Dissertation
To wait or not to wait: the dynamics of lexical access adapt to suit listener needs
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Autumn 2020
DOI: 10.17077/etd.005701
Abstract
One key challenge in word recognition is the temporary ambiguity in the signal created by the fact that speech unfolds over time. Research with normal hearing (NH) listeners reveals that listeners cope with this temporary ambiguity by incremental processing and competition among possible lexical candidates. Effective communication requires that this process of lexical access be completed efficiently and accurately. Our understanding of how this process is affected by adverse listening conditions is still incomplete. Previous research suggests that when the listener receives spectrally degraded input (as is the case for cochlear implant [CI] users, or normal hearing [NH] listeners hearing noise-vocoded stimuli) mechanisms of lexical access are altered. In this dissertation, I ask when and how these mechanisms of lexical access change.
To measure lexical access, I assess fixations in the Visual World Paradigm (VWP). In this task, listeners select the picture corresponding to a spoken word from among four items that include the target word (e.g., mustard), a cohort competitor (e.g., mustache), a rhyme competitor (e.g., custard), and an unrelated item (e.g., penguin). During this task, eye movements are tracked as an index of the relative lexical activation of each object type during word recognition. Early in the trial, listeners fixate primarily the target and the cohort, but suppress the cohort within a few hundred milliseconds after disambiguating input is received. Rhyme items may be fixated, to a lesser degree, shortly after cohorts and are also suppressed quickly.
Prior work with the VWP has shown that post-lingually deafened CI users show a delay in processing when single words are presented in isolation. In the first study, across two experiments, I asked if CI users exhibit this delay when words are presented in sentence context. I compared patterns of fixations between CI users and NH controls when both groups listened to blocks of words presented in isolation or embedded in semantically neutral sentences (e.g., “click on the…”) that varied in length. CI users continued to show delays in processing relative to NH listeners in both sentence and isolation conditions. In addition, CI users showed lower peak levels of competition from cohorts but preserved activation of cohorts late in the trial. Both groups of listeners weighted early information more heavily when words were presented in sentence context, leading to increases in cohort competition as well as changes in the slope of target fixation curves.
In the second study, across two experiments, I asked if previously observed changes in lexical processing are solely driven by the quality of the input from the auditory periphery, or if these changes are modulated by cognitive mechanisms. To do this, I used noise-vocoded stimuli presented to NH listeners and varied the number of vocoding channels (which determines the difficulty). Listeners heard levels of vocoding either blocked together consistently or randomly interleaved from trial-to-trial. There was a small, but significant, effect of blocking. Listeners in the blocked condition showed more differentiation between the two levels of vocoding conditions; this suggests that some level of learning is in effect to best adapt to the varying levels of uncertainty in the input. In addition, exploratory analyses suggested that when less intelligible speech is being processed there is a cost to switching processing modes. A simple visual cue to inform listeners of the level of difficulty of the upcoming auditory input was enough to attenuate this cost, suggesting a role for cognitive adaptations in advance of adverse conditions.
Across these two studies the bottom-up input is critical in causing changes to lexical access. But there is also evidence that task demands (such as lexical access during sentences) and cognitive factors (such as learning, task switching, and effort) also play a critical role in shaping how lexical access proceeds. I also found evidence that listeners can exhibit several profiles of lexical access (such as “wait-and-see” and sustained competitor activation) simultaneously. The lexical system is flexible and can adapt to suit a wide variety of listener needs.
Details
- Title: Subtitle
- To wait or not to wait: the dynamics of lexical access adapt to suit listener needs
- Creators
- Francis X Smith Jr
- Contributors
- Bob McMurray (Advisor)Eliot Hazeltine (Committee Member)Shaun Vecera (Committee Member)Kristi Hendrickson (Committee Member)Jiefeng Jiang (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Psychology
- Date degree season
- Autumn 2020
- DOI
- 10.17077/etd.005701
- Publisher
- University of Iowa
- Number of pages
- xiv, 135 pages
- Copyright
- Copyright 2020 Francis X. Smith, Jr
- Language
- English
- Description illustrations
- color illustrations
- Description bibliographic
- Includes bibliographical references (pages 128-135).
- Public Abstract (ETD)
To effectively communicate, listeners must quickly and accurately map incoming speech signals onto meaning. This process of lexical access has been well studied in normal hearing listeners. There is broad consensus that lexical access relies on immediate activation of concepts consistent with the speech input and is followed by competition among multiple active concepts. However, listeners need to cope with a variety of adverse listening conditions frequently and it is currently unclear how, and when, adverse listening conditions may impact these processes. In this dissertation, I asked how one specific type of adverse listening condition impacts real-time processing of speech. Specifically, I asked if the type of input that cochlear implant users receive (which is missing information, relative to input to normal hearing listeners) leads to changes in when and how lexical access proceeds.
In one study, I asked if cochlear implant users change how they engage lexical access when words are presented in sentences (compared to presenting single words in isolation). In a second study, I investigated if cognitive mechanisms play a role in changes to mechanisms of lexical access or if these changes are solely based on the quality of the input. The findings suggest that while the bottom-up signal quality is critical, the demands of the task and cognitive factors both play a role in how lexical access proceeds. Our language systems are flexible and can adjust processing to suit the needs of the listener in the moment.
- Academic Unit
- Psychological and Brain Sciences
- Record Identifier
- 9984035794202771
Metrics
25 File views/ downloads
73 Record Views