This particular e-mail discussion has not been a very successful event, in spite of the quality of the contributions. It attracted only a small number of participants and never seemed to take on a life of its own, eventually fading away to silence. It seems worth trying to analyse some of the reasons for this.
First, if there is any blame, it must of course rest with me. At a critical point I was unable to access the e-mail network regularly, and this is when my intervention was probably most needed. Evidently there is more to making an electronic discussion work than simply posting some controversial ideas and hoping that they will stimulate a lively discussion. In retrospect it is clear that I should have been quicker to intervene directly with questions, comments and general promptings (and to have more actively sought contributions from those who had earlier expressed an interest in the topic but never, in fact, became actively involved.)
Rather as at a successful party, the host cannot simply join in the festivities, but must be constantly scanning for potential problems or lapses in the joi de vivre and ready to intervene deftly in order to keep things going. The really good host, like the really good moderator, does this so cleverly that it appears effortless; this self-effacing skill is the most successful when it is almost invisible, which misleads the naive observer into thinking that there is nothing to it. But to organise a successful electronic meeting demands at least as much skill and energy as organising a good party. (One particular problem: I found it difficult to take part in the ongoing discussion with my usual zeal, as such a degree of partiality and enthusiasm seemed to be at odds with the overall sympathetic objectivity expected from a moderator. It is difficult to be simultaneously a player and a referee. )
However, there were several more objective reasons why this discussion was probably doomed to fail. A very successful electronic discussion organized by Stevan Harnad on almost the identical topic had recently been published. Several of the contributors to that discussion had gone on to develop their particular views in more conventional papers, or were now even working on books. Several of the contributors to this discussion therefore began by referring the reader to these publications. While reasonable under the circumstances, this is probably not a good way to begin a lively conversation. E-mail communication lies somewhere between writing and talking, and seems to be best when it can bring together some of the the precision of written text with the liveliness and spontaneity of a face-to-face conversation. If we begin by handing one another reprints, it takes an unusual degree of determination and self-control to read them all before saying anything.
There seems to be a natural progression revealed here. A topic appears which sparks widespread interest and on which not much has been written. It may arise from a notorious academic assault on some established position (as this did from Searle's claim to have reduced computationalism to triviality), or by a publication which gives academic blessing to a topic which has been déclassé for many years (as Dennett and Chalmers recently did to consciousness): but whatever, many people have views and enjoy a debate; and at this stage, debating is indeed a useful mechanism. It allows the rapid exchange of many starting ideas and counterarguments, exposes many initial misunderstandings, and allows people to sort out the major ``positions'' which are sensible and coherent. This is when electronic discussions are uniquely useful and productive.
But then there comes a more mature stage, when the various positions need time to develop in a more careful and scholarly fashion. This is slower, produces longer texts and tends to display firmer opinions. While electronic communication will always be uniquely useful between individuals, an e-mail forum does not seem to be so natural at this stage. Many of the people who might have made it interesting now find it a distraction. I think that this question of how best to characterise computation was already well into this second stage even before we began.
Editor: Patrick J. Hayes
University of West Florida