But is it science?

Recently there has been a succession of scientists, some quite famous, claiming that philosophy is dead and has been replaced by science, with particular disdain for the branch of the discipline named Philosophy of Science. (I saw a comment by a university professor where he said that philosophy had absolutely no relevance to his teaching of the scientific method. I think he was missing the point completely.) One criticism of philosophy of science is that the philosophers try to say what scientists do and this doesn't describe what actually goes on in laboratories and research establishments. This is more missing of the point, because the philosophy is really about what science is as distinct from other forms of human activity rather than how it is actually done.

This is not the place for a robust discussion of philosophy of science because, as a famous mathematician once said, the margin is too narrow to contain it. One principle, however, which helps to define what science is and how can be distinguished from pseudoscience or nonsense is Popper's idea of falsification. This is often misconstrued as meaning that Popper suggested that scientists spend their time trying to falsify existing hypotheses or theories but that's nonsense. What the principle really means is that for a theory or hypothesis to be considered part of science it must be possible to think of an experiment or a test which can prove it to be false. Put another way, you might not be able to prove that something is universally true but there must be the possibility that it could be proved false.

The area of science which is most confusing for nonscientists is undoubtedly physics, where weird predictions are made from mathematics with some predictions seeming to lie well outside Popper's rules of what is and isn't science.

Sometimes these predictions can be verified by observation, although it may take some time. Johann Galle knew where to point his telescope in 1846 to locate the planet Neptune because of calculations done by Urbain Le Verrier on perturbations in the orbit of Uranus, and it was a wobble in Neptune's orbit which led to the discovery of Pluto, although it was about 30 years from the observation until Pluto was actually located and identified.

The existence of the Higgs boson was first postulated in 1964 but anyone watching the telecast from CERN in March 2013 would have seen the emotion on Peter Higgs' face when it was announced that the results from the Large Hadron Collider had severely increased the probability that his prediction was correct, the famous "5 sigma" result. (The mainstream media reported this as "Higgs boson found", but the actual statement said that it "strongly indicates that it is a Higgs boson", the correct scientific position where things can be known to a very high probability but never certainty.)

It might have taken 30 years to locate Pluto and almost 50 years to pinpoint the Higgs boson, but took a century from Einstein's predictions before gravitational waves were detected at the level of probability that satisfied scientists.

What is common to these three predictions is that even though they were based on mathematical calculations they all had the possibility of being tested when advances in technology and engineering allowed it. Telescopes had to get better to detect faint and small objects at the fringes of the solar system. Massively complex and expensive test equipment like the Large Hadron Collider and the Laser Interferometer Gravitational-Wave Observatory had to be invented, built and tested to look deep inside the atom or to detect the effects of colliding black holes.

Then there are some elegant mathematical calculations which seem to explain some of what we know but require hypotheses which can never be tested. As one example, I saw a prominent physicist who claimed to have solved the problem of what happened before the Big Bang. Relying on the fact the time can run backwards (there are no universal laws which would be broken if physical or chemical reactions were run backwards in time) he postulated that space-time is like two cones touching at their points, with time running in the opposite direction before the Big Bang. People living there would see a finite past and an infinite future just as we do. There is no possible way that this could be tested to be true or false.

Another example is the idea of multiverses. The mathematics that supposedly requires this may be solid, but as by definition there can be no communication between any of the multiple universes it is again a hypothesis which is impossible to test and falsify. It may very well be true but we will never know.

And as for the idea that we live in a computer simulation like something in a bad science fiction novel, that just seems to replace the theist's creator God with a creator programmer. I think I can reject that with a 5-sigma probability.

This article was published as the Naked Skeptic column in the January/February 2019 edition of Australasian Science