One of the staples of popular science journalism is that every development and every discovery tend to be breathlessly reported as a 'breakthrough'. My own favourite is the article in the British newspaper, The Sunday Times, reporting the discovery of a 'gene' for asthma which, the reporter went on to say, would lead to a cure for asthma within five years. The report appeared 15 years ago.
The dynamics of newspapers and of science are always in tension with each other. Newspapers seldom have a time-frame that stretches much beyond tomorrow. They also focus on people and 'human interest'. Science on the other hand tends to de-emphasise the personal and to generalise rather than concentrate on the particular and the human. Science is also a process of a gradual unfolding of knowledge, not of breakthrough discoveries. In the words of the late, great Sir Peter Medawar, it is the 'art of the soluble'. Researchers ask very small questions, because these are ones that they can solve.
Ray Girvan's article on computational fluid dynamics illustrates the point very well. The equations which lie at the heart of this subject are more than 180 years old, and were first derived (albeit incorrectly) by Navier in 1821. But in a curious way, this case also highlights the importance of the human element in science. Research is not an impersonal, disembodied process; it is carried out by real people whose importance is sometimes overlooked.
Despite the antiquity of the problem of fluid dynamics, mathematicians have at best only a primitive understanding of these equations, even today, with no certainty that solutions exist for some cases. The problems could hardly be more important: fluid flow arises in many situations. Cardiovascular disease is one of the greatest killers in Western developed societies (and is rising in importance in many developing countries). One basic pre-requisite in understanding this class of disorders is to model the flow of blood in veins and arteries. Fluid flow is obviously important in the design of aircraft, both in terms of the flow of air over the wings and fuselage and also in terms of the design of the engines. My personal favourite is the application of computational fluid dynamics in the making of pasta in Italy. But the equations have applications in many situations which do not, at first sight, appear to be fluid flow. Yet crowds of people and traffic 'flowing' down the motorway also behave in ways that can be described by these equations.
Across many areas of science, not just fluid dynamics, the advent of numerical methods, packaged into easy to use and easy to understand software products, has helped to broaden the range of application of theoretical calculations. Numerical methods are comparatively cheap; experimental methods and empirical approximation can be very expensive. But these packages should not be applied blindly, especially where the underlying scientific understanding is less than perfect. The human element, the exercise of judgment, is just as necessary in today's environment of scientific computing as it has ever been. Perhaps the popular science journalists' focus on human interest is not so wide of the mark, after all.
Dr Tom Wilkie
