Inferring native language from early bio-electrical activity

Abstract

Early semantic effects (<200 ms) for single word processing have been reported inconsistently, though not rarely, in the ERP literature. It may be hypothesized that the frequently reported later lexical effects are related to the lengthy processing of long/scarcely familiar words or task factors. In this paper, we have investigated the timing of brain activation during processing of native vs. later-acquired languages in simultaneous interpreters. The data show a clear-cut difference between native and foreign languages (proficiency being equal) in a silent letter detection task not requiring semantic processing. Although interpreters were equally proficient in L1 (Italian) and L2 (English), only L1 showed early lexical effects at occipito/temporal sites at about 160–180 ms. L2 words were distinguished from pseudo-words at about 260–320 ms (N2 level), while L3 (German) words showed a lexical effect only at about 320–380 ms (N3 level) at posterior sites. The fact that only L1 words were discriminated from pseudo-words at the earliest processing stage suggests a faster/more efficient access to lexicon for L1 than for later-acquired languages, regardless of proficiency.