Generics with normative force (Latest draft here, under review)

Some generic generalizations have both a descriptive and a normative reading. The generic sentence “Philosophers care about the truth,” for instance, can be read as describing what philosophers in fact care about but can also be read as prescribing philosophers to care about the truth. Similarly, the generic stereotype “Men are tough” can also be understood in a normative way, saying that men ought to be tough. According to the standard account, these normative generics are ambiguous between expressing a descriptive generalization and a normative generalization. This paper argues against the ambiguity thesis, focusing in particular on Leslie’s most recent articulation of this view. In response, an alternative view is developed according to which generics are not ambiguous but rather semantically indeterminate with respect to the type of generic relation that exists. A generic has a normative reading when an utterance of the generic can be used to convey the existence of a functional relation between the kind and property.

The strong meaning of generics (Draft available upon request)

Generics are generalizations like “Tigers are striped” and “Birds can fly.” There is a puzzling and unexplained discrepancy at the heart of the meaning of these generalizations. By default, generics like these have stronger meaning than what is required for a generic sentence to be true. “Tigers are striped” conveys that one can expect a particular tiger to be striped and that when a tiger is striped, it is by virtue of being a tiger. Other generic sentences, however, are true even though there is no such inferential or explanatory relation between the kind and the property. This paper develops a new integrated semantic-pragmatic theory that accounts for the strong meaning of many generics, and explains the difference in meaning with other generic sentences. According to this theory, all generics have weak disjunctive truth conditions, saying that at least one of three conditions is fulfilled. People are cognitively biased, however, and assume these three conditions will co-occur, provided this is compatible with the nature of the kind and the property. As a result of this commonly shared assumption, many generics conversationally implicate that several conditions hold.

A kind theory for generics (Draft available upon request)

Bare plural generics like “Tigers are striped” and “Ravens are black” are one way in which our language allows us to characterize kinds. Yet according to most semantic theories about generics, the fact that these sentences describe kinds (or at least their members) has no bearing on their truth-conditions. Although unpronounced, generics are thought to hide ‘normality’, ‘probability’, or ‘typicality’ quantifiers in their logical form. In this paper, I argue that such semantic analyses of generics are unsuccessful. Instead, I offer a metaphysical theory about the nature of kindhood and argue that this theory also provides us with the truth-conditions for generics, since generics characterize kinds. On the metaphysical view that is defended, kinds are clusters of properties that are held together by at least one (but often several) of three relations; a probabilistic, causal, or functional one. A generic sentence, furthermore, says that the predicated property is part of the cluster of properties constitutive of a kind. It follows that generic sentences can be true because a probabilistic, causal, or functional relation exists between the predicated property and the other properties constitutive of the kind.

In what sense are mental disorders natural kinds? (Latest draft here, under review)

This paper argues that the difficulty in determining whether mental disorders are natural kinds is partly due to the ambiguity of the phrase ‘natural kind’ itself. This phrase is used to talk about several different phenomena. On the account defended in this paper, ‘natural kind’ is ambiguous between several combinations of three more basic concepts, namely that of a kind, that of a representation corresponding to reality, and that of a phenomenon belonging to the domain of the natural sciences. By using the phrase ‘natural kind’ for various conjunctions of these basic concepts, the phrase is not only ambiguous but also less precise than desirable. This paper argues that it would be less ambiguous and more precise for philosophers of psychiatry to ask whether mental disorders are kinds, whether their representation corresponds to reality, and whether they belong to the domain of the natural sciences.

The history of natural kinds (Full text here)

This is the first chapter of my PhD dissertation on ‘The scientific classification of natural and human kinds’. This chapter provides an overview of the history of the philosophical concept of ‘natural kinds’, and develops a particular interpretation of the debate between Whewell and Mill on natural classification. Perhaps this will some day be rewritten as a journal article.