Let $X$ and $Y$ be normed spaces and $T:X\to Y$ a bounded linear operator. The adjoint operator $T^\times:Y'\to X'$ of $T$ is defined by $$(T^\times g)(x)=g(Tx),\,g\in Y'$$ where $X'$ and $Y'$ are the dual spaces of $X$ and $Y$, respectively. I'm trying to prove that $\|T^\times\|=\|T\|$. I already got that $\|T^\times\|\leq\|T\|$, but I'm struggling to show that $\|T^\times\|\geq\|T\|$. In Kreyszig's Functional Analysis book, the author uses the following argument:

enter image description here

However, Theorem 4.3-3 (which resulted from the Hahn-Banach Theorem) states that:

If $X$ is a normed space and $x_0\neq 0$ is any element of $X$, then there exists a bounded linear functional $\tilde{f}$ on $X$ such that $$\|\tilde{f}\|=1\quad \text{and} \quad \tilde{f}(x_0)=\|x_0\|.$$

We know $x_0\neq 0$ does not imply that $T(x_0)\neq 0$, so how is this theorem applicable in this case?


1 Answer 1


You have then $$||T^*||=\sup_{||g||=1} ||Tg||=\sup_{||g||=1} \sup_{||x||=1}|(Tg)(x)|=\sup_{||g||=1} \sup_{||x||=1}|g(Tx)|\geq \sup_{||g||=1} \sup_{||x||=1}||Tx||=\sup_{||x||=1}||Tx||=||T||.$$ When $Tx=0$ then of course for any functional $g$ we have $g(Tx) =0 =||Tx||, $ so the above consideration is true.


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.