Whamit!

The Weekly Newsletter of MIT Linguistics

Syntax Square 5/15 - Suzana Fong (MIT)

Speaker: Suzana Fong (MIT)
Title: Bare nominals in Wolof
Date and time: Tuesday May 13, 1-2pm
Location: 32-D461
Abstract:

Even though Wolof is a language with overt determiners, it also allows for nominals to occur without any determiner at all.

(1) Awa defar na oto bi/yi/Ø.
Awa fix NA.3sg car the.SG/the.PL/Ø
‘Awa fixed the car/the cars/a car.’

In this research in progress, I try to explore the syntax and semantics of these bare nominals. The properties considered so far are the following:

(2) Syntactic properties

  • Bare nominals can the be the subject of transitive and intransitive verbs.
  • They can also be the object of transitive verbs.
  • When in subject position, bare nominals are compatible with singular or plural verbal agreement.
  • They are also compatible with a singular or plural genitive suffix.
  • They are compatible with singular or plural relative clauses.
  • They can be referred back to by a singular or plural clitic.

(3) Semantic properties

  • Bare nominals cannot be followed up with the question How many? when they are the subject of a singular verb, though this is possible when the verb is plural. Something along these lines holds of bare nominals in object position too.
  • They can be the antecedent of a singular reflexive, though not of a plural one.
  • They cannot be the argument (subject or object) of a collective predicate.
  • They can occur in the existential construction Am na… (‘there is…’), which displays definiteness effects.
  • They can scope above or below intensional predicates and an iterative verbal affix.

In order to account for these properties, I tentatively draw a distinction between syntactic and semantic number that is loosely based on Landau (2001)’s analysis of partial control. Specifically, I try to explain the semantic properties in (3) by suggesting that bare nominals in Wolof are semantically singular. However, they would have no syntactic number feature. Coupled with a fallible (Preminger 2014) version of feature checking (Chomsky 1995), this proposal could be consistent with the syntactic properties in (2).