What Is Epistemology? How Philosophy Looks at Science

One term, different objects

The word “epistemology” joins two Greek roots: episteme (knowledge) and logos (account, reasoning). That’s how the Stanford Encyclopedia of Philosophy frames it, describing the discipline as the study of cognitive success, that is, of what makes an act of coming to believe something succeed or fail. Treccani credits the first use of the term to the Scottish philosopher James Frederick Ferrier, and adds a detail that explains a lot of cross-language confusion: in Italian usage, “epistemologia” originally names the part of the broader theory of knowledge devoted to the foundations and limits of scientific knowledge. In English-speaking countries the same word covers wider ground, equal to the whole theory of knowledge: an Italian essay on “epistemologia” is almost always about science, while an English chapter on “epistemology” can just as easily cover everyday knowledge.

A reader used to the Italian sense who picks up an English-language book on the subject finds a much broader field than the title would suggest.

The split between the forms of argument science itself relies on traces back to inductive and deductive reasoning: the difference with historical examples, and the twentieth-century push to put the logical analysis of science at the center of the debate runs through logical positivism.

What counts as knowing, according to the classic definition

The Stanford Encyclopedia of Philosophy and the Internet Encyclopedia of Philosophy both revisit the traditional analysis of knowledge, broken into three pieces that all have to hold at once. A person has to believe something is true; it has to actually be true, because a false belief can never count as knowledge; and a third ingredient is needed, justification, which the Internet Encyclopedia of Philosophy describes as the result of arriving at that belief the right way.

Practical example: a horoscope predicting “you’ll have a demanding but manageable day” never clashes with any observation, since any day at all confirms it. A forecast like “it will snow in Turin tomorrow” can instead be proven wrong by a clear sky: it’s the kind of statement that, later in this Recap, Popper’s criterion classifies as scientific.

This three-part scheme has a known weak point, the Gettier problem: the Internet Encyclopedia of Philosophy describes cases where a belief is true and properly supported, and yet philosophy hesitates to call it full-blown knowledge. The Stanford Encyclopedia of Philosophy ties the same knot to the difficulty of explaining why a cognitive act truly succeeds, beyond sheer luck.

Readers who trace this thread back to its modern roots find common ground in Kant’s Critique of Pure Reason and in David Hume’s empiricism, which questioned how far past experience can really guarantee what happens next.

Skepticism, the most radical doubt

If knowing requires belief, truth and adequate support, a more radical question stays open: do we ever have knowledge at all, in any field? The Internet Encyclopedia of Philosophy defines philosophical skepticism as the position that there is nothing we know, or can know. It isn’t a side issue: the Internet Encyclopedia of Philosophy assigns epistemology two parallel tasks, determining what knowledge is and determining how far what we actually know extends.

The line with pseudoscience

Telling science and pseudoscience apart belongs to the same family of questions, aimed at a different target: what makes a body of knowledge scientific, beyond simply sounding confident? The most discussed answer is the demarcation criterion Karl Popper proposed, reported by the Stanford Encyclopedia of Philosophy: a statement or system of statements is scientific only if it can, in principle, conflict with possible or conceivable observations. The criterion concerns only the logical structure of the statements, not their specific content, which is why an idea considered metaphysical at one stage can later become a testable scientific theory — for Popper, this happens often. Popper adds a clause about how scientists behave, beyond the shape of the theories themselves: a doctrine loses its scientific status once its supporters refuse to accept falsification when it shows up against the facts.

The Stanford Encyclopedia of Philosophy flags a limit of the criterion, though. Several pseudosciences were testable in principle, the facts have already contradicted them more than once, and their supporters keep defending them anyway. The mere theoretical possibility of being refuted separates testable statements from the rest; it doesn’t explain why a theory the evidence has already rejected stays in circulation.

The falsifiability criterion meets actual history in Galileo Galilei and the Scientific Revolution, while the drive to make the method of the natural sciences the model for all knowledge defines positivism.