Whamit!

In this research in progress, I try to analyze some properties of the syntactic behavior of bare nominals (i.e. nominals without an overt determiner) in Wolof (Niger-Congo), (1c).

 

(1)

a.

Awa defar na oto bi/yi.
Awa fix NA.3sg car the.SG/the.PL

‘Awa fixed the car/the cars.’ 

b. 

Awa defar na ab/ay oto.
Awa fix NA.3sg some.SG/some.PL car
‘Awa fixed some car/some cars.’

c. 

Awa defar na oto.
Awa fix NA.3sg car
Lit.: ‘Awa fixed car.’

 

Bare nominals in Wolof are apparently singular in that (i) they can bind a singular reflexive, though not a plural one; (ii) they cannot saturate a collective predicate, and (iii) a sentence containing a bare nominal cannot be followed up by the question how many targeting the bare nominal.

However, added plural morphology elsewhere in the sentence “turns” bare nominals plural. A plural relative clause or plural genitive agreement render a bare nominal (i) an adequate binder for a plural anaphor and (ii) an appropriate argument for a collective predicate. Conversely, plural verbal morphology has no effect in the licensing of (i) and (ii). The first asymmetry to explain, then, is why nominal-internal plural morphology in relative clauses and genitive agreement are able to “turn” a bare nominal plural, but verbal morphology does not have the same effect.

A second asymmetry that I will describe is that (iii) a how many follow-up becomes felicitous due to the addition of plural morphology not only from a relative clause or genitive agreement, but also from verbal agreement.

I will present a tentative analysis where bare nominals in Wolof come in two varieties, a singular and a plural one, number being determined by the type of Phi-P (Sauerland 2003) that tops off the bare nominal. With an additional stipulation of a featural difference between a singular and a plural Phi-P and some assumptions about binding, I will try to derive the binding and how many data. I will also comment on how the analysis falls short of accounting for the collective predicate data.