Unified Information Processes and Instructions in Sumerian Initial Structures Part I of the Voynich–Sumerian–Medieval Music Structural Reconstruction Series

Sumerian is described as deciphered at the level of lexical correspondence, yet the layer of procedure and operation recorded in the texts remains unexplained. Taking as its starting point the concentration of unresolved elements in the verbal prefixes, this paper applies a single state-transition model to three bodies of material — accounting, tuning, and sentence-initial structure — and tests whether it reproduces across them. A six-state model was extracted from Ur III balanced accounts. The assignment rules were fixed before the sources were examined, and were then applied without any modifi- cation to MVN 15, 108 (104 lines) and SET 130 (391 lines). Excluding damaged lines, the residual unassigned count was zero in both. State 3 ⃝ was never observed independently in MVN, but in SET 130 it appeared nine times on fixed lexical grounds alone, showing that the 3 ⃝/ 4 ⃝ collapse in MVN is an observation boundary specific to that source. On the musical side, separating the fourteen dichords of CBS 10996 into two columns while preserving the tablet order shows that the sign of the difference reverses at the point where the second string of the primary series wraps mod 7. The same wrap appears again under regrouping by shared string, and again in linear representation. Because no reordering was performed, this structure is contained in the arrangement of the source itself. The tuning text UET VII 74 records a bidirectional cycle directly. Each tuning state con- sists of six clear dichords and one that is not clear, and this remainder supplies the name of the next tuning. This is isomorphic to the carry-forward of residuals in the accounts (la₂-ia₃ → si-i₃-tum). For the sentence-initial prefixes, 3,133 finite verbs and 2,461 transitions were measured across the full ETCSRI corpus. Since i-, mu- and ba- co-occur within a single verb, they cannot be treated as one exclusive series and must be handled as multi-layered edge oper- ators. All three show significant within-document persistence. Comparing the retention of prefixes on identical stems between royal inscriptions and literary texts, no large-scale loss was observed. The cross-genre test on the administrative corpus, and the one-to-one correspondence be- tween the six states and the two directions, remain incomplete. Throughout, observation, support, prediction and non-verification are kept distinct.