University of Miami
Recent research demonstrates a link between acquisition and iconicity, or the resemblance between word forms and meanings, in natural languages. For example, I have found statistical regularities in the lexicons of English and Spanish: early-learned words in both languages tend to be higher in iconicity than later-learned words, (even after controlling for related factors, e.g., concreteness). To better understand the statistical structure of the speech children produce and hear, I also conducted a corpus analysis of children?s speech, child-directed speech, and adult-directed speech in English. I found that the higher a word was rated in iconicity, the more frequently it appeared in children?s speech and adults? child-directed speech. On the other hand, the lower a word was rated in iconicity the more frequently it appears in adult-directed speech. Together, findings suggest iconicity is not distributed equally across the lexicon but specifically characterizes early-learned words and early parent-child conversations. This statistical distribution suggests that iconicity is particularly important early in development and may serve to bootstrap language acquisition. In this talk I will discuss how understanding the statistical regularities present in children?s language environment (including speech, storybooks, and songs) can inform our understanding of language development mechanisms.