The influence of speech formulation and articulation on language switch costs

Abstract

Aims and objectives:

This empirical study aimed to investigate which process of speech production, particularly formulation and articulation, plays a crucial role in the occurrence of language switch costs (i.e., slower and/or impaired performance when switching to another language).

Methodology:

To dissociate the roles of formulation and articulation, we employed a cued language-switching task, in which long or short signals were manipulated to ensure complete or incomplete involvement of the formulation process, and Go or Nogo signals to indicate articulation or no-articulation.

Data and analysis:

Data were gathered from two groups of Chinese-English bilingual university students differing in L2 proficiency. An ANOVA analysis was performed, incorporating the formulation/articulation status of the preceding trial, the bilinguals’ L2 proficiency, and the direction of language switching.

Findings:

Only when the preceding trial was articulated that language switch costs were observed in the current articulation trial, and switching to L1 incurs larger costs than switching to L2 (i.e., asymmetric switch costs typically found in the literature). When the preceding trial was not articulated, however, repeating in L1 incurs larger costs than repeating in L2, which can be referred to as asymmetric rearticulation costs.

Originality:

This study manipulated L2 proficiency and considered the factors of cognate status and language families, and yet the main findings about language switch costs remained the same as those revealed by Philipp and Koch. Moreover, it is the first time when asymmetric rearticulation costs were observed, reinforcing the critical role of articulation in language switch costs.

Significance:

With language switch costs replicated in a more diverse context, and with the novel finding of asymmetric rearticulation costs, the critical role of articulation was consistently revealed in language switch costs and probably, in language production.