Podcast with Bjorn Merker on brain systems and brain architecture

How many systems does the mammalian brain actually have, and what is each one really doing? Neuroscientist Bjorn Merker challenges conventional anatomical boundaries and proposes that the brain's major subdivisions, neocortex, cerebellum, basal ganglia, and brainstem, each perform a distinct generic function, running in parallel all the time rather than switching on and off. Subscribe for more from the Convergent Science Network podcast series. Bjorn Merker joins Paul Verschure and Tony Prescott for a wide-ranging tutorial on brain systems architecture. He begins by questioning how we define a system at all, showing that textbook divisions like midbrain and diencephalon dissolve under embryological and molecular scrutiny. Instead, he argues that genuine systems should exhibit redundant internal structure reflecting a generic function , as the crystalline circuitry of the cerebellum or the uniform laminar organization of neocortex clearly do. From this principle, he derives a functional decomposition: neocortex performs veridical source reconstruction across all sensory afferents, solving the inverse problems that plague perception; basal ganglia handle action selection and policy; cerebellum contributes decorrelation and calibration. The discussion challenges the standard view that higher brain systems replace lower ones. Merker advocates a Jacksonian layered control model where every level runs its generic computation continuously in parallel, with higher levels adding new capacities rather than suppressing old ones. He illustrates this with eye-blink conditioning, where anticipatory and reflexive responses coexist, and with the evolutionary persistence of the superior colliculus alongside cortical vision. The conversation also explores why the brain's massive learning structures, cortex, cerebellum, basal ganglia, scale together in evolution, and why hippocampus sits at the apex of cortical hierarchy as a hinge converting feedforward into feedback processing. Key topics include the bowtie architecture of cortical connectivity, why volumetric scaling predicts learning capacity, the developmental sensitivity versus adult robustness of brain systems, and how frontal-limbic-hippocampal circuits form the densely interconnected hub of the mammalian brain. Part of the Convergent Science Network podcast series from the BCBT Summer School.