Fluency versus accuracy in advanced spoken learner language

Abstract

In this paper we present a possible multi-method approach towards the description of a potential correlation between errors and temporal variables of (dys-)fluency in spoken learner language. Using the German subcorpus of the Louvain International Database of Spoken English Interlanguage (LINDSEI) and the native control corpus Louvain Corpus of Native English Conversation (LOCNEC), we first analysed errors and temporal variables of fluency quantitatively. We detected lexical and grammatical categories which are especially error-prone as well as problematic aspects of fluency for all learners in the LINDSEI subcorpus, e.g. confusion in tense agreement across clauses or an overuse of unfilled pauses. In the ensuing qualitative analysis of five prototypical learners, no trend for a possible correlation of accuracy and fluency could be observed. Fifty native speakers’ ratings of these five learners revealed that the learner with an average performance across the investigated variables received the highest ratings for overall oral proficiency.

Exploring Connections Between First and Second Language Fluency: A Mixed Methods Approach

“…To gain insights to individual differences in the connections between L1 and L2 fluency, six participants’ samples in L1 and L2 were examined with a qualitative analysis (the third research question). Three participants with the most stalling mechanisms in their L2 from both groups (participants G1‐008, G1‐012, G1‐013, G2‐301, G2‐303, and G2‐306) were included in the analysis, as stalling mechanisms (especially filled pauses and repetitions) have been associated with individual variation in previous studies (e.g., Brand & Götz, ; Cucchiarini et al., ; Götz, ; Kahng, ; Peltonen & Lintunen, ). The number of participants was narrowed down to six to enable in‐depth analyses of each participants’ L1 and L2 productions from the perspective of stalling mechanisms; yet the number of participants (three subjects from both groups) enabled some comparisons across the participants both within and between the groups.…”

“…The number of participants was narrowed down to six to enable in‐depth analyses of each participants’ L1 and L2 productions from the perspective of stalling mechanisms; yet the number of participants (three subjects from both groups) enabled some comparisons across the participants both within and between the groups. In some previous L2 fluency studies the participants for the qualitative analysis have been chosen based on “extreme case sampling” (Dörnyei, , p. 128; e.g., Brand & Götz, ; Peltonen, ; Peltonen & Lintunen, ), targeting participants, for instance, with the least and most repair phenomena. However, the participant selection in the present study can be characterized as “critical case sampling” (Dörnyei, , p. 128), targeting subjects with the most stalling mechanisms, as the purpose was to examine specifically what their use can reveal about the connections between L1 and L2 fluency.…”

“…However, the participant selection in the present study can be characterized as “critical case sampling” (Dörnyei, , p. 128), targeting subjects with the most stalling mechanisms, as the purpose was to examine specifically what their use can reveal about the connections between L1 and L2 fluency. The analysis was based on earlier L2 fluency studies that have included a qualitative component (e.g., Brand & Götz, ; Hilton, ; Peltonen, ; Peltonen & Lintunen, ; Towell et al., ). More specifically, the use of stalling mechanisms in the participants’ L1 and L2 productions was compared to examine whether learners’ potential idiosyncratic patterns in the L2 could be explained with reference to their L1 productions.…”

“…Individual differences affecting L1 speech fluency are likely to be reflected in L2 fluency measures as well (e.g., De Jong et al, 2015;Huensch & Tracy-Ventura, 2017). For instance, although not explicitly studying within-group variation, some L2 fluency studies have demonstrated relatively large standard deviations especially in the use of filled pauses and repetitions both in native and L2 speech (e.g., Brand & Götz, 2011;Cucchiarini, Strik, & Boves, 2002;Götz, 2013;Kahng, 2014;Peltonen & Lintunen, 2016), which suggests that at least these aspects of fluency are subject to individual variation (see also Fehringer & Fry, 2007). It has also been postulated that pause length is more related to individual speaking style than to L2 proficiency (e.g., Cucchiarini et al, 2002, Kolly et al, 2015but cf.…”

Application of a Corpus to Identify Gaps between English Learners and Native Speakers

“…Learners of English as a foreign language (EFL) frequently demonstrate a level of language ability that differs from that of native speakers (gap between learner and native speaker). Compared with native speakers, learners more frequently generate grammatically incorrect sentences and speak at a slower rate (Brand and Götz 2011, Chang 2012, Thewissen 2013. To develop tools and methods for effective learning of EFL, gaps in the four basic linguistic skills (reading, writing, pronunciation, and listening) need to be clearly identified and bridged.…”