The Stage/Individual Distinction and (In)alienable Possession

The stage-level/individual-level distinction, which has so far been limited to verbal and adjectival predicates, should be extended to (underived) nominal predicates as well. Specifically, while simple nominals are individual-level predicates, event nominals and inalienable possession nominals are stage-level predicates. The currently prevailing distinction between the stage-level and individual-level predicates, that is, whether a predicate denotes a transitory state or an inherent and unchangeable stage, must be reconsidered. Current suggestions in line with this move will be discussed.*

1. Two types of predicates

Kratzer (1989, 1996) proposes dividing predicates into two types: stage-level predicates and individual-level predicates. While the individual-level predicates express inherent or enduring properties of their subjects, the stage-level predicates denote transient properties of their subjects. For verbal and adjectival predicates, several diagnostics have been proposed to distinguish stage-level predicates from individual-level ones:

1. 1.

   1. a. Only stage-level predicates can be modified by spatio-temporal modifiers.

   2. b. Only stage-level predicates can introduce a variable to be bound by adverbs of quantification. (Kratzer 1989, 1996)

   3. c. Only stage-level predicates can occur as secondary predicates. (Rapoport 1991)

My main goal in this article is to show that the syntactic distinction between stage-level and individual-level predicates should be extended to nominal predicates. Since Abney (1987), a number of linguists have argued that noun phrases involve the same subject-predicate relationship as clauses: the NP selected by D is a predicate; a subject in the spec of DP and the NP complement of the D enter into a predication relation, just as a subject in the spec of TP and VP do (Stowell 1987, Safir 1987, Vergnaud & Zubizarreta 1992). Given their insight, it is not surprising to find a stage-level/individual-level distinction among the nominal predicates derived from verbs or adjectives. Endo’s (1995) proposal to extend the distinction to adjectival nouns in Japanese is reasonable for the same reasons. The originality of my proposal is that the distinction also holds in noun phrases headed by an underived nonadjectival nominal. More specifically, I propose that while inalienable possession (IAP) nominals are stage-level, simple (alienable possession) nominals are individual-level. Under this proposal, noun phrases headed by IAP nominals are expected to behave like those headed by event nominals. This proposal will be motivated on the basis of the three diagnostics noted in 1 and by consideration of the multiple object construction in Korean, possessor raising in Japanese, the multiple genitive construction in Japanese, and the quantificational possessive construction in English, among others. Given this proposal, the currently prevailing distinction between stage-level and individual-level predicates—whether a predicate denotes a transitory state or an inherent and unchangeable state—should be reconsidered.

2. Preliminary assumptions

To begin the discussion I introduce two motivations for the presence of a syntactic entity (specifically, the event place) that distinguishes stage-level predicates from individual-level predicates: the possibility of predicates being bound by adverbs of quantification (Kratzer 1989, 1996) and the possibility of secondary predication (Rapoport 1991).

2.1. Motivations for the event arguments

Kratzer (1989, 1996) proposes that the stage-level/individual-level distinction is represented in the argument structure of each predicate, and that stage-level predicates have an argument position for spatiotemporal location, whereas individual-level predicates do not have this position.¹ Following Heim (1988), Kratzer first assumes tripartite quantifier structures consisting of a quantifier, its restriction, and the nuclear scope, and assumes further that the adverb of quantification must be indexed with free variables in its restrictive clause. This means that the adverb of quantification binds all free occurrences of these variables in its entire scope and that failure of this binding relation leads to a violation of the prohibition against vacuous quantification. With this in mind, consider the asymmetry in 2.

1. 2.

   1. a. When a Moroccan knows French, she knows it well.

   2. b. *When Mary knows French, she knows it well.

   3. c. When Mary speaks French, she speaks it well.

Ex. 2a is ruled in since the nonovert operator always, which restricts the domain...