Tots Take Rhythmic Stock before Talk
Bower, Bruce, Science News
Scientists for the first time have evidence that, beginning as early as 2 months of age, babies intermittently babble approximately 3-second-long "prelinguistic phrases" characterized by a rhythm and structure that later underlie speech.
Both healthy infants and those with Down's syndrome, which usually includes severe language delays, vocalize in this way, assert Michael P. Lynch, a psychologist at Purdue University in West Lafayette, Ind., and his colleagues. However, for as yet unclear reasons, babies with Down's syndrome take considerably longer to finish prelinguistic phrases.
The new study is slated to appear in an upcoming DEVELOPMENTAL PSYCHOBIOLOGY. Other evidence gathered by Lynch's group, and accepted for publication in the AMERICAN JOURNAL OF MENTAL RETARDATION, indicates that babies with Down's syndrome utter their first speechlike syllables at an average age of 9 months, 2 months later than neurologically healthy infants. Previously, the earliest documented speech problems in Down's syndrome appeared at age 2.
These findings open the possibility of eventually using acoustic measures of the sounds babies make to identify those most likely to experience childhood speech and language problems, according to Lynch.
"The organization of prelinguistic phrases corresponds to the way in which adults use intonation and rhythm to segment a stream of speech," the Purdue scientist argues. "Babies may develop parsing mechanisms for vocalizing that later get modified for language use."
Lynch and his coworkers made monthly audio recordings of 16 infants from age 2 months to 1 year, eight of them healthy and eight with Down's syndrome. During each 20-minute session, babies played with a parent and an experimenter in a special acoustic chamber that held a number of quiet toys.
Acoustic data were gathered on infant utterances, defined as vocalizations bounded by an audible inhalation or at least 1 second of silence. Printouts showing selected sequences of utterances and types and numbers of syllables in each utterance were then given to seven adults with no training in coding infant vocalizations. …