Scientific generics

Generics are generalizations that are not explicitly quantified, like ‘Ravens are black’ and ‘Electrons have negative charge’. Generic sentences are not only ubiquitous in everyday conversations but are also commonly found in every scientific discipline. Several philosophers of science, however, have worried that these unquantified generalizations are prescientific and should ideally be replaced by more precise quantified generalizations. Recent research in the philosophy of language bolsters this worry that generics are not suitable to play the same epistemic role in science that quantified law-statements do. Philosophers of language have argued that generic sentences express normality claims or perhaps even give voice to psychologically primitive and biased generalizations. After all, generics have decidedly strange truth-conditions, as can be seen by the following examples:

  1. Tigers are striped.
  2. Lions have manes.
  3. Ticks carry Lyme disease.

It is in order to account for the fact that each of these three sentences is typically judged true, that some philosophers of language have argued that generics express normality claims and others have argued that they give voice to psychologically primitive generalizations. There is no simple quantificational analysis that could explain what these generics say about the world. Sentences like (2) and (3) are true despite only a minority of the kind instantiating the generalized property, yet a sentence like ‘Humans are right-handed’ is false despite the majority of humans being right-handed. As a result of their complex meaning, it is unclear which, if any, scientific inferences true generics support and how their truth can be (dis)confirmed.

The goal of this research project is to explore the role of generic generalizations in science and to respond to the worry that they are prescientific. On the view that I defend, generics do have a respectable epistemic role in every scientific discipline. I defend this view by developing two novel positions about their semantic content. Firstly, I argue that generics express kindhood generalizations. According to this semantic theory, a generic sentence says that the generalized property is part of what makes the designated individuals a kind. A sentence like (1), for instance, says that being striped is part of what makes tigers a kind. Secondly, I defend a contextualist position about ‘kindhood’, according to which what it means to be a kind depends on the epistemic context. Within a particular epistemic context, like a scientific discipline, the notion of a ‘kind’ denotes a grouping that accommodates the contextually relevant epistemic concerns. It is this context-sensitive notion of ‘kindhood’ that explains the variety in meaning between generics of various scientific disciplines, as well as their fundamental epistemic role in each of these disciplines.