Where does the delay in L2 picture naming come from? Psycholinguistic and neurocognitive evidence on second language word production

Abstract

Bilinguals are slower when naming a picture in their second language than when naming it in their first language. Although the phenomenon has been frequently replicated, it is not known what causes the delay in the second language. In this article we discuss at what processing stages a delay might arise according to current models of bilingual processing and how the available behavioural and neurocognitive evidence relates to these proposals. Suggested plausible mechanisms, such as frequency or interference effects, are compatible with a naming delay arising at different processing stages. Haemodynamic and electrophysiological data seem to point to a postlexical stage but are still too scarce to support a definite conclusion.

Keywords:

• Bilingual word production
• Picture naming

Acknowledgements

The first author was supported by the grant agency VEGA, Grant No. 2/0204/09.

Notes

¹This phenomenon is sometimes referred to as the “bilingual disadvantage” (Gollan et al., 2005; Ivanova & Costa, 2008). However, it is mostly used to refer to the observation that bilinguals tend to be slower in comparison to monolinguals in their dominant language. Therefore, we will use a more general term “naming delay” where appropriate.

²For an overview of processing stages see Figure 1. Note that Levelt, Roelofs, and Meyer (1999) assume distinct lemma and lexeme (word-form) levels of lexical processing. For a different view see e.g., Caramazza (1997), Caramazza and Miozzo (1998), and Starreveld and La Heij (1996).

³N200 no-go effect refers to a frontally distributed negativity observed on trials where subjects withhold/inhibit a response (no-go trials) in comparison to trials where the response is executed (go trials) (Pfefferbaum, Ford, Weller, & Kopell, 1985; Thorpe, Fize, & Marlot, 1996). The onset and/or peak of the effect can be used to time different cognitive processes even before they show up behaviourally. More on the effect and its use in experiments on bilingual production follows in Section “Studies on the time-course of bilingual lexical access”.

⁴As noted above, participants in the bilingual group were “switched-dominance bilinguals”, and the task was carried out in their dominant language.