Bunge characterises scientific knowledge as rational, systematic, precise, verifiable and therefore fallible, and separates the formal sciences — logic and mathematics, which construct their own objects — from the factual ones, which have to be confronted with the world. It is a short pamphlet, written with a clarity the genre has largely lost since.
What I took from it
The word fallible is the one that stayed. A statement that cannot turn out false cannot turn out true either: it says nothing. I sell a maturity diagnostic, and that word commits you. So every claim I put in a report now has to be able to name how it would be shown wrong; if it cannot, it comes out, or it gets rewritten as what it is, an impression.
Out of that came a sorting habit I use when gathering information: separate what I observed working, what somebody told me, and what I am assuming. Three different things, and in most consulting documents they appear mixed inside the same paragraph, so that nobody coming back six months later can tell which was which.
The formal-versus-factual distinction has a use Bunge did not have in mind: a model can be flawless inside and say nothing about the world until somebody checks it against the world. That is exactly what happens to a spreadsheet full of return-on-investment arithmetic — the sums are right and the assumptions were never tested.
What I left out
It is a text from 1960 and it shows: the faith in progress, the vocabulary, the confidence with which it dismisses what it dislikes. The polemical passages against pseudoscience convince the already convinced. Read it as what it is, a short popular essay, rather than the treatise on philosophy of science it never claims to be.