| Summary: | Grounded-embodied theories hold that language is understood and remembered through perceptual and motor simulations (i.e., activations and re-activations of sensorimotor experiences). This thesis aims to illustrate simulations of space in memory for language. In four experiments, we explored (1) how individuals encode and re-activate word locations and (2) how word meanings activate locations in space (e.g., "bird" - upward location). Experiment 1 reveals that the propensity to refer to the environment during retrieval correlates with individual's visuospatial memory capacity. Experiment 2 shows that words which are more difficult to remember and, particularly, words that are more difficult to visualise in mind lead to more reliance on the environment during word retrieval. Experiment 3, which is a norming study, demonstrates that there is a high degree of agreement among individuals when linking words to locations in space although there are no explicit conventions with regard to word - space associations. Experiment 4, in which recognition memory for words with spatial associations was probed, shows that both language-based simulation of space and simulation of word locations dictate memory performance even if space is irrelevant and unnecessary for successful retrieval. Results are discussed within grounded-embodied and extended approaches to memory and language.
|
|---|
