The domain of figurative language comprehension was used to probe the developmental relation between language and cognition in typically developing individuals and individuals with Williams syndrome. Extending the work of Vosniadou and Ortony (1983), the emergence of non-literal similarity and category knowledge was investigated in 117 typically developing children aged between 4 and 12, 19 typically developing adults, 15 children with Williams syndrome between 5 and 12 years of age, and 8 adults with Williams syndrome. Participants were required to complete similarity and categorization statements by selecting one of two words (e.g., either "The sun is like...?" or "The sun is the same kind of thing as...?") with word pairs formed from items that were literally, perceptually, or functionally similar to the target word, or else anomalous (e.g., 'moon', 'orange', 'oven', or 'chair', respectively). Results indicated that individuals with Williams syndrome may access different, less abstract knowledge in figurative language comparisons, despite the relatively strong verbal abilities found in this disorder.
3Although metaphor and analogy have traditionally been viewed as a relatively rare linguistic ornament that complement literal language, recent research suggests that metaphorical language is, in fact, commonplace in everyday communication (e.g., Graesser, Long & Mio, 1989;Pollio, Barlow, Fine & Pollio, 1977). For example, various strands of linguistic evidence suggest that metaphor is important for communicating, and perhaps reasoning, about abstract concepts (e.g. Gibbs, 1994;Lakoff & Johnson, 1980). Understanding metaphorical language necessitates certain degrees of proficiency in both cognition and language, and relies on several component abilities, such as processing capacity, metalinguistic skill, an understanding of communicative pragmatics and semantic knowledge (see Vosniadou, 1987aVosniadou, , 1987b.Although definitions of what constitutes a metaphorical comparison vary considerably (e.g., Bowdle & Gentner, 2005), it seems uncontroversial that understanding verbal metaphor involves accessing some similarity between two terms, while recognizing that two terms belong to different conventional categories. This, of course, necessitates knowledge of such categories. If so, then one can take an initial step in ascertaining whether children are capable of understanding metaphor at a given stage in development by investigating their ability to understand non-literal similarity statements, that is, that items that fall in different semantic or conceptual categories can nevertheless be similar in certain ways. Vosniadou and Ortony (1983) required children between the ages of 3 and 6, as well as adults, to complete similarity statements by selecting one of two words from either (1) a metaphorical/literal pair of alternatives, (2) a literal/anomalous pair, or (3), a metaphorical/anomalous pair. For example, the experimenter would say "A river is like a...?" and the participant was required to respond with...