A third of the classic is a confessed graft

Claim (verbatim)

The Carakasaṃhitā confesses to being one-third replacement part. Its final redactor, Dṛḍhabala of Pañcanadapura, states in the text itself that he found the saṃhitā mutilated and reconstituted it, supplying seventeen chapters of the Cikitsāsthāna and the entire Kalpasthāna and Siddhisthāna — forty-one chapters of one hundred and twenty, grafted over a lost original third. The foundational text of Ayurveda thus carries an internal loss-horizon: the closing books of the Agniveśa-Caraka text had already perished in antiquity, and what stands in their place is a named later author's prosthesis. The tradition even disagrees about which seventeen Cikitsā chapters are his — the commentators transmit more than one enumeration — which converts the confession into a computable problem, because a replacement layer written by one hand in one period should carry a stylistic signature separable from the base text it patched. Prediction: profiling all one hundred and twenty chapters on function-word and particle frequencies and clustering them, the Cikitsāsthāna will split into two stylistic populations of approximately thirteen and seventeen chapters, the seventeen-chapter cluster will group with the Kalpa- and Siddhi-sthāna chapters rather than with the rest of the Cikitsāsthāna, and its membership will agree with one of the traditional enumerations of Dṛḍhabala's chapters at 80% or better (primary clause: the seventeen-with-Kalpa-and-Siddhi clustering; the verdict follows it). Exact computation: run the profile on the digital Carakasaṃhitā text with chapters as units and report the dendrogram. Kill: the GRETIL e-text of the Carakasaṃhitā (Trikamji's Nirṇaya Sāgara edition as base), with Dṛḍhabala's self-statements and the commentators' chapter-lists as set out in G. Jan Meulenbeld, A History of Indian Medical Literature, vol. IA.

Prediction clause (verbatim)

Prediction: profiling all one hundred and twenty chapters on function-word and particle frequencies and clustering them, the Cikitsāsthāna will split into two stylistic populations of approximately thirteen and seventeen chapters, the seventeen-chapter cluster will group with the Kalpa- and Siddhi-sthāna chapters rather than with the rest of the Cikitsāsthāna, and its membership will agree with one of the traditional enumerations of Dṛḍhabala's chapters at 80% or better (primary clause: the seventeen-with-Kalpa-and-Siddhi clustering; the verdict follows it). Exact computation: run the profile on the digital Carakasaṃhitā text with chapters as units and report the dendrogram.

Kill-dataset (verbatim)

Kill: the GRETIL e-text of the Carakasaṃhitā (Trikamji's Nirṇaya Sāgara edition as base), with Dṛḍhabala's self-statements and the commentators' chapter-lists as set out in G. Jan Meulenbeld, A History of Indian Medical Literature, vol. IA.

Novelty / leakage triage

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

Drdhabala's 41-chapter supplement and the commentators' divergent enumerations of his seventeen Cikitsa chapters are fully documented, and scholars assert the styles of the layers are distinguishable in principle; but no stylometric or function-word clustering of the 120 chapters testing the seventeen-with-Kalpa-and-Siddhi grouping was located - the layer question is explicitly still open research.

• G.J. Meulenbeld, A History of Indian Medical Literature, vol. IA (Groningen, 1999), on Drdhabala's contribution and the chapter-lists
• P.A. Maas, 'On What Became of the Carakasamhita after Drdhabala's Revision', eJournal of Indian Medicine 3 (2010)

Predictions

No prediction registered yet.