TOPICS
Search

Symmetric Bilinear Form


A symmetric bilinear form on a vector space V is a bilinear function

 Q:V×V->R
(1)

which satisfies Q(v,w)=Q(w,v).

For example, if A is a n×n symmetric matrix, then

 Q(v,w)=v^(T)Aw=<v,Aw>
(2)

is a symmetric bilinear form. Consider

 A=[1 2; 2 -3],
(3)

then

 Q((a_1,a_2),(b_1,b_2))=a_1b_1+2a_1b_2+2a_2b_1-3a_2b_2.
(4)

A quadratic form may also be labeled Q, because quadratic forms are in a one-to-one correspondence with symmetric bilinear forms. Note that Q(a)=Q(a,a) is a quadratic form. If Q(a) is a quadratic form then it defines a symmetric bilinear form by

 Q(a,b)=1/2[Q(a+b)-Q(a)-Q(b)].
(5)

The kernel, or radical, of a symmetric bilinear form is the set of vectors

 KerQ={v:Q(v,w)=0 for all w in V}.
(6)

A quadratic form is called nondegenerate if its kernel is zero. That is, if for all v in V, there is a w in V with Q(v,w)!=0. The rank of Q is the rank of the matrix (a_(ij))=Q(e_i,e_j).

The form Q is diagonalized if there is a basis {v_i}, called an orthogonal basis, such that (b_(ij))=Q(v_i,v_j) is a diagonal matrix. Alternatively, there is a matrix C such that

 Q(Cv,Cw)=(Cv)^(T)A(Cw)=v^(T)(C^(T)AC)w
(7)

is a diagonal quadratic form. The jth column of the matrix C is the vector v_j.

A nondegenerate symmetric bilinear form can be diagonalized, using Gram-Schmidt orthonormalization to find the v_i, so that the diagonal matrix C^(T)AC has entries either 1 or -1. If there are p 1s and q -1s, then Q is said to have matrix signature (p,q). Real nondegenerate symmetric bilinear forms are classified by their signature, in the sense that given two vector spaces with forms of signature (p,q), there is an isomorphism of the vector spaces which takes one form to the other.

A symmetric bilinear form with Q(v,v)>0, for all nonzero v, is called positive definite. For example, the usual inner product is positive definite. A positive definite form has signature (n,0). A negative definite form is the negative of a positive form and has signature (0,n). If the form is neither positive definite nor negative definite, then there must exist vectors w!=0 such that Q(w,w)=0, called isotropic vectors.

A general symmetric bilinear form Q can be diagonalized with diagonal entries 1, -1, or 0, because the form Q is always nondegenerate on the quotient vector space V/KerQ. If V is a complex vector space, then a symmetric bilinear form can be diagonalized to have entries 1 or 0. For other fields, there are more symmetric bilinear forms than in the real or complex case. For instance, if the field has field characteristic 2, then it is not possible to divide by 2 since 2=0. Hence there is no correspondence between quadratic forms and symmetric bilinear forms in characteristic 2.

The symmetric bilinear forms on a vector space, whose field k is not real, have been classified for some fields. There are also theorems about symmetric bilinear forms on free Abelian groups, for example Z^n.

A symmetric bilinear form Q corresponds to a matrix A by giving a basis e_i and setting a_(ij)=Q(e_i,e_j). Two symmetric bilinear forms are considered equivalent if a change of basis takes one to the other. Hence, A∼CAC^(T), where C is any invertible matrix. Therefore, the rank of the symmetric bilinear form is an invariant.

Also, detA can change by (detC)^2detA. The coset of detA in k^*/k^*^2 is a well-defined invariant of Q, called the discriminant. For real forms, it is either 1 or -1. For Q, the discriminant can be any rational number a/b where a and b are squarefree. A symmetric bilinear form on a finite field is determined by its rank and its discriminant.

A symmetric bilinear form on the p-adic numbers Q_p is characterized by its rank, discriminant, and another invariant epsilon(Q). Given a basis e_i, orthogonal for Q, define a_i=Q(e_1,e_i), then

 epsilon(Q)=product_(i<j)(a_i,a_j)
(8)

where (a_i,a_j) is the Hilbert symbol.

Two symmetric bilinear forms are equivalent on the rationals iff they are equivalent in every Q_p as well as the reals (also called Q_infty.) The data in Q_p can be thought of as "local" information, which can be patched together to yield "global" information in Q. So rational forms have a countable number of distinct invariants, three for every prime number, and two for the reals.


See also

Diagonal Quadratic Form, Field, Hilbert Symbol, Inner Product, Matrix Index, p-adic Number, Quadratic Form, Signature, Vector Space

This entry contributed by Todd Rowland

Explore with Wolfram|Alpha

Cite this as:

Rowland, Todd. "Symmetric Bilinear Form." From MathWorld--A Wolfram Web Resource, created by Eric W. Weisstein. https://mathworld.wolfram.com/SymmetricBilinearForm.html

Subject classifications