A Cornell survey suggests that the brain deals with language (at least) in a fluid way, with the output of processes being distributed as they take place rather than after each is completed. A person is shown images on a screen of object including two with similar-sounding names, such as “candle” and “candy”. They are told to click on “candle”. Rather than waiting until the brain has decided which of the two sounds was heard, or indeed starting to move the mouse one way and then correcting themselves, people appeared to make arcing movements. The study suggests that initially the meaning of the word is in a cognitively ambiguous state — good enough to start moving the mouse towards the candle/candy — until the ambiguity is resolved, and thence the mouse trajectory.
Could have interesting implications for information and HCI design, let alone our understanding of cognition.
(From Technology Review)