Language Science Lunch Talk: Gary Oppenheim (Bangor University)

Choosing Your Words Carefully? Abstract: A typical adult knows tens of thousands of words in their first language, thousands more in any additional languages they might speak, and will continue to learn new words in each language throughout their life. Yet in spontaneous speech they can find the right words to express their ideas with remarkable ease, speaking with 99.9% accuracy at a rate of 2-3 words per second. Such success has inspired a theoretical model that has dominated language production research for the past several decades: it posits a strong, hard-wired role for real-time cognitive control in comparing candidates to choose the very best words and actively suppressing alternatives. But the empirical evidence for such controlled selection and suppression in word production has largely been limited to laboratory tasks that explicitly demand it. In this talk, I will discuss empirical and computational work from my lab that challenges the notion that comparison-based selection and suppression are baked into the production process. Instead, I will argue that speakers achieve fluent, apparently successful production largely through the application of a cognitively frugal, activation-based selection heuristic (any word that is highly activated should be close enough to convey an intended meaning) and demonstrate that key empirical phenomena that are often claimed as evidence of comparison-based selection naturally emerge from the same incremental learning that supports vocabulary development in the first place. Learn more about Gary Oppenheim Lunch served at 12:15 PM. Vegetarian options availalble. Let us know if you have other dietary restrictions.