The notion of decoupling, suggested by Cosmides and Tooby as a way to expla
in how audiences are protected against the misleadingly false aspects of na
rratives, is itself crucially misleading. The behavioral and linguistic evi
dence suggests, rather, that the processes by which we protect ourselves fr
om misinformation a) don't work reliably enough to suggest that they are au
tomatic, and b) all seem to be culturally dependent in ways that evolved ad
aptations could not be. The evolutionary argument itself suggests that it w
ould actually be maladaptive if the making of such distinctions were automa
tic and decontextualized.