The noisy channel model and sentence processing in individuals with simulated hearing loss

Recent research (Gibson, Bergen, & Piantadosi, 2013; Levy, 2011; Levy, Bicknell, Slattery, & Rayner, 2009) provides empirical evidence that language users maintain uncertainty about perceived linguistic input and in order to increase the likelihood of a successful communicative exchange, where the meaning intended matches the meaning perceived, may interpret a perceived sentence in a way that is unfaithful to the literal syntax. Gibson, Sandberg, Fedorenko, Bergen, & Kiran (2015) and Warren, Dickey, & Liburd (2015) found that individuals with aphasia may be aware of the increased noise in their language processing mechanism, and as a result, rely more on semantic information as a means to increase the likelihood of a successful communicative exchange. The present study aims to further examine how properties of the comprehender, such as presence or absence of simulated hearing loss, may affect one’s reliance on a perceived linguistic signal. 40 participants with a simulated high frequency hearing loss and 40 participants without a simulated hearing loss were administered the Gibson task, a forced choice picture task that asks participants to select which of two illustrations best represents a sentence they heard. One illustration represents the literal syntax while the other represents an alternate interpretation that may be obtained through edits or distortions of the literal syntax. The sentences presented vary in structure (double object, prepositional object, active, and passive) and plausibility (plausible, implausible, and impossible). Participants had their eyes-tracked while listening to sentences and making decisions. Both groups of participants partook in rational sentence inferencing. Participants in the simulated hearing loss exhibited lower accuracy scores and longer reaction times. Both groups of participants exhibited more competition in less reliable sentence conditions as evident through eye-tracking; however, participants with simulated hearing loss showed more competition between the target and competitor image than individuals with no hearing loss. Furthermore, participants with simulated hearing loss appeared the reach the ceiling in terms of available processing resources as evident through both reaction time and eye-tracking data