Taxonomy / Divergent precedent / Building and using a case record

Building and using a case record

Decisions get made one case at a time and kept as a record that later decisions draw on. Two genuinely different disciplines live under that substrate: writing each outcome as narrowly as possible, so parties who disagree about the general principle can still sign the same specific holding (the legal idea of an incompletely theorized agreement), and arguing a new case out of the record by citing the closest past cases, distinguishing the ones cited against you, and manufacturing hypothetical variants to find where a position breaks. These pull in opposite directions on one point, and the regime should show that rather than hide it: the first deliberately leaves the principle unsettled, the second is a search for the principle's boundary, and a system can do either without the other. A benchmark here asks whether language models can tell when one Supreme Court decision overrules an earlier one -- a capability any accumulating case record presupposes, though neither of the two disciplines above actually supplies it.

The method, against Divergent precedent

Scroll the diagram sideways to see all of it.

Concept Analysis: Theoretical Foundations

Each concept is read twice: whether the approach carries it, and whether the approach's own sources claim it. A concept that is absent and was never claimed is a gap in the field rather than a failure of the work, and is marked out of scope.

Recorded dissent

PartialClaimed · partial

Definition · Appellate practice

When judges disagree, the losing view is not erased — it is published alongside the majority's, signed and reasoned. That does two things at once: it forces the majority to answer a stated objection, which is what makes the majority's own reasoning legible, and it keeps the rejected theory on the shelf, available to become tomorrow's majority.

Analysis

The closest mechanism is HYPO's three-ply argument: a point is met with a full counter-argument that distinguishes the cited precedent and counter-cites another, so the opposing position is stated in reasoned form rather than suppressed, and both lines become legible. Everything the word 'recorded' carries is missing — HYPO never rules, so there is no majority for a minority view to stand against, and the counter-argument is manufactured for the occasion rather than entered into the case base where a later argument could pick it up. AI-A4's bracketing preserves only that a disagreement exists, not a signed and reasoned account of the losing side, and the overruling benchmark works on case pairs while never asking a model to read the dissents those opinions actually carry. No source asserts published dissent, but the pluralism pitch — parties who disagree still acting together, with the disagreement surviving the decision — pays off only if the rejected view stays on the shelf.

Overruling with provenance

AbsentNot claimed · out of scope

Definition · Common-law practice

The law can reverse itself when a precedent has become unworkable, unjust or simply overtaken by events — but only in the open: the overruling decision names what it overturns and why, so the change has a traceable history instead of being quietly written out of the record.

Analysis

Nothing in this subcategory reverses a holding it has itself made. HYPO's case base is a static index, and its characteristic move against an unwelcome precedent is distinguishing, which confines that case to its facts and deliberately leaves it standing; AI-A4 supplies no procedure by which a settled agreement becomes unsettled. The benchmark's 236 pairs are genuine reversals, but they belong to the Supreme Court's record rather than to any artifact here, and what it retains is the bare relation while the stated reasons — the very thing that makes a change traceable — are dropped. Grading whether a model can spot that one case overruled another is a claim to measure the capability, not to perform an overruling with provenance.

Incompletely theorized agreement

PartialNot claimed

Definition · Cass R. Sunstein, Legal Reasoning and Political Conflict (1996)

Agreeing on what to do in the case at hand while explicitly leaving the deeper principle unresolved. The agreement can be on an outcome, a rule or a mid-level principle; what gets bracketed is the comprehensive worldview underneath. Courts do this constantly: it lets people who disagree about theory still act together, one case at a time.

Analysis

The bracketing half is genuinely carried: HYPO reasons entirely in dimensions and factor comparisons drawn from prior cases and never ascends to a general theory, and AI-A4 makes leaving the deep principle unresolved its explicit design aim. The agreement half — the part that carries the weight in Sunstein, where disagreeing parties commit to a narrow outcome and act on it — is nowhere built: HYPO issues no holding at all, generating arguments for both sides and stopping, and AI-A4 offers no enforcer, institutional or computational, that holds anyone to the bracket. The result is preserved disagreement without the action it is supposed to enable, and the third paper runs the opposite way, fixing one ground-truth label per pair. AI-A4 names the concept, but as a goal for a system it does not build, so no source claims an artifact that realizes it.

Papers

Argued for, not built

These make the case for this approach, or sketch a design for it, but leave nothing built and tested behind.