How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
Unbounded Self Adjoint Operators and Stones Theorem
1 · Prerequisites
- Absolute and Conditional Convergence; Rearrangement; Products
- Absolute Continuity and the Sharp Fundamental Theorem of Calculus
- Analyticity of Holomorphic Functions; Liouville and Morera
- Approximation and Compactness in C(K)
- Arc Length and Rectifiable Curves
- Areas of Elementary Plane Figures
- Banach Alaoglu Goldstine and Krein Milman
- Banach Algebras Spectrum and Holomorphic Functional Calculus
- Banach Valued Integration and the Radon Nikodym Property
- Binary Operations, Monoids, Groups and Subgroups
- Bounded Linear Operators and Quotient Spaces
- Bounded Variation and the Riemann–Stieltjes Integral
- Cardinal Arithmetic, Cofinality and the Alephs
- Compact Operators and Riesz Schauder Theory
- Compactness
- Compactness in Metric Spaces
- Completeness, Completion, and Uniform Continuity
- Complex Differentiability and the Cauchy–Riemann Equations
- Complex Lp Spaces and Test-Function Conventions
- Complex Power Series and Analytic Functions
- Connectedness
- Construction of the Natural Numbers
- Construction of the Real Numbers via Cauchy Sequences
- Construction of the Real Numbers via Dedekind Cuts
- Continuity, IVT, EVT, and Uniform Continuity
- Continuous Functional Calculus for Self Adjoint and Normal Operators
- Contour Integration
- Convergence: Nets and Filters
- Cosets, Index and Lagrange's Theorem
- Countability and Uncountability
- Countability Axioms and Cardinal Functions
- Density Separability and Convolution in Lᵖ
- Determinants of Matrices over a Commutative Ring
- Differentiation of Monotone Functions and the Vitali Covering Theorem
- Divisibility, Euclidean Domains, Principal Ideal Domains and Unique Factorisation
- Dual Spaces Adjoint Operators and Annihilators
- Dual Spaces, Bilinear and Quadratic Forms, and Sylvester's Law of Inertia
- Filters and Ultrafilters
- Finite Counting, Factorials and Binomial Coefficients
- Finite Dimensional Normed Spaces and Riesz Lemma
- Foundations of the Real Numbers for Analysis
- Fubini and Change of Variables
- Function Space Topologies and the Exponential Law
- Fundamental Trigonometric Identities
- Gaussian Elimination, Elementary Matrices and Reduced Row Echelon Form
- Gelfand Theory and Commutative C Star Algebras
- Goursat's Theorem and Cauchy's Theorem in a Convex Domain
- Group Actions, Orbits, Stabilisers and Cayley's Theorem
- Hausdorff via the Diagonal
- Hilbert Space Geometry and Riesz Representation
- Ideals, Quotient Rings and the Isomorphism Theorems for Rings
- Inner Product Spaces, Gram-Schmidt, Projections and Adjoints
- Lebesgue Measure on Euclidean Space
- Lebesgue-Stieltjes Measures and Distribution Functions
- Limits of Real Functions
- limsup, liminf, and Subsequential Limits
- Line Integrals and the Gradient Theorem
- Linear Independence, Bases and Dimension
- Linear Transformations, Rank-Nullity and Quotient Spaces
- Matrices, the Matrix of a Linear Map, and Change of Basis
- Measurable Functions and Simple Approximation
- Measures and Their Basic Properties
- Metric Spaces
- Mixed Partials, Taylor Formulae, and Extrema
- Modes of Convergence Egorov and Lusin
- Monotone Functions, Discontinuities, and Continuity Sets
- Monotone Sequences, Bolzano-Weierstrass, and Cauchy Completeness
- Normal Subgroups and Quotient Groups
- Normed and Banach Spaces
- Norming and Separation under Hahn–Banach
- Order, Zorn's Lemma, and the Axiom of Choice
- Ordinal Arithmetic and the First Uncountable Ordinal
- Ordinals, Cardinals, and Transfinite Recursion
- Orthonormal Bases, Parseval and Fourier Series
- Outer Measure and the Caratheodory Extension Theorem
- Partitions of Unity and Paracompactness
- pi: the Equivalent Characterizations
- Polynomial Rings, the Division Algorithm and Roots
- Power Series and Real-Analytic Functions
- Product Measures and the Fubini Tonelli Theorems
- Properties of the Integral and the Working FTC
- Radon Measures and the Riesz Markov Kakutani Theorem
- Relations, Functions, and Quotients
- Rings, Subrings, Integral Domains and Fields
- Rⁿ as a Normed Space; Vector-Valued Functions
- Roots, Rational Powers, and Classical Inequalities
- Separation Axioms: the Hierarchy
- Sequences and Limits
- Sequences and Series of Functions; Uniform Convergence
- Series: Convergence and the Nonnegative Tests
- Sigma Algebras and Borel Sets
- Signed and Complex Measures Hahn and Jordan
- Simple Field Extensions and the Construction of the Complex Numbers
- Sine, Cosine, and the Definition of Pi
- Spectral Measures and Borel Functional Calculus
- Stone–Weierstrass in General
- Subspaces, Products, and Quotients
- Suprema and Infima
- Symmetric Groups, Cycle Decomposition and the Sign Homomorphism
- The Analytic Hahn Banach Theorem
- The Baire Principles of Functional Analysis
- The Cantor Set, Baire Category, and Measure Zero in ℝ
- The Complex Exponential and Euler's Formula
- The Derivative and the Mean Value Theorems
- The Exponential Function
- The Fundamental Theorems of Calculus
- The Lebesgue and Riemann Integrals Compared
- The Lebesgue Integral and the Convergence Theorems
- The Logarithm and General Powers
- The Lᵖ Spaces Holder Minkowski and Riesz Fischer
- The Maximal Function and Lebesgue Differentiation
- The Radon Nikodym Theorem and Lebesgue Decomposition
- The Riemann Integral in Rᵐ and Jordan Content
- The Riemann Integral: Definition and Integrability
- The Topology of Euclidean Space
- The Total Derivative in ℝᵐ → ℝⁿ
- The ZFC Axioms and the Basic Set Constructions
- Topological Spaces and Continuity
- Topology of ℝ
- Triangularisation, Generalised Eigenspaces and Jordan Canonical Form
- Urysohn's Lemma and the Tietze Extension Theorem
- Vector Spaces, Linear Subspaces, Span and Direct Sums
- Weak and Weak Star Topologies
2 · Summary
This page extends the bounded PVM calculus of the preceding pair to unbounded self-adjoint operators and ends with Stone's theorem and the min-max principle. It fixes at the outset that a linear operator carries its domain as part of its data, that means containment of graphs, and that closedness is closedness of the graph in ; the graph norm makes closedness a completeness statement. The adjoint is defined only for densely defined operators, by representability of through Hilbert space Riesz representation, and is proved closed with ; a graph rotation identifies closability with density of and yields . Symmetric, self-adjoint and essentially self-adjoint operators are then defined by , and self-adjointness of , and the minimal derivative with vanishing endpoint conditions is proved closed and symmetric but not self-adjoint, with the periodic operator in between.
The resolvent is used in the library's sign convention. For self-adjoint the identity off the real axis gives with , and the resulting range criterion characterises self-adjointness by . The Cayley transform is unitary with and is proved to be a bijection from self-adjoint operators onto such unitaries, the inverse recovering from . The unbounded PVM integral has domain and is closed, with adjoint ; the spectral theorem is proved by transporting the bounded normal spectral theorem along , and completeness of the calculus is recorded as the product, sum and spectral-mapping rules on the correct domains.
Stone's theorem is proved in both directions: a self-adjoint generates by the Borel calculus, with derivative domain exactly , while the generator of a strongly continuous unitary group is reconstructed from the Laplace resolvents , proved to be bounded inverses of ; the generator is closed, symmetric and skew-adjoint, and a group is determined by its generator, so is a bijection. The deficiency subspaces and the von Neumann parameterization of self-adjoint extensions by unitaries are proved, with the explicit domain and action, and equal deficiency indices are equivalent to existence of a self-adjoint extension. The page closes with the canonical decomposition by spectral type, with relative boundedness, the second resolvent identity, the Kato-Rellich theorem, the discrete and essential spectrum with the Weyl criterion and Weyl's invariance theorem, norm and strong resolvent convergence with the continuous calculus and unitary-group corollary, and the form domain and min-max principle below the essential spectrum.
3 · Logical flowchart
4 · Definitions, theorems and proofs
Unbounded linear operators: domain, graph and extension
Definition
Throughout this page is a complex Hilbert space (Hilbert space) with inner product linear in the first variable and conjugate-linear in the second (Real and complex inner-product spaces and their induced length).
A (possibly unbounded) linear operator on is a linear map (Linear map between vector spaces over the same field) whose domain is a linear subspace (Linear subspace of a vector space). The domain is part of the data: means and for all . One writes , and calls an extension of , when and for every .
The graph of is The direct sum is read as the complex vector space of pairs with coordinatewise operations and the inner product , whose induced norm is it is complete because is (Complete metric space: every Cauchy sequence converges in the space). The graph norm on is so that is an isometric isomorphism of onto with the norm restricted from .
is closed when is a closed subset of . A linear operator is determined by its graph, and the following elementary translations are used silently below: is a linear subspace of ; ; the image of under the first coordinate projection is ; and if and only if . In particular a closed operator is exactly one whose graph is a closed subspace of .
Notation. No boundedness of is assumed, and is frequently called unbounded to stress this; a bounded everywhere defined operator on is the special case in which is a closed subspace by continuity. The letter denotes the identity operator on with domain .
Completeness of . If is a Cauchy sequence in , then and show that and are Cauchy in ; let be their limits. Then , so is complete. Consequently a sequence in converges exactly when its two coordinate sequences converge in .
Densely defined, closed and closable operators, and cores
Definition
Assume Countable Choice (The Axiom of Countable Choice ()). Let be a linear operator on with domain and graph (Unbounded linear operators: domain, graph and extension).
is densely defined when is a dense subset of (Dense, nowhere dense and codense subsets of a topological space, and the criterion by basic open sets); closed when is a closed subset of ; and closable when has a closed extension. Recall that the graph norm is .
The graph-norm dictionary. The following three statements are part of the definition's content and are proved in the remarks below rather than assumed:
- is a norm on , and is an isometric isomorphism of onto ;
- is closed if and only if is a complete metric space (Complete metric space: every Cauchy sequence converges in the space);
- if is closed, then is a Hilbert space for the inner product whose induced norm is .
A linear subspace is a core for a closed operator when is dense in , equivalently when , where the closure is taken in and denotes the restriction of to .
Remarks
Claim 1. holds on , and always. The map is linear, and . Hence homogeneity and the triangle inequality for are the corresponding norm properties in pulled back along . It is definite because forces and . The map is therefore linear and isometric, and its image is with ; a linear isometry is injective, so it is a bijection onto .
Claim 2. The space is complete with (Complete metric space: every Cauchy sequence converges in the space, Hilbert space), while carries the subspace metric. If is closed, every Cauchy sequence in it converges in and its limit remains in , so the graph is complete. Conversely suppose is complete and . Countable Choice selects with for every . Then is Cauchy, so it converges to a point of ; uniqueness of metric limits makes that point . Thus is closed. Since the map of claim 1 is an isometry onto , the space is complete exactly when is closed, that is, exactly when is closed.
Claim 3. If is closed then is a closed subspace of the Hilbert space , hence a Hilbert space in its own right, and the inner product transfers to through the isometry of claim 1, making a Hilbert space with inner product and induced norm .
Core. The isometry of claim 1 is a homeomorphism from onto and maps onto . Hence it carries the closure of onto the closure of inside . Therefore is graph-norm dense exactly when ; this topological argument does not replace density by sequential density.
Closure of a closable operator
Statement
Assume Countable Choice (The Axiom of Countable Choice ()). Let be a linear operator on with domain and graph (Unbounded linear operators: domain, graph and extension). Then the following are equivalent:
- is closable (Densely defined, closed and closable operators, and cores);
- whenever , and , one has .
If either condition holds, then the closure of in is the graph of a linear operator , the closure of ; it is the least closed extension of , and is closed if and only if . Necessity of the hypothesis is never claimed: both conditions hold automatically for a closed operator.
Facts & Assumptions
is a linear subspace of ; exactly when ; is closed exactly when is closed; and (Unbounded linear operators: domain, graph and extension).
is closable when it has a closed extension; the closure of a subset of a metric space is closed, is contained in every closed set containing , and every point of is the limit of a sequence in (Interior, closure, boundary, limit point, isolated point and dense subset of a metric space, Densely defined, closed and closable operators, and cores).
A sequence in converges exactly when its two coordinate sequences converge in (Unbounded linear operators: domain, graph and extension).
Proof
Given: A linear operator on and the two conditions (1) and (2).
If has a closed extension , then with closed, and for with , we get by [A3] with ; closedness of gives , and by the last clause of [A1] applied to this forces . Thus (1) implies (2).
Now assume (2), and let . Then is closed and, being the closure of the linear subspace , is itself a linear subspace. If then by [A2] there are with , hence and by [A3], so by (2). Therefore .
By step 1.2, determines at most one second coordinate per first coordinate: if then since is a subspace, so . Hence is a linear subspace of , the formula for defines a linear operator with , and is closed with because .
By step 2.1 the operator is a closed extension of , so is closable, and (2) implies (1). With 1.1 this proves the equivalence of (1) and (2), and it shows that whenever either holds the closure is the graph of the closed extension .
If is any closed extension of , then is closed and contains , so by [A2], that is, . Thus is the least closed extension of .
If is closed then is already closed, so and ; conversely if then is closed because is closed by step 2.1. Hence is closed if and only if , and the closure of the graph is the graph of the least closed extension.
Adjoint of a densely defined operator
Definition
Assume the Axiom of Countable Choice (The Axiom of Countable Choice ()). Let be a densely defined linear operator on , with domain dense in (Densely defined, closed and closable operators, and cores, Dense, nowhere dense and codense subsets of a topological space, and the criterion by basic open sets). Since is dense, and since is complete (Hilbert space), every bounded linear functional on the normed space (A bounded linear operator between normed spaces) extends uniquely to a bounded linear functional on of the same norm, by setting for any sequence with in ; the limit exists because is bounded on a dense subspace and is complete, and it does not depend on the sequence because a bounded functional is uniformly continuous.
A vector belongs to the adjoint domain when the linear functional , , is bounded on . In that case Hilbert space Riesz representation (Riesz representation for Hilbert spaces) applied to the extension produces a unique vector with equivalently for all . The map , , is the adjoint of .
Well-definedness. The functional is linear in for each , so its domain of boundedness is a linear subspace: if are bounded so is for scalars , because the first-variable-linear convention gives . On the map is linear: if are represented by , then is represented by , since ; the representing vector is unique. By Riesz representation for the Hilbert space a vector is determined by the values with ranging over , and those values are determined by the functional . Equivalently, if both represent , then for every ; density of and continuity of the inner product extend this equality to every , and taking gives . Finally only ambient-norm boundedness of on is required: no extension, no closure and no closedness of is presupposed.
The adjoint is well defined, closed, and reverses inclusions
Statement
Assume Countable Choice. For a densely defined linear operator on the adjoint is well defined and linear with linear domain ; is closed; for every and if and both are densely defined, then .
Facts & Assumptions
holds exactly when is bounded on , and then is the unique with for all ; is linear and is a linear subspace (Adjoint of a densely defined operator, The Axiom of Countable Choice ()).
is closed exactly when is closed, and (Unbounded linear operators: domain, graph and extension, Adjoint of a densely defined operator).
Proof
Given: A densely defined linear operator on .
By [A1] the adjoint is well defined, is a linear subspace and is linear.
Let with and . For every we have , both limits being scalar limits; hence is bounded, with . So and by [A1], and therefore is closed.
Let and . If , then , that is , for every ; this exhibits as a bounded functional, so and by uniqueness in [A1]. Conversely if , then for every , so . Hence .
If are densely defined, then for every , so every pair lies in by [A3]; hence .
Claims collected: well-definedness and linearity from step 1.1, closedness from step 1.2, the kernel-range identity from step 1.3, and the inclusion reversal from step 1.4. ∎
Closability is equivalent to density of the adjoint domain
Statement
Assume Countable Choice. Let be a densely defined linear operator on . Then is closable if and only if is dense in . In that case and ; moreover if , then is closable.
Facts & Assumptions
For densely defined the adjoint is defined by for , , and . Putting defines a bijective isometry of with , and maps orthocomplements to orthocomplements: (Adjoint of a densely defined operator, Hilbert space, Orthogonality and the orthogonal complement).
If is a linear subspace of the Hilbert space , then (The double orthogonal complement of a subspace is its closure).
For a densely defined the operator is closed. If in addition is dense, then is defined, is closed, and contains (The adjoint is well defined, closed, and reverses inclusions, Adjoint of a densely defined operator).
is closable when it has a closed extension; if is closable then is the least closed extension of and (Closure of a closable operator, Densely defined, closed and closable operators, and cores).
Proof
Given: A densely defined linear operator on .
By [A1] we have : indeed means for all , which is exactly for all , that is .
Assume in addition that is dense, so that is defined. Replacing by in step 1.1 gives , so by step 1.1 and unitarity of , with and , one has .
Conversely assume is closable, and let be its closure, a closed densely defined operator with and by [A4]. Then is dense: if then , so by step 1.1 applied to and closedness of we get , that is for some , forcing and .
Under the hypothesis of step 2.1 the set is the graph of the operator , which is closed by [A3]; hence has a closed extension and is closable, and its closure is by minimality in [A4].
With closable as in step 2.2, apply step 1.1 to and to : using and [A2] one gets , so and is dense by step 2.2.
The two implications are steps 3.1 (density of gives closability, with closure ) and 3.2 (closability gives density of , and ). If , then is a closed extension of by [A3], so is closable.
Symmetric, self-adjoint and essentially self-adjoint operators
Definition
Assume Countable Choice. Let be a densely defined linear operator on .
- is symmetric when , that is, and for all ; equivalently for all .
- is self-adjoint when , that is, the domains and the values agree.
- is essentially self-adjoint when its closure is self-adjoint; by Closability is equivalent to density of the adjoint domain this presupposes that is closable.
Consequences, with proofs. These are part of the content of the definition.
- A symmetric is closable. and is closed (The adjoint is well defined, closed, and reverses inclusions), so is a closed extension of . In particular the closure exists and .
- The closure of a symmetric operator is symmetric. If , the inclusion reversal of The adjoint is well defined, closed, and reverses inclusions applied to gives , that is . Applying it once more to the inclusion gives , that is . Since and (Closability is equivalent to density of the adjoint domain), this reads : the closure is symmetric. (Each application is legitimate because the adjoint is defined once its operator is densely defined, and is dense because , so that contains the dense domain .)
- A self-adjoint operator has no proper symmetric extension. If and , then by inclusion reversal, so and all inclusions are equalities.
- A self-adjoint operator is closed, being equal to the adjoint of the densely defined , and an essentially self-adjoint operator has exactly one self-adjoint extension, namely : a self-adjoint extension of is closed, hence contains the least closed extension , and then by item 2 and symmetry, so .
A symmetric closed operator that is not self-adjoint
Statement refuted
Assume the Axioms of Countable Choice and Dependent Choice. On the Hilbert space (The space as the quotient by null functions) let with domain where complex-valued absolute continuity and the derivative are read on real and imaginary parts (Absolute continuity on a compact interval). Domain notation means the classes having the indicated absolutely continuous representative; endpoint values refer to that representative. Then:
- is densely defined, closed and symmetric, so refutes the reading "closed and symmetric implies self-adjoint";
- and , so is a proper closed extension of ; itself is not symmetric, and ;
- the periodic domain carries a closed symmetric extension of , strictly between and .
Facts & Assumptions
A densely defined operator is symmetric when and self-adjoint when ; the adjoint of a densely defined operator is always closed (Symmetric, self-adjoint and essentially self-adjoint operators, The adjoint is well defined, closed, and reverses inclusions).
For real-valued absolutely continuous functions the fundamental theorem of calculus and integration by parts hold, and the indefinite integral of an function is absolutely continuous; applied to real and imaginary parts this gives the same calculus for complex-valued absolutely continuous functions (Absolute continuity on a compact interval, Fundamental theorem of calculus for absolutely continuous functions, Integration by parts for absolutely continuous functions, The indefinite integral of an function is absolutely continuous, The indefinite integral of an function is differentiable almost everywhere).
is dense in under Countable Choice, applied componentwise for complex functions. There exists a smooth on with , equal to one on and zero off . Dominated convergence applies under an integrable majorant ( is dense in for , Explicit compactly supported smooth cutoffs ↗, Dominated convergence).
exactly when is bounded on , and then for all (Adjoint of a densely defined operator, The adjoint is well defined, closed, and reverses inclusions).
Under Countable Choice, complex is a Hilbert space of almost-everywhere classes with first-variable-linear pairing ( with the integral pairing is a Hilbert space, The space as the quotient by null functions). The pairing satisfies Cauchy–Schwarz; taking and on an interval of length at most one gives (The complex pairing is well-defined and satisfies Cauchy–Schwarz).
Counterexample
Given: and on the domain above.
An absolutely continuous representative is continuous and unique in its almost-everywhere class: a nonzero difference at a point would stay nonzero on a relative interval of positive length. It is bounded on [0,1] and thus belongs to ; its a.e. derivative is independent of the representative. The stated domains are linear and define linear operators. Complex absolutely continuous calculus is obtained componentwise from the real theory. If are complex-valued with absolutely continuous real and imaginary parts, then for all , and ; if then is of this kind with derivative almost everywhere by the first fundamental theorem, using the inclusion in [A5]. These are the uses of the stated Countable Choice and Dependent Choice through the calculus and Hilbert-space suppliers.
Given , extend it by zero to on . By [A3] and Countable Choice, choose complex with , using real and imaginary approximations if necessary. For put on and extend it by zero outside. It vanishes in neighborhoods of both endpoints, so this extension is smooth with compact support in . Also and for every . Then and by domination by for the squared error. Thus is dense.
Symmetry: for the boundary term in 1.1 vanishes because , so ; hence , and is symmetric.
Closedness: let with and in . Then in , so by 1.1 the functions converge uniformly to on , since by [A5], with ; hence pointwise and in , so , that is, is absolutely continuous with almost everywhere, and . Thus and , so is closed. The sequential graph criterion applies in this metric product under the assumed Countable Choice.
Adjoint computed. If has , then for every the identity in step 1.1 read backwards gives , so and .
Conversely let and put ; by 1.1 the function is absolutely continuous with and . For every integration by parts in the form of 1.1 gives , while . Hence for every .
The derivatives of elements of are exactly : one inclusion follows from step 1.1, and conversely has derivative , vanishes at both endpoints, and belongs to . Set and . Step 2.5 says for every . Taking gives , since . Thus as a class and . This supplies an absolutely continuous representative of with a.e., so . No unproved orthogonal-hyperplane assertion is needed.
By steps 2.4 and 3.1, and . This is strictly larger than : the constant function lies in but not in . Hence , and is a proper closed extension of by [A4]. Moreover is not symmetric: for , one computes from step 1.1 that .
Let on the periodic domain . It is densely defined because it extends . For the boundary form in step 1.1 vanishes, so is symmetric. If , the adjoint identity restricted to gives and by [A4] and step 4.1. For arbitrary , integration by parts therefore gives . Choose : then , so . Conversely the boundary form vanishes for every periodic , giving and . Hence with equality of domains; is self-adjoint and therefore closed by [A1]. The constant function lies in , and lies in , proving both strict inclusions.
Every claim is witnessed: is densely defined and closed by steps 2.1 and 2.3, symmetric by step 2.2, and by step 4.1; the failure of "symmetric implies self-adjoint" is therefore established, and no claim is made that a self-adjoint extension does not exist: the periodic domain of step 5.1 provides one by its explicitly computed adjoint.
Resolvent and spectrum of an unbounded operator
Definition
Let be a complex Hilbert space and a linear operator on with domain . Write for . The resolvent set of is where is the space of bounded everywhere defined operators on (A bounded linear operator between normed spaces); the spectrum is . For the bounded operator is the resolvent of at . This is the library convention used throughout the page, matching the bounded resolvent of Spectrum and resolvent of a bounded operator: the shift is , with coefficient on .
The resolvent determines back. If and , then and for ; equivalently with , and for , for .
Nonempty resolvent set forces closedness, without the closed graph theorem. Assume and fix with . Choose a bound such that . The map is continuous, since . Its zero set is closed. Because and for every , the graph of is where is a continuous linear bijection of with continuous inverse , since and . A homeomorphism carries closed sets to closed sets, so is closed and is closed. No closed graph theorem and hence no choice principle is used here.
For the zero operator on a nonzero , : the nonzero shifts have inverse , whereas the zero shift is not bijective. If , every shift is the unique bijection of the zero space, so and .
Resolvent of a self-adjoint operator: nonreal resolvents and the estimate
Statement
Assume Countable Choice. Let be a self-adjoint operator on . Then every nonreal number belongs to : . More precisely, for with , , and every , and consequently . In particular .
Facts & Assumptions
; thus is dense, , and is closed, being closed (Symmetric, self-adjoint and essentially self-adjoint operators, The adjoint is well defined, closed, and reverses inclusions).
For one has whenever , and is a real number: it equals (Symmetric, self-adjoint and essentially self-adjoint operators, Real and complex inner-product spaces and their induced length).
For the identity holds (The adjoint is well defined, closed, and reverses inclusions, [A1]).
If is a linear subspace of a Hilbert space, then (The double orthogonal complement of a subspace is its closure, Orthogonality and the orthogonal complement).
means that is a bijection of onto with bounded inverse, and then is the operator norm of that inverse (Resolvent and spectrum of an unbounded operator, The operator norm as the least bound and as the unit-sphere or unit-ball supremum).
Proof
Given: Countable Choice, a self-adjoint operator on , and a number with .
For , expanding gives ; by [A2] the two middle terms combine to , so .
Adding and subtracting and using , step 1.1 becomes .
By step 2.1, for every ; in particular is injective and its range is closed: if , then is Cauchy, hence for some and , and closedness of gives and , that is .
Also by [A3] applied to , , and step 2.1 with replaced by shows is injective, so the kernel is . Hence the closed range of step 3.1 satisfies .
By steps 3.1 and 4.1 the map is a bijection, and step 2.1 gives for every : applying step 2.1 to yields . Thus and by [A5].
Since was an arbitrary nonreal number, , that is, ; the identity and the bound of the statement are steps 2.1 and 5.1.
Range criterion for self-adjointness
Statement
Assume Countable Choice. Let be a densely defined symmetric operator on . Then the following are equivalent:
- is self-adjoint;
- is closed and for every ;
- is closed and ;
- for every ;
- ;
- .
In particular a closed symmetric operator is self-adjoint if and only if , and the same criterion holds with replaced by for any real .
Facts & Assumptions
, is dense, and is self-adjoint exactly when ; a self-adjoint operator is closed, and for in the domain of a symmetric operator the number is real (Symmetric, self-adjoint and essentially self-adjoint operators, The adjoint is well defined, closed, and reverses inclusions, Real and complex inner-product spaces and their induced length).
For a linear subspace of a Hilbert space, ; in particular if is closed and then (Orthogonality and the orthogonal complement, The double orthogonal complement of a subspace is its closure).
For self-adjoint one has , hence ; conversely every gives that is a bijection (Resolvent of a self-adjoint operator: nonreal resolvents and the estimate, Resolvent and spectrum of an unbounded operator).
If then is closed (Resolvent and spectrum of an unbounded operator).
Proof
Given: A densely defined symmetric operator on .
Preparatory identity. Let be densely defined and symmetric, let with and let . Expanding and using that is real by [A1] gives, as in the proof of Resolvent of a self-adjoint operator: nonreal resolvents and the estimate,
(1) implies (2): if then is closed by [A1]. If is nonreal and , then and by [A1], so .
(4) implies (5) is immediate, since are nonreal.
(5) implies (1): let . Since , choose with . Then , because for ; and by [A2] with the kernel of is . Hence . So , and with this gives .
(1) implies (6) by [A4]. Conversely (6) implies (4): if then every nonreal lies in , so is surjective and .
Closed range for closed symmetric . If in addition is closed, then is closed for nonreal : given , step 1.1 makes Cauchy with limit , so , and closedness of gives with , that is .
(5) implies (3): by 1.1 with and the map satisfies , so its inverse on its range is bounded by ; since the range is by (5), because , and is closed by [A5]. The two kernels vanish by [A2] and (5).
(2) implies (4): for nonreal , [A2] and (2) give , while (2) and step 2.1 show is closed; hence the range is all of by [A3].
(3) implies (5): by [A2], , and by step 2.1 applied to the closed with the two ranges are closed; hence they equal by [A3]. So (3) and (5) are equivalent.
Collecting: by steps 1.2, 3.1, 1.3 and 1.4; and by steps 2.2 and 3.2; and by step 1.5. Thus all six statements are equivalent. The final clause follows because only nonreality of the parameters was used, so with may replace .
Cayley transform of a self-adjoint operator
Definition
Assume Countable Choice (The Axiom of Countable Choice ()). Let be a self-adjoint operator on . By Resolvent of a self-adjoint operator: nonreal resolvents and the estimate the points lie in , so and are bijections of onto with bounded inverses, and is a bounded everywhere defined operator, the Cayley transform of . In the resolvent convention of Resolvent and spectrum of an unbounded operator one has
Its properties, with proofs. Writing :
- is isometric. For put , so that with ; then , and the symmetry computation of Resolvent of a self-adjoint operator: nonreal resolvents and the estimate gives .
- is unitary, with . Using and the adjoint rule for the self-adjoint one gets ; the elementary resolvent identity then gives , as follows also from applying the computation of item 1 to and to and using that a surjective isometry of onto is unitary (A bounded linear operator between normed spaces). Concretely for by direct substitution, and both sides are continuous.
- , and , as maps on . Indeed is injective with inverse , while for .
- Domain recovery. : by item 3, .
Cayley correspondence between self-adjoint operators and unitaries
Statement
Assume Countable Choice. The map is a bijection from the set of self-adjoint operators on onto the set of unitary operators on with . The inverse map assigns to such a the operator For this both and map onto , so is self-adjoint, and .
Facts & Assumptions
For self-adjoint the Cayley transform is unitary, , and ; also for (Cayley transform of a self-adjoint operator, Resolvent of a self-adjoint operator: nonreal resolvents and the estimate).
A densely defined symmetric operator is self-adjoint if (Range criterion for self-adjointness).
For unitary one has and is bijective, with ; moreover for bounded , so is dense exactly when (Hilbert-adjoint identities, Kernel–range orthogonality for Hilbert adjoints, Orthogonality and the orthogonal complement, Dense, nowhere dense and codense subsets of a topological space, and the criterion by basic open sets).
Proof
Given: A unitary with , and the map of [A1].
The assignment is well defined: if satisfy , then because ; hence is well defined and defines a map on .
is dense: by [A3], , and has kernel because means , that is .
Conversely for every self-adjoint : by [A1] and , so for one has and ; hence the inverse construction sends to .
is symmetric: for , expanding both pairings and using , one gets .
Ranges: for every one has and ; hence and, since is onto by [A3], .
By steps 1.2, 2.1 and 2.2 the operator is densely defined, symmetric, and has both ranges equal to , so is self-adjoint by [A2].
: by [A1] applied to the self-adjoint and by step 2.2 one has , so ; since is onto , .
By steps 1.3 and 4.1 the two constructions are mutually inverse, and every assignment above is a bijection by construction; hence is a bijection onto the stated class.
Integral of a measurable function against a projection-valued measure
Definition
Assume Countable Choice. Let be a projection valued measure on the measurable space acting on the complex Hilbert space (Projection valued measure), and let be -measurable. Put where the scalar measures are those of Scalar and complex measures from a pvm, and for set For , the bounded integrals are those of Bounded borel pvm integral, and the limit is taken in the norm of . For , define to be the unique operator on for every bounded measurable . Then is the zero measure, , and . Thus this case is defined directly without applying a theorem requiring a nonzero space.
Well-definedness. is a linear subspace: the estimate for scalar measures and follow from and . Writing the sets are measurable, so is bounded and measurable. For , The integral of the right side against tends to zero by Dominated convergence, with the integrable majorant and pointwise limit zero since is finite-valued. Linearity of the bounded calculus (Pvm integral is a star homomorphism) and its quadratic identity give These clauses hold directly on the zero space as well. Hence the truncation vectors are Cauchy and converge uniquely by Hilbert-space completeness (Hilbert space). The limit is linear in because every truncation operator is linear and addition and scalar multiplication are norm-continuous. Countable Choice is inherited from the bounded PVM suppliers; no further choice is needed to take these specified limits.
Finally, if two -measurable functions agree outside a measurable set with , then for every . Thus the integrals of and agree, so their domains agree. At every truncation level their bounded truncations agree outside ; the quadratic identity applied to the difference gives equal truncation vectors for every . Taking limits gives on their common domain.
The unbounded PVM integral is densely defined, closed and normal
Statement
Assume Countable Choice. Let be a projection valued measure on on the complex Hilbert space and let be -measurable. Then the operator of Integral of a measurable function against a projection-valued measure is densely defined and closed, and it is normal, in the sense and equal norms for the operator and its adjoint; in particular and for all such . Moreover and ; in particular is self-adjoint whenever takes real values. Finally, if are bounded -measurable functions with pointwise and pointwise, then for every .
Facts & Assumptions
is a linear subspace, with , and the limit is linear in ; here and below for bounded -measurable (Integral of a measurable function against a projection-valued measure).
For and bounded -measurable : , , , products of bounded -measurable functions multiply as , and is linear (Bounded borel pvm integral, Pvm integral is a star homomorphism, Scalar and complex measures from a pvm).
Dominated convergence on the finite measure : if pointwise and with , then . In particular, if and with integrable , then by applying the theorem to the squared differences (Dominated convergence, Scalar and complex measures from a pvm).
Scalar monotone convergence: for nonnegative measurable the integrals increase to (Monotone convergence for the integral).
The adjoint of a densely defined operator is closed, and consists of those for which is bounded, with the representing vector (Adjoint of a densely defined operator, The adjoint is well defined, closed, and reverses inclusions).
Projection values satisfy , , and (Projection valued measure). Closedness means the graph is closed, and density means the domain is dense (Densely defined, closed and closable operators, and cores).
Proof
Given: A PVM and a -measurable function as in the statement.
If , [A1] defines all integrals on as the unique zero-space operator. Its domain is all , its graph is the whole , its adjoint is itself by the representing identity, and every scalar integral and norm in the statement is zero. All approximation vectors are zero. This proves every assertion in that case; henceforth assume , as required by [A2].
For any measurable and , norm convergence of the defining truncations and monotone convergence give . Also is integrable against the finite measure . Thus dominated convergence in the bounded quadratic pairings gives , with the conjugate required by the first-variable-linear inner product.
Fix a bounded measurable and . For each , bounded linearity and the quadratic identity give . Let ; the left side converges by [A1], while the right side converges by [A3], since , an integrable majorant. Consequently . For the sequence in the statement, choose its bound ; the majorant and [A3] now imply the claimed convergence. This argument applies to any measurable target function in place of .
If is bounded measurable, [A2] and [A6] give , and hence . Thus for all . For this implies , and also , by the bound . Bounded multiplication gives . The first two limits follow from [A1] and bounded continuity; the last tends to by step 1.3 applied to the target , because . Therefore on .
Put for . For every , step 2.1 gives , so . Moreover by [A2], [A6] and dominated convergence, since is finite-valued. Hence the domain is dense. For every , the defining truncations on are constant for : . Therefore .
The domains of and coincide because their defining squared moduli agree. For in this domain, bounded adjoints and the defining limits yield . Since density is established in step 3.1, [A5] gives .
Conversely let and . For every , step 3.1 and the adjoint identity give . Thus . By [A2], [A4] and the projection bound, . Hence , and step 4.1 proves with equal domains.
Apply step 5.1 to the measurable function , whose integral has dense domain by step 3.1. It gives , so [A5] proves closed directly. Step 1.2 and give equal norms for and its adjoint on their common domain, establishing normality. If is real-valued, step 5.1 gives . Together with steps 1.2, 1.3 and 3.1 this proves every assertion.
Spectral theorem for unbounded self-adjoint operators (PVM form)
Statement
Assume the Axiom of Choice. Let be a self-adjoint operator on the nonzero complex Hilbert space . Then there is a unique regular projection valued measure on the Borel -algebra of such that
Conversely, if is a regular projection valued measure on , then the operator with that domain is self-adjoint and its spectral projection valued measure is again.
Facts & Assumptions
For a Borel function the integral of Integral of a measurable function against a projection-valued measure has domain , is closed, satisfies and , and is self-adjoint for real (The unbounded PVM integral is densely defined, closed and normal).
Every bounded normal operator has a unique regular spectral PVM with ; the support of is , and is unique among regular PVMs representing on a compact set (Spectral theorem for bounded normal operators pvm form, Support and uniqueness of the spectral measure).
is unitary with , , and on (Cayley transform of a self-adjoint operator, Cayley correspondence between self-adjoint operators and unitaries).
For a unitary one has : forces by the Neumann series, so lies in the closed unit disk, and with equivalent to (Neumann series, Spectrum and resolvent of a bounded operator).
For bounded Borel on the identity holds. If a PVM on satisfies and is a Borel isomorphism, then defines a PVM on with ; regularity is preserved by this transport (Bounded borel pvm integral, Projection valued measure, Regular Borel measure on an LCH space).
Proof
Given: A self-adjoint operator on and .
By [A3] the operator is unitary with , and by [A4] its spectrum lies in . Let be the regular spectral PVM of on supplied by [A2], and extend it to by for Borel . Then is a regular PVM on , is carried by , and .
: for a bounded normal operator the spectral projection at a point is the orthogonal projection onto . Indeed, if , then is carried by , so the bounded calculus gives Conversely, if , the same identity shows that is carried by ; hence and . Thus ; with this kernel is by [A3], so the projection is zero.
Let . Then is a homeomorphism of onto , and is a PVM on the Borel sets of with ; it is regular because is and is a homeomorphism, and .
Let be the self-adjoint operator with domain , given by [A1]; write , a bounded Borel function. Then , and because and on the natural domain; hence and , so on .
Using step 3.1 in the formula of [A3] gives ; now , so on , because .
Uniqueness: let be a regular PVM on with and for , and put for Borel . Then is a regular PVM on , , and , where is proved as in step 3.1 with in place of . By the uniqueness clause of [A2] applied to the bounded normal operator , is carried by and agrees there with , hence on . Therefore .
Conversely, if is a regular PVM on , then is self-adjoint by [A1] and represents ; by step 4.2 the representing PVM is unique, so is the spectral PVM of .
Unbounded Borel functional calculus: domains, products, spectral mapping
Statement
Assume the Axiom of Choice. Let be a self-adjoint operator with spectral projection valued measure on acting on a complex Hilbert space (Spectral theorem for unbounded self-adjoint operators (PVM form)) and let be Borel. Write , with the truncation definition below. If , use its unique PVM and unique full-domain operator directly. Then:
- ;
- has domain and equals the restriction of to that domain, and its closure is ;
- on the sum equals the restriction of , and the closure of is ;
- the spectrum of is the essential range for every of with respect to ;
- if is continuous then that essential range is the closure of , with the closure redundant when is closed.
Facts & Assumptions
For every finite-valued measurable , has domain , is densely defined and closed, and satisfies and . It is the norm limit of , where . Bounded approximants dominated by and converging pointwise to converge on (The unbounded PVM integral is densely defined, closed and normal, Integral of a measurable function against a projection-valued measure).
On nonzero the bounded measurable PVM calculus is linear, unital, multiplicative and adjoint preserving, and . The scalar measure is finite with total mass (Bounded borel pvm integral, Pvm integral is a star homomorphism, Scalar and complex measures from a pvm).
Dominated convergence holds for an integrable majorant, and monotone convergence holds for increasing nonnegative measurable functions (Dominated convergence, Monotone convergence for the integral).
A spectral PVM represents as the integral of the identity function with its exact squared-integrability domain; the cited spectral theorem assumes nonzero and AC (Spectral theorem for unbounded self-adjoint operators (PVM form), The Axiom of Choice). The resolvent convention is , required bounded and everywhere defined (Resolvent and spectrum of an unbounded operator).
, projection values are contractive, and is strongly countably additive with (Projection valued measure). Rational intervals form a countable base of , by countability and density of the rationals ( is countably infinite, Both and are dense in , and every nonempty open subset of is uncountable).
Proof
Given: The self-adjoint operator and spectral PVM in the statement, under AC.
For all integrals are the unique full-domain operator by [A1]. Sums, products, adjoints and closures are that operator; its resolvent is all and its spectrum is empty. Every spectral projection is zero, so every essential range here is empty, as is . Thus all claims hold. Henceforth , and all bounded calculus uses have the hypothesis required by [A2].
For bounded Borel and arbitrary , bounded multiplication gives , so . For , truncate : norm convergence and boundedness of give by monotone convergence. Hence , including when is unbounded.
We will also use a dominated approximation with a general square-integrable majorant. Fix and bounded Borel . Comparing with using [A2] and then passing gives ; the scalar limit is dominated by . Therefore if bounded pointwise and with , then is square integrable and , by dominated convergence with .
For , since . Bounded linearity and step 1.3 applied to with show . In particular and , since constants are integrable against finite .
If is bounded and , step 1.2 shows , while . The bounded identities pass to limits: the last uses [A1] with target since . Thus for .
The definition of composition and step 1.2 yield . For in that domain, step 2.2 gives . The left side tends to by [A1]. To control the right side without pretending bounded, note that : apply step 2.2 with there replaced by and the bounded indicator. The projections converge strongly to , since their complementary squared norms are integrals of decreasing indicators against finite scalar measures. Hence the right side tends to , proving the product value.
For use . Step 1.2 shows , and gives and . Step 2.1 and closedness in [A1] prove the sum closure. These cutoff ranges also show the sum domain dense.
For put . Step 1.2 shows . The same step and dominated convergence give and . Thus every point of the graph of is a limit of graph points of . The reverse graph inclusion follows from step 3.1 and closedness in [A1], proving the product closure. The composition domain is dense as well: the ranges of , contained in that domain, approximate every vector by the same indicator estimate.
Suppose for some . Define the Borel function piecewise: when , and otherwise. It is bounded by , and off an -null set. By the domain and value identities of step 3.1, has domain all and equals ; has domain and equals the identity there. Null-set invariance is supplied by [A1]. By step 2.1, and (linearity with constants, or multiplication by in step 2.2). Thus is its bounded inverse and in the convention of [A4]. Only the inverse is asserted bounded.
Conversely suppose for every . Using AC in [A4], choose unit in the range of for each . The scalar measure of is supported there by [A5], and there, so . Hence by [A1] and step 2.1. A bounded inverse with bound would give for every , impossible. This proves the essential-range formula.
Apply steps 4.2 and 5.1 to , which represents by [A4]. Nonreal have a ball disjoint from , so ; for real , membership in is equivalent to every interval about having nonzero projection. The union of all rational intervals with zero projection is exactly , by the rational base in [A5] and projection monotonicity from . Enumerate pairs of a fixed rational enumeration by increasing sums of their indices, giving an enumeration of rational intervals. In that enumeration retain such intervals and replace the rest by the empty set. Disjointify this sequence by subtracting previous intervals. Each resulting set has zero projection, and their union is , so strong countable additivity gives . In particular is open and measurable.
Let be continuous. If , a ball about has preimage contained in of step 6.1, so that preimage has zero projection. Conversely, for and , choose with . Continuity supplies an interval about whose image lies in ; its projection is nonzero by step 6.1, so the whole preimage has nonzero projection. This proves the continuous spectral-mapping formula. The adjoint identity is [A1], and steps 3.1, 4.1, 3.2, 4.2 and 5.1 prove the remaining claims.
Strongly continuous one-parameter unitary group
Definition
A strongly continuous one-parameter unitary group on the complex Hilbert space is a map such that , for all , each is unitary (that is, onto and norm preserving, equivalently , A bounded linear operator between normed spaces, The operator norm as the least bound and as the unit-sphere or unit-ball supremum), and the orbit maps are continuous at every for every , from the usual metric on to the norm metric of (Continuity of a map between metric spaces, at a point and globally, in the - form).
Continuity at the origin suffices, and weak continuity is equivalent to strong continuity. These two clauses are part of the definition's content and are proved here. If is continuous at and , then and is isometric, so as ; the group law and isometry turn continuity at one point into continuity everywhere. Likewise, if is merely weakly continuous at , then for the expansion shows that norm convergence at follows from ; and at an arbitrary one has , so weak continuity at for every follows from the case . Conversely, norm continuity of an orbit implies its weak continuity, since for each fixed , Cauchy--Schwarz gives Finally, a group with is automatically invertible with , so the unitarity and group clauses are symmetric in .
Infinitesimal generator of a unitary group
Definition
Let be a strongly continuous one-parameter unitary group on (Strongly continuous one-parameter unitary group). Its infinitesimal generator is the linear operator with domain where this is the two-sided norm limit over real : it has value precisely when for every there is such that Equivalently, defining the quotient's value at to be makes it continuous there from the usual metric on to the norm metric on (Continuity of a map between metric spaces, at a point and globally, in the - form). Such a is unique by the triangle inequality.
is a linear subspace and is linear: if and are scalars, then converges with limit , because the operations of are continuous and both estimates use the same punctured real parameter ; the restriction to is therefore well defined and linear (Linear subspace of a vector space, Linear map between vector spaces over the same field).
Sign convention. This page writes for the generator, so that Stone's theorem reads with self-adjoint; equivalently . In the convention of Teschl's book one has with self-adjoint, so and hence ; every formula below is written in the convention and the translation is recorded where a source is cited.
A self-adjoint operator generates a strongly continuous unitary group
Statement
Assume the Axiom of Choice. Let be a self-adjoint operator on with spectral projection valued measure , and set for . Then is a strongly continuous one-parameter unitary group, and on , and the derivative exists exactly for , where it equals . Thus the generator of is , with .
Facts & Assumptions
For bounded Borel the operator is bounded with , , and ; the truncation definition gives for (Unbounded Borel functional calculus: domains, products, spectral mapping, The unbounded PVM integral is densely defined, closed and normal, Integral of a measurable function against a projection-valued measure).
, and for one has for real ; also and on , since commutes with and products of functions multiply (Spectral theorem for unbounded self-adjoint operators (PVM form), Unbounded Borel functional calculus: domains, products, spectral mapping).
Scalar dominated convergence and Fatou's lemma apply to the finite measures (Dominated convergence, Fatou's lemma).
The generator is defined by the difference quotients of the statement, and is a linear subspace (Infinitesimal generator of a unitary group).
Proof
Given: A self-adjoint with spectral PVM , and .
Unit and group law: is bounded, and by [A1], and is unitary because and .
Strong continuity: for and , by [A3], the integrand being bounded by and tending to pointwise.
Derivative at for : by [A3], since and is -integrable exactly because by [A2].
Converse: if for some sequence , then by [A2] and Fatou's lemma , so ; step 1.3 then identifies the full limit as .
By steps 1.1, 1.2, 1.3 and 2.1 the family is a strongly continuous one-parameter unitary group whose generator satisfies and ; the invariance and commutativity claims for on are [A2].
Laplace resolvents of a unitary group
Statement
Assume Countable Choice and Dependent Choice. Let be a strongly continuous one-parameter unitary group on with infinitesimal generator , and let . Then is a Bochner integral depending linearly and boundedly on , with and ; moreover and .
Facts & Assumptions
Strong measurability means pointwise almost-everywhere norm approximation by measurable simple functions. Such a function is Bochner integrable when its norm is integrable, and . The integral is the limit of integrals of simple approximations in integral norm (Strongly measurable Banach-valued function, Banach-valued simple function and integral, Bochner-integrable function, Bochner integrability criterion, Bochner integral norm inequality).
Bounded linear maps commute with Bochner integrals and Bochner dominated convergence holds under Countable Choice (Bounded linear maps commute with Bochner integration, Bochner dominated convergence theorem, The Axiom of Countable Choice ()).
The group law, norm preservation and strong continuity hold, and is the norm derivative at zero on its linear domain (Strongly continuous one-parameter unitary group, Infinitesimal generator of a unitary group, Convergence of a sequence in a metric space: iff in ). Cauchy--Schwarz gives continuity of the inner product (Cauchy–Schwarz: , with equality exactly for dependent pairs).
Compact Newton--Leibniz and the Riemann/Lebesgue bridge (under Countable Choice), followed by scalar monotone convergence, apply to the continuous exponential weight (Newton–Leibniz needs only continuity on , differentiability on , and a Riemann-integrable extension of the interior derivative, A bounded Riemann integrable function on a closed bounded interval is Lebesgue measurable and has the same integral, Monotone convergence for the integral).
Lebesgue measurability and measure are invariant under translation (Lebesgue outer measure, Lebesgue measurability and Lebesgue measure are unchanged by translation).
Proof
Given: Countable Choice, Dependent Choice, a strongly continuous unitary group with generator , and .
For either sign put on . It is continuous. For integer , approximate it on by its values at the left endpoints of intervals of length and put zero elsewhere. These are measurable simple functions with finite support measure, converging at every fixed to by continuity. Thus is strongly measurable. Compact Newton--Leibniz with primitive and the bridge in [A4] give . Monotone convergence as integer gives . Since , [A1] defines and gives . Linearity follows first for simple integrals and then by adding their approximating sequences in integral norm. Therefore these are bounded linear operators.
For write for real . Norm preservation and the inner-product expansion give . Since a convergent family is bounded near zero, the limit gives . Hence . If , positivity of gives ; both shifts are injective. This argument requires no density or closedness theorem for .
Translation of a Bochner integral is legitimate here: for simple integrable functions it follows termwise from [A5]; for their integral-norm limits the scalar change-of-variables identity follows first for nonnegative simple functions, then by monotone convergence, and shows that translation preserves the approximation error. Thus the simple identities pass to the Bochner integral by [A1]. Subdivision and linearity follow in the same way from simple integrals. Also as . The tail tends to , since the norm of the omitted integral is at most .
For the plus sign and , commuting with integration and translating gives by step 2.1. To obtain the required two-sided derivative, if has right quotient , then : its error is bounded by . Thus and . For , substitution in the two-sided derivative definition gives and . Applying the proved plus-sign argument to gives and .
For let . Step 3.1 places in and gives . Injectivity from step 1.2 implies , proving on . Together with step 3.1 this shows and both inverse identities with their stated domains.
From obtain and . Multiplication on the left by is legitimate on every vector because . Step 4.1 gives , hence .
The norm bound, range and inverse claims are steps 1.1, 3.1 and 4.1, and the sum identity is step 5.1. For the same formulas directly concern its unique full-domain operator; zero vectors give zero integrals. The strict condition ensures integrability and injectivity. Countable Choice supplies the compact integration bridge and the Bochner framework; the declared Dependent Choice is not additionally needed by this proof. No half-line fundamental theorem for a merely bounded derivative is invoked.
Source notes
Schnaubelt, Lemma 1.18 and Proposition 1.20(a)-(b), printed pp.11-13, supplies the translated-integral route to the resolvent. Here unitarity proves injectivity directly, so the left inverse follows from the right inverse without any half-line scalar fundamental theorem or a prior closedness theorem for the generator. Both signs and the two-sided derivative are checked explicitly.
The generator of a unitary group is closed and skew-adjoint
Statement
Assume Countable Choice and Dependent Choice. The generator of a strongly continuous one-parameter unitary group is densely defined, closed, and skew-adjoint: . Consequently is self-adjoint with .
Facts & Assumptions
Both have range and satisfy with ; also and on (Laplace resolvents of a unitary group).
and is differentiable at with derivative when , by the definition of and sesquilinearity and continuity of the inner product (Infinitesimal generator of a unitary group, Strongly continuous one-parameter unitary group, Hilbert space, Cauchy–Schwarz: , with equality exactly for dependent pairs).
A densely defined symmetric operator with is self-adjoint (Range criterion for self-adjointness, Symmetric, self-adjoint and essentially self-adjoint operators).
A bounded everywhere-defined inverse to puts in the resolvent of and forces its graph closed, with convention (Resolvent and spectrum of an unbounded operator, Densely defined, closed and closable operators, and cores). The exact limit argument is also given below.
The Bochner integral is linear by passage from simple integral approximations, and its norm is bounded by the integral of the norm (Bochner-integrable function, Bochner integral norm inequality). Compact Newton--Leibniz, the Countable Choice Riemann/Lebesgue bridge and monotone convergence compute for , , by the primitive on followed by (Newton–Leibniz needs only continuity on , differentiability on , and a Riemann-integrable extension of the interior derivative, A bounded Riemann integrable function on a closed bounded interval is Lebesgue measurable and has the same integral, Monotone convergence for the integral).
The adjoint domain consists of vectors making bounded on , with in the first-variable-linear convention (Adjoint of a densely defined operator).
Proof
Given: Countable Choice, Dependent Choice, and a strongly continuous unitary group with generator .
is dense: for and , [A1] gives and Given , strong continuity supplies such that for ; the integral norm is then at most . Letting and then proves . Taking positive integer supplies an approximating sequence in for every . Thus is dense.
is skew-symmetric: the function is constant with value for , so its derivative at vanishes, that is ; equivalently .
is closed: with , [A1] and [A4] already imply closedness. Explicitly, if , and , then since is bounded. Uniqueness of limits gives and , hence . The sequential graph criterion is valid in the norm metric under the assumed Countable Choice.
On set . For in this domain, step 1.2 gives , so is symmetric. It is densely defined by step 1.1. It is closed: convergence of and implies convergence of , so step 1.3 applies. The correct signed formulas are and . Both ranges equal by the two signs of [A1] at ; multiplication by a nonzero scalar preserves surjectivity. Hence [A3] makes self-adjoint.
The adjoint domain of equals that of : multiplication of the scalar functional in [A6] by preserves boundedness in both directions. For in this domain, ; uniqueness of the representing vector gives . Step 2.1 gives including domains, hence and .
The conclusions are density, closedness and skew-adjointness from steps 1.1, 1.3 and 3.1, and self-adjointness of from step 2.1. The zero space and constant identity group satisfy the same identities directly. The declared Countable Choice and Dependent Choice match the Laplace supplier; Countable Choice also licenses the range/adjoint and integration interfaces. Only strictly positive Laplace parameters are used.
Stone's theorem: unitary groups and self-adjoint generators
Statement
Assume the Axiom of Choice. The map , computed by the Borel functional calculus of (Unbounded Borel functional calculus: domains, products, spectral mapping), is a bijection from the set of self-adjoint operators on onto the set of strongly continuous one-parameter unitary groups on . Its inverse assigns to its generator and the self-adjoint operator ; here is exactly the set of vectors at which is differentiable at , and for .
Facts & Assumptions
For every self-adjoint , is a strongly continuous one-parameter unitary group with generator and (A self-adjoint operator generates a strongly continuous unitary group).
The generator of a strongly continuous unitary group is densely defined and skew-adjoint, so is self-adjoint with (The generator of a unitary group is closed and skew-adjoint, Infinitesimal generator of a unitary group).
If then , , and is differentiable with derivative (Laplace resolvents of a unitary group, Infinitesimal generator of a unitary group).
A skew-symmetric operator satisfies for , because . The generator in [A2] is skew-adjoint and hence skew-symmetric. The generator of a unitary group is closed and skew-adjoint Hilbert space
Proof
Given: A self-adjoint , and a strongly continuous unitary group with generator .
Applying [A1] to produces a strongly continuous unitary group with generator , so the map is well defined, and its derivative at exists exactly on where it equals .
Applying [A2] to produces the self-adjoint operator with ; applying [A1] to gives the strongly continuous unitary group , whose generator is .
Uniqueness for a fixed generator: if are strongly continuous unitary groups with the same generator and , then is differentiable with by [A3], so by [A4], and gives on ; since is dense by [A2] and are isometries, for every .
Hence in step 1.2, that is for the self-adjoint ; combined with step 1.1 this makes a bijection with the stated inverse.
The derivative characterisation is the one from [A1] applied to the self-adjoint : the limit exists exactly on and equals .
Deficiency subspaces and deficiency indices
Definition
Assume the Axiom of Choice. Let be a densely defined closed symmetric operator on a complex Hilbert space (Symmetric, self-adjoint and essentially self-adjoint operators, Densely defined, closed and closable operators, and cores). Its deficiency subspaces are and its deficiency indices are the Hilbert dimensions , that is, the cardinalities of orthonormal bases of (Orthonormal families, complete orthonormal systems and Hilbert bases).
The subspaces are the orthocomplements of the ranges. By The adjoint is well defined, closed, and reverses inclusions , so and . Each is a closed linear subspace by Orthogonal complements are closed, hence is itself a Hilbert space. For , symmetry and conjugate symmetry make real. With the inner product linear in its first variable Real and complex inner-product spaces and their induced length, Thus is injective. If , applying the identity to makes Cauchy. Completeness gives , and . Closedness of now gives and , proving both ranges closed. The closed-subspace decomposition theorem Orthogonal decomposition by a closed subspace therefore gives Also , since membership forces . The two deficiency spaces need not be orthogonal to each other in .
Dimension convention and well-definedness. An orthonormal basis exists in each by Existence of a maximal orthonormal family, and maximality as completeness, using full AC. Its cardinality is independent of the basis, as follows. Let and be two orthonormal bases of the same Hilbert space. By The Bessel inequality for an arbitrary orthonormal family, for each and integer at most indices satisfy . Thus the support in of each row is countable. Each column has a nonzero entry: otherwise is orthogonal to the dense span of all , hence to itself, contradicting norm one. Here orthogonality passes to the closure by Orthogonal complements are closed. If is infinite, full AC lets us enumerate the row supports and assign each to one row containing it; this gives an injection . Cardinal absorption Absorption: for cardinals with infinite and , , and when and cardinal comparison Commutativity, associativity, distributivity and monotonicity of and , the unit laws, the two exponent laws, and if and only if injects into give . If has finite size , the residual is orthogonal to the dense span of the and so vanishes. Taking its norm gives . For every finite , Bessel in the other direction yields Hence . Exchanging the two bases proves equality of cardinalities in all cases. In the finite case each basis also spans algebraically by the same residual argument, so this is the ordinary linear dimension. For the zero space the basis is empty and the dimension is zero. AC is used for basis existence and the simultaneous choices in the infinite comparison.
Cayley sign and domain convention. Define Injectivity of makes this well defined; the norm identity makes it an isometry onto the stated range. On its domain, , whence , since is a complex linear subspace. If a unitary extends , it maps the orthogonal complement of the initial range onto the orthogonal complement of the final range, by preservation of inner products and surjectivity. Conversely, any unitary gives the unitary extension on the two displayed orthogonal decompositions. This describes the free part of a unitary extension and fixes the signs; it does not assert that such a always exists.
Von Neumann parameterization of self-adjoint extensions
Statement
Assume the Axiom of Choice. Let be a densely defined closed symmetric operator on a complex Hilbert space H, with first-variable-linear inner product, with deficiency subspaces (Deficiency subspaces and deficiency indices), and let be a unitary operator. Then defines a self-adjoint extension of ; the sum is direct and is dense. The map is a bijection from the set of unitary operators onto the set of self-adjoint extensions of .
Facts & Assumptions
For the given closed densely defined symmetric T, and are closed, and orthogonally. The linear map is an isometric isomorphism between these ranges, and . Full AC licenses this deficiency-space interface, including its Hilbert-dimension convention. Deficiency subspaces and deficiency indices
Under Countable Choice the Cayley correspondence sends a unitary U with to the self-adjoint operator , with domain , and recovers . For self-adjoint S its Cayley transform satisfies on D(S). Cayley correspondence between self-adjoint operators and unitaries Cayley transform of a self-adjoint operator
D(T) is norm dense in H. Orthogonal decompositions have zero intersection and their squared norms add. A vector orthogonal to a dense subspace is zero: continuity of pairings follows from Cauchy-Schwarz. Densely defined, closed and closable operators, and cores Orthogonality and the orthogonal complement Cauchy–Schwarz: , with equality exactly for dependent pairs
Full AC is assumed to use [A1]. It implies the Countable Choice required by [A2] directly: AC supplies a choice function for the range family of any given sequence of nonempty sets, and composing that choice function with the sequence gives the required indexed choices. No additional family of choices is made in the construction from the supplied unitary V. The Axiom of Choice The Axiom of Countable Choice ()
Proof
Given: T as in the statement and a unitary . Put and .
Define for and . The orthogonal decomposition in [A1] makes this a uniquely defined linear map on H. Its two output terms lie in the orthogonal subspaces and , so . Since both component maps are onto their corresponding summands, U is onto H. Thus U is unitary and extends C_T; the minus sign on K_+ is necessary for the displayed plus sign in u+Vu.
For , , hence contains D(T) and is dense. If Uz=z, then for every y in H, , since a unitary preserves inner products. Consequently z is orthogonal to the dense D(T), so z=0 by [A3]. This proves .
Apply [A2], licensed by [A4], to get the self-adjoint operator on , with . Because , that domain equals ; scalar multiplication by 2i maps D(T) onto itself. To prove the sum direct, suppose . Then , and injectivity from step 2.1 gives . The two sides lie in M_+ and K_+, whose intersection is zero. Hence u=0 and x=0. Also u+Vu=(I-U)u shows the parametrization of the second summand is injective. Thus every vector has a unique representation x+u+Vu, and the domain contains the dense D(T).
On D(T), , so Sx=Tx. On the second summand, . Linearity yields . Thus S is exactly the well-defined operator T_V in the statement and is a self-adjoint extension of T.
Let R be any self-adjoint extension of T and put W=C_R. For x in D(T), , so . Thus W agrees with C_T on M_+ and maps M_+ onto M_-. For u in K_+ and m in M_+, , so Wu belongs to K_-. Conversely, for v in K_-, take the unique y with Wy=v. For every m in M_+, , hence y belongs to K_+. This proves W(K_+)=K_-, without treating W* as the Cayley transform of R. Consequently V=-W restricted to K_+ is unitary from K_+ onto K_-, and the construction of step 1.1 returns U=W. The inverse correspondence [A2] then gives T_V=R.
If T_V=T_{V'}, their Cayley transforms agree by [A2]. Step 3.1 identifies these transforms with the constructed U and U', whose restrictions to K_+ are -V and -V'. Hence V=V'. Along with steps 4.1 and 5.1, this proves the bijection. This includes empty parameter sets: if no such unitary exists, step 5.1 rules out every self-adjoint extension. If K_+=K_-={0}, the unique unitary of the zero spaces gives D(T_V)=D(T) and T_V=T, so T is already self-adjoint. If H={0}, every displayed map is its unique zero-space map and the same conclusion holds directly. No finite-dimensional or separability assumption is used. AC is used only through [A4].
Existence of self-adjoint extensions is equality of deficiency indices
Statement
Assume the Axiom of Choice. Let be a densely defined closed symmetric operator on , with deficiency indices (Deficiency subspaces and deficiency indices). Then has a self-adjoint extension if and only if . Moreover is self-adjoint if and only if , and a densely defined symmetric (not necessarily closed) operator is essentially self-adjoint if and only if , equivalently ; if , then has self-adjoint extensions.
Facts & Assumptions
Unitary operators correspond bijectively to self-adjoint extensions of ; every such unitary is onto by definition, and the Cayley transform of a self-adjoint extension restricts to a unitary (Von Neumann parameterization of self-adjoint extensions).
Every Hilbert space has a complete orthonormal family, and two Hilbert spaces are unitarily isomorphic exactly when their orthonormal bases have the same cardinality; a unitary exists exactly when (Existence of a maximal orthonormal family, and maximality as completeness, A Hilbert space with a given orthonormal basis is of the index set, Orthonormal families, complete orthonormal systems and Hilbert bases).
A closed symmetric operator is self-adjoint if and only if , equivalently (Range criterion for self-adjointness).
For a densely defined symmetric one has and ; is essentially self-adjoint exactly when is self-adjoint (Closability is equivalent to density of the adjoint domain, Symmetric, self-adjoint and essentially self-adjoint operators).
Proof
Given: A densely defined closed symmetric operator , and a densely defined symmetric operator .
If , then by [A2] there is a unitary (equal Hilbert dimensions), and [A1] produces a self-adjoint extension of . Conversely, if has a self-adjoint extension , then by [A1] its Cayley transform restricts to a unitary , so by [A2].
is self-adjoint if and only if : if both deficiency subspaces are zero then and [A3] applies; conversely a self-adjoint has and injective by the estimate , so both kernels vanish.
For symmetric, is closed and symmetric by [A4], and because ; applying steps 1.1-1.2 to gives: is essentially self-adjoint, meaning self-adjoint, if and only if ; if then has a self-adjoint extension, which is also an extension of .
All the stated equivalences are steps 1.1, 1.2 and 2.1. ∎
Pure point, absolutely continuous and singular continuous spectral subspaces
Definition
Assume the Axiom of Choice. Let be a self-adjoint operator on . If , let be its spectral projection valued measure on from Spectral theorem for unbounded self-adjoint operators (PVM form); if , let for every Borel , the unique PVM on the zero space. In either case put (Integral of a measurable function against a projection-valued measure). Call a finite Borel measure on purely atomic (equivalently, discrete) when it is concentrated on a countable subset of . Then where atoms are as in An atom of a measure on , absolute continuity is that of Absolute continuity of a signed or complex measure with respect to a positive measure and singularity that of Mutual singularity for signed or complex measures.
Well-definedness. Every finite Borel measure on has a unique decomposition into a discrete (hence, by the convention above, purely atomic), an absolutely continuous and an atomless singular part (Every finite Borel measure on R has a unique absolutely continuous, discrete, and singular-continuous decomposition), and the three classes of nonzero measures are mutually exclusive; hence each either satisfies one of the three defining conditions or none, and the zero vector lies in all three subspaces. Whether the three subspaces do cover and are closed is not part of this definition and is the content of the canonical spectral type decomposition theorem below.
Conventions. For each type one writes for the restriction of to the closed invariant subspace constructed in the canonical spectral type decomposition theorem below; these restrictions are self-adjoint there. The point spectrum is not : is the closure of the set of eigenvalues of , and the three sets may overlap, so they do not partition .
Canonical decomposition into pure point, absolutely continuous and singular continuous parts
Statement
Assume the Axiom of Choice. Let be a self-adjoint operator on a complex Hilbert space H with spectral projection valued measure on and let , , be the subspaces of Pure point, absolutely continuous and singular continuous spectral subspaces. Then are closed, mutually orthogonal, -reducing subspaces with canonically determined by ; the restrictions of to them are self-adjoint and their spectral measures are respectively purely atomic, absolutely continuous with respect to Lebesgue measure, and atomless and singular. If is separable and is a maximal scalar spectral measure with disjoint Borel supports of its discrete, absolutely continuous and singular continuous parts, then .
Here a support means a Borel carrier (zero mass off the set), not necessarily topological support. A maximal scalar spectral measure is a finite positive Borel measure mu with for every x (in particular a scalar spectral measure with this domination property qualifies). The last assertion is conditional on the supplied mu. On the zero Hilbert space use the unique PVM and full-domain zero operator directly.
Facts & Assumptions
The three types are defined by the scalar measures : discrete means concentrated on a countable set, absolutely continuous means vanishing on Lebesgue-null Borel sets, and singular continuous means atomless and carried by a Lebesgue-null Borel set. Pure point, absolutely continuous and singular continuous spectral subspaces Absolute continuity of a signed or complex measure with respect to a positive measure Mutual singularity for signed or complex measures An atom of a measure on
Each finite positive Borel measure on the line has a unique decomposition into discrete, absolutely continuous and atomless singular measures. Its atoms form a countable set. Lebesgue measure gives zero mass to a singleton, hence to a countable set by countable subadditivity. Every finite Borel measure on R has a unique absolutely continuous, discrete, and singular-continuous decomposition Every finite Borel measure on splits as an atomic part plus an atomless part A box in with parameters is Lebesgue measurable of measure , whichever of its faces are included
E is a regular PVM, its projection values commute and satisfy E(B)E(C)=E(B intersect C), and each is contractive and self-adjoint. Orthogonality uses the first-linear inner product. Projection valued measure Orthogonality and the orthogonal complement Spectral theorem for unbounded self-adjoint operators (PVM form)
and . Bounded integrals are operator-norm limits of integrals of uniform simple approximations. Conversely the integral of the real coordinate against a regular PVM is self-adjoint on this domain. Integral of a measurable function against a projection-valued measure Bounded borel pvm integral Spectral theorem for unbounded self-adjoint operators (PVM form)
AC supplies the measure-decomposition and spectral interfaces and directly supplies the countable choices of null carriers and all dependent or countable witness choices used below. The Axiom of Choice
Proof
Given: AC, a complex Hilbert space H, and self-adjoint T with its regular spectral PVM E; use the direct zero-space convention when needed.
For any finite measure nu, take its decomposition [A2]. Let P be the countable set of atoms, carrying the discrete part, and choose a Lebesgue-null Borel carrier N for the singular continuous part. The three disjoint Borel sets , , partition the line and carry their corresponding parts: both atomless parts vanish on P, the absolutely continuous part vanishes on P union N, and the discrete part vanishes off P. Thus for each type r. More generally, for any disjoint carriers S_r of the three components, their union carries nu and the same restriction identity holds. If eta<<nu is finite positive, its restrictions to these carriers are of the respective types: they are dominated in the sense of null sets by nu_r, so inherit a countable carrier, Lebesgue absolute continuity, or a null carrier and zero singleton masses. They sum to eta, so uniqueness in [A2] implies eta has type r exactly when eta is carried by S_r. The zero measure has all three types, consistently with this assertion.
For every Borel B, [A3] gives and . Consequently each type set is linear: two countable carriers have countable union, two null carriers have null union, and zero masses on null sets or singletons pass through this inequality. If x_n of one type converge to x, the contraction inequality shows E_x(B)=0 whenever all E_{x_n}(B)=0. For the discrete case choose countable carriers P_n and use their countable union P; then E_x(P^c)=0. For the singular case choose null Borel carriers N_n and use their null union N. For the atomless condition apply the same argument to each singleton; for absolute continuity apply it to each fixed Lebesgue-null Borel set. Hence all three subspaces are closed. The countable carrier choices and countable unions use the declared AC [A5].
Measures of different types are mutually singular: a discrete carrier is countable and both other types give it zero mass; a singular-continuous null carrier has zero absolutely continuous mass. Thus for x,y of different types there is a Borel S carrying E_x with E_y(S)=0. Since , one has E(S)x=x, while E(S)y=0. Self-adjointness of E(S) yields .
For arbitrary x, apply step 1.1 to E_x and set x_r=E(S_r)x. The PVM identities give , so x_r belongs to the indicated type. The partition gives . By step 2.1 this decomposition is orthogonal and unique. The component maps Q_r are linear by uniqueness, contractive by the Pythagorean identity for this finite orthogonal sum, and self-adjoint because . They are orthogonal projections onto the closed type subspaces. As these subspaces were defined from E_x, they and the projections are canonical, independent of the carriers chosen for individual vectors.
For each Borel B the equality shows that E(B) preserves each type. Applying it to the unique decomposition in step 3.1 yields Q_r E(B)=E(B)Q_r. Therefore , so x in D(T) implies Q_rx in D(T) by [A4]. Commutation with every E(B) gives commutation with every simple integral and then every bounded integral by [A4]. Passing to the coordinate truncation limit gives TQ_rx=Q_rTx for x in D(T). Thus each type subspace reduces T, with its domain carried along.
In the separable clause let the supplied finite maximal measure mu have disjoint carriers B_r of its three components. For every x, E_x<<mu, so step 1.1 says x has type r exactly when E_x(B_r^c)=0. The latter is equivalent to E(B_r)x=x, since E(B_r^c)=I-E(B_r) and . Hence H_r=ran E(B_r). This proves the assertion for every supplied maximal mu, not just for a specially constructed one, and needs no circle-to-line transport.
On a nonzero type subspace K, E_K(B)=E(B)|_K is a regular PVM: its projection and strong countable-additivity properties restrict from E, and its scalar measures are the same regular E_x for x in K. The coordinate integral against E_K has domain K intersect D(T); bounded simple integrals and their limits agree with the restrictions of those for E, so its value is Tx. The converse spectral theorem in [A4] makes this restriction self-adjoint. On K={0}, self-adjointness is direct since its unique densely defined operator equals its adjoint. The scalar measures of each restriction have exactly the specified type by [A1].
If H={0}, every scalar measure is zero, the three subspaces are {0}, and all conclusions including the carrier formula hold directly. Vanishing components on a nonzero H also give zero subspaces by the same arguments, and no measure is divided by its mass. Nonseparability causes no difficulty in the preceding arguments, since only a single scalar measure or a sequence of vectors is used at a time. Full AC is used exactly as in [A5], including countable carrier choices for closedness.
Source notes
Teschl, Section 3.3, Lemma 3.18, printed pp.118–119, gives the canonical type spaces and their spectral projections from maximal-measure carriers. The direct scalar-measure argument here proves the decomposition without a separability assumption; the maximal-measure carrier formula is asserted conditionally as in the statement. No change-of-variables or Cayley transport is needed.
Relative boundedness with respect to an operator
Definition
Let be a linear operator on with domain . An operator is -bounded, or relatively bounded with respect to , when and there are finite constants with The infimum of the admissible constants is the -bound of , and one says the -bound is below one when some is admissible.
Equivalent form. is -bounded exactly when is bounded on for the graph norm of Unbounded linear operators: domain, graph and extension: each estimate gives , and conversely a graph-norm bound gives the estimate with . The constant is not intrinsic, and no closedness, density or resolvent hypothesis is needed for the definition. If is closed with nonempty resolvent set, then -bounded operators are exactly those with for which is bounded for some, equivalently every, (Resolvent and spectrum of an unbounded operator). Indeed, an -bound makes bounded because ; conversely, if is bounded, then gives an -bound. Thus boundedness for one resolvent implies relative boundedness and hence boundedness for every resolvent. This is used in the Kato-Rellich theorem below and recorded here as an interface.
Second resolvent identity for a closed perturbation
Statement
Assume Dependent Choice. Let and be closed operators with , put , and assume is bounded for the graph norm of (Relative boundedness with respect to an operator). Then for every and every product here is defined on all of and is bounded.
Facts & Assumptions
For the operator maps bijectively onto and on , since (Resolvent and spectrum of an unbounded operator).
is bounded for the graph norm of : there are with for ; each is therefore everywhere defined and bounded, since and both terms are bounded in (Relative boundedness with respect to an operator, [A1]).
and on , so and , because has range (Resolvent and spectrum of an unbounded operator).
The graph norms of two closed operators with the same domain are equivalent: both domains are Banach, and the identity map from the -graph norm to the -graph norm has closed graph, hence is bounded by the closed graph theorem (Closed graph theorem, The axiom of dependent choice: a relation in which every element is related to something admits an -indexed chain).
Proof
Given: Closed with , graph-norm bounded for , and .
The operator is bounded by [A1] and [A2]. By [A4] the -graph norm is bounded by a constant times the -graph norm, so the same relative bound makes bounded for the -graph norm; applying [A1] with in place of shows that is bounded as well.
, that is : by [A3], , and expanding gives , because is minus the identity on and takes values in .
By the same computation with the roles of and interchanged (so that the perturbation is ), , that is .
Rearranging steps 2.1 and 2.2 gives and , which is the stated identity; all products are bounded by step 1.1. ∎
Kato-Rellich theorem
Statement
Assume the Axiom of Choice. Work on a complex Hilbert space with inner product linear in the first variable. Let be self-adjoint and let be symmetric with and -bound less than one (Relative boundedness with respect to an operator). Then with domain is self-adjoint. If is merely essentially self-adjoint and is symmetric with and -bound less than one, then on is essentially self-adjoint and its closure is the self-adjoint operator obtained by applying the first part to and the graph-norm extension of to . If and is an admissible pair for with , then .
Facts & Assumptions
Relative boundedness supplies finite a,b>=0 with a<1 and on D(A). The graph norm is ; graph closure defines the operator closure. Relative boundedness with respect to an operator Unbounded linear operators: domain, graph and extension Densely defined, closed and closable operators, and cores
Symmetry is the identity on the domain; self-adjoint operators are closed and densely defined. A densely defined symmetric S is self-adjoint if both ranges of S plus and minus i mu equal H for some mu>0. Nonreal points are resolvent points of a self-adjoint operator. The convention is R_T(z)=(z-T)^{-1}. Symmetric, self-adjoint and essentially self-adjoint operators Range criterion for self-adjointness Resolvent of a self-adjoint operator: nonreal resolvents and the estimate Resolvent and spectrum of an unbounded operator
If a bounded operator C has norm less than one, I-C has a bounded inverse given by the Neumann series. Neumann series
On a nonzero H, a self-adjoint T has a regular PVM E, domain , and , . Bounded integrals satisfy the norm bound and quadratic identity. Projections multiply by intersection; strong countable additivity holds. The Borel calculus has the product rule with domain D(g(T)) intersect D((fg)(T)), and the spectrum of T is the set of v for which every neighborhood has nonzero projection (apply its essential-range assertion to f(v)=v). Spectral theorem for unbounded self-adjoint operators (PVM form) The unbounded PVM integral is densely defined, closed and normal Bounded borel pvm integral Projection valued measure Unbounded Borel functional calculus: domains, products, spectral mapping
AC supplies the spectral theorem's choices and directly supplies the countable witness choices used by the range/adjoint interfaces and by sequences approximating a fixed point in a graph closure. The Axiom of Choice
Proof
Given: the operators and admissible pair (a,b) in the statement, with 0<=a<1 and b>=0.
If H={0}, all domains and graphs are zero and all conclusions hold directly. Otherwise use [A4]. For z=plus or minus i mu, mu>0, put r_z(v)=(z-v)^{-1}. The bounds |r_z(v)|<=1/mu and |v r_z(v)|<=1 show that r_z(A) maps H into D(A), that , and that . The product rule gives (z-A)r_z(A)=I on H and r_z(A)(z-A)=I on D(A), so this is R_A(z). Thus when, for example, .
For any nonzero-space self-adjoint T and real c, the spectral-measure equivalence if and only if follows directly. If T>=c and has a nonzero projection, a nonzero x in its range belongs to D(T), has E_x carried by J_n, and satisfies , a contradiction. These increasing sets exhaust (-infinity,c), so countable additivity gives zero projection. Conversely zero projection below c gives for every x in D(T). Also if every real t<c is a resolvent point, the essential-range characterization in [A4] supplies a zero-projection open neighborhood of each t. The rational intervals contained in such neighborhoods form a countable cover of (-infinity,c); their union has zero projection by countable subadditivity of each E_x. Therefore again T>=c. This does not require a finite spectral infimum or any resolvent-distance formula.
Put S=A+B on exactly D(A). For z=plus or minus i mu the identity holds on D(A), since R_A(z)(z-A)x=x there. The first factor is boundedly invertible by [A3], and z-A is bijective D(A) to H. Hence both nonreal shifts of S are onto. S is symmetric by summing the two symmetry identities and densely defined because D(A) is dense. The range criterion makes S self-adjoint. No second-resolvent identity with a previously closed S is assumed.
Suppose A>=gamma and let lambda+gamma=d>0. By step 1.2 E_A is carried by [gamma,infinity). Define r(v)=(-lambda-v)^{-1} on that half-line and zero outside. For v>=gamma, and : for v>=0 the ratio v/(v+lambda) is monotone with its maximum at an endpoint or its limiting value 1; for gamma<=v<0, (-v)/(v+lambda) decreases with v since lambda>0 in that case. The spectral product rule shows r(A)=R_A(-lambda), exactly as in step 1.1, and hence . Put . For every d>C this last bound is strictly below one: if d>=|gamma| it equals a+b/d<1 (also when b=0); if d<|gamma| it equals (a|gamma|+b)/d<1. The factorization of step 2.1 therefore proves that every real number t<gamma-C is in rho(S), with bounded inverse . Apply step 1.2 to the already self-adjoint S to obtain S>=gamma-C, the stated bound including its endpoint.
If A is essentially self-adjoint, put T=closure(A). For each x in D(T) choose x_n in D(A) with x_n to x and Ax_n to Tx, using [A5]. The inequality in [A1] applied to x_n-x_m makes Bx_n Cauchy. Define Btilde x as its limit. Two such approximations give the same limit by the same inequality applied to their difference. Approximating x and y and their linear combinations proves linearity, , and symmetry by passing to the limit in . It extends B restricted to D(A); no extension of B's possibly larger domain is claimed. Step 2.1 makes T+Btilde self-adjoint. The inclusion A+B subset T+Btilde and closedness give closure(A+B) subset T+Btilde. Conversely the same approximating sequences satisfy (A+B)x_n to (T+Btilde)x, giving the reverse graph inclusion. This proves essential self-adjointness and the exact closure formula. If A>=gamma in this case, taking limits of its quadratic inequality gives T>=gamma; step 3.1 then gives the bound for the closure and its restriction A+B.
The choice use is exactly [A5]. The cases B=0 or a=b=0 are admitted by the same estimates and return the original lower bound. Dimension one requires no change. The strict hypothesis a<1 is used in the positive choice of mu and in b/(1-a); no conclusion at a=1 is asserted. The endpoint gamma-C is included by the zero-projection argument, without asserting that the spectrum is nonempty on the zero space.
Source notes
Teschl, Section 6.1, Lemma 6.3 and Theorem 6.4, printed pp.158–159 (PDF pp.169–170), supply the imaginary and real resolvent perturbation method. Signs here are computed for the library convention (z-A)^{-1}. The numerical bound is derived above directly from equation (6.3); the proof does not rely on a spectral-distance claim or endpoint continuity of a concave function.
Discrete and essential spectrum of a self-adjoint operator
Definition
Assume the Axiom of Choice. Let be a self-adjoint operator on a complex Hilbert space with spectral projection valued measure on (Spectral theorem for unbounded self-adjoint operators (PVM form)). The discrete spectrum is the set of eigenvalues of that are isolated points of and whose eigenspace is finite dimensional; the essential spectrum is with as in Resolvent and spectrum of an unbounded operator.
For use its unique PVM directly: the spectrum and both parts are empty and every projection has rank zero. Below suppose , as required by the cited spectral theorem.
Spectral-projection description, with proofs. Fix and write for .
Here rank means the algebraic dimension of the range when finite; rank means the range is not finite dimensional. The calculus Unbounded Borel functional calculus: domains, products, spectral mapping gives the support facts: , and every open interval about a spectral point has nonzero projection. Projections on disjoint sets have orthogonal ranges, and Projection valued measure.
- is the projection onto . If , its scalar measure is supported on by the projection identity. Thus , so , and The domain and norm identities are supplied by the unbounded calculus and The unbounded PVM integral is densely defined, closed and normal. Hence . Conversely, for an eigenvector (or the zero vector) the same norm identity gives zero integral. On this bounds the measure by times that zero integral; taking the countable union shows . Since equals this scalar measure, .
- A finite-rank interval contains only finitely many spectral points. If has rank and its interval contained distinct spectral points, choose disjoint small open intervals about those finitely many points, all contained in the given interval. Each has a nonzero projection by support. Choose one unit vector in each range. They are orthonormal vectors in , hence linearly independent (take inner products with each vector), contradicting its dimension . Thus there are at most spectral points there. In particular any with such a finite-rank is isolated: take a smaller interval around excluding the other finitely many points. For that interval the projection is by support, is nonzero by support, and has finite rank since its range lies in . By item 1, is an eigenvalue of finite multiplicity.
- The two rank characterizations. If , an isolating interval has projection , of finite rank by item 1. Conversely, if is an eigenvalue and some has finite rank, item 2 proves it discrete. Hence If , item 2 excludes every finite-rank interval. Conversely, if every has infinite rank, each is nonzero, so support puts in , and the just-proved discrete characterization excludes it from . Therefore
- is closed. If and , then for every some has , so and by item 3.
The closure conclusion is in , and also in since is closed there. A spectral accumulation point has infinitely many spectral points in every surrounding interval, so item 2 forces infinite rank. An isolated eigenvalue of infinite multiplicity has its infinite-dimensional eigenspace inside every interval range by item 1. Consequently an accumulation point of and an isolated eigenvalue of infinite multiplicity both lie in , and is the disjoint union of and .
Weyl criterion for the essential spectrum
Statement
Assume the Axiom of Choice. Let be a self-adjoint operator on a complex Hilbert space H and let . Then if and only if there is a sequence with , weakly (Weak convergence of nets and sequences) and . The sequence may be chosen orthonormal; such a sequence is called a singular Weyl sequence for .
Sequences below are indexed by all n in N={0,1,2,...}; the shrinking radii are 1/(n+1). Inner products are linear in the first variable.
Facts & Assumptions
For every real lambda, membership in the essential spectrum is equivalent to infinite rank of every projection , epsilon>0. This equivalence includes real resolvent points and isolated finite-multiplicity eigenvalues. On the zero Hilbert space the essential spectrum is empty. Discrete and essential spectrum of a self-adjoint operator
On nonzero H the spectral theorem gives the domain . Projections multiply by intersection and . The integral norm identity and Borel sum rule give on D(A). Spectral theorem for unbounded self-adjoint operators (PVM form) Projection valued measure The unbounded PVM integral is densely defined, closed and normal Unbounded Borel functional calculus: domains, products, spectral mapping
Weak convergence means convergence against every bounded linear functional. Under Countable Choice every such functional on a Hilbert space is ; these pairings are bounded by Cauchy-Schwarz. Bessel bounds the sum of squared coefficients against an orthonormal family by the squared norm. Weak convergence of nets and sequences Riesz representation for Hilbert spaces Cauchy–Schwarz: , with equality exactly for dependent pairs The Bessel inequality for an arbitrary orthonormal family Orthonormal families, complete orthonormal systems and Hilbert bases
The declared AC supplies the spectral theorem and directly chooses a successor for every extendible finite orthonormal list; iterating that fixed choice function from the empty list gives the required sequence. The Axiom of Choice The axiom of dependent choice: a relation in which every element is related to something admits an -indexed chain
Proof
Given: the self-adjoint A, real lambda and the hypotheses of the direction under consideration.
If H={0}, there is no unit vector and the essential spectrum is empty, so both sides are false. Otherwise use the PVM of [A2]. For any bounded interval J and x in ran E(J), the projection identity gives E_x(R\J)=0 and , hence x belongs to D(A). For any x in D(A) and epsilon>0, [A2] gives . The closed complement includes both interval endpoints.
A finite-dimensional range of an orthogonal projection P has a finite orthonormal basis e_1,...,e_r: from a finite linear basis, successively subtract its projections onto the previously obtained vectors and normalize the nonzero residuals (nonzero follows from linear independence). Then , since its difference from that sum lies in the range and is orthogonal to its basis. If x_n is weakly null, [A3] gives each coefficient tending to zero and . For rank zero this is the empty sum and Px=0.
Every orthonormal sequence is weakly null: Bessel gives for each y. If infinitely many coefficients had modulus at least epsilon>0, finite partial sums of arbitrarily many such terms would exceed this bound. Thus their moduli tend to zero; conjugate symmetry gives , and Riesz gives convergence against every bounded linear functional.
Suppose a singular Weyl sequence exists. If any P_epsilon had finite rank, step 1.2 would give P_epsilon x_n to zero, while step 1.1 and the residual hypothesis would give (I-P_epsilon)x_n to zero. The triangle inequality would contradict norm x_n=1. Thus every P_epsilon has infinite rank, and [A1] proves lambda belongs to the essential spectrum. This proves the implication for all real lambda, including exclusion of real resolvent points, rather than merely excluding the discrete spectrum.
Suppose lambda is in the essential spectrum. For n>=0 write V_n=ran E((lambda-1/(n+1),lambda+1/(n+1))), infinite dimensional by [A1]. Given a finite list of n previously chosen orthonormal vectors x_0,...,x_(n-1), there is a nonzero vector in V_n orthogonal to them: choose n+1 linearly independent vectors in V_n and solve the n homogeneous linear equations for their pairings with the preceding vectors; a nonzero coefficient solution exists by finite-dimensional elimination, and independence makes its vector nonzero. Normalize it. For n=0 choose any nonzero vector in V_0 and normalize; there are no orthogonality equations. On the set of finite lists meeting these conditions, the relation of adjoining such a vector is entire. Apply DC from [A4] with the empty list as initial point; the compatible lists define x_n for every n>=0. The sequence is orthonormal and lies in D(A) by step 1.1. Its scalar measure is carried by the stated interval, so . Step 1.3 gives weak nullity. This constructs the required sequence and proves the converse.
The two implications are steps 2.1 and 2.2. Finite-dimensional H (including dimension one) has only finite-rank interval projections, so neither side holds there. Infinite multiplicity at an isolated point and spectral accumulation points are both covered by the same infinite-rank construction. Lambda=0 is allowed, since only the positive radii n+1 and epsilon are inverted. The empty initial list and the index-zero vector are included in step 2.2; the Choice use is [A4] and the spectral/Riesz interfaces, with no separability assumption.
Source notes
Teschl, Lemma 6.17, printed pp.170–171 (PDF pp.181–182), gives the singular Weyl criterion and its complete projection-estimate proof. The present proof uses the infinite-rank characterization directly in both directions and a DC construction on shrinking interval ranges, with the library's zero-based indexing.
Relative compactness with respect to an operator
Definition
Assume the Axiom of Choice. Let be a self-adjoint operator on a complex Hilbert space , and let be linear and bounded for the graph norm of (Relative boundedness with respect to an operator). Then is -compact, or relatively compact with respect to , when is a compact operator (Compact linear operator) for one, equivalently for every, .
The resolvent set is nonempty: by Resolvent of a self-adjoint operator: nonreal resolvents and the estimate. If , all operators here are the unique operator, compact with bound zero; the assertions hold directly. Below suppose .
Well-definedness, with proofs.
-
is everywhere defined and bounded. maps into and is graph-norm bounded there, so using , so that (Resolvent and spectrum of an unbounded operator).
-
Independence of . Write . For , and . Applying gives since is the identity on . Thus . Interchanging also gives . This derives both orders without an unproved resolvent identity; the two sides of the latter identity have values in , so applying the linear map gives Compactness at implies compactness at by composition with the bounded and finite linear combinations Compositions with a compact operator are compact Linear combinations of compact operators are compact. Exchanging proves the converse, including the trivial case .
-
Vector space. If are graph-norm bounded and -compact, then is graph-norm bounded and is compact, being a linear combination of compact operators (Linear combinations of compact operators are compact).
-
An -compact has -bound zero. For and , the inverse identity gives . Hence, with , It suffices to prove along positive integers . The spectral theorem Spectral theorem for unbounded self-adjoint operators (PVM form) and product/domain rule Unbounded Borel functional calculus: domains, products, spectral mapping identify with the bounded function of : multiplication by gives the two inverse identities, with range in since both and are bounded. Consequently is the bounded function of . For real and , The bounded PVM calculus and its adjoint rule Bounded borel pvm integral Pvm integral is a star homomorphism give . For every both squared norms and equal , which tends to zero by Dominated convergence, dominated by in the finite measure of mass . In particular both families converge strongly to zero.
Put , compact by item 2. The inverse identity on gives . Suppose its norm does not tend to zero. There exist , a strictly increasing integer subsequence , and, using the declared AC, vectors with and . For any , Riesz representation Riesz representation for Hilbert spaces therefore proves . A compact operator sends a weakly null sequence to a norm-null sequence under AC Compact operator sends weakly convergent sequences to norm convergent sequences, contradicting the displayed lower bound. Thus . Given any , choose with in the first estimate: is finite and the coefficient is below . Its infimum is therefore zero. This does not assert that the zero coefficient itself is attained. The full AC assumption covers the spectral theorem and the compactness/sequence argument.
Weyl's theorem: invariance of the essential spectrum
Statement
Assume the Axiom of Choice. Let be self-adjoint operators such that is compact for one nonreal (equivalently, for every ). Then . In particular a bounded self-adjoint compact perturbation preserves the essential spectrum, and if is symmetric and -compact, then with .
Facts & Assumptions
exactly when has an orthonormal singular Weyl sequence at (Weyl criterion for the essential spectrum, Discrete and essential spectrum of a self-adjoint operator).
For and one has on , equivalently (Resolvent and spectrum of an unbounded operator).
A compact operator maps weakly convergent sequences to norm convergent sequences, and is bounded (Compact operator sends weakly convergent sequences to norm convergent sequences, Compact linear operator).
For self-adjoint and nonreal , the bounded resolvent identity gives , where and the inverse first factor is . Hence compactness of transfers to , and conversely by exchanging (The resolvent star algebra is dense in C_0(R), Compositions with a compact operator are compact).
An -compact symmetric has -bound zero, so Kato-Rellich makes self-adjoint on ; the second resolvent identity holds for in the common resolvent set (Relative compactness with respect to an operator, Kato-Rellich theorem, Second resolvent identity for a closed perturbation).
Proof
Given: Self-adjoint with compact resolvent difference at a nonreal .
Let and let be the orthonormal Weyl sequence of [A1]. By [A2] and one has ; since is compact and , [A3] gives as well.
Parameter independence is [A4].
Then by [A2] and step 1.1, and ; the normalized vectors lie in , have unit norm, converge weakly to and satisfy , so they form a singular Weyl sequence and by [A1]. Interchanging the roles of and gives equality.
Bounded compact perturbations: if is bounded, symmetric and compact, then is self-adjoint with by Kato-Rellich applied with the admissible pair , and the second resolvent identity gives , compact as a product of the compact with bounded factors; so by step 2.1.
-compact perturbations: for symmetric that is -compact, [A5] makes self-adjoint with and gives , a product of the bounded operator with the compact operator , hence compact; then step 2.1 applies.
The claims are steps 1.1, 1.2 and 2.1 (compact resolvent difference), 3.1 (bounded compact perturbations) and 3.2 (-compact perturbations). ∎
Norm and strong resolvent convergence
Definition
Assume Countable Choice (The Axiom of Countable Choice ()). Let and be self-adjoint operators on the same Hilbert space and fix a nonreal . One writes in the norm resolvent sense when in operator norm, and in the strong resolvent sense when that is, strong operator convergence (The operator norm as the least bound and as the unit-sphere or unit-ball supremum, A bounded linear operator between normed spaces, Weak convergence of nets and sequences).
Both notions are well posed for every choice of nonreal : by Resolvent of a self-adjoint operator: nonreal resolvents and the estimate each nonreal number belongs to , so all resolvents occurring are bounded with . The definition deliberately does not assert independence of the parameter : that independence is a theorem, proved for the norm case by the resolvent-star-algebra density lemma below and used in the continuous-calculus-under-resolvent-convergence theorem below. Norm resolvent convergence implies strong resolvent convergence, and both are notions about the resolvents rather than about the operators: no convergence of the operators themselves is asserted or implied.
The resolvent star algebra is dense in C_0(R)
Statement
Fix and let . Let be the -algebra generated by in , that is, the uniform closure of the linear span of the products with . Then . Under Countable Choice (The Axiom of Countable Choice ()), consequently, if self-adjoint satisfy strongly (respectively in norm) at one nonreal , then the same convergence holds at every nonreal , in particular at .
Facts & Assumptions
is a self-adjoint algebra of continuous functions vanishing at infinity, is injective on , and does not vanish at any point of ; the one-point compactification is a compact Hausdorff space and (The one-point (Alexandroff) compactification , whose open sets are the open sets of together with the complements in of the closed compact subsets of , is compact and contains as an open subspace; is dense in exactly when is not compact; and is Hausdorff exactly when is locally compact and Hausdorff, Locally compact topological space: every point has a compact neighbourhood; and what this says in a metric space, Continuity of a map of topological spaces at a point and globally).
Let be a compact Hausdorff space and let be a self-adjoint complex function algebra containing the constants, separating points, with no common zero. Then is uniformly dense in (Complex Stone–Weierstrass dichotomy for separating self-adjoint algebras; the unital case is dense).
Under Countable Choice, with for self-adjoint and nonreal , the resolvent identity gives for nonreal , the first factor being and the second ; these affine transforms of resolvents are bounded with norms at most (Resolvent and spectrum of an unbounded operator, Resolvent of a self-adjoint operator: nonreal resolvents and the estimate, Norm and strong resolvent convergence, The Axiom of Countable Choice ()).
Proof
Given: , and the -algebra generated by .
Extend to by ; then and separates the points of : it is injective on , and .
Parameter independence: for nonreal the difference factors as in [A3] through , with both outer factors of norm at most , since they are and and ; hence in norm whenever in norm, and for every whenever for every , because bounded operators preserve both modes of convergence.
The algebra contains the constants and , is self-adjoint, separates points by step 1.1, and has no common zero because of the constant function ; hence is uniformly dense in by [A2].
Therefore : given and , density of gives with and ; evaluating at , where , gives , so and .
The density claim is step 3.1 and the consequence is step 1.2, which also covers . ∎
Continuous functional calculus under resolvent convergence
Statement
Assume the Axiom of Choice. Let be self-adjoint operators on a complex Hilbert space and suppose in the strong resolvent sense. Then for every bounded continuous and every . If in the norm resolvent sense, then for every bounded continuous with . In both cases the conclusion does not depend on the nonreal parameter used in the definition of convergence.
Facts & Assumptions
On nonzero complex , AC supplies the spectral PVM of each self-adjoint , representing as the integral of the identity on its squared-integrability domain. The unbounded calculus has the exact product domain , and sums and products agree with their pointwise counterparts on their domains (Spectral theorem for unbounded self-adjoint operators (PVM form), Unbounded Borel functional calculus: domains, products, spectral mapping).
The bounded PVM calculus is linear, unital, multiplicative and conjugation preserving, satisfies and , and . Its Countable Choice assumptions are supplied by AC (Bounded borel pvm integral, Pvm integral is a star homomorphism, Scalar and complex measures from a pvm, The Axiom of Choice).
exists for nonreal with . Convergence is initially assumed at one fixed nonreal parameter only (Resolvent of a self-adjoint operator: nonreal resolvents and the estimate, Resolvent and spectrum of an unbounded operator, Norm and strong resolvent convergence).
For fixed nonreal , the linear span of , , , is uniformly dense in (The resolvent star algebra is dense in C_0(R)). Only its function-algebra density assertion is used; parameter independence is proved below.
Scalar dominated convergence holds with an integrable majorant (Dominated convergence). Operator norm bounds give and, by applying this twice, (A bounded linear operator between normed spaces, The operator norm as the least bound and as the unit-sphere or unit-ball supremum).
Proof
Given: AC, the self-adjoint operators and convergence at a fixed nonreal in the indicated mode.
If , all operators, resolvents and bounded functions of them are the unique full-domain operator, so all assertions hold. Suppose . For nonreal put . Both and are bounded on the real line. The product-domain rule in [A1] shows that maps into and on , while on . Thus with the convention of [A3]. The bounded calculus agrees with its unbounded truncation definition because the truncations are eventually the same bounded function.
Put . The scalar identity and [A2] give these identities for each single operator . Define and , which is independent of . Expanding the product using the preceding identities yields : its two terms are and , whose mixed terms cancel. The left factor has norm at most . Consequently in the strong case, since is fixed, and in the norm case. This proves parameter independence without taking a strong limit on a varying vector.
In particular convergence holds at and , and , , with the analogous formulas for . Both sequences are uniformly bounded. If and strongly and , then . In the norm case the same inequality with operator norms proves convergence of products when the factors are uniformly bounded. Iteration and finite linear combinations, using [A2], therefore prove convergence for every polynomial in with zero constant term.
Given and , choose one such polynomial with by [A4]. The bounded calculus gives . For fixed the last term tends to zero by step 3.1; since is arbitrary, strong convergence follows for each (including directly). The operator-norm inequality is , proving the norm version as well.
In the norm case, if has a common finite limit at both ends, then . Unital linearity gives , so step 4.1 proves the assertion.
For the strong case let be any bounded continuous function and set . For integers define . These are continuous, compactly supported, between zero and one, equal to one on , and converge pointwise to one. Thus by dominated convergence with majorant , integrable against the finite measure of mass . For fixed , both and belong to , so their calculi converge strongly by step 4.1.
Bounded multiplicativity and linearity give the exact four-term decomposition . Applying it to , its norm is at most . If the claim is immediate. Otherwise choose so that the first term is less than half a prescribed positive error, using step 5.2, and then so that the other two together are less than its other half. This proves strong convergence for bounded continuous .
Step 2.1 proves independence of every nonreal parameter, step 5.1 the norm conclusion and step 6.1 the strong conclusion. AC supplies the spectral and countable-choice calculus hypotheses; the cutoffs are explicit. The zero Hilbert space and zero function have been treated in steps 1.1 and 6.1; no limit at either infinity is required in the strong case.
Source notes
Teschl, Theorem 6.31 and its proof, pp.179-180, gives the polynomial approximation and four-term cutoff route, with Corollary 6.32 giving parameter independence. Here the latter is proved first using a resolvent-difference factorization with a fixed right factor. The library convention is , so . No general continuity of adjoints for strongly convergent bounded operators is assumed.
Unitary groups converge under strong resolvent convergence
Statement
Assume the Axiom of Choice. Let be self-adjoint operators on a complex Hilbert space with in the strong resolvent sense. Then If, in addition, and all are bounded below by a common real constant (meaning for and ), then strongly for every .
Facts & Assumptions
For a self-adjoint with spectral PVM, is the Borel calculus of and is a strongly continuous unitary group, with infinitesimal generator (A self-adjoint operator generates a strongly continuous unitary group, Infinitesimal generator of a unitary group, Strongly continuous one-parameter unitary group).
Under AC, strong resolvent convergence gives for every bounded continuous and every (Continuous functional calculus under resolvent convergence, The Axiom of Choice).
On nonzero , each self-adjoint has a spectral PVM with and . For domain vectors, , using the quadratic pairing identity and the reality of in the first-variable-linear convention. Functions agreeing off a measurable -null set have the same integral operator and domain; the zero-space calculus is defined directly (Spectral theorem for unbounded self-adjoint operators (PVM form), The unbounded PVM integral is densely defined, closed and normal, Integral of a measurable function against a projection-valued measure).
PVM projections satisfy and ; the scalar measures are positive of mass , and scalar monotone convergence holds (Projection valued measure, Scalar and complex measures from a pvm, Monotone convergence for the integral).
Proof
Given: AC and the self-adjoint operators with strong resolvent convergence in the statement.
If , all operators in either conclusion are its unique operator, so both conclusions hold. Otherwise the spectral PVMs exist by [A3]. Fix any , including negative times. The function is continuous and has absolute value one everywhere. Thus [A2] gives for every ; [A1] identifies these as the stated unitary groups. At each operator is .
For the second claim suppose is one of and satisfies the common lower bound. Let for integers , with an empty interval interpreted as empty. If , choose for some . The projection identities give , so the scalar measure of is supported on . As is bounded, [A3] gives and , a contradiction. Hence each . The sets increase to ; monotone convergence gives for every . Since , the projection itself is zero.
Now fix and define . It is continuous and bounded by , and it agrees with on . Step 1.2 and null-set invariance in [A3] give equality of the integral operators , including domains, for . In particular each exponential here has full domain and is bounded: its defining squared integral is at most , and its quadratic norm identity gives the same operator bound. Applying [A2] to proves for every .
The first conclusion holds for every real time by step 1.1, and the second for every nonnegative time by step 2.1. At time zero both reduce to the identity. The lower bound is assumed for the limit and every approximant; no preservation-of-lower-bound theorem is assumed. AC supplies the spectral and calculus hypotheses, including their Countable Choice assumptions.
Spectral form domain and core of a semibounded operator
Statement
Assume the Axiom of Choice. Let be self-adjoint on a complex Hilbert space H, with meaning for every , for some real . Put Then and do not depend on the choice of the constant (the form is itself unchanged, while the summand changes by the constant when is replaced by ), for , the domain is dense in for the norm , and is a closed quadratic form with .
Here the square root is the Borel calculus of ; the proof shows that E is carried on . A closed semibounded quadratic form means the diagonal of a Hermitian sesquilinear form on a dense linear domain, complete in the displayed shifted form norm. If H={0}, use the unique PVM and operator and the convention inf(empty spectrum)=+infinity.
Facts & Assumptions
The spectral theorem gives and A as the coordinate integral on nonzero H. The unbounded integral is closed, has linear domain, squared norm integral, and real-function pairing . Real f gives a self-adjoint operator. The zero-space integral is defined directly. Spectral theorem for unbounded self-adjoint operators (PVM form) The unbounded PVM integral is densely defined, closed and normal Integral of a measurable function against a projection-valued measure
Projections multiply by intersection, and is a finite measure of mass . Consequently , using and the norm formula; complementary projections give the analogous complementary restriction. The spectrum of A is the essential range of the coordinate function. Projection valued measure Unbounded Borel functional calculus: domains, products, spectral mapping
Scalar dominated convergence applies to the finite measures E_x. Dominated convergence
H is complete and its inner product is first-linear. Self-adjoint operators have dense linear domains. The assumed AC directly supplies every choice function required by the PVM, closed-integral and spectral-theorem interfaces. Hilbert space Symmetric, self-adjoint and essentially self-adjoint operators The Axiom of Choice
Proof
Given: AC, self-adjoint A and its lower bound c.
If H={0}, every domain and form consists of zero, all norms vanish, and every assertion follows directly from the zero-space convention in [A1]. Otherwise obtain E from [A1], under the choice assumption in [A4]. For (empty intervals allowed), a vector belongs to D(A) by [A2] and boundedness of lambda on J_m. If v were nonzero then , contradicting the lower bound. Thus E(J_m)=0 for all positive integers m. Their union is , whose scalar measures therefore vanish by countable subadditivity. The projection norm formula gives E((-infinity,c))=0. The essential-range description in [A2] implies .
Put . It is closed and self-adjoint by [A1], and step 1.1 gives and . Integrals here and below can be restricted to [c,infinity). Since there, lambda is absolutely integrable for x in Q(A). Hence and .
For any other lower spectral bound c'<=c, on the carrier. Since E_x has finite mass, the two domain integrals are finite simultaneously. Adding the appropriate constant times the mass gives the same q_A, while the square-root squared norm increases by . Two arbitrary admissible lower bounds can be compared in their numerical order, so this proves full independence. Their squared form norms differ by that same multiple of , hence are equivalent since each dominates .
If x belongs to D(A), then on the carrier, so x belongs to Q(A). By [A1] and step 2.1, .
For x in Q(A), set , n>=1. By [A2], , so x_n belongs to D(A). The same restriction identity gives by dominated convergence, with nonnegative integrable majorant on the carrier. This is the asserted form-norm density, with the exact identity .
The form is the diagonal of on the linear domain D(B); this is Hermitian and sesquilinear by [A4]. Its domain is dense in H because it contains D(A) by step 3.2. For a Cauchy sequence in the form norm, both x_n and Bx_n are Cauchy in H. Completeness gives limits x and y. Closedness of B implies x in D(B) and Bx=y. Therefore , proving completeness and closedness in the stated sense.
Steps 2.1 and 3.1 establish the domain, integral identity, lower bound and independence; steps 3.2 and 3.3 give the operator-domain identity and core, and step 4.1 gives the closed quadratic form. Positive integer cutoffs are specified without choices. AC is inherited through [A4]; negative and zero lower bounds are allowed without taking a square root of q_A itself. The zero Hilbert space was handled in step 1.1.
Min-max principle below the essential spectrum
Statement
Assume the Axiom of Choice. Let A be a self-adjoint operator on a complex Hilbert space H, bounded below by c, with form domain Q(A) and form q_A of Spectral form domain and core of a semibounded operator. Put , with inf(empty)=+infinity. List the eigenvalues below Lambda in nondecreasing order with multiplicity as E_1,E_2,..., setting E_n=Lambda after the list is exhausted. Then for every integer n>=1, All infima over empty sets are +infinity. Finite dimensions here are ordinary linear dimensions, equivalently Hilbert dimensions for these finite-dimensional subspaces. The second outer family is always nonempty, since it contains {0}. Whenever Q(A) contains an (n-1)-dimensional subspace, the same value is obtained by requiring dim F=n-1. The at-most convention includes the exhausted finite-dimensional case for all n without taking a supremum over an empty outer family.
The first infimum is unchanged if its trial spaces L are restricted to D(A). If E_n<Lambda, both outer values are attained, respectively by spans of the first n and the first n-1 orthonormal eigenvectors (the latter span is {0} when n=1).
Facts & Assumptions
The essential spectrum is closed, and a real v is in it exactly when every interval about v has infinite-rank spectral projection. A finite-rank interval contains only finitely many spectral points, and each such point is an isolated finite-multiplicity eigenvalue. The singleton projection is the eigenspace projection. The spectrum carries E. Discrete and essential spectrum of a self-adjoint operator
The spectral PVM has intersection products, orthogonal disjoint ranges and strong countable additivity. The spectral domain is . For a vector in the range of a bounded interval projection its scalar measure is carried by that interval, so it belongs to D(A). The calculus identifies spectral support with spectrum. Projection valued measure Spectral theorem for unbounded self-adjoint operators (PVM form) Unbounded Borel functional calculus: domains, products, spectral mapping
is finite on Q(A), , and E is carried on [c,infinity). The form domain is a dense linear subspace containing D(A); its form is Hermitian with on D(A). Self-adjoint operators are symmetric. Spectral form domain and core of a semibounded operator Symmetric, self-adjoint and essentially self-adjoint operators
Orthogonality uses the first-variable-linear inner product; a finite orthonormal family is linearly independent and the squared norm of its linear combination is the sum of squared coefficient moduli. Pairings are continuous by Cauchy-Schwarz. A closed bounded real interval has the finite-open-subcover property. Orthogonality and the orthogonal complement Orthonormal families, complete orthonormal systems and Hilbert bases Cauchy–Schwarz: , with equality exactly for dependent pairs Heine-Borel by bisection: every closed bounded interval is compact
AC is declared for the spectral and form interfaces and, if the eigenvalue list is infinite, for choosing orthonormal bases in its countably many finite-dimensional eigenspaces. All variational subspace and kernel arguments below are finite-dimensional. The Axiom of Choice
Proof
Given: A>=c, its spectral measure E, and the variational quantities in the statement.
Suppose first H is nonzero. For any real r<Lambda with r>=c, each point of [c,r] has a finite-rank interval neighborhood by [A1]. Use the family of all such intervals and the finite subcover property [A4]. Each chosen interval contains finitely many spectral points, so sigma(A) intersect [c,r] is finite. These points are discrete eigenvalues of finite multiplicity by [A1]. Since E is carried by sigma(A) intersect [c,infinity), is the finite sum of their singleton projections. For r<c it is zero. In particular the total multiplicity below each r<Lambda is finite. The eigenvalues below Lambda can therefore be ordered from below: if any remain, choose one such u; the nonempty finite set of remaining values <=u has a smallest member, which is the smallest remaining value altogether. Repeat with its finite multiplicity. No value can be omitted forever, since only finitely many terms lie below any fixed u<Lambda. A countable increasing sequence of bounds r approaching Lambda (or infinity) covers all values; [A5] licenses the associated orthonormal eigenvector choices.
Write alpha_n for the first outer value over Q(A), beta_n for the second (dim F<=n-1), and alpha_n^D for the first over D(A). If dim L=n and dim F=r<=n-1, choose finite bases and impose the r equations on x in L. Finite-dimensional elimination gives a nonzero solution, since there are fewer equations than unknowns. Normalize it to obtain a unit vector in L intersect F-perp. Thus the supremum on L is at least the infimum on Q(A) intersect F-perp. For each F take the infimum over L, then the supremum over F, giving beta_n<=alpha_n; this remains true when there is no L because alpha_n=+infinity. The trial-space inclusion gives alpha_n<=alpha_n^D.
If Lambda is finite, it belongs to the essential spectrum: that set is nonempty in this case, bounded below by c, and closed; for every positive integer m choose a point between its infimum and Lambda+1/m and use closedness. Hence every interval around Lambda has infinite-rank projection by [A1]. If Lambda=+infinity and only k eigenvalues with multiplicity occur, there is no other spectrum: every finite spectral point lies below Lambda and is one of these eigenvalues. Support then makes their eigenspaces sum to all of H, so dim H=k and D(A)=Q(A)=H. The same last assertion holds directly with k=0 for H={0}.
If E_n<Lambda, choose orthonormal eigenvectors phi_1,...,phi_n ordered with multiplicity as in step 1.1. Distinct eigenspaces are orthogonal because symmetry gives ; within each eigenspace choose an orthonormal basis by finite Gram-Schmidt. For , contained in D(A), , so its unit-sphere supremum is E_n. For , all eigenspaces strictly below E_n are contained in F_0. The projection E((-infinity,E_n)) therefore annihilates every x in F_0-perp; repetitions of E_n need not be removed. Its scalar measure is carried on [E_n,infinity), so q_A[x]>=E_n for unit x in Q(A) intersect F_0-perp. Equality holds at phi_n. Consequently E_n<=beta_n<=alpha_n<=alpha_n^D<=E_n by step 1.2. This proves the formulas, the D(A) version and both attainments in this case.
If E_n=Lambda is finite, exactly k<n eigenvalues occur below Lambda with multiplicity. Let F_0 be the span of all their orthonormal eigenvectors (zero if k=0). It is an admissible space of dimension k<=n-1. By support, vectors perpendicular to it have no spectral mass below Lambda, so the inner infimum is at least Lambda and beta_n>=Lambda. For any epsilon>0, the range of E((Lambda-epsilon,Lambda+epsilon)) is infinite dimensional by step 2.1. Choose n independent vectors there and use finite Gram-Schmidt to get an n-dimensional subspace L_epsilon in the same range. Bounded spectral support puts it in D(A) and gives q_A[x]<=Lambda+epsilon on its unit sphere. Hence alpha_n^D<=Lambda+epsilon. Let epsilon decrease to zero in the inequalities of step 1.2 to conclude beta_n=alpha_n=alpha_n^D=Lambda. No form-cross-term estimate or Weyl-sequence approximation is needed.
If E_n=Lambda=+infinity, step 2.1 gives dim H=k<n, D(A)=Q(A)=H, with k the exhausted total multiplicity. There is no n-dimensional trial space, so alpha_n=alpha_n^D=+infinity. F=H has dimension k<=n-1 and is admissible for beta_n. Its orthogonal complement contains no unit vector, so its inner infimum is +infinity and beta_n=+infinity. This includes the zero Hilbert space for every n>=1.
Finally suppose Q(A) has a subspace of dimension n-1. Any finite-dimensional F contained in Q(A) with dim F<n-1 can be enlarged inside Q(A) to dimension n-1: as long as its dimension is smaller, choose a vector from the given (n-1)-dimensional subspace not in the current span and adjoin it. Enlarging F shrinks Q(A) intersect F-perp, so its inner infimum cannot decrease. Taking suprema shows that restricting the second outer family to exact dimension n-1 leaves beta_n unchanged; the reverse inequality is family inclusion. When n=1 the only space is F={0}. When H is finite dimensional and n exceeds dim H+1, the at-most convention remains necessary. The real lower bound c was never shifted away, so negative eigenvalues and threshold zero require no special argument. AC is used exactly in [A5] and the infimum-approaching sequence in step 2.1.
Source notes
Teschl, Section 4.4, printed pp.139–141 (PDF pp.150–152), equations (4.37)–(4.41) and Theorem 4.12, gives the max-min argument using n-1 trial vectors, whose span can have smaller dimension. Theorem 4.14's min-max proof is assigned as Problem 4.11 and its printed trial count is inconsistent with (4.43). The full projection proof above supplies both formulas directly, with n-dimensional min-max spaces and explicit finite-dimensional exhaustion conventions.
5 · Examples, counterexamples and false statements
None yet.