Perceptual expertise affects visual short-term memory and the time-course of perceptual encoding

Abstract

Shelves are stocked with endless books offering advice on how to increase one’s memory capacity, but it is unclear if all types of memory are open to improvement. For example, according to one prominent theory, visual short-term memory (VSTM) may have a fixed object-based limit. The current results challenge this theory by bridging two large but as yet unconnected literatures—those on VSTM and visual expertise—to demonstrate that visual expertise can influence VSTM capacity for complex objects. In particular, I focus on the influence of an encoding bias associated with expertise, namely the tendency to process faces or other objects of expertise holistically rather than in a more feature-based manner. A probe recognition match-to-sample task was used in seven experiments. A VSTM advantage for holistically processed faces over more featurally processed non-face objects and inverted faces emerged—but only with sufficient encoding time. This advantage extended to non-face expert categories, such as cars among car experts, which have also been shown to recruit holistic processing. Again, this effect only emerged with sufficient encoding time. Additional studies refuted alternative possibilities that this VSTM advantage might be dependent on verbal short-term memory, long-term memory, or eye-movement strategies. Experiments 8 and 9 further explored whether the dependency of the expert VSTM advantage on encoding time is related to capacity limitations occurring in early visual processing. Performance/encoding time functions were examined using a backward-masking sequential matching paradigm with a variable stimulus-mask-onset asynchrony. The processing of faces and other objects of expertise experienced a "head-start" over that of non-expert objects, requiring less encoding time to exceed chance-level performance. However, once initiated, performance increased at a similar rate. Therefore, the reliance of the expert VSTM advantage on encoding time was not a result of early perceptual limitations. Alternative explanations are discussed instead. In sum, experience can impact VSTM capacity and the time-course of perceptual processing. These findings suggest some plasticity in the VSTM system and highlight the potential of specialized expert perceptual mechanisms to impact higher-level cognitive functions that utilize their output.