One of the fascinating things about observing the effects of computers on people is their effect on so-called computer experts. They do not seem to realise, even as they pontificate on the potentials and advantages of systems, how bizarre and illogical their own behaviour has become.

Consider, for example, a university where a department recently opted for an network of Apple Macs, in part because of their user-friendliness. Not for long. Within a few days, the departmental computer expert (there is always one) had added enough automatic start-up features to baffle the most sophisticated evolutionary organism, and we are dealing, let us not forget, with academics.

There was the funny message about mail involving ampersands and other rightly nameless typewriter symbols, the reminder to change your password, the warning about such and such, the need to make back-ups immediately, and not to forget the mysterious little square which apparently was there in case anyone felt the need to treat the computer as a memo pad (instead of using the more conventional sort).

Needless to say, the department members were very pleased with their new computers and one or two of them even tried using them for email, "e" standing for, of course, ersatz, substitute, inferior. However, it then transpired that no one read the mail being sent to them, at which point a warning note was produced reminding people to "check their mail boxes" regularly. This note, it was found after several false starts, worked best if it was distributed by traditional paper means.

ÖйúAƬ

Traditional paper means were not however, encouraged. For instance, in order to pay for computer support (computers need a lot of support, otherwise they don't feel appreciated) the department had axed the money for printers. Instead of everyone having their own printer in their office, there would be one in the main office, with which people could communicate electronically. The repeated trips to and from the office could be looked on as useful exercise combating the effects of too much staring at computer screens.

Unfortunately, this printer had an annoying habit of breaking down, at which point the only way to communicate was email. But then, as any computer enthusiast will tell you if you let them, the only way to communicate these days is email.

ÖйúAƬ

In the 1960s, Joseph Weizenbaum developed a number of programs that accepted natural language from a user typing on a keyboard, and responded with what appeared to be natural language via a printer. The most celebrated of these programes was Eliza, a program that was modelled on psychotherapy. Eliza generally returned whatever people typed in a slightly different order, perhaps having picked out a keyword. The random responses were interpreted as evidence of deep thought (much as they are in philosophy seminars). Simple though the program was, it became very popular and psychiatrists adopted it as the basis for actual therapy sessions. Users became attached and dependent on Eliza and computer pundits cited it as an example of how computers could learn to "talk".

Aghast at this, Weizenbaum decided that computers should not be allowed to give responses that appear human. He later wrote, "what I had not realised is that extremely short exposure to a relatively simple computer program could induce powerful delusional thinking in quite normal people".

Computers in education make the teacher "fall into the habit of merely training". Their classes, and indeed themselves, become "less than fully autonomous persons . . . mere followers of other people's orders, and finally no better than the machines that might someday replace them in that function".

At another university where I had the sad task of teaching, there were not one but two computer enthusiasts - and if one makes for trouble, two makes for . . . rather less. They never actually agree on anything, and as a result increasingly grandiose plans are continually being delayed or put off. In this case, the proposal was to utilise not just email but interactive hyper multimedia to produce a sort of virtual reality conference.

ÖйúAƬ

No longer would members of staff have to meet in the seminar room for those long dreary talks on Friday afternoons. Now it could all be done by sitting in your office staring at the computer screen on Friday afternoons. Instead of talking to people, you would simply click on a picture of your colleague, and look at what they thought - in black and white. Then if you disagreed you would simply delete their views and write in your own.

One can easily see the attractions of such a system. Unfortunately, it never really got going, the future was obstructed by the refusal of the organic academics to progress beyond learning how to update the dates at the bottom of their lecture handouts for students each September.

Martin Cohen Visiting lecturer in philosophy at University College, Scarborough.