In Life on the Screen, Sherry Turkle provides a first-hand report on the uses being made of the computer world, particularly by those who become absorbed in MUDs, or multiple-user-dimensions. She shows how users are able to develop alternative identities, playing at roles that may extend to changing gender, and they build imaginary spaces that they can inhabit on screen. Her evaluation of these phenomena is judicious, alert to both the problems and the possibilities of this medium, acknowledging both its addictive dangers and its liberating and transforming powers. Her discussion is set within a helpful perspective, that of the development of computers and computer networks over the last twenty to thirty years and their influence on thinking about the nature of the mind, culture, and politics. For example, she shows how the early phases of cognitive science and artificial intelligence grew out of a commitment to the information processing capabilities of the computers of the 1950s and 60s, and how this has been replaced by connectionist and network models, with their apparently more humanizing metaphors for the mind.

At the same time, there are perhaps two underlying limitations in the book. First, while she often acknowledges the influence of postmodern conceptions, referring to writers such as Foucault and Baudrillard, she never seriously considers postmodern thought or seems fully to grasp its implications. While she mentions simulation frequently, she does not engage critically with the possibility posed by Baudrillard of culture being now a tissue of simulation and nothing else. Thus the theme of the "convergence" of computing with postmodern conceptions, which occurs frequently in her book, is never more than a loose and sometimes misleading analogy. Second, her appeals to the sciences of psychology and biology, which inform her ideas about identity, usually draw on half-understood ideas that seem to derive more from popularizations of these sciences than any sustained thought about their implications. Thus on both grounds the notion of identity central to the book tends to be unreliable or vague. As a witness of the experience of modern computing, however, Turkle provides a rich and suggestive archive of reports.

The postmodern theme is struck early in the book. The new world of the internet and MUDs is said to evoke ideas that can be characterized as "decentered," "fluid," "nonlinear," and "opaque," all suggestive of postmodern debate. This is contrasted with the way of thinking called "modernism" that has existed since the Enlightenment (p. 17). Another way of putting it is that "we are moving from a modernist culture of calculation toward a postmodernist culture of simulation" (p. 20). While this offers a useful approximation for the issues Turkle proposes to discuss in the book, the first term "modernism" is not, of course, characteristic of discussions of literary history, where modernism appears around 1900 (in fiction with James, Conrad, Woolf, Lawrence, etc., and in poetry with Eliot, Pound, etc.). If the Enlightenment was (among other things) a culture of calculation, a view supported by Blake's critique of Newton and Locke, then the Romantic movement (c. 1790-1830) comes to challenge it long before postmodernism does. (Let it also be noted that the term "romantic" is used freely by Turkle, but not in reference to the ideas of the Romantic writers: e.g., "the romantic reaction to information processing": p. 147.)

Turkle positions the rise of cognitive science in the 1960s, motivated in particular by the computer models of the day, as central to the "modernist" culture of calculation (p. 128), an insistence on the "information processing" model of the mind. The shift to a postmodernist paradigm comes with the rise of connectionism, the theory that information is distributed in networks across the brain. Thus brains, and the computers programmed to simulate connectionism, are now seen as "nondeterministic," "spontaneous," and "nonprogrammed" (p. 133). Although connectionism has developed largely in independence of empirical psychological study, its perspective gains some validation from theories such as that of Antonio Damasio, the neuropsychologist, who has proposed multiple "convergence zones" as the primary locus of brain activity. "Emergent" programming on computers shows how complex behaviour can emerge from a few simple rules (p. 166), analogous to ant or termite colonies, where intelligence is distributed across thousands of separate and semi-autonomous organisms. Again, there is now wide acceptance among neuropsychologists that the mind has no central executive (or homunculus) but rather an array of mental modules, which communicate with each other to the degree necessary to co-ordinate behaviour.

See Damasio, 1989.

mental modules: see Hirschfeld and Gelman, 1994.

Thus, contemplating recent developments in computing, Turkle can talk of a "postmodern convergence" (p. 136) -- echoing George Landow (although apparently unaware of his work). If the postmodern world is one of surface only, then our explorations of it call for metaphors of navigation rather than of depth (p. 47-8) -- a reminder of Landow's construal of Roland Barthes. I will examine the postmodern claim in relation to three major themes of Turkle's book: simulation, animation, and identity. In each case I will question whether a part of what appears to be new in Turkle's account is actually based on long-standing human capacities or tendencies. To the extent that we inhabit a life "on the screen," have we adapted to this as readily as we have because it offers another avenue for behaving in familiar ways?

Although the present currency of "simulation" is due in particular to Baudrillard's influence, Turkle seems uninterested in the postmodern notion that simulation is ubiquitous. She is happy to distinguish RL (real life) from life on the MUD in many of her reports on users (people who also seem to find no difficulty in making this distinction). Thus it is worth considering how far the issues that Turkle raises about MUD simulations are also raised by earlier forms of simulation. When we watch a stage play, read a novel, or see a movie we are also caught up in a simulation, perhaps involving some of the same challenges to identity that Turkle invites us to consider. I certainly remember as an adolescent taking on the identity of the hero of films that I saw as I walked down the street from the cinema. Coleridge's well-known phrase "the willing suspension of disbelief" refers to the state of accepting, for the time being, the scenario of a fictional or dramatic work as though it were real. If I become absorbed in the fate of Pip as I read Great Expectations, I take on, or try out, his (fictional) identity as though it were my own, with a resultant "decentering" of my everyday self (whatever that may be). So what is new about the challenge to identity on the internet? The fantasized others on a MUD answer my statements and respond in real time, unlike the characters in a novel, introducing a degree of indeterminacy at the level of the unfolding "text." But as Pip in this novel (that I am reading for the first time) do I not experience the same degree of indeterminacy from the characters I engage with? Their responses are equally unpredictable, even though they are fixed on the page; and I have given them credit for their existence, just as I "believe in" the characters I encounter on the MUD, although some may be bots, and others are men acting as women, or vice versa.

disbelief: Coleridge, 1983, Vol 2,
p. 6.

Perhaps the chief difference between MUD and novel lies not in the quality of the experience as such, which is captured by the concept of simulation, but in the processes of understanding that occur. A novel -- at least, a significant literary novel such as that of Dickens -- is designed to effect a shift in the reader's understanding. Dickens invites us to experience the world of nineteenth-century London through Pip in order to raise troubling questions about the role of class and money. If a shift in understanding occurs on a MUD, this is more likely to be fortuitous: Stewart failed to experience it (p. 198), whereas Ava learnt to cope with disability and to love again (p. 262-3). In this instance, virtuality can be a "transitional space," in Turkle's words (p. 263). But this possibility is also available to those who open themselves to change through engaging the virtual world of a poem, novel, or play.

Our readiness to engage in simulations is perhaps due to the provisional and flexible nature of our consciousness. As the neuropsychologist Llinas argues, perhaps only slightly overstating the case, consciousness itself is a continuous illusion: "perception is a dream modulated by sensory input" (Llinas, p. 351). Perhaps this facilitates our "culture of simulation," as Turkle calls it, fostering our liability to addiction from television, internet surfing, and the like (p. 235). Do the evils of simulation come, as her footnote hints, from our tendency to separate mind from body, which has been endemic for 2500 years or more in the thinking of philosophers from Plato onwards (p. 315 n6)? As she points out, the edited world of simulation, such as television, may serve to devalue the real world, especially for children. To observe life in the wild cannot be as exciting as film footage of the wild shown on television (p. 237). But in this respect "life on the screen" seems to offer little that is entirely new: the medium does not determine what we do with it; and simulation (contra Baudrillard) is not all. The postmodern world may at times seem too thick with overlapping simulations, an impenetrable hall of mirrors (p. 72). But this too is an illusion: the body still breathes; we must still ingest food; our immune system is busy defending us against invasive germs. At any time our embodied nature can break through the artifice of our simulations and reassert itself. The postmodernist will argue that, even so, we have only representations for construing the body and its processes, so we don't escape the world of simulations. But the psychologist will counter that the forms of thought we bring to bear are shaped by evolutionary history and adapted to the realities of the body and its environment. Thus, although our thinking may not correspond point to point with reality, it must approximate reality closely enough to have permitted the survival of our species.

Further examples of this "natural" endowment which underlies our capacity for simulation are raised by the question of prosthetics. Discussion of virtual reality (the helmeted, gloved variety), or MUDs, might lead us to overestimate the significance of "cyborg prosthetics." Thus Turkle points to the extensions to our self understanding and capabilities offered by machines: MRI scanners, for example, enable us, as she puts it, to "read our emotions" (p. 177). But this confuses the possibilities that are available to us: we may be able to locate an emotion through a brain scan, but this gives no information about what the emotion is about; an attentive reading of a person's face and body language is more likely to offer that. And this is made possible in part by our instinctive and immediate ability to simulate another's emotions (the capacity we normally term empathy), a vital skill in a social environment where our effectiveness, even our survival, may depend on accurately assessing the emotional state of the people around us. Then think of what the body can achieve with a simple stick as an extension of the hand: as you probe, it is as though your tactile nerves are no longer in your hand and fingers but at the end of the stick. If a stick is a form of artificial intelligence, it is one that has been available to us for a million years. Virtual reality builds on those same capabilities, but is it a new dimension?

One tendency highlighted often in Turkle's book is to think of ourselves as more machine-like, "plugged-in technobodies" (p. 177) -- halfway to the simulations of Gibson's science fiction novel Neuromancer. Turkle asks us to consider how far, with recent developments in understanding of genetics and the use of therapeutic drugs to alter mind states, we have learned to see the human being as a machine. If we can reengineer a biological organism through tinkering with its DNA (perhaps eliminating some human diseases in this way soon), does this mean that we are "programmed beings"? (p. 25). Are we just another piece of digital code? In this respect, Turkle seems to suggest, like other writers on electronic media, that a simulation possesses agency and is an end in itself (the gene controls me, not I the gene). This is too simple, however, since no code interprets itself. In the phrase of the biologist Jesper Hoffmeyer, we require "code duality," that is, an entity that can read the code (whether the code is DNA, or the lines of a program in a computer). A code refers to something for somebody: whether we deal with computer codes or the codes disseminated by television, we still choose the framework within which a given code operates and confers meaning.

The argument over DNA and other codes is fuelled by our tendency to animation, another important theme of Turkle's book. This is well illustrated by Turkle's account of how young children view computers. For instance, if a program enables the computer to cheat at a game then, in the eyes of the children observed by Turkle, it is alive (p. 80). As Turkle suggests, with the advent of computers, the child's criterion for determining what is alive seems to have shifted from motion to thought or intentionality (the computer, says the child, "meant to cheat"). This offers another example of the "theory of mind" that psychologists now see as central to children: the ability (developed in infancy) to attribute thoughts and intentions to others. Young children animate objects too inclusively (the sun; a cloud: p. 282, n3), but then, so does the adult who refers to a car or boat as "she," or attributes malevolence to a faulty computer. Endowing objects in the world with animation (inappropriately or not) is the oldest and most persistant mode of understanding, and one that clearly has adaptive value in the ancestral environment. Like Pascal's wager about the existence of God, the price of conferring animation is less than the price of not animating if you should turn out to be mistaken (that patch of stripes is a tiger, after all). In fact, animation is pervasive: for example, modern advertising draws upon this tendency frequently, attributing human feelings to cars that encourage us to assimilate them to our need for companionship. Similarly, hypertext theorists characteristically attribute agency to language, as Grusin points out. Whether this means we are "trivializing human life" in so doing, as Wiezenbaum thought when users overestimated the powers of his therapy simulation ELIZA (p. 105), depends on what you believe it means to be human: animation is so important (abused by advertisers as it may be), that it can be seen as a defining quality of human nature. It endows us with "objects to think with," to borrow Turkle's useful phrase (p. 260).

However, since postmodern theory has set its face against the assumption that human thought interacts in any direct sense with nature, there is also an interesting paradox here. On the one hand, a New Historicist such as Alan Liu (to cite one known to me from Romantic studies) proposes that our feelings are determined by our local culture, making "nature" a construction of the prevailing ideology; sociologists and anthropologists have argued that emotions are socially constructed, that is, they are determined by our local culture and bear no relation to some supposed universal or "natural" process. On the other hand, neuropsychologists have recently asserted that our forms of feeling are comparable in structure and complexity to those of animals. Jaak Panksepp writes: "The weight of existing evidence suggests that all mammals share essentially identical primary process emotional circuits, subcortically situated" (Panksepp, p. 63); thus "anthopomorphism may be a beneficial guide to empirical inquiries as opposed to the mentalistic sin it has always been deemed to be" (p. 60). In this perspective, animation allows direct insight into nature, and, indeed, constitutes a mode of inquiry for productive research. It is thus inevitable that animation should form a primary mode of construing machines, once these have become complex enough to support it (that is, I don't animate my electric pencil sharpener, but I sometimes animate my computer). But the implications of this tendency are the reverse of those feared by Turkle (e.g., pp. 265-6): rather than reductively assimilating the human to the computer, this elevates the computer to the human, a participant in that endemic tendency to endow with human meaning the environment around us that is the material of our thoughts and the stage of our actions -- a tendency that Panksepp's claim would authorize, if our emotions resonate in parallel with the world we are attempting to understand. The danger of reading too much into the object in front of us is present with an animal as much as it is with a computer.

Liu, p. 104.

social construction of
emotions: e.g., Geertz, p. 50; Harré.

Panksepp, 1991.

The third theme of this book is identity. Turkle raises the possibility that as we engage with the computer and other electronic media we have become a collocation of multiple identities, or at least (a less novel proposal) that we play a variety of roles (p. 180), behind which no central, organizing self exists. Turkle's discussion falls well short of the latest postmodern movement to consider identity itself as outmoded, since it has been the basis of oppression; perhaps we are now in a post-identity world, as it has been called. At the same time, Turkle is still able to appeal to Erikson's model of a "core self" (p. 203), and suggests that role playing on a MUD may facilitate its development. While Turkle seems unable to decide where she stands on this issue, her discussion of the addictive qualities of MUDS implies that her beliefs are closer to Erikson. Discussing the example of Stewart who invested much time developing an alternate personality through a MUD, she notes that Stewart felt it had failed to develop his real self in any way (p. 198). This seems to suggest a core self, resistant to whatever therapeutic values the MUD had to offer: in a word, identity is not easily moulded or transformed by our computer extensions. Those who are ready to change will be able to do so through some appropriate medium, but on the evidence she provides a MUD is probably neither unique nor a notably effective medium.

How did people make such changes before the computer? A hundred years ago the catalyst might have been joining a church, or taking up bicycle riding, or attending a university. And just as Stewart failed to develop because, in Turkle's useful distinction, he merely "acted out" his problems on the MUD instead of "working through" them (p. 200), so joining a church might often have failed to resolve the problems of those who turned to it. An analogy with reading comes readily to mind: perhaps acting out is offered more readily by fantasy fiction (romances, westerns); working through is more likely to be a consequence of reading literary novels which challenge our understanding. Since Stewart appears to have behaved in ways familiar to him off screen as well as on, the MUD evidently failed to simulate a new identity that he could then learn to inhabit in the real world; his ability to empathize with (i.e., to animate) the other characters he encountered seems to have been limited or non-existent. Stuck with his existing self, he remained unreconstructed by the machine to which he had devoted so much time (p. 198).

Whether this speaks to the endurance of a core self is unclear, but one issue this narrative raises is what agent motivates shifts in identity, once we see that the mere identity games played on the MUD offer no guarantee of development. Since we bring to this situation an existing and complex set of psychological processes, which are apparently not so readily set aside or moulded by the computer, what process enables changes in identity? Most therapists and psychoanalysts would nominate emotion. An emotion appears to represent an aspect of the self that we postulate: it embodies a potential character in the process of being experienced, created, or resisted -- which is why emotions feel important, as though the self is at stake in how we express or handle them. The MUD world and other computer-based self-representations are built on this preexisting substratum in which the self is already and necessarily in flux. Thus those who develop new identities, such as Ava, are able to use the MUD to experience their emotions in a new and ultimately constructive setting. Ava, as Turkle puts it, "found a way to love her own virtual body" (p. 263). Once again, it seems that when the computer is most radical, it is drawing on and helping to shape long-standing and central features of our human endowment.