Proceedings of the Annual Meeting of the Cognitive Science Society

Mounting evidence suggests that people make use of non-arbitrary relationships between word form and meaning (e.g.,rounded vowels and rounded shapes) when determining the meaning of a novel word. Typically, these studies use carefullyselected materials to maximize iconic relationships between word forms and meanings. Can people make use of form-meaning resemblances for randomly selected word-forms? We gave 21 groups of undergraduates 40 randomly generatednonce words and asked them to draw a creature for each word such that a nave viewer could reliably match the creature-drawing back to the word that motivated it. Despite the words being selected randomly and filtering out any reliance onexisting English words, drawings were routinely matched back by nave participants (n=222) at rates well above chance.We discuss possible explanations for what makes certain words fit an especially good fit for certain drawings.