The Corpus of Irish English Speech

Abstract

The past years have witnessed the collection of various corpora for the study of Irish English (IE). Most have been developed and driven by diverse research foci with a specific aim in mind, however, at present data sources consist of unavailable and/or outdated audio files of English spoken primarily in Dublin and Belfast. Additionally, a very limited number of investigations on the prosodic features of IE varieties have been conducted to date. As a result, a comprehensive overview of IE prosodic diversity is still missing and existing speech corpora do not allow for the analysis of intonation patterns, which requires more controlled, purpose-built data sets. A prosodic corpus devoted to the analysis of IE varieties needs to be incorporated into the research agenda. This contribution presents the corpus of Irish English Speech (IES) with the following objectives: to collect recordings of spoken IE across present-day Ireland under a unified protocol in order to guarantee comparisons among different datasets; to obtain an initial phonological inventory of each variety examined; to compare the phonological systems of diverse IE varieties; and to provide researchers with accessible and open data sources. The core of the corpus has been gathered in accordance with the guidelines of the Interactive Atlas of Romance Intonation project (Prieto, Borràs-Comes & Roseano, 2011-2014) via a questionnaire based on the Discourse Completion Task, which was translated and readapted for Irish English speakers, and a Map Dialogue Task designed to obtain spontaneous speech productions. This method has yielded the collection of a wide range of intonation patterns concerning different types of context-specific utterances, such as statements, questions, imperatives and vocatives. After an overview of the segmental phonology of IE, previous studies on the prosodic features of IE varieties and the speech corpora of IE will be examined with the purpose of identifying the gaps in existing literature, which will then be followed by a detailed outline of the development of the corpus of IES. This contribution will provide an illustrative example for fully exploiting the potential of the IES database and call for further in-depth investigations on IE prosody.