Pierre Duhem, Popper’s predecessor

Karl Popper

In a previous post in this blog, I mentioned how Karl Popper defines what is, and what is not, a scientific theory:

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

I also said there that, according to Popper, a scientific theory can never be considered completely confirmed. In other words, we can never be absolutely sure that it is true.

After writing that post, which I published almost one year ago, I have discovered that these two Popper’s fundamental ideas had been anticipated by Pierre Duhem, to whom I dedicated the previous post in this blog.

Popper detailed both these ideas in his book The Logic of Scientific Research, published in German in 1934 and in English in 1959, both versions written by himself. Duhem, however, had anticipated them in 1913, in a letter addressed to Réginald Garrigou-Lagrange, which the latter published in his book Dieu. Son existence et sa nature (1914). By then, Popper was about 12 years old. Did Popper read Duhem? Perhaps not this letter, although other works certainly, since in The Logic of Scientific Research he quotes Duhem five times, usually to show his discrepancy. The problem is that misunderstanding Duhem is quite easy. To avoid it, one should probably read his whole work, which not everybody can do.

Pierre Duhem: Realist or anti-realist?

Pierre Duhem

Pierre Duhem (1861-1916) can be considered one of the last nineteenth-century physicists. Specialist in Thermodynamics, the branch of physics that dominated the second half of the nineteenth century, introduced the idea of the ​​chemical potential at the same time as William Gibbs, expressed in the Gibbs-Duhem equation, which connects the chemical potential with magnitudes such as the volume, pressure, entropy and temperature of a chemical mixture. He is considered one of the creators of physical chemistry and was nominated twice for the Nobel Prize in Physics.

In his scientific work, Duhem confronted Marcellin Berthelot, whose principle of maximum work he opposed. This led to his doctoral thesis being rejected, and he was denied a teaching position at the University of Paris. Duhem was finally shown to be right, as Berthelot’s principle is not generally applicable, for it has many exceptions.

Anti-realist answers to the no-miracles argument

Hilary Putnam

The previous post described the no-miracles argument, proposed by Hilary Putnam. The article ended thus:

What do anti-realists answer to this argument? Are they convinced?

I guess the readers have deduced that the answer to the second question must be negative, otherwise the debate between realism and anti-realism would have ended. Let us look, therefore, at the answer to the first question. Faced with the abductive argument of no-miracles, anti-realists answer in two different ways:

1.      Bas van Fraassen is an anti-realist American philosopher who criticizes Putnam’s argument, arguing that scientific theories are successful because unsuccessful theories have been eliminated by natural selection (i.e. scientists have ruled them out). Therefore, asking why science is successful is similar to asking why basketball players are tall: because they have been selected. Let us see how Fraassen describes his theory, which is called constructive empiricism:

Abduction and the no-miracles argument

The Cheshire cat,
famous invisible cat

In an earlier post in this blog, I explained with an example the mode of reasoning based on abduction. Although not as strong as deduction and induction, abduction reaches high degrees of confidence in fields such as history, art criticism and others, less scientific than mathematics or natural science.

In another post published in March 2016, I described the fallacy of the invisible cat, which confuses a sufficient condition with a necessary condition for something to happen. This situation occurs when there are several possible causes that may have given rise to the same phenomenon.

In some cases, if we apply abduction to a situation where the fallacy of the invisible cat could occur, a conclusion can be reached. Think of the example I proposed to describe this fallacy: