Noun translation equivalents that share orthographic and semantic features, called "cognates", are generally recognized faster than translation equivalents without such overlap. This cognate effect, which has also been obtained when cognates and noncognates were embedded in a sentence context, emerges from the coactivation of representations in two languages. The present study examined whether cognate facilitation in sentences is subject to effects of word class, reading proficiency in a second language (L2), and task demands. We measured eye movements (Experiment 1) and self-paced reading times (Experiment 2) for Dutch-English bilinguals reading L2 sentences that contained either a noun or a verb cognate. Results showed that cognate effects were smaller for verbs than for nouns. Furthermore, cognate facilitation was reduced for readers with a higher proficiency in L2 as expressed by self-ratings or reading speed in L2. Additionally, the results of the eye-movement study and the self-paced reading study indicated that the likelihood of observing cognate facilitation effects also depends on task demands. The obtained pattern of results helps to identify some of the boundaries of the cognate effect.

This study investigated two prominent issues in the comprehension of language switches. First, how does language switching direction affect switch costs in sentence context? Second, are switch costs modulated by L2 proficiency and cross-linguistic activation? We conducted a self-paced reading task involving sentences that switched between participants’ L1 Dutch and L2 English. The cognate status of the main verb was manipulated to examine the influence of co-activation on intra-sentential switch costs. The reading times indicated the influence of switch direction: a cost was observed for switches into L2 but not for switches into L1, and the magnitude of the costs was correlated with L2 proficiency, indicating that switch costs in language comprehension depend on language dominance. Verb cognates did not yield a cognate facilitation effect nor did they influence the magnitude of switch costs in either direction. The results are interpreted in terms of an activation account explaining lexical comprehension based on L2 proficiency.

In bilingual processing, cognates are associated with facilitatory processing, while switching between languages is associated with a processing cost. This study investigates whether co-activation of cognates affects the magnitude of switch costs in sentence context. A shadowing task was conducted to examine whether verb cognates reduce switch costs in sentences that switched between participants' LI and L2. In addition, we considered whether these effects were influenced by L2 proficiency, switching direction and cross-linguistic overlap in syntactic structure. Bilinguals were presented with LI and L2 sentences that contained a language switch preceded by a cognate. Shadowing latencies showed that switching to L2 was more costly than switching to LI. Switch costs in both directions were not modulated by the presence of a verb cognate, and this effect was not affected by syntactic structure or L2 proficiency. The results are informative for the field of bilingual processing and the lexical trigger hypothesis.

This paper presents further evidence that cognates may facilitate code switching. In three corpora of natural speech, code switches occur more often directly following or in the same clause as a cognate (or 'trigger word') than elsewhere. Triggered code switching is found between typologically similar languages, with Dutch L1-English L2 speakers in New Zealand and Australia, and between typologically distant languages, with a Russian L1-English L2 speaker in the USA. We find that words following but not words preceding a trigger word have an increased chance of being code switched, that form overlap without meaning overlap may be sufficient for triggering to take place, and that the attachment of extensive Russian morphology to a trigger word stem does not diminish its triggering potential. We do not find an effect of the interlocutor's use of trigger words. Further, discourse connectors are often used in the vicinity of code switches.

No abstract