, 2012). The representational-hierarchical theory emphasizes the importance of the organization of representations in a hierarchical continuum throughout the ventral visual processing LY294002 in vivo stream (Cowell et al., 2010b). Under this view, anterior regions such as the PRC contain complex conjunctive representations (e.g., object ABC), whereas more posterior regions contain representations of lower-level features (e.g., features A, B, and C) (Figure 1). At the beginning of the High Ambiguity condition in experiment 3, individuals with PRC damage may have successfully used a single-feature strategy, supported by intact regions posterior to their damage

(by definition, the objects in the discrimination of ABC versus ABD contained a single unambiguous feature: C versus D). However, as the condition progressed, more and more perceptually similar features were processed and represented in these posterior regions. Over time (after

approximately 36 trials), irrelevant single features from previous trials created interference, and the single-feature strategy became less successful. Whereas individual object features were very similar from trial-to-trial, Crizotinib mouse the objects themselves were trial unique and could be uniquely represented by an intact PRC. The cases with MTL damage including PRC, however, lacked these unique conjunctive PRC representations to disambiguate the single features, and thus, impairments emerged relative to controls and relative to individuals with a damaged hippocampus but an intact PRC. Intermixing perceptually dissimilar objects rather than perceptually similar objects in experiment 4 minimized the degree of interference. When the same number of stimuli were interspersed as in

experiment 3—but the stimuli were perceptually dissimilar rather than perceptually similar—the MTL cases were no longer impaired. However, once consecutive trials involving perceptually similar stimuli were introduced, the deficit Bay 11-7085 re-emerged. Thus, we propose that the present findings, and related ones in the animal literature, are best explained in terms of a representational deficit, rather than an impairment in a given psychological process, be it memory or perception. Impoverished representations will lead to deficits in all of these processes, and thus, a representational account may provide a more parsimonious explanation for the deficits observed on a wide range of tasks—both mnemonic and perceptual. Interestingly, although cases with MTL damage including PRC were impaired, cases with selective hippocampal lesions performed normally on the present tasks. This suggests that the effect of interference is dependent on which MTL region is damaged and the specific stimuli that are used. Thus, although vulnerability to object-based perceptual interference may explain visual memory impairments in some cases of MTL amnesia, it is not a general mechanism underlying visual memory impairments in all cases.