Tasha Young¹ (), Joan Danielle K. Ongchoco¹; ¹The University of British Columbia

Natural scenes are distinguishable not just by their color (i.e., more green hues) or edges (i.e., more entropy), but also by their *rhythms* — moving shadows cast from sunlight through trees, or ripples in flowing water. Here, we ask what spatiotemporal properties define these ‘visual rhythms’ of nature. Is it the unpredictable *spatial* patterns as light hits moving surfaces? Or is it the unpredictable *tempos* at which things change — especially in contrast to the more predictable and persistent rhythms in urban settings, such as blinking traffic lights or the relentless ticking clock? Across experiments, observers viewed dynamic displays of randomly distributed pairs of discs that alternated between light and dark shades of gray at varied tempos — to create a sense of ‘visual rhythm’. Sometimes, the colors the discs alternated between were resampled at given intervals (mimicking unpredictable spatial patterns). Other times, the tempos at which discs alternated were resampled at given intervals (mimicking unpredictable tempos). We then systematically investigated the degree of unpredictability in these featural and temporal properties: (1) maximum unpredictability, where both colors and tempos were resampled; (2) featural unpredictability, where colors were resampled, but tempos remained constant; (3) temporal unpredictability, where tempos were resampled, but colors remained constant; and (4) maximum predictability, where neither were resampled. After the rhythm played, an image of a scene appeared, and observers had to report whether it was natural or urban as quickly as possible. Results showed that only the maximum unpredictability case facilitated priming of natural (versus urban) scenes — and this was reliably greater than in all other conditions that were predictable in at least one dimension, suggesting that predictability may cue us *out* of nature. For the first time, we isolate a specific configuration of spatiotemporal properties in visual rhythms that can activate natural scene perception.

Acknowledgements: This work was supported by the Hampton Research Grant awarded to J.D.K.O.