Dawkins begins with a strong analogy: the Pythagorean Theorem, he writes, is a fact. I doubt that mathematicians talk this way, but the Pythagorean Theorem is a fine example of an ironclad truth, and it allows Dawkins to establish the sense in which he is using the term “fact,” viz. a proposition or statement that “is beyond sensible doubt.”* Jerry Coyne takes a similar position. “A theory,” he writes, “becomes a fact (or a ‘truth’) when so much evidence has accumulated in its favor - and there is no decisive evidence against it - that virtually all reasonable people will accept it.” Evolution, they assert, is just such a fact. According to Dawkins, evolution “is a fact in the same sense as it is a fact that Paris is in the Northern Hemisphere.” * Not really.
These notions are seriously confused. They muddle the relationship between fact and theory in scientific investigations. For empirical scientists, facts are merely the raw material of scientific explanation; theory is the real name of the game. Facts are what science begins with; theory is its intended destination. So what is a fact? In simplest terms a fact is a statement about an observation or a summary of a number of observations. The facts that are of interest to scientists are facts that raise questions, the ones that require an explanation. For example, here’s a fact: ‘Dinosaur fossils have been found in many parts of the U.S.’ How old are they? How are they related to other fossils? How did they come to be where they are? Etc. A theory is intended to answer such questions, that is, to explain the facts. If it is to do that, a theory cannot itself be a fact. The job of evolution, then, is to explain the observed facts contained in the fossil record and other phenomena. However, on the Dawkins/Coyne interpretation, if evolution is a fact and facts require explanation, then we are confronted with the awkward question ‘What explains evolution?’ There are no scientific books or papers on that topic, because evolution is a theory, not a fact. It explains but does not itself call for an explanation.
Theories and facts belong to different logical categories. Facts are particular, specific; theories are very general. Secondly, facts can be strung together in conjunctive sentences, but a theory cannot be a bead on the string. The sentence "We have a rich fossil record scattered over billions of years (fact) and life has evolved from a single ancestor" (theory) is syntactically correct but scientifically incoherent. We have a rich, diverse fossil record because life has evolved from a single ancestor. The clauses of the latter sentence cannot exchange places, unlike a simple conjunction of facts; they have a fixed logical relationship. Also, facts are observed, theories are constructed. Another difference is that a fact is stuck with being what it is - it produces no offspring. A theory may lead via prediction and experimentation to the discovery of new facts (in a stepfatherly sort of way).
Coyne is right to say that a theory can become a fact, as for example when a detective’s theory of a murder case is confirmed by a video showing the murderer in the act of killing his victim. However, that occurs rarely in science, where theories deal mostly with the unobservable, and never by the mere accumulation of confirming evidence. It certainly will never occur with evolution owing to the impossibility of observing extinct species. Coyne would be on firmer ground if he were to stick with “evolution is a ‘truth’ ” - not ideal but better than ‘fact.’ Evolution can be characterized modestly as true in the pragmatic sense as a set of ideas that work, that is, it explains a large range of phenomena and allows for predictions which have been borne out by subsequent investigations, and it has not been overthrown by decisive evidence.
Why is this important? First, for our own understanding - always a supreme value - but also because the cause of defeating creationism is not well served in the long run by the tactical use of falsehoods or distortions. Above all, we want everyone to be possessed of clear and accurate ideas about the nature of science.
* Not an effective analogy. Mathematical theorems are not facts as science understands the term. They are necessary truths arrived at by deduction from a priori axioms, whereas scientific theories are contingent upon empirical evidence and are therefore unavoidably probabilistic.
- C. Marxer
References: Jerry A. Coyne, Why Evolution Is True, 2009.