Abstract

Mary Hesse’s work on the role of analogical reasoning in science has set the disciplinary standard. Her classic Models and Analogies in Science (1966) spearheaded the emergence of the philosophy of scientific modelling and enjoys lasting prominence in the field. Hesse’s account famously divided analogy into three kinds: positive, negative, and neutral. In this chapter we aim to bring attention to the one element that has been relatively least explored, the negative analogy. We draw on an example from medical practice, namely the use of the Neer Classification system in the medical prognosis and treatment of shoulder fractures. We argue that there are externally and internally functioning negative analogies, and that, importantly, the latter have a ‘positive’ inferential role. We describe three types of externally functioning negative analogies that have already been to some extent discussed in the literature—which we refer to as ‘crucial’, ‘irrelevant’, and ‘requisite’—as well as a new fourth type that we call ‘scope-fixing.’ We then apply Hesse’s distinction between horizontal and vertical relations to the clinical case study and suggest that there are two different ways in which internally functioning negative analogies can serve as premises in analogical inference.