This is Lemma 1.4.4 from Scharlau's book Quadratic and Hermitian Forms.

Let $GL(V)=Aut(V)$ denote the group of invertible endomorphisms of $V$. Further one can check that $$\mathbb H \colon GL(V) \to O(\mathbb H(V)), \qquad \alpha\to \alpha\oplus(\alpha^*)^{-1}$$ is an injective group homomorphism.

enter image description here

$H$ is hyperbolic space on $V$ and one can define bilinear form on this like below:

Let $V$ be a vector space and $V^*$ its dual space. On the vector space $V\oplus V^*$ we consider the following symmetric bilinear form: \begin{gather*} h=h_V\colon (V\oplus V^*)\times(V\oplus V^*) \to K\\ h((x,f),(y,g)):= fy+gx \end{gather*}

enter image description here

Also, I know from this lemma that

4.4. Lemma. Let $\alpha\colon V\to W$ be a vector space isomorphism and $\alpha^*\colon W^*\to V^*$ be the dual isomorphism. Then $$\mathbb H(\alpha):= \alpha\oplus(\alpha^*)^{-1} \colon \mathbb H(V) \to \mathbb H(V)$$ is a bijective isometry.

enter image description here

Question is how to show it is an injective group homomorphism. First of all how to show $O(H(V))$ is a group.


Why is $O(\mathbb H(V))$ a group? If you already know that $\mathbb H(V)=(V\oplus V^*, \mathbb h_V)$ is a bilinear space, then this follows from the fact that $O(V,b)$ is a group for any bilinear space $(V,b)$.

$\mathbb H(\alpha)$ belongs to $O(\mathbb H(V))$. The fact that $\mathbb H(\alpha)\in O(\mathbb H(V))$ for $\alpha\in GL(V)$ is exactly the claim in Proposition 4.4.

$\mathbb H$ is a group homomorphism. We just want to show that $\mathbb H(\beta\circ\alpha)=\mathbb H(\beta)\circ \mathbb H(\alpha)$ for any $\alpha,\beta\in GL(V)$. We just compute \begin{align*} (\beta\circ\alpha)\oplus ((\beta\circ\alpha)^*)^{-1} &= (\beta\circ\alpha)\oplus (\alpha^*\circ\beta^*)^{-1} \\ &= (\beta\circ\alpha)\oplus ((\beta^*)^{-1}\circ(\alpha^*)^{-1}) \\ &= (\beta,(\beta^*)^{-1})\circ(\alpha,(\alpha^*)^{-1}) \end{align*}

$\mathbb H$ is injective. It suffices to check that $\mathbb H$ has trivial kernel. Clearly, $$\mathbb H(V)=\alpha\oplus(\alpha^*)^{-1}=id_V\oplus id_{V^*}$$ implies $\alpha=id_V$.

  • $\begingroup$ How the third step in homomorphism is true? $\endgroup$ – maths student Jul 12 '18 at 7:25
  • $\begingroup$ By "the third step" do you mean $(\beta\circ\alpha)\oplus ((\beta^*)^{-1}\circ(\alpha^*)^{-1}) = (\beta,(\beta^*)^{-1})\circ(\alpha,(\alpha^*)^{-1})$? $\endgroup$ – Martin Sleziak Jul 12 '18 at 7:28
  • $\begingroup$ yes that step only. $\endgroup$ – maths student Jul 12 '18 at 7:29
  • $\begingroup$ Is it just component wise breaking of direct sum? $\endgroup$ – maths student Jul 12 '18 at 7:31
  • $\begingroup$ Perhaps (depending on your notation and on how you interpret $\oplus$) you might prefer the notation $\beta\oplus(\beta^*)^{-1}$ instead of $(\beta,(\beta^*)^{-1})$. $\endgroup$ – Martin Sleziak Jul 12 '18 at 7:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.