Events

Event Information:

  • Wed
    06
    Nov
    2013

    Language statistics explain good-enough representations

    4:00pm - 5:00pm D18, Fylde, Lancaster University

    Speaker: Max Louwerse, Professor Cognitive Psychology and Artificial Intelligence, Tilburg Center for Cognition and Communication, Tilburg University

    A vast amount of literature has demonstrated that cognitive processes can be explained by a perceptual simulation account. Oftentimes such studies interpret an effect for perceptual simulation as the only explanation. This talk will discuss whether there are alternatives. One such alternative is provided by the Symbol Interdependency Hypothesis, which proposes that language encodes perceptual information. Language has evolved into a system of regularities that allows for a symbolic shortcut to perceptual relations in the world around us. Language users can rely on symbolic and perceptual relations depending on the nature of the stimulus, the cognitive task, and the individual, as well as the time course of processing. The Symbol Interdependency Hypothesis thereby emphasizes the role of language statistics in cognitive processing: Various studies ranging from conceptual metaphor, iconic relationships, geographical orientation, and the Spatial Numerical Association of Response Codes (SNARC) will be discussed showing that what might seem to be best explained by a perceptual simulation account is in fact best explained by statistical linguistic frequencies.