UC Riverside

The perception of speech is supported by both multisensory (e.g. Sumby & Pollack, 1954) and lexical information (e.g. Miller, Heise, & Lichten, 1951). How the mechanism for speech perception processes these two sources of information is an open theoretical question. Evidence indicates that multisensory information is integrated early in speech processing (e.g. Musacchia et al., 2006) and some theories assume that integration precedes lexical processing (e.g. Fowler, 2004; see also Rosenblum et al., 2016). Accordingly, these theories assume that lexical processing is performed on the integrated multisensory information. In contrast, some have recently proposed that lexical processing is performed on unintegrated unisensory information (e.g. Ostrand et al., 2016; Samuel & Lieblich 2014). This dissertation provides a careful investigation into these claims to address the potential interactions between lexical processing and multisensory integration. Chapter 1 investigates if semantic processing of McGurk stimuli is consistent with the unperceived (and putatively unintegrated) auditory information (i.e. Ostrand et al., 2016) or the perceived (audio-visually integrated) information. Chapter 2 investigates if selective adaptation, a perceptual phenomenon known to be sensitive to low-level sensory information (e.g. Samuel & Newport, 1979) but also to lexically supported illusory percepts (e.g. Samuel, 1997), is sensitive to multisensory illusions. Finally, Chapter 3 investigates if lexical information influences the integration of auditory and visual speech information. The results of this dissertation indicate that lexical processing is sensitive to integrated multisensory information. However, this dissertation found no indication that lexical information influenced the multisensory integration process.