Deciding new cases against retrieved precedents

Definition · Edward H. Levi, An Introduction to Legal Reasoning (1949)

This procedure decides a new case by comparison with already decided ones. The judge states which features of the earlier case are relevant and which are not, and that selection is generally where the reasoning happens.

Analysis

The comparison half of Levi's procedure is genuinely here and, in CARO, trained rather than approximated: the model generates an analogy chain from the new case to retrieved ones and is optimized on the chain, NEW-S10 puts decided cases in the context beside the written rules so the comparison runs in readable text, and SIROCCO does hand-built structural matching against 500 NSPE decisions. What is missing is the step Levi says the reasoning consists of — the statement of which features of the earlier case are relevant and which are not. CARO's own divergence records that the selection stays inside the weights; NEW-S10's chains invoke cases without naming what was discarded; SIROCCO's relevance is a computed similarity, not a stated ground. First half carried, load-bearing half not.

Definition · Arthur L. Goodhart (1930)

This distinction separates the part of a past decision that binds later cases, the holding, from the remarks made along the way, the dicta, which do not. Without a way to tell the two apart, a body of precedent cannot indicate which of its own decisions are actually binding.

Analysis

Two of the three papers give the model whole cases as text with an outcome attached, so nothing marks what in a past decision governs and what was said in passing. SIROCCO is the exception that makes this thin rather than empty: its case base indexes each NSPE decision by the code provisions the Board applied to the facts, which locates a decision's operative ground rather than treating the opinion as an undifferentiated blob. But locating the applied provision is not Goodhart's test — nothing separates a provision necessary to the result from one merely canvassed, no material-fact statement is extracted, and since no output of any of these systems constrains a later one, the distinction has nothing to do work on.

Definition · Common-law practice

These two moves let a later court revise the record: distinguishing confines an earlier precedent to its own facts, and overruling sets it aside altogether. Both are done openly and with reasons stated, which is what keeps a body of case law open to revision.

Analysis

In all three systems the case base is a fixed input — supplied at training and inference for NEW-S10, bootstrapped once and then optimized over for CARO, hand-encoded from Board opinions for SIROCCO — and no decision any of them reaches is written back into it. SIROCCO's divergence puts it exactly: it reads a body of precedent it can never join. Because the record never grows through use, there is no later precedent of the system's own to confine to its facts or to set aside, and neither paper offers a mechanism for stating reasons that would narrow or retire a case already in the base.

Definition · Albert R. Jonsen & Stephen Toulmin, The Abuse of Casuistry (1988)

This method anchors judgment in paradigm cases, i.e., clear cases on which agreement is easy, and in the short maxims drawn from them. A hard case is then settled by asking how far it sits from the paradigms, rather than by deduction from a general principle.

Analysis

The casuist's direction of travel is real across all three: NEW-S10's headline result is that demonstrating safety through decided cases beats spelling the code out at length, which degrades helpfulness; CARO reasons from clear retrieved cases outward to the ambiguous one; and SIROCCO gives abstract code provisions content only through the cases in which they were applied. What is absent is the structure that makes casuistry work — no case is designated a paradigm, nothing orders the base from central to marginal, and a hard input therefore sits near whatever the retriever happened to return rather than at a measured distance from an agreed clear case. The maxim is missing too: the trained systems draw nothing short and portable out of their cases, and SIROCCO's principles arrive ready-made from an external professional code rather than being distilled from the paradigms.

The grid's two locus categories are treated as alternatives; this approach supplies both at once and lets the model reason across them. That is how real legal systems work and how neither column of this taxonomy describes: the codified text and the decided cases are inputs to one act of reasoning rather than competing sources of authority.

Levi describes analogy as something a judge does and can be held to. Making it a training objective means the model is optimized toward analogical form, which produces the form reliably and the accountability not at all: a chain of analogy that reads correctly is rewarded whether or not the feature selection behind it would survive scrutiny.