An adjoint operator is an operator associated with a linear map that transfers its action from one argument of an inner product to the other. For an operator between Hilbert spaces, its adjoint is characterized by
In finite-dimensional spaces with orthonormal coordinates, this is exactly the operation of taking the conjugate transpose of a matrix. In infinite-dimensional spaces, the existence and domain of the adjoint require additional care, especially for unbounded operators. Adjoints are fundamental in functional analysis, differential equations, optimization, and quantum mechanics.
Definition for bounded operators
Let and be real or complex Hilbert spaces, and let be a bounded linear operator. There is a unique bounded linear operator satisfying
Throughout this article, complex inner products are taken to be linear in the first argument and conjugate-linear in the second. The alternative convention is also common; the defining adjoint identity remains the same when used consistently.
Existence and uniqueness follow from the Riesz representation theorem: for each fixed , the map
is a continuous linear functional on , so it has a unique representation as for some . Setting defines the adjoint.
The notations and, less commonly, are also used. The dagger notation is particularly common in physics, although in numerical mathematics it can instead denote a pseudoinverse.
The adjoint depends on the chosen inner products. It is therefore not determined by the underlying linear map alone unless those inner products are specified.
Matrix representation
Suppose and are finite-dimensional, and represents in orthonormal bases. Then is represented by
Thus the adjoint is the transpose over the real numbers, and the conjugate transpose over the complex numbers.
For example,
The orthonormality assumption matters. If coordinate inner products are specified by positive-definite Hermitian Gram matrices and , so that
then the coordinate matrix of the adjoint is
Consequently, taking a conjugate transpose without accounting for the inner products can give the wrong adjoint in nonorthonormal coordinates.
The adjoint operator should not be confused with the adjugate matrix, the transpose of the cofactor matrix used in determinant and inverse formulas. Older terminology sometimes calls the adjugate the “classical adjoint.”
Algebraic and geometric properties
For bounded operators with compatible domains and codomains,
Thus taking adjoints reverses the order of composition and is conjugate-linear in the operator. If has a bounded inverse, then
For the operator norm,
Moreover,
The operator is therefore self-adjoint and positive.
Adjoints also express the relationship between an operator’s kernel and range. Writing for the range,
where denotes the orthogonal complement. Equivalently,
The closures cannot generally be omitted in infinite dimensions. For example, has dense range exactly when is injective.
Important classes of operators
Several major classes of operators are defined through their adjoints.
- A self-adjoint operator satisfies .
- A skew-adjoint operator satisfies .
- A normal operator satisfies .
- A unitary operator satisfies and .
- An isometry satisfies , but need not be onto.
- An orthogonal projection satisfies .
For real Hilbert spaces, surjective inner-product-preserving operators are usually called orthogonal rather than unitary.
These identities connect adjoints to the spectral theorem. In finite-dimensional complex inner-product spaces, normal operators admit an orthonormal basis of eigenvectors. Self-adjoint operators have real eigenvalues, while unitary operators have eigenvalues of absolute value one.
For a general matrix , the singular value decomposition is governed by the positive operators and . Their nonzero eigenvalues agree, and their square roots are the nonzero singular values of .
Examples
Multiplication operators
On the space , with
let , where is an essentially bounded measurable function. Then
In particular, is self-adjoint exactly when is real-valued almost everywhere.
Integral operators
Suppose and are finite measure spaces and . The integral operator
defines a bounded map from to . Its adjoint is
Thus taking an adjoint exchanges the input and output variables and conjugates the kernel, paralleling the conjugate transpose of a matrix.
The unilateral shift
On the Hilbert space , define
Then
Here , but : the latter removes the zeroth coordinate. This illustrates why an isometry need not be unitary in infinite dimensions.
Unbounded operators and their domains
Many operators arising from differential equations are unbounded. Such an operator is specified not merely by a rule but also by a linear domain:
Assume is dense in . The domain of the adjoint is
Density makes unique, and is defined to be this vector.
Equivalently, precisely when
is bounded on with respect to the norm of .
Unlike the bounded case, need not be all of , or even dense. The adjoint is always a closed operator. A densely defined operator is closable exactly when is dense; in that case,
where denotes its closure.
For unbounded operators, equality includes equality of domains. Merely showing
establishes that is symmetric, meaning . Self-adjointness requires the stronger statement
This distinction is essential in infinite-dimensional spectral theory.
Differentiation and boundary conditions
Consider
on , with
the Sobolev space of functions whose endpoint traces vanish. By integration by parts,
The boundary term vanishes for every , yielding
The adjoint therefore has a different domain from the original operator. If instead is defined on all of , its adjoint is on .
A formal adjoint is the differential expression obtained by integration by parts before domains and boundary conditions are fully specified. It need not be the complete Hilbert-space adjoint: the latter includes its precise domain.
Relation to dual-space adjoints
There is a related construction that does not require an inner product. For a bounded linear map between Banach spaces, its transpose, also called its Banach-space adjoint, is
where and are the continuous dual spaces.
For complex spaces, this transpose is linear in :
whereas the Hilbert-space adjoint satisfies
The two constructions are connected by the Riesz identifications. Under the inner-product convention used here, define the conjugate-linear maps
Then
The conjugate-linearity of these identifications explains why the Hilbert-space adjoint introduces complex conjugation.
Applications
Least squares and optimization
For a bounded operator and data , consider the least-squares objective
Its gradient is
In a complex Hilbert space, this statement uses the underlying real inner product .
Any minimizer satisfies the normal equation
Conversely, every solution of this equation minimizes . In infinite dimensions, a minimizer need not exist if the range of is not closed.
Adjoint-state methods extend the same principle to optimization constrained by partial differential equations. They transfer derivatives through the governing equations, allowing gradients of scalar objectives to be computed without separately differentiating the full state with respect to every parameter.
Quantum mechanics
In quantum mechanics, observables are represented by self-adjoint operators, often unbounded ones. Domain conditions are part of their mathematical specification: a formally symmetric differential expression alone does not determine a self-adjoint observable. Unitary evolution also uses the adjoint through .
Numerical computation
Algorithms for large-scale inverse problems often require only the actions and , rather than an explicitly stored matrix. An implementation can be checked using the adjoint identity
on test vectors, allowing for floating-point error. If discretized spaces use weighted inner products, those weights must be included; an ordinary matrix transpose may not represent the intended adjoint.