---

Effects of Age of Acquisition on Cortical Organization for American Sign Language:
an fMRI Study

---

Introduction:
  We have reported (1) that when native speakers of English read English sentences there is repeated and robust cortical activation (as indexed by fMRI) within classical language areas and the superior temporal sulcus of the left hemisphere (LH), but little activation within the right hemisphere (RH). Parallel studies of native signers viewing ASL (2) reveal similar activation within the LH, as well as extensive activation within frontal and temporal regions of the RH. Several lines of evidence suggest that LH specialization for English is dependent upon early exposure to that language. The present study was designed to assess the hypothesis that early exposure to ASL is essential for the recruitment of RH regions for its processing.

Methods:
  Changes in blood oxygenation/flow level (MR imaging at 4T: 8 sagittal slices (5mm), gradient echo MBEST< 64x64, TR=4sec, TE=28ms, FOV=16cm) were recorded while 21 normally hearing, right-handed, young adults viewed filmed ASL sentences that alternated with nonsigns formally similar to ASL signs, and in separate runs, written English sentences alternating with consonant strings. Ten of the subjects were native signers, born to deaf parents from whom they acquired ASL at the normal age for primary language acquisition. These participants also acquired aural-oral English. Eleven subjects learned English as a first language and then acquired ASL at the age of >15 years in the course of becoming trained as interpreters of ASL. All subjects were trained interpreters. Subjects were asked recognition questions at the end of each run to ensure that their attention was focused on the stimuli. The data were analysed in two ways. Statistical Parametric Mapping (SPM'96, 3) was applied to find clusters of activation, across subjects, via spatial normalization of each brain to a standard template. We were unable to apply this method to the data of the early learners of ASL, however, because we were able to acquire only 8 structural slices from each hemisphere of each subject, which was insufficient for the SPM normalization algorithm. Thus we employed a region-of-interest (ROI)-based averaging of activations detected by a cross-correlation technique (1, 4), to compare the two groups of subjects (early vs. late learners). This latter method also allowed comparison of the present results with those from related studies (1,2) from our lab,

Results:
When reading English sentences, all subjects showed asymmetrical activation most pronounced within the LH. When processing ASL, hearing native signers displayed bilateral activation within inferior frontal and superior temporal regions. In contrast, late learners showed patterns of activation related to ASL processing which were very similar to those seen when they read English. Thus while early learners of sign language demonstrate quite distinct patterns for each of their two native languages (English and ASL), late learners do not show this difference. These results imply that, as has been noted for other natural languages, delayed exposure to ASL disrupts the normative (i.e. native-like) pattern of cerebral organization for that language. Specifically, early exposure to ASL may be necessary for the recruitment of specialized RH regions for the processing of that language.

handout of the poster in PDF format (with images)

References:
1. Bavelier, et al., 1997, J. Cognitive Neuroscience, 9:664-686.
2. Neville et al., 1998, Proceedings of the National Academy of Sciences, 95, 922-929.
3. Wellcome Department of Cognitive Neurology, London, UK
4. Bandettini et al., 1993, Magnetic Resonance in Medicine, 30:161-173.

Supported by NINDS DC00128 and the McDonnell-Pew Foundation. A.N. supported by a postgraduate scholarship from the Canadian Natural Sciences and Engineering Research Council.