Valsø, A. M. & Behne, D. M.
Speech Lab, Psychology Department, Norwegian University of Science and Technology
Behavioral research on audio-visual speech perception has long shown an increased use of visual information when speech is imbedded in white noise (e.g., Sumby & Pollack, 1954). More recently, use of visual speech information has been shown to be even greater in babble than white noise (e.g. Alm et al., 2009). The current study investigates temporal effects of white noise and cafeteria babble on speech processing, and the effect of visual speech on the respective conditions. Continuous EEG was recorded while thirteen healthy Norwegians were presented with audio-visual, audio-only, and visual-only productions of the syllable /ba/ in quiet or masked with white noise or babble. Latency analyses of the N1 component showed that early speech processing was later in noise conditions compared to quiet, and to a greater degree in cafeteria babble than in white noise. Results also showed that the access to visual speech cues increased processing speed in the noisy conditions and reduced the gap in processing speed of syllables between quiet and noise. These findings will be discussed in terms of the principle of inverse effectiveness