Abstract
Face recognition recruits both a different processing style and neural substrate compared to typical non-face object processing (Kanwisher, McDermott, & Chun, 1997; Tanaka & Farah, 1993). Does this influence the perceptual and/or visual short-term memory (VSTM) capacity for faces compared to other object categories? In Exp. 1, we used a probed recall match-to-sample task to estimate VSTM capacity for faces and two categories of non-face objects (cars and watches). Encoding time was manipulated to explore the influence of perceptual encoding limitations on VSTM capacity. A concurrent articulatory suppression task prevented verbal rehearsal. VSTM capacity generally increased with additional encoding time. VSTM capacity was smaller for faces than non-face objects at shorter encoding durations. However, at longer encoding durations VSTM capacity was equivalent for faces and objects. In Exp. 2, we compared the VSTM capacity for upright and inverted faces. Similar to Exp. 1, VSTM for upright faces benefited more from additional encoding time than that for inverted faces. However, VSTM capacity for inverted faces, unlike VSTM for non-face objects, did not reach the same level as that for upright faces. This suggests that experience can influence VSTM capacity. In Exp. 3, we explored the influence of perceptual expertise on VSTM and perceptual capacity. Preliminary results suggest that car experts demonstrate a similar pattern for cars as that found for faces: Car experts benefit from additional encoding time more than car novices. Car experts also appear to have a greater VSTM capacity for cars than do novices. Such results would suggest that object complexity and expertise both influence VSTM capacity. However, effects of expertise emerge with increasing encoding time, suggesting that the manner in which experts encode information may be more efficient and lead to more objects being stored in VSTM.
Supported by grants from NSF (BCS-0091752), NEI (EV13441-01) and by the James S. McDonnell Foundation.