Branzi, F. M. 1 , Fitzpatrick, I. 1 , Della Rosa, P. A. 2 , Abutalebi, J. 2 & Costa, A. 1, 3
1 Universitat Pompeu Fabra, Barcelona, Spain
2 Università Vita-Salute San Raffaele, Milan, Italy
3 Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain
The ability of highly-proficent bilinguals to control their languages seems to depend on cognitive processes and brain areas largely shared with the domain general cognitive control system (Abutalebi & Green, 2007; Costa et al., 2008). However, recent findings could be taken to suggest that certain cognitive control processes are unique to the bilingual language domain (e.g., Abutalebi et al., 2007). The present study aimed to evaluate whether language switching incurs an additional processing cost compared to domain general task switching by investigating the effects of changes in task set and stimulus-response associations in both a linguistic and a non-linguistic task.
We tested Spanish-Catalan bilinguals undergoing: (1) a three block, picture naming task; and (2) a three block, non-linguistic task that required subjects to make a size judgment (cf., Dobbins et al., 2004) on the stimuli by means of a two button, forced-choice response. In the former task the naming language (i.e., Spanish or Catalan) was alternated between blocks, in the latter task the classification instruction was alternated (i.e., "Is the object bigger than a shoebox?" or "Is the object smaller than a shoebox?") while the response mapping remained fixed (i.e., "Left button for yes, right button for no"). Both tasks used the same picture stimuli, half of which were Spanish-Catalan cognates. In the second and third blocks half the stimuli were repeated (once) from a previous block, and were thus already associated with a prepotent response (language or response hand). The order of the tasks was counterbalanced across participants.
We obtained a cognate facilitation effect in the language switching task, but no effect in the non-linguistic task. Moreover, our results revealed a greater overall cost for language switching compared to task-set switching, suggesting that producing language switches requires greater recruitment of cognitive control resources compared to domain general switching.