6
$\begingroup$

Let $X$ be an infinite-dimensional Banach space and let $p\in(1,+\infty)$. We may define $L^p(\mathbb R;X)$. Is it always true that the topological dual of $L^p(\mathbb R;X)$ is $L^{p'}(\mathbb R;X^*)$? Maybe some reflexivity or separability is needed for the Radon-Nikodym argument to work. Is there a simple realization of the dual of $L^1(\mathbb R;X)$?

$\endgroup$
1
  • $\begingroup$ Typo: you mean $L^{p'}({\mathbb R}, X^*)$, right? $\endgroup$
    – Yemon Choi
    Jan 28, 2015 at 17:58

1 Answer 1

9
$\begingroup$

Diestel-Uhl, Vector measures, Section IV.1., Theorem 1:

Let $(\Omega,\mu)$ be a $\sigma$-finite measure space, $1\leq p<\infty$ and $\frac{1}{p}+\frac{1}{p'}=1$.

The dual of $L^p(\mu;X)$ is $L^{p'}(\mu,X^\ast)$ if and only if $X^\ast$ has the Radon-Nikodym property.

This is especially the case if $X$ is reflexive.

$\endgroup$
9
  • 1
    $\begingroup$ Just to clarify: presumably there is a missing quantifier, and the full result is: the dual has the desired form for all $\sigma$-finite $(\Omega,\mu)$ if and only if $X^*$ has RNP? $\endgroup$
    – Yemon Choi
    Jan 28, 2015 at 18:03
  • $\begingroup$ Yes, I add the details to the answer. Thank you. $\endgroup$ Jan 28, 2015 at 18:04
  • $\begingroup$ Another important special case (where $X^*$ has the RNP) is when $X^*$ is separable. For example, if $X$ is separable and $X^*$ is not separable, we would expect the dual of $L^{p}(\mu, X)$ to include some non-Bochner-measurable functions. $\endgroup$ Jan 28, 2015 at 18:25
  • $\begingroup$ What is the Radon-Nikodym property? Is it easy to say? $\endgroup$ Jan 29, 2015 at 11:14
  • 2
    $\begingroup$ That Diestel--Uhl is a fine book. Have a look. $\endgroup$ Jan 29, 2015 at 13:09

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