A partial solution to the homonym problem: parents' linguistic input to young children.

During early word learning, children may assume a unique form-meaning relationship: that a unique form corresponds to each meaning (Clark, 1987) and vice versa (Slobin, 1985). Homonyms appear to violate this one-to-one mapping, and therefore might prove problematic; for example, all things labeled bat might be put into the odd category 'things that fly and are used to play ball.' We asked whether parents' descriptions of homonyms to their young children provide linguistic information that could help children to differentiate conceptually and linguistically these cases. Fifteen parents described pictured sets of concrete object homonyms and categorically related objects to their young children. Parents described homonyms by using explicit statements of category inclusion, subordinate forms (compounds, e.g., kiwi-bird), and specific forms that flagged the unusual homonym relationship. Parents especially provided distinctive alternating forms for homonym pairs (e.g., iceskate vs. skate-fish) suggesting that they were attempting to preserve the unique form-meaning relationship violated by homonyms. We discuss how such linguistic information might be used by children to differentiate homonyms.