Biological movement and the encoding of approach

We investigated the role of biological movement in the encoding of walker-signalled approach. Two stimuli were used: a walking mannequin and an oscillating ovoid, both with chequerboard textures. The ovoid had the same approximate dimensions as the mannequin and oscillated about its minor, horizontal axis at the same frequency as the gait cycle of the mannequin. Stimuli could be rotated in the horizontal plane, with 0° indicating that the mannequin or ovoid directly faced the observer, and positive angles indicating leftwards rotation. Subjects reported whether stimuli headed to their left or their right, and we calculated their angle of subjective approach (ASA). Using an adaptation-top-up method with on-screen adaptors 25% larger than test stimuli, we measured ASAs during adaptation to adaptors angled at -25° and at +25°. Typically, due to adaptor repulsion, -25° adaptation results in a positive ASA whilst +25° adaptation results in a negative ASA. We defined heading aftereffect as half the difference between these two ASAs. When adaptor and test were the same stimulus type we found robust heading aftereffects. When they differed, the aftereffect was abolished – so the aftereffect is not a simple response bias. Next we compared our mannequin heading aftereffects for forward and reversed motion mannequin adaptors. All test sequences were standard forward motion. Both forward and reversed adaptors elicited robust aftereffects, with the latter significantly weaker. Finally, we measured direction contingent heading aftereffects by adapting subjects to interleaved segments of forwards mannequin at -25° and reversed mannequin at +25° (and vice versa). We found that heading aftereffects were contingent upon the direction (forward or reversed) of the test stimulus. Our results show that, beyond influences such as body orientation and direction of limb trajectory, the actual biological motion itself factors into the percept of direction in our encoding of approach.