Current influential normative accounts of analogical inference—like Hesse’s model, Bartha’s articulation model, and Gentner’s structure-mapping model—assume that an analogical inference’s plausibility is determined by it having two essential characteristics, which I will call Source Domain Salience and Overlap. I show that these two characteristics are neither sufficient nor necessary for plausibility. First, these two characteristics do not capture all of the criteria that scientists use to eva…
Read moreCurrent influential normative accounts of analogical inference—like Hesse’s model, Bartha’s articulation model, and Gentner’s structure-mapping model—assume that an analogical inference’s plausibility is determined by it having two essential characteristics, which I will call Source Domain Salience and Overlap. I show that these two characteristics are neither sufficient nor necessary for plausibility. First, these two characteristics do not capture all of the criteria that scientists use to evaluate analogical inferences. Second, some analogical inferences can be plausible because they lack these two characteristics. Using examples from physics, archaeology, and astrobiology, I show that an analogical inference’s plausibility depends on a further essential characteristic, Target Domain Salience, that is not captured by Hesse, Bartha, and Gentner’s two essential characteristics. I argue that a successful normative account of analogical inference should establish criteria for this further characteristic.
