Jenna Steiner

Date of Award

Spring 2019

Document Type

Open Access Dissertation



First Advisor

Mila Tasseva-Kurktchieva


This study aims to identify whether the acquisition of the English copula by Arabic- speaking learners of English provides evidence for a performance or representational- based account of errors. The representational theory tested in this study is the Interpretability Hypothesis (Tsimpli & Mastropavlou, 2007) which proposes that language learners have only partial access to Universal Grammar (UG), making some language structures unacquirable for second language learners. The performance theory tested in this study is Feature Reassembly (Lardiere, 2008), which proposes that the source of errors lies with the mapping of features onto morphology rather inside the core computational component of grammar.

In Arabic, there is no overt copula in present tense sentences; however, in past tense, the copula is overt. This stands in contrast to English in which an overt copula is required in both present and past tenses. Therefore, copula omissions in English committed by Arabic speakers learning English can be analyzed in order to determine whether a representational or performance-based theory best accounts for the omissions. Previous research looked exclusively at copula omission in written production only under the condition of tense and framed results only in light of L1 transfer without considering the role of access to UG. This study builds upon previous research by providing a more thorough analysis of copula omission and errors than previous studies and by framing these results in a more modern theoretical context of performance versus representational deficit.

The participants of this study consisted of 45 Arabic-speaking students studying in English language programs at large public U.S. universities. The participants were divided into two proficiency groups, with 24 participants in the “High” proficiency group and 21 participants in the “Low” proficiency group. Data were collected via a time- pressure grammaticality judgement task and an elicited imitation task. With these two tasks, the effects of tense, verb type, and number were investigated across several syntactic environments. The conditions of tense, verb type, and number are analyzed statistically, and a descriptive analysis is presented for syntactic environments and non- omission copula errors.

Feature Reassembly is best able to account for the data of this study. The Low Group performed better on past tense items than present tense items, better on main verb items than auxiliary items, and better on the singular items than the plural items. All of these results are predicted by Feature Reassembly as constrained by Slabakova (2009). Moreover, these deficits were overcome by the High Group, which is also predicted by Feature Reassembly. Additionally, the descriptive results of this study such as the omission of -ing, the doubling of the copula, and the tense and agreement errors on the copula are also predicted by FR. This all suggests that copula errors by highly proficient Arabic speakers learning English are performance-based, not representational, in nature.

Included in

Linguistics Commons