Compared to studies of the effects of formal training in translation, little is known about the psycholinguistic impact of the experience of informal translation, or language brokering. The present study examined this issue in the context of idiom comprehension. Bilingual adults differing in prior brokering experience read English idioms and judged whether target words presented in English or Spanish were related to the idiom's meaning. For brokers, relatedness judgments were not affected by whether the targets were in the same or different language as the idiom; however, non-brokers were faster for same-language than different-language idiom-target pairings. The findings suggest that language brokering experience facilitates idiom meaning comprehension even across language boundaries, with further differences related to idiom decomposability. More generally, the findings underscore the importance of considering systematic sources of variability in language practice among bilinguals, aside from differences related to proficiency, in theorizing effects associated with bilingualism.

The present research examined whether conceptual divergence is reduced in bilinguals with extensive informal translation experience. Across two experiments, Spanish–English bilinguals (brokers vs. non-brokers) generated exemplars for 10 categories, using the same or different language across sessions. Both groups demonstrated more divergence for different than same language responses across sessions a week apart. More convergence was found in both groups for no delay compared to delayed responses. Brokers showed significantly more convergence in exemplars than non-brokers; for both immediate and delayed sessions Findings indicate exemplars are differentially accessible depending on language and timing of response, but also individual differences in brokering experience. Extensive brokering experience may lead to a more integrated conceptual representation for features of concepts shared across languages. Findings support concept models that emphasize the dynamic and distributed nature of concepts, and underscore the need to consider the cognitive impact of systematic sources of variability among bilinguals.

This study examined whether training in translation/interpretation leads to a reliance on a ‘vertical’ translation strategy in which the source language text is comprehended before the message is reformulated. Students of translation/interpreting and untrained bilinguals were given an idiom translation judgment task with literal (form and meaning) or figurative equivalents (meaning only). Dependent measures included the time taken to comprehend the first presented sentence and the accuracy and speed of judging if the second presented sentence was a translation of the first sentence. The groups did not differ in their speed of reading the first presented sentence but translation verification times differed by group and translation type: untrained bilinguals were significantly faster at verifying literal than figurative translations while trained bilinguals were equally fast for the two types. The pattern of findings is consistent with the view that training in translation fosters a processing-for-meaning-before reformulating, or vertical, translation strategy.

Evidentiality refers to the linguistic marking of the nature/directness of source of evidence of an asserted event. Some languages (e.g., Turkish) mark source obligatorily in their grammar, while other languages (e.g., English) provide only lexical options for conveying source. The present study examined whether or under what conditions firsthand source information is relied on more than nonfirsthand sources in establishing discourse coherence. Turkish- and English-speaking participants read a series of somewhat incongruous two-sentence narratives and were to come up with a way of completing each narrative so that it would form a coherent story. Each narrative contrasted two source types (firsthand vs. hearsay, firsthand vs. inference, or inference vs. hearsay) and two information types (general vs. particular information) each presented first or second. Analysis of story completions showed greater overall reliance on firsthand information when it was presented second and referred to a particular event. When the firsthand source occurred first and the particular event occurred second, the latter was favored, especially by Turkish participants. Taken together, the findings suggest that evidentiality interacts with information type in establishing discourse coherence and that both firsthand and particular information are relied on more when presented later rather than earlier in discourse.