1 Definition
The minimal polynomial is a fundamental invariant that records the simplest algebraic relation satisfied by an element, matrix, or linear operator. In the setting of field theory, it identifies the least-degree monic polynomial over a chosen base field that vanishes at a given algebraic element. In linear algebra, it is the least-degree monic polynomial that sends an operator or matrix to the zero transformation or zero matrix.
1.1 Minimal polynomial of an algebraic element
Let \(\alpha\) be an algebraic element over a field \(F\). Its minimal polynomial over \(F\) is the unique monic polynomial \(m_{\alpha,F}(x) \in F[x]\) of smallest positive degree such that \(m_{\alpha,F}(\alpha)=0\). This polynomial captures the exact algebraic dependence of \(\alpha\) on \(F\). Every polynomial in \(F[x]\) that has \(\alpha\) as a root must be divisible by \(m_{\alpha,F}(x)\).
1.2 Minimal polynomial of a matrix or linear operator
For a square matrix \(A\) over a field \(F\), the minimal polynomial is the monic polynomial \(m_A(x)\in F[x]\) of least degree such that \(m_A(A)=0\). The same definition applies to a linear operator \(T\) on a finite-dimensional vector space: \(m_T(x)\) is the least monic polynomial with \(m_T(T)=0\). This polynomial encodes essential structural information about the operator, including its eigenvalues and the sizes of its Jordan blocks.
1.3 Uniqueness and monic normalization
The minimal polynomial is uniquely determined once it is required to be monic. Without this normalization, any nonzero scalar multiple of a minimal polynomial would also satisfy the defining annihilation property. Requiring the leading coefficient to be 1 removes this ambiguity and makes the invariant canonical.
1.4 Existence criteria
For an algebraic element \(\alpha\), existence follows from the fact that some nonzero polynomial in \(F[x]\) vanishes at \(\alpha\). Among all such polynomials, one can choose one of least degree and then make it monic. For a matrix or linear operator, existence is guaranteed in finite dimensions because the operator acts on a finite-dimensional space and satisfies a nonzero polynomial relation, for example by the Cayley-Hamilton theorem.
2 Basic properties
The minimal polynomial is characterized by strong divisibility and factorization properties. It is the smallest annihilating polynomial, but it also controls all other polynomial relations satisfied by the same element or operator. In many cases, it reflects the underlying field in an essential way.
2.1 Divisibility properties
If \(f(x)\in F[x]\) satisfies \(f(\alpha)=0\), then the minimal polynomial \(m_{\alpha,F}(x)\) divides \(f(x)\). Likewise, if \(f(A)=0\) for a matrix or linear operator \(A\), then \(m_A(x)\mid f(x)\). This makes the minimal polynomial the universal divisor among annihilating polynomials.
2.2 Irreducibility for algebraic elements
The minimal polynomial of an algebraic element is irreducible over the base field. If it factored nontrivially, one of the factors would also vanish at the element, contradicting minimality. This property is central in field theory, where minimal polynomials provide a direct link between algebraic elements and irreducible polynomials.
2.3 Relation to annihilating polynomials
Every annihilating polynomial is a multiple of the minimal polynomial. Consequently, the set of all polynomial relations satisfied by an element or operator forms an ideal in \(F[x]\), generated by the minimal polynomial in the algebraic-element setting or by the operator’s minimal polynomial in the linear-algebra setting.
2.4 Dependence on the base field
The minimal polynomial depends on the chosen field \(F\). A polynomial irreducible over one field may factor over another, changing the minimal polynomial accordingly. As a result, the same element or matrix can have different minimal polynomials when considered over different base fields.
3 Computation
Computing a minimal polynomial often proceeds by identifying a polynomial relation of sufficiently low degree and then verifying that no smaller monic polynomial works. In practice, matrix computations and algebraic manipulations use structural information to reduce the search.
3.1 Computing from algebraic relations
For algebraic elements, one typically begins with a known polynomial relation and factors it over the base field. The minimal polynomial is the unique irreducible factor that vanishes at the element. If the element is expressed using radicals, roots, or other algebraic constructions, elimination methods may be used to obtain a polynomial relation first.
3.2 Computing from matrices
For a matrix \(A\), one may exploit powers of \(A\), eigenvalue data, or canonical forms. The goal is to determine the smallest monic polynomial \(m(x)\) such that \(m(A)=0\). Since the minimal polynomial divides the characteristic polynomial, its degree is bounded above by the matrix size.
3.2.1 Using characteristic polynomials
The characteristic polynomial provides an upper bound for the minimal polynomial. One can factor the characteristic polynomial and determine the least powers of its irreducible factors needed to annihilate the matrix. This is often done by testing successive divisors until the smallest annihilating polynomial is found.
3.2.2 Using invariant factors
If the matrix is brought into rational canonical form, the invariant factors determine the minimal polynomial directly. Specifically, the minimal polynomial is the largest invariant factor. This method is especially effective over arbitrary fields, where diagonalization or Jordan form may not be available.
3.3 Examples of explicit calculations
A matrix that is already diagonal has minimal polynomial equal to the product of distinct linear factors corresponding to its distinct eigenvalues. A nilpotent matrix has minimal polynomial \(x^k\), where \(k\) is the smallest positive integer with \(A^k=0\). For an algebraic number such as \(\sqrt{2}\), the minimal polynomial over \(\mathbb{Q}\) is \(x^2-2\).
4 Relation to field theory
Minimal polynomials are among the most important tools in the study of field extensions. They describe how adjoining a single algebraic element enlarges a field and provide a bridge between polynomial factorization and extension structure.
4.1 Minimal polynomial in field extensions
If \(\alpha\) is algebraic over \(F\), then \(F(\alpha)\) is a simple extension generated by \(\alpha\). The minimal polynomial of \(\alpha\) determines the algebraic relations in this extension and gives a concrete description of elements of \(F(\alpha)\) as polynomials in \(\alpha\) of bounded degree.
4.2 Degree and algebraic extensions
The degree of the minimal polynomial of \(\alpha\) equals the dimension of \(F(\alpha)\) as a vector space over \(F\). Thus, the degree of the extension \(F(\alpha)/F\) is the degree of the minimal polynomial. This equality is a key result in the theory of algebraic extensions.
4.3 Conjugates and Galois theory
The roots of the minimal polynomial are the conjugates of the algebraic element over the base field, at least within an algebraic closure. These conjugates are permuted by field automorphisms that fix the base field. In Galois theory, the minimal polynomial therefore provides a starting point for understanding the symmetries of algebraic elements.
4.4 Splitting fields
The splitting field of the minimal polynomial is the smallest field extension over which the polynomial decomposes completely into linear factors. This field often plays a central role in constructing normal extensions and in studying the Galois group associated with the polynomial.
5 Relation to linear algebra
In linear algebra, the minimal polynomial acts as a concise summary of the operator’s structure. It controls which polynomial expressions in the operator vanish and strongly influences canonical forms, eigenvalue behavior, and decomposition into invariant subspaces.
5.1 Cayley-Hamilton theorem
The Cayley-Hamilton theorem states that every square matrix satisfies its own characteristic polynomial. This guarantees the existence of a nonzero annihilating polynomial and implies that the minimal polynomial divides the characteristic polynomial. It is one of the key links between determinant-based invariants and operator theory.
5.2 Minimal polynomial and characteristic polynomial
The characteristic polynomial and minimal polynomial are related but generally not equal. The minimal polynomial divides the characteristic polynomial and often has lower degree. The characteristic polynomial records eigenvalue multiplicities, while the minimal polynomial records the largest sizes of the Jordan blocks associated with each eigenvalue.
5.3 Diagonalizability criteria
A linear operator is diagonalizable over a field if and only if its minimal polynomial splits into distinct linear factors, with no repeated roots. In that case, the operator satisfies a polynomial with no repeated irreducible factors. Repeated factors indicate the presence of nontrivial nilpotent behavior within the operator’s structure.
5.4 Jordan canonical form
Over an algebraically closed field, the minimal polynomial is determined by the sizes of the largest Jordan blocks for each eigenvalue. For each eigenvalue, the exponent of the corresponding linear factor in the minimal polynomial equals the size of the largest block attached to that eigenvalue. This makes the minimal polynomial a compact summary of the Jordan form.
5.5 Eigenspaces and generalized eigenspaces
The minimal polynomial helps describe the decomposition of a vector space into generalized eigenspaces. Each distinct eigenvalue contributes a generalized eigenspace on which a power of \(T-\lambda I\) acts trivially. The exponents in the minimal polynomial measure how far the operator is from being semisimple on each such component.
6 Advanced topics
Beyond basic computation and classification, minimal polynomials appear in more refined structural settings. They interact with canonical matrix constructions, module theory, and comparisons between different base fields.
6.1 Minimal polynomials over different fields
Changing the base field can alter both irreducibility and factorization patterns. An operator that has a simple minimal polynomial over one field may have a more complicated decomposition over a larger field, or vice versa. This dependence is especially visible when moving from \(\mathbb{Q}\) to \(\mathbb{R}\) or \(\mathbb{C}\).
6.2 Companion matrices
Every monic polynomial can be realized as the minimal polynomial of its companion matrix. Companion matrices provide a concrete bridge between polynomial data and linear transformations. They are useful in constructing examples and in proving existence results for matrices with prescribed invariants.
6.3 Rational canonical form
In rational canonical form, a matrix is decomposed into blocks associated with invariant factors. The minimal polynomial is the least common multiple of the elementary divisors and equals the largest invariant factor in the decomposition. This perspective is particularly effective over fields where the matrix is not necessarily diagonalizable.
6.4 Minimal polynomial of an endomorphism on a module
More generally, an endomorphism of a finitely generated module over a principal ideal domain has a minimal polynomial in suitable settings. The idea extends from vector spaces to module structures, where polynomial relations describe module decomposition and torsion behavior. This broader context unifies linear algebra with module-theoretic classification.
7 Examples
Examples illustrate how the minimal polynomial reflects the simplest nontrivial relation satisfied by an element or transformation. They also show how the same notion adapts across algebraic and linear settings.
7.1 Algebraic numbers
The number \(\sqrt{2}\) has minimal polynomial \(x^2-2\) over \(\mathbb{Q}\). The number \(\sqrt{2}+\sqrt{3}\) has a higher-degree minimal polynomial over \(\mathbb{Q}\), obtained by eliminating the radicals. In each case, the polynomial is irreducible over the base field and uniquely identifies the algebraic dependence.
7.2 Polynomial matrices
A matrix such as \[ A=\begin{pmatrix} 0 & 1\\ 0 & 0 \end{pmatrix} \] satisfies \(A^2=0\) but \(A\neq 0\), so its minimal polynomial is \(x^2\). By contrast, a diagonal matrix with distinct diagonal entries has minimal polynomial equal to the product of the corresponding distinct linear factors.
7.3 Nilpotent and idempotent operators
If \(N\) is nilpotent, then its minimal polynomial is a power of \(x\), with the exponent equal to the nilpotency index. If \(E\) is idempotent, meaning \(E^2=E\), then its minimal polynomial divides \(x(x-1)\). When \(E\) is neither the zero nor identity operator, the minimal polynomial is typically \(x(x-1)\).
7.4 Rotations and other standard linear transformations
A planar rotation by angle \(\theta\) has minimal polynomial determined by its eigenvalue behavior over the chosen field. Over \(\mathbb{R}\), a nontrivial rotation often has minimal polynomial \(x^2-2(\cos\theta)x+1\), provided it is not \(\pm I\). More generally, standard linear transformations are often classified by their minimal polynomials together with their invariant factors.