← Research

Beyond a defensible answer.

What makes an interpretation admissible—and why a plausible answer is not enough.

A simplified scope distinction from “The shifted map.”

An answer can be defensible under one interpretation and inadmissible under another. The relevant object is therefore not an isolated proposition, but the relation between a proposition, its evidential context, and the interpretive commitments under which that relation is evaluated.

Assessment design sits inside this distinction. A passage does not merely contain facts; it distributes commitments across scope, modality, reference, and qualification. A response may preserve the vocabulary of those commitments while altering their logical force. Conversely, a response may depart substantially from the wording while preserving the relation that matters.

Neither lexical proximity nor explanatory fluency is sufficient to resolve the distinction. The evaluation problem is relational: whether an interpretation is licensed by the relevant evidence without importing an unacknowledged premise, relaxing a material restriction, or substituting a weaker conclusion for the requested one.

𝒜=⋂i∈I𝒞i
Satisfying several conditions jointly is different from averaging their individual assessments. This notation describes a general logical relation, not an internal acceptance rule.
Correctness is a judgment about an answer.
Admissibility is a judgment about the conditions of that judgment.
Interpretation-relative closure

Closure is meaningful only relative to a specified inferential regime. Treating an evidential context as closed under arbitrary background knowledge can enlarge the defensible conclusion set beyond the intended task; treating it as closed under literal restatement can exclude legitimate abstraction. The resulting boundary is neither purely syntactic nor reducible to a single scalar similarity measure.

This creates a distinction between the extension of a response under a selected reading and its stability across the readings the task reasonably permits. Apparent agreement at the extension level can coexist with disagreement about the admissible interpretive domain.

This perspective describes a research problem and its interpretive commitments. It is not a claim of calibrated difficulty, demonstrated learning gains, or deployment of each formalism discussed.