Original text: CC0 1.0. Prerequisite proofs and component terms.

Closing a domain after spectral localization

OA-FLOW.GRAPH.CONTRACT — Functional calculus used here

Let PP be self-adjoint on a Hilbert space HH, of any Hilbert dimension, with spectral measure EE. The spectral theorem gives the strongly continuous unitary group Ut=eitPU_t=e^{itP} and the closed normal operator B=f(P)B=f(P) for any Borel f:R→Cf:\mathbb R\to\mathbb C finite EE-almost everywhere. Its domain is

Dom⁡(B)={ξ∈H:∫R∣f(p)∣2 d⟨E(p)ξ,ξ⟩<∞}.\operatorname{Dom}(B)=\left\{\xi\in H:\int_{\mathbb R}|f(p)|^2\,d\langle E(p)\xi,\xi\rangle<\infty\right\}.

The spectral projections commute with BB on its domain and with UtU_t; BUtξ=UtBξBU_t\xi=U_tB\xi there. If a closed subspace reduces every UtU_t, it reduces every E(S)E(S) for Borel SS. This last assertion is the spectral-theorem consequence that the spectral projections of a self-adjoint generator belong to the von Neumann algebra generated by its unitary group. These are the exact functional-calculus inputs to the proof.

OA-FLOW.GRAPH.MAIN — The graph-localization theorem

Theorem. In the preceding setting, let DD be a linear subspace of Dom⁡(f(P))\operatorname{Dom}(f(P)). Assume that DD is dense in HH and UtD⊂DU_tD\subset D for every real tt. Then DD is a core for f(P)f(P): the closure of the graph of f(P)∣Df(P)|_D is the full graph of f(P)f(P).

There is no requirement that D⊂Dom⁡(P)D\subset\operatorname{Dom}(P), nor that ff be continuous, polynomially bounded or real-valued. Finiteness almost everywhere refers to the spectral measure, not Lebesgue measure.

Proof. Write B=f(P)B=f(P) and let

K={(d,Bd):d∈D}‾ H⊕H.K=\overline{\{(d,Bd):d\in D\}}^{\,H\oplus H}.

Closedness of BB gives K⊂Graph⁡(B)K\subset\operatorname{Graph}(B). Since BUt=UtBBU_t=U_tB on Dom⁡(B)\operatorname{Dom}(B), the hypothesis on DD implies

(Ut⊕Ut)(d,Bd)=(Utd,BUtd)∈Graph⁡(B∣D). (U_t\oplus U_t)(d,Bd)=(U_td,BU_td)\in\operatorname{Graph}(B|_D).

The same holds with −t-t, so KK reduces every Ut⊕UtU_t\oplus U_t. The generator on H⊕HH\oplus H is P⊕PP\oplus P. Its spectral projection for a Borel set SS is E(S)⊕E(S)E(S)\oplus E(S). The functional-calculus contract therefore gives

(E(S)⊕E(S))K⊂K.(G1) (E(S)\oplus E(S))K\subset K. \tag{G1}

For n≥1n\geq1, set

Sn={p:∣p∣≤n, ∣f(p)∣≤n},Qn=E(Sn).S_n=\{p:|p|\leq n,\ |f(p)|\leq n\},\qquad Q_n=E(S_n).

The projections increase strongly to 11, since ff is finite spectral-almost everywhere. Moreover BQnBQ_n is bounded with norm at most nn. Fix ξ∈H\xi\in H and choose dj∈Dd_j\in D with dj→ξd_j\to\xi in HH. Formula (G1) shows that each

(Qndj,BQndj)=(Qndj,QnBdj) (Q_nd_j,BQ_nd_j)=(Q_nd_j,Q_nBd_j)

belongs to KK. For fixed nn, boundedness of QnQ_n and BQnBQ_n implies convergence in H⊕HH\oplus H to (Qnξ,BQnξ)(Q_n\xi,BQ_n\xi). Therefore this pair belongs to KK.

Now take ξ∈Dom⁡(B)\xi\in\operatorname{Dom}(B). Strong convergence gives Qnξ→ξQ_n\xi\to\xi and, because BQnξ=QnBξBQ_n\xi=Q_nB\xi, also BQnξ→BξBQ_n\xi\to B\xi. Hence (ξ,Bξ)∈K(\xi,B\xi)\in K. Thus Graph⁡(B)⊂K\operatorname{Graph}(B)\subset K, giving equality. □\square

The proof deliberately places QndQ_nd in the closed graph KK. The hypothesis does not say that DD itself is invariant under discontinuous spectral cutoffs, and the proof never needs that stronger assertion.

OA-FLOW.GRAPH.POWERS — Imaginary-power invariance controls real powers

Corollary. Let AA be a positive, nonsingular self-adjoint operator on HH, and let r∈Rr\in\mathbb R. If D⊂Dom⁡(Ar)D\subset\operatorname{Dom}(A^r) is dense in HH and AitD⊂DA^{it}D\subset D for all t∈Rt\in\mathbb R, then DD is a core for ArA^r.

Proof. Nonsingularity means ker⁡A=0\ker A=0; it does not mean A−1A^{-1} is bounded. The spectral calculus defines the densely defined self-adjoint operator P=log⁡AP=\log A and gives Ait=eitPA^{it}=e^{itP} and Ar=f(P)A^r=f(P) with f(p)=erpf(p)=e^{rp}. Apply the theorem. When r=0r=0, Ar=1A^r=1 and the result is the original norm density; when r<0r<0, the same proof handles the possible singular behavior of the inverse near zero. No change of domain convention is made. □\square

Application to a dual Hilbert algebra. Suppose a proposed left Hilbert algebra has dense underlying vector domain DD, an antilinear operator S0S_0 on DD, and operators J,ΔJ,\Delta with JJ antiunitary, Δ\Delta positive nonsingular self-adjoint, and

S0d=JΔ1/2d(d∈D),D⊂Dom⁡(Δ1/2),ΔitD⊂D. S_0d=J\Delta^{1/2}d\quad(d\in D),\qquad D\subset\operatorname{Dom}(\Delta^{1/2}),\quad\Delta^{it}D\subset D.

The corollary with r=1/2r=1/2 implies that DD is a graph core for Δ1/2\Delta^{1/2}. Applying the isometry 1⊕J1\oplus J to the graphs then gives

S0‾=JΔ1/2.\overline{S_0}=J\Delta^{1/2}.

The isometry is real-linear on H⊕HH\oplus H if JJ is antilinear; this suffices to preserve closure and convergence of graphs. It also proves closability directly. This closes the analytic bridge once all the displayed hypotheses have been proved for the dual Hilbert algebra. It does not establish its density, invariance or the displayed formula; those remain concrete obligations in the general dual-weight construction.

Editable source · Proof dependencies and component terms