The Cohomology Language: How the Compiler Specifies Its Output

Papers 135–138 established the quintic compiler (the mechanism by which A₅ non-solvability converts algebraic input into topological output) and showed that the Riemann Hypothesis, BSD, the Tate conjecture, and Bloch–Kato occupy rows of a single compilation table. This paper identifies the language of the compiler. The BSD conjecture Part II specifies the exact leading term of the L-function of an elliptic curve as a product of five factors: period, regulator, Tamagawa numbers, the order of the Tate–Shafarevich group, and torsion. We show that the same five-factor format appears at every row of the compilation table — BSD (β₁), Tate (β₂), and Bloch–Kato (β₃) — with each factor playing the same structural role. This universal output format is the cohomology language: a grammar in which periods are literal declarations, regulators are variable bindings, Tamagawa numbers are exception handlers, the Tate–Shafarevich group is a type checker (whose finiteness is a consistency condition on the language), and torsion elements are null values. The Bloch–Kato conjecture (1990) is recognised as the language specification itself. The Langlands programme, in this framework, is syntax equivalence: the assertion that two different source representations compile to identical output. Proving the language subsumes BSD Part II, Tate, Bloch–Kato, and the Langlands programme as consequences.