MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

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

For some infinite-dimensional Banach spaces $E$, it is easy to find sequences $\langle x_i:i\in\mathbb N_0\rangle$ which converge to zero weakly but not in the norm topology, i.e. we have $\lim_{i\to\infty}y(x_i)=0$ for all $y\in E^*$ but $\lim_{i\to\infty}\|x_i\|\not=0$. For example, taking $E=c_0(\mathbb N_0)$ or $E=\ell^p(\mathbb N_0)$ with $ 1 < p < +\infty $, the sequence of standard unit vectors provides such an example. For $C([0,1])$, the sequence $\langle\langle (i+1)(1-t)t^{i+1}:0\le t\le 1\rangle:i\in\mathbb N_0\rangle$ is an example. However, I cannot find such an example for $\ell^1(\mathbb N_0)$. So I ask

Are weak and strong convergence of sequences equivalent in $\ell^1(\mathbb N_0)$? What about $L^1([0,1])$? Is there possibly a general result saying that in any infinite-dimensional Banach space weak and strong convergence of sequences are not equivalent?
share|cite|improve this question
up vote 17 down vote accepted

Banach spaces where all weakly convergent sequences are norm convergent are said to have the Schur property. A classical Theorem of Schur says that $\ell^1(I)$ has the Schur property for every set $I$. $L^1([0,1])$ does not have it. I guess that you can find this in P. Wojtaszczyk's "Banach Spaces for Analysts".

share|cite|improve this answer
See also Wikipedia: I can also recommend Albiac and Kalton's book. – Matthew Daws Mar 1 '12 at 15:39
As to $L^1[0,1]$ and more generally $L^1(X,\mu)$ the complete statement is: $f_n$ converges strongly if and only if it converges weakly and in measure. (Of course in a discrete measure space such as $\mathbb{N}$, for a bounded sequence, weak convergence, convergence in measure and pointwise convergence all coincide). Check e.g. Dunford&Schwartz. – Pietro Majer Mar 1 '12 at 17:20
@Pietro Majer: In $\ell^1(\mathbb N_0)$, the sequence of standard unit vectors converges to zero pointwise but neither weakly, nor (globally) in measure. Anyway, thanks for the additional reference. – TaQ Mar 1 '12 at 19:35
Now that Schur has been mentioned, I found from Jarchow's Locally Convex Spaces, Theorem 10.5.2, proof pp. 205−206, which is the result of Schur Jochen Wengenroth referred to, however without the phrase "Schur property". – TaQ Mar 1 '12 at 20:38
I have now also observed that Schur's result is contained in Corollary 4.21.6 (p. 276) in R. E. Edwards' Functional Analysis. An example of a weakly but not in norm to zero convergent sequence in $L^p([0,1])$ for $1\le p<+\infty$ is $\langle[\langle \sin(2\pi(i+1)t):0\le t\le 1\rangle]:i\in\mathbb N_0\rangle$. Thanks again to Pietro Majer for providing refreshment to my memory. – TaQ Mar 3 '12 at 0:52

A remark in H. Brezis, Functional Analysis, Sobolev Spaces and Partial Differential Equations, Chapter 3:

In any infinite dimensional space the weak topology is strictly coarser than the strong topology.

In the same place there are two examples:

  1. The unit sphere $S=\{x \in E : \|x \|=1\}$, with $E$ infinite dimensional, is never closed in the weak topology $\sigma(E,E^*)$

  2. The unit ball $U=\{x \in E : \|x\|<1\}$, with $E$ infinite dimensional is never open in the weak topology $\sigma(E,E^*)$

The proofs of these two facts can be found in the book.

You can find some useful facts and examples in the following documents:

share|cite|improve this answer
This says the topologies are different... It doesn't mean that convergence of sequences is different. – Gerald Edgar Mar 1 '12 at 15:38
@Gerald Edgar: Yes. I realize that now. Thank you for your comment. – Beni Bogosel Mar 1 '12 at 17:23

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.