The theory of quadratic forms: algebra and geometry

The theory of quadratic forms is a fundamental pillar at the intersection of algebra and geometry. It offers a powerful perspective for deciphering the properties of homogeneous polynomials of degree two, as well as their connections to the structure of vector spaces. Long confined to abstract mathematical treatises, this discipline now permeates a variety of fields ranging from number theory to contemporary physics, illustrating the richness of a multidimensional approach. The main concepts that emerge concern the classification of forms by their invariants, the diagonalization of associated matrices, and geometric analysis through the notions of orthogonality and signature.

Observing a quadratic form is not merely grasping a polynomial: it is entering a world in which coefficients and variables weave a web connecting linear algebra and geometry. Understanding these objects therefore means opening up to fine classifications based on their invariants, discovering how each form expresses a unique geometric structure, and viewing linear transformations in a new light. As the theory develops, it also reveals its major applications, particularly in the context of special relativity and algebraic topology, demonstrating its vital importance to modern mathematicians and physicists.

The powerful relationship between a quadratic form and its associated symmetric matrix multiplies the possibilities for analysis. Diagonalization and spectral theory then become indispensable tools for characterizing the nature of a form and its impact in the underlying vector space. Understanding this interaction between algebraic theory and geometry provides an ideal framework for studying the classification and manipulation of quadratic forms, essential elements of any in-depth study.

This leads to a precise study of concepts such as index, radical, and discriminant, based on the intrinsic properties of forms and their behavior under a change of basis. These concepts, far from being purely abstract, express geometric realities that can be visualized and interpreted, such as light cones in relativity or inertia ellipsoids in mechanics.

The study of quadratic forms thus serves as a bridge between algebraic rigor and geometric intuition, offering a comprehensive view of an indispensable mathematical tool, rich in both theoretical and practical implications.

Key points:

• A quadratic form is a homogeneous polynomial of degree 2 that characterizes algebraic and geometric structures in a vector space.
• The correspondence with a unique symmetric bilinear form makes it possible to define the polar form, the heart of the theory.
• Diagonalization of the associated matrices plays a central role in classifying forms according to their invariants.
• Applications are wide-ranging, from special relativity to algebraic topology, as well as classical mechanics.
• The signature of a real quadratic form is an essential invariant for characterizing its geometric behavior.

Algebraic foundations of quadratic forms and associated matrices

At the heart of the theory of quadratic forms lies the notion of a vector space defined over a field, often of characteristic other than 2 to ensure division by 2. A quadratic form is then defined as a homogeneous map of degree two, characterized by the existence of a unique associated symmetric bilinear form, called the polar form. This duality between a quadratic form and a bilinear form is formalized by the polarization identity, an essential tool for reconstructing one from the other.

The associated bilinear form B makes it possible to manipulate the quadratic form Q without relying on individual variables, providing a solid basis for abstraction. In a vector space of dimension n, a fixed basis allows B to be expressed by a symmetric n × n matrix. The value of Q at a vector is then given by a homogeneous polynomial of degree 2 in the coordinates, succinctly written as Q(u) = uᵗBu, where u represents the column matrix of coordinates.

This matrix representation makes linear transformations easier to understand. A change of basis yields a congruent matrix B′ = PᵗBP, where P is the change-of-basis matrix, emphasizing that the intrinsic properties of quadratic forms do not depend on the choice of basis. This invariance under congruence is crucial for classification and for understanding invariants such as rank or discriminant.

A classic example illustrates the concept: the quadratic form x² + y² + z² over ℝ³ corresponds to a 3×3 identity matrix, which is already diagonal. This form defines the Euclidean norm, providing an immediate geometric interpretation in terms of distances and angles, and serving as a prototype of a positive-definite quadratic form.

We also encounter more complex forms, such as the Minkowski form on ℝ⁴, defined by x² + y² + z² – t², which is fundamental to special relativity. The structure of its associated matrix reveals a negative sign, introducing the concept of signature and highlighting the variety of possible quadratic forms beyond the positive-definite ones. This transition from algebra to geometry illustrates how the matrix and its eigenvalues determine the geometric properties of a vector space, as well as its behavior under transformations.

Diagonalization of symmetric matrices, closely related to spectral theory, emerges as a decisive step. It simplifies the quadratic form with respect to an orthogonal basis, reducing the expression to a weighted sum of squares, whose coefficients are the eigenvalues. This simplification opens the door to studying invariants. Eigenvalues and rank emerge as essential measures for classifying forms, distinguishing, for example, between positive-definite, negative-definite, and indefinite forms.

Here is a summary table of the fundamental properties of a quadratic form in relation to its matrices and invariants:

These fundamental concepts form the basis for the geometric study and classification of quadratic forms, and guide solution methods in a variety of contexts.

Classification and invariants of quadratic forms in vector spaces

The classification of quadratic forms relies on invariants that remain constant under a change of basis, thus making it possible to group these forms into equivalence classes. Rank, discriminant, and signature are central to this approach. Rank, related to the dimension of the image of the associated bilinear form, measures the degree to which the form is nondegenerate, while the discriminant, an element of the object K*/(K*)², expresses the fundamental nature of the coefficients, independently of the chosen basis.

For real vector spaces, the signature, arising from Sylvester’s law of inertia, becomes the main invariant. It is defined as the pair (p, q), where p is the number of strictly positive eigenvalues and q is the number of strictly negative eigenvalues of the associated matrix. This pair not only identifies the nature of the form (positive-definite, negative-definite, or indefinite), but also allows us to deduce geometric properties such as the type of curvature or orientation in a pseudo-Riemannian space.

A major result in 2025 concerns the equivalence of quadratic forms: two forms are said to be equivalent (or isometric) if there is a linear isomorphism transforming one into the other. This notion of equivalence classifies forms into orbits under the action of the general linear group GL(V). Thus, complete classification consists of determining these orbits, which in practice can be achieved by analyzing their invariants and diagonalizing them.

Several cases are distinguished according to the base field:

• Algebraically closed field (for example, ℂ): every quadratic form is determined by its rank; two forms of the same rank are equivalent.
• Real field ℝ: Sylvester’s law of inertia dictates classification by signature; two forms are equivalent if they have the same signature.
• Finite fields: classification is based on the discriminant and rank, and two main classes arise.
• Rational field ℚ: the situation is infinitely richer, with an infinite variety of equivalence classes.

These distinctions have a profound impact on how problems involving quadratic forms are approached in different mathematical or physical contexts. For example, in relativity, the signature (3,1) or (1,3) corresponds to the Minkowski metric, shaping the structure of spacetime.

A final invariant, the isotropy index, measures the maximum size of a totally isotropic subspace (where the form vanishes). This index characterizes the relative “neutrality” of a form and plays an important role in the theory of quadratic spaces, particularly in the study of nondegenerate forms and their finer classification.

To summarize:

1. Invariants structure the classification of quadratic forms through properties that remain unchanged under a change of basis.
2. Diagonalization with respect to an orthogonal basis facilitates the study of eigenvalues, which are at the heart of the invariants.
3. The theory offers a unified framework for understanding a variety of situations in algebra, geometry, and physics.

Geometric applications of quadratic forms and connections with conics

Quadratic forms occupy a central place in geometry, particularly through their role in defining algebraic curves and surfaces called quadrics. These objects include the classical conics (circles, ellipses, parabolas, hyperbolas) and their multidimensional analogues. Quadratic forms therefore provide a unified basis for approaching these figures through their homogeneous polynomial equation of degree two.

Each quadric can be expressed as the set of points v in a vector space V satisfying Q(v) = 0, where Q is a quadratic form with an associated matrix. The geometry of these forms, particularly their signature and discriminant, determines the geometric nature of the quadrics: ellipsoids, paraboloids, cones, and so on. The topological and differential properties of a quadric follow directly from the algebraic characteristics of Q.

For example, in ℝ³, the quadratic form x² + y² + z² = 1 defines a sphere (ellipsoid), while the form x² + y² – z² = 0 defines a double cone. The signature makes it possible to predict the general shape of the surface without requiring an explicit calculation. This geometric interpretation makes it possible to explore the diversity of quadrics and understand their classification in connection with algebra.

In practice, reducing the form by diagonalization (Gaussian reduction) makes it possible to simplify the equation of the quadric and determine its type. This method is used in the local analysis of surfaces in differential geometry to study curvature, which itself is based on quadratic forms arising from the Taylor expansions of functions defining the surfaces.

Quadratic forms also arise in the analysis of critical points, which are essential in optimization and modeling. The Taylor expansion of a function f at a critical point introduces the Hessian matrix, which can be viewed as the matrix associated with a quadratic form. The nature of the critical point (minimum, maximum, or saddle) is then determined by studying the eigenvalues of the matrix, connecting analytic geometry with the theory of quadratic forms.

Finally, the notion of an orthogonal space defined by the associated bilinear form opens the way to studying geometric configurations, making it possible to define notions of angles, projections, and orthogonal decompositions in spaces more general than classical Euclidean spaces.

Interactions with modern physics: relativity and mechanics

Quadratic forms play a prominent role in the mathematical formulation of several modern physical theories, particularly special and general relativity, as well as solid mechanics. Their ability to encode the geometry and symmetries of a physical space makes them essential tools for expressing fundamental laws.

In the context of special relativity, the quadratic form Q(x, y, z, t) = x² + y² + z² – t² describes the light cone, the boundary between events accessible and inaccessible to an observer according to the principle of causality. This cone represents the vanishing of the quadratic form and defines a so-called pseudo-Euclidean or Minkowski geometry. The sign of the form expresses temporal and spatial orthogonality relations, making it possible to distinguish temporally connected events.

More broadly, in general relativity, the metric on the spacetime manifold is given by a pseudo-Riemannian quadratic form, whose signature and eigenvalues govern gravitational dynamics. The matrix associated with this form corresponds to the metric tensor, a central object in Einstein’s geometric description of gravity.

Quadratic forms also arise in classical mechanics to define physical quantities such as kinetic energy, which is often written as the square of a norm derived from a positive-definite quadratic form. In solid mechanics, the quadratic form makes it possible to characterize inertia ellipsoids, determining the dynamic behavior of rigid bodies around their principal axes.

The theory of quadratic forms thus provides a common language for expressing equilibrium, stability, and fundamental interactions in a variety of physical systems. Algebra ensures the rigorous manipulation of quantities, while geometry provides an intuitive and visual interpretation of phenomena. This duality makes their study indispensable in 2025 for training researchers and engineers who model the real world.

Quadratic forms in modern areas of mathematical research

Beyond their historical role, quadratic forms continue to drive advances in several contemporary branches of pure and applied mathematics. Their growing usefulness in number theory, algebraic topology, and functional analysis illustrates their power.

In number theory, integral quadratic forms, particularly binary ones, are used to solve classical Diophantine equations, such as representing integers as sums of squares. The work of Lagrange, Gauss, and their successors enabled a comprehensive classification that remains a solid foundation for research in this field to this day. The algebraic structure of equivalence classes over ℚ opens the way to finer analyses of arithmetic invariants.

In algebraic topology, quadratic forms arise in the study of intersections and cobordism, making it possible to identify deep topological invariants of manifolds. Their connection with orthogonal groups, Clifford algebras, and spinor theory provides a rich framework for describing spaces with complex symmetries and nontrivial behavior.

Functional analysis, particularly in Hilbert spaces, uses Hermitian quadratic forms to study symmetric operators, which are essential in spectral theory. These developments apply to the study of partial differential equations and integrals, as well as their numerical solutions, which are crucial in many physical and engineering models.

Exploring these fields requires an in-depth understanding of the properties of quadratic forms, their classification, and the algebraic tools used to manipulate them. Researchers also make use of generalizations of quadratic forms to modules over more complex rings, adapting the theory to the constraints of characteristic 2 or to nonclassical structures.

The current influence of quadratic forms underscores their central place in the mathematical landscape of 2025, connected to growing needs in modeling, analysis, and the resolution of fundamental problems.

What is a quadratic form?

A quadratic form is a homogeneous polynomial function of degree two on a vector space, represented by a polynomial in several variables where each term has degree two, often associated with a symmetric matrix.

What is the relationship between a quadratic form and a bilinear form?

Every quadratic form is associated with a unique symmetric bilinear form, called the polar form, which makes it possible to derive the quadratic form via a reversible polarization identity.

How are quadratic forms classified?

Classification is based on invariants such as rank, discriminant, and signature, which remain invariant under a change of basis, allowing forms to be grouped into equivalence classes or orbits under the action of the linear group.

What is the signature of a real quadratic form?

The signature is the pair (p, q), denoting the number of strictly positive and negative eigenvalues of the matrix associated with the quadratic form, characterizing its geometric behavior and the nature of the form.

What are the connections between quadratic forms and physics?

Quadratic forms model essential physical structures, particularly in special relativity via the light cone and in classical mechanics to describe kinetic energy or the inertia ellipsoid of a solid.