In science, a theory is “a well-substantiated explanation of some aspect of the natural world that is acquired through the scientific method and repeatedly tested and confirmed”–according to Wikipedia. This is frequently contrasted with the colloquial meaning of “theory”, which usually refers to something speculative and unconfirmed. It is suggested that in a scientific context, it is more appropriate to refer to a speculative idea as a “hypothesis”.

However, in my experience as a physicist, this is not how the word “theory” is used in practice. Generally, the word “theory” is contrasted with “experiment”, describing the kind of work rather than the quality of the work. Since theories are carefully crafted by experts, it is fair to say that they are more than mere speculation, but that doesn’t mean that every theory has been thoroughly tested and confirmed. Some theories are untested, some theories are in direct competition with other equally viable theories, some theories intentionally model things that do not presently exist, and some theories are just poorly crafted.

So, basically, Wikipedia–and most dictionaries as well–appear to be in conflict with my understanding as a fluent speaker of English physics. That probably means there’s some bad lexicography going on.

But we should start out being charitable to the dictionary. Wikipedia indeed describes one of the ways that “theory” is used in science. It may even be the most common usage, for all I know. For example, it is possible that condensed matter physicists mostly use “theory” in the way I understand, while every other field of science mostly uses the definition given by Wikipedia.  I doubt it though, since I do occasionally read papers in other fields.

And to be honest, the given definition of “theory” just strikes me as impractical in real scientific work. Just imagine that every time I cite a theoretical work, I got into a pointless argument with a referee who strongly feels that it either should or should not be called a “theory”. There are enough pointless arguments with referees as it is.

But obviously this question requires some serious lexicographical research, and I’m not equipped to provide that.

So, granting that there is a discrepancy between commonly given definition of “theory”, and its actual usage, how are we to explain it? I do not know the correct explanation, but it is easy to come up with several possibilities.

First, there’s the whole pro-science/anti-science debate. Creationists have said many times that evolution is “just a theory”, and to any scientist the argument sounds absurd. Evolution is a theory, but that does not mean it is idle speculation. It is very important to explain to the public that “theory” does not, in a scientific context, mean “speculation”. And one way to convey this to the public is to provide a prescriptive definition of “theory” that errs in the opposite direction, describing theories as pinnacles of scientific achievement.

Of course, it helps that evolutionary theory, in particular, really is a pinnacle of scientific achievement. In general, the public is only really aware of the most established scientific theories, like evolutionary theory, gravitational theory, quantum theory, germ theory, and plate tectonics theory. It is very easy to generalize from these famous examples to a definition that holds theories in very high regard. But most theories are not famous. It would not be unusual if I produced some unusual observation in the compound Bi₂Sr₂CaCu₂O_8+δ, and then collaborated with a theorist to generate a new theory to explain it. (Yes, this has happened, and no I do not deserve congratulations for it.) There’s really no reason for these kind of theories to get public recognition unless they are very broad, well-established, and dominant in their fields.

Finally, you can never underestimate the power of prescriptivist thinking. I was highly amused by the Wikipedia talk page, where someone defends Wikipedia’s definition of “theory” by saying the following:

The problem arises because nearly every scientist pushes to have their pet hypothesis referred to as a theory and often times through political campaigning they accomplish this. The examples that you give – Lemarckian evolution, phlogiston theory, vortex theory – are not and never were scientific theories. It was wrong to call these theories when they were never anything more than hypotheses, and now they are nothing more than failed hypotheses. Even string theory is really only an hypothesis, not a scientific theory.

In other words, they admit that many scientists use an “incorrect” definition of theory. But if the usage is so common, on what basis is it incorrect? The purpose of Wikipedia and the dictionary is describe things as they are, not how they should be.

(Note: I wrote this as a tangential reaction to an article by Crip Dyke. Here, Crip Dyke contrasts the usage of “theory” in science with its usage in social theory. I think the difference is not as large as supposed, and I always find it odd when other fields seem to have an inferiority complex based on a rather idealized view of the hard sciences.)

(Second note: Following best practices, I use permanent links to Wikipedia articles as they were at time of writing.  They may have been revised since then.  I do not think it would be feasible to “fix” the Wikipedia article at this time, since Wikipedia’s definition is similar to the one given by many science organizations, and Wikipedia can only be as good as its sources.)