Quadratic Space¶

A vector space equipped with a quadratic form, carrying isotropy, radical, orthogonality, and equivalence structure that depends essentially on the base field.

Version: v1 · 2026-08-30 · History
Domain-specific #: 2583
Origin domain: mathematics
Aliases: Formed quadratic space

Core Idea¶

A quadratic space is a pair \((V,Q)\) consisting of a vector space \(V\) over a field \(F\) and a quadratic form \(Q:V\to F\). The form satisfies \(Q(av)=a^2Q(v)\), while its polarization \(B_Q(u,v)=Q(u+v)-Q(u)-Q(v)\) is bilinear. Voight uses precisely this definition and calls \((V,Q)\) a quadratic space.^[1] The pair—not the form alone and not the vector space alone—is the abstraction.

Once \(Q\) is installed, vectors acquire form-relative predicates and relations: isotropic vectors satisfy \(Q(v)=0\); orthogonality is read from \(B_Q\); the radical records vectors bilinearly orthogonal to all of \(V\); and isometries preserve \(Q\). In characteristic not two, symmetric bilinear forms and quadratic forms are tightly linked through polarization. In characteristic two, that recovery fails and the distinction becomes structural, not cosmetic.^[1]

Structural Signature¶

Recognition roles:

Base field \(F\): controls scalar arithmetic, characteristic, square classes, and allowable normalizations.
Vector carrier \(V\): supplies addition, scalar multiplication, dimension, and subspaces.
Quadratic form \(Q\): assigns degree-two scalar values and obeys homogeneous scaling.
Polar form \(B_Q\): makes pairwise orthogonality and the radical available.
Isotropic cone: vectors with zero quadratic value, including the distinction between trivial and nontrivial zeros.
Nondegeneracy or radical status: records whether the form loses directions.
Equivalence notion: normally isometry, similarity, or Witt equivalence, which must be declared.
Invariants: dimension, discriminant, signature, Witt index, and field-specific data used only where defined.

A candidate is recognized when the vector operations and quadratic form are jointly load-bearing. Merely writing a quadratic polynomial in coordinates is insufficient unless basis changes are understood as descriptions of the same formed vector space.

What It Is Not¶

A quadratic space is not a quadratic field: the latter is a degree-two field extension. It is not a quadratic equation, whose identity is a one-variable root problem. It is not simply a normed vector space, because a general norm need not arise from a quadratic form or satisfy the parallelogram law. It is not a symmetric bilinear space without qualification; in characteristic two, distinct quadratic forms may share polar behavior that does not recover their diagonal values.^[1]

The word “space” also must not be inflated to mean a geometric manifold. A quadratic space is algebraic. Projective quadrics and orthogonal groups are constructions from it, not synonyms for the pair \((V,Q)\).

Scope of Application¶

Quadratic spaces organize classical quadratic-form theory, orthogonal geometry, arithmetic of forms, and algebraic groups. Over \(\mathbb R\), signature classifies nondegenerate forms up to real linear change of coordinates. Over finite fields and number fields, isotropy, discriminants, local invariants, and Witt decomposition take the leading roles; Lam provides the specialist field-based treatment behind this scope.^[2] Projectivizing \(Q(v)=0\) yields a quadric whose points encode isotropic lines. Orthogonal groups consist of linear automorphisms preserving the form.

Applications in geometry and physics often start with a quadratic space but add structure: Euclidean inner-product spaces arise from positive-definite forms over \(\mathbb R\), while Lorentzian tangent spaces use indefinite forms. Those applications do not erase the field and equivalence conventions that determine the algebraic abstraction.

Clarity¶

Naming the quadratic space prevents three substitutions. First, it keeps the carrier and form paired, so a change of basis does not create a new object. Second, it separates bilinear orthogonality from quadratic value, especially in characteristic two. Third, it forces a declared equivalence relation before “same form” is asserted.

The operational test is: identify \(F\) and its characteristic; verify homogeneous scaling and bilinear polarization; state whether degeneracy is allowed; and specify whether comparisons are by isometry, similarity, or Witt class. If any item is missing, invariants such as discriminant or signature can be misapplied.

Manages Complexity¶

The pair \((V,Q)\) compresses a coordinate polynomial plus its transformation law into a basis-independent object. A matrix represents the polar form after choosing a basis, but the space makes clear which matrix changes preserve identity. Orthogonal decomposition reduces a complicated formed space into manageable summands. Witt decomposition separates hyperbolic planes from an anisotropic remainder, exposing the part that cannot be canceled by isotropic pairs.

The abstraction deliberately retains field dependence. There is no single invariant list that classifies all quadratic spaces over all fields. It also does not make every quadratic-form computation easy; arithmetic classification may require local-global arguments and field-specific invariants.

Abstract Reasoning¶

If \(W\subseteq V\), one can restrict \(Q\) to \(W\), form its orthogonal complement, and test whether direct orthogonal decomposition is available. An isometry carries isotropic vectors to isotropic vectors and preserves any invariant correctly defined for the setting. If \(Q(v)\ne0\) and characteristic is not two, reflection formulas generate familiar orthogonal transformations.

Coordinate calculations become invariant deductions only after transformation behavior is checked. For example, a diagonal coefficient list over \(\mathbb R\) yields signature after rescaling nonzero entries to \(\pm1\), but the same move over a general field is blocked by nonsquare classes. This contrast is part of the structure.

Knowledge Transfer¶

Transfer inside the domain is exact: the same carrier–form–polarization roles recur in real, finite-field, and arithmetic cases. Techniques such as restriction, orthogonal complement, and isotropy testing travel, while the final classification invariants change with \(F\). A geometric or physical problem instantiates the abstraction only when its local or global vector carrier genuinely bears a quadratic form.

Cross-domain transfer beyond mathematics usually belongs to the parent Vector Space or to generic invariant-preserving representation. Calling a cost landscape “quadratic space” because it contains squared terms is metaphor unless the formed-vector-space roles and isometries are present.

Examples¶

Let \(V=\mathbb R^2\) and \(Q(x,y)=x^2+y^2\). Then \(B_Q((x,y),(u,v))=2xu+2yv\). The only isotropic vector is zero, the form is positive definite, and rotations preserve it. This is a quadratic space, not merely the equation \(x^2+y^2=0\).

Let \(Q(x,y)=x^2-y^2\). The nonzero vectors \((1,1)\) and \((1,-1)\) are isotropic, while the form is nondegenerate with signature \((1,1)\). In coordinates \(p=x+y\), \(q=x-y\), the form is \(pq\), exposing a hyperbolic plane. The example shows why nondegeneracy does not imply anisotropy.

Over a field of characteristic two, take \(Q(x,y)=xy\). Its polar form is \(B_Q((x,y),(u,v))=xv+yu\). One cannot recover \(Q(v)\) by halving \(B_Q(v,v)\), because that diagonal is zero. The example marks the characteristic boundary rather than treating it as a technical footnote.^[1]

Structural Tensions¶

Quadratic form versus polar form. They correspond cleanly away from characteristic two but diverge in characteristic two. Diagnostic: record the characteristic before replacing one by the other.
Coordinate matrix versus intrinsic pair. Matrices enable calculation yet change under basis choice. Diagnostic: state the congruence transformation and compare invariants, not raw entries.
Nondegenerate versus anisotropic. A zero radical does not rule out nonzero isotropic vectors. Diagnostic: test \(Q(v)=0\) separately from testing the polar matrix’s rank.
Uniform language versus field-sensitive classification. The same definition spans fields, while signature or square-class data may not. Diagnostic: cite only invariants defined for the declared field.
Autonomy versus reduction. Vector Space plus a generic scalar map can describe ingredients, but the polarization, isotropy, and isometry package is residual. Diagnostic: require at least one form-specific relation or invariant to do real inferential work.

Structural–Framed Character¶

Quadratic Space is structural inside algebra, surviving basis replacement and coordinate renaming. It remains domain-framed because field characteristic, quadratic homogeneity, polarization, and algebraic equivalence are indispensable. The mathematical substrate cannot be replaced by arbitrary objects while keeping the candidate’s vocabulary literally intact.

Structural Core vs. Domain Accent¶

The portable skeleton is a carrier equipped with a value-assignment whose preserving maps and invariants organize comparison. The domain accent is the vector-space law, degree-two scaling, bilinear polarization, and field-sensitive isotropy theory. Once those are removed, only a generic structured carrier remains; that residue belongs to broader primes.

Quadratic Space strictly composes Vector Space: every instance has a vector carrier, while most vector spaces do not come with a selected quadratic form. It also relates to Metric only in restricted positive-definite real settings; a general quadratic space need not define a metric, so that edge is declined.

Relationships to Other Abstractions¶

Current abstraction Quadratic Space Domain-specific

Parents (1) — more general patterns this builds on

Quadratic Space presupposes Vector Space Prime

Quadratic Space strictly composes Vector Space: every instance has a vector carrier, while most vector spaces do not come with a selected quadratic form.

Hierarchy path (1) — routes to 1 parentless root

Quadratic Space → Vector Space → Set and Membership

Neighborhood in Abstraction Space¶

Quadratic Space sits in a sparse region of the domain-specific corpus (64^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Applied Linear & Special Functions (18 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Multilinear form: the polar form is bilinear, but \(Q\) itself scales quadratically.
Norm form: a specific form produced by a field extension’s norm, not the general pair.
Inner-product space: closely related over \(\mathbb R\) with positive definiteness, but narrower and characteristic-sensitive.
Quadratic field: a degree-two field extension, not a formed vector carrier.
Quadratic equation: a root problem rather than a space with preserving maps.
Quadratic form: the map component; “quadratic space” explicitly includes its carrier and base field.

References¶

[1] John Voight, Quaternion Algebras, Graduate Texts in Mathematics 288, chapter 4 “Quadratic Forms,” Springer, 2021, DOI 10.1007/978-3-030-56694-4_4. registry ↩a ↩b ↩c ↩d

[2] T. Y. Lam, Introduction to Quadratic Forms over Fields, Graduate Studies in Mathematics 67, American Mathematical Society, 2005, ISBN 978-0-8218-3109-1. registry ↩