The Philosopher Disagreeing with Oneself: On Feldman’s View about Reasonable Religious Disagreement

Abstract
Richard Feldman’s “Reasonable Religious Disagreements” (2011) deploys social epistemology to reject the possibility of reasonable religious peer disagreement. By eliminating all potential cases—differing conclusions from shared evidence (via Uniqueness Thesis), differing foundational beliefs, private evidence, and self-assessed superiority—Feldman concludes that awareness of such disagreement requires suspending religious belief, defending limited modest skepticism. This paper critically examines Feldman’s eliminative argument. It identifies two core problems: first, an inconsistency with Feldman’s later positions (2014, 2022) on second-order evidence, where no general principle mandates belief revision and steadfastness can be rational, allowing reasonable religious disagreement; second, the argument generalizes to common knowledge propositions (e.g., external world realism), entailing overarching skepticism rather than limited skepticism, as Feldman claims. These tensions stem from Feldman’s shifting concepts of rationality—from a strict evidentialist one in 2011 to a permissive view incorporating other epistemic values later. The analysis reveals unnoticed self-disagreement in Feldman’s oeuvre and its implications for religious epistemology. 
Keywords: Religious Epistemology, Epistemology of Disagreement, Religious Skepticism, Rationality, Evidentialism, Richard Feldman
Introduction 
The epistemology of disagreement, a key area of social epistemology, examines how epistemic agents should respond when they discover that equally qualified peers hold conflicting beliefs about a proposition. Positions on the spectrum range from conciliatory views (e.g., Elga, Christensen) that require significant belief revision under peer conditions, to steadfast views (e.g., van Inwagen, Wedgwood) that permit retention of one’s original doxastic attitude. Religious disagreement occupies a prominent place in these debates, long predating technical disagreement literature. While some philosophers interpret widespread religious disagreement as evidence against justified religious belief, others accommodate or even leverage it in support of religious rationality. Richard Feldman’s 2011 paper “Reasonable Religious Disagreements” offers one of the most precise arguments in this domain, employing evidentialist and uniqueness principles to reject reasonable religious peer disagreement outright and defend modest skepticism. This extended abstract articulates Feldman’s eliminative strategy and subjects it to critical scrutiny, revealing internal tensions with his later work and unintended broader skeptical implications.
Modest Skepticism: Feldman’s Eliminative Argument 
Feldman restricts his analysis to original, peer-level religious disagreements in which parties take each other’s views seriously, have shared all relevant evidence, and are epistemic peers in intelligence, reasoning ability, and background knowledge. He asks whether it is possible for both parties to remain rational in their conflicting beliefs after learning of the disagreement. Systematically enumerating four candidate cases for reasonable disagreement, Feldman dismisses each.
First, when parties draw different conclusions from identical evidence, the Uniqueness Thesis entails that a single body of evidence supports at most one doxastic attitude; hence at least one party is irrational. Second, when foundational beliefs or weighting principles differ, Feldman insists these starting points must themselves be rationally grounded; once shared and discussed, the Uniqueness Thesis again forces convergence. Third, private evidence (e.g., incommunicable religious experiences) is neutralized by the principle “evidence of evidence is evidence”: awareness of the other’s private evidence becomes second-order evidence that must be treated symmetrically. Fourth, cases in which parties regard the disagreement as unreasonable (each rationally deeming the other incompetent) collapse once competence judgments, which must also be evidence-based, are shared.
Having eliminated all possibilities, Feldman concludes that modest skepticism is the only rational stance: each party knows that one of them is mistaken but has no reason to suppose it is the other; suspension of judgment on the disputed proposition is therefore required. The argument is entirely general and non-religious-specific in its premises.
Inconsistencies with Later Work and Overarching Skepticism
Feldman’s later writings on disagreement—especially “Evidence of Evidence is Evidence” (2014) and “Is There Something Special about Religious Disagreement?” (2022)—introduce a markedly more permissive account of rationality that directly contradicts the 2011 prescription. He now denies both additivity of evidence and any general rule linking changes in total evidence to rational belief revision. A gap opens between justification (what the total evidence supports) and rational doxastic change (how one ought to update). Non-evidential intellectual values (e.g., the value of inquiry, attractiveness of a view for future research) can render steadfastness rational even when new second-order evidence is acquired. Religious disagreement, moreover, is declared to have “no special difference” from other disagreements. Consequently, no general principle mandates suspension upon peer disagreement, and steadfast retention of belief can be rational—directly opposing the 2011 conclusion that suspension is obligatory.
When the same 2011 eliminative argument is applied beyond religion to common-knowledge propositions (e.g., existence of the external world, other minds, ordinary objects such as clouds or hills), the result is overarching rather than modest skepticism. Anti-realist philosophers who deny these entities on metaphysical or epistemological grounds qualify as epistemic peers once evidence is shared; the four-case elimination again forces suspension. Feldman’s 2011 claim of “limited” skepticism is therefore untenable. His later permissive rationality—allowing non-evidential values to justify steadfastness—avoids this global result, but Feldman never addresses the resultant self-inconsistency. The root lies in two incompatible concepts of rationality: the narrow, strictly evidentialist notion operative in 2011 versus the broader, pluralistic notion embraced later.
Conclusion
Zahra Zargar’s analysis demonstrates that Feldman’s 2011 defense of modest religious skepticism is internally unstable and over-generalizes. The eliminative argument presupposes a rigid evidentialist rationality that Feldman himself later relaxes, permitting reasonable religious disagreement and avoiding global skepticism when applied to common knowledge. These unnoticed shifts reveal a philosopher disagreeing with himself. While the later view is more nuanced and inquiry-friendly, it leaves unresolved how exactly non-evidential values interact with evidentialism and the Uniqueness Thesis. Future work in religious epistemology must either reconcile these conceptions or choose between them explicitly. The case underscores the importance of consistency across an author’s corpus when deploying disagreement arguments to support or undermine religious rationality.