The simple truth about complexity (Geoff Mulgan, É«ºÐÖ±²¥ View, É«ºÐÖ±²¥S, December 22) is that Mr Mulgan is talking through his hat. For instance, he accuses proponents of theories such as complexity and chaos of "trying to make a religion out of it". What are the grounds for this accusation? There is no religion in the technical journals.

However, in the popular books you will find opinions about how the new theories affect out view of humanity's place in the universe. Those opinions are there because the public wants to know about such questions. Similar opinions are expressed in many established areas of science - cosmology (Hawking), particle physics (Weinberg, Gell-Mann), developmental biology (Wolpert) - and they are bread and butter to writers in the humanities or the arts.

But apparently imaginative scientists who think they have some new insights are not permitted to express their excitement, or to hazard some guesses at the wider implications of their ideas. Of course, if they do not, they will be criticised for dehumanising science.

Mr Mulgan is wrong to dismiss these theories as having no useful applications. However, any new paradigm suffers from "the Rolls-Royce problem". One person's new idea is being asked to compete with established work done over decades by thousands of people. It is as if a new concept in car design is being asked to compete with a Rolls Royce - instantly.

É«ºÐÖ±²¥

It took 80 years for the photocopier to pass from the laboratory bench to the office. It took over a century for the fax machine to do the same. Remarkably, the ideas of chaos and complexity have led to useful applications over a far shorter period. They range from quality control of springs to the management of coral reef ecosystems.

Mr Mulgan says "there is not one usable theory" in economics. False, as the work of Brian Arthur shows. Perhaps these theories are not actually being used at the moment, but that is largely because of the Rolls Royce problem, and it owes far more to ingrained conservatism among conventional scientists than it does to an inherent fault of the new viewpoint.

É«ºÐÖ±²¥

In any case, the main reason for getting excited about the new theories is the manifest failure of the old ones.

Ian Stewart Professor of Mathematics, University of Warwick