What Is A Scientific Theory?

How does science work? (At least in the case of physics and other mathematical sciences.) I touched on this subject before (in a post entitled Does Nature “Obey” The Laws Of Physics?), but I would like to take a more general perspective here.

So how does science work? First you observe the world, and you also do experiments. You also abstract from the many observations and experiments the key quantities (such as position, velocity, force, etc.) that will appear in your theories.

Then you create theories that relate the key quantities in ways that help you explain the phenomena that you observed  and/or appeared in your experiments. You create the theories by guessing; that’s right, guessing. (OK, if you want a fancy word for it, you can call it inductive logic. But it’s still guessing.)

Of course, I don’t mean random guessing. It’s a creative process that requires a deep knowledge of the current scientific understanding of the world, and it helps if you know the history of the development of science. What I’m trying to say is that you don’t logically derive the laws of physics; you just create them. Then, once they are created you test them using logic, and if they survive these tests, then you test them using observations and experiments. Ultimately, the vast majority of scientific theories are discarded; few survive to form part of the ever-evolving currently generally accepted body of science.

You test your theories against the phenomena that you observed/experimented on. If the equations of your theory predict results that agree with your observations or experiments, then good. If not, you will have to modify your theory, or maybe discard it and start from scratch.

Then you use deductive logic to try to derive consequences of the theory that were not observed before. If you can do this, and if subsequent experiments or observations agree with the predictions of the theory, then that is very good. Otherwise, you will have to modify your theory, or maybe discard it and start from scratch.

Logic plays a key role in testing physical theories. A theory of physics must be logically consistent; for example, it must not be possible to derive two contradictory predictions from the theory. But the creation of a theory is not necessarily a logical process, at least not in the same sense. Intuition, analogy, “feeling,” play a greater role in creation; logic plays the primary role in testing the theory for consistency and for deriving consequences and predictions. But the ultimate test of a physical theory is observation and experimental verification.

No amount of experimental or observational testing can ever prove a scientific theory correct. Though scientists sometimes use such terms (saying a theory is right or true or correct) colloquially, they are not meaningful because a scientific theory can never be proved correct, because it’s impossible to test the theory at all points in space and at all times.

Asking whether a scientific theory is correct is like asking whether your marriage is red or green (which would be truly confusing if you are married to Red Green). Or asking whether a sculpture by Modigliani is true or false. Such questions are meaningless. Although Picasso once said that “Art is a lie that helps you see the truth.” Beautiful, isn’t it? And a scientific theory is something like an art work as well: A human creation that is somehow false (has approximations built in, has oversimplifications, idealizations, has limited applicability, etc.), but yet helps us gain insight into our wonderful world.

Some people try to denigrate science using phrases such as, “it’s only a theory.” That demonstrates a profound misunderstanding of science (or perhaps a willful attempt to mislead). There is a difference between the every-day use of the term “theory,” to mean uninformed speculation, and the scientific use of the term theory. If science were like the Olympic games, then achieving the status of “theory” would be analogous to winning a gold medal. Becoming a theory (successfully tested by observation and experiment) is the pinnacle of achievement for a scientific idea.

Others twist the nature of science around to market their products, or to sell their ideas. When I was young, there was a brand of floor wax called Aerowax, and their TV commercials trumpeted the fact that it represented SPACE AGE TECHNOLOGY! (They sponsored this TV show (the product is promoted at about the 1-minute mark), but back then it was called the “jet-age floor wax.”) And the truisms “scientific fact” and “scientifically proven” were popular advertising catch phrases in the 1970s. Nowadays, one finds the word “quantum” attached to many, many products and companies, to lend them a cachet of some kind of scientific validity (beware!).

Although scientific theories can’t be proven correct, they are nevertheless a precious part of human culture. They represent the highest achievements in scientific thought. They represent the most successfully tested, hardened-by-trials products of the scientific enterprise. The vast majority of scientific ideas end up in the slag heap; the best theories are the survivors.

Reflect on the words of Henri Poincaré, which emphasize the role of creativity:

A science is no more a collection of facts than a house is a heap of stones.

Also reflect on the words of Isaac Asimov:

Consider some of what the history of science teaches. First, since science originated as the product of men and not as a revelation, it may develop further as the continuing product of men. If a scientific law is not an eternal truth but merely a generalization which, to some man or group of men, conveniently described a set of observations, then to some other man or group of men, another generalization might seem even more convenient. Once it is grasped that scientific truth is limited and not absolute, scientific truth becomes capable of further refinement. Until that is understood, scientific research has no meaning.

If he were writing today, Asimov would no doubt have used the word “person” instead of “man,” but I’m sure you get the idea: Laws of physics are not to be obeyed, but are rather convenient generalizations of nature’s workings. The collection of all physical theories is like a vast work of art; nobody would call it correct, but it’s beautiful, and absolutely useful. The bridges engineers design using Newton’s laws don’t fall down, do they? And the MP3 players made using principles of electromagnetism and quantum theory are rather functional as well.

So physical theories are not “true,” but they are tightly constrained to apply very closely to this world. But some day, maybe tomorrow, maybe next century, someone (maybe one of you?) may create a new theory, that is somehow more beautiful, or more useful, or somehow of value, so that it may supersede or replace an existing theory of physics.

PS: I occasionally encounter people who feel they have invented a tremendous new theory; it has usually been relatively straightforward to set them straight, although I have read stories about famous scientists who have had persistent pestering from people who don’t understand how science works, and resist any efforts at correction. Check John Baez’s Crackpot Index and Warren Siegel’s “Are you a quack?” for their remedies. Gerard ‘t Hooft also has a number of responses to crackpots at his site; for example, this clarifies and corrects some typical thinking errors in relativity.

(This post first appeared at my other (now deleted) blog, and was transferred to this blog on 22 January 2021.)