Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

This problem has been giving me some troubles. Does anyone have any ideas on how to go about proving this?

Let $X$ and $Y$ be Banach spaces. If $T: X \to Y$ is a linear map such that $f \circ T \in X^*$ for every $f \in Y^*$, then $T$ is bounded.

Thanks in advanced!

share|cite|improve this question
up vote 2 down vote accepted

Let $A=\{f\circ T \mid f\in Y^* \land \Vert f \Vert = 1\}$ (so $A\subset X^*$). Let $x\in X$. Then ${|(f\circ T)x|}\leq {\Vert f \Vert} {\Vert Tx \Vert}$, so $$ \sup_{\phi\in A} |\phi(x)| \leq \Vert Tx\Vert $$ which is finite. By the uniform boundedness principle, we have $$\sup_{\phi\in A} \Vert \phi \Vert<\infty $$ and $$ \begin{align*} \sup_{\phi\in A} \Vert \phi \Vert &= \sup_{\substack{f\in Y^* \\ \Vert f \Vert = 1}} \Vert f\circ T\Vert\\ &= \sup_{\substack{f\in Y^* \\ \Vert f \Vert = 1}} \left(\sup_{\substack{x\in X \\ \Vert x \Vert = 1}} |(f\circ T)x|\right)\\ &= \sup_{\substack{x\in X \\ \Vert x \Vert = 1}}\left(\sup_{\substack{f\in Y^* \\ \Vert f \Vert = 1}} |(f\circ T)x|\right) \end{align*} $$

But $$ \forall x\in X,\; \sup_{\substack{f\in Y^* \\ \Vert f \Vert = 1}} |(f\circ T)x| \ge \Vert Tx \Vert, $$ therefore $$ \Vert T \Vert = \sup_{\substack{x\in X \\ \Vert x \Vert = 1}} \Vert Tx \Vert < \infty. $$ So $T$ is bounded.

share|cite|improve this answer

By the uniform boundedness principle applied to the set of all $f \circ T$ with $|f| = 1$ (operator norm from $Y$ to ${\bf C}$), there are two possibilities: 1) there is a constant $C$ such that $|f(T(x))| \leq C||x||$ for all $x \in X$ and all $f$ with $|f| = 1$, or 2) there exists at least one $x$ such that ${\displaystyle \sup_{|f| = 1} |f(T(x))| = \infty}$.

The second option cannot hold since $|f(T(x))| \leq ||T(x)||$ whenever $|f| = 1$. So the first option must occur; there is a constant $C$ such that $|f(T(x))| \leq C||x||$ for all $x \in X$ and all $f$ with $|f| = 1$. By the Hahn-Banach theorem, for any $y = T(x) \in Y$ one can create an $f$ with $|f| = 1$ such that $|f(T(x))| = ||T(x)||$. So we have $||T(x)|| \leq C||x||$.

share|cite|improve this answer

First, I claim that the unit ball of $Y^*$ is mapped into a bounded subset of $X^*$. This follows from the Banach-Steinhaus theorem. If $x \in X$, then $(f \circ T)(x) = f(T(x))$ is bounded as $f$ ranges over the elements of $Y^*$ of norm at most one. So we have the collection $\mathcal{C}$ of functionals $f \circ T$ on $X$, such that for each $x \in X$, $\sup_{r \in \mathcal{C}} ||r(x)|| < \infty$. This implies that $\mathcal{C}$ is a bounded set and that the transpose of $T$ is bounded.

Now, if the transpose of a linear transformation $T$ is bounded by some $C$, then $T$ is itself bounded by $C$ (to see this, suppose $x \in X$ is of norm at most one; then the claim is that $|\ell(T(x))| \leq C$ for $\ell$ a functional on $Y$ of norm at most one, which is equivalent by Hahn-Banach. But this is $ |T^*(\ell)(x)|$, which by assumption is of norm at most $C$).

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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