What’s a scientific theory

Karl Popper

Although it is fashionable to assert that Karl Popper’s theories about the evolution of science are outdated, his definition of what is a scientific theory is unassailable:

A theory is scientific if and only if it is possible to design an experiment that proves that this theory is false.

A paradigmatic case is the Copenhagen Interpretation of Quantum Mechanics. In 1935, Einstein, Podolsky and Rosen designed an experiment that could prove this theory false. A few months later, Niels Bohr published another article in the same magazine, in answer to the previous article. Almost 30 years later, as I explained in another post in this blog, the EPR experiment, which up to that point had been mental, could be carried out and confirmed Bohr’s predictions, rather than Einstein’s. As this theory was able to resist an attempt to prove it false, it must be considered a scientific theory.

Of course, this success of the theory does not imply that it should automatically be considered correct or true. Scientific theories (always according to Popper) never become so. This theory has successfully withstood an attempt to prove it false, but the next attempt could do it.

Alan Guth

Let us look at another example: last week a posthumous article was published authored by Stephen Hawking and Thomas Hertog, who propose a new version of the Inflationary Cosmological Theory. Initially proposed by Alan Guth in 1979, and later (in the nineties) corrected by Andrei Linde and Paul Steinhardt, this theory asserts that between 10^-36 and 10^-33 seconds after the Big Bang, the cosmos experienced an inordinate expansion. Some versions of the theory argue that the end of inflation is the origin of time, and before that, there is eternal inflation. This interpretation of the theory is compatible with the inflationary multiverse and the M theory. Hawking and Hertog’s theory simplifies the passage from eternal inflation to normal expansion. 
The inflationary theory predicts that the cosmic background radiation should show traces of primordial gravitational waves, which would have arisen at the limit of eternal inflation. For now, this prediction has not been confirmed. If it were shown that there are no such gravitational waves, some versions of inflationary theory would have been proven false. Thus the theory must be considered scientific, since it is possible to prove its falsity. But if the existence of such waves were confirmed, the theory would not be proven, for the origin of those waves could have been different, although it would receive a boost, having resisted an attempt to prove it false.

Another example is dark matter, whose existence has been proposed to explain the movements of galaxies and the data provided by the cosmic background radiation, although at the moment both sources of information give rise to incompatible results when they are used to calculate the value of the Hubble constant. The theory of dark matter makes several predictions, some of which have not been proven, such as the existence of certain elementary particles, or the arrangement of dwarf galaxies around a large central galaxy, as in this recent Science News article. In spite of all its problems, or precisely because of them, the theory of dark matter must be considered scientific, since it is possible to prove it false. Perhaps this is about to be done, although we still don’t know.

As I explained in another post, the existence of the mysterious dark energy is also subject to experimentation, even in the form of computer simulations. It is true that this way of contrasting a theory is not as strong as direct experimentation, because a simulation, to be useful, must be validated, and it isn’t easy to see how to validate a simulation of the whole universe. Anyway, since those simulations could prove false the prediction of the existence of dark energy, this should also be considered a scientific theory.

Stellar paralax

When it is not possible to design an experiment to prove a theory false, this theory is not scientific. At most, it could be considered philosophical. A clear example of this is any of the six independent and mutually contradictory theories about the existence of a multiverse, about which I spoke in another post in this blog. Until now, at least, none of them has managed to propose a single experiment, not even mental, to prove them false. Therefore, the existence of the multiverse, in any of its forms, is not a scientific theory. These theories are thus far below Ptolemy’s cosmology, which was able to make predictions (for instance, about eclipses and the positions of the planets) and therefore could be proved false, something that in fact was not achieved until the nineteenth century, with the discovery of stellar parallax.

According to Popper, a scientific theory can never be considered completely confirmed, which means that we can never be absolutely sure that it is true. The confirmation of the theories is always provisional, since at any time we can find new experimental data that do not fit with the theory, as happened with Newton’s Gravitation when an error was detected in the precession of Mercury’s orbit. The problem was not solved until Einstein proposed the General Theory of Relativity to replace Newton’s, which was reduced to a less precise first approximation, enough for many applications.

The same post in Spanish
Thematic Thread on Science in General: Previous Next

Manuel Alfonseca