Reason: Proof of lem:finite-orthonormal-basis-operator-bounded-2026b. Carried over from the proof of the 2026a version with the comparison families written as tuples in R^n and references updated to def:orthonormal-basis-2026b, def:orthonormal-family-2026b and thm:orthonormal-expansion-parseval-2026b. The initial segment [n] is now introduced at the top of the proof, since the revised statement no longer introduces it and the argument quantifies over it. No step of the argument changed.
Proof
Throughout, [n] is the initial segment determined by n, and by the definition of a tuple an n-tuple in a set X is a map from [n] to X, so the finite sums below are formed from maps on [n] as required.
Two preliminaries. First, if a,b∈Rn satisfy ak≤bk for every k∈[n], then ∑k=1nak≤∑k=1nbk. Indeed 0≤bk−ak for every k, so claim 5 of Properties of Finite Sums gives 0≤∑k=1n(bk−ak), while claims 2 and 3 of that lemma give
adding ∑k=1nak to 0≤∑k=1nbk−∑k=1nak gives the assertion. Second, if x,y,c are real numbers with x≤y and 0≤c, then 0≤c(y−x)=cy−cx by the second order axiom of Ordered Field, so cx≤cy; we call this multiplying an inequality by a nonnegative number.
By Triangle Inequality for Finite Sums of Vectors and the absolute homogeneity of the norm, noting that the n-tuples in R with components ∥⟨ek,u⟩T(ek)∥ and ∣⟨ek,u⟩∣∥T(ek)∥ coincide and therefore have the same finite sum,
Multiplying by the nonnegative number ∥T(ek)∥ gives ∣⟨ek,u⟩∣∥T(ek)∥≤∥u∥∥T(ek)∥ for every k∈[n], so by the first preliminary and then claim 3 of Properties of Finite Sums together with the commutativity of multiplication,