Communication is central to our species’ success, from facilitating collective action to supercharging cumulative cultural evolution. Yet across human evolutionary history and to the present day, communication carries with it the threat of misinformation and manipulation. For communication to remain adaptive, theorists propose that the mind contains a suite of cognitive mechanisms for evaluating speakers and their messages. This dissertation presents 7 experiments (N = 1,681) investigating a key function hypothesized of these “epistemic vigilance” adaptations: The selective linkage of messages that violate prior beliefs with “meta-data” specifying their source or the context of their acquisition. Remembering such links permits the ongoing evaluation of communication and preserves the integrity of existing knowledge.

In the experiments reported here, participants read a series of stories associated with different sources and each containing counterintuitive (which violate prior beliefs) and ordinary concepts. As predicted, participants in Exp. 1 more accurately attributed counterintuitive versus ordinary concepts to their speakers. Exp. 2a-b replicated this finding and found that this attribution advantage extended to places and dates associated with counterintuitive concepts. Exp. 3 then investigated the relative strength of these links over time, finding that links between counterintuitive concepts and speakers were differentially durable compared to those with places. Exp. 4 explored the mechanisms that might link epistemically suspect messages to their source. A memory advantage was again found for links between counterintuitive concepts and persons, but only when the messages were framed as told by others (“incoming”) and not when told to others (“outgoing”). Exp. 5a-b attempted to replicate this pattern but found an advantage for matching counterintuitive versus ordinary concepts with their associated speakers or recipients, along with an overall advantage for matching incoming versus outgoing messages.

Together, these results demonstrate that the mind selectively tags misleading messages with meta-data that facilitates the ongoing evaluation of their source and content. The results also outline the memory mechanisms – including metarepresentation and elaborative processing – involved in forming links between the source and content of communication. Finally, implications for communication and the representation and social diffusion of counterintuitive concepts broadly, using the case study of those found in pseudoscience, are discussed.