Time: 11.00 - 12.30
Venue: Department of Psychology and Cognitive Science, Meeting Room, Corso Bettini 31, Rovereto
Speaker: Prof. Eva Reinisch - Ludwig Maximilian University, Munich (DE)
Abstract: Speech is inherently variable. On an everyday basis, we listen to different speakers, of different gender, from different regional backgrounds, who might be in a hurry or talk about something emotional. Despite the vastly different utterances that speakers produce in these situations, listeners appear to deal with this variation effortlessly. I will present different angles from which my research addresses how listeners use context information to deal with this variability in the speech signal and tune their perception to different listening situations. I will first present some work on how listeners use acoustic information such as speech rate and spectral context to interpret upcoming sounds in the continuous speech stream. These processes are typically considered low-level automatic processes and my research confirms that they apply during early stages of speech perception. I will then move on to perceptual learning paradigms in which lexical or visual information is used to recalibrate speech sound categories. I show that such category recalibration is likely one of the mechanisms that we use to adapt to different foreign-language accents. If time permits, I will also illustrate how perceptual learning paradigms have contributed to the debate about the smallest (abstract) units that listeners extract from the speech signal to map onto mental representations of sounds and words.
Scientific Coordinator: Francesco Vespignani