0
$\begingroup$

Are there two normed spaces such that there is no nontrivial bounded linear operator between them: $$\nexists T:X\to Y: T\text{ nontrivial, linear and bounded}$$

$\endgroup$
  • 1
    $\begingroup$ Take a one-dimensional subspace of $Y$ and use a non-trivial bounded linear functional on $X$ to construct a non-trivial bounded operator from $X$ to this subspace. $\endgroup$ – David Mitra May 31 '14 at 15:10
  • $\begingroup$ Only for $X = \{0\}$ or $Y = \{0\}$. $\endgroup$ – Daniel Fischer May 31 '14 at 15:11
  • $\begingroup$ So there always exist nontrivial operators? $\endgroup$ – C-Star-W-Star May 31 '14 at 15:14
  • $\begingroup$ Can you explain the construction? $\endgroup$ – C-Star-W-Star May 31 '14 at 15:15
  • $\begingroup$ If you allow your norm to take value $\infty$, you could endow $Y$ with the (basically useless) norm $$ \| y\| = \begin{cases} 0 & y=0_Y \\ \infty & y \neq 0_Y. \end{cases}$$ In this case, the only bounded linear operator is trivial as long as there is some $0_X \neq x \in X$ such that $\|x\|< \infty$. $\endgroup$ – Tom May 31 '14 at 15:18
2
$\begingroup$

The Hahn-Banach Theorem gives us a wealth of nontrivial bounded linear functionals on (nontrivial) normed linear space $X$. We can for example take the "coordinate" map on a one-dimensional subspace $\{ru | r \in \mathbb{R} \}$, $f(ru)=r$, and extend it to a map $F:X \to \mathbb{R}$ also of the same norm as $f$.

Then another nontrivial bounded map $G:\mathbb{R} \to Y$ can be defined by $G(r) = ry$ for any nonzero vector $y \in Y$. The composition $T = G \circ F$ is then a nontrivial bounded linear transformation from $X$ to $Y$.

$\endgroup$
  • $\begingroup$ As answered in the Comments by David Mitra and Daniel Fischer. $\endgroup$ – hardmath May 31 '14 at 15:40
  • $\begingroup$ This doesn't show that any normed spaces admit nontrivial bounded linear operators! $\endgroup$ – C-Star-W-Star May 31 '14 at 15:49
  • $\begingroup$ Why doesn't it? $\endgroup$ – hardmath May 31 '14 at 15:53
  • $\begingroup$ Oh I'm sorry I overread the last sentence... $\endgroup$ – C-Star-W-Star May 31 '14 at 16:31

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.