In-group-trust patterns prove trust is relational, not based on the trusted’s traits.

Evidence

That some people trust you and others do not means, on this account from characteristics of the trusted, that some are mistaken in their assessment of your dispositions and commitments. Oddly, in their own experimental results, the Yamagishis (1994) find that Japanese subjects tend to trust one another within cliques but not across cliques. How might we explain this as merely the result of mistaken assessments of others’ characteristics? It seems trivially easy to explain it as relational, as grounded not merely in characteristics of the trusted but in relationships with them. — Russell Hardin, Trust and Trustworthiness, 390

Categorization

• XREF: Directly supports the argument that trust requires the observable relational constraints Hardin (and the reader’s integration with Luhmann) is building toward: cliques create observability and shared interests.
• QUESTION: Worth chasing: does the Japanese clique-trust finding connect to Luhmann’s idea that familiarity is the precondition for trust within bounded social systems?