• Sections 1–2 serve to locate Sugden’s topic with respect to other thinking about economics. Comparing what he says about Friedman on pp. 622f with what Gibbard and Varian say (in their §II on p. 671) can give you some sense of the relation of Sugden’s discussion to theirs.

The concept of the “hypothetico-deductive” that he mentions in §2 refers to a common method in the sciences according to which hypotheses are tested by deducing observable consequences and comparing these to observations (which may be observations of the results of experiment). The deduction would follow the pattern Hempel saw in explanation and prediction, and the hypothetico-deductive method can be described as the testing of hypotheses by checking the accuracy of their predictions of observations.

This method can be opposed to an “inductive” method according to which scientific investigation begins with observations rather than hypotheses, and laws are discovered by analyzing observations rather than framing and testing hypotheses. Although the two methods might be combined, it was common in the first half of the 20th century to argue that the hypothetico-deductive method was the only really methodical approach to scientific investigation, and Sugden’s quotation from Friedman is in keeping what many others said at the time.

• In §3, Sugden introduces the concept of “exhibits” that will be his topic for the rest of the paper. You can find a definition near the top of p. 624, but his examples may be just as useful for getting a sense of what he has in mind.

Notice in particular that exhibits are not the single observations that are contrasted with theoretical hypotheses in the hypothetico-deductive or inductive methods described above. They bear some resemblance to what are sometimes called “empirical generalizations” in discussions of the natural sciences. These differ from theoretical hypotheses only by degree, in departing relatively little from the vocabulary employed in direct reports of observations.

• The idea of “refinement” introduced in §4, is one of the forms of investigation Sugden sees that can focus on exhibits without reference to theoretical hypotheses. It, too, has an analogue in the natural sciences: it has been noted by historians of science that one common response to observations that don’t fit well with accepted theories is to work not only to be sure that they can be reproduced (and are not one-time flukes) but also to develop good ways of reproducing them.

• The term ‘construct’ used in the title of §5 has a special meaning that Sugden explains on the next page (p. 630). Its use in psychology derives from its use by the psychologist Paul Meehl (1920–2003) to describe ideas that could not be defined directly by observational tests (i.e., could not be given simple “operational definitions”). That could suggest a similarity to the terms that appear in theoretical hypotheses; but, as Sugden uses the idea, constructs acquire their meaning from ties to experiment rather than from a role in stating theoretical hypotheses. Notice how he presents constructs as a sort of alternative to such hypotheses.

• Sugden’s brief concluding §6 seems to suggest that exhibits might be seen not merely as an alternative to theoretical hypotheses but as a replacement for them as the central focus of science. (You might recall here Dore’s suggested “piece-meal” approach to sociology.) Do you think anything would be lost if “our knowledge of the world” (to use the phrase in Sugden’s final sentence) was reorganized in this way?