Skip to main content

Augustina Owusu Passed Her Dissertation Defense

Congratulations to Augustina Owusu, who passed her dissertation defense on Friday, Nov. 20th!!

Title“Cross-Categorial Definiteness/Familiarity”

Chair: Simon Charlow

Committee: Viviane Deprez, Veneeta Dayal (Yale), Malte Zimmermann (External, Potsdam University)


This dissertation examines the semantics of the definite determiner nó in Akan. The Akan definite determiner occurs in the DPs as usual and occurs outside the nominal domain in simple declarative clauses, among other clausal contexts. I argue that across categories, nó encodes a familiarity presupposition — it presupposes that a discourse referent with the descriptive content of its complement exists in the discourse. I propose that nó is a cross-categorial definite determiner. Specifically, I argue that it takes both NP and TP (and other propositional nodes, including NegP) complements. This dissertation contributes to the emerging literature on cross-categorial definite determiners (Lefebvre 1998; Larson 2003; Wespel 2008; Renans 2016,2019; Korsah 2017).

In the nominal domain, I propose a reanalysis of the definite determiner. Specifically, I propose that nó encodes two presuppositions: a familiarity and a non-uniqueness presupposition. The familiarity presupposition captures the determiner’s use in Hawkin’s (1978) anaphoric and immediate situation contexts.

I adopt the notion of weak familiarity by Roberts (2003) to characterize the kind of familiarity encoded by nó. The non-uniqueness presupposition proposed by Robinson (2005) and Dayal & Jiang (2020) for demonstratives captures the definite determiner’s inability to be licensed by inherently unique nouns such as sun, president of Ghana, and superlatives in out-of-the-blue contexts. Further, in anticipation of the non-nominal use of the determiner, I posit that the meaning of nó does include iota. I contend that iota-shift is available (Partee & Rooth, 1983; Partee, 1986; Chierchia, 1998).

Also, in the nominal domain, the dissertation presents novel empirical evidence supporting the specific indefinite analysis of the Akan indefinite marker bí. I show that contrary to Arkoh (2011), bí does not take obligatory wide scope in all contexts. The determiner is ambiguous between opaque and transparent readings in intensional contexts and has “narrow scope” readings in conditional sentences. I propose that bí is a skolemized (parameterized) choice function with individual and world variables. The world variable gets bound on the opaque reading and remains free on the transparent reading. The semantics of the definite determiner proposed captures the fact that the definite and indefinite determiners can co-occur.

In the non-nominal domain, I provide evidence that nó clauses in Akan are propositions. I propose that nó-clauses are definite propositions, where definiteness corresponds to a familiarity presupposition. A proposition is familiar so long as it has an antecedent in the context. The non-nominal analysis extends to DP-headed relative clauses, a domain that straddles the nominal and clausal domains. Because DP-headed relative clauses are DPs, they license the nominal determiner. Simultaneously, since relative clauses are clauses, they also license the clausal determiner.

Congratulations, Dr. Owusu!!!!
We wish you all the best as you continue on your journey.

Augustina zoom defense