Researching phonological variation in child, learner and adult speech: The case of schwa in French

Abstract

Phonological variation can be defined as describing “a situation in which a single morpheme can be realized in more than one phonetic form in a single environment” (Coetzee & Pater, 2011, p. 401). In the latter decades, research on this integral part of grammar has become increasingly important, resulting not only in a growing literature on the topic, but also in large datasets on which hypotheses can be tested.

In order to capture and understand the patterns observed in the group(s) of speakers observed, research on phonological variation simultaneously requires rigor and an open mind: Rigor in order to develop a trustworthy empirical basis, an open mind in order to reveal all influencing factors and their interplay.

In this talk, I reflect on how phonological variables can behave differently across various groups of speakers, and what this requires with regard to methods and analysis of the results. To illustrate my talk, I focus on schwa in French. A multifaceted phenomenon, many lines have been written throughout history about this phonological variable, and it continues to intrigue phonologists, phoneticians, as well as sociolinguists.

With reference to the recent literature on the topic, primarily developed within the frame of the research programmes PFC: Phonologie du français contemporain (Durand, Laks, & Lyche, 2002, 2009) and IPFC: Interphonologie du français contemporain (Racine, Detey, Zay, & Kawaguchi, 2012), as well as during my own PhD project (Andreassen, 2013), I discuss and compare research on schwa in the speech of children acquiring French as their mother tongue, in the speech of learners of French as a foreign language, and in the speech of adult native French speakers coming from various parts of the Francophone world.

These studies together illustrate the complexity of phonological variation and how different internal and external factors influence – or do not influence – the way we produce language.