A well-respected exposition on the problems facing scientists (and philosophers) contemplating models of scientific theories. A third of the books is a dialogue between a Campbellian (proponent of models as more than psychological aids, as lending predictive power) and a Duhemist (formalist, arguing for models as merely psychological aids). The remainder analyses types of material analogies, introducing four types: A) According to common properties (eg Earth and Moon); B) A scientific analogy (eg properties of sound and those of light); C) A classification analogy (eg Birds and Fish); D) A political rhetoric analogy (eg father to son and state to citizens).
Mary Hesse’s most prominent contribution to the language of metaphor analysis is that of positive, negative, and neutral analogy, where positive analogies are defined as properties verifiable shared between the explicandum and the explanans, negative as those verifiably not shared, and neutral are the most interesting: they provide the precious unknown territory from which we can draw predictive power.
The last third of the book revisits Aristotle’s thoughts on analogy and is rather dry.
Overall, no literary masterpiece, but a useful and much cited work indeed.
An interesting consideration of certain modes of scientific thinking—but if you're not used to (or in love with) what I'll call the style of mid-twentieth-century analytic philosophy, it'll probably come across as pretty dry.