Esther Janse, Associate Professor at the Centre for Language Studies, Radboud University, is the guest speaker at the ACLC seminar on Friday 1 November 2019 from 16:15 until 17:30 (followed by drinks).
Poor listeners and speakers (and the things we ask our participants to do)
In language research we often present speakers and listeners with challenging tasks eliciting their ‘maximum performance’ where listening and speaking may be effortful. By studying language users’ performance on such challenging tasks we can investigate how language processing is grounded in language users’ sensory, cognitive and motor control systems. Obtaining such knowledge is crucial for understanding how and why language processing may change over the adult life span, which has been a key question driving my research.
I will present two lines of research in which I have investigated how participants’ abilities (e.g., hearing acuity or working memory capacity) interact with properties of the speech material they have to process. The first line of research is on the identification and recall of nonwords where we asked how older listeners’ hearing might modulate the effects of neighbourhood density and phonotactic probability on nonword processing. I will show that participants’ hearing loss also affects their recall of visuallypresented nonwords. The second (more recent) line of research is on speech production. Speakers differ in the maximum speech rate they can attain and in how prone they are to getting their tongue twisted. As tongue twister phrases (e.g., she sells sea shells on the sea shore) typically involve having to switch back and forth between similar syllable onsets, we investigated whether aspects of executive functioning (shifting ability in particular) relate to tongue twister performance. Results indeed showed that switching ability, more than other aspects of executive function, predicted tongue twister performance as well as maximum performance on diadochokinesis tasks (being a clinical measure of articulatory control). For both lines of research, I will elaborate on how performance on these ‘artificial’ or ‘construed’ tasks may generalize to more naturalistic settings of language use. Lastly, I will elaborate on the implications of these results for models of spoken language processing.