![]() The model’s conception of WM also assumes a central pool of cognitive resources that may be allocated to either storage or processing of information. Working memory is central to goal-directed behaviors where information must be stored and manipulated ( Chai et al., 2018), and contains a general executive attention component ( Chung et al., 2018) linked to inhibitory control ( Getzmann et al., 2018 Tiego et al., 2018). The phonological mismatch can result in increased reliance on WM to process degraded speech through an interplay with semantic long-term memory and episodic long-term memory (ELTM) in which linguistic inferences are formed based on prior knowledge. Presbycusis (age-related hearing loss) or interference from background speech can cause a mismatch in phonological binding of input and the representations in long-term memory. An initial sensory module that Rapidly, Automatically, and Multimodally Binds Phonological information (RAMBPHO) allows for easy lexical access and understanding of speech. The Ease of Language Understanding (ELU) model ( Rönnberg et al., 2013, 2019, 2021) highlights the importance of sensory-cognitive integration in speech-on-speech perception. Hence, higher-order cognitive functions like selective attention ( Oberfeld and Klöckner-Nowotny, 2016) and working memory (WM Sörqvist and Rönnberg, 2012) are also engaged when speech is attended in the presence of competing speech. In segregating target and masker components, the listener also needs to actively inhibit interference from the masker ( Tun et al., 2002). For example, increased spatial separation ( Viswanathan et al., 2016) and speaker voice differences ( Darwin et al., 2003 Başkent and Gaudrain, 2016) seem to improve performance. In contrast, in speech-on-speech perception - a special case of “informational masking” ( Pollack, 1975)- where the masker can be a single-talker, multitalker, or babble (gibberish), further interference can occur due to the overlapping temporal and spectral properties of the target and masking speech, and linguistic and semantic content of the masker speech.ĭuring speech-on-speech perception, spectral and temporal components of the speech streams can be used to segregate target and masker streams to better understand a target speaker. In such “energetic masking” ( Brungart, 2001), speech perception relies on the construction of the speech meaning from the audible temporal and spectral sections of the target speech, using linguistic and semantic constraints, but otherwise no further interference comes from the masker. ![]() When the background masker is non-speech, such as a steady noise, the speech acoustic cues are limited due to direct obliteration from the masker. With increasing complexity, such as in background noise, it is an active cognitive process ( Heald and Nusbaum, 2014 see Mattys et al., 2012), and one’s ability to parse information relies on combined sensory and cognitive capacities ( van Knijff et al., 2018). Under ideal listening conditions, speech perception is considered automatic ( Johnson and Ralston, 1994). Introduction: The Ease of Language Understanding Model and Working Memory
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |