9
$\begingroup$

Is the number $$\sum_{p\text{ prime}}p^{-2}$$ known to be irrational?

The limit exists, since $$\sum_{p\text{ prime}}p^{-2}<\sum_{i=1}^{\infty}i^{-2}=\frac{\pi^{2}}{6}$$.

$\endgroup$
10
  • 5
    $\begingroup$ relevant answer here: mathoverflow.net/questions/53443/… $\endgroup$ – Francesco Polizzi Jun 15 '16 at 15:58
  • 10
    $\begingroup$ No it does not, but that's because a fool can ask more questions than 1000 wise men can answer. $\endgroup$ – Franz Lemmermeyer Jun 15 '16 at 16:10
  • 5
    $\begingroup$ I would consider this question as part of transcendental number theory (as one really expects this number to be transcendental). The main takeaway from Franz's last comment is that in fact very little is known in this area: we don't even know whether $e - \pi$ is irrational. But the question might be answered conditionally upon assuming some standard conjecture in the area (e.g. Schanuel's conjecture), just as one sometimes proves things conditional on the truth of the Riemann hypothesis if no unconditional proof is known. Is this question important to you? $\endgroup$ – Todd Trimble Jun 15 '16 at 16:20
  • 6
    $\begingroup$ If it's important to you, then could you please give some details why? Giving some context can only improve this question, and indicate why the question is not just one of idle curiosity (that any "fool" could ask, harking back again to Franz's comment). $\endgroup$ – Todd Trimble Jun 15 '16 at 16:31
  • 2
    $\begingroup$ Here might be my own question on this: is this number known to be a period (in the sense used by Kontsevich and Zagier)? I'm going to quote from their paper ihes.fr/~maxim/TEXTS/Periods.pdf (page 16): "... many, if not almost all proofs of irrationality and transcendence results use periods and their associated differential equations in one form or another". $\endgroup$ – Todd Trimble Jun 15 '16 at 16:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.