Читать книгу The Handbook of Speech Perception - Группа авторов - Страница 34

The notion that the speech mechanism may be sensitive to a form of information that is not tied to a specific sensory modality has been discussed for over three decades (e.g. Summerfield, 1987). This construal of multisensory speech information has been alternatively referred to as amodal, modality‐neutral (e.g. Rosenblum, 2005), and supramodal (Fowler, 2004; Rosenblum et al., 2016, 2017). The theory suggests a speech mechanism that is sensitive to a form of information that can be instantiated in multiple modalities. Such a mechanism would not need to contend with translating information across modality‐specific codes, or to involve a formal process of sensory integration (or merging), as such. From this perspective, the integration is a characteristic of the relevant information itself. Of course, the energetic details of the (light, sound, tactile‐mechanical) input and their superficial receptor reactions are necessarily distinct. But the deeper speech function may act to register the phonetically relevant higher‐order patterns of energy that can be functionally the same across modalities.

The supramodal theory has been motivated by the characteristics of multisensory speech discussed earlier, including: (1) neurophysiological and behavioral evidence for the automaticity and ubiquity of multisensory speech; (2) neurophysiological evidence for a speech mechanism sensitive to multiple sensory forms; and (3) neurophysiological and behavioral evidence for integration occurring at the earliest observable stage; and (4) informational analyses showing a surprising close correlation between optic and acoustic informational variables for a given articulatory event. The theory is consistent with Carol Fowler’s direct approach to speech perception (e.g. Fowler, 1986, 2010), and James Gibson’s theory of multisensory perception (Gibson, 1966, 1979; and see Stoffregen & Bardy, 2001). The theory is also consistent with the task‐machine and metamodal theories of general multisensory perception which argue that function and task, rather than sensory system, is the guiding principle of the perceptual brain (e.g. Pascual‐Leone & Hamilton, 2001; Reich, Maidenbaum, & Amedi, 2012; Ricciardi et al., 2014; Striem‐Amit et al., 2011; see also Fowler, 2004; Rosenblum, 2013; Rosenblum, Dias, & Dorsi, 2017).

It will be argued that this thesis, which was presented in the first version of this chapter (Rosenblum, 2005), continues to gain supportive evidence (Rosenblum, Dorsi, & Dias, 2016; Rosenblum, Dias, & Dorsi, 2017). Throughout, the term supramodal information will be used instead of modality‐neutral information, which was used in the previous version of the chapter. This change has been made largely to be consistent with other theories outside of adult speech perception (neurophysiological; behavioral), which now make very similar claims (e.g. Papale et al. 2016; Ricciardi et al., 2014; Zilber et al. 2014).