When a moving disc stops right next to a second disc that then starts moving in the same direction, we typically perceive a causal launch. Increasing spatial overlap between the two discs reduces the perception of launches such that one disc appears to pass over the other. This basic phenomenology is subject to contextual influences. The impression of a pass can be switched to a launch by presenting a launching context event simultaneously with the test event (Scholl & Nakayama, 2002). Here we investigated the impact of visual adaptation to launches (Rolfs, Dambacher, & Cavanagh, 2013) on these contextual influences.
We presented test events with parametrically varying amounts of disc overlap and asked observers to report whether they perceived a causal launch or non-causal pass—allowing us to obtain psychometric functions for the perception of causality. Before adaptation, a simultaneous launch context event increased (while a pass context event decreased) the proportion of reported launches in ambiguous events. Subsequent adaptation significantly decreased the proportion of reported launches in trials without a context event. However, simultaneous launch context events counteracted the influence of visual adaptation, such that the proportion of reported launches was unaffected by the visual adaptation.
Thus, attenuating the perception of causality did not abolish contextual influences. On the contrary, contextual influences emerged strongest after adaptation. We conclude that weaker evidence for the presence of a causal interaction at the test event location—here induced by adaptation to launches—increased the influence of context on the perception of causality.