Abstract

In parallel with other species, humans possess a remarkable ability to encode detailed spatial information about our environments, forming cognitive maps that enable inference and generalisation for goal-directed behaviour. Long linked to the hippocampal-entorhinal system, growing evidence now suggests that the neural mechanisms supporting spatial navigation also extend to abstract domains, involving a broader network of cortical regions. In this talk, I will propose that the default mode network (DMN), traditionally associated with mind-wandering and self-referential processing, plays a domain-general role in constructing and traversing cognitive maps: structured representations of relational knowledge that span both spatial and non-spatial domains. I will present recent findings from our lab using ultra-high-field 7T fMRI, showing how spatial learning and memory are encoded across the DMN during navigation in virtual environments, and how these same regions organize conceptual knowledge along interpretable representational axes to support abstract mental navigation. Together, these results suggest that the DMN implements a unified computational architecture for mapping space, memory, meaning, and value. This framework bridges classical theories of cognitive maps with contemporary systems neuroscience and offers translational insights into disorders such as Alzheimer’s disease, where both spatial navigation and DMN function are compromised.

Host: Prof Trevor Robbins (twr2@cam.ac.uk)

This talk will be recorded and uploaded to the Zangwill Club Youtube channel in due course.