Proceedings of the Annual Meeting of the Cognitive Science Society

Prior research suggests that a small vocabulary of meaningful words (a semantic seed) aids distributional learning. In two experiments, we show that adults who are exposed to a complex artificial language are better at inferring the meaning of previously unknown pseudowords when they were taught a semantic seed prior to distributional exposure. We further show that the benefit of a semantic seed is driven primarily by using seed words to discover the relationship between distributional and semantic classes. These results have implications for how syntactic bootstrapping begins.