Abstract ID: 1220
Part of Session 171: Experimental methods in the study of social meaning (Other abstracts in this session)
Authors: Watt, Dominic J.L. (1); Llamas, Carmen (1); Docherty, Gerard J. (2)
Submitted by: Watt, Dominic James L. (University of York, United Kingdom)
A principal objective of the Accent and Identity on the Scottish/English Border (AISEB) project was to shed light on the interrelationship of language variation and identity factors by triangulating data gathered using a variety of analytical methods from people living near the border. Firstly, data on variability in the production of selected phonological variables (rhoticity; pronunciations of (r); Voice Onset Time in /p t k/; Scottish Vowel Length Rule; NURSE, etc.) were collected from recordings of 160 speakers using auditory and acoustic techniques. Secondly, sociopsychological attitudes towards the border, national and regional identity labels, and local language issues were elicited from informants via a structured questionnaire, with the intention of then relating their responses to the same speakers’ phonological production patterns. The third category of data (from a subset of 40 speakers) was informants’ responses to two types of association test: two ‘Explicit Association Tests’ and two ‘Implicit Association Tests’.
The Explicit Association Tests (EATs) were based on methods developed by Clopper & Bradlow (2009). In the first (EAT1), informants classified audio samples into categories in a free, unguided way by moving clickable tiles linked to the samples into clusters on a laptop screen, while describing their motives for grouping the samples as they had. The second task (EAT2) was identical except that a grid of lines representing the national border, and finer east/west/central and north/south distinctions, was superimposed on the working area of the screen. The classification task was thus guided in a way predetermined by the investigator.
The Implicit Association Test (IAT) paradigm is well established in social scientific and psychological research, and it has recently been used to good effect by researchers interested in the evaluation of linguistic varieties and phonological variants by listeners (e.g. Redinger 2010, Pantos 2010, Campbell-Kibler 2011). In the present experiment we tested the extent to which listeners associate the categories ‘Scotland’ and ‘England’ with (a) positive and negative personality traits (this was IAT1), and (b) members of pairs of minimally different audio stimuli representing variants of (r) and the FLEECE and GOOSE vowels (IAT2). The responses for IAT1 were in the form of reaction time latencies, differences between which are taken to represent the level of automaticity of a participant’s association of a positive or negative trait with one or other of the two countries. Responses for IAT2, where participants listened to audio samples they were to associate with either Scotland or England as reflexively as possible, were pooled across participants so that the level of agreement within speaker groups (defined by town of residence, age, and gender) could be gauged.
We present the results of the EAT and IAT experiments in this paper, and relate these to the production patterns exhibited by the 40 speakers who took part in the perceptual experiments. We argue that the innovative multi-strand approach taken in AISEB is a means of bringing us closer to an understanding of speakers’ conscious and subconscious motivations for adopting or resisting sound changes in this linguistically sensitive region.