Closed Linear Operator¶

A partially defined linear operator whose graph is a closed subset of the product of its domain's ambient space and codomain.

Version: v1 · 2026-09-28 · History
Domain-specific #: 8501
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomains: Functional Analysis, Operator Theory → Mathematics

Core Idea¶

A closed linear operator is a partial linear map T:D(T)⊆X→Y whose graph is closed in X×Y, so convergent input–output pairs cannot escape the declared operator. Closedness concerns paired limits: if domain vectors converge and their images converge, the limit pair must still lie on the graph. This does not make a proper-domain unbounded operator continuous on the ambient space. Closability is weaker—the graph closure must be an operator graph—and the closure is then a closed extension. Because the domain controls adjoints, spectra, and boundary behavior, a formula without its domain does not specify the operator.

How would you explain it like I'm…

No faithful explanation at this level. All three generators judged eli5 unreachable: the only child-level picture is a well-behaved machine where close inputs give close outputs, which is continuity or boundedness, exactly what closed operators need not have; the real condition is about limits of input-output pairs landing back in the domain.

The No-Gaps-at-Limits Machine

Think of a machine that takes in certain allowed inputs and gives outputs, following a straight-line (linear) rule. Suppose you feed it a list of allowed inputs that settle down toward some point, and the outputs also settle down toward some answer. The machine is closed if, whenever that happens, the point it was settling toward is also an allowed input, and the machine gives exactly that answer for it. Which inputs are allowed is part of what the machine is. This does not mean close inputs always give close outputs.

Closed-Graph Linear Operator

A closed linear operator is a linear map T from part of a space X (its domain) into a space Y, such that its graph, the set of pairs (x, Tx), is closed in X×Y. In terms of sequences: if x_n are in the domain, x_n converges to x, and Tx_n converges to y, then x must be in the domain and Tx must equal y. The domain is part of the operator's identity; the same formula on a different domain can be closed or not. Closed does not mean continuous: a continuous operator keeps outputs close whenever inputs are close, while a closed one only requires that limits which exist are consistent. On Banach spaces, a closed operator defined on the whole space is automatically bounded, so the important unbounded closed operators, like many derivatives, are defined only on part of the space, often a dense part.

A closed linear operator is a linear partial map T: D(T) ⊆ X → Y between normed (typically Banach) spaces whose graph {(x, Tx) : x ∈ D(T)} is closed in X × Y. Equivalently, whenever x_n ∈ D(T), x_n → x, and Tx_n → y, it follows that x ∈ D(T) and Tx = y. Note the hypothesis requires both the inputs and their images to converge; closedness does not assert that convergent inputs have convergent images, which would be continuity. The domain is constitutive: restricting or extending it changes the operator and can change whether it is closed. By the closed graph theorem, a closed operator defined on all of a Banach space is bounded. Hence important unbounded operators are closed only on proper, often dense, domains. Closedness is the weaker regularity that survives in the unbounded setting and makes such operators tractable.

Scope of Application¶

The concept governs unbounded-operator analysis when spaces, topology, action, and domain are all explicit. Use the concept only with ambient spaces, exact domain, action, and topology stated; distinguish a closed operator from a merely closable one and from a bounded full-domain map.

Functional analysis. Controls limit stability.
Differential operators. Uses boundary-conditioned domains.
Spectral theory. Supports resolvents and spectra.
Quantum theory. Treats unbounded observables mathematically.
Semigroup theory. Analyzes generators.

Clarity¶

Closedness is not the same as boundedness or continuity. State X, Y, D(T), topology, and the operator action before applying the graph criterion.

Manages Complexity¶

A graph packages domain and output simultaneously. Closedness preserves this relation under limits even when the action grows without a uniform bound on the ambient unit ball. Closedness is a stability statement about simultaneous limits of inputs and outputs, not continuity on the ambient space. If x_n in the operator domain converges to x and Tx_n converges to y, a closed operator requires x to remain in the domain and Tx=y. An unbounded differential operator can meet this test on a proper dense domain even though it cannot be bounded on all of X. A merely closable operator has a graph whose closure is itself the graph of an operator; its closure is the smallest closed extension, so closed and closable must not be interchanged. Domain declarations are therefore mathematical data: two operators using the same formula but different domains may have different closedness, adjoints, spectra, and boundary conditions. The central unbounded action–limit stability tradeoff is this: Growth can be unbounded while graph limits remain valid.

Abstract Reasoning¶

Use three linked moves: declare ambient topological vector spaces; specify the exact linear domain and action; take a convergent sequence in the graph. As a collapse test, the identity fails when a convergent graph sequence has a limit pair outside the graph.

Knowledge Transfer¶

Graph-closedness transfers to other partial maps, but linear-operator theorems stop without linearity, Banach hypotheses, or the declared domain. No canonical parent prime is currently asserted; broader structural comparisons remain related-prime analogies until separately adjudicated in the DAG. A closable operator may acquire a closed extension.

Relationships to Other Abstractions¶

Current abstraction Closed Linear Operator Domain-specific

Parents (1) — more general patterns this builds on

Closed Linear Operator is a kind of Linear Operator Domain-specific

Closed Linear Operator satisfies the defining boundary of Linear Operator: A linear operator is a map from a linear subspace of a vector space to another vector space that preserves vector addition and scalar multiplication, with domain, codomain, topology, boundedness, closure, and adjoint conditions declared when relevant.

Children (1) — more specific cases that build on this

Affiliated operator Domain-specific is a kind of Closed Linear Operator

An affiliated operator is a closed densely defined linear operator with the additional differentia of commuting with unitaries in a von Neumann algebra's commutant.

Hierarchy path (1) — routes to 1 parentless root

Closed Linear Operator → Linear Operator → Mathematical Operator → Function (Mapping)

Neighborhood in Abstraction Space¶

Closed Linear Operator sits in a moderately populated region (49^th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Dynamical Systems & Differential Structures (37 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-08