AS we probe ourselves with an increasingly impressive array of sensors linked to computers, we are beginning to gain insights into just how complex an organism we are. But most of our studies fail to take full account of causality. We know that smoking cigarettes rewards us with a one in 200 chance of death by cancer, but is it the cigarette that is the killer or some predisposition to smoke? Despite decades of warning signs we continue to ignore this vital aspect of statistics.

Modest amounts of exercise and indulgence in good foods are unlikely to be harmful, but extremes of exercise and eating will adversely affect our lives. Sometimes this will transcend the physical and incur some mental penalty. Stress, for example, is thought to be detrimental, and yet without it we do and achieve nothing. It seems that some can lead highly stressful lives and even feed off it and prosper. In contrast, others are prone to heart attacks and seizures and should not be put into stressful situations. Running on the edge of a heart attack or a nervous breakdown for any reason is unwise, and I can think of little that would persuade me to do so. And yet information technology is pushing us to do more in ever-shorter times.

In recent experiments a number of heart patients were set mathematical problems in a stressful environment. As they solved each problem they were asked to speed up their response until they started to make mistakes, and then errors were harshly corrected. This element of harassment showed that these patients used a smaller volume of the right side of the brain than normal. They also displayed hyperactivity in their left lobes and the blood flow associated with pain perception decreased. Physically, it is not at all clear that people in this class get any physical warning that they are actually prone to damaging themselves.

I think most of us are aware of damaging ourselves physically at some time during our lives, through accidents and over-stressing during sporting activities. Asking our bodies to do more than they can deliver generally invokes obvious warning signs. What is not so evident is the over-stressing and battering the human brain will receive in a normal day's education, experience or work. Would we treat machines this way? I think not. We understand the operational limits of machines, and put them into environments where they can perform well without failure.

We have never had the facility to characterise humans in such a manner, to make sure they are situated in the right environment where they can perform without any form of damage or breakdown. It might just be that new sensor and scanner technologies will allow us to characterise people under standardised conditions. We could then create some bounded zone in which they could really perform. Certainly, as computers continue to develop apace there is going to be an increasing opportunity for the creation of working environments that are not suited to large numbers of the population. It would be cruel to subject these people to the stress and physical damage that would ensue.

In a sane world we would make the most economic of choices by selecting people and fitting them to the task rather than tyre-levering them into some life-threatening operation.

Peter Cochrane holds the Collier Chair for the Public Understanding of Science & Technology at the University of Bristol. His home page is:
http://cochrane.org.uk