Purchase this article with an account.
Andrew persichetti, Joseph Denning, Jiongjiong Yang, Stephen Gotts, Alex Martin; Modality and category selectivity in the anterior temporal lobes. Journal of Vision 2020;20(11):371. doi: https://doi.org/10.1167/jov.20.11.371.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The anterior temporal lobe (ATL) is a large and anatomically heterogeneous region of cortex with connections to memory, language, and perceptual systems. Given the connectivity profile of the ATL, different theories claim that it serves as a general convergence zone for all semantic information, while other models posit more selective roles in object recognition, language processing, and social cognition. To investigate the functional roles of the ATL, we used a rapid event-related fMRI design to measure the responses across the ATL while 32 adults were asked to either name or categorize visual and auditory stimuli from two categories: common animals (e.g., dog, cat) and tools (e.g., hammer, saw). Each stimulus was represented in four ways: as a grayscale picture, a printed name, a spoken name, and a stimulus-associated sound. We analyzed the patterns of responses in functionally distinct cortical parcels within the ATL that were obtained using resting-state fMRI in a separate large cohort of adults (N=88). Within the most anterior portions of the ATL (i.e., the temporal pole), we found a bias toward visual stimuli in the medial and inferior temporal, and fusiform gyri. By contrast, we found a clear auditory bias in the superior temporal gyrus and sulcus. Within the visually selective portions of the temporal poles, we found a lateral bias towards animal pictures and a ventromedial bias towards tool pictures that is similar to the organization of more posterior category-selective cortex in the ventral visual stream. These results suggest that the temporal pole, and the ATL more generally, can be functionally dissociated into category- and modality-selective regions. Therefore, the ATL does not seem to be a domain-general semantic hub. Rather, it seems to comprise functionally distinct regions that are part of domain-specific cognitive systems.
This PDF is available to Subscribers Only