The latest flare up in the evolution versus intelligent design dispute has brought out some questionable assertions regarding the role of theory in science. One such claim is that a theory is a hypothesis that has passed a round of tests. That is, there are different degrees of certainty in science; ‘theory’ is simply the intermediate state between ‘hypothesis’ and ‘law’.
That is certainly good news for the proponents of reigning theories. But is it good for science? The history of science shows that relatively few theories endure—and even fewer endure without modification. More important, theory-as-partially-verified-hypothesis confuses the distinction between the facts that a theory attempts to explain and the explanation.
The French scientist Claude Bernard, widely considered the Father of Physiology, recommended that researchers treat all theories with skepticism. He had good reasons for this attitude. Bernard was at the forefront of the movement to liberate medicine from scholasticism and refashion it as a dynamic, experimental science. His classic work An Introduction to the Study of Experimental Medicine is worth quoting at length (all of the following quotes are found in Part 1, Section III):
The first condition to be fulfilled by men of science, applying themselves to the investigation of natural phenomena, is to maintain absolute freedom of mind, based on philosophic doubt. Yet we must not be in the least skeptical; we must believe in science, i.e., in determinism; we must believe in a complete and necessary relation between things, among the phenomena proper to living beings as well as in all others; but at the same time we must be thoroughly convinced that we know this relation only in a more or less approximate way, and that the theories we hold are far from embodying changeless truths. When we propound a general theory in our sciences, we are sure only that, literally speaking, all such theories are false. They are only partial and provisional truths which are necessary to us, as steps on which we rest, so as to go on with the investigation; they embody only the present state of our knowledge, and consequently they must change with the growth of science, and all the more often when sciences are less advanced in their evolution. On the other hand, our ideas come to us, as we said, in view of facts which have been previously observed and which we interpret afterword. Now countless sources of error may slip into our observations, and in spite of all our attention and sagacity, we are never sure of having seen everything, because our means of observation are often too imperfect. The result of all this is, then, that if reasoning guides us in experimental science, it does not necessarily force its deductions upon us. Our mind can always remain free to accept or to dispute these deductions. If an idea presents itself to us, we must not reject it simply because it does not agree with the logical deductions of a reigning theory. We may follow our feelings and our idea and give free rein to our imagination, as long as all our ideas are mere pretexts for devising new experiments that may supply us with convincing or unexpected and fertile facts.
Bernard is not saying that scientists should be extreme skeptics. Nor is he saying that theories are mere guesses. What he is saying is that theories are attempts to connect the factual dots within a logically consistent model. Theories can help lead us to new discoveries. But we must not let them blind us to facts that don’t fit the model.
Bernard proceeds to describe how unwarranted attachment to theories can lead to error:
If a doctor imagined that his reasoning had the value of a mathematician's, he would be utterly in error and would be led into the most unsound conclusions. This is unluckily what has happened and still happens to the men whom I shall call systematizers. These men start, in fact, from an idea which is based more or less on observation, and which they regard as an absolute truth. Then they reason logically and without experimenting, and from deduction to deduction they succeed in building a system which is logical, but which has no sort of scientific reality. Superficial persons often let themselves be dazzled by this appearance of logic; and discussions worthy of ancient scholasticism are thus sometimes renewed in our day. The excessive faith in reasoning, which leads physiologists to a false simplification of things, comes, on the one hand, from ignorance of the science of which they speak, and, on the other hand, from lack of a feeling for the complexity of natural phenomena. That is why we sometimes see pure mathematicians, with very great minds too, fall into mistakes of this kind; they simplify too much and reason about phenomena as they construct them in their minds, but not as they exist in nature.
He goes so far as to suggest that science can give rise to a unique form of superstition:
...We must trust our observations or our theories only after experimental verification. If we trust too much, the mind becomes bound and cramped by the results of its own reasoning; it no longer has freedom of action, and so lacks the power to break away from that blind faith in theories which is only scientific superstition.
Or as the great American inventor Edwin H. Armstrong was fond of saying, “It ain’t ignorance that causes all the trouble in this world. It’s the things people know that ain’t so”:
It has often been said that, to make discoveries, one must be ignorant. This opinion, mistaken in itself, nevertheless conceals a truth. It means that it is better to know nothing than to keep in mind fixed ideas based on theories whose confirmation we constantly seek, neglecting meanwhile everything that fails to agree with them. Nothing could be worse than this state of mind; it is the very opposite of inventiveness...
Educators would do well to heed Bernard’s words:
In scientific education, it is very important to differentiate, as we shall do later, between determinism which is the absolute principle of science, and theories which are only relative principles to which we should assign but temporary value in the search for truth. In a word, we must not teach theories as dogmas or articles of faith. By exaggerated belief in theories, we should give a false idea of science; we should overload and enslave the mind, by taking away its freedom, smothering its originality and infecting it with the taste for systems.
Finally, he points out that not only religious beliefs, but scientific beliefs, can obstruct the search for truth:
To sum up, two things must be considered in experimental science: method and idea. The object of method is to direct the idea which arises in the interpretation of natural phenomena and in the search for truth. The idea must always remain independent, and we must no more chain it with scientific beliefs than with philosophic or religious beliefs; we must be bold and free in setting forth our ideas, must follow our feeling, and must on no account linger too long in childish fear of contradicting theories...
Today, Bernard is revered as a champion of vivisection but shunned as an opponent of human clinical studies. Though he was wrong to assume animal research is sufficient, his skepticism towards the use of statistics in human studies was not entirely unwarranted. Scientists would do well to read his book today.
UPDATE May 13, 2008 - 10:00am Eastern:
According to Wikipedia, even the National Academy of Sciences has decided it's time to redefine theory. The Academy reasons that if a theory purports to explain a sufficiently large number of facts, then somehow that validates the explanation.
Never mind that there can be any number of theories to explain the same facts. Never mind that the Academy is essentially tossing out the bulk of philosophy of science.
NOTE: This post was originally titled Claude Bernard’s Take on “What is a Theory?”