Research · Preprint
Comparisons between modern computational systems and older technical practices are everywhere, and mostly unreliable. They either read a modern result back into an old source and call the resemblance descent, or refuse the comparison altogether and call the two incommensurable. This paper sets out a third path: how to make one responsible comparison, with explicit guardrails.
The paper proposes eight elements for making a cross-temporal comparison responsibly: separate the questions of attestation, interpretation, reconstruction, comparison and transmission; grade every claim in the sentence that makes it; keep the layers of evidence visible; apply source-tier discipline; write an explicit analogy contract; run an anti-overclaim checklist; publish the reconstruction step reproducibly; and read living traditions in their own setting.
None of the eight elements is new on its own; each has precedent in the history and philosophy of science. The contribution is their selection, ordering and packaging for a specific, recurring use case: claims that draw a line from an old technical practice to a modern computational one.
The protocol is demonstrated on one case: an independent reconstruction of the competing readings of Mānava 11.14–11.15 and the value 25/8. Two coherent reconstructions produce near-mirror area errors of about 0.53 percent, and small changes of assumption move the implied value from 3.06 to 3.19. Numerical closeness to π cannot, on its own, decide the textual dispute, and that is the point. The protocol exists because a striking number is not evidence of what the source text meant.
The book The Rope and the Algorithm: Evidence, Procedure, and Trustworthy AI (Mohan Das, 2026, Mindful Machines Press) applies this protocol at length. Starting from the Śulba Sūtras, the Vedic construction texts this paper's worked example is drawn from, it asks what makes a procedure trustworthy: a stated purpose, known conditions, visible checks, a defined sense of error, and an accountable route for correction. This paper sets out the method and the guardrails; the book applies them to the harder question of what a modern reader is licensed to say about the relationship between an old technical tradition and a contemporary computational one.
Every source behind the book's claims is logged and open to inspection in the book's source log, which separates evidence, interpretation and comparison in the same way this paper's protocol requires.
Read this before citing. This paper is a preprint and has not been through peer review. Treat the argument as provisional.
Mohan Das, Viveka, Provenance, Reconstruction, and the Limits of Analogy: A Protocol for Cross-Temporal Comparison of Technical Practice (September 09, 2026). Available at SSRN: https://ssrn.com/abstract=7437420 or http://dx.doi.org/10.2139/ssrn.7437420
The author identifier for citation and disambiguation is ORCID 0009-0007-9715-0951. The correct family name form is Mohan Das, two words.