Toward a pluri-component, multimodal, and dynamic organization of the ventral semantic stream in humans: lessons from stimulation mapping in awake patients

International audienceFor many decades, neural basis underlying cognitive functions was conceived in a localizationist framework. Owing to the development of connectomics, an alternative hodotopical account was proposed, in which brain functions are sub-served by the interactions of large-scale distributed and parallel subnetworks (Catani, 2007; de Benedictis and Duffau, 2011). In this setting, the processing of visual information is divided in a dorsal stream dedicated to the analysis of the spatial position (“where”) and in a ventral stream specialized in object identification (“what”; Underleider and Haxby, 1994). By analogy, a dual-stream model for auditory language processing was suggested, with a dorsal stream involved in mapping sound to articulation and a ventral stream involved in mapping sound to meaning (Hickok and Poeppel, 2004). Nonetheless, the neural structures supporting the ventral route is still controversial. We have recently proposed a new model of language, in which the subcortical anatomical constraints have been incorporated (Duffau et al., 2013): beyond a well-recognized dorsal phonological/articulatory stream underlain by the superior longitudinal fascicle, the neuroanatomy subserving a parallel ventral stream involved in multimodal semantics was described. Here, our purpose is to detail the dynamic functional anatomy of this multi-component ventral route, constituted by direct and indirect pathways (explaining a possible compensation following brain injury) and implied in pluri-modal semantic processes—i.e., in verbal and non-verbal comprehension, control and noetic consciousness