Priming of plane-rotated objects depends on attention and view familiarity

Three experiments investigated the role of attention in visual priming across rotations in the picture plane. Experiment 1 showed that naming latencies increased with the degree of misorientation for objects commonly seen in an upright view (base objects) but not for objects seen familiarly from many views (no-base objects). In Experiment 2, no-base objects revealed a priming pattern identical to that observed previously for left-right reflections (Stankiewicz, Hummel, & Cooper, 1998): Attended objects primed themselves in the same and rotated views, whereas ignored images primed themselves only in the same view, with additive effects of attention and orientation. In Experiment 3 ignored base objects only primed themselves in a familiar (upright) view, indicating that priming only obtains when that image makes contact with object memory. These data challenge theories of object recognition that rely on any single representation of shape and contribute to evidence suggesting holistic (view-like) representations for ignored and analytic (view-insensitive) representations for attended objects.