The folded fingers of Gargas

Claim (verbatim)

Many hand stencils at Gargas in the Pyrenees, and a set at the now-drowned Cosquer Cave near Marseille, show fingers apparently shortened or missing above a joint. Three families of explanation compete: ritual or punitive amputation, cold- or disease-caused loss (frostbite, Raynaud's, Dupuytren's), and deliberate finger-folding - the hand held in a bent posture, perhaps a gesture or a signing code, while the pigment was blown. The competition is decidable, not by asking what the makers meant, but by combinatorics, because a folded living hand can only produce a specific, enumerable subset of 'missing-segment' configurations, whereas amputation or pathology can leave patterns that no fold can imitate (an intact fingertip above a missing middle segment, say). If folding is the mechanism, the observed configurations must lie almost entirely inside the flexion-accessible set, and their frequencies should track which folds are easy rather than being uniform over anatomically possible losses. The missing knowledge is the meaning of the gesture; but the mechanism behind the 'mutilated' hands is testable from the catalogue of which fingers, and which segments, are absent - a distribution-geometry problem, not a leap of empathy.

Prediction clause (verbatim)

Prediction: coding each incomplete Gargas and Cosquer stencil by exactly which finger segments are absent, at least 90 percent of distinct observed configurations will be reproducible by voluntary flexion of an intact hand (the flexion-accessible set), and the frequency ranking of configurations will fit an ease-of-folding model significantly better (by a chi-square or likelihood ratio) than a uniform-over-possible-losses amputation model (primary clause: the >=90 percent flexion-accessible fraction; the model comparison is corroborating). Disambiguation: a configuration counts as flexion-accessible only if a living hand can hold it with the remaining fingers extended for stencilling. Coverage guard: partial or smudged stencils where segment boundaries are unreadable are excluded from the denominator, not scored as anomalies.

Kill-dataset (verbatim)

Kill: the Gargas hand-stencil corpus (Barriere's and Leroi-Gourhan's catalogues; the Grotte de Gargas publications) and the Cosquer inventory (Clottes & Courtin), coded by missing-segment configuration and compared against the enumerable flexion-accessible set - the operation is the fraction of observed configurations inside that set and the fit of an ease-of-folding versus uniform-amputation frequency model. See also McCauley, Maxwell & Collard's cross-cultural analysis of missing-phalange hand images.

Novelty / leakage triage

anticipated in the literature — this exact test has never been run

The three-way debate (folding, amputation, pathology) and the folding hypothesis's combinatorial core are published - Leroi-Gourhan argued the Gargas hands are bent-finger gestures, and McCauley, Maxwell & Collard (2018, Journal of Paleolithic Archaeology) surveyed 121 amputation-practising societies to argue instead for deliberate finger-segment removal. But neither line runs the conjecture's decidable operation: coding each Gargas/Cosquer stencil by exact missing-segment configuration, measuring the fraction inside the enumerable flexion-accessible set (>=90%), and fitting an ease-of-folding versus uniform-over-possible-losses model. McCauley et al. argue interpretation ethnographically, not by that distribution-geometry test; the combinatorial audit was not located.

• A. Leroi-Gourhan, 'Les mains de Gargas: essai pour une etude d'ensemble,' Bulletin de la Societe Prehistorique Francaise 64 (1967)
• B. McCauley, D. Maxwell & M. Collard, 'A Cross-cultural Perspective on Upper Palaeolithic Hand Images with Missing Phalanges,' Journal of Paleolithic Archaeology 1 (2018), doi:10.1007/s41982-018-0016-8
• J. Clottes & J. Courtin, The Cave Beneath the Sea: Paleolithic Images at Cosquer (New York: Abrams, 1996)

Predictions

No prediction registered yet.