Theses Doctoral

Defining Marks: A Defense of the Predicate View of Proper Names

Moss, Matt

At the start of the last century philosophical consensus was that names were more or less like descriptions, and, at its end, that names were utterly unlike descriptions. The former view, Classical Descriptivism, had it that each individual’s name was its name in virtue of the individual uniquely fitting some implicit characterization. Names were thus believed to have structure at the level of content: they expressed properties an object can have or fail to have. This view was in turn challenged beginning in the 1970s, most notably by Saul Kripke. Kripke’s claim was that an individual’s name has no structure at the level of content: it simply stands for a given individual. A name cannot characterize anything, and has no “meaning” save what it names. Kripke’s view, Referentialism, in turn became the new orthodoxy.
In my dissertation, I challenge the arguments that have lead us to believe names and descriptions are expressions of two different kinds. But I do not vindicate the old orthodoxy. I chart a middle path between Classical Descriptivism and Referentialism that can recapture many virtues of the former view, while respecting the linguistic data that lead to its abandonment. I do this in defending a competing theory, one that has recently grown in prominence: the Predicate View of names. The Predicate View offers a radically different conception of what a name is, one tied neither to an individual referent (as with Referentialism), nor to some set of properties an individual might uniquely bear (as with Classical Descriptivism). Instead, on the Predicate View a name such as “Bambi” expresses a property, bearing-“Bambi”, satisfied by all and only Bambis. To fully substantiate this approach requires an investigation of how definiteness — a linguistic marker of something being unique relative to some context, as when “the cat” refers to some specific cat thanks to “the” — is realized cross-linguistically, and how this bears on the way a name like “Bambi” successfully picks out some particular Bambi. I take the proper formulation and defense of the Predicate View to be a preliminary contribution to such an investigation. What it promises is a more refined understanding both of how language expresses thoughts about individuals, and how this language is related to the language of properties, i.e. ways individuals can be.
Names are not proprietary to individuals on the Predicate View. They express shareable properties, they have structure at the level of content, and they have a meaning, which can be characterized schematically: for any name “N”, its meaning is given by bearing-“N”. The Predicate View does not assume the tight connection between name meaning and name denotation than both Referentialism and Classical Descriptivism do (albeit in different ways). The name “Bambi” corresponds to some set of individuals that satisfy bearing-“Bambi”, but does not “refer” to that set or indeed any member of it. Which individuals gets referred to with the name “Bambi” by speakers is a distinct matter, and the denotation of a name is to be understood in terms of acts of referring. In this respect, the Predicate View differs fundamentally from both Referentialism and Classical Descriptivism. Indeed, it differs more from both of these accounts than they do from one another. The Predicate View assumes a very different structure for name bearing, i.e. how names themselves are individuated and how names are related to what they name.
I show in Chapters 1 and 2 that the semantic behavior of names — especially with respect to time and modality — provides evidence that the Predicate View gets the structure of name bearing right. I argue further that data which were taken to support Referentialism are equally well explained by the Predicate View. This runs counter to a common assumption — that the Predicate View faces a serious problem with modality, since it cannot deliver the result that names are rigid. I show that, on the contrary, the Predicate View offers a more nuanced and explanatory account of name rigidity than Referentialism. The Predicate View also explains a neglected fact that Referentialism cannot: that there are non-rigid occurrences of names. The picture that emerges is one on which names are predominantly rigid, but where they occur non-rigidly as a result of certain presuppositions being satisfied. I conclude that we should abandon Referentialism and embrace the Predicate View.
In Chapter 3, I defend the View against a challenge due to John Hawthorne and David Manley in The Reference Book. There they argue that the most dramatic data favoring the Predicate View may have nothing to do with names at all, being adequately explained by an all-purpose mechanism of metalinguistic ascent. Why not say that in constructions like “I know three Caitlins” we are quantifying over what are strictly speaking ambiguous names (in the Referentialist’s sense), rather than revise our semantics in the way the Predicate View proposes? I argue that in fact on a very natural elaboration of the basic idea behind the Predicate View, the cases of “metalinguistic ascent” that Hawthorne and Manley have in mind are naturally explained on the Predicate View itself, blunting their dialectical force.


  • thumnail for Moss_columbia_0054D_13914.pdf Moss_columbia_0054D_13914.pdf application/pdf 1.08 MB Download File

More About This Work

Academic Units
Thesis Advisors
Peacocke, Christopher
Ph.D., Columbia University
Published Here
July 24, 2017