Language-modulated event processing examined through interruption detection and boundary tapping

Languages differ in how they encode object-state changes, potentially biasing attention to phases of unfolding events. Three experiments compared English and Mandarin speakers viewing cutting and blending videos, whose endpoints both groups rated as more delimited and salient for cutting than for blending. In Experiment 1, real-time descriptions showed English speakers using more quantised noun phrases and Mandarin speakers more resultative constructions for blending. English descriptions were also more often telic for cutting, but less often telic for blending, than Mandarin descriptions. Experiment 2 tested whether these encoding preferences aligned with implicit phase sensitivity. During blending, English speakers detected early interruptions more accurately, whereas Mandarin speakers detected late interruptions more accurately. Experiment 3 measured boundary localisation. English speakers located cutting endpoints closer to the physical offset than Mandarin speakers did, while no language difference emerged for blending. Together, the findings suggest language-related attentional biases operating within a shared perceptual structure.