Abstract
Ambiguous motion stimuli can be disambiguated in accord with one’s intentional control (e.g., Kohler, Haddad, Singer, & Muckli, 2008). However, past studies have relied on simple ambiguous test stimuli in which motion could be disambiguated in a limited number of directions (e.g., vertical or horizontal). Recently, Davidenko and colleagues (Davidenko, Heller, Cheong, & Smith, 2017) reported illusory apparent motion (IAM), in which random pixel textures refreshing between 1–3Hz elicit percepts of globally coherent motion that can be perceived as moving in any number of directions (diagonal, rotating, contracting, etc.). The current study investigates intentional control of ambiguous motion within the context of IAM. Stimuli consisted of 15 frames of randomized 140 × 140 pixel textures refreshing at 1.5 Hz. In each 10-second trial, participants (n=49) were instructed to intentionally hold horizontal or vertical motion, or change between vertical and horizontal motion as fast as possible, while reporting any vertical or horizontal motion percepts as they occurred by holding down one of two buttons. To account for experimental demand, we confirmed in catch trials that participants correctly reported motion when it was present, regardless of instruction. Results showed a robust effect of instructions on reported percepts, with participants reporting much more vertical (M: 6.2 s) than horizontal (0.3 s) motion during vertical hold trials (t(48)=11.6, p< 0.00001), and much more horizontal (4.7 s) than vertical (0.8 s) motion during horizontal hold trials (t(48)=5.88, p< 0.00001). In contrast, during change trials, participants reported horizontal (2.0 s) and vertical (3.3 s) motion at more comparable rates (although there was a significant vertical bias, t(48)=3.64, p=0.0007). Our results demonstrate that intentional control is possible within the context of IAM.