A dual route of prediction-by-production and prediction-by-association during simultaneous interpreting: Evidence from the visual world paradigm

Despite growing interest in prediction during simultaneous interpreting (SI), the real-time processing mechanisms supporting it remain underexplored. This study employed the visual world paradigm to investigate whether interpreters can predict upcoming content while simultaneously interpreting multi-sentence paragraphs and to examine the mechanisms underlying prediction. Interpreting students and professionals simultaneously interpreted four paragraphs embedded with sentences containing a critical verb that manipulates the predictability of the target noun, while viewing visual displays containing a target object, two semantic competitor objects and one distractor object. Both groups made predictive eye movements to the target objects before hearing the corresponding word, indicating interpreters’ ability to predict in a challenging task. The observed fixation patterns further suggest the involvement of both prediction-by-production and prediction-by-association during SI. Crucially, professionals showed more flexible attention shifts and efficient cue use, whereas students shifted attention less and used a more cautious prediction strategy.