As with many other skills, the acquisition of a second language is characterised by a gradual shift from effortful, conscious processing to increased automaticity and fine-grained control. Fluency requires not just the acquisition of a second language's words and structures, but the ability to efficiently allocate processing resources to effectively access, select, and manipulate such representations. Because semantic priming is believed to reflect both automatic and controlled mechanisms in comprehension (Neely, 1991), it provides an optimal lens through which to examine the development of such skilful behaviour.
Thus, sixty bilinguals of low and high proficiencies (30 Italian-English and 30 English-Italian) were tested on an auditory sentence-priming task in their L1 and L2. Additionally, 20 native English speakers were tested to serve as a monolingual baseline. Sentences varied in their contextual constraint and their semantic congruency with target words, thereby generating five priming conditions (congruent-strong, congruent-weak, neutral, incongruent-weak, incongruent-strong). Facilitation priming was calculated by finding the difference between congruent and neutral lexical decision reaction times, while inhibition priming reflected the difference between incongruent and neutral trials.
Bilinguals performed similarly as monolinguals in their L1, demonstrating graded sensitivity to contextual constraint in facilitatory, but not inhibitory, contexts. Priming patterns in their L2 differed significantly, however, depending on proficiency and language. High proficiency Italian-English bilinguals replicated monolingual patterns while low proficiency Italian-English bilinguals demonstrated facilitation and inhibition but no sensitivity to contextual constraint. In contrast, for English-Italian bilinguals, high proficiency speakers were not sensitive to constraint, and low proficiency speakers showed no inhibition. Priming was greater for both groups in English.
These unexpected differences between the two bilingual groups, equated in proficiency, likely reflect their greater relative activation of English. Given that the bilinguals tested live in England, it appears English was more strongly activated for both groups and thus benefited more from priming. The lesser sensitivity to contextual constraint in L2 suggests weaker representations and/or links in its semantic network, particularly when low proficiency. The inconsistent inhibition experienced by low proficiency speakers further suggests the attentional demands of L2 comprehension limited resources available for the full set of word recognition mechanisms. Moreover, the controlled mechanisms underlying inhibition priming appear to develop later than the automatic processing generating facilitation effects.
References
Neely, J. H. (1991). Semantic priming effects in visual word recognition: A selective review of current findings and theories. In D. Besner & G. Humphreys (Eds.) Basic Processes in Reading: Visual Word Recognition (pp.264-336). Hillsdale, NJ: Erlbaum.