What’s a Science?

(some ruminations from the work in *How to Reason + Reasoning in the Sciences*)

Some like to say that science is determined by its methods. But that’s not clear enough.

It seems that we have two necessary criteria for a subject to be classified as a science.

• There is a theory or theories.
• From the theories, predictions can be made that can be tested (shown to be true or false).

Are biological classifications science?

“The purpose of taxonomic classification is improved communication. If scientists are in general agreement with respect to the proper “scientific” name of an organism, anyone who reads the name knows which animal the writer means. Consistency precludes confusion. Most important earlier classifications were based on a description of form; that is, animals that looked alike were placed together and the classification was then given a scientific name. This classifying and naming process is what we mean by ‘taxonomy.’ ”
Claud A. Bramblett, Patterns of Primate Behavior, 2nd edition

If there is no theory, only a method of classifying, this doesn’t seem to be science. You can show that someone has classified incorrectly, if the criteria are objective. But you can’t show that the criteria are wrong. They’re only more or less useful for the purposes of communication. If someone finds a specimen that doesn’t fit into the classification system, the system can be expanded.

By these criteria, grammar as taught in secondary schools is also not a science. The rules of grammar are all prescriptive. Science is meant to describe and explain the world. The prescriptions of science are the same as the prescriptions of any study: if you agree with the theories, then you should accept their consequences. So we have another necessary criterion for a subject to be a science.

• The claims that can be deduced from the theory are descriptive.

Is astronomy, like taxonomy, only classification and hence not a science? No. Astronomers aren’t simply tracking the movements of planets and stars and classifying those. They are developing theories they test by making predictions. These are predictions about what happened—at least for stars, since the light astronomers observe has taken years to arrive here. Predictions can be made about the past, too. What about the Big Bang theory? That seems to be just a speculative explanation. But scientists say that predictions can be made with it. They can test it, and so far it seems correct.

There does seem to be a difference between experimental science and historical science. Chemistry is an experimental science. Astronomy is an historical science. With an historical science, theories and explanations can be offered, and predictions can be made from those about what one will find in the historical record. But no experiments can be made, for the past cannot be manipulated. Or perhaps we have different kinds of experiments, for sending a satellite to a comet to discover the composition of the material on it is a kind of experiment to test a theory.

Is evolution an historical science? Not entirely. Scientists have manipulated some species, such as the silver fox in Russia, to show how certain traits evolve.

Is history a science? Historical theories have been propounded, predicting the future development of human society. They’re always wrong. There are too many differences to make the analogies (abstractions) good. You can say that some historians, such as Marx, were trying to be scientists, using the methods of scientists. But if so, they were doing bad science.

What about economics? That’s not just history. Economists propound theories. Some have even been given power to manipulate the economic system of a country or even the whole world for experiments. Their theories are always proved wrong for the same reason that historical theories are wrong: there are too many differences between the past and the present. So if economics is a science, it’s a bad one.

What about the work of ESP researchers and Freudian psychoanalysts? They use the methods of science. But their theories are too vague or too general to be tested. We can never show that a (purported) claim deduced from the theory is false, for anything will confirm those theories.

To comment on this post, please visit our Facebook page: