Hemidemisemi science

A quarter note in music is classically known as a quaver. A sixteenth note is half of a quaver, which is called a semiquaver. For a thirty-second note the prefix “demi” is used instead of a second “semi” to make a demisemiquaver. Similarly, a sixty-fourth note is a hemidemisemiquaver. As we shall see, these prefixes will come in handy.

“Science” is a term for what was called in centuries past “natural philosophy” or what Isaac Newton called “experimental philosophy.” Physics is the pre-eminent science and has always set the highest standard for anything else to be called “science.” It is primarily characterized by highly controlled experiments. The particular value of controlled experiments has been precisely described recently by the study of causality (see Judea Pearl’s book, Causality: models, reasoning, and inference. Cambridge, 2000.)

In order to trace out a causal chain it is best to have sufficient control over all variables and then vary one individual factor at a time while holding all other variables constant. Such highly controlled experiments are only available for hard science such as physics and chemistry. That plus inductive generalization is science. Anything else that claims to be science should be measured by such a standard.

By relaxing the standards somewhat one can engage in fairly controlled experiments in other subjects. Medical science uses limited experiments to validate results of treatments on patients. However, unlike electrons and molecules, patients are individuals that vary and whose particular history cannot be controlled so statistical approaches are required for more predictable results. But medical science does not have the precision of physics or chemistry. One may call medical science a “semi-science.”

There is great interest in making psychology and sociology into sciences. But greater limits on experiments prevent them from reaching the standards of medical science, much less physics or chemistry. One may call psychology or sociology a “demisemi-science.” It is a similar situation with fields such as astronomy and botany in which the main source of data is not experiment but field observation. There may be much observational data and empirical generalization but causality is elusive without experiments. They are demisemi-sciences, too.

There is great interest in making the study of the past into science but there are many reasons why the subject of history is not science and should remain with the humanities. For one thing, the role of documentation is much more important. The best sources of history are trustworthy documents based on direct experience and observation of past happenings. Understanding and evaluating such texts is primarily an activity that must consider wider aspects of life — religion, culture, philosophy, and the like. There is a limited place for science.

Some say that natural history is science because its sources are objects such as rocks rather than texts. But again the best sources are documents based on direct experience and observation. A written account of a volcano observed, for example, is much superior to trying to estimate the date from rocks of today. Interpretation is required which takes us away from the controlled world of science. One may call natural history a “hemidemisemi-science.”

So while many would try to wrap themselves with the prestige of science, there is nothing that has the rigor of physics and chemistry because they cannot control all variables. Other disciplines are half or a quarter or an eighth of a science. Our terminology should reflect this if we are trying to avoid exaggeration.

October 2011