SUMMARY: Legibility asymmetry gives institutions detailed visibility into participants while participants receive limited visibility into institutional state, reasoning, recipients, and correction. One side becomes continuously measurable; the other appears through notices, interfaces, and selective disclosures.
Digital administration asks people to remain legible across identity checks, behavioral signals, transaction histories, device attributes, location, documents, and linked accounts. The institution’s own decision path is usually far less available to inspection.
Observation and Explanation Separate
A participant may be scored from hundreds of attributes but receive one reason code. The institution can reconstruct patterns across time while the person sees only the latest outcome. More data does not produce reciprocal understanding.
Interfaces Curate Institutional Visibility
Dashboards expose the states an organization is prepared to acknowledge. They may omit queue position, unresolved conflicts, recipient disagreement, model uncertainty, manual review, and exception use. The interface becomes a policy boundary around what counts as knowable.
The auditability gap is the evidentiary expression of this imbalance: internal reconstruction remains possible while participant verification is constrained.
Visibility Changes Negotiating Power
A person who cannot see the disputed input, rule version, or downstream recipient must appeal broadly. The institution can demand specificity while withholding the details required to provide it. Delay and uncertainty become bargaining advantages.
Transparency Can Still Be One-Way
Publishing general policies, model cards, annual reports, or aggregate error rates may improve public knowledge without helping an individual verify a live case. System-level disclosure and case-level auditability solve different problems.
Appeal asymmetry shows how institutions control time and evidence during challenge. Legibility asymmetry determines who knows enough to begin that challenge effectively.
Reciprocity Requires Bounded Visibility
Expose current state, material inputs, decision authority, rule or model version, recipient delivery, uncertainty, exceptions, and correction progress. Protect other people’s data and genuine security-sensitive detail, but document every withholding category and provide independent review.
Test reciprocity through digital checkpoint audits. If the system can demand evidence from a participant, it should produce evidence for the consequential claims it makes in return.
\nMeasure reciprocity directly: compare the fields collected from a participant with the fields disclosed about the institution’s decision. Track unresolved unknowns, withheld categories, explanation delays, and whether access to evidence changes appeal outcomes.
\n\nIntel assessment: legibility becomes domination when one side must remain observable while the other can govern through abstraction.