2

I seem to recall that the prime number theorem (PNT) is equivalent to the fact that the Riemann zeta function $\zeta(s)$ is non-zero on all of $\text{Re}(s) = 1$ (see https://math.stackexchange.com/questions/1379583/why-is-zeta1it-neq-0-equivalent-to-the-prime-number-theorem or https://math.stackexchange.com/questions/706934/riemann-zeta-function-non-vanishing-on-the-line-mathrmre-z-1), where I suppose "equivalent" means that PNT and the non-vanishing of $\zeta(s)$ along $\text{Re}(s)=1$ can be both be proven in a relatively short manner from the other.

Is it the case that Dirichlet's theorem on primes in arithmetic progressions is also equivalent to the fact that all the Dirichlet $L$-functions $L(s,\chi)$ are non-zero at $s=1$, in any sense similar to the equivalence described above? Perhaps another way of phrasing the question is: is it "possible" that $L(s,\chi)$ could equal 0 (for some non-trivial character $\chi$) but still Dirichlet's theorem could hold?

It has come to my attention that this has already been discussed here Analytic equivalents for primes in arithmetic progressions. Apologies all, and thanks for your comments.

GH from MO
  • 98,751
D.R.
  • 741
  • 7
    For Dirichlet's theorem on the infinitude of primes in arithmetic progressions, you only need that $L(1,\chi) \neq 0$ for each Dirichlet character $\chi$. For the prime number theorem in arithmetic progressions, you need $L(1 + it,\chi) \neq 0$ for all $t \in \mathbb{R}$. – Peter Humphries Sep 11 '21 at 04:08
  • 1
    To clarify the previous comment, “Dirichlet’s theorem for primes in arithmetic progressions” means ${p : p \equiv a \bmod m}$ is infinite whenever $\gcd(a,m) = 1$, where $p$ runs over primes, and “the prime number theorem for arithmetic progressions” means $|{p\leq x:p\equiv a \bmod m} \sim (1/\varphi(m))x/\log x$ as $x\to\infty$ whenever $\gcd(a,m)=1$, where $p$ runs over primes. – KConrad Sep 11 '21 at 06:12
  • @PeterHumphries thank you and KConrad for your comments, but I did not mean "a version of Dirichlet's theorem for arithmetic progressions (DTAP) in the form of the PNT" (which would be the result you mentioned); I meant "a version of the equivalence between DTAP and the non-vanishing of the $L$-functions at $s=1$ in the form of the equivalence between PNT and the non-vanishing of $\zeta(s)$ along $\text{Re}(s)=1$". – D.R. Sep 11 '21 at 06:47
  • Unless I'm missing something, the argument showing that the usual prime number theorem implies the nonvanishing of zeta should apply mutatis mutandis to showing that the PNT for APs in the form $\sum_{n\leqslant x}\Lambda(n)\chi(n)=o(x)$ is equivalent to the nonvanishing of $L(1+it,\chi)$... – Anurag Sahay Sep 11 '21 at 07:04
  • @AnuragSahay this is promising. What about for just DTAP, not PNT for APs? – D.R. Sep 11 '21 at 07:25
  • 4
    I don't think they are quite equivalent in any reasonable way. Nonvanishing of the L-functions implies not just infinitude of primes in APs, but also that their logarithmic/Dirichlet density is positive. One plausible way to make this into a formal statement is to say that for general L-functions (say in a Selberg class) nonvanishing at a point is equivalent to some density result. An analogous result holds for PNT statements, and I believe is due to Kaczorowski et al. – Wojowu Sep 11 '21 at 15:41
  • 3
    @Wojowu please see my answer below. There is no need to prove a sweeping claim about everything in the Selberg class to show a certain density version of Dirichlet's theorem is equivalent to nonvanishing of certain Dirichlet $L$-functions at $1$. Note that the equivalence is not with mere positivity of the Dirichlet density, but with the expected value of it (namely $1/\varphi(m)$). – KConrad Sep 11 '21 at 21:45
  • Please use a high-level tag like "nt.number-theory". I added this tag now. – GH from MO Sep 11 '21 at 22:04

1 Answers1

9

Theorem: Fix a positive integer $m$. The following two conditions are equivalent:

(1) $L(1,\chi) \not= 0$ for all nontrivial Dirichlet characters $\chi \bmod m$.

(2) For all $a \in (\mathbf Z/m\mathbf Z)^\times$, the set of primes $\{p \equiv a \bmod m\}$ has Dirichlet density $1/\varphi(m)$.

Proof. When $m$ is $1$ or $2$, the condition (1) is vacuously true (there are no nontrivial Dirichlet characters mod $m$) and condition (2) is obvious, so from now on we can let $m \geq 3$.

The usual proof of Dirichlet's theorem shows (1) implies (2). It remains to show (2) implies (1).

For each Dirichlet character $\chi \bmod m$, $L(s,\chi)$ is analytic for ${\rm Re}(s) > 0$ except that $L(s,\mathbf 1_m)$ has a simple pole at $s = 1$, where $\mathbf 1_m$ denotes the trivial character mod $m$ (we have $L(s,\mathbf 1_m) = \zeta(s)\prod_{p \mid m} (1-1/p^s)$ and $\zeta(s)$ is analytic on ${\rm Re}(s) > 0$ except for a simple pole at $s = 1$). Set $$ r_\chi := {\rm ord}_{s=1}(L(s,\chi)) $$ so $r_{\mathbf 1_m} = -1$ (the simple pole at 1) and $r_\chi \geq 0$ for all nontrivial $\chi$. Condition (1) is equivalent to $r_\chi = 0$ for all nontrivial $\chi$, so we want to show condition (2) implies the numbers $r_\chi$ vanish for all nontrivial $\chi$.

For a Dirichlet character $\chi \bmod m$ and ${\rm Re}(s) > 1$, define $$ \log L(s,\chi) := \sum_{p,k} \frac{\chi(p^k)}{kp^{ks}} = \sum_p \frac{\chi(p)}{p^s} + \sum_{p,k\geq 2} \frac{\chi(p^k)}{kp^{ks}}. $$ This is a logarithm of $L(s,\chi)$ (meaning the exponential of that series is $L(s,\chi)$. The second series on the right is absolutely convergent for ${\rm Re}(s) > 1/2$, with absolute value bounded by $\sum_{p,k \geq 2} 1/(kp^{k\sigma})$, so for $s > 1$ we can say $$ \log L(s,\chi) = \sum_p \frac{\chi(p)}{p^s} + O(1), $$ where the $O$-constant is $\sum_{p,k \geq 2} 1/(kp^k)$. In the usual proof of Dirichlet's theorem, for each $a \in (\mathbf Z/m\mathbf Z)^\times$ and $s > 1$ we write $$ \sum_{p \equiv a \bmod m} \frac{1}{p^s} = \frac{1}{\varphi(m)}\sum_{\chi \bmod m} \overline{\chi}(a)\left(\sum_p \frac{\chi(p)}{p^s}\right), $$ so $$ \sum_{p \equiv a \bmod m} \frac{1}{p^s} = \frac{1}{\varphi(m)}\sum_{\chi \bmod m} \overline{\chi}(a)\log L(s,\chi) + O(1), $$ where the $O$-constant on the right is an overall term (outside the sum).

Now let's bring in the order of vanishing $r_\chi$. For all $s$ near $1$, $L(s,\chi) = (s-1)^{r_\chi}f_\chi(s)$ where $f_\chi(s)$ is a holomorphic function in a neighborhood of $s = 1$ (in fact it is holomorphic on ${\rm Re}(s) > 0$, or even $\mathbf C$) and $f_\chi(1) \not= 0$. Therefore $f_\chi(s)$ has a logarithm around $s = 1$ (well-defined up to adding an integer multiple of $2\pi$), so for $s > 1$ $$ \log L(s,\chi) = r_\chi\log(s-1) + \ell_{f_\chi}(s) $$ where $\ell_{f_\chi}(s)$ is a suitable logarithm of $f_\chi(s)$. Thus $\log L(s,\chi) = r_\chi\log(s-1) + O(1)$ for $s$ near $1$ to the right, and plugging this into the formula for $\sum_{p \equiv a \bmod m} 1/p^s$ we can say $$ \sum_{p \equiv a \bmod m} \frac{1}{p^s} = \frac{1}{\varphi(m)}\sum_{\chi \bmod m} \overline{\chi}(a)(r_\chi\log(s-1)) + O(1). $$ Let's extract the term for the trivial character mod $m$: since $r_{\mathbf 1_m} = -1$, $$ \sum_{p \equiv a \bmod m} \frac{1}{p^s} = -\frac{1}{\varphi(m)}\log(s-1) + \frac{1}{\varphi(m)}\left(\sum_{\chi \not= \mathbf 1_m} \overline{\chi}(a)r_\chi\right)\log(s-1) + O(1). $$

In order to bring in a Dirichlet density, we want to divide both sides by $\sum_p 1/p^s$ for $s$ near $1$ to the right. For such $s$, $$ \log \zeta(s) = -\log(s-1) + O(1) $$ from the simple pole of $\zeta(s)$ at $s = 1$ and $$ \log \zeta(s) = \sum_p \frac{1}{p^s} + O(1) $$ from the Euler product for $\zeta(s)$ when $s > 1$. Therefore $\sum_p 1/p^s = -\log(s-1) + O(1)$ as $s \to 1^+$, so $-\log(s-1) \sim \sum_p 1/p^s$ as $s \to 1^+$. Dividing through the last (big) formula above for $\sum_{p \equiv a \bmod m} 1/p^s$ by $\sum_p 1/p^s$ and letting $s \to 1^+$, we get $$ \frac{\sum_{p \equiv a \bmod m} 1/p^s}{\sum_p 1/p^s} \to \frac{1}{\varphi(m)} - \frac{1}{\varphi(m)}\left(\sum_{\chi \not= \mathbf 1_m} \overline{\chi}(a)r_\chi\right) $$ as $s \to 1^+$. So we have shown, without assuming condition (1) in the theorem, that for all $a \in (\mathbf Z/m\mathbf Z)^\times$ the set of primes $\{p \equiv a \bmod m\}$ has Dirichlet density $$ \frac{1}{\varphi(m)}\left(1 - \sum_{\chi \not= \mathbf 1_m} \overline{\chi}(a)r_\chi\right). $$

Finally it is time to assume condition (2) in theorem, which implies $$ \sum_{\chi \not= \mathbf 1_m} \overline{\chi}(a)r_\chi = 0 $$ for all $a \in (\mathbf Z/m\mathbf Z)^\times$. When $\chi = \mathbf 1_m$, $\overline{\chi}(a)r_\chi = 1(-1) = -1$, so condition (2) implies $$ \sum_{\chi} \overline{\chi}(a)r_\chi = -1 $$ for all $a \in (\mathbf Z/m\mathbf Z)^\times$, where the sum runs over all Dirichlet characters mod $m$ (including the trivial character). We want to show the above equation, for all $a$, implies $r_\chi = 0$ for all nontrivial $\chi \bmod m$.

Using vectors indexed by the Dirichlet characters mod $m$, let $\mathbf r_m = (r_\chi)_\chi$ and $\mathbf v_a = (\chi(a))_\chi$ for each $a \in (\mathbf Z/m\mathbf Z)^\times$. The space of all complex vectors $\mathbf z = (z_\chi)_\chi$ has a Hermitian inner product $\langle \mathbf z, \mathbf w\rangle = \frac{1}{\varphi(m)}\sum_{\chi} z_\chi\overline{w_\chi}$ for which the vectors $\mathbf v_a$ are an orthonormal basis by the usual orthogonality of Dirichlet characters mod $m$. The equation $\sum_\chi \overline{\chi}(a)r_\chi = -1$ above says $\langle \mathbf r_m,\mathbf v_a\rangle = -1/\varphi(m)$ for all $a$ in $(\mathbf Z/m\mathbf Z)^\times$, so $$ \mathbf r_m = \sum_{a} \langle \mathbf r_m,\mathbf v_a\rangle\mathbf v_a = -\frac{1}{\varphi(m)}\sum_{a}\mathbf v_a. $$ For nontrivial Dirichlet characters $\chi \bmod m$, the $\chi$-component of $\sum_{a} \mathbf v_a$ is
$\sum_a \chi(a)$, which is $0$ (the $\mathbf 1_m$-component is $\varphi(m)$, but that's irrelevant). Since $\mathbf r_m$ has $\chi$-component $r_\chi := {\rm ord}_{s=1}L(s,\chi)$, we have $r_\chi = 0$ for all nontrivial $\chi$, so $L(1,\chi) \not= 0$ for all nontrivial $\chi$. QED

KConrad
  • 49,546
  • 1
    Nice answer. But I'd like to revisit Wojowu's remark about the Selberg class. I always feel some vague discomfort with claims that two theorems are equivalent. Not that I think that there's anything wrong with such a claim, but I'm always left wondering if there's some stronger sense of equivalence that can be proved. In the example at hand, I wonder if the equivalence can be proved for various kinds of fake integers? – Timothy Chow Sep 12 '21 at 03:24
  • 1
    @TimothyChow do you have discomfort now with the statement of the Riemann Hypothesis (about nontrivial zeros lying on a line) being proved "to be equivalent" to each of a long list of other statements (a bound on $|\pi(x) - {\rm Li}(x)|$, a bound on the growth of $\sum_{n \leq x} \mu(n)$, a half-plane of convergence of $\sum_{n \geq 1} \mu(n)/n^s$, etc.), and if not would you have discomfort with such claims of equivalence in a hypothetical future where RH is proved? – KConrad Sep 12 '21 at 04:29
  • 1
    I don't have discomfort now with such statements. The discomfort would kick in after RH is proved. A claim that "RH is equivalent to X" is implicitly a claim that the equivalence is easier to prove than "Axioms for math imply RH." As long as nobody knows how to prove "Axioms for math imply RH," we have a more or less objective basis for claiming that "RH is equivalent to X" is easier. Once RH is proved, it becomes a judgment call. Admittedly, we expect that RH will be very hard to prove; if so, then the judgment call may be a no-brainer. But there is still a qualitative difference. – Timothy Chow Sep 12 '21 at 11:49
  • But this is a bit of a digression. In the case at hand, it's not the distinction between a conjecture and a theorem that's at play. For the sake of argument, let's say that I had replied that I experienced discomfort with "RH is equivalent to X." Then the question still stands whether a stronger statement can be proved in the case at hand. – Timothy Chow Sep 12 '21 at 11:54
  • Maybe a better example to have used than equivalences with RH would be equivalences for an elliptic curve over $\mathbf Q$ to be modular. See the 6 conditions in the middle of the third page of the paper proving the modularity theorem (https://www.ams.org/journals/jams/2001-14-04/S0894-0347-01-00370-8/S0894-0347-01-00370-8.pdf). The equivalence among those conditions was known before modularity could be proved outside of individual examples (i.e., pre-Wiles). [contd] – KConrad Sep 12 '21 at 16:24
  • Even after that paper established that one (and hence all) of those equivalent conditions is provable from "axioms for math", most people understand things like "these two conditions are equivalent" in a down to earth way: you can derive each condition from the other by "really using" the assumption of each condition being true to show the other condition is true without needing to prove either condition is true along the way: they are just hypotheses. [contd] – KConrad Sep 12 '21 at 16:25
  • Or consider a much simpler example: for all positive integers $a$ and $b$, the conditions $(a,b) = 1$ and $(a,b)[a,b] = ab$ are equivalent. I don't think anyone would question whether such a proof of equivalence is meaningful because we can't prove either condition really holds (obviously: the equivalence holds for all $a$ and $b$, whether or not they are actually relatively prime). Whether or not the conditions actually hold is a separate mathematical issue from showing the conditions imply each other. [contd] – KConrad Sep 12 '21 at 16:25
  • Concerning your question about whether there is a stronger statement of equivalence between $L$-function nonvanishing and Dirichlet density calculations by proving a theorem of that type in a broader context, my only comment is that knowing a result is true more broadly could be losing some information, namely that the equivalence in the original case of interest only has to use the characters mod $m$ and the arithemtic progressions of primes mod $m$ for one specific $m$ at a time. [contd] – KConrad Sep 12 '21 at 16:25
  • That is, you don't need to prove a result for a bigger collection of objects in the way you sometimes are only able to prove a theorem by proving a theorem for a much bigger set of objects than the original theorem is about (in order to have technical flexibility in the course of the argument). [Sorry for the long chain of comments!] – KConrad Sep 12 '21 at 16:25