Sir

M. Piattelli-Palmarini, in his review (Nature 411, 887–888; 2001) of the book Pathways to Language, criticises authors Kyra Karmiloff and Annette Karmiloff-Smith for misrepresenting the linguistic and cognitive abilities of children with Williams syndrome (WMS). Contrary to the book's authors, your reviewer claims that “children with Williams syndrome have a barely measurable general intelligence” but “an exquisite mastery of syntax and vocabulary”, although they are “unable to understand even the most immediate implications of their admirably constructed sentences”.

Based on our own extensive research on WMS in English and Italian, we disagree with your reviewer. First, intelligence is certainly testable in WMS. In older children, the average performance IQ is around 60, and many score much higher. By the age of 8–10 years, children with WMS are typically functioning at a mental age of 5–6, an age at which normally developing children display sophisticated vocabulary and complex grammar.

Second, syntax is far from perfect in WMS. In younger children, lexical and grammatical development are delayed; in older children, difficulties and errors persist. Overall levels of syntax never exceed mental age.

Third, comprehension abilities in WMS are often puzzling, but they are much more sophisticated than Piattelli-Palmarini implies. Individuals with WMS are extremely interesting for research on language, cognition and social functions, because they display an unusual profile of strengths and weaknesses that may be linked to the genetic alterations responsible for this syndrome. But they are not language savants, and do not provide evidence for intact language in the absence of measurable intelligence.

In short, contrary to your reviewer, we believe that Karmiloff and Karmiloff-Smith 'got it right'.