Innovation in Language Learning

A primary goal of many language learners is to communicate effectively with native speakers of the target language, rendering the ability to decode native speech of paramount importance. Aural decoding (also known as speech decoding) is the process by which listeners convert the speech signal into sounds, words, clauses, and literal meaning (Field, 2008). Recent work by Authors (under review) indicates that speech rate significantly predicts L2 listeners’ speech-decoding performance, suggesting that delivery speed plays an important role in aural decoding accuracy, with faster speech rates associated with lower decoding performance. However, the specific locations of decoding breakdowns within a sentence (i.e., initial, medial, or final position) at different speech rates are not yet understood. It remains unclear, therefore, whether certain sentence positions are more susceptible to speech-rate effects than others. The present study examines whether speech rate moderates the relationship between error location and L2 speech-decoding performance. Native English-speaking learners of L2 Spanish (N = 31) enrolled in upper-level Spanish courses in a Southeastern university listened to sentences articulated by a native-Spanish speaker. To measure their speech-decoding ability, participants were asked to orthographically transcribe aural input in Spanish. Speech decoding was operationalized as the percent of syllables accurately decoded in the sentence. Of sixteen discrete sentences (controlled for length), half were delivered at a normal rate (M = 5.17 syllables per second, sps) and half were delivered at a fast rate (M = 7.13 sps). Orthographic transcriptions will be analyzed to determine decoding error location (initial, medial, or final position in the sentence). To explore the extent to which speech rate, error location, and their interaction predict speech decoding, linear-mixed effects models will be fit in R. The analyses will determine whether decoding performance varies according to error location and whether this relationship is moderated by speech rate. These findings will guide developers in creating L2 listening materials and strategies that facilitate the decoding of native-speaker speech by identifying sentence positions that are particularly vulnerable to speech-rate effects.

Keywords: listening, decoding, speech rate, error location

REFERENCES

[1] Field, J. (2008). Listening in the language classroom. Cambridge University Press. https://doi.org/10.1017/CBO9780511575945