Dissertations, Theses, and Capstone Projects

Date of Degree

9-2026

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy

Program

Linguistics

Advisor

Jason Kandybowicz

Committee Members

Jon Nissenbaum

Christina Tortora

Subject Categories

Language Description and Documentation | Linguistics | Semantics and Pragmatics | Syntax | Typological Linguistics and Linguistic Diversity

Keywords

locality, Ā-movement, Grassfields Bantu, syntax, focus, topic

Abstract

Universal characteristics of human language have fascinated scholars for millennia, a trend that continues in various schools within the modern discipline of linguistics. Potential universals indicate that learners across speech communities converge on very similar rules despite the fact that the linguistic input made available to them through communicative interactions is compatible with a variety of abstract grammatical rules. Analyzing the nature of our capacity for language necessarily means determining whether the bias toward seemingly exceptionless characteristics of human languages is hardwired to a Language-specific capacity (sometimes called Universal Grammar) or whether only general cognitive principles are at work in language learning and processing. In certain circles it is widely acknowledged that a combination of both underlies our capacity for Language, but the division of labor between the language-specific and the general is subject to debate. Theories of language universals are consequently key to our understanding of human language as a cognitive and behavioral phenomenon.

Some of the strongest candidates for language universals are types of sentences known as "islands". In language after language, islands appear to be isolated from surrounding linguistic expressions for syntactic connectivity as made evident by acceptability judgment tasks. For example, wh- questions (e.g., 'what did the child break?') are acceptable as natural questions in English, but form them across the island of relative clauses, and the result is judged to be unacceptable (* `what did you see [the child that broke?') (as a question about the broken object). For the past sixty years, the generality of island effects within and across languages has had linguists debate the Language-specific or general cognitive source of their ubiquitous limiting effect. Various counterexamples to the universality of islands were argued to obviate their degrading effect without truly violating them, giving rise to "surface variation" (Phillips, 2013) in keeping with the purported universality of islands as structures that are isolated from their surrounding linguistic material for the purpose of syntactic connectivity.

The dissertation stems from the observation that in Shupamem (a Grassfields Bantu language of Cameroon), six suspected islands do not feature the blocking effect expected of their kind. My research on the language establishes that islands in the language surprisingly facilitate syntactic dependencies that islands otherwise block across languages. Testing this hypothesis has involved elicitation-based studies with a language consultant, a native speaker of Shupamem. Based on minimally contextualized acceptability judgment tasks, the evidence for syntactic connectivity comes from nine standard and novel diagnostics applied to six suspected islands. Structural considerations and the evidence from three types of dependencies (wh-, topic, and negative) favor direct syntactic connectivity over "surface variation". Strengthened by comparative evidence from additional sub-Saharan languages without one or more islands, I conclude that islands are not universal. However, I argue that the conditions that determine the status of islands may very well be universal. Competing theories of islands (e.g., Bianchi & Chesi, 2014; Goldberg, 2013; Johnson, 2003) arguably tap into the same intuition, capturing crosslinguistically common island effects and their exceptions from sub-Saharan Africa. This conclusion has implications for linguistic theory, acquisition, and processing. As before in the history of the discipline, linguistics and the cognitive sciences clearly stand to gain from an Africanist perspective in testing and refining syntactic theory and its underlying universal forces.

Download

This work is embargoed and will be available for download on Saturday, August 26, 2028

Share

COinS