Cognitive Authority Networks

Cognitive authority networks are the social structures that determine whose judgment counts as evidence in a given epistemic community. Unlike formal hierarchies — which are explicitly designated — cognitive authority emerges from the network of deference relationships: who cites whom, who reviews whom, whose methods are adopted and whose are ignored. The structure of these deference networks shapes what questions get asked, what methods get funded, and what results get believed.

The concept extends social epistemology by treating authority not as a property of individuals but as a property of network position. A scientist's authority is not merely a function of their track record; it is a function of their centrality in the citation network, their editorial positions, their mentorship relationships, and their institutional affiliations. Two scientists with identical publication records can have dramatically different authority depending on their network position — one may be a hub in a dense cluster, the other an isolated node with high individual output but low structural influence.

Cognitive authority networks exhibit the same topological properties as other complex networks: clustering within subfields, short paths between distant specialties (the small-world property), and heavy-tailed degree distributions where a small number of authorities receive disproportionate deference. These properties are not design features; they are emergent consequences of how scientists allocate attention under constraints of time and cognitive load.

The normative question is whether cognitive authority networks produce reliable collective beliefs. The answer depends on the coupling between authority structure and evidence quality. In a well-functioning network, authority is correlated with reliability: the most-cited scientists are also the most careful. In a captured network, authority decouples from reliability: centrality is achieved through political skill, institutional position, or rhetorical dominance rather than through epistemic virtue. The replication crisis can be read as a decoupling event — a moment when the authority network continued to function while its correlation with reliable methods broke down.

The connector's provocation: the problem with expertise is not that experts are wrong. It is that the network structures that make experts credible are invisible to the people who rely on them. Until we can see the authority network, we cannot judge whether the expertise it produces is worth trusting.