## Welcome! ## We hope you'll leave a comment or join [our video chat][1] during Poster Session #1, Thursday September 3rd at 16:00–17:30 CEST (7:00am-8:30am PDT). See the files below for a quick [3-minute video presentation of the poster (Poster Spiel.mp4)][2], as well as the poster and the abstract. You can also visit http://spellout.net/ibexexps/jmduff/FR-90-Maze-Sample/experiment.html for a Maze demonstration. ---------- ## Short Abstract ## In a landmark eyetracking study, Frazier & Rayner (1990) demonstrated that **polysemy** (overlapping meanings; e.g. *newspaper* as printed object vs. organization) and **homonymy** (distinct meanings; e.g. *jam* as fruit spread or traffic blockage) have different profiles in online reading. Modern accounts built on their results suggest that homonyms engender an immediate commitment to one of the meanings, potentially a garden path, while polysemes temporarily permit a flexible, underspecified representation. We fail to observe these critical differences in a Maze task replication, and take this as evidence that maintenance of an underspecified representation is not a necessary component of the processing of polysemy. [1]: https://meet.jit.si/AMLaP2020_poster_320 "our video chat" [2]: https://osf.io/anwdm/ "3-minute video presentation of the poster"