This thesis presents an investigation of bilingualism as a motor learning skill, with
success ultimately measured in terms of strength of a foreign accent, in contrast to
the many studies of bilingualism in terms of linguistic competence. My research used
functional magnetic resonance (fMRI) imaging to investigate feedforward (motor) and
feedback (auditory and somatosensory) systems involved in the production of foreign
speech and how these systems are modulated by proficiency levels. I investigated
the function of the frontal operculum and the temporo-parietal junction (TPJ) –
planum temporale (posterior auditory association cortex) and parietal operculum
(somatosensory association cortex) – during speech. The frontal operculum, strongly
lateralised to the left, has been associated with speech since Broca performed his
classic post mortem lesion-deficit analysis. Interest in the TPJ has arisen because of
recent publications proposing the posterior half of the left planum temporale (±
adjacent parietal operculum) as a ‘sensorimotor interface’ for speech production. My
research compared activity within the frontal operculum and the TPJ during overt and
covert speech. A second fMRI study examined retrospective proficiency based on
existing language skills in people with English as a foreign language who were
scanned during speech production in their native language and in English. A third
fMRI study manipulated proficiency by training monolingual native English
participants in the production of foreign speech sounds, with scanning pre- and post-training.
This allowed measures of changes in activity (indicating rapid plasticity)
following a short period of behavioural training in articulating novel foreign speech
sounds. Training effects were observed predominantly in the striatum, and further
analyses indicated that striatal activity in vocal learning is modulated by proficiency.