Abstract
One of the unsolved debates in the prolonged exchanges between face specificity and perceptual expertise hypothesis of FFA (e.g., Gauthier et al., 2007; 2017; Kanwisher et al., 2017) has been on whether the BOLD activities in FFA for objects of expertise correlate with their behavioral expertise. While early studies found positive evidence (Gauthier et al., 2000; Xu, 2015), later replications by other labs did not (Grill-Spector et al., 2004; Rhodes et al., 2004; Moore et al., 2006). Until today, the reason behind these disparate expertise-FFA correlation results remain unsolved. In this study, we recruit bird experts, and look from the opposite side of the expertise-FFA correlation: whether there is an adaptation-like negative correlation between face selectivity (faces vs. objects) and bird expertise in FFA. 16 Taiwanese birders and 17 American birds were both evaluated with behavioral expertise index, and later scanned with various fMRI tasks, including FFA localizers (Taiwan only), passive viewing and 1-back identity tasks of 4 conditions : Asian faces, familiar birds, unfamiliar birds, and objects. While the localizer-defined rFFA showed insignificant FFA-expertise correlations (r14=.3, p=.3), consistent with much of the majority of previously mentioned literature, the face selectivity (faces vs. objects) shows significant negative correlation (r14 = −606, p=.022). In other words, the default face-responsive fusiform area, when sandwiched between blocks of objects of expertise, could become face un-responsive! Complementary whole-brain correlations with “birds vs. objects” and “faces vs. objects” contrasts (N=33) also showed adjacent but distinctive rFG regions for both positive and negative expertise-correlated regions, not only rapport to the ROI analysis results, but also explains that when focusing on the suboptimal bird stimuli in the default face-selective FFA, the brain-behavior correlation might be sub-optimal. Together, these results once again support the Flexible mid-Fusiform Area (FFA, Tarr and Gauthier, 2000) under various tasks.