

For me, the most satisfying work requires undivided concentration, in which case interruptions are an aggravation. I suspect that no woman caring for small children ever spent 12 hours in uninterrupted knitting, but it remains an activity that can be done half-unconsciously -- as Ellen says, with gradations of attention; you can talk to someone at the same time. And as Ellen's phone example points out, reading and talking (or at least really being present in a conversation) are mutually exclusive.

Sadie's right that we shouldn't assume that programming a computer or managing a corporation are the most desirable or socially significant things to do, but they are the sorts of work that lead to positions of power in the high-tech world.

Sadie Plant: I do take Ellen's point about multi-tasking (and I really don't take any kind of Turkle-ish position on this -- I think that kind of work is naive). It certainly is true that computers are currently serial systems which do one thing at a time, effectively reproducing the most bland, orthodox ways of working. But even at their worst, serial computers can do several very different things -- spreadsheets, communicate, make pictures, music, play games, etc. -- even if they currently have to do them one by one.

It's where things are going which really interests me. We've learned from the Net that issues which seemed so important only a few years -- or even months -- ago can quickly lose their pertinence. The near future of computing lies in parallel processing, or its equivalent in simultaneous processes, and this does begin to map onto more varied modes of activity. It may well be that a particular person will be absorbed in a particular process for a particular length of time quite regardless, but I do think that the differences computers make to the wider culture (by which I don't just mean computer workers, users or even people with an interest in computing, but the subtle, pervasive influences which developments like the Net have right across the board) are such that the possibility of doing more than one thing -- at least in a lifetime, if not in the moment -- is becoming increasingly viable and, indeed, necessary. True, anyone can spend their lives watching a browser count, just as they can spend them having babies or whatever. But the point is that there was a time when it was difficult to do anything but one thing. And, for women, that one thing was pretty much prescribed. Watching the counter may be addictive, but it's not compulsory.

It's also the case that the kind of concentration demanded by, say, computer programming, or even getting to grips with using particular software, is a little like learning to read, in that one's not learning to read a specific book, but gaining a skill which is extremely transferable and can open up a vast range of other possibilities.

Computing obviously does reproduce the old paradigms of work -- we know that all too well. What's crucial is to look out for the ways in which it changes the ways things are done.

I very much agree with the point about engineering going its own way -- indeed, this is precisely why getting into supposed positions of power isn't necessarily where it's at for either women or men. The whole notion that technology is intentionally developed as a human tool for human use -- i.e. that we know what it's for, that we're in control of it, and can predict the outcomes and effects it has -- all this is no longer obvious at all. And yes, the Net is an example of the extent to which the people who think they have all the cultural power can be taken by surprise. Machines, men, women -- we're all components of that vast engineering process known as reality.

3. The authoritarian net

1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1

Ellen Ullman: Sadie, you say that computing does reproduce the old paradigms of work but also changes the way things are done. I'd like to look a little further at that point, especially as it concerns the Internet.

I'm constantly amazed that the popular imagination has seized upon the Net as some harbinger of a brave, new freedom, or at least a metaphor for it. On the surface, the Net does seem democratic and multivalent -- the whole notion of "each user a publisher," and so forth. But I must say that -- internally, in the technical configuration -- the Net represents a return to the most restrictive, authoritarian model of early computing: the control of the central server.

When I watch users try the Web, it slowly becomes clear to me that this part of the Net represents the ultimate dumbing-down of the computer. Think about it. In your six-pound laptop resides more computing power than once fit on whole floors of buildings. But when it's connected to the Net, what is it doing? Waiting for each screen to come down the line.

I once told a consultant friend that I thought the Net was a plot by corporate central MIS to regain control of the desktop. He scoffed at first, but after working on the Net for a year or so, he called me and said, "You know, I think you're right. They just want to control what everybody has running."

Strangely (or not so strangely), the structure of the Net -- everything of interest on the server, the user waiting for a chance to talk -- reproduces the sort of computing system women have been working on for years as clerks. I'm exaggerating to make a point, but I can't resist saying it: The Net is a pink-collar terminal with prettier pictures and better sound.

Laura Miller: There's some truth to the idea that the Web limits the degree to which the user -- the surfer -- can shape or use the data at hand. It's also been championed as a medium that makes the Net accessible to all sorts of people -- many of whom are women -- who found telnet and FTP daunting and overly technical. It's a lot like the difference between the old command-line DOS OS and the Macintosh graphical user interface.

The Web has helped many new users leap over their impression that accessing the Net requires membership in a mandarin group of people with years of technical training and orientation. That's a democratic effect, but its price seems to have been that users' involvement with their computers is more passive. More people have access, but their experience is more limited and controlled. (Unless they're publishing on the Web, which is another matter entirely.)

Sadie Plant: I don't disagree at all with these comments about how the Net is actually developing -- like any emergent phenomenon, it's the ambivalent result of tendencies pulling in very different directions, some trying to maintain the old routines, others making new modes of connection and communication possible. The guy who said, "They just want to control what everybody has running" is absolutely right. But the question is, obviously, what can be done to counter this, or rather what tendencies are already in motion which do counter this?

It has to be remembered that insofar as there were deliberate intentions behind the development of computing in general and the Net in particular, they were solely concerned with command, surveillance and control. If this had worked, we'd be living in an absolutely authoritarian global state by now, and we're not. Again, yes, there are plenty of interests in this direction, but there are also potential and actual disruptions of such controlling tendencies. The Net is only interesting to the extent that it's part of these disruptions. If or when it becomes a corporate mall, then it won't be interesting at all.

I also agree with the point that there are many current attempts to confine users to preprogrammed packages. But this is a little like the supposed backlash against feminism -- not a sign of paralyzing defeat, but on the contrary, evidence that there is something there which needs to be contained, recuperated, confined, packaged and sold yet again. Power always wants to reproduce itself and its strategies -- the trick is to get in its way.

The notion that the Net is a brave free adventure is absolute bunk, and I've tried to ridicule this in my book. Any more thoughtful engagement with these issues makes it clear that the Net doesn't provide new freedoms for individuals at all; the whole notion of individuality is brought into question instead. Nor does it encourage democracy as we've known it in the modern West -- after all, our notions of democracy are crashing round our ears, partly because of the (again possible, potential, as well as actual) tendency of electronic communications to override national boundaries with huge implications for fiscal policy, trade, etc., but also because the ways in which technological changes creep up on us make it clear that we're not, and never have been, freely choosing the directions in which our cultures develop.

Ellen Ullman: Well, if I were nit-picking, I'd have to say that it's true that early computers were first used to calculate shell trajectories (still an engineering 101 assignment), but the Internet we are all trying to salvage was actually created in a fortuitous, serendipitous cooperation of government infrastructure project and individual free-for-all. The Department of Defense built it, then let scientists across the country talk to each other, more or less openly. Maybe this is why we are all having such a battle over what the Internet means and where it should go. For a brief moment, its very structure seemed to contain a delicate balance between collective and individual authority.

My biggest concern is the way computing culture is spreading outward. What began as a back-room operation created by a weird breed of human called the programmer is now a nearly intimate part of life in the developed world. Like it or not, programmers are re-creating the world. With every new version of an operating system, browser or application program goes an assumption about how human beings understand their existence and how they wish to organize it. These assumptions are not easily recognized or understood, yet they operate anyway, and powerfully. The explicit logic and structures of a computer system can't help but become implicit ways of being.

The only course, I think, is for us to understand the embedded logic and the demimonde from which it came. And then to ask ourselves, is this what we wish to be?

Laura Miller: We've taken a pretty wide detour from our initial topic. Is there something that really needs to be said about women and computers, something that doesn't apply equally to many men?

Sadie Plant: Here are some points I'd really like to see raised: There's nothing peculiarly masculine about new technologies, which, if anything, offer a great deal to women and everyone for whom access to information and means of communication were severely restricted in the past. This is not only the case in terms of the immediate functions of computers, but also because of the extent to which, for example, new conceptions of the human and its place in the world have emerged amidst these technological changes. The traditionally masculine notions of man as the author, creator, originator of "his" machines are eroded by the emergence of machine intelligence; as is the old insistence on definite boundaries between different genres, disciplines, modes of expression and art forms. There has also been a crucial shift away from the old notion of large-scale, top-down, centralized organization to the molecular, the microscopic and other microprocesses, and small-scale, piecemeal engineering. This too coincides with and encourages the undermining of -- or at least unprecedented pressure on -- all notions of centralized authority and, with them, the old patriarchal cultural structures which once made women the "second sex."

To confine discussions of women and technology to the obvious issues of who's making what and working where is to miss all this. It would also be good to reference the rich history of women's involvement in the emergence of the new technologies, not least because any women feeling they are coming to it late can see the extent to which they've always been involved. And also because it does any remaining boys who believe it's all in their hands good to know that a young woman (Ada Lovelace) was the first programmer.

4. A computer is not a metaphor

1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1

Ellen Ullman: Some of your points, Sadie, are very interesting metaphorically. But as the resident engineer, I feel I have to point out the obvious: that computers are actual objects doing quantifiable tasks, and a social, analytical, metaphorical understanding of them is not the same thing as their functioning existence.

First, "machine intelligence," which is somehow supposed to erode old notions. Put nakedly, there is no such thing as "machine intelligence." A computer and its program are realizations of the two things that make us most human: tool-making and language. Maybe that is why computers so fascinate us -- they are us to the nth degree. The "intelligence" of a computer is our attempt to understand and codify our own intelligence.

To give a concrete example: The game of chess between the grandmaster Kasparov and the IBM program Deep Blue was widely discussed in the press as a "man vs. machine" contest. But that was inaccurate. The more accurate view is that one man, Kasparov, was playing against an entire team of engineers and programmers, most of whom had studied Kasparov's past play and incorporated that knowledge into Deep Blue's program. (Kasparov was, in part, playing against himself.)

The contest was not between human and machine; it was not a contest at all. The game asked, What is the relationship between "natural" intelligence in biological practice -- its knowledge-ways not well understood, the question of how one knows left unaddressed -- and intelligence that is self-aware, codified, structured, where the issue of how one knows is at least as important (to computer scientists, more important) as the actions prompted by that knowledge. In other words, Kasparov knows how to play chess. Deep Blue poses the question: How does Kasparov know how to play chess?

I am going on about this point because I think it's crucial in overcoming a kind of idealism about computers which I fear quite acutely. "Machine intelligence" is not some sort of fast track to a better future. If anything, as an attempt to codify in structured ways what we only half understand about ourselves and our world, computing can just as well lead to the worst sort of reactionary thinking. Notions of "his" and "her" can change and develop; or they can be frozen into the yet more stereotyped personas assumed by people in chat rooms. We have to decide which way events will unfold. Machines won't save us from ourselves.

[Pause. Off the soapbox. Back on.]

Next, "piecemeal engineering." This is a perfect example of how a computing reality can seem to mean one thing and yet, in practice, mean something else altogether. Object-oriented programming has been touted as a sort of holy grail: It would make programming faster, it would let us reuse code, it would let us put bits of stuff all around networks. And now it's supposed to break down central authorities and, I presume, somehow bring down the patriarchy.

Sigh. Without even going into the technical problems (no one can really figure out how to test little pieces of code running everywhere at any time, for instance), I have to say that object-oriented programming eventually passes all processes through even a narrower funnel of central control: what is called the "object request broker," or similar registration and control authority. In other words, it won't do to see all this as only metaphor. Multiple processors need to be synchronized. Bits of code need to get in touch.

When we talk about computers, we are talking about the most authoritarian object you can imagine. Somewhere in any system there is a locus of control. I cannot see how such a thing will free anybody from anything.

Laura Miller: I tend to agree with Sadie that tech culture isn't inherently "masculine," but Ellen points out that the tools we use to create it do have certain mind-sets imbedded in their design that profoundly influence what we can do in that culture. I just don't see the potential problems as having a whole lot to do with gender, necessarily. Women can be just as invested in centralized control as men are.

What's frustrating for all of us about this discussion, I think, is that we're trying to deal with what I see as a diversionary issue. It's not about male culture vs. female culture -- it's about alive, imaginative, original culture vs. formulaic, predictable, dull culture. Uninteresting, conventional people are always the ones who most conform to "masculine" and "feminine" gender stereotypes. In the social world of the early Net, there was a predominant voice that was full of puffed-up bravado and excessively rationalistic -- all the stereotypes of what "men" are supposed to be like. And now that there are more women online, we find the hypersensitivity, passive-aggression and emotional navel-gazing that are the worst side of conventional femininity.

But the really interesting people -- male and female -- have always been around and have always operated outside those stereotypes; they're just fresh, engaging minds and personalities. That's what matters to me, when I'm looking for a culture to inhabit, virtually and actually.

By Laura Miller

Laura Miller is the author of "The Magician's Book: A Skeptic's Adventures in Narnia."

MORE FROM Laura Miller

Related Topics ------------------------------------------