Riemann hypothesis - Consequences and Equivalent Criteria

Consequences and Equivalents of the Riemann Hypothesis Introduction The Riemann Hypothesis is one of the most important open problems in mathematics, but its significance lies not just in the hypothesis itself—it lies in the many other statements it implies. This section explores what we would know if the Riemann Hypothesis were true. These consequences range from precise bounds on prime-counting functions to elegant inequalities in number theory. Remarkably, some of these consequences are not merely implied by the hypothesis; they're actually equivalent to it. Proving any one of them would prove the Riemann Hypothesis itself. The Prime Number Theorem and Its Error Term The Prime Number Theorem tells us that the number of primes up to $x$, denoted $\pi(x)$, is asymptotically equal to $\operatorname{li}(x)$, the logarithmic integral. But how good is this approximation for finite values of $x$? The error term depends on the zeros of the Riemann zeta function. More precisely, if $\Theta$ denotes the supremum of the real parts of all non-trivial zeros, then the error satisfies: $$|\pi(x) - \operatorname{li}(x)| = O(x^{\Theta} \log x)$$ What the Riemann Hypothesis gives us: If the hypothesis is true, then $\Theta = \tfrac{1}{2}$. This yields the optimal error bound: $$|\pi(x) - \operatorname{li}(x)| = O(\sqrt{x} \log x)$$ This was proven by von Koch and later refined by Schoenfeld. Crucially, this error bound is the best possible we could hope for—it cannot be improved without knowing something stronger than the Riemann Hypothesis itself. The hypothesis thus gives us the tightest possible control over how the prime-counting function deviates from the logarithmic integral. Precise Bounds on Chebyshev's Function Chebyshev's psi function, denoted $\psi(x)$, is another way to count primes (technically, it counts prime powers with multiplicity). Assuming the Riemann Hypothesis, Schoenfeld proved an explicit bound: $$|\psi(x) - x| \le \frac{1}{8\pi}\sqrt{x}\log^{2}x \quad \text{for all } x \ge 599$$ This shows that $\psi(x)$ stays very close to $x$ for all sufficiently large values. This level of precision would be impossible to prove without assuming the hypothesis. Primes in Short Intervals One beautiful consequence of the Riemann Hypothesis concerns the gaps between consecutive primes. Dudek proved that assuming the hypothesis, we can guarantee the existence of at least one prime in surprisingly short intervals. The consequence: For any $x > 0$, there exists a prime between $x$ and $x + x^{1/2}\log x$. To understand how strong this is, consider that the "average gap" between consecutive primes near $x$ is approximately $\log x$. This result says we can find a prime in an interval of length roughly $x^{1/2}\log x$, which is much shorter than naive spacing would suggest when $x$ is large. The Möbius Function and an Equivalent Formulation The Möbius function $\mu(n)$ is defined as: $\mu(n) = 1$ if $n$ is a product of an even number of distinct primes $\mu(n) = -1$ if $n$ is a product of an odd number of distinct primes $\mu(n) = 0$ if $n$ has a squared prime factor The sum of the Möbius function, $\sum{n \le x} \mu(n)$, oscillates around zero. The Riemann Hypothesis is equivalent to: $$\sum{n \le x} \mu(n) = O(x^{1/2+\varepsilon}) \quad \text{for every } \varepsilon > 0$$ This means the oscillations of the Möbius function are bounded by a function slightly larger than $\sqrt{x}$. This equivalence is important because it means proving any better bound on the sum of the Möbius function would automatically prove the Riemann Hypothesis. Robin's Inequality: A Concrete Equivalence One of the most striking equivalences involves an elementary number-theoretic function. Let $\sigma(n)$ denote the sum of all positive divisors of $n$ (including 1 and $n$ itself), and let $\gamma$ denote the Euler–Mascheroni constant (approximately 0.5772). Robin's Inequality (1984): The Riemann Hypothesis is equivalent to: $$\sigma(n) < e^{\gamma} n \log\log n \quad \text{for all } n > 5040$$ Why is this remarkable? This inequality involves only elementary number theory—just the divisor sum function and logarithms—yet it's equivalent to one of the deepest conjectures in mathematics. Robin proved that the inequality fails for infinitely many $n$ if and only if the Riemann Hypothesis is false. This gives an alternative way to think about the hypothesis: instead of studying zeros of the zeta function, you could try to prove (or disprove) this inequality about divisors. Lagarias's Criterion Another elegant equivalence comes from Lagarias. Define $Hn$ as the $n$-th harmonic number: $$Hn = 1 + \frac{1}{2} + \frac{1}{3} + \cdots + \frac{1}{n}$$ Lagarias's Criterion: The Riemann Hypothesis is true if and only if for every integer $n > 1$: $$\sigma(n) \le Hn + \exp(Hn) \log(Hn)$$ Like Robin's Inequality, this gives an elementary criterion that avoids directly discussing the zeta function or its zeros. These elementary equivalences are valuable because they suggest the hypothesis might be provable using techniques from classical number theory rather than complex analysis. <extrainfo> The Lindelöf Hypothesis The Riemann Hypothesis implies a weaker statement known as the Lindelöf Hypothesis. On the critical line (the line $\operatorname{Re}(s) = \tfrac{1}{2}$), this hypothesis states that for any $\varepsilon > 0$: $$\zeta\left(\tfrac{1}{2} + it\right) = O(|t|^{\varepsilon}) \quad \text{as } |t| \to \infty$$ This bounds the growth of the zeta function on the critical line. The Lindelöf Hypothesis is weaker than the Riemann Hypothesis (it doesn't require all zeros to be on the critical line), but the RH does imply it. In fact, proving the Lindelöf Hypothesis would be progress toward the Riemann Hypothesis, even though the two are not equivalent. </extrainfo> Functional-Analytic Approaches to the Riemann Hypothesis Beyond the number-theoretic consequences, the Riemann Hypothesis has been reformulated using techniques from functional analysis. These approaches connect the zeta function to abstract mathematics in surprising ways. Nyman–Beurling Criterion In 1950, Bertil Nyman discovered a functional-analytic reformulation. Consider functions of the form: $$f(x) = \sum{k=1}^{N} \frac{ck}{k} \, \mathbf{1}{[0,1/k)}(x)$$ where $\mathbf{1}{[0,1/k)}(x)$ is the indicator function (equal to 1 on $[0, 1/k)$ and 0 elsewhere), and the coefficients satisfy $\sum ck = 0$. The Nyman–Beurling Criterion states: The Riemann Hypothesis holds if and only if the set of all such functions is dense in the Hilbert space $L^2(0,1)$. "Denseness" means you can approximate any function in $L^2(0,1)$ arbitrarily well using these special step functions. This connects the zeta function to functional analysis, providing another angle of attack on the problem. Li's Criterion Another approach comes from working directly with the coefficients of the zeta function's logarithmic derivative. Define $\lambdan$ by: $$\lambdan = \sum{\rho} \left[1 - \left(1 - \frac{1}{\rho}\right)^n\right]$$ where the sum runs over all non-trivial zeros $\rho$ of the zeta function. Li's Criterion (1997): The Riemann Hypothesis is equivalent to the statement that $\lambdan > 0$ for every positive integer $n$. This remarkable criterion reduces the Riemann Hypothesis to checking the positivity of infinitely many numbers defined explicitly from the zeros. While this doesn't make the hypothesis easier to prove in practice, it offers a different perspective: the hypothesis is equivalent to a positivity condition rather than a statement about the location of zeros. Summary: Why These Equivalences Matter The existence of these many equivalences tells us something profound about the Riemann Hypothesis. It's not an isolated statement—it's deeply connected to: The distribution of primes and prime-counting functions Elementary divisor-sum inequalities Oscillations of the Möbius function Functional-analytic properties of spaces Explicit inequalities involving harmonic numbers This interconnectedness suggests that the hypothesis, if true, is a fundamental principle about how integers and their divisors behave. Proving any one of these equivalent statements would prove them all, and would give us extraordinary precision in understanding the distribution of primes—one of the central questions in mathematics.