Von Neumann's Closed-Operator Theorem¶
A closed densely defined Hilbert-space operator has a densely defined positive self-adjoint adjoint-product on its correctly restricted composition domain.
Core Idea¶
If \(T:D(T)\subseteq G\to H\) is a closed densely defined operator between Hilbert spaces, von Neumann's \(T^*T\) theorem says that \(T^*T\) is densely defined, positive and self-adjoint on \(G\). The composition domain is \(\{x\in D(T):Tx\in D(T^*)\}\), not automatically all of \(D(T)\). This makes the positive square root \(|T|=(T^*T)^{1/2}\) available through spectral calculus.[^ref-b0f8efebaf27]
Scope of Application¶
The theorem applies to bounded and unbounded closed densely defined operators. The constructed diagonal map \(T(a_n)=(na_n)\) on its weighted \(\ell^2\) domain gives \(T^*T(a_n)=(n^2a_n)\) on the stricter \(\sum n^4|a_n|^2<\infty\) domain. On \(L^2(0,1)\), the constructed weak derivative \(T:H^1_0(0,1)\to L^2(0,1)\) has \(T^*=-d/dx\) on \(H^1(0,1)\), hence \(T^*T=-d^2/dx^2\) on \(H^2\cap H^1_0\), the Dirichlet realization. The differential symbol alone would not choose this boundary domain.[ref-b0f8efebaf27][ref-715ba40d2ea1]
Clarity¶
For \(x\) in the composite domain, \(\langle T^*Tx,x\rangle=\|Tx\|^2\ge0\). This shows positivity there, not by itself equality of operator and adjoint domains. Self-adjointness is the theorem's nontrivial conclusion. A formal \(T^*T\) symbol with no domain or proof of closedness and density cannot invoke it.[^ref-b0f8efebaf27]
Manages Complexity¶
Once closedness and density of \(T\) are established, the theorem packages a difficult domain-sensitive self-adjointness proof into a reusable result. It then supports the modulus and polar decomposition, while leaving concrete boundary and physical interpretations to be checked separately.[^ref-b0f8efebaf27]
Abstract Reasoning¶
Declare \(G,H,D(T)\); prove \(D(T)\) dense and the graph closed; characterize \(T^*\) and form the correct composition domain. Only then conclude that \(T^*T\) is positive self-adjoint and apply spectral calculus. Do not infer that \(T\) itself is self-adjoint or that a product is automatically a physical observable.[^ref-b0f8efebaf27]
Knowledge Transfer¶
The same closed/dense/domain test transfers across sequence operators and PDE operators. The concrete adjoint domain does not transfer: it depends on the chosen \(T\). “Von Neumann's theorem” is therefore disambiguated here as the \(T^*T\) result, not the bicommutant or Stone–von Neumann theorem.[^ref-6fd96b5f3434]
[^ref-b0f8efebaf27]: Christian Budde, Operator Algebras and Unbounded Self-Adjoint Operators, master's thesis, Radboud University Nijmegen (2015), §4.1, Theorem 4.1; later exposition. [^ref-6fd96b5f3434]: “On the adjoint of Hilbert space operators” (2018). [^ref-715ba40d2ea1]: Grenoble notes on unbounded-operator domains.
Relationships to Other Abstractions¶
Current abstraction Von Neumann's Closed-Operator Theorem Domain-specific
Parents (1) — more general patterns this builds on
-
Von Neumann's Closed-Operator Theorem is a kind of Formal theorem Domain-specific
The closed-operator adjoint-product result is a proved formal theorem.
Hierarchy paths (2) — routes to 2 parentless roots
- Von Neumann's Closed-Operator Theorem → Formal theorem → Formal System → Formalization → Representation → Abstraction
- Von Neumann's Closed-Operator Theorem → Formal theorem → Formal System → Formalization → Transformation → Function (Mapping)
Neighborhood in Abstraction Space¶
Von Neumann's Closed-Operator Theorem sits in a sparse region of the domain-specific corpus (93rd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Functional Analysis & Operator Theory (16 abstractions)
Nearest neighbors
- Closed Linear Operator — 0.82
- Affiliated operator — 0.79
- Nilpotent Operator — 0.78
- Closed Immersion — 0.78
- Limiting Absorption Principle — 0.78
Computed from structural-signature embeddings · 2026-10-08