Combining and integrating multiple linguistic cues during spoken language comprehension: A focus on semantics and coarticulation.

This research examines how adults process and integrate a combination of higher-level semantic cues (i.e., semantic context) which are followed by lower-level acoustic cues (i.e., coarticulatory cues) during online spoken comprehension. Previous studies investigating cue integration within Martin (2016)'s framework found that listeners can flexibly use and integrate a variety of available cues across linguistic representations. The current pre-registered study used an eye-tracking paradigm and tested how listeners process coarticulation (a lower-level cue) in the presence of preceding semantic information (a higher-level cue). Adult listeners were sensitive to both semantic and coarticulatory cues; moreover, in an exploratory analysis, adults' processing of later acoustic cues varied depending on the earlier semantic context. These results demonstrate that listeners can flexibly use and weigh cues across multiple levels of linguistic representations during language comprehension. Earlier semantic information may be maintained over time and can influence the processing of later lower-level acoustic cues.