🔢 mathematics

Some uniform effective results on André--Oort for sums of powers in $\mathbb{C}^n$

This paper establishes uniform and effective André–Oort-type bounds for linear combinations of powers of singular moduli in $\mathbb{C}^n$ , and provides a complete unconditional classification for the specific case of three singular moduli summing to a rational number.

Original authors: Guy Fowler

Published 2026-04-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Guy Fowler

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a detective trying to solve a very strange mystery involving a special club of numbers called Singular Moduli.

The Characters: The Special Numbers

Think of these "Singular Moduli" as VIP members of a mathematical club. They aren't just any numbers; they are incredibly rare and special, born from a specific type of geometric shape called an "elliptic curve."

Every VIP member has an ID card called a Discriminant. This ID card is a negative number that tells us exactly how "complex" or "deep" that member is.

Small ID numbers (like -3, -4, -7) are the "local celebrities." They are easy to find, and there are only a few of them.
Huge ID numbers (like -1,000,000) are the "superstars." They are incredibly rare, and the deeper you go, the harder they are to find.

The Mystery: The Magic Equation

The paper investigates a specific rule: Can you mix these VIP members together to get a simple, ordinary number?

Imagine you have three VIPs: $x$ , $y$ , and $z$ . You mix them with some secret ingredients (coefficients $A, B, C$ ) using a power rule (like squaring them or cubing them). The question is:

If you mix them up ( $A \cdot x^m + B \cdot y^m + C \cdot z^m$ ) and the result is a simple, rational number (like 5 or 1/2), how big can the ID cards (discriminants) of $x, y, z$ be?

For a long time, mathematicians knew the answer was "not infinite," but they couldn't put a specific limit on it. It was like saying, "The suspect is somewhere in the universe," without knowing if they are in the next town or in another galaxy.

The Breakthrough: Setting a Speed Limit

Guy Fowler, the author of this paper, acts like a traffic cop. He proves that there is a strict speed limit on how "complex" these VIPs can be if they are to mix together to form a simple number.

He says: "If you see three VIPs mixing to make a simple number, their ID cards cannot be bigger than X. If they are bigger than X, it's mathematically impossible."

This is a huge deal because:

It's Effective: He doesn't just say "there is a limit." He actually calculates what that limit is (or gives a formula to find it).
It's Uniform: The limit works regardless of how many VIPs you have or what power you raise them to.

The "Exceptional" VIP

There is one tiny catch. In the world of these numbers, there is one specific "rogue" field (a special type of mathematical universe) that behaves differently. The paper says: "Unless you have more than one VIP from this specific rogue field, our speed limit holds."

If you have at most one VIP from this rogue field, the speed limit applies perfectly. If you have two or more from that specific field, the math gets tricky, but the paper handles that too by saying, "If they are all the same rogue VIP, we can still bound them."

The Grand Finale: The Case of Three

The paper's most exciting achievement is solving the mystery completely for the case of three VIPs ( $x, y, z$ ) mixed linearly (just adding them up, no powers).

Fowler didn't just set a limit; he found every single possible combination that works. He essentially wrote down a "Wanted Poster" for every trio of these numbers that can mix to make a simple number.

He found that for three numbers to mix successfully, they must fall into one of five specific patterns:

The Boring Trio: All three are already simple numbers (no mystery).
The Twin Pair: One is simple, and the other two are "twins" (conjugates) that cancel each other out perfectly.
The Triplets: All three are twins of each other in a specific way.
The Perfect Balance: All three are distinct but perfectly balanced (like a scale).
The Nested Twins: One is simple, and the other two are a specific type of twin pair with a special relationship.

The Analogy: The Lock and Key

Think of the "Simple Number" (like 5) as a locked treasure chest.
The Singular Moduli are keys.
Usually, if you take three random, complex keys and try to jam them together to open the chest, it won't work. The keys are too big and too weird.

Fowler's paper proves that only very specific, small keys can ever fit together to open that chest. He didn't just say "only small keys work"; he made a mold of every single key shape that could possibly fit, and he showed that if a key is too big (too complex), it physically cannot fit into the lock, no matter how you twist it.

Why Does This Matter?

In the world of mathematics, this is like finding a universal rule for how rare things interact. It helps mathematicians understand the hidden structure of numbers. It's the difference between guessing that a monster lives in the forest and actually having a map that shows exactly where the monster lives and what it looks like.

In short: This paper proves that if you mix these special, rare numbers to get a simple result, they can't be arbitrarily complex. There is a hard ceiling on their complexity, and for the case of three numbers, we now know exactly which ones pass the test.

1. Problem Statement

The paper addresses the André–Oort conjecture in the context of the modular curve $Y(1)$ and its powers $\mathbb{C}^n$ . Specifically, it investigates the distribution of special points (points where all coordinates are singular moduli, i.e., $j$ -invariants of elliptic curves with Complex Multiplication) lying on hypersurfaces defined by linear combinations of powers of singular moduli.

The core problem is to determine effective bounds on the discriminants of singular moduli $x_1, \dots, x_n$ satisfying an equation of the form:
$a_1 x_1^m + \dots + a_n x_n^m = b$
where $a_i, b \in \mathbb{Q} \setminus \{0\}$ and $m \in \mathbb{Z}_{>0}$ .

While the André–Oort conjecture (proven by Pila) guarantees that such points are finite in number, Pila's proof is ineffective (it does not provide a computable bound on the size of the discriminants). Previous effective results were limited to small $n$ (specifically $n \le 2$ ) or required dependence on the height of the coefficients $a_i, b$ . Fowler aims to provide uniform (independent of coefficient height) and effective bounds for general $n$ , and explicitly for the case $n=3$ .

2. Methodology

The author combines tools from Diophantine geometry, complex analysis, and computational number theory.

Pila's Uniform André–Oort Theorem: The paper utilizes Pila's uniform version of the conjecture, which provides an ineffective bound depending on the degree of the field of definition of the coefficients.
Binyamini's Effective Results: The author leverages Binyamini's effective bounds for the André–Oort conjecture, which are effective but depend on the height of the coefficients.
Siegel–Tatuzawa Exceptional Field: A critical component is the handling of the Siegel–Tatuzawa lower bound for class numbers. This bound is effective for all imaginary quadratic fields except possibly one exceptional field, denoted $K^*$ . The paper isolates the difficulty to cases where the discriminants correspond to $K^*$ .
Galois Conjugacy and Determinants: For singular moduli of the same discriminant, the author uses the fact that they form a complete set of Galois conjugates. By constructing determinants involving these conjugates and applying bounds on the size of singular moduli (Proposition 2.2), the author derives contradictions if the discriminants are too large.
Class Number Bounds: The proof relies heavily on explicit bounds for the class number $h(\Delta)$ of imaginary quadratic orders, distinguishing between the "generic" case (where $K \neq K^*$ ) and the "exceptional" case ( $K = K^*$ ).
Computational Verification (PARI/GP): For the specific case of $n=3$ , the author performs extensive computations to enumerate all possible discriminant triples and verify linear independence over $\mathbb{Q}$ .

3. Key Contributions and Results

A. Uniform Effective Bounds (Theorems 1.2 & 1.3)

The paper establishes that for equations over $\mathbb{Q}$ ( $d=1$ ), the discriminants of the singular moduli are effectively bounded, provided a condition on the exceptional field $K^*$ is met.

Theorem 1.2: For any $m, n$ , there exists an effective constant $c(m, n)$ such that if $a_1 x_1^m + \dots + a_n x_n^m = b$ with $x_i$ distinct singular moduli, then $\max |\Delta_i| \le c(m, n)$ , unless there are multiple distinct discriminants $\Delta_i$ such that $\mathbb{Q}(\sqrt{\Delta_i}) = K^*$ . Specifically, the result holds if at most one $\Delta_i$ corresponds to $K^*$ .
- Significance: This removes the dependence on the height of coefficients $a_i, b$ , achieving uniformity.
Theorem 1.3: Under stricter conditions (specifically, if for every $k$ , there is only one $\Delta_i$ such that $\mathbb{Q}(\sqrt{\Delta_i}) = \mathbb{Q}(\sqrt{\Delta_k})$ ), the author provides explicit constants $c_1(n)$ and $c_2(n)$ that are uniform in $m$ .
- This improves upon Binyamini's results by removing the dependency on the height of coefficients and handling the exceptional field $K^*$ more robustly.

B. Intermediate Theorems on Same-Discriminant Moduli

The author proves two auxiliary results of independent interest regarding singular moduli of the same discriminant $\Delta$ :

Theorem 1.4: If $a_1 x_1^m + \dots + a_n x_n^m \in K$ (where $K=\mathbb{Q}(\sqrt{\Delta})$ ), then either $|\Delta|$ is small (bounded by a function of $n$ ) or the class number $h(\Delta) = n$ and the coefficients are equal.
Theorem 1.5: A result on the degree of the field generated by linear combinations of powers of singular moduli of the same discriminant.

C. Complete Classification for $n=3$ (Theorem 1.6)

The paper provides a completely explicit and unconditional classification for the case $m=1, n=3$ .

Theorem 1.6: It lists all triples of distinct singular moduli $(x, y, z)$ such that $Ax + By + Cz \in \mathbb{Q}$ for non-zero rational $A, B, C$ .
The classification covers five cases based on the degrees of the fields $\mathbb{Q}(x), \mathbb{Q}(y), \mathbb{Q}(z)$ $Q (x), Q (y), Q (z)$ and their Galois conjugates:
1. All three are rational.
2. One is rational, the other two generate the same quadratic field with a specific ratio of coefficients.
3. All three generate the same quadratic field with a specific linear relation.
4. All three generate the same cubic field (conjugate over $\mathbb{Q}$ ) with equal coefficients.
5. One generates a quadratic field, the other two generate a quartic field (conjugate over $\mathbb{Q}$ ) with a specific relation.
Computational Aspect: The proof involves checking a finite list of discriminants (derived from class number bounds) using PARI/GP to ensure no other solutions exist.

4. Significance

Effectivity and Uniformity: This is a major step forward in the effective André–Oort program. Previous effective results either depended on the height of the coefficients (making them non-uniform) or were limited to $n \le 2$ . This paper provides uniform effective bounds for general $n$ (with a caveat regarding the exceptional field $K^*$ ) and explicit bounds for $n=3$ .
Resolution of the $n=3$ Case: The paper completely solves the problem of finding all rational linear relations between three singular moduli, a problem previously open for general coefficients.
Handling the Exceptional Field: The paper explicitly addresses the "Siegel–Tatuzawa" obstruction, showing that the only potential failure of uniformity arises from the single exceptional field $K^*$ , and even then, the result holds if the discriminants are not distinct in their associated fields.
Computational Number Theory: The work demonstrates the power of combining deep theoretical bounds (Pila, Binyamini, Tatuzawa) with modern computational verification (PARI) to solve specific Diophantine problems.

In summary, Guy Fowler's paper significantly advances the understanding of linear relations between singular moduli by providing the first uniform effective bounds for general $n$ and a complete explicit classification for $n=3$ , effectively bridging the gap between the qualitative André–Oort conjecture and concrete Diophantine applications.