***Apologies for multiple postings***
The first Cognitive Area Seminar this semester will take place on January 16th, 2015 15h30-17h00 in W7/21, Stewart Biology Building. There will be snacks!
Learned Whorfian Effects and Dimensional Reduction
Stevan Harnad
(presenting the ongoing work of Xi-Wei Kang [U Southampton], Phlippe Vincent-Lamarre [UQAM, U Ottawa], Fernanda Perez Gay [UQAM, McGill] and Daniel Rivas [McGill)]
According to the Whorf-Sapir Hypothesis, language influences how we perceive the world. All the well-known evidence for this (e.g., color terms, Inuit snow terms, Hopi perception of the future, Chinese perception of counterfactuals) has turned out to be false. I will present some new evidence that perceived similarity is changed by learning to sort stimuli (digital drawings of fish) into named categories ("Limfish" and "Ailfish"). Within-category similarity increases and between-category similarity decreases after successful category learning
(an effect called "categorical perception" CP). This Whorfian effect is stronger the harder it is to learn the category, because fewer features co-vary with category membership, making it harder to find which features to selectively attend to and which features to ignore. Our hypothesis is that the CP effect arises from this dimensional reduction. In closing I will make some links with the symbol grounding problem of how words get their meanings.