We indeed perceive—and are aware of seeing—the face of a particular person rather than the combination of pixels and specific features that compose the person’s face. This process of extracting meaning involves categorizations and perceptual decisions (Beale and Keil, 1995, Freedman et al., 2001, Freedman et al., 2002, Fabre-Thorpe, 2003, Palmeri and Gauthier, 2004, Rotshtein et al., 2005 and Heekeren et al., 2008), where similar visual inputs, like the front view of two different faces, can lead to different percepts and, conversely, disparate images, like the front and profile view

of a person, give the same Everolimus solubility dmso percept. Converging evidence has demonstrated the involvement of the ventral visual pathway—going from primary visual cortex to inferotemporal cortex—in visual perception (Logothetis and Sheinberg, 1996, Tanaka, 1996 and Tsao and Livingstone, 2008). At the top of the hierarchy along the ventral visual pathway, high-level

visual areas have strong connections to the medial temporal lobe (MTL) (Saleem http://www.selleckchem.com/products/BKM-120.html and Tanaka, 1996, Suzuki, 1996 and Lavenex and Amaral, 2000), which has been consistently shown to be involved in semantic memory (Squire and Zola-Morgan, 1991, Nadel and Moscovitch, 1997 and Squire et al., 2004). It is precisely in this area where we previously reported the presence of “concept cells”—i.e., neurons with highly selective and invariant responses that represent the meaning of the stimulus. In fact, concept cells are selectively activated by different pictures

of a particular person, by the person’s written or spoken name, and even by internal recall, in the absence of any external stimulus (Quian Quiroga et al., 2005, Quian Quiroga et al., not 2008a, Quian Quiroga et al., 2009, Gelbard-Sagiv et al., 2008 and Quian Quiroga, 2012). In the quest to understand how the brain constructs meaning from sensory information, several works have studied the firing of single neurons in monkeys using identical but ambiguous stimuli that elicit different perceptual outcomes (for reviews, see Logothetis, 1998, Kanwisher, 2001 and Blake and Logothetis, 2002). One such experimental manipulation is the use of face adaptation, where the perception of an ambiguous face is biased by the presentation of another face shortly preceding it (Leopold et al., 2001, Leopold et al., 2005, Webster et al., 2004, Moradi et al., 2005, Jiang et al., 2006, Fox and Barton, 2007 and Webster and MacLeod, 2011). In this work, we used the unique opportunity of recording the activity of multiple single neurons in awake human subjects—who were implanted with intracranial electrodes for clinical reasons—to study how neurons in the MTL respond to face adaptation. In particular, starting from two pictures of persons known to the subject (for which we had a neuron firing to one of them but not to the other), we created ambiguous morphed images that were briefly flashed, immediately following the presentation of an adaptor image (one of the two pictures).