Between roughly the sixth and tenth centuries CE, a community of Jewish scholars in Tiberias and Babylon — the Masoretes — produced the authoritative pointed text of the Hebrew Bible. They did not merely copy it. They counted it: the middle letter of the Torah, the middle word of each book, the total verses in a section, tallies recorded in the margins as a standing audit. Any scribe anywhere, checking his copy against those tallies, could detect a single dropped or added letter arithmetically, without having to find the error by eye.
The Canon chapter names this outright: the Masoretic letter-count and the checksum on your downloaded file are the same operation. Not metaphorically — structurally. Both are content-hashes: a fixed fingerprint of the text derived from the text itself, stored independently, so that drift screams rather than silently accumulates. The purpose in both cases is to make a read word stay read — to prevent the slow ouroboros by which a text bends half a degree toward each holder’s wish until it has become a mirror.
The Vedic pāṭha — permuted-order recitation, forward and reversed and braided — is the oral equivalent: the same text checked from several directions at once, any slip clashing against the others and caught on the spot. Two communities, one architecture: freeze the word so the consulting party cannot quietly become its author.
Sources. The Masoretes of Tiberias (~6th–10th c. CE) and their marginal letter/word tallies (the Masorah — e.g. the counted middle letter of the Torah); the Vedic pāṭha permuted recitation (jaṭā-pāṭha, ghana-pāṭha) as the oral equivalent. Search: Masoretes Masorah letter count middle letter Torah; Vedic patha recitation jata ghana.
Appears in: The Canon — the word they would not let drift