Wake-up call for science

In theory, science is a failure. In practice, it works; in product, it soars. Yet because scientific product depends on practice, which in turn depends on theory, it is vital that scientists understand how and why science fails.

Science has a long, rich history that practising scientists often overlook. Modern scientists proudly herald the enterprise of science as logical (reasoned), yet since science's earliest stirrings more than 2000 years ago, various thinkers have maintained this logic is flawed.

Even so, it wasn't until 1739 that Scottish thinker David Hume drove the point home by specifying that modern science is inherently illogical, because it relies in part on assumed theories that must reach beyond what we can ever observe.

This has profound implications: the very foundations of science are infirm. Yet understandably, in his day Hume's concern fell largely on deaf ears of those more interested in reaping rewards of practical science at the birth of the industrial revolution.

It wasn't until the mid to late-20th century that science historians Thomas Kuhn and Paul Feyerabend looked back across 400 years of astonishing scientific achievement to conclude that it was precisely the theoretical shortcomings of science that forced scientists to be more liberal in practice.

They proposed that Copernicus and Galileo, for example, boldly stepped outside the limits of logic to garner support for their theories within the science community.

Such activities imply an inherently social aspect to science, although the degree to which "opposing communities of scientists" drive theory change remains contentious. But, at least, such work shows that science does not simply equate with reason.

Today, practising scientists may dismiss these issues as unimportant, but this is a sheer mistake.

The standard view is that science's problems may be generalised thus: we can never be 100 per cent sure in truth, even of scientific theories, because of the limits imposed by our human observations of the world.

Yet if we give this fundamental problem credence, our pursuit of scientific "truth" becomes challenging indeed.

Most science practitioners conclude that paying anything more than lip service to Hume's problem is futile, a stymie to progressing knowledge.

More controversially, if we push the notion of a "social" aspect to science championed by Kuhn and Feyerabend to its limit, we may view every scientist as inherently subjective, plying any given scientific theory only via their private world. We then have cause to wonder: if each person has his or her own valid mode of doing science, could there be as many methods of science as there are scientists?