We are poor at distinguishing knowledge that is in our heads from knowledge that resides in the community (KC); we overestimate how much we know or understand merely by participating in a KC

Individual Representation in a Community of Knowledge. Nathaniel Rabb, Philip M. Fernbach, Steven A. Sloman. Volume 23, Issue 10, October 2019, Pages 891-902. https://doi.org/10.1016/j.tics.2019.07.011

Highlights
.  The knowledge that supports many of our beliefs and attitudes resides not in our own heads, but in a community of knowledge constituted by other people, artifacts, and information repositories (e.g., libraries or the Internet).
.  Individuals are poor at distinguishing knowledge that is in their heads from knowledge that resides in the community. This leads them to overestimate how much they know or understand merely by participating in a community of knowledge.
.  This failure of differentiation has implications for public discourse. Extreme views about science and politics have been found to covary with knowledge overestimation.
.  Modeling individual cognition under collective knowledge is an emerging challenge for cognitive science.

Abstract: An individual’s knowledge is collective in at least two senses: it often comes from other people’s testimony, and its deployment in reasoning and action requires accuracy underwritten by other people’s knowledge. What must one know to participate in a collective knowledge system? Here, we marshal evidence that individuals retain detailed causal information for a few domains and coarse causal models embedding markers indicating that these details are available elsewhere (others’ heads or the physical world) for most domains. This framework yields further questions about metacognition, source credibility, and individual computation that are theoretically and practically important. Belief polarization depends on the web of epistemic dependence and is greatest for those who know the least, plausibly due to extreme conflation of others’ knowledge with one’s own.

Keywords: collective cognitionknowledge representation