Enough blanket metaphysics, time for data-driven heuristics

Bruineberg and colleagues criticisms’ have been received but downplayed in the FEP literature. We strengthen their points, arguing that the Friston blanket discovery, even if tractable, requires a full formal description of the system of interest at the outset. Hence, blanket metaphysics is futile, and we postulate that researchers should turn back to heuristic uses of Pearl blankets.