Finally, I’ll offer a brief remark concerning Martin’s positive suggestion for reifying levels. Martin’s central idea—that we often pick out things as being on a level because they are detectable with the same or a similar kind of apparatus or with the same or a similar set of procedures— seems to me correct and important. Among the many possible ways of decomposing a spatio-temporal whole into component parts, some of these correspond to items that for one reason or another are readily detectable as such by one or more experimental apparatus. This way of putting things highlights how pragmatic factors, such as available apparatus, determine what we will take to be an appropriate “landing” in a hierarchy of levels. And it raises useful questions about, for example, when two different apparatus, or two different tasks, or two different procedures in fact target the same items, or items at the same level. On a fine-grained characterization of our experimental instruments, no two experiments are the same. On a course grain, even superficially quite distinct experiments can be targeting the same phenomenon (consider, e.g., implicit bias or spatial memory). As noted above, it seems to me that many other epistemic, theoretical, and psychological factors enter into these decisions as well.
This good point, however, is obscured by an ontology in which properties apparently come into existence during acts of detection. Martin’s view is that objects have disposition profiles, and these profiles are turned into properties when they are measured. It is unclear to me, however, why one would not want to say that properties are there to be detected all along, or perhaps that properties just are the dispositional profiles of things. Martin doesn’t do much to motivate this experimental idealism about properties, but the thought seems hard to motivate, at least for macroscopic phenomena. It is an apparent consequence of Martin’s account that levels don’t exist until they are detected. And it is an apparent consequence of his view that new levels come into existence when we develop new instruments to detect them. And, to reiterate the thought in the last paragraph, we will have to wrestle with the question of when two techniques detect the same thing or different things. Martin’s positive proposal makes this question much more pressing, given that, for Martin’s view, the structure of the world– specifically, the distribution of properties in space and time– apparently hangs in the balance. But if we abandon the idea that the placement question must have a uniquely correct answer, these questions are less pressing for thinking about ontology; nonetheless, questions about how we coordinate different experimental tasks, protocols, and procedures are at the heart of the epistemological challenge faced by any experimentalist (see Sullivan 2009, 2010).