Classification Theorems
Classification theorems for bilinear and quadratic forms provide complete invariants determining when two forms are equivalent. The answer depends crucially on the underlying field.
Two real symmetric bilinear forms (or quadratic forms) are congruent if and only if they have the same signature .
This means every real quadratic form is congruent to exactly one of:
The rank is , and signature completely classifies the form.
Over complex numbers, symmetric bilinear forms are classified only by rank. Every non-degenerate form of rank is congruent to:
The distinction between positive/negative disappears since is a square in .
If as spaces with bilinear forms (where is non-degenerate), then .
This allows "canceling" common summands when comparing forms, analogous to cancellation in arithmetic.
In , quadratic forms classify conics via discriminant :
- : Ellipse (signature or )
- : Hyperbola (signature )
- : Parabola (degenerate, rank 1)
The signature provides geometric classification.
A skew-symmetric bilinear form is called symplectic if non-degenerate. Every symplectic form on is congruent to:
This canonical form underlies Hamiltonian mechanics and symplectic geometry. Dimension must be even; symplectic forms exist only on even-dimensional spaces.
Classification theorems transform infinite families of matrices into finite lists of canonical forms. Over , geometry matters: signature distinguishes elliptic from hyperbolic behavior. Over , only rank survives. These results extend to more general settings (p-adic fields, function fields) with increasingly subtle invariants, forming the foundation of algebraic K-theory and quadratic form theory.