The claim
The manuscript’s §5, as CLAIM-050 quoted it: for translation, Blackwell’s order is restricted to a family Q of tasks, and the manuscript “proposes deficiency for the approximate case (Torgersen, LIT-778)”. Shannon (THEORY-602) supplies rate-distortion, and “decoder-only audience knowledge makes it a Wyner-Ziv problem (LIT-767)”.
Exact garbling is rare between two renderings, so a usable order needs a measure of how far one is from being a garbling of the other. Le Cam’s deficiency is that measure. The audience brings knowledge the translator cannot condition on in the text, which is the decoder-side information of Wyner and Ziv.
Why it is Active
The first step’s link is held: in the finite setting of CLAIM-381, a deficiency of ε is exactly ε-approximation in every [0, 1]-loss problem (CLAIM-503), and ε = 0 is Blackwell’s order (THEORY-717). For a family Q of tasks the deficiency restricted to Q is CLAIM-381’s finite computation, bounded by the unrestricted one; only the unrestricted one has the kernel form. The second step is the definition of a Wyner–Ziv problem, under the condition its defeated_if names. Torgersen and Wyner and Ziv stay cited, unread.
These steps were split from CLAIM-050 when it was narrowed on 2026-10-28 to what its read grounds carry, so that the narrowing drops nothing silently.