• Okasha gives a fairly full discussion of considerations on both sides of the issue. I’ll comment on only one point. At the end, he says that underdetermination is only an aspect of the problem of induction and that it follows that there is no special problem about unobservable entities. However, that is so only if there is no special problem of induction regarding unobservable entities, and that is dubious claim. It’s in the case of such entities that it seems that ordinary enumerative induction—i.e., generalization from instances—is not sufficient and some form of inference to the best explanation is needed. And underdetermination calls into question any claim that there is ever a single best explanation.

• Laudan’s “Grünbaum on ‘The Duhemian Argument’” is a short paper but I’m asking you to read only part of it, the first two pages or first four paragraphs, in which Laudan discusses some ideas of the physicist and historian and philosopher of science Pierre Duhem (1861-1916). Laudan does use some special symbols here (in a rather free way). The large tilde ~ indicates denial, so ~O reports the failure of an observational prediction. The arrow stands for “if … then …” or “implies,” so “H → O” says that the hypothesis H implies the observational prediction O. He uses both the multiplication dot · and the addition sign + for “and,” so, for example, the “modus tollens” argument he speaks of concludes that the hypothesis H is false on the basis of the fact that H implies the observational predication O and this prediction fails. Duhem’s criticism of the idea of a “crucial experiment” is widely accepted even by those who do not accept broad claims of underdetermination.

• It is the American philosopher Willard van Orman Quine (1908-2000) who is principally responsible for extending Duhem’s ideas to a broad claim about science and, indeed, our knowledge generally. He first did this decades before he wrote the paper we will discuss; it is a considered careful argument for these ideas. I’ll just fill in some background. First the terms “ostensive” and “ostension” (see p. 316) may not be familiar; as Quine uses them, they refer to pointing, so learning “ostensively” is learning from examples someone has pointed to. The Poincaré who is the source of an example Quine considers is Henri Poincaré (1854-1912), a mathematician who made important contributions to physics and had views on the philosophy of science that were somewhat similar to Duhem’s. (Quine’s discussion of Craig’s ideas on p. 325 gets somewhat technical at points, but it is safe to ignore it.) The German phrase on p. 327 may be translated roughly as “everything else is a human creation”; Kronecker, 1823-1891, was a mathematician who used it (or a phrase like it—the quotation didn’t appear in print and is recalled differently) to contrast the integers, as a product of God, with other numbers, which were due to humans.