Efficiency of spoken word recognition slows across the adult lifespan

Abstract

Spoken word recognition is a complex cognitive process that underpins efficient language processing. When recognizing words, listeners must quickly map the spoken input to stored lexical candidates as speech unfolds over time. Older adults experience two declines that could impact word recognition: hearing loss and cognitive decline. To recognize words, listeners must be able to accurately hear the input, and they must resolve competition between lexical candidates, though it is unclear whether this resolution derives from domain-general or language specific mechanisms at any age. Thus, spoken word recognition may provide a crucial mediator between hearing loss and general cognitive declines in older adults. We examined online spoken word recognition in a continuous age sample from adolescence to older adulthood (N=104, ages 11 – 79). Hearing thresholds were collected from all participants at octave frequencies between 0.25 - 8 kHz. Participants completed a Visual World Paradigm (VWP) task to examine the dynamics of lexical competition and a non-linguistic analogue of the VWP to capture domain-general visual cognition and speed of processing. Spoken word recognition increased in efficiency through the 20’s, but began to decline around age 45. A hierarchical regression showed that age predicted the efficiency of activating target words over and above visual cognition and peripheral hearing. This suggests that age uniquely affects word recognition, separate from hearing ability and age-related changes to visual cognition. This highlights spoken word recognition as a potential early marker for more severe cognitive decline in the future.

Engaging effort improves efficiency during spoken word recognition in cochlear implant users

“…The resolution of competition (Sustained Activation) decreases with older age. Our findings replicate the findings of Colby and McMurray (2023), who documented similar patterns with normal hearing people across the lifespan, and we extend this pattern of results with a group of listeners with severe hearing loss. This suggests that age, separate from hearing acuity, consistently impacts the dynamics of spoken word recognition.…”

“…For example, frequency is instantiated via stronger bottom-up connections between phonological representations and words (Dahan et al, 2001) and density is a product of direct (lateral) inhibition between words. Colby and McMurray (2023), included for visual comparison. The onset of the target stimulus is at 0 msec.…”

“…Design. The VWP task is described in detail in and Colby and McMurray (2023). There were 60 item sets in this task (30 monosyllabic, 30 bisyllabic), each comprised of a target, cohort competitor, rhyme competitor, and unrelated word.…”

“…Solid bold lines are for the CI users. Dashed faint lines are average data from a sample of normal hearing listeners reported inColby and McMurray (2023), included for visual comparison. The onset of the target stimulus is at 0 msec.…”

Decoupling Language Processing from Time

“…This pattern has been colloquially termed sustained activation (Figure 5C, D) [79; 84; 86]. Elderly listeners also show evidence of this sustained activation even with normal hearing [88]. [83] In moderately challenging conditions, listeners do not fully suppress competitors, a profile termed sustained activation.…”