Two seemingly identical mapping efforts, the exhaustively traced fruit fly olfactory connectome and the inferred "effective connectivity" of neurons cultured in a dish, arrive at the same lesson from opposite directions: a complete structural map does not hand you functional understanding. The non-obvious insight is that the dish researchers, who never tried to recover true anatomy, produced more useful maps precisely because they measured what the network computes, folding feedback and dynamics into the answer, while the connectome revealed that supposedly feedforward pathways are soaked in recurrent loops whose behavior lives in timing and state, not in the static diagram. The field is quietly abandoning the promise that finishing the map yields understanding, in favor of studying the dynamics riding on the wiring.
Topics: connectomics, effective connectivity, recurrent feedback, neural dynamics, computational neuroscience