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

I would like to know,
Why is it true that $e_n$, the nth unit vector in $\ell^p(\mathbb N)$ converges weakly to $0$. $1 < p < \infty$ According to Mazur's lemma, which says that $y_n$ is convex combination of $e_n$ converges to $0$ in norm topology. I am trying to construct an explicit convex combination but i failed.

How do i construct it, any example would be nice. Thanks.

share|cite|improve this question
You need $1<p<\infty$, here. Take an element $x\in\ell_q$. Note its coordinates converge to $0$. Then, what must $x(e_i)$ converge to? For the convex combinations, consider $y_n={1\over n}\sum_{j=1}^n e_j$. – David Mitra Dec 13 '12 at 17:09
@DavidMitra : Its will give co-ordinate wise convergence to $0$, at every co-ordinate. But i am getting confused because for some $x \in \ell_q $, nth co-ordinate ie 1 will not be zero ? Where is the flaw in what i am thinking. – Theorem Dec 13 '12 at 17:22
You get $\lim_i x(e_i)=\lim_i x_i=0$. This is true for all $x\in\ell_q$, so $(e_i)$ is weakly convergent to $0$ for $1<p<\infty$. When $p=1$, $(e_i)$ is not weakly convergent and any convex combination of them will have norm 1. I'm not sure what happens when $p=\infty$. The $(y_n)$ I defined previously do tend to $0$ in norm. I'm not sure if $(e_i)$ converges weakly or not. – David Mitra Dec 13 '12 at 17:31
@David Mitra : Thank you :) – Theorem Dec 13 '12 at 17:32
up vote 2 down vote accepted

I suppose here $1<p<\infty, and \frac{1}{p}+\frac{1}{q}=1$.

Note that by definition, $e_n\overset{w}{\to}0$ iff $f(e_k)\to 0, \forall f\in (l^p)'=l^q$, so for any $f=(a_k)\in l^q$, we check that $|f(e_k)|=|a_k|\to 0.$

For the second part, just take $y_n=\sum_{k=n+1}^{2n}\frac{1}{n}e_k$, then calculation shows that $||y_n||^p=\sum_{k=n+1}^{2n}\frac{1}{n^p}=\frac{1}{n^{p-1}}\to 0$, since $p>1$.

share|cite|improve this answer
$a_k \to 0$ is because of that fact that if it belongs to $\ell_q$ then it has to converge to $0$ if it is a infinite sequence , and if it is a finite sequence then its obvious . – Theorem Dec 13 '12 at 17:31
@Theorem Yes, you are correct. – ougao Dec 13 '12 at 17:33

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.