Spectro-temporal cues for perceptual recovery of reduced syllables from continuous, casual speech

Function words may be highly reduced, with little to no discontinuity marking their onsets to cue their segmentation from continuous speech. The present study investigated whether reduced function words lacking onset discontinuities have residual timing cues that could be used for word segmentation....

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:The Journal of the Acoustical Society of America 2017-05, Vol.141 (5), p.3700-3700
Hauptverfasser: Dilley, Laura, Arjmandi, Meisam K., Ireland, Zachary
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Function words may be highly reduced, with little to no discontinuity marking their onsets to cue their segmentation from continuous speech. The present study investigated whether reduced function words lacking onset discontinuities have residual timing cues that could be used for word segmentation. Participants (n = 51) briefly viewed sentences and spoke them from memory to elicit casual speech. They were randomly assigned to either a “function-word present” condition (n = 29) in which experimental items contained a critical function word expected to frequently blend spectrally with context, or a “function-word absent” set (n = 22) with phonetically matched items lacking the critical word. Acoustic analyses confirmed that in “function-word present” sentences, critical words lacked detectable onset discontinuities 60% of the time. Critically, in the “function-word present” condition, portions of speech containing critical function words were longer, both in terms of absolute duration and normalized for context speech rate, compared with matched portions in the “function-word absent” condition, even when the former were highly reduced and lacked onset discontinuities. These findings suggest that relative duration cues provide substantial information which may be used by listeners for segmentation of highly reduced syllables from continuous speech. [Work supported by NSF Grant BCS-1431063.]
ISSN:0001-4966
1520-8524
DOI:10.1121/1.4988070