Abstract
The classic debate over the face specificity of prosopagnosia was recently revived by a meta-analysis that found 42-80% of developmental prosopagnosics (DPs) to have mild to major object recognition deficits (Geskin and Behrmann, 2017). However, nearly all studies used real-world objects (e.g., cars) where recognition depends on object-specific expertise, raising the question of whether impairment in a single object category is representative of general object abilities. Recently, recognition tests using novel objects have shown to correlate highly with general object abilities. No studies to our knowledge have tested DPs using novel object memory tests (NOMTs), and the majority of studies have tested fewer than 20 DPs. In the current study, we tested 30 DPs and 30 age- and gender-matched controls (TD) on a NOMT (“Ziggerins”) and the Cambridge Face Memory Test (CFMT). DPs were impaired on the CFMT (DP: 38.3, TD: 59.4, p<0.01) but performed normally on the NOMT (DP: 58.2, TD: 57.5, p=0.75). We also compared DP NOMT performance to a matched sample of 275 web-based controls and found no difference in reaction time (DP: 4.2, TD: 4.1, p=0.58) and only a trend in accuracy difference (DP: 58.2, TD: 61.1, p=0.06). Additionally, 12 out of the 30 DPs met the stringent criteria for a classical dissociation between face and object recognition. These results are consistent with our fMRI findings in a subset of 23 DPs and 23 controls, which showed reduced face-selectivity in DPs across all face areas, but no difference for the object-selectivity between the two groups. Taken together, these results suggest that DPs are essentially normal at novel object recognition. Future investigations would be useful to determine whether some DPs have specific impairments with acquiring longer-term object expertise.