The cognate facilitation effect in bilingual lexical decision is influenced by stimulus list composition

Abstract

A B S T R A C TCognates share their form and meaning across languages: "winter" in English means the same as "winter" in Dutch. Research has shown that bilinguals process cognates more quickly than words that exist in one language only (e.g. "ant" in English). This finding is taken as strong evidence for the claim that bilinguals have one integrated lexicon and that lexical access is language non-selective. Two English lexical decision experiments with Dutch-English bilinguals investigated whether the cognate facilitation effect is influenced by stimulus list composition. In Experiment 1, the 'standard' version, which included only cognates, English control words and regular non-words, showed significant cognate facilitation (31 ms). In contrast, the 'mixed' version, which also included interlingual homographs, pseudohomophones (instead of regular non-words) and Dutch-only words, showed a significantly different profile: a non-significant disadvantage for the cognates (8 ms). Experiment 2 examined the specific impact of these three additional stimuli types and found that only the inclusion of Dutch words significantly reduced the cognate facilitation effect. Additional exploratory analyses revealed that, when the preceding trial was a Dutch word, cognates were recognised up to 50 ms more slowly than English controls. We suggest that when participants must respond 'no' to non-target language words, competition arises between the 'yes'-and 'no'-responses associated with the two interpretations of a cognate, which (partially) cancels out the facilitation that is a result of the cognate's shared form and meaning. We conclude that the cognate facilitation effect is a real effect that originates in the lexicon, but that cognates can be subject to competition effects outside the lexicon. PsychINFO classification code (Cognitive Processes)

Studies of cross-lingual long-term priming

“…For the unprimed trials, we make the following predictions: (1) we will not find a cognate facilitation effect for the identical or non-identical cognates in either version of the experiment but (2) we will find evidence for an interlingual homograph inhibition effect in the +IH version. The predictions for the identical and non-identical cognates follow from the results of Experiment 1 and Poort & Rodd's (2017, May 9) findings discussed previously. The prediction that we will find evidence for an interlingual homograph inhibition effect also follows from Poort and Rodd's (2017, May 9) experiments, as well as many other experiments (e.g.…”

“…The predictions for the identical and non-identical cognates follow from the results of Experiment 1 and Poort & Rodd's (2017, May 9) findings discussed previously. The prediction that we will find evidence for an interlingual homograph inhibition effect also follows from Poort and Rodd's (2017, May 9) experiments, as well as many other experiments (e.g. Dijkstra et al, 1998), that have shown that inhibition for interlingual homographs is found most consistently when the experiment includes non-target language words.…”

“…Again, the Bayes factors provided strong evidence for the null hypothesis. With regards to the latter finding, this may not be as surprising as it seems, as recent research suggests that the cognate facilitation effect is significantly smaller (even non-existent) when the experiment includes non-target language words (Poort & Rodd, 2017, May 9), as our experiment did. However, the fact that we did not find evidence of priming is unexpected given Poort et al's (2016) results.…”

“…First, we did not find evidence for a cognate facilitation effect for either the identical cognates or the non-identical cognates in either version of the experiment, neither based on our traditional frequentist p-values nor based on the Bayes Factors. Although a great number of studies has found evidence for a cognate facilitation effect in lexical decision (Cristoffanini et al, 1986;De Groot & Nas, 1991;Dijkstra et al, 1999;Dijkstra et al, 2010;Dijkstra et al, 1998;Font, 2001;Lemhöfer & Dijkstra, 2004;Lemhöfer et al, 2008;Peeters et al, 2013;Sánchez-Casas et al, 1992;Van Hell & Dijkstra, 2002), recent research conducted by Poort and Rodd (2017, May 9) indicates that this effect can disappear when the experiment includes nontarget language words like our Dutch words that the participants are required to respond 'no' to. Our data are fully consistent with their findings and further show that the same appears to be true for non-identical cognates.…”

Inhibitory and Facilitatory Effects of Phonological and Orthographic Similarity on Bilingual Word Recognition

“…Despite a large amount of literature exploring the cognate effect, the focus has been quite narrow, placing the emphasis on orthographic similarity and using visual word recognition paradigms [11][12][13][14][15] . Orthographic similarity manipulations in the visual modality consistently lead to a facilitation effect of cognates in perception.…”