Causal chains as a distinct type of event in visual perception

Poster Presentation 53.320: Tuesday, May 19, 2026, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Perceptual Organization: Intuitive physics

Emily Wedin1, Joan Danielle K. Ongchoco1; 1University of British Columbia

The task of perception is to come up with a description of experience. Accordingly, visual processing seems to traffic in properties beyond color or motions, such as *causality*—e.g., one object launching another into motion. But what caused *you* to be here? The answer may be your flight to Florida, your December submission, or other choices along the way—but ultimately any causal chain can be traced all the way back to the Big Bang. Do causal chains belong solely to the domain of reasoning, or might they also play a unique role in visual processing, stringing together events in time? Here, observers enumerated discs in fast, irregular sequences, with the display panning continuously to track up to 30 collisions. In the “causal chain” condition, each disc launched the next, while in the non-causal control, motions were identical, except for small spatial gaps between discs. Observers estimated *more* discs in the causal chain, demonstrating that chain structure helps retain more motions in working memory. To further isolate the “causal chain” from mere causality and from mere spatial continuity, we ran independent experiments with “paired” and “overlap” conditions. In the paired condition, collisions occurred between alternating disc pairs, preserving causality without chain structure. Estimates fell between causal chain and non-causal estimates, suggesting that causality alone does not allow the same degree of retention. In the overlap condition, each disc took over the next disc’s spatial position, providing continuity without causality. Estimates were again reliably different from the causal chain, in fact, even higher—suggesting that spatial continuity may generate a “one is more” illusion, but causality still provides additional structure. Causal chains—as sequences of temporally-linked causal interactions—may then be a distinct type of event in vision, allowing us to *hold on* to more of what we see.