AI-generated conjecture · below the evidence/publication boundary
← All conjectures · Engineering, materials & building
Buttress gradient descent
Status is derived only from the shepherd-authored triage/prediction data above -- community submissions and claims are a separate overlay and can never change it (see the participation panel below).
Claim (verbatim)
Buttress gradient descent. Gothic churches were built in successive campaigns over decades, and masons could watch how earlier bays and earlier buildings cracked, leaned, or stood. Modern limit analysis can compute, for any buttress geometry, how far it sits from the thrust-line optimum — the configuration that contains the vault's forces with the least material and the greatest safety margin. The conjecture is that Gothic construction was literally an optimization algorithm: successive campaigns should move monotonically toward the thrust-line optimum, each generation's buttresses measurably closer than the last, like gradient descent on a structural loss function. Dated buttress geometries evaluated under limit analysis should show that loss decreasing step by step — iterative structural learning, executed in stone across generations of builders.
Prediction clause (verbatim)
For each geometry in the dated buttress geometries under limit analysis, compute a structural loss: the deviation of the built form from the thrust-line optimum (e.g. geometric safety-factor shortfall under limit analysis); order campaigns chronologically within each building sequence. Primary clause: the loss decreases monotonically across successive campaigns, with a significant negative trend (Spearman ρ ≤ −0.5, p < 0.05) and no more than one non-monotonic step per sequence. The verdict follows the primary clause.
Kill-dataset (verbatim)
dated buttress geometries under limit analysis.
Nobody has run this test. The kill-data is named above. If you can run it — or you know the paper that already settles it — claim the kill or submit the prior. Kills and priors are credited here, by name, as they come in.
In the atlas
This conjecture is bridged, as an L1 lead, onto these Inferpedia subject pages.
Provenance
Run: Imported conversation (verbatim harvest) · model: claude-fable-5
Origin: operator conversation with Claude Fable 5 at max effort, conducted 2026-07-03, relayed verbatim by the operator into the shepherd session on 2026-07-04. No ModelRun exists for the original generation (it happened outside the pipeline); this transcript file is the canonical capture. Transcript path: docs/generated/conjecture_harvest_fablemax_20260703.md. Model (operator-attested, not pipeline-recorded): claude-fable-5. Novelty disclaimer (verbatim, load-bearing -- rule 4): "Same caveat as before, doubled: at 100 items across all of archaeology and history, some of these will have cousins in the literature I can't check. What I can guarantee is the format — each links two things not normally linked, and each names the dataset or measurement that would kill it."
Novelty / leakage triage
anticipated in the literature — this exact test has never been run
Limit analysis has been applied parametrically to a DATED series of early Gothic flying buttresses (Nikolinakou & Tallon: geometry variables and failure modes across the series), and learn-from-failure narratives across campaigns are standard; framing the trajectory as a monotonic quantified loss-decrease toward the thrust-line optimum was not located.
- Nikolinakou & Tallon, 'Structure and form of early Gothic flying buttresses' — Limit analysis across a dated flyer series
Predictions
No prediction registered yet.
Weigh in
No community feedback yet.
Add your take
Posted immediately (spam is removed). Community feedback is never an adjudicated verdict and never changes this conjecture's triage label or status above.
Working on this?
Sign in to claim this conjecture and let others know you're working on it.