Quadratically Closed Field¶

A quadratically closed field contains a square root of every one of its elements, eliminating missing roots of quadratic polynomials.

Version: v1 · 2026-10-03 · History
Domain-specific #: 13539
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomains: Field Theory, Quadratic Form Theory → Mathematics
Aliases: Quadratic-closed field

Core Idea¶

A field k is quadratically closed when every element of k is a square of an element already in k: for every a∈k, there is x∈k with x²=a. In the characteristic-not-two convention used here, this is equivalent to saying that every quadratic polynomial over k has a root there, since completing the square reduces its root question to the existence of a square root of its discriminant.^[1]^[2]

The property lies between several familiar closures. An algebraically closed field is certainly quadratically closed, because it has roots of all nonconstant polynomials. A quadratically closed field need not have roots of every cubic or higher-degree polynomial. An ordered Euclidean field such as the reals has square roots for its nonnegative elements but is not quadratically closed: −1 is not a real square. Its extension by i, the complex field, is quadratically closed.^[2]

The property also collapses a quadratic-form invariant in a precise way. Ormsby's original lecture notes prove that the dimension map from the Grothendieck–Witt ring GW(k) to the integers is an isomorphism exactly when k is quadratically closed; the ordinary Witt ring W(k) is then Z/2Z. These are different rings. The seed's wording merged them and would make the invariant false if read as W(k)≅Z.^[1]

Structural Signature¶

Field carrier: addition, multiplication and division by nonzero elements.
Square map: x↦x² within the field.
Surjectivity condition: every field element has a preimage under that map.
Quadratic-polynomial consequence: under characteristic not two, no irreducible quadratic remains.
Quadratic-form consequence: one-dimensional coefficients become square-equivalent, collapsing GW(k) to dimension.
Optional closure construction: iteratively adjoining missing roots to a base field yields a quadratically closed overfield; that construction is not the definition of the property.

Condensed: field k + surjective square map → no missing quadratic roots, with all-polynomial closure still a separate, stronger condition.

Sig role-phrases: field carrier; square map; universal surjectivity; characteristic-not-two quadratic-polynomial consequence; quadratic-form invariant; optional closure construction.

What It Is Not¶

Not an algebraically closed field by definition. Square-root completeness does not automatically solve every higher-degree polynomial.
Not a Euclidean ordered field. In an ordered field only nonnegative elements can be squares; its square-root condition is correspondingly narrower.^[2]
Not a Pythagorean field alone. Closure of sums of squares does not imply that −1 or every other element is a square.
Not the same thing as a quadratic closure of a given base field. The adjective describes a field's intrinsic property; a quadratic closure is a selected extension constructed relative to another field.
Not the statement that the ordinary Witt ring is Z. It is the Grothendieck–Witt ring whose dimension map gives Z here; the Witt quotient is Z/2Z.^[1]
Not unrestricted across characteristic conventions. This entry uses the standard characteristic-not-two quadratic-polynomial and form-theory setting; characteristic two needs separate formulations.

Scope of Application¶

In field theory, the criterion eliminates quadratic root obstructions. If a is already a square, adjoining √a does not enlarge the field. If every element is a square, no nontrivial extension can be obtained by adjoining a root of X²−a under the stated characteristic assumption. This is a narrower result than asserting that no algebraic extension of any degree exists.

In quadratic-form theory, coefficients of one-dimensional forms become square-equivalent. The resulting dimension invariant is strong enough to characterize the field property through GW(k)≅Z, while passing to the Witt ring identifies hyperbolic contributions and gives Z/2Z instead.^[1]

In constructibility, one can adjoin square roots repeatedly within a chosen algebraic closure. For the rationals, the field K of complex constructible numbers is the resulting quadratic closure under finite quadratic towers. It is quadratically closed but not the full field of algebraic numbers. The word complex matters: the field of real constructible numbers lacks a square root of −1 and is not quadratically closed.^[3]^[4]

Clarity¶

Test all elements, not only positive ones or a sample of familiar values. One nonsquare disproves the property. In R, −1 is the immediate witness; in C, every element has a complex square root. When using polynomial language, state characteristic not two so completing-the-square arguments apply as written.

State whether “closure” means a property of the present field or a construction extending a base field. Also distinguish the ordinary Witt ring from the Grothendieck–Witt ring when quoting a collapse theorem. A wrong ring label changes Z to Z/2Z.

Manages Complexity¶

Instead of separately asking whether every quadratic polynomial has a solution, one tests the square map on the field. This single invariant organizes quadratic extensions and drastically simplifies quadratic-form classification. It does not compress arbitrary algebraic equations into quadratic ones; remaining cubic and other odd-degree problems retain independent content.

Abstract Reasoning¶

Given k of characteristic not two, ask whether k×=(k×)². If not, find a nonsquare a. Then X²−a has no root in k, and adjoining one gives a proper quadratic extension. If the square map is surjective, a quadratic aX²+bX+c with a≠0 can be completed to a square; its discriminant is a field element and thus has a square root, so the polynomial has a root in k.

Do not invert the implication from a stronger class. Algebraically closed implies quadratically closed; the reverse is not automatic. Likewise, a real-closed or Euclidean ordered field is not quadratically closed, since ordering excludes √−1; adjoining i can change the situation.^[2]

For quadratic forms, apply the proved dimension-map characterization to GW(k). If discussing W(k), account for the hyperbolic quotient and state its separate result. This avoids a common but material ring-name conflation.^[1]

Knowledge Transfer¶

The square-map test transfers across characteristic-not-two fields, regardless of whether the field is numerical, functional or abstract. The “all elements” quantifier and characteristic convention must remain. Methods that rely on order, on the real/complex example, or on the constructible-number interpretation are setting-specific, not universal consequences of the definition.

Examples¶

A quadratic over the complex field¶

In C, take a=3+4i. The explicitly exhibited element x=2+i satisfies x²=3+4i, so the equation X²−(3+4i)=0 has a root in the field. This is one execution of the square-map test, not by itself a proof of the universal property; C passes the all-elements test because it is algebraically closed.^[2]

Mapped back: the field carrier is C; the square map sends 2+i to 3+4i; surjectivity is supplied by the stronger all-polynomial closure theorem, and the displayed quadratic shows the polynomial consequence on an actual input.

Real-number near miss¶

Every nonnegative real has a real square root, but −1 does not. Hence R is an ordered Euclidean field but not a quadratically closed field.^[2]

Mapped back: one nonsquare witnesses failure of the all-element quantifier.

Complex constructible numbers: quadratic but not algebraic closure¶

Let K be the field of complex numbers obtainable by finitely many straightedge-and-compass steps from rational starting data. Elman, Karpenko and Merkurjev identify it as the quadratic closure of Q; Lipman's original notes give the complex quadratic-tower characterization and square-root closure.^[3]^[4] For a displayed member, i∈K, and x=(1+i)/√2 also belongs to K by another square-root step; x²=i. The same closure argument works for any a∈K: extend a finite tower containing a by one quadratic step to contain a root of X²−a. Yet ∛2∉K: its minimal polynomial X³−2 has degree three, whereas an element in a finite quadratic tower has degree over Q dividing a power of two.^[4] This is a concrete positive case that does not inherit full algebraic closure.

Mapped back: the field carrier is K; the square map has a preimage for each member by a further tower step (shown for i); the surjectivity condition holds, but the quadratic-polynomial consequence stops at degree two and does not solve X³−2.

Two rings, two answers¶

Over a quadratically closed field, dimension classifies GW(k) as Z, whereas W(k) is Z/2Z. Saying simply “the Witt ring becomes the integers” would confuse the invariants.^[1]

Mapped back: an equivalent characterization must preserve the exact ring.

Structural Tensions¶

No universal intrinsic two-sided optimization tension is established by this field property: either the square map is surjective or it is not. The quadratic-versus-arbitrary-algebraic distinction, intrinsic property versus chosen extension, and GW versus W are classification boundaries, not opposed objectives with costs on both sides. Their diagnostic tests belong in the examples and exclusions above; labeling them tradeoffs would misdescribe the mathematics.

Structural–Framed Character¶

On the structural–framed spectrum this entry lies strongly toward the structural end: whether every element is a square is a formal invariant of a field, independent of anyone's approval or preferred example. Its evaluative weight is essentially nil; a quadratically closed field is not inherently better than another field for every purpose. Human practice chooses which questions make the property useful, but does not create the square-map condition. The term and its distinction from quadratic Closure have a mathematical-community origin; that naming convention is framed, while the quantified algebraic relation is not. The vocabulary can travel among different fields and quadratic-form settings, but not intact into arbitrary networks or social systems that merely have something called closure. An algebraist recognizes the property by checking an existing field; deliberately adjoining roots imports it into an extension and changes the carrier. Its character: a formal, high-structure field property with a conventional name and domain-bound algebraic tests.

Structural Core vs. Domain Accent¶

The skeletal relation is a surjective operation on a carrier: each target element has a preimage under squaring. A possible cross-domain prime about closure or surjectivity would require its own independent admission; none is asserted here. The domain-bound mechanism is arithmetic in a field: multiplication defines x², and characteristic-not-two algebra turns that universal preimage condition into the quadratic-polynomial and quadratic-form consequences. The named entry fails the prime bar because removing the field, its square operation and those exact tests removes its identity; a workflow's metaphorical “closed loop” cannot instantiate quadratically closed Field (Algebraic). The Field relation is a taxonomic domain-specific parent, not evidence that this named subclass is portable across domains.

This entry is a kind of Field (Algebraic).

Field is the strict parent because every quadratically closed field is a field with an additional universal root property. Algebraically Closed Field is a prospective child because solving all polynomials entails solving quadratics; no edge to it is installed here. Euclidean Ordered Field and Real Closed Field are comparison neighbors, not parents of a class in which −1 is a square.

Relationships to Other Abstractions¶

Current abstraction Quadratically Closed Field Domain-specific

Parents (1) — more general patterns this builds on

Quadratically Closed Field is a kind of Field (Algebraic) Domain-specific

A quadratically closed field is a field with the additional requirement that every element is a square in that field.

Neighborhood in Abstraction Space¶

Quadratically Closed Field sits in a moderately populated region (58^th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

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

Nearest neighbors

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

Not to Be Confused With¶

Quadratic field usually means a degree-two extension of a base field, which need not be quadratically closed. Quadratic closure is an extension generated by iterated root adjunction. Algebraic closure solves all nonconstant polynomials and is stronger. Pythagorean field only requires sums of two squares to be squares.

References¶

[1] Kyle Ormsby, Quadratic Forms lecture notes, Proposition 14.2. Original author proof of the GW and W claims. registry ↩a ↩b ↩c ↩d ↩e ↩f

[2] “An Algebraic Approach to the Fundamental Theorem of Algebra,” Rendiconti del Circolo Matematico di Palermo (2024), Sections 3–4. Original research article; only stated definitions and results are used here. registry ↩a ↩b ↩c ↩d ↩e ↩f

[3] Richard Elman, Nikita Karpenko and Alexander Merkurjev, The Algebraic and Geometric Theory of Quadratic Forms, Chapter V §31. Authors' manuscript identifies the complex constructible field with the quadratic closure of Q and gives the ordinary Witt-ring statement. registry ↩a ↩b

[4] Joseph Lipman, Notes and Exercises on Constructibility, Theorem 2 and corollary. Original Purdue notes identify complex constructible points with finite quadratic towers over a conjugation-closed base and give the power-of-two degree obstruction. registry ↩a ↩b ↩c