Friday, October 1, 2010
2001 LSA Building
Most real-world speech communication takes place in environments with some degree of background noise, and very often this noise includes competing speech signals that may or may not be in the same language as the target speech. Thus, even if the interlocutors in a given dialogue are both monolingual speakers of the target language, they must frequently function in a “polyglot soundscape” (i.e. an auditory environment that includes background speech in multiple languages). In this talk, I will present a series of studies that focuses on speech-in-speech recognition in the context of a polyglot soundscape. This series forms part of a research agenda whose overall goal is to understand speech communication under a variety of real-world communicative conditions, particularly situations that involve contact between native and non-native speakers of English. Speech recognition in the presence of background noise is challenged by a combination of energetic/peripheral and informational/central masking. Energetic masking is related to target audibility. Informational masking depends on linguistic, attentional, and cognitive factors (Cooke, Garcia-Lecumberri, & Barker, 2008).