Is speech special? : insights from neonates and neuroimaging

Language is a uniquely human adaptation that is hypothesised to require specialised anatomical substrates and dedicated processing mechanisms. Speech, the primary medium for language, is argued to rely on specialised substrates and processing as well. Yet, to date, evidence for the speech specialisation hypothesis has been equivocal. The four experiments in this thesis aim to advance the discussion in two ways. The differential processing of speech by humans was investigated through the use of functional neuroimaging tools (Experiment One), and through developmental studies of young infants' listening biases (Experiments Two-Four). In Experiment One, functional neuroimaging tools are used to investigate the specificity of neural substrates recruited in detecting speech compared with closely matched non-speech controls. This study takes advantage of an event-related imaging design that provides a narrow window of observation for neural recruitment during individual stimulus events. The results of the first study demonstrate that adults activate specific neural substrates when detecting speech sounds, indicating that specialised substrates are involved from the early processing stages. Experiments Two through Four take an ethological approach to speech specialisation and investigate whether young infants show a bias for listening to speech as compared to matched non-speech sounds. In Experiment Two, behavioural methods probe whether young infants of 2 to 7 months show listening preferences for speech compared with non-speech. Experiment Three seeks to establish the roots of a speech bias in the neonatal period. Finally, Experiment Four investigates the origin of the bias, to determine whether the speech bias originates from prenatal experience, or is independent of specific experience. The results of these studies show that differential processing has its roots in early infancy, with infants demonstrating a preference for listening to speech from birth. The speech bias shown by neonates appears not to be based on specific experience with speech sounds, but instead is rooted in human biology. Human infants are prewired to preferentially attend to speech, granting speech a special status in relation to other sounds.Medicine, Faculty ofGraduat