"The future is here, it's just not evenly distributed yet."
--William Gibson
March 2004: I originally wrote this paper for MacHack 2000, a computer conference in Detroit, Michigan. Chris Russ, the papers chair for the conference, had urged me to submit something. I puzzled over what I could write that would be philosophical but also of interest to educated non-philosophers. This was the result. In the period that followed the conference, the paper simmered slowly on the back burner. I revised it a couple of times and shopped it around, but never really dedicated myself to it. Now, I think that it is a relic. The term `Virtual Reality' was coined by Jeron Lanier circa 1989, and discussions of it in the 1990s had the reality-as-immersion conception that I argue against here. By now, however, the usage I advocate has become commonplace. I still think the arguments are fundamentally sound, but I suspect that they have become unobjectionable and banal.
Typical discussions of virtual reality (VR) fixate on technology for providing sensory stimulation of a certain kind. They thus fail to understand reality as the place wherein we live and work, misunderstanding it instead as merely a sort of presentation. The first half of the paper examines popular conceptions of VR. The most common conception is a shallow one according to which VR is a matter of simulating appearances. Yet there is, even in popular depictions, a second, more subtle conception according to which VR is a matter of facilitating new kinds of interaction. The latter half of the paper turns to questions about the contemporary technology of Internet chatrooms. The fact that chatrooms can be used in certain ways suggests something about the prospects for VR. The penultimate section asks whether chatrooms may legitimately be thought of as places. (In a sense, they may.) The final section asks whether cybersex may legitimately be thought of as sex. (Again, yes.) Chatroom technology thus provides an argument for the second conception of VR over its much ballyhooed rival.
1. Two conceptions of virtual reality
2. Talking in cyberspace
3. Talking plus
4. Conclusion
One commentator writes that virtual reailty (VR) "prides itself on being a computer-generated environment that can simulate the entire interactive experience, including its visual, audio, and tactile dimensions" [Kuj01]. This VR he personifies is not any technology we actually have, however. Using present technology, it would not be possible for a mad scientist to style himself a Cartesian demon and deceive you with a completely realistic virtual world. If VR has this aspiration, then the concept of VR cannot be explicated by considering its extension-- its extension is empty. So I propose to explicate VR by considering fictional portrayals. Although they are no doubt implausible forecasts of the future, popular science fiction programs provide evidence of how writers and audiences think about virtual reality. In fiction, we can find two conceptions of virtual reality: one according to which VR merely simulates interactive experience and another according to which VR fosters new kinds of interaction.
Talk of virtual reality conjures up images of a helmet and gloves that plunge the user into a computer-generated world, so it's tempting to say that the world the user is plunged into is virtually real. Here `virtually real' sounds similar to `virtually dry,' as in the sentence: "An hour ago my shirt was wet, but now it's virtually dry." This latter sentence means that the shirt is nearly dry, and that if it were any drier then it would really be dry. Yet it would be wrong to think that virtual reality is virtually real in this sense or that adding just a bit more reality to the helmet and gloves would make the computer-generated world really real. We make a wet shirt virtually dry, in effect, by moving it along a continuum of moistness from the wet end towards the dry end. To make it really dry, we push it a bit further along that continuum. Suppose we make a computer-rendered landscape `virtually real' by moving it along a continuum of convincingness. To make it really real, however, we would need to do an entirely different sort of thing. To make a real landscape, we'd have to go out into the world and build it.
VR systems like the Matrix, from 1999's science-fiction blockbuster of the same name, would provide an experience sufficiently like a real-world experience that people inside the system would accept it as real. Similarly, numerous Star Trek episodes have presented characters being caught in holodeck programs which, initially at least, passed for their real world.1 Whether it fools its users or not, the Matrix and the holodeck are not real worlds, and making them more or less convincing would not make the simulation any more or less a real world.
Although the popular goal for VR is to have technology that could pass for real, nobody actually wants to copy the real world in every detail. In The Matrix, the system builders of some future century designed VR that would convince the users that they were on late-20th-century Earth. In Star Trek, characters enter into holonovels, fight the battle of the Alamo, and watch old baseball games.2 In these stories, VR simulates places that can't actually be experienced.
Although it might seem real, any virtual reality you might imagine would not be real; whether or not it is able to deceive you into thinking it's real is beside the point. The intuitions of science fiction authors and hence, I suggest, of ordinary folk in the audience seem to accord with this observation. Consider a few examples from Star Trek: A crewman's interaction with a virtual woman is unable to sate his lust for a real woman. A tailor's claustrophobia is not calmed in the slightest by a virtual horizon.3
Computer constructs, while appearing for all the world like a woman or a seascape, cannot satisfy even the characters' subjective need for sex or open spaces. Although some holographic simulations in the show adopt a measure of reality, they do so only when they find a role in the non-holographic world: The holographic Doctor becomes real only by becoming a contributing member of the crew. The holographic lounge singer Vic Fontaine becomes real only when he works to better the lives of his flesh-and-blood friends and begins to live a life off stage.4 They become real only when they cease to be mere presentations. If we think about VR merely as a presentation that fools our perception, it's not clear that what it presents is even virtually real.
We may also approach the concept of VR by way of its other term, asking about the reality in virtual reality. In The Matrix, Morpheus asks: "What is `real'? How do you define `real'? If you're talking about what you can feel, what you can smell, what you can taste and see, then `real' is simply electrical signals interpreted by your brain." This suggests a view of reality whereby nothing is real except insofar as it impinges on our perception. If we think of reality in this way, then VR technology that can make it seem as if we are (for instance) eating roast beef is sufficient to produce a reality in which we are eating roast beef. Although this might put Bishop Berkeley or a young A.J. Ayer in an awkward position, it's a sort of vulgar phenomenalism that has fallen out of philosophical fashion.5
Later in the film, the traitor Cypher insists that the virtual world-- the Matrix-- can be more real than the physical world, since people in the Matrix can die there. Cypher raves, "I think the Matrix can be more real than this world. All I do is pull a plug here, but there you have to watch Apoc die." This suggests a funny criterion: Any place where you can die is a real place. On this criterion, the Matrix is real because people can live and die there-- reality is what has practical upshot and what holds a place in peoples' lives.6
These reflections on science fiction reveal two conceptions of VR: the shallow conception we anticipated at the outset on which VR is merely a show of reality convincing enough to trick us and deeper conception on which VR is new reality that plays a role in human lives. In these fictional worlds, it is VR in the deep sense that proves to be lasting and important. VR in the shallow sense would require technology years or decades beyond what we have now, but it would also miss the point. In the remaining sections, I will argue that technology we already have today provides VR in the latter, deeper sense and thus promises as much reality as we could ever hope to program into a holodeck. I will consider Internet chatrooms, first as places and then as fora for the chat activity of cybersex.7
Compare chatrooms to e-mail: E-mail is faster than mailing a letter, but it's not so much different in its power. Indeed, e-mail is rather like the postal network that bound together 19th-century Europe. Correspondents within England could trade overnight mail everyday throughout the week, and letters from as far away as St Petersburg could reach London in only two weeks. Historian Martin Rudwick comments: "Such speed and reliability gave the exchange of scientific correspondence an immediacy and vitality that it had never had in earlier generations and, arguably, that it has never had since" [Rud85,p. 36-7]. E-mail does offer something no prior postal network offered, in that a single message is sent at effectively no cost. Regardless, the advance of e-mail is only a difference in degree. It may be faster and cheaper, but it is still more or less just mail. We occasionally have pen pals whom we know only in a postal context, but it is far more usual to send mail only to people we have met in another social or commercial context-- either to someone who has given us their address or to a business that has made their address publicly available. People keep up with old friends by e-mail, but can meet and make new friends in chatrooms. If we enter a chatroom, there will probably be people there whom we don't know and with whom we would not interact were it not for the chatroom.
For this same reason, chatrooms are not like telephones either. Phone calls are made to one number with the expectation that there will be some particular person or organization at the other end. An analogy can be made between chatrooms and the telephone chat lines which used to be advertised late at night on UHF television in the midwest United States. The ads encouraged insomniacs to call in and talk with other people who had called in. With such chat lines, however, any noise that one user made was equally audible to any other user. In chatrooms, a user can issue commands to send private messages, query user information, or search other rooms at the site. When she does so, other users know only that she isn't saying something publicly. This allows users to carry on private conversations in the midst of the public conversation of the chatroom. Sometimes, of course, one interlocutor may communicate by private message while the other responds by public message. This can be frustrating for everyone else, since the rest of the room only sees half the conversation.
Although I have not provided an exhaustive argument, I want to suggest that chatrooms are disanalogous to other communication technologies in at least these two important respects. First, they encourage users' meeting and interacting with people whom they only know on-line. Second, they allow users to switch seamlessly between public and private messaging.
We speak of chatrooms using the usual place vocabulary, and thus it is tempting to think that chatrooms are places. A user enters, moves from room to room, and eventually leaves. When she sits down at her computer, she goes to the chatroom. Of course, she is still sitting at her computer. How can she both be in the chatroom and be sitting at home?
An obvious answer is to dismiss chatrooms by insisting that our use of place vocabulary to describe them is just an abuse of language. One may insist: Science says what is a place and what is not. The only real places are specifiable in the language of physics as points or regions of spacetime. Chatrooms are not in spacetime, so they are not places. Although one could use the word `place' in this scientistic way, it would have rather odd consequences.
Consider-- for the sake of concreteness-- the November, 2002 meeting of the Philosophy of Science Association.8 It was held at the Hyatt hotel in Milwaukee, Wisconsin. That is not in the spacetime language of physics, but academic conferences are familiar phenomena; surely the PSA was held someplace. One might hope to locate the hotel in spacetime by exploiting physical theory, cosmology, or whatever else.9 Suppose you did that. Would you include the lobby as part of the meeting? Some discussion spilled over into the lobby, so perhaps you would. What about the restaurant adjacent to the lobby? What about the front desk of the hotel, where people checking in were as likely to be in town on other business as to be conference participants?
It is natural to say that people off the street could come to the hotel without coming to the conference, so the PSA meeting can't be mapped cleanly onto any region of spacetime. Rather, it was a location in social space. One might note, of course, that the meeting did have a corresponding base of objects in spacetime even if the precise base can only be vaguely specified. The meeting rooms and all the rest were-- to put it loosely-- the physical basis for the meeting as a location in social space. Insofar as chatrooms lack such a physical base, one might think, they are not places. There are objects in the physical world that are preconditions for the existence of a chatroom: servers connected to networks, ISP's connected to phone systems, personal computers connected to keyboards, people sitting at desks typing, and so on. I could argue that this collection of objects makes up a physical basis of a chatroom in the same way that the hotel was the basis of the meeting.
One might still object that there is an important difference. The meeting's location in social space was strongly correlated with the location of the hotel in physical space; to go to the PSA you had to go to Milwaukee. Being near the server or a phone line, however, won't get you any closer to being in the chatroom. Being in the chatroom requires logging on, not simply sitting near certain pieces of equipment. Indeed, it seems natural to say that although the people are in the chatroom, the server and the phone lines are not.
All this is just to say that a chatroom is not a place in exactly the same sense as the meeting, but we have already seen that the meeting is not a place in exactly the same sense as (e.g.) a geosynchronous orbit. If we accept a distinction between physical and social space, we can admit that there are some places which have locations in both and some places that have a location in only one or the other.10 The meeting had a location in social space and, roughly speaking, a location in physical space. There are vast regions of deep space which, although specifiable as regions of spacetime, are not places in our social space at all. Chatrooms, I suggest, may best be thought of as places in social space with no clear analogue in physical space.
Whereas telephones may be thought of as connections between the places at either end, chatrooms become distinct places in social space because of the special features I noted earlier. Chatrooms not only allow interaction between people widely distributed in physical space, but foster the formation of communities which only meet in an abstract location in social space. They make for a kind of reality, then, because they make possible a kind of interaction not possible with other technology.
Of course, there is a sense in which cybersex is not sex. It is not actual intercourse-- the one does not physically penetrate the other. Although we call a range of things `sex' beyond straight intercourse, one might argue for this restricted usage on the grounds that straight intercourse is the sex act that allows for reproduction and reproduction is the natural function of sex. However, this would yield the perverse result that infertile men and women could not have `sex' even by straight intercourse, since their act could not fulfill its function of reproduction. It would allow that a couple might have `sex' merely by exchanging the requisite biological material through the post. I will contend without further argument that any definition so restrictive that it excludes cybersex peremptorily will be too restrictive to capture common usage.
Let's consider the definition that Thomas Nagel offers in `Sexual Perversion' [Nag79]. He asks us to imagine two people, whom he calls Romeo and Juliet. Romeo sees Juliet and is sexually aroused by the sight of her. Juliet looks back at Romeo and is aroused not only by the sight of an aroused man, but also by the very fact that she is what has aroused him. Romeo, in turn, is aroused by the fact that Juliet is aroused by his arousal. This process of feedback and mutual arousal advances to states of heightened arousal that become harder and harder to express: Romeo is aroused by Juliet's arousal at his arousal at her arousal at his arousal, and so on. Such escalation, Nagel contends, is at the heart of sexual experience. He writes:
Physical contact and intercourse are natural extensions of this complicated visual exchange... . Ordinarily, of course, things happen in a less orderly fashion-- sometimes in a great rush-- but I believe that some version of this overlapping system of distinct sexual perceptions and interactions is the basic framework of any full-fledged sexual relation and that relations involving only part of the complex are significantly incomplete. [Nag79,p. 46]Nagel explicitly considers visual and tactile exchange, but observes rightly that this is a general schema.
Imagine Romeo and Juliet are sitting in their respective apartments at their computers. They enter a chatroom, exchange pleasantries, and begin to chat. At some point in that conversation, Juliet says something risqué which arouses Romeo. Romeo gives a clever reply which not only arouses Juliet but makes her aware that Romeo was aroused by her original comment. She is aroused, we may say, on two levels: first by Romeo's reply, second by the fact that she has aroused Romeo. The conversation continues with Juliet's sultry rejoinder, and so on. They have thus found in a chatroom the escalating mutual awareness that Nagel argues is characteristic of sexual relations. Although this is true whether Romeo and Juliet pursue the encounter to climax or not, I think it is reasonable to argue that whatever they are doing with their hands besides typing is a natural extension of their on-screen interaction.
Cybersex is not, as one pseudonymous interlocutor suggested to me in a chatroom, "just talking." Neither is it, as she conceded a moment later, "just talking and masturbation." It may be those things, but is not just those things. The `talking' in cybersex functions in a way that talking in ordinary chat situations does not. It exhibits the complex psychological structure exhibited by real, in-the-flesh sex. The `masturbation' in cybersex functions in a way that the masturbation in response to a pornographic web-site does not. The arousal when viewing porn is purely voyeuristic, while the arousal in cybersex is mutual. The former is in response to a picture, while the latter is in response to a person.12
One may resist the conclusion that cybersex is sex in a number of ways. Perhaps people do not interact during cybersex as we have imagined Romeo and Juliet interacting. Instead, they lie, dissimulate, and confabulate. This paper is not an anthropological study of chatrooms, I have said nothing robust about what goes on in them, and there is no denying that it is easier to fake an orgasm on-line than in person. Nevertheless, the conclusion that cybersex could be sex doesn't rely on the fact that people usually experience the full range of mutual arousal when they interact in adult chatrooms. It need only be a phenomenological possibility. If the full-fledged interaction is both harder to establish and easier to fake on-line, then this may make cybersex on average less satisfying than in-person sex, but it ought not obscure the fact that cybersex could attain that level of complexity.13
Another response is to note that Nagel has only proposed a necessary condition for full-fledged sexual relations. Even if he is right, there may be other things that are also required. I have no principled objection to such a move, but I cannot for my part imagine what the additional criteria would be. One could attempt to confine Nagel's schema to particular modalities of interaction, but this is a nonstarter. If we said that the escalating awareness must involve a visual component, for instance, then it would be impossible to have sex in the dark.14 Even if there were some further criteria which excluded cybersex from being a full-fledged sexual relation, it is important to note how cybersex got so close in the first place. It was not because it provides physical stimulation that resembles sex in any way. If physical stimulation were the criterion, cybersex would just be talking and masturbation. Instead, the critical feature of cybersex is that it establishes a particular kind of relation between people.
There are attempts to push the teledildonic envelope by engineering computer interfaces and network clients for sex toys. As a consequence of such gizmos, one commentator suggests, "the definition of sex has become a whole lot more complicated" [Kad99]. However, these complications arise even without new technology. Perhaps more intimate hardware will make cybersex easier or more satisfying, but it will not change the nature of the beast. Cybersex is sex, if it is sex at all, not because it provides stimulation like what can be had in person but because it is the right form of human interaction. Similarly: chatrooms are places, if they are places at all, not because they exhibit the geometry of physical space but because they are places where people can go.
Of course, it is in part only a terminological dispute as to whether the label `VR' should be applied to technology that impersonates reality or technology that creates new reality. The point here is about what we should expect technology to do and what we should think when it's done it. Call it by whatever name you like, but the real potential of VR is not merely the power to clone reality. Instead, new technology will make new reality as it finds a place in peoples' lives, as it empowers them in new ways, and as it offers them new possibilities. I have argued that technology we have today is already doing that. Virtual reality in this important sense is not something to wait for in the 25th century; it is something we have already brought with us into the 21st.
* <http://www.fecundity.com/job> I'd like to thank L. Bridget Magnus and Ryan Hickerson for helpful feedback on earlier drafts.
1Holodecks first appeared in the first season Star Trek: The Next Generation episode "The Big Goodbye"-- they have been a recurring feature in the subsequent series Star Trek: Deep Space 9 (DS9) and Star Trek: Voyager. The holodeck is closer to reality than the Matrix, in that space in the simulation corresponds to real space on the holodeck and peoples' bodies in the simulation are their own bodies. That caveat aside, the Matrix and the holodeck provide similar barometers of intuitions about VR.
2And the Cartesian demon is presumed to misrepresent the world. Would it have mattered to Descartes if there were a benign genie who made everything appear exactly as it actually was?
3Respectively, the third season Voyager episode "Blood Fever" and the seventh season DS9 episode "Afterimage."
4The characters are from Voyager and DS9, respectively.
5On a typical retelling, phenomenalism received a sound beating from Sellars [Sel56] and Austin [Aus62]. One recent retelling of the demise of phenomenalism is given by Psillos [Psi99]. The reader is invited to recall their own reasons for not being a phenomenalist.
6The movie may be seen as posing the two views without taking sides: Morpheus proposes the phenomenalist view, but does not endorse it. Cypher chooses the phenomenal allure of the Matrix over the difficult life outside it, so his suggestion might be mere provocation or irony. The conflict of the movie is resolved with kung fu action rather than with philosophy.
7Although I use the phrase chatroom in what follows, the lessons apply equally well to IRC channels, MUD's, and other on-line chat environments.
8One may worry that the PSA was an event rather than a place. Since it is not an instantaneous event, this worry would be misplaced. Regions of spacetime (considered in the reference frame of some observer) and uncontroversial places (like Times Square) also have temporal boundaries.
9At least since Strawson [Str59], it has been clear that places cannot be picked out merely by description-- there is an essential indexical element in picking out here and now. Set that concern aside for the nonce.
10Strawson [Str59] argues that all reference piggybacks on reference to physical places. I'm not sure this is true for reference to people met in chatrooms, but it doesn't bankrupt talk of social space in any case. Once we can reference people, we can talk about their contexts of social interaction.
11Although he says that people are most excited by the idea of teledildonics and electronic LSD, Rheingold argues that the greatest near-term value of VR research will be for training and for operations in hostile or restrictive environments. Although training simulations should mimic reality as closely as possible (and thus require VR in the shallow sense), this application is the exception rather than the rule. The objective of teleöperations is to accomplish a task at a distance. The operator should be given the best and most efficient interface for accomplishing the task, which will often mean representing the data and allowing him to manipulate the remotely controlled device in ways which make the experience unlike being there.
12This presumption about pornography might be contested. Kibby and Costello write that the consumption of pornography depends on "a belief in the `real person' beyond the image, with the possibility of a connection" [KC01,p. 357]. Even if they are right, this belief when confronting pornography is necessarily false. In a case of interaction with a real person (without scare quotes) the possibility of connection is genuine.
13Interestingly, a study of cybersex observes that partners who met on-line "got to know each other much better and quicker, based on actual characteristics" [KW98,p. 448].
14Goldman [Gol77] suggests that the necessary modality is tactile.