How can you trust experts without surrendering your judgment?

The thesis that non-experts can rationally defer to specialists through "second-order evaluation"—assessing track records, institutional credentials, and consensus rather than raw technical content—rests on a complex framework of unstated premises. Grounded in social epistemology (notably developed by theorists such as Alvin Goldman), this approach attempts to rescue individual agency from the impossible demand of universal domain mastery. However, excavating what this framework must take for granted reveals its profound dependencies. ## Conceptual Assumption: The Separability of First- and Second-Order Evidence * **The Assumption:** It is conceptually possible to cleanly separate the evaluation of an expert's *credentials and environment* (second-order) from the evaluation of their *arguments* (first-order). * **Contestability:** Moderate to high. Critics within the philosophy of science note that recognizing a valid consensus or assessing whether a peer-review process is robust often requires a baseline of substantive domain knowledge. Without it, the novice risks committing "epistemic trespassing," misinterpreting what constitutes a genuine consensus versus a manufactured one. * **If Challenged:** If first- and second-order criteria bleed into one another, the core promise of the framework collapses: the non-expert cannot evaluate the proxy without already possessing elements of the expertise they are trying to bypass. ## Factual Assumption: The Truth-Conduciveness of Institutional Proxies * **The Assumption:** Standard institutional markers—such as peer-review status, academic affiliation, and professional consensus—reliably correlate with truth-conduciveness rather than systemic groupthink, regulatory capture, or institutional self-preservation. * **Contestability:** High. Empirical investigations into publication biases, replication crises across the biomedical and social sciences, and industry-funded research demonstrate that institutional incentives frequently diverge from truth-seeking. * **If Challenged:** If institutional proxies track compliance, prestige, or safety rather than objective validity, then relying on them is not rational calibration, but rather systemic risk-absorption. Deference ceases to be truth-tracking and becomes compliance with institutional power. ## Value Assumption: Procedural Agency Constitutes Genuine Autonomy * **The Assumption:** Retaining intellectual autonomy does not require substantive comprehension of a claim; rather, choosing *whom* to trust based on procedural transparency and disclosed conflicts of interest satisfies the moral and cognitive requirements of agency. * **Contestability:** Moderate. Philosophers in the Kantian tradition might argue that true intellectual autonomy requires an agent to grasp the reasons for a belief internally. Deferring based on administrative disclosures might be practically necessary, but calling it "judgment" could be dismissed as a semantic sleight-of-hand. * **If Challenged:** If procedural checking is disqualified as true agency, then non-expert deference must be categorized as a form of intellectual surrender, invalidating the text's central attempt to reconcile trust with autonomy. ## Contextual Assumption: A Stable and Pluralistic Epistemic Ecosystem * **The Assumption:** The surrounding sociopolitical environment features a plurality of independent verification channels, transparent conflict-of-interest disclosures, and a baseline of public trust in expert institutions. * **Contestability:** Very high in contemporary public life, where institutional credibility is deeply polarized and expertise is frequently weaponized in culture-war conflicts. * **If Challenged:** In an environment of structural epistemic collapse—where verification channels are viewed as partisan actors—the secondary markers proposed by Goldman's framework (such as tracking institutional scrutiny) become circular or contested themselves. The novice can no longer look to an independent referee because the referees are perceived as contestants.