The introduction also offers other areas for which this research is useful, for example, “language acquisition, cortical plasticity and performance theory, might be advanced by an increased understanding of the ways in which multiple languages are represented in the bilingual speakers’ brain/mind” (Perani et al. 1998, p. 1841). The focus of the research is on high proficiency, late acquisition groups (HPLA), high proficiency, early acquisition groups (HPEA) in comparison to low proficiency, late acquisition groups (LPLA).
The research involved two experiments. The first included “nine right-handed male native speakers of Italian” whose L2 was English; their story comprehension proficiency was analyzed during a PET scan experiment with “two regional cerebral blood flow (rCBF) measures taken (Perani et al. 1998, p. 1843-1844). The second experiment designed to study “the effect of age of [L2] acquisition” with subjects who spoke Spanish and Catalan; these subjects were analyzed by an ANOVA as well as a PET scan (Perani et al. 1998, p. 1843). Upon performing the research, Perani et al. discovered that there were very different brain pattern results in LPLA groups, but were surprised to discover “no such major difference . . . found in high proficiency subjects (HPLA and HPEA groups), regardless of the age of L2 acquisition (1988, p. 1845, 1848). In general, the results of the study show that as long as high proficiency is achieved in a language, it does not matter at what age learning a L2 begins.
The method used to analyze the collected data for experiment one with the Italian-English speakers involved analyzing the PET data “using Statistical Parametric Mapping (SPM95)” utilizing a general linear model to compare the “differences of activation patterns between the HPLA and the LPLA English-Italian bilinguals” (Perani et al. 1998, p. 1844-1845). The researchers used the baseline tasks of backward speech and attentive silence as baselines.
The method used to analyze the collected data for experiment two with the Spanish-Catalan speakers employed similar techniques as experiment one, except the experimenters eliminated the use of attentive silence as a baseline control condition because they found in experiment one that “backward speech had proved to be adequately informative” (Perani et al. 1998, p. 1844). SPM95 was also used to analyze the PET data in experiment two.
This research and results provided by this article will be of great interest to language teachers who wish to promote the idea of learning a new language at any age, since the research results demonstrate that it is possible to obtain a true bilingual proficiency at any age. The sample sizes for experiments one and two were both small, with nine subjects in experiment one and twelve subjects in experiment two (Perani et al. 1998, p. 1842-1843). However, because of the specific nature of this study, such as with experiment two which required testing subjects that knew Spanish and Catalan equally well, this small sample was taken from a larger sample of over 80 individuals “who were selected through behavioral tests and directed interviews” to make sure they were suitable for the study (Perani et al. 1998, p. 1843). Therefore, the sample size was adequate for this study which used a rigorous selection process.
The independent variables were the L1 and L2 languages spoken by the subjects, while the dependent variables were the results of the PET scans. Tables 1A and 1B show the areas of the brain tested by the PET scan during the experiment, while Figure 2 provides a graphic illustration of “Brain areas of greater activation in HPLA subjects compared with LPLA subjects for L2 (English)” (Perani et al. 1998, p. 1846-1847).
The validity of the data is not so much questioned by the researchers, who designed a rigorous study with quantifiable results, but it surprised the researchers, because they write, “The lack of differences between the cerebral activations associated with L1 and L2 in highly proficient individuals is surprising, because it seems to mesh poorly with behavioural data obtained in many different experimental settings” (Perani et al. 1998, p. 1848). For instance, they discuss research involving infants raised “in a bilingual setting from birth” in comparison to infants who are “not precociously exposed to two languages” that develop a dominance in only one of the languages he or she is exposed to (Perani et al. 1998, p. 1849). The authors find it hard to “reconcile the discrepancy [observed] between the [PET] imaging data (largely similar activations with L1 and L2 in highly proficient individuals, regardless of age and acquisition)” with the behavioral findings that in reality, L1 and L2 are utilized differently even in subjects who know both languages well (Perani et al. 1998, p. 1849). The apparent surprise of the researchers is not a weakness of the study, but opens up further questions for research that have not previously been explored before.
The study provides some interesting results for language teachers, but the primary audience is researchers in the field of language acquisition and neurology. This study provides a lot of technical information not as useful to a language teacher as it will be to a researcher looking for ways to contribute to the field of research surrounding learning a second language.
Perani, Daniela, Paulesu, Eraldo, Sebastien Galles, Nuria, Dupoux, Emmanuel, Dehaene, Stanislas, Bettinardi, Valentino, Cappa, Stefano F., Fazio, Ferruccio, & Mehler, Jacques (1998). The Bilingual Brain: Proficiency and Age of Acquisition of the Second Language. Brain 121. 1841-1852.