​​​​​Talk from  Simone Vigano, Max Planck Institute for human cognitive and brain sciences

Short abstract:​​

A fascinating hypothesis in cognitive neuroscience is that of a phylogenetic continuity between brain circuits for spatial navigation and human declarative knowledge, where “concepts” stored in memory are conceived as points of internal “cognitive maps” within the hippocampal formation. This idea, rooted in ancient philosophy and recently re-popularized by cognitive neuroimaging based on findings in rodent electrophysiology, has become a foundational framework for studying human memory and inspiring models in artificial intelligence. In my talk, I will argue that this framework may need to be revised to account for the complexity of human cognition as it emerges from the specific functional and biological integration of our mind, brain, and body. First, its current focus, centered mostly on the hippocampal-entorhinal system, might underestimate the distributed and interactive nature of knowledge representation across the human brain: hippocampal maps are only one node in a larger network that might represent knowledge across different reference frames. I will show recent fMRI evidence supporting this view. Second, it focuses heavily on how knowledge is structurally represented, but not on how we search for goal information in it, nor what is the underlying physiology of this process - a crucial gap which we are trying to fill in an ongoing stereo-EEG experiment. Third, it overlooks the ecological specificity of primates’ interaction with the environment: unlike rodents, we rely heavily on vision and eye movements to sample information from our surroundings and guide action. In line with this fundamental difference, I will summarize recent eye-tracking findings indicating that spontaneous gaze behavior during mental search tasks are quite revealing about internal organization of memory spaces. Finally, the framework seems to implicitly assume, or at least suggest, that the brain’s internal geometry is Euclidean, perhaps simply as a consequence of a literal interpretation of the term “map”. Based on recent studies, I will argue this might be incorrect, and I will briefly introduce an ongoing project and a future research line aimed at unveiling the actual neural geometry behind knowledge representation and, arguably, human cognition​.​

​