Independent Adjudication and Philosophical Convergence: A Prospective Research Design

Why do some philosophical disagreements persist while others diminish? The question bears on whether we can determine which philosophical disputes are resolvable and how. Beyond its foundational arguments concerning the two logical obstacles, NP2, The Two Logical Obstacles of Philosophy and the Response of Natural Science, proposes an empirical extension hypothesis pending determination: in a target class of philosophical disputes, the absence of commonly accepted independent adjudication of load-bearing factual premises is a principal structural contributor to persistent expert non-convergence. Prediction 1 holds that when such premises begin to be constrained by accessible, discriminating, independently checkable evidence, and the community accepts that the relevant methods can adjudicate the question—accepted methodological jurisdiction—longitudinal convergence should be greater. This paper designs a prospective study of that prediction to evaluate the empirical relationship between independent adjudication and convergence. The foundational arguments that valid deduction cannot itself supply the truth of empirical premises and that natural science supplies true or approximately true premises are assessed by their respective arguments and scientific evidence; they do not depend for their validity on this study’s convergence outcomes. An independent panel will assemble ongoing disputes. At baseline, coders will register whether independently documented research is expected to produce adjudicative evidence within the observation window and whether the community accepts its methodological jurisdiction. The primary comparison, within the high-jurisdiction-acceptance stratum, contrasts disputes prospectively identified as expecting evidence events with matched disputes lacking such events. Evidence already available will be analyzed separately; expected evidence that fails to arrive will be handled through preregistered feasibility rules rather than retrospective recoding. Expert-position dispersion will be compared between baseline and follow-up. Temporal ordering prevents future outcomes from defining the predictors, although baseline disagreement and unmeasured rival factors can still introduce bias. Motivated reasoning, institutional and social factors, diversity of problems, field size, and dispute duration explain the same outcome at the same level and are therefore modeled as rivals to the hypothesis. The preregistered jurisdiction interaction will be tested in a two-by-two framework only when all cells are adequately populated. Historical cases illustrate the logic; a randomized individual-updating experiment tests a narrower mechanism. One of two alternative follow-up routes must be selected before registration. When all conditions are met, precise adverse results count against the empirical prediction; imprecise findings and operational failures leave it untested. The contribution is to fix predictors before outcomes arise, treat the principal rivals as competing models of the same outcome, and specify in advance which findings support Prediction 1 and which count against it.