Bourbaki

Proposition 9

Tag 0DJK, permanent. Cite it as [Bourbaki, Tag 0DJK].

Algebra, chapter IX, § 1. Formes sesquilinéaires

Let AA be a commutative field endowed with an automorphism JJ, Ei,FiE_i, F_i finite-dimensional vector spaces over AA, and Φi\Phi_i a sesquilinear form for JJ on Ei×FiE_i \times F_i (1⩽i⩽m1 \leqslant i \leqslant m). If the forms Φi\Phi_i are non-degenerate, the same is true of their tensor product Φ\Phi. In this case the inverse form Φ^\widehat{\Phi} of Φ\Phi is the tensor product of the inverse forms Φ^i\widehat{\Phi}_i.

Indeed, since AA is a field, it follows from props. 6 and 7 of chap. III, § 1, no. 3 that a tensor product of injective (resp. surjective) linear mappings of AA-modules is an injective (resp. surjective) linear mapping. Since the sΦis_{\Phi_i} are bijective by assumption (prop. 6, no. 6), the same is therefore true of their tensor product. On the other hand the canonical mapping jsj_s of ⨂iFi∗\bigotimes_i F_i^* into (⨂iFi)∗(\bigotimes_i F_i)^* is bijective (chap. III, § 1, no. 5, prop. 11). Therefore, by virtue of (36), sΦs_\Phi is bijective, and this establishes our first assertion (prop. 6, no. 6). Analogously dΦd_\Phi is bijective.

In the second assertion we have implicitly identified ⨂iFi∗\bigotimes_i F_i^* with (⨂iFi)∗(\bigotimes_i F_i)^* and ⨂iEi∗\bigotimes_i E_i^* with (⨂iEi)∗(\bigotimes_i E_i)^* by means of the mappings jsj_s and jdj_d, which are here isomorphisms. The inverse forms in the statement exist since the sΦis_{\Phi_i}, the dΦid_{\Phi_i}, sΦs_\Phi and dΦd_\Phi are bijective (no. 7). Let us then put x′=x1′⊗⋯⊗xm′x' = x'_1 \otimes \cdots \otimes x'_m, y′=y1′⊗⋯⊗ym′y' = y'_1 \otimes \cdots \otimes y'_m (xi′∈Ei∗x'_i \in E_i^*, yi′∈Fi∗y'_i \in F_i^*, i=1,…,mi = 1, \ldots, m). By definition of the inverse forms, and viewed (36), we have

Φ^(js(y′),jd(x′))=Φ(sΦ1−1(y1′)⊗⋯⊗sΦm−1(ym′),dΦ1−1(x1′)⊗⋯⊗dΦm−1(xm′)) \widehat{\Phi}(j_s(y'), j_d(x')) = \Phi(s_{\Phi_1}^{-1}(y'_1) \otimes \cdots \otimes s_{\Phi_m}^{-1}(y'_m), d_{\Phi_1}^{-1}(x'_1) \otimes \cdots \otimes d_{\Phi_m}^{-1}(x'_m))
=∏i=1mΦi(sΦi−1(yi′),dΦi−1(xi′))=∏i=1mΦ^i(yi′,xi′), = \prod_{i=1}^m \Phi_i(s_{\Phi_i}^{-1}(y'_i), d_{\Phi_i}^{-1}(x'_i)) = \prod_{i=1}^m \widehat{\Phi}_i(y'_i, x'_i),

whence our second assertion.

C. Q. F. D.

Let E and F be two modules over the commutative ring A, and Φ\Phi a sesquilinear form for JJ on E×FE \times F. The mapping

(x1,…,xm,y1,…,ym)→det⁡(Φ(xi,yk))(xi∈E,yi∈F,i=1,…,m) (x_1, \ldots, x_m, y_1, \ldots, y_m) \to \det(\Phi(x_i, y_k)) \quad (x_i \in E, y_i \in F, i = 1, \ldots, m)
of Em×(Fj)mE^m \times (F^j)^m into A is A-multilinear. It therefore defines a bilinear form Φ′\Phi' on (⨂mE)×(⨂mFj)(\bigotimes^m E) \times (\bigotimes^m F^j) characterized by
Φ′(x1⊗⋯⊗xm,y1⊗⋯⊗ym)=det⁡(Φ(xi,yk)). \Phi'(x_1 \otimes \cdots \otimes x_m, y_1 \otimes \cdots \otimes y_m) = \det(\Phi(x_i, y_k)).
Since the first member is zero when xi=xkx_i = x_k or when yi=yky_i = y_k (i≠ki \neq k), Φ′\Phi' defines, by passing to the quotients, a bilinear form on (∧mE)×(∧mFj)(\wedge^m E) \times (\wedge^m F^j), or again, since ∧mFj\wedge^m F^j is identified with (∧mF)j(\wedge^m F)^j, a sesquilinear form Φ(m)\Phi_{(m)} for J on (∧mE)×(∧mF)(\wedge^m E) \times (\wedge^m F). This one is characterized by
{Φ(m)(x1∧…∧xm,y1∧…∧ym)=det⁡(Φ(xi,yk))(xi∈E,yi∈F,i=1,…,m). \left\{ \begin{array}{l} \Phi_{(m)}(x_1 \wedge \ldots \wedge x_m, y_1 \wedge \ldots \wedge y_m) = \det(\Phi(x_i, y_k)) \\ (x_i \in E, y_i \in F, i = 1, \ldots, m). \end{array} \right.
(37)

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Nothing.

Translated with machine help, not yet reviewed by a person.