Networks for proper versus common name retrieval: insights from causal mapping and new hypotheses.

Over the years, the anatomical correlates of common name and proper name retrieval have been studied and discussed side by side in the field of cognitive neuroscience. Due to the particular semantic status of proper names (i.e. meaningless marker referring to unique entities with a token link), it is generally acknowledged that their lexical retrieval is cognitively more demanding than that of common names. Thus, the neural networks engaged for accessing the phonological output lexicon of common names and proper names could be partially segregated, centred on different cortical regions and underpinned by distinct structural pathways. However, this possible anatomo-functional organization has not yet been explored in an in-depth review. Hence, the objective of our review is 3-fold: (i) to highlight the cortical networks underlying common name and proper name retrieval; (ii) to summarize the findings from recent studies using causal methods of tract-oriented functional mapping (i.e. direct electrostimulation and disconnection-symptom mapping) with the aim of isolating ventral structural connections that contribute to common name and/or proper name retrieval; (iii) to develop hypotheses regarding the anatomical connectivity implicated in lexical access to common names and proper names, in relation to the input modality and semantic category (for proper names, in the latter case). Our critical review first suggests that, although both categories appear to rely on a widely distributed and overlapping left fronto-temporo-parietal cortical network, the mid-to-posterior part of the ventral temporal cortex (especially the fusiform gyrus) and the anterior temporal structures (including the temporal pole), seem more specifically involved in common name retrieval and in proper name retrieval respectively, regardless of their input modality. Second, the left inferior longitudinal fasciculus appears to play a crucial role in picture-based lexical retrieval for both common names and proper names. Based on these key findings, we propose tentative models of anatomical connections underlying lexical retrieval. We suggest that distinct components of the inferior longitudinal fasciculus are implicated in lexical retrieval for common names and proper names and that another similar segregation mediates the retrieval of face names and landmark names. We further hypothesize that the left middle longitudinal fasciculus plays a role in auditory-guided lexical retrieval, owing to its projections into the anterior temporal structures and its close relationship with the auditory associative cortex. Within the framework of the dual-language model, we discuss the heuristic value of these anatomical hypotheses to understand the mechanisms by which anomia arises in various neurological conditions, providing insights and directions for future studies.