Shhh! a Study Comparing the Production of Whispered Segments by Dutch L2 Speakers of English

Shhh! A study comparing the production of whispered segments by Dutch L2 speakers of English and native L1 speakers.

Marita Everhardt, Matt Coler, and Wander Lowie

In this paper we present the results of our study on whispered speech in L1 and L2. We compare production strategies used by native Dutch and English speakers to distinguish between phonated and whispered voiced/voiceless final-obstruents of English words. In English, voiced final-obstruents are distinguished from voiceless final-obstruents by both the primary cue of voicing and secondary acoustic cues like preceding vowel length. Yet, in Dutch speech the primary cue of the voicing distinction between underlying voiced and voiceless final-obstruents is neutralized, as voiced final-obstruents are realized voiceless due to final devoicing (Ernestus & Baayen, 2003; Jongman et al., 1992). The question is to what extent native Dutchspeakers make use of secondary cues of the voicing distinction in English. To investigate this, we used whispered speech,since voicing is impossible in whispered speech and only secondary cues can be used to create the voicing distinction (Raphael, 1971; Sharf, 1964).

To compare how native Dutch versusnative English speakers distinguish between voiced and voiceless final obstruents in whispered English speech, we devised an experiment to test the production of final-obstruent voicing in phonated and whispered speech. Participants included twentynative Dutch speakerswith English proficiency levels ranging from intermediate to advanced and eight native English speakers. Using English stimuli consisting of minimal pairs which differed only by the voicing specification of the final-obstruent (e.g. beat vs bead), we analysed recordings of both groups. Results indicate that, whereas native English speakers use the secondary cue in both phonated and whispered speech, Dutch L2 speakers of English do not as no significant vowel length difference was found between words with voiced and words with voiceless final-obstruents in either phonated or whispered speech. This seems to indicate that native Dutch speakers do not make use of secondary cuesof the voicing distinction in English.

References

Ernestus, M., & Baayen, R. H. (2003). Predicting the unpredictable: Interpreting neutralized segments in Dutch. Language, 79(1), 5-38.doi: 10.1353/lan.2003.0076

Jongman, A., Sereno, J. A., Raaijmakers, M., & Lahiri, A. (1992). The phonological representation of [voice] in speech perception. Language and Speech, 35(1,2), 137-152. doi:10.1177/002383099203500212

Raphael, L. J. (1971). Preceding vowel duration as a cue to the perception of the voicing characteristic of word- final consonants in American English. The Journal of the Acoustical Society of America, 51(4), 1296-1303. doi:10.1121/1.1912974

Sharf, D. J. (1964) Vowel duration in whispered and in normal speech. Language and Speech, 7(2), 89-97. doi:10.1177/002383096400700204