(Un)dressing the interface

Brahnam, S., Karanikas, M., Weaver, M., 2011. (Un)dressing the interface: Exposing the foundational HCI metaphor “computer is woman.” Interacting with Computers 23, 401–412. doi:10.1016/j.intcom.2011.03.008

p.401

This focus on dialogue—interaction with a ‘‘second person’’ interface—has eroded the boundaries separating human beings from machines, calling into question the uniqueness of human intelligence and the sacredness of human personality and identity.

What both metaphors fail to acknowledge is that computers are not simply tools or personas. They are complex. In our interaction with them, they define who we are.

p.402

Gender norms, inscribed in the tools we use, define who we are as human beings.

Beneath the screen-based metaphors that cloak the interface are unspoken gendered subtexts that have the power to bind or to liberate. We propose to un/dress one of HCI’s screen- based metaphors, HCI is communication, that cloaks the interface.

Ultimately, we discover that an unacknowledged and invisible metaphor has served as the foundation for HCI is communication and continues to constrict the full range of roles and actions imaginable for both computers and the people interacting with them. This foundational metaphor is computer is woman.

Our goal is to render visible the foundational metaphor of computer is woman and to explore how this metaphor has served as an incredibly binding straitjacket that has limited the possibilities of both computers and women.

Harding (1991) and Rose (1994) have shown how much of women’s work has remained invisible because women have traditionally been given the role of looking after the body so that men can be free to pursue more cerebral activities.

Most of these studies, as Tympas et al. (2010) point out, focus on male–female relationships in computer advertisements, exploring for example, how women are often positioned looking up to men (and, indirectly, to computers).

We propose an alternative perspective: because electronic computers were introduced as the untiring mechanical counterparts of real female calculators this has limited the possibilities for both computers and women; both have been ‘‘feminized’’ in the pejorative sense.

Rather than treating gender as a biological ‘‘given,’’ we advocate approaching gender as an embodied performance.

Gendered metaphors have real consequences when they are used to design female agents who are subjected to abuse.

For Dourish (2004) the history of HCI is the story of how people have come to engage greater ranges of human skills (primarily cognitive and visual) when interacting with computers.

A feminist perspective, however, reveals that it is also a story that is enmeshed in and constricted by a vision that equates the work performed by computers (especially as computing in the 1950s moved outside the military and hard sciences into business) with the low-level clerical skills associated with women’s work.

p.403

By the late 19th century, computing was considered so simple that the astronomer Pickering is reported to have remarked one day that he could transform his maid, Williamina Fleming, into a computer. In 1876 she became the first of many female computers he hired in what later came to be called Pickering’s harem.

The choice to automate computing in the 1940s was prompted in part by a shortage of young women (Goldstine, 1972).

It was also the case that many of the sexual stereotypes that dispatched women into careers as human computers were capacities that could readily be given to the new computing machines. Attributes, such as patience, alertness, tirelessness, and precision, were often described in mechanical terms.

Steele (1943), editorial director of Vocational Guidance Research, a second glance: ‘‘It is generally agreed that women do well at painstaking, tedious work requiring patience and dexterity of hands. Also, women adapt themselves to repetitive jobs requiring constant alertness, nimble fingers and tireless wrists. They have the ability to work to precise tolerances, can detect variations of ten-thousandths of an inch, [and] can make careful adjustments at high speed with great accuracy.’’

Because computer was an occupation held by women, this coupling of women and electronic computers helped businessmen understand what the machines could do for them: electronic computers were the untiring mechanical counterparts of female calculators, only these machines were faster and more accurate.

In the minds of the computer engineers at this time, there was a clear division of labor in computing between ‘‘the headwork of the (male) scientist or ‘planner,’ and the handwork of the (largely female) ‘coder’’.

p.404

Only when it was realized that programming was creative, intellectually demanding, and valuable did men begin to take over the profession.

That the computer was built by men originally to do women’s work for them has constricted the vision of what it is possible for computers to do. 

They simply performed the massive, monotonous calculations demanded by the military, the government, and the hard sciences. Because these repetitive and mundane tasks were figured as women’s work, the capacities of computers were viewed as limited.

This close coupling of woman and computer is reflected from the 1960s to the 1980s in computer advertisements that were in- tended for business managers. In our reading of Hicks’ (2009, 2010) analysis of these advertisements, we observe computers and the women who operate them becoming nearly indistinguishable.

The ads at this time were also careful to separate the masculine profession of programmer from the feminine profession of operator.

Since the 1960s, the unspoken, invisible metaphor, computer is woman, has continued to influence popular perceptions of the computer, as well as interface design.

p.405

As Weiser (1991) has noted, there is no reason for computer development to have taken the path it did. Why, he wonders, do electronic computers not function more invisibly like the solenoids that control nearly all our everyday technologies? He surmises that one of the reasons computers took the path they did is the fact that computer technology was introduced first in our work environments. We surmise it is because computers were once women, and they still are, metaphorically.

Feminization, however, does not imply a move toward feminist positions. As Schell (2003) points out, ‘‘feminization’’ has a double-edged meaning: ‘‘it defines the work as both literally ‘female’ and ‘feminized’ in the pejorative sense’’ (p. 552). While women have populated the computer workforce, many of their contributions have been marginalized and in some cases even erased.

Only through analyzing the ways that women have been de/ constructed materially and metaphorically can we begin to challenge the master metaphor that has kept women in a feminized position.

The metaphor of computer is woman, based as it is on the Western definition of woman in terms of man, has served as nothing other than an incredibly binding straitjacket that has limited the possibilities of both computers and women.

[discussion of Plato and logos]

Logocentrism assumes that words have some specific/determinate (singular) meaning that does not change—a transparent relation between the signifier (language) and the referent that can be discerned through reason.

p.406

Haraway (1990) and Holbrook (1991) both emphasize that such professions are ‘‘feminized’’ regardless of whether the work is performed by men or women. This phenomenon, Miller (1991) explains, ‘‘codes the individual of either sex as a woman’’ (p. 139) and endows the individual with ‘‘qualities much like those of the mythologized mother: self-sacrifice, dedication, caring, and enormous capacities for untheorized attention to detail’’ (p. 46).

With the introduction of conversational agents and personified agents in HCI, this form of feminization is all the more visible. Believability has become tied to gender personification.

The classic ideal of woman as faithful servant is repackaged in images of female virtual assistants who are eulogized as doing woman’s work even better than women.

p407

JULIA was badgered repeatedly by ‘‘Brandi’’ to give him her clothes. Although his side of the conversation is not visible because he was ‘‘whispering to her,’’ her responses in this context suggest a violation:

Julia says, ‘‘Just try to take ‘em, buster!’’
􏰀 Julia says, ‘‘Help! Rape!’’
􏰀 Julia says, ‘‘No, I won’t.’’
􏰀 Julia disrobes fearfully and hands her clothes to Brandi (Foner, 1993, p. 12).31

The techno-body is simply another inflection of the self, no less real and certainly no less significant. What happens to the techno-body happens to the self.

p.408

‘What exactly, we wonder does it mean to be feminized?’’ Traschsel (1995) asks. ‘‘We are more comfortable, we sometimes decide, with the moral position and the language of victimization. To be feminized, then, is to be disempowered within the patriarchy, to be the object of systematic oppression. (p.26)

Computers have complicated our relationships with our bodies. We’ve listened too intently to the patriarchical messages we’ve been given.

In an effort to serve/help, women have attempted to imitate the images that we have been given about what it means to be female. But these stereotypical specifications are impossible to meet. As a result, we punish ourselves, and ‘‘we tend to accept lower salaries, poorer working conditions, lower status’’ (Harris, 1990, p. 20). Most of all, we’ve accepted shame.

Ironically, even though the Turing Test posits gender as two clearly defined categories, the very nature of the Turing Test hints that gender may be less a function of mirroring actual characteristics than it is a function of some constructed ideal.

p.409

Male HCI designers, in other words, are not simply dressing as females in the form of embodied agents. They are instead performing a pastiche, a copy of a copy, without an original. These men-in-drag overperform femininity in their design of female agents and ‘‘out-woman woman’’ (Butler, 1990, p. 132).

The more these supposed ‘‘norms’’ are repeated, the more that they take on what Baudrillard (1983) calls a ‘‘hyperreality.’’ This hyperreality before long becomes the criterion for the real itself, and the boundary between hyperreality and reality is erased.

Femininity is always considered drag, no matter who is doing it.

The more we disrobe personified agents, the more we see that beneath their feminine dress is the privileging of the phallus and the dismissal of the feminine.

These interfaces deny the lived experiences of real persons: the bodies of women who try to conform to the unrealistic standards of beauty and femininity embraced and promoted in these interface designs, and the gender-variant individuals in our society who are not willing to constitute themselves according to oppressive gender norms.

HCI metaphors, both spoken and unspoken, may influence human–computer interactions as well as how human beings treat each other. Unlike the insensate computer, the human user has a body and an embodied mind. We feel pleasure and pain. Unlike the metaphor, computer is woman, which can only figuratively be undressed, the embodied user can literally dress and be undressed, and may experience empowerment or abuse.

p.410

Similarly, as we develop personified agents, we must continue to guard against feminization as it exploits women who are ‘‘seen less as workers than as servers’’ (Haraway, 1991, p. 166) .

Personified interfaces, in presenting an embodiment of woman that is bodiless, yet interactive and conversational, trivialize the violence experienced in real life against our real bodies and our real selves.

Playing violent word games with JULIA and other personified interfaces trivializes crimes against women and perpetuates the denial of dysfunction in families and the culture at large.

We are concerned that personified agents, precisely because they are not sensate, encourage violence and aggression that can later be more readily extended, through this practice, to women in the flesh.

Once computers enter the social realm as autonomous social actors, speaking now for themselves, usurping the privileging of the masculine speaking subject, we observe an overwhelming anxiety about human identity that is tantamount to a sexual identity crisis.

Like homosexuals, gays, lesbians, and transsexuals, computers set into motion the massive patriarchal mowing machines enforcing heterosexual norms and punishing those who do not conform. Relief from this anxiety is purchased by playing drag games—by pegging computers into the second person (object) status of woman.

Personified interfaces (mostly female) are often punished, as are so many women, for both the ‘‘gross indecency’’ of pretending to be human and of failing to be properly feminine, as attested by the verbal aggression prevalent in the interaction logs of loquacious interfaces.

The assumptions that clothe the interface can bind or empower us. Especially in the creation of personified interfaces, gender demands reflection, for these animations both mirror and imperil fragile human flesh.

Advertisements