The Impact of Language Control State on the Cognate Effect: A Replication – Experiment 3

Cognates are translation-equivalent words that are similar or even identical across languages in phonological and orthographic form (e.g., BABY/BABY [English/German]). Previous research has shown that pictures of cognates are more easily named (i.e., fewer errors, faster RTs) by bilinguals than non-cognates. However, such cognate effect may depend on the exact task demands and language control state a bilingual is in. To investigate further the mechanisms behind the cognate effect, we manipulated the proportion of cognates versus non-cognates within blocks of a L2 picture naming task (preregistration: https://doi.org/10.23668/psycharchives.14636). As hypothesized, we observed a larger cognate effect in blocks with more cognates, suggesting that people proactively adapt languages activation to optimize ease of language production. That is, if more cognates are encountered, bilinguals may be able to maintain activation of both languages without much cost, while down-tuning co-activation if the proportion of non-cognates is high. However, these results stand in contrast with the findings by Spinelli and Sulpizio (2024), which did not observe such a modulation of the cognate effect. Thus, we will conduct a replication of our experiment (with some slight adjustments) to better understand why different result patterns were observed across studies.

Cognates are translation-equivalent words that are similar or even identical across languages in phonological and orthographic form (e.g., BABY/BABY [English/German]). Previous research has shown that pictures of cognates are more easily named (i.e., fewer errors, faster RTs) by bilinguals than non-cognates. However, such cognate effect may depend on the exact task demands and language control state a bilingual is in. To investigate further the mechanisms behind the cognate effect, Benini et al. (in preparation) manipulated the proportion of cognates versus non-cognates within blocks of a L2 picture naming task (preregistration: https://doi.org/10.23668/psycharchives.14636). As hypothesized, they observed a larger cognate effect in blocks with more cognates, suggesting that people proactively adapt languages activation to optimize ease of language production. That is, if more cognates are encountered, bilinguals may be able to maintain activation of both languages without much cost, while down-tuning co-activation if the proportion of non-cognates is high. However, these results stand in contrast with the findings by Spinelli and Sulpizio (2024), who did not observe such a modulation of the cognate effect. In the present study, we intend to shed light on the factors that may have produced such an inconsistent pattern of results. In Experiment 1 in this series (preregistration: https://doi.org/10.23668/psycharchives.21221), we set out to replicate the findings of Benini et al. introducing some minor methodological changes. As expected, we replicated the larger cognate facilitation in majority-cognate blocks—a proportion-cognate effect. In the present experiment (Experiment 2 in the series), we used non-identical cognates to investigate the role of cognate similarity in producing the proportion-cognate effect.