MathOverflow is a question and answer site for professional mathematicians. 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

How is the proof that $$[L^2(0,T;X)]' = L^2(0,T;X')$$ looking like, where $X$ is a Hilbert space? I am asking for the proof that the dual space of $L^2(0,T;X)$ is the space $L^2(0,T;X^*)$.

Is the proof much different to the $L^p(a,b)$ case?

I can't find any easy to understand proofs.

share|cite|improve this question
I am curious: what is what? – Włodzimierz Holsztyński May 16 '13 at 18:07
@Wlodzimierz I added some details. These are Bochner spaces. – lollypop May 16 '13 at 18:09
I am a bit confused. If $X$ is a Hilbert space, then $L^2(0,T;X)$ is a Hilbert space (complete + norm comes from a scalar product). Hence if you also identify $X$ with its dual (as you do with $L^2$), then the statement follows. Maybe this is not what you ask? – András Bátkai May 16 '13 at 19:09
@Andras you are right. But I am interested in more general $X$, but your answers caters for this. Thanks. – lollypop May 17 '13 at 19:23
@AndrasBatkai Is it as simple as: $L^2(0,T;X)' = L^2(0,T;X) = L^2(0,T;X')$? Where the first equality is by Riesz representation theorem (RRT) for the Bochner space (which is Hilbert), and the second is by RRT for $X$ which is also Hilbert. Would this be a good proof?? (The longer way to do this requires us to show that the map is isometric, which seems difficult to do even in this case.) – lollypop May 18 '13 at 14:29
up vote 6 down vote accepted

To give you a reference: Diestel-Uhl, Vector measures, page 98, Chapter 4, Theorem 1:

$$L^p(\mu,X)^\ast = L^q(\mu,X^\ast)$$

if and only if $X^\ast$ has the Radon-Nikodym property with respect to $\mu$.

Here $\mu$ is a finite measure and $p$ and $q$ as usual. Proof can be read there.

ADDED: It should be noted that $L^q(\mu,X^\ast) \subset L^p(\mu,X)^\ast $ always holds, without any condition on the Banach space $X$. The proof of this is quite the same as in the scalar case.

share|cite|improve this answer
You should also mention that reflexive Banach spaces (hence, Hilbert spaces) possesses the Radon-Nikodym property. – gerw May 17 '13 at 19:03

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.