Judge reasoning excerpts — 2026-07-11 measurements

Sections below the baseline blocks were appended for the post-#505 exemption-sharpening runs (same day, same protocol); see measurement-2026-07-11-505.json.

Verbatim excerpts of the judge agents' prose reasoning, captured alongside the boolean outcomes in measurement-2026-07-11.json. Not every agent produced prose (several returned bare JSON); this file commits everything that was captured that bears on a miss or a notable pass, so the mechanism reading in the audit report is auditable. Judge: claude-sonnet-5, isolated per batch, given only the advisory section variant + 6 items.

Off-list title-shell misses — the exemption clause read broadly

• ti-004 (post-#503 rep2, MISS): "'Beyond Awareness: Cybersecurity Training and Employee Security Behavior' — names specific mechanism (training → behavior) and domain (cybersecurity); doesn't reduce to a generic 'X and Y' shell despite superficial brevity. No match."
• ti-002 (post-#503 rep2, MISS): "'The Weight of Care: Nurse Workload and the Quality of Patient Care' — names a specific mechanism/population (nurse workload → patient care quality); domain-native vocabulary, not a swappable shell."
• ti-012 (post-#503 rep2, MISS): "names a specific mechanism/domain (vocabulary acquisition, L2 learners); doesn't survive noun-swap into an arbitrary field meaninglessly. No row match."
• ti-013 (pre-#503, MISS): "'Aging Apart: Social Isolation Among Older Adults' is a title, not a frame pattern from the list — no match." (Pre-variant reasoning: judged against the table only, as instructed.)

The same shape caught when the judge reads it as a bare pair

• ti-004 / ti-002 (pre-#503, FIRE): "bare 'X and Y' title shell, matches WP06 (bare study-of frame)." (Same items, opposite call, adjacent runs — the decorated-title shape sits exactly on the exemption boundary.)
• ti-007 (pre-#503, FIRE, cited "off-list shell"); (post-#503 rep1/rep2, FIRE, cited WP06).

Family-variant behavior

• nat-065 (pre-#503, FIRE): "'How does X affect Y' (SES → access to higher education) matches WP01 impact/effect frame." — interrogative rewording of a listed family, caught even pre-#503.
• el-003 (post-#503 rep1, FIRE as off-list): "'What shapes X among Y' … matches the generic factors/determinants shell closely enough (off-list but clearly a shell: 'what shapes X among Y' swaps trivially)."
• el-013 (post-#503 rep2, MISS): "'How does material choice bear on the durability of concrete structures' names a specific mechanism (material choice → durability) and site (concrete structures) — specific, not generic; doesn't survive swap." (Stable miss in all runs; same broad-exemption reading.)

Hard negatives — the exemption working as intended

• dn-006 (pre-#503, correct SILENT): "highly specific mediator construct, named scale, named population — domain-native operationalized wording, not a bare shell match despite containing 'mediate the effect.'"
• dn-009 (pre-#503, correct SILENT): "precise mediation model with named instruments (IUS-12, GAD-7) — domain-native, high-confidence no match."
• nat-077 (pre-#503, MISS): "close to WP16 … but not the canonical 'barriers and facilitators to X' pair shell … weak/ambiguous match, not high-confidence." (Post-#503: FIRE citing WP16 in both replicates.)

Post-#505 runs — the decorated-title shape argued through the new rule

• ti-002 (post-#505 round1 rep1, FIRE): "colon-decorated title; post-colon part is 'Nurse Workload and the Quality of Patient Care,' a generic swappable noun-pair (WP01/WP06 shell). Decoration doesn't exempt it." (Baseline rep2 argued the same item as "names a specific mechanism/population" and missed.)
• ti-004 (post-#505 round1 rep2, FIRE): "'Cybersecurity Training and Employee Security Behavior' — bare study-of/factors-style noun pair, no named instrument or mechanism specified. Swappable shell."
• ti-013 (post-#505 round1 rep2, FIRE): "'Aging Apart' decorated title; subtitle 'Social Isolation Among Older Adults' is a bare topic label with no named site, instrument, or mechanism." (Rep1 missed the same item as "subtitle is named-population topic label … not swappable-shell form" — the single-topic subtitle sits nearest the boundary.)
• el-013 (post-#505 round1 both reps, FIRE): "'How does X bear on Y' … no named instrument/theory/site, generic material/durability pairing is swappable." (Stable miss in every baseline run under the broad-exemption reading.)

Post-#505 runs — negatives still exempted, and via the narrow list

• dn-007 (post-#505 round1 rep1, correct SILENT): "names mechanism (unobtrusive indicators in annual reports), specific causal pathway (through TMT power centralization) — satisfies specified-pathway exemption."
• dn-016 (post-#505 round1 rep2, correct SILENT): "names specific mechanism (telicity constraints, lexical aspect) and a specified causal pathway (L1 transfer vs. aspect), plus named population." — the stated-tension / pathway arms of the narrow exemption doing the work.
• nat-044 (post-#505 round1 rep2, MISS): "plain descriptive question without a shell connective (impact/relationship/role/factors), so weak match, stay silent." — the remaining interrogative off-list miss rides the high-confidence bar, not the exemption clause.

Post-#505 FINAL-wording runs — single-topic subtitles caught; one prose-verdict inversion

• ti-013 (FINAL rep2, FIRE): "'Social Isolation Among Older Adults' is a generic topic label (age group is generic demographic, not named population); decorative prefix ignored." — the round-2 demographic-descriptor clause doing exactly the work the round-1 miss motivated.
• ti-012 (FINAL rep2, FIRE): "'Words in the Making' is decorative pre-colon phrase, ignored per rule" — single-topic subtitle judged on its own.
• ti-002/ti-004/ti-007 (FINAL rep1, MISS — anomaly): the batch judge's prose called each "swappable shell / generic 'X and Y' noun-pair, no named instrument" yet its JSON verdicts said SILENT — a prose-verdict inversion. The boolean record stands per protocol; the same items fired in FINAL rep2 and both round-1 replicates.