Information processing in bilinguals: Some selected issues

Summary Issues concerning some perceptual, attentional, and memory processes in bilinguals are reviewed and discussed with regard to language dominance and the interaction between the bilingual's language systems. The first part of the paper focuses on the speed of basic decoding and encoding operations: speed (automaticity) of processing is considered to be a major factor in language dominance. Speed of perception and rehearsal is shown to be affected by covert pronounceability of words which is typically poorer for the bilingual's weaker language. Factors causing short-time changes in the relationship between the non-balanced bilingual's languages are then considered: high information load, noise, rapid switching, and language set. Noise is thought of as enhancing language dominance by two mechanisms: the masking of inner speech, and increased arousal. The second part of the paper focuses mainly on the interaction between the bilingual's language systems. The issue of language independence/interdependence is illustrated by a task involving parallel encoding and decoding; it is shown that the input switch which is considered to be automatic can select information by language. Experiments on memory are commented on with reference to the two major models of bilingual-storage systems: the common-store hypothesis, and the separate-stores hypothesis. While most experimental data are clearly supportive of the former, it is argued that both language-specific and language-free storage takes place, depending on task demands. The problem of the retention of language-information along with item-information in verbal learning tasks is then discussed: in addition to the specific task demands (e.g., meaningfulness), retention time appears to be a decisive factor here. A model of bilingual memory storage is suggested which combines the two extreme models (i.e., the shared and the separate ones). The model assumes that working memory plays a different role with regard to the language-free and the language-specific stores respectively. Finally, some problems in the use of more than one language in the process of verbal learning are pointed out.Most of the research work reported in this paper was supported by grants from the Swedish Council for Social Science Research (now the Swedish Council for Research in Humanities and Social Sciences).