Response to Walker's article

By Warren Sack


Warren Sack's response to my essay on Online Caroline in Wardrip-Fruin and Harrigan's book First Person, MIT Press, forthcoming.


This essay is published as part of jill/txt.

ISSN: 1502-8003

Last modified
9 Oct 2002 12:13

Written and published with Tinderbox.

This essay is a response, by Warren Sack, to Jill Walker's essay "How I was Played by Online Caroline". Adrianne Wortzel also responded (not online), and Jill Walker wrote a capstone in response to the responses. The following is reproduced here with Warren Sack's permission, and will be published in First Person, MIT Press in 2003.

The founding essay of artificial intelligence (AI) is British mathematician Alan Turing's 1950 article for the journal Mind entitled "Computing Machinery and Intelligence." In the 1950 essay, Turing proposes a game to determine if a computer is intelligent. His game is based upon a parlor game. I will use Turing's words to describe the game and the role he thinks it could play in determining the answer to the question "Can machines think?"

The new form of the problem can be described in terms of a game which we call the "imitation game." It is played with three people, a man, a woman, and an interrogator who may be of either sex. The interrogator stays in a room apart from the other two. The object of the game for the interrogator is to determine which of the other two is the man and which is the woman.


It is [the man's] object in the game to try and cause [the interrogator] to make the wrong identification.


The object of the game for [the woman] is to help the interrogator.


We now ask the question, "What will happen when a machine takes the part of [the man] in this game?" Will the interrogator decide wrongly as often when the game is played like this as he does when the game is played between a man and a woman? These questions replace our original [question], "Can machines think?" (Turing, 1950, pp. 433-434)

Turing intended the man, in the original game, to play the role of a woman. The point of this being that both the man and the woman are suppose to try to convince the interrogator that they are the woman. In other words, Turing's original proposal was, essentially, to build a machine to function as a man pretending to be a woman.

The first question that Turing proposes the simulacrum might be given by the interrogator is as follows:

[The interrogator asks]: Will [you] please tell me the length of [your] hair?

It is the [simulacrum's] object in the game to try and cause [the interrogator] to make the wrong identification. [The simulacrum's] answer might therefore be

"My hair is shingled, and the longest strands are about nine inches long." (Turing, 1950, pp. 433-434)

Remarkably -- at least from the perspective of standard AI readings of the text -- the first thing that Turing posits with this question is the existence of a "virtual body" for the simulacrum. The "virtual body" invoked here is that of the simulacrum, i.e., either the machine or man pretending to be a woman and/or the woman playing the role of a woman.

I will argue that it is more than coincidence that Walker starts her essay with an analogous reference to her hair and, in general, her embodiment at the other end of a computer terminal. Walker starts her essay with this sentence: "My hair is still wet from the shower when I connect my computer to the network, sipping my morning coffee." With this sentence we -- the readers -- are convinced of Walker's embodiment. We now know that Walker is non-fictional, i.e., "real" while Caroline, according to Walker, "…isn't actually real. She's a fictional protagonist in a 24 part online drama called Online Caroline." Of course, what do we -- the readers of Walker's piece -- really know? Do we know Walker is real? How do we know she isn't just another Caroline in print form? We are, I believe, suppose to understand Walker's more concrete status as "proven" by the fact that she is embodied, she drinks coffee, she has hair.

If one has hair one is, necessarily, a mammal. Hair is a feature unique to mammals. But, humans are also characterized as those mammals that possess a small amount of hair. These are assertions of modern biology, but their history can be seen in, for example, Charles Darwin's 1871 book, The Descent of Man and selection in relation to sex. Darwin made extensive use of comparisons of hair and the attribute of hairiness to distinguish humans from animals, men from women, and one race from another. In popular advertising we learn that hairiness is a manly characteristic and near hairlessness an ideologically coded feminine ideal -- witness the many products for hair replacement for men and hair removal intended for women. Walker, according to her statement, can be understood as having enough hair to be characterized as embodied and thus human and "real," but not too much hair that she drifts out a certain category of primates and people (i.e., the presupposition of her statement "My hair is still wet from the shower..." is that her hair is not so thick that it does not eventually dry later in the day).

So, I argue, Turing's first question for his simulacrum and Walker's first sentence of her essay both very efficiently encode an understanding of embodiment that has been repeatedly used to distinguish the human from the non-human, women from men, etc. In general, a reference to hair encodes not just embodiment, but a very specific kind of biologically and culturally coded embodiment.

Within the literature of AI, discussions of Turing's imitation game have focused on the role of the machine and the role of the interrogator. The role of the woman has been almost entirely ignored. Yet, if one looks more closely at the woman's role in Turing's game, it is clear that variants of this role have been reiterated in popular art and performance for thousands of years. The woman's role, in Turing's game, is to compete with the simulacrum for an identity which is properly hers to begin with (i.e., the role of "woman").

These fears of an AI machine are, specifically, the fears of the "double" as it has been explored in psychoanalytic theory (e.g., Rank, 1971) and in the arts, for instance, in literature (e.g., The Sandman by E.T.A. Hoffmann, 1855) and film (e.g., The Student of Prague written by Hanns Heinz Ewers and directed by Paul Wegener, 1912). More generally, these fears can be described as those associated with the uncanny aesthetic discussed by Sigmund Freud (Freud, 1919) and others (e.g., Julia Kristeva, 1991; Anthony Vidler, 1992). Fears of the uncanny are often associated with machines, automata, and artificially produced "doubles." The frequently reiterated, popular fears surrounding AI and its cultural and specifically artistic precedents are the fears of the woman's role in Turing's imitation game, the fears of the uncanny: i.e., the fears of loss of identity, fears of replacement by machine, fears of disfiguration, dismemberment, and death (Sack, 1998).

In Walker's analysis of Online Caroline there are at least two women's roles: one for Caroline and one for Walker. But, is Walker's role obviously a woman's role? I would suggest not for two reasons. Or, rather, more specifically I would argue that Walker gets displaced out of a woman's role. Firstly, as an experiment, Walker enrolls herself to play Online Caroline both as a woman, Jill, and as a man, Jack: "I decided to see how Caroline would react to a different kind of reader. I started over, using a different email address and inventing a reader I called Jack, making him the opposite of my original character, who had been an honest rendition of my real self. But the emails barely changed, and Caroline's response to Jack the childless bachelor shows how changing a phrase needn't change the story at all…"

So, reading as a woman (Jill) compared to reading as a man (Jack) elicits from the Online Caroline site almost exactly the same story. But this comparison is a strictly formalist approach; i.e., it is an attempt to compare the story not according to variations in reception, but rather from variations in its form; e.g., the words and phrases contained in the story and their ordering. Is this formalist analysis of the difference in structure elicited by a male reader versus that structure elicited by the female reader enough to fairly distinguish a man's role from a woman's role? Walker doesn't answer this directly – at least not in this piece of her writing.

However, one of the luxuries of the Internet is that one can often quickly and easily read an author's other writings simply by plugging the author's name into a search engine. gave me the following fragment posted by Walker as a review of another article, "The Art of Being There," written by Lisbeth Klastrup. In her response to Klastrup Walker describes the inadequacy of a reading based purely on the structure of a piece and mentions specifically her own reading of Online Caroline:

jill skrev en kommentar den 25.06.2001 kl. 07:34:

"It is no longer the writer's prose, but her architecture of interaction that should be our object of study", skriver Lisbeth. Det blir ofte sånn. Når jeg prøver å skrive om Online Caroline lurer folk på hvorfor jeg bare skriver om strukturene og ikke gjør en tematisk analyse, som av en roman. I blant er det bare strukturen som er interessant, som Kevin skriver om Here to there. I blant er det tomt inni.

Strukturer kan være spennende, men det er også lett å rote seg inn i en strukturalistisk skog uten utveier, med uendelige stier som kan analyseres eller realiseres i nye former for narrativitet eller interaktivitet - uten at det er så mye mer der enn det at det er nytt.

Er det nok? Det er noe. Men ikke nok for meg….

Consequently, it is necessary to search for a second, non-structuralist reason to explain why Walker's role as a woman in reading Online Caroline is negated or displaced. A feminist analysis of reading as a woman can highlight the fact that the reading as a woman is not an act that is achieved purely through biological means, but is, rather, a difficult and even heroic process. Hegemonic means of narration and reception often constrain women to read like men and so reading as a woman can be a difficult act to achieve. Elaine Showalter points out the coercion implicit in various literatures: "Women are expected to identify with a masculine experience and perspective which is presented as the human one." (Showalter, 1971, p. 856). Laura Mulvey has made a similar observation viz-a-viz the viewing of Hollywood cinema (Mulvey, 1975). "Mulvey originally argued that if the spectator is a women, she has to assume the male position…" (Kaplan, 1992, p. 263).

My point is that it is not just the software that makes the reading of Online Caroline a similar experience for Jack and for Jill, it is also a host of other processes that have been insinuated, incorporated, and unconsciously invoked by all of us, both male and female.

That Walker has been constrained, even coerced, into a particular position, or, rather, through a particular sequence of movements is readily apparent. Walker draws an analogy between visiting Emanuel Vigeland's mausoleum and reading Online Caroline. "He designed the space to direct visitors' movements in several ways. The door is so low that I have to bow my head to enter. Inside the light is dim, it takes several minutes for my eyes to adjust…Online Caroline reminds me of this manipulation of the audience: it forces me to act in certain ways. The program behind Online Caroline doesn't just track my movements; it also makes me move." In other words, participating in Online Caroline is like being pushed into a mausoleum, a place of death and burial.

Walker is even more explicit about this in her final concluding comments in which she states "Reading, in Online Caroline, is being an accomplice to murder." It is like being an accomplice to murder because, we discover, Caroline is – by the end of the story – incapacitated to such an extent that she can no longer send email and is, thus, presumably dead.

Thus, Online Caroline is doubly uncanny in its reading by Walker. It is an uncanny narrative firstly because the protagonist, Caroline, suffers death and displacement. More succinctly, she is replaced: Walker states that email stops coming from Caroline and is replaced by an email sent by the president of XPT, Caroline's boyfriend's boss. Secondly, the uheimlich is manifest in Walker's own loss of position or, rather, what she identifies as a relationship "defined by its impotence." Walker is coerced into a reading position that renders her powerless, makes her feel guilty as an accomplice to murder. This, indeed, seems to fulfill the conditions of a perverse and narrow understanding of Turing's imitation game: the women have been displaced by the simulacra.

Yet, if such is the case, we must ask what sort of intelligence is incorporated in Online Caroline? This is a question that Walker answers graphically in a turn of phrase that pushes me to understand the "intelligence" of Online Caroline as a matter of physical force: "I'm not in charge of reading Online Caroline. I'm not a disinterested reader or viewer. I'm involved. This is a simulation, and simulations make their audiences participants: I bow my head as I enter, speak with a hushed voice, feel guilty at letting Caroline down. I'm the raw material for a simulation: it's carving itself into my flesh and my emotions. I'm being played."


Darwin, Charles. The Descent of Man, and Selection in Relation to Sex. New York, Humbolt, 1871.

Ewers, Hanns Heinz. Der Student von Prag. Directed by Paul Wegener. (Videocassette release of the 1912 German motion picture). Essex Films, 1980.

Freud, Sigmund. "The 'Uncanny'." In The Standard Edition of the Complete Psychological Works of Sigmund Freud, Volume XVII. Translated by James Strachey. London, 1919.

Hoffmann, Ernst Theodor Amadeus. Hoffmann's Strange Stories. Translated by L. Burnham. Boston, 1855.

Kaplan, E. Ann. "Feminist Criticism and Television." In Channels of Discourse, Reassembled: Television and Contemporary Criticism, Second Edition, Edited by Robert C. Allen. Chapel Hill, NC, 1992.

Kristeva, Julia. Strangers to Ourselves. Translated by Leon S. Roudiez. New York, 1991.

Mulvey, Laura. "Visual Pleasure and Narrative Cinema," Screen 16, number 3 (1975): 6-18.

Rank, Otto. The Double: A Psychoanalytic Study. Translated by Harry Tucker, Jr.. Chapel Hill, NC, 1971.

Sack, Warren. "Artificial Intelligence and Aesthetics," in The Encyclopedia of Aesthetics, Volume 1, Michael Kelly, editor-in-chief (New York: Oxford University Press, 1998)

Showalter, Elaine. "Women and the Literary Curriculum." College English, 32 (1971), 855-62.

Turing, Alan. "Computing Machinery and Intelligence," Mind, Volume LIX, No.

236 (1950): 433-460.

Vidler, Anthony. The Architectural Uncanny: Essays in the Modern Unhomely. Cambridge, MA, 1992.