From: randy@xanadu.com (Randy Farmer -- A survivor of the Lost Patrol) Subject: 2Cyberconf: An article Date: Fri, 10 May 91 17:00:03 GMT Organization: AMIX, The American Information Exchange The Second International Conference on Cyberspace: Literary Criticism Collides With Software Engineering by F. Randall Farmer This April saw the Second International Conference on Cyberspace; it was even more colorful and controversial than its predecessor. The collected abstracts listed 98 papers, covering a wide range of topics like implementation, representation, 'wiring up', AI, hermaneutics, artistry, religion, sex, fractals, cinema, anthropology, cychology (sic), ghosts, mummies, architecture, post-modernism, jazz, supercomputing, photorealism, dimensionality, space and time. Only 15 papers were actually presented. And, as you might expect, the content, style and state of preparation of the papers varied widely. Over half the presentations were given by software engineers about the cyberspaces they were building and what they learned from them. These talks were relatively clear, even if sometimes a little disorganized. Some of them contained technical material, often prefaced with the disclaimer "I'm sorry, but I'm going to get technical for a few minutes". I saw some eyes glaze over in the audience until the jargon was over. The remainder of the papers were presented by academics, in the traditional language of the literary critic, examining everything from cyberspace as master narrative to a character by character analysis of Gibson's Neuromancer trilogy. I'm certain these presentations were professional enough, and I truly believe that there were some points they were trying to get across, but, frankly, I couldn't figure out what they were. After talking with other software engineers, I discovered that I was not alone. The title of one of the papers helps to illustrate my confusion: "Cyberspace and the Proprioceptive Coherence: A Proposal." This sent me scrambling for my dictionary as soon as I got home. The language of literary criticism left me playing catch-up with the presenter, and falling three words further behind every paragraph. One programmer quipped that to his untrained ear these presentations sounded like "polysyllabic word salad." So, these two worlds collided due to confusions in purpose, language, and even in the definition of cyberspace. The software engineers were looking for information about where to go, and what to do next. I presume (and hope) that the literary critics were trying to bring artistic, literary, social, and humanistic concerns to cyberspace. It is clear that both groups will benefit from understanding the purpose of the other. But understanding the purpose is useless if the message itself is not also understood by the audience. I am one of the many software engineers in the audience who was bewildered by the language of the literary critics at this conference. Perhaps an explanation of how we think might shed some light on why. I'll use myself as an example. I am one of those lucky few who have actually implemented a cyberspace system and survived to tell the tale. Like many others, I have a few years of college, and lots of hands-on experience. Like many others, I don't spend much time studying the humanities or arts or reading the great French philosophers. My thought processes are instead dedicated to debugging. Debugging is usually defined as finding the failure points in a computer program, but software engineers also debug concepts and their implementations. Our emphasis is on finding an adequate initial design, and modifying it based on feedback until we get one that works--not a something perfect, just one that is functional. The advantage to this approach is that we can start working right away, and therefore have a working prototype done more quickly. Of course this also means that we are prone to make mistakes early on, and unlikely to get a solution that is optimal or even correct. In complicated systems, it is a fundamental reality that perfect solutions are a practical impossibility anyway. So like the scientist, we need gobs and gobs of input early on, to shape our systems, and help us improve them over time. Software engineers want input! This is very important to us because we are building cyberspace now. We want insights from people who are non-engineers: artists, psychologists, sociologists, economists, archaeologists, historians, and philosophers. This kind of communication is essential if cyberspace is fulfill its potential as a powerful medium for interpersonal communications instead of becoming just another rich boy's toy, sold to the wealthy consumer through places like "The Sharper Image Catalog." However undesirable we find this latter outcome, it is a very real possibility because cyberspace systems are consumer products: they want to be built, packaged, and shipped. As in the development of all consumer products, time is a most precious commodity. Time is so valuable that several well known cyberspace implementors have stopped attending conferences--except when they can be used as advertising vehicles--in favor of getting their systems built. This trend is likely to continue if the conferences don't offer something tangible. Presentations in the style of the literary critic aren't very tangible to us because the language used is not concrete enough for swift or accurate comprehension, extension or refutation. In short, software engineers can't debug literary criticism, so we don't get it. We can't even tell if there is any 'it' to get! Conferences are for sharing information and insights. They should be very important to the cyberspace researcher. It is this assertion that led me to write this article. But at this year's conference we didn't share very well. We collided with each other, confused in purpose and in language. So, given that software engineers debug systems, are busy building cyberspace now, are still making efforts to hear others' concerns, and given that literary critics are ready to offer their insights on how worlds work, how can we bridge this communications gap? Perhaps we could try using one or more of the tools that other conferences have found effective for dealing with these problems. The community could create 'Conference Submissions Guidelines' requiring clear statements, in plain language that avoids jargon, of both the paper's purpose and applicability to current or future cyberspace systems. The guidelines committee should encourage diversity: the request for clarity is intended to make papers understandable across disciplines, not to restrict the participants to a single style or approach. The chief drawback of this proposal is that it introduces the problems of a review process. Alternatively, the conference could split into a number of tracks. This would allow more papers to be presented, published, and would not require any standards of language. This would allow attendees to customize use of their time, but would not increase inter-disciplinary communications. It could also reduce the intimacy that the conference has enjoyed thus far. These measures are a matter for the cyberspace community to discuss and decide upon. To that end, I propose a multi-disciplinary panel for discussion of these and other suggestions the community may have. The Usenet newsgroup sci.virtual-worlds might well serve the purpose, considering both the origins of this conference and the wide dispersion of the participants. Last year, I was able to take at least some germ of an idea away from each and every presentation. Sadly, that was not the case this year. If this article touches the community in the way it was intended- -to encourage open and plain communications--I eagerly look forward to next year's conference in Montreal.