Research into non-native (L2) speech perception has increased the need for specialized experimental materials. The Non-Native Speech Recognition (NNSR) sentences are a new large-scale set of speech recognition materials for research with L2 speakers of English at CEFR level B1 (North, Ortega, & Sheehan, 2010) and above. The set comprises 439 triplets of sentences in three related conditions: semantically predictable, neutral, and anomalous. The sentences were created by combining a strongly or weakly contextually constrained sentence frame with a congruent or anomalous final keyword, and they were matched on a number of factors during development, to maintain consistency across conditions. This article describes the development process of the NNSR sentences, along with results of speech-in-noise intelligibility testing for L2 and native English speakers. Suggestions for the sentences' application in a range of investigations and experimental designs are also discussed.
The intelligibility of an accent strongly depends on the specific talker-listener pairing. To explore the mechanism of this influence, we investigated the relationship between acoustic-phonetic similarity and accent intelligibility across native (L1) and non-native (L2) talker-listener pairings. We also used online measures to observe processing differences in quiet. Method: English (N=16) and Spanish (N=16) listeners heard Standard Southern British English, Glaswegian English and Spanish-accented English in a speech recognition task (in quiet and noise) and an EEG task (quiet only) designed to elicit the Phonological Mismatch Negativity (PMN) and N400 effect. Stimuli were drawn from the Non-native Speech Recognition sentences (Stringer & Iverson, in preparation). The acoustic-phonetic similarity between listeners' accents and the three accents was calculated using the ACCDIST metric (Huckvale, 2004, 2007). Results: Talker-listener pairing had a clear influence on accent intelligibility. This was linked to the phonetic similarity of the talkers and listeners, but similarity could not account for all findings. The influence of talker-listener pairing on processing in quiet was less clear, but it seems that a close match between talkerlistener accents is required to elicit a PMN response. The N400 is more robust to accent mismatches, with some relationship to intelligibility. Conclusion: These findings suggest that the influence of talker-listener pairing on intelligibility may be partly attributable to accent similarity in addition to accent familiarity. Online measures also show that differences in talker-listener accents can disrupt processing in quiet even where accents are highly intelligible.
The intelligibility of accented speech in noise greatly depends on the pairing of speaker and listener, where two important factors are a listener’s familiarity with a speaker’s accent and the acoustic similarity between their accents. In this study, we present patterns of the intelligibility of standard British English, Glaswegian English and Spanish English accents for British and high and low proficiency Spanish listeners. We predict intelligibility will correlate with acoustic-phonetic similarity across accent pairings, in line with previous findings. As such, findings are expected to provide further support that accent similarity can predict patterns of accent intelligibility, even if listeners have little experience of a speaker’s accent. Electrophysiological measures relating to phonological and semantic integration stages of word recognition will also allow the investigation of the influence of accent on the time course of word recognition, which has not previously been directly compared to accent intelligibility or applied in studies of accent processing by non-native listeners. Findings will be discussed in the context of previous exploratory findings in this field, and in reference to other factors influencing accent intelligibility.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2025 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.