In response to a previous posting, commenter Sarah Taylor made several important points. She clearly articulated the view that evolutionary theory is a complex edifice that is built on many observations that fit into a general pattern that is largely chronologically consistent.

She also notes that one distinguishing feature of science is that there are no questions that it shirks from, that there are no beliefs that it is not willing to put to the test. She says that “What makes scientific theories different from other human proposals about the nature of the universe are their courage. They proclaim their vulnerabilities as their strengths, inviting attack.�

I would mostly agree with this. Science does not shy away from probing its weaknesses, although I would not go so far as to claim that the vulnerabilities are seen as strengths. What is true is that the ‘weaknesses’ of theories are not ignored or covered up but are seen as opportunities for further research. Since there is no such thing in science as infallible knowledge, there is no inherent desire to preserve any theory at all costs, and the history of science is full of once dominant theories that are no longer considered credible.

But having said all that, it is not necessarily true that finding just one contradiction with a theory is sufficient to overthrow the theory. In the context of the challenge to Darwinian theory by intelligent design (ID) advocates, Sarah’s statement that “All that any ID devotee has to do is to show ONE fossil ‘out of place’, to prove the theory doesn’t work. Just one horse shoulder blade in a Cambrian deposit somewhere in the world, and we can say goodbye to Darwinâ€? is a little too strong.

Sarah’s view seems to be derived from the model of falsificationism developed by the philosopher of science Karl Popper (see his book Conjectures and Refutations: The Growth of Scientific Knowledge), 1963) who was trying to explain how science progresses. After showing that trying to prove theories to be true was not possible, Popper argued that what scientists should instead do is try to prove theories false by finding a single counter-instance to the theory’s predictions. If that happens, the theory is falsified and has to be rejected and replaced by a better one. Hence the only status of a scientific theory is either ‘false’ or ‘not yet shown to be false.’

But historians of science have shown that this model, although appealing to our sense of scientific bravado, does not describe how science actually works. Scientists are loath to throw away perfectly productive theories on the basis of a few anomalies. If they did so, then no non-trivial theory would survive. For example, the motion of the perigee of the moon’s orbit disagreed with Newton’s theory for nearly sixty years. Similarly the stability of the planetary orbits was an unsolved problem for nearly 200 years.

Good theories are hard to come by and we cannot afford to throw them away at the first signs of a problem. This is why scientists are quite agreeable to treating such seeming counter-instances as research problems to be worked on, rather than as falsifying events. As Barry Barnes says in his T.S. Kuhn and Social Science (1982): “In agreeing upon a paradigm scientists do not accept a finished product: rather they agree to accept a basis for future work, and to treat as illusory or eliminable all its apparent inadequacies and defects.�

Dethroning a useful theory requires an accumulation of evidence and problems, and the simultaneous existence of a viable alternative. It is like a box spring mattress. One broken spring is not sufficient to make the mattress useless, since the other springs can make up for it and retain the mattress’s functionality. It takes several broken springs to make the mattress a candidate for replacement. And you only throw out the old mattress if you have a better one to replace it with, because having no mattress at all is even worse. The more powerful and venerable the theory, the more breakdowns that must occur to make scientists skeptical of its value and open to having another theory replace it.

After a theory is dethroned due to a confluence of many events, later historians might point to a single event as starting the decline or providing the tipping point that convinced scientists to abandon the theory. But this is something that happens long after the fact, and is largely a rewriting of history.

So I do not think that finding one fossil out of place will dethrone Darwin. And ID does not meet the necessary criteria for being a viable alternative anyway, since it appeals to an unavoidable inscrutability as a factor in its explanatory structure, and that is an immediate disqualification for any scientific theory.