Algebraic Extension¶

A field extension in which every element of the larger field satisfies a nonzero polynomial with coefficients in the embedded base field.

Version: v1 · 2026-10-03 · History
Domain-specific #: 12976
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomain: Field Theory → Mathematics
Aliases: Algebraic Field Extension

Core Idea¶

An algebraic extension \(L/K\) is a field extension for which every \(\alpha\in L\) is algebraic over the embedded base field \(K\): some nonzero polynomial \(f_\alpha(x)\in K[x]\) satisfies \(f_\alpha(\alpha)=0\). The witness may differ from element to element. Among the polynomials that vanish at an algebraic element there is a unique monic irreducible minimal polynomial, but an arbitrary nonzero witness need not be minimal.^[1]

The all-elements quantifier is the classification. A finite-dimensional field extension is necessarily algebraic, since enough powers of any one element become linearly dependent over \(K\). The reverse implication fails: an infinite tower can contain only algebraic elements while having infinite total degree over its base. Thus “algebraic” means polynomial dependence relative to a base, not “finite,” “closed,” or “solvable by radicals.”^[1]

Structural Signature¶

Sig role-phrases:

Embedded base field \(K\): supplies coefficients and fixes the comparison; algebraicity is relative to this base.^[1]
Extension field \(L\): contains a specified copy of \(K\) and supplies all elements to test.^[1]
Universal polynomial witness condition: for every \(\alpha\in L\), a nonzero \(K\)-coefficient polynomial vanishes at \(\alpha\). One transcendental element is enough to make \(L/K\) nonalgebraic.^[1]

A finite \(K\)-basis is a sufficient way to prove the condition, not a necessary structural role. Neither a single common polynomial nor a single generator is required in the general case.

What It Is Not¶

A bare field extension only embeds \(K\) in \(L\); adjoining an indeterminate \(t\) gives a counterexample \(K(t)/K\), because \(t\) is transcendental. The degree \([L:K]\) measures vector-space dimension and may be infinite even for an algebraic extension. An algebraic closure is an algebraic extension with the additional root-completeness requirement that it is algebraically closed; a typical quadratic number field does not have that property.^[1]

Nor does algebraic automatically mean normal or separable. Those are separate field-theoretic refinements. A polynomial with coefficients in \(L\), instead of \(K\), does not by itself establish algebraicity over \(K\).

Scope of Application¶

Over characteristic zero, \(\mathbb Q(\sqrt2)/\mathbb Q\) is a simple quadratic extension: \(\sqrt2\) has minimal polynomial \(x^2-2\), and \(\{1,\sqrt2\}\) is a \(\mathbb Q\)-basis. Thus any \(a+b\sqrt2\) is algebraic over \(\mathbb Q\), even if its own minimal polynomial differs from \(x^2-2\). The finite basis establishes the universal condition, not merely the generator's one witness.^[1]

Over characteristic \(p\), \(\mathbb F_{p^n}/\mathbb F_p\) is another algebraic extension. A field of order \(q=p^n\) has degree \(n\) over its prime subfield, and every element \(a\) satisfies \(a^q=a\). The common polynomial \(x^q-x\in\mathbb F_p[x]\) therefore witnesses algebraicity of every element, though it is not generally each element's minimal polynomial. This finite-field route differs from the quadratic irrational example in carrier, characteristic, and witness structure.^[2]

Clarity¶

The quantifiers matter: for every \(\alpha\in L\), there exists a nonzero \(f_\alpha\in K[x]\) with \(f_\alpha(\alpha)=0\). Reversing these quantifiers to require one common nonzero polynomial for all \(\alpha\) would wrongly exclude many infinite algebraic extensions. Checking only a single \(\alpha\) would wrongly admit a larger field containing another transcendental element.^[1]

For \(L=\mathbb Q(\sqrt2)\), finite dimensionality makes the check short. For any \(\alpha\in L\), the vectors \(1,\alpha,\alpha^2\) lie in a two-dimensional \(\mathbb Q\)-space and are linearly dependent; their dependence coefficients form a nonzero polynomial over \(\mathbb Q\). By contrast, \(t\) in \(\mathbb Q(t)\) is an indeterminate, so no nonzero polynomial in \(\mathbb Q[x]\) vanishes at it.

Manages Complexity¶

The identity lets field theory distinguish extensions built entirely from polynomial roots from those containing free transcendental parameters. For a finite extension, one can use vector-space dimension to certify every element at once rather than write infinitely many explicit polynomials. In a finite field, \(x^q-x\) gives a single convenient witness for the entire carrier.^[1]^[2]

The distinction remains useful in towers. If \(K\subseteq E\subseteq L\), then \(L/K\) is algebraic exactly when both \(E/K\) and \(L/E\) are algebraic. But one must check the base at each step; “root of a polynomial somewhere” is not enough.^[1]

Abstract Reasoning¶

The form is field embedding plus universal element-wise polynomial dependence. The extension relation comes first; the algebraic condition is an added predicate on that relation. For a simple extension \(K(\alpha)\), algebraicity of \(\alpha\) makes \(K(\alpha)=K[\alpha]\) and yields finite degree equal to the degree of the minimal polynomial. That useful consequence is not a definition of all algebraic extensions; an algebraic extension need not be simple or finite.^[1]

An infinite counterexample to the converse of “finite implies algebraic” is the field of all complex numbers algebraic over \(\mathbb Q\). It contains roots of arbitrarily high polynomial degree, so its total degree is infinite, while each of its elements retains a finite polynomial witness over \(\mathbb Q\).^[1]

Knowledge Transfer¶

The number-field and finite-field cases transfer the same questions: Which base field supplies coefficients? What is the larger field? How do we prove every element has a polynomial witness? The answer may be a basis/dependence argument, a common finite-field polynomial, or a union of finite subextensions. The test does not transfer numerical degrees or characteristic-specific formulas unchanged.^[1]^[2]

The relative base also matters. An element can be algebraic over one field and transcendental over another; the phrase “algebraic element” is incomplete without saying over which field.

Examples¶

Quadratic number field. Mapped back: base = \(\mathbb Q\); extension = \(\mathbb Q(\sqrt2)\); witness for generator = \(x^2-2\); universal check = the basis \(\{1,\sqrt2\}\) gives finite \(\mathbb Q\)-dimension, so each \(a+b\sqrt2\) has some nonzero rational-coefficient polynomial. The nearby \(\mathbb Q(t)/\mathbb Q\) fails because \(t\) is transcendental.^[1]

Finite field. Mapped back: base = \(\mathbb F_p\); extension = \(\mathbb F_{p^n}\); universal check = every \(a\) obeys \(a^{p^n}=a\); witness = \(x^{p^n}-x\). For a concrete \(n=2,p=2\), \(\mathbb F_4\) can be represented by adjoining a root of the irreducible \(x^2+x+1\) over \(\mathbb F_2\); each of its four elements also obeys \(x^4-x=0\). The common polynomial is not asserted to be every element's minimal polynomial.^[2]

Structural Tensions¶

One generator versus every element. A visible generator's minimal polynomial may prove the case for a simple extension, but an arbitrary larger \(L\) may contain other elements. Diagnostic: Does the demonstrated generator actually generate all of the claimed extension?^[1]

Finite degree versus infinite algebraic extent. Finite degree is an easy sufficient check; infinite total degree is compatible with element-wise algebraicity. Diagnostic: Is the claim about the dimension of \(L/K\), or about a polynomial witness for each member of \(L\)?^[1]

Relative dependence versus root completeness. Every element already in \(L\) may be algebraic over \(K\), yet additional polynomial roots can still be missing from \(L\). Diagnostic: Is the argument proving algebraicity or the stronger algebraically-closed condition?^[1]

Structural–Framed Character¶

Evaluative weight. “Algebraic” is a yes/no mathematical property relative to a specified base field, not praise for an extension's usefulness or simplicity. Its status can change when the base changes.^[1]

Human-practice bound. Mathematicians choose the embedding and base, but polynomial witnesses determine membership once those choices are fixed. Institutional origin. Field theory supplies the term; number fields and finite fields instantiate it without one application defining the identity.^[1]^[2]

Vocabulary travel. Extension and dependence are broad language; a nonzero polynomial with coefficients in the base field that annihilates each extension element is not. Import versus recognition. A new field pair qualifies by checking every element against a base-coefficient polynomial. Saying a theory is an “algebraic extension” of another without field operations imports only the words.^[1]

Its character: structural within algebra, but domain-framed by field operations and base-relative polynomial dependence.

Structural Core vs. Domain Accent¶

Portable skeleton. A general “all members of an extension satisfy a base-relative constraint” is only a future-prime candidate, not an established cross-domain parent. The staged strict genus is live Field Extension: an algebraic extension is a field extension satisfying the stronger all-elements polynomial condition.^[1]

Domain-bound mechanism. With base embedding \(K\hookrightarrow L\), every \(\alpha\in L\) must be annihilated by some nonzero polynomial over \(K\). Number fields make minimal polynomials salient; finite fields use Frobenius identities and may offer a common polynomial witness in a finite extension. Neither finite degree nor one shared witness is required in the general definition.^[1]^[2]

Why not prime. The phrase “constrained extension” might transfer, but without field addition/multiplication, coefficient rings and polynomial evaluation there is no literal algebraicity test. A broad future skeleton cannot replace the exact field-theoretic identity; the live Field Extension node supplies the valid immediate genus.

This entry is a kind of Field Extension.

A staged strict subsumption edge to live domain-specific Field Extension is proposed: every algebraic extension is a field extension, and the all-elements condition makes this a genuine narrower kind. Live Degree of a Field Extension is an invariant, not the genus; infinite algebraic extensions show why degree is not a strict parent. No canonical edge has been applied.

Relationships to Other Abstractions¶

Current abstraction Algebraic Extension Domain-specific

Parents (1) — more general patterns this builds on

Algebraic Extension is a kind of Field Extension Domain-specific

Every algebraic extension is a field extension satisfying an additional all-elements polynomial condition.

Hierarchy path (1) — routes to 1 parentless root

Algebraic Extension → Field Extension

Neighborhood in Abstraction Space¶

Algebraic Extension sits in a sparse region of the domain-specific corpus (73^rd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Field Extensions & Galois-Theoretic Structures (18 abstractions)

Nearest neighbors

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

Not to Be Confused With¶

An algebraic element is one element relative to a base; an algebraic extension quantifies over the whole larger field. A finite extension is always algebraic but does not exhaust algebraic extensions. A transcendental extension contains at least one element without a polynomial witness over the base. An algebraic closure additionally contains roots needed to be algebraically closed.^[1]

References¶

[1] Romyar Sharifi, Abstract Algebra, Chapter 6: “Field Theory and Galois Theory,” UCLA, §§6.2 and 6.8; especially Example 6.2.6, Corollary 6.2.8, Definition 6.2.17, Propositions 6.2.18–19, Example 6.2.21, and Definition 6.8.11. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w

[2] Sophie Huczynska, with small changes by Max Neunhoffer, Finite Fields course notes (2011/12), Chapter 3 §6, Lemmas 6.1 and 6.3–6.4 and Theorem 6.5, printed pp. 23–25. registry ↩a ↩b ↩c ↩d ↩e ↩f