The following question came up while I was working through an example:

Does there exist an $\ell^1$ sequence of complex numbers $a_n$, not all zero, such that $\sum_n a_n n^{-p} = 0$ for all $p \in [0,1]$?

  • $\begingroup$ You mean, aside from the zero sequence? $\endgroup$ Mar 8, 2013 at 11:20

1 Answer 1


Let $U\subset \Bbb C$ be the set of complex number with positive real part and $f\colon U\to\Bbb C$ given by $f(z):=\sum_{n=1}^{+\infty}a_n\exp(-z\log n)$. Since there is uniform convergence on compact subsets of $\Bbb C$, $f$ is holomorphic. We have $f(z)=0$ if $z\in (0,1]$, hence $f$ vanishes identically on the connected set $U$. In particular, the initial assumption is valid for any $p>0$.

Then we can prove by induction that $a_n=0$ for all $n$.

We have $|a_1|=\left|\sum_{n\geqslant 2}a_n\exp(-p\log n)\right|\leqslant \exp(-p\log 2)\lVert a\rVert_{\ell^1}$ for each $p$, hence $a_1=0$.

Assume that for $n\geqslant 2$, $a_0=\dots=a_{n-1}=0$, then $$|a_n| \leqslant \exp\left(-p\log \frac{n+1}{n}\right)\lVert a\rVert_{\ell^1},$$ giving what we want.

Note that we can relax the initial assumption "for all $p\in [0,1]$" by for "all $p$ in an non-discrete subset of the unit interval".

  • $\begingroup$ That's a very nice solution. Thank you. $\endgroup$
    – Andre
    Mar 9, 2013 at 6:45
  • $\begingroup$ You are welcome. (and I thank you for the interesting problem) In which context did it appear? $\endgroup$ Mar 9, 2013 at 8:56

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.