Remembering times ahead: The effect of linguistic framing on representational momentum in state-change events

When people remember scenes, they can draw from different sources to reconstruct their memory, such as what they saw, and linguistic descriptions of what they saw. How human memory for events is impacted by linguistic encoding has been a long-standing question in cognitive science. Here, we ask whether language permeates memory of state-change events. We use the visual phenomenon of Representational Momentum, in which people misremember an event they saw (e.g., a melting ice cube) as more completed than it actually was. After three successful replications of the Representational Momentum effect for state changes (total N = 150), in two further experiments, participants (total N = 400) watched partial videos of state change events that were followed by linguistic descriptions in either imperfective aspect (“the ice was melting”) or perfective aspect (“the ice has melted”). As predicted by linguistic theories, perfective aspect increased the Representational Momentum effect, compared to imperfective aspect. In addition, linguistic encoding improved overall memory accuracy. Our work presents the first evidence that the Representational Momentum effect for state change events is sensitive to linguistic framing, specifically the use of grammatical aspect. More generally, our data contribute to a model of cognition in which different kinds of representations interact in memory.