Bilinear form

In mathematics, more specifically in abstract algebra and linear algebra, a bilinear form on a vector space V is a bilinear map V × V → K, where K is the field of scalars. In other words, a bilinear form is a function B : V × V → K which is linear in each argument separately:

• B(u + v, w) = B(u, w) + B(v, w)
• B(u, v + w) = B(u, v) + B(u, w)
• B(λu, v) = B(u, λv) = λB(u, v)

The definition of a bilinear form can be extended to include modules over a commutative ring, with linear maps replaced by module homomorphisms.

When K is the field of complex numbers C, one is often more interested in sesquilinear forms, which are similar to bilinear forms but are conjugate linear in one argument.

Coordinate representation

Let V ≅ Kⁿ be an n-dimensional vector space with basis {e₁, ..., e_n}. Define the n × n matrix A by A_ij = B(e_i, e_j). If the n × 1 matrix x represents a vector v with respect to this basis, and analogously, y represents w, then:

$B(\mathbf{v}, \mathbf{w}) = x^\mathrm T Ay = \sum_{i,j=1}^n a_{ij} x_i y_j.$

Suppose {f₁, ..., f_n} is another basis for V, such that:

[f₁, ..., f_n] = [e₁, ..., e_n]S

where S ∈ GL(n, K). Now the new matrix representation for the bilinear form is given by: S^TAS.

Maps to the dual space

Every bilinear form B on V defines a pair of linear maps from V to its dual space V^∗. Define B₁, B₂: V → V^∗ by

B₁(v)(w) = B(v, w)

B₂(v)(w) = B(w, v)

This is often denoted as

B₁(v) = B(v, â‹…)

B₂(v) = B(â‹…, v)

where the dot ( â‹… ) indicates the slot into which the argument for the resulting linear functional is to be placed (see Currying).

For a finite-dimensional vector space V, if either of B₁ or B₂ is an isomorphism, then both are, and the bilinear form B is said to be nondegenerate. More concretely, for a finite-dimensional vector space, non-degenerate means that every non-zero element pairs non-trivially with some other element:

$B(x,y)=0\,$ for all $y \in V$ implies that x = 0 and

$B(x,y)=0\,$ for all $x \in V$ implies that y = 0.

The corresponding notion for a module over a commutative ring is that a bilinear form is unimodular if V → V^∗ is an isomorphism. Given a finite-dimensional module over a commutative ring, the pairing may be injective (hence "nondegenerate" in the above sense) but not unimodular. For example, over the integers, the pairing B(x, y) = 2xy is nondegenerate but not unimodular, as the induced map from V = Z to V^∗ = Z is multiplication by 2.

If V is finite-dimensional then one can identify V with its double dual V^{∗∗}. One can then show that B₂ is the transpose of the linear map B₁ (if V is infinite-dimensional then B₂ is the transpose of B₁ restricted to the image of V in V^{∗∗}). Given B one can define the transpose of B to be the bilinear form given by

^tB(v, w) = B(w, v).

The left radical and right radical of the form B are the kernels of B₁ and B₂ respectively;^[1] they are the vectors orthogonal to the whole space on the left and on the right.^[2]

If V is finite-dimensional then the rank of B₁ is equal to the rank of B₂. If this number is equal to dim(V) then B₁ and B₂ are linear isomorphisms from V to V^∗. In this case B is nondegenerate. By the rank–nullity theorem, this is equivalent to the condition that the left and equivalently right radicals be trivial. For finite-dimensional spaces, this is often taken as the definition of nondegeneracy:

Definition: B is nondegenerate if B(v, w) = 0 for all w implies v = 0.

Given any linear map A : V → V^∗ one can obtain a bilinear form B on V via

B(v, w) = A(v)(w).

This form will be nondegenerate if and only if A is an isomorphism.

If V is finite-dimensional then, relative to some basis for V, a bilinear form is degenerate if and only if the determinant of the associated matrix is zero. Likewise, a nondegenerate form is one for which the determinant of the associated matrix is non-zero (the matrix is non-singular). These statements are independent of the chosen basis. For a module over a commutative ring, a unimodular form is one for which the determinant of the associate matrix is a unit (for example 1), hence the term; note that a form whose matrix is non-zero but not a unit will be nondegenerate but not unimodular, for example B(x, y) = 2xy over the integers.

Symmetric, skew-symmetric and alternating forms

We define a form to be

• symmetric if B(v, w) = B(w, v) for all v, w in V;
• alternating if B(v, v) = 0 for all v in V;
• skew-symmetric if B(v, w) = −B(w, v) for all v, w in V;

  Proposition: Every alternating form is skew-symmetric.

  Proof: This can be seen by expanding B(v + w, v + w).

If the characteristic of K is not 2 then the converse is also true: every skew-symmetric form is alternating. If, however, char(K) = 2 then a skew-symmetric form is the same as a symmetric form and there exist symmetric/skew-symmetric forms that are not alternating.

A bilinear form is symmetric (resp. skew-symmetric) if and only if its coordinate matrix (relative to any basis) is symmetric (resp. skew-symmetric). A bilinear form is alternating if and only if its coordinate matrix is skew-symmetric and the diagonal entries are all zero (which follows from skew-symmetry when char(K) ≠ 2).

A bilinear form is symmetric if and only if the maps B₁, B₂: V → V^∗ are equal, and skew-symmetric if and only if they are negatives of one another. If char(K) ≠ 2 then one can decompose a bilinear form into a symmetric and a skew-symmetric part as follows

$B^{+} = \tfrac{1}{2} (B + {}^{\text{t}}B) \qquad B^{-} = \tfrac{1}{2} (B - {}^{\text{t}}B) ,$

where ^tB is the transpose of B (defined above).

Derived quadratic form

For any bilinear form B : V × V → K, there exists an associated quadratic form Q : V → K defined by Q : V → K : v ↦ B(v,v).

When char(K) ≠ 2, the quadratic form Q is determined by the symmetric part of the bilinear form B and is independent of the antisymmetric part. In this case there is a one-to-one correspondence between the symmetric part of the bilinear form and the quadratic form, and it makes sense to speak of the symmetric bilinear form associated with a quadratic form.

When char(K) = 2 and dim V > 1, this correspondence between quadratic forms and symmetric bilinear forms breaks down.

Reflexivity and orthogonality

Definition: A bilinear form B : V × V → K is called reflexive if B(v, w) = 0 implies B(w, v) = 0 for all v, w in V.

Definition: Let B : V × V → K be a reflexive bilinear form. v, w in V are orthogonal with respect to B if B(v, w) = 0.

A form B is reflexive if and only if it is either symmetric or alternating.^[3] In the absence of reflexivity we have to distinguish left and right orthogonality. In a reflexive space the left and right radicals agree and are termed the kernel or the radical of the bilinear form: the subspace of all vectors orthogonal with every other vector. A vector v, with matrix representation x, is in the radical of a bilinear form with matrix representation A, if and only if Ax = 0 ⇔ x^TA = 0. The radical is always a subspace of V. It is trivial if and only if the matrix A is nonsingular, and thus if and only if the bilinear form is nondegenerate.

Suppose W is a subspace. Define the orthogonal complement^[4]

$W^{\perp}=\{\mathbf{v} \mid B(\mathbf{v}, \mathbf{w})=0\ \forall \mathbf{w}\in W\} \ .$

For a non-degenerate form on a finite dimensional space, the map W ↦ W^⊥ is bijective, and the dimension of W^⊥ is dim(V) − dim(W).

Different spaces

Much of the theory is available for a bilinear mapping to the base field

B : V × W → K.

In this situation we still have induced linear mappings from V to W^∗, and from W to V^∗. It may happen that these mappings are isomorphisms; assuming finite dimensions, if one is an isomorphism, the other must be. When this occurs, B is said to be a perfect pairing.

In finite dimensions, this is equivalent to the pairing being nondegenerate (the spaces necessarily having the same dimensions). For modules (instead of vector spaces), just as how a nondegenerate form is weaker than a unimodular form, a nondegenerate pairing is a weaker notion than a perfect pairing. A pairing can be nondegenerate without being a perfect pairing, for instance Z × Z → Z via (x,y) ↦ 2xy is nondegenerate, but induces multiplication by 2 on the map Z → Z^∗.

Terminology varies in coverage of bilinear forms. For example, F. Reese Harvey discusses "eight types of inner product".^[5] To define them he uses diagonal matrices A_ij having only +1 or −1 for non-zero elements. Some of the "inner products" are symplectic forms and some are sesquilinear forms or Hermitian forms. Rather than a general field K, the instances with real numbers R, complex numbers C, and quaternions H are spelled out. The bilinear form

$\sum_{k=1}^p x_k y_k - \sum_{k=p+1}^n x_k y_k$

is called the real symmetric case and labeled R(p, q), where p + q = n. Then he articulates the connection to traditional terminology:^[6]

Some of the real symmetric cases are very important. The positive definite case R(n, 0) is called Euclidean space, while the case of a single minus, R(n−1, 1) is called Lorentzian space. If n = 4, then Lorentzian space is also called Minkowski space or Minkowski spacetime. The special case R(p, p) will be referred to as the split-case.

Relation to tensor products

By the universal property of the tensor product, bilinear forms on V are in 1-to-1 correspondence with linear maps V ⊗ V → K. If B is a bilinear form on V the corresponding linear map is given by

v ⊗ w ↦ B(v, w)

The set of all linear maps V ⊗ V → K is the dual space of V ⊗ V, so bilinear forms may be thought of as elements of

(V ⊗ V)^∗ ≅ V^∗ ⊗ V^∗

Likewise, symmetric bilinear forms may be thought of as elements of Sym²(V^∗) (the second symmetric power of V^∗), and alternating bilinear forms as elements of Λ²V^∗ (the second exterior power of V^∗).

On normed vector spaces

Definition: A bilinear form on a normed vector space (V, ‖·‖) is bounded, if there is a constant C such that for all u, v ∈ V,

$B ( \mathbf{u} , \mathbf{v}) \le C \left\| \mathbf{u} \right\| \left\|\mathbf{v} \right\| .$

Definition: A bilinear form on a normed vector space (V, ‖·‖) is elliptic, or coercive, if there is a constant c > 0 such that for all u ∈ V,

$B ( \mathbf{u} , \mathbf{u}) \ge c \left\| \mathbf{u} \right\| ^2 .$

See also

• Bilinear map
• Bilinear operator
• Inner product space
• Linear form
• Multilinear form
• Quadratic form
• Positive semi-definite
• Sesquilinear form
• Polar space

Notes

1. Jacobson 2009 p. 346
2. Zhelobenko, DmitriÄ­ Petrovich (2006). Principal Structures and Methods of Representation Theory. Translations of Mathematical Monographs. American Mathematical Society. p. 11. ISBN 0-8218-3731-1.
3. Grove 1997
4. Adkins & Weintraub (1992) p.359
5. Harvey p. 22
6. Harvey p 23

References

• Jacobson, Nathan (2009). Basic Algebra I (2nd ed.). ISBN 978-0-486-47189-1.
• Adkins, William A.; Weintraub, Steven H. (1992). Algebra: An Approach via Module Theory. Graduate Texts in Mathematics 136. Springer-Verlag. ISBN 3-540-97839-9. Zbl 0768.00003.
• Cooperstein, Bruce (2010). "Ch 8: Bilinear Forms and Maps". Advanced Linear Algebra. CRC Press. pp. 249–88. ISBN 978-1-4398-2966-0.
• Grove, Larry C. (1997). Groups and characters. Wiley-Interscience. ISBN 978-0-471-16340-4.
• Halmos, Paul R. (1974). Finite-dimensional vector spaces. Undergraduate Texts in Mathematics. Berlin, New York: Springer-Verlag. ISBN 978-0-387-90093-3. Zbl 0288.15002.
• Harvey, F. Reese (1990). "Chapter 2: The Eight Types of Inner Product Spaces". Spinors and calibrations. Academic Press. pp. 19–40. ISBN 0-12-329650-1.
• M. Hazewinkel ed. (1988) Encyclopedia of Mathematics, v.1, p. 390, Kluwer Academic Publishers
• Milnor, J.; Husemoller, D. (1973). Symmetric Bilinear Forms. Ergebnisse der Mathematik und ihrer Grenzgebiete 73. Springer-Verlag. ISBN 3-540-06009-X. Zbl 0292.10016.
• Shilov, Georgi E. (1977). Silverman, Richard A., ed. Linear Algebra. Dover. ISBN 0-486-63518-X.
• Shafarevich, I. R.; A. O. Remizov (2012). Linear Algebra and Geometry. Springer. ISBN 978-3-642-30993-9.

External links

• Hazewinkel, Michiel, ed. (2001), "Bilinear form", Encyclopedia of Mathematics, Springer, ISBN 978-1-55608-010-4
• Bilinear form at PlanetMath.org.

This article incorporates material from Unimodular on PlanetMath, which is licensed under the Creative Commons Attribution/Share-Alike License.