Additive Rigidity for $x$-Coordinates of Rational Points on Elliptic Curves

This paper establishes that the number of rational points on an elliptic curve whose $x$-coordinates lie in a $d$-dimensional generalized arithmetic progression is bounded by a constant raised to the power of the curve's Mordell-Weil rank, a result derived by combining gap principles for large canonical heights with spherical code bounds and which implies restrictions on sets of points with small sumsets.

The Gist

Imagine you are looking at a giant, invisible grid floating in space. This grid represents the rational points on an elliptic curve. In the world of mathematics, these points are like special "dots" that follow very strict, complex rules. If you pick two dots and combine them using the curve's special "group law," you get a third dot that also lives on the grid.

Now, imagine you have a different kind of structure: a Generalized Arithmetic Progression (GAP). Think of this as a perfectly organized shelf, or a set of shelves, where items are arranged in a very predictable, linear pattern. If you pick a number from this shelf, the next one is just a fixed step away. It's like a row of dominoes or a ladder with evenly spaced rungs.

The Big Question:
Can these two very different worlds coexist? Specifically, can a large number of the special "dots" from our elliptic curve grid happen to land exactly on the rungs of our perfectly organized ladder?

The Paper's Answer:
No, not really. The paper by Seokhyun Choi proves that these two structures are fundamentally incompatible. If you try to force a huge number of elliptic curve points to line up on a simple arithmetic ladder, the curve simply won't allow it. The more points you try to force onto the ladder, the more the "geometry" of the curve fights back.

Here is a breakdown of the paper's logic using everyday analogies:

1. The Two Opposing Forces

• The Ladder (Additive Structure): This represents the rational numbers ($\mathbb{Q}$). They love to form neat, predictable patterns like $1, 2, 3, 4$or$10, 20, 30, 40$.
• The Curve (Group Structure): This represents the elliptic curve ($E(\mathbb{Q})$). Its points are governed by a complex, curved geometry. They don't like to sit in neat, straight lines.

The paper asks: If I have a ladder with a million rungs, and I find that 10% of the rungs are occupied by points from an elliptic curve, how many points can I possibly have?

2. The "Crowded Room" Analogy (The Mordell-Weil Lattice)

To understand why the curve resists the ladder, the author uses a concept called the Mordell-Weil Lattice.

Imagine the elliptic curve points are people in a crowded room.

• The "Height" (Canonical Height): This is like how "tall" or "important" a person is. The taller they are, the more space they need.
• The "Angle" (Gap Principle): The paper proves that if two people (points) are roughly the same height, they cannot stand too close to each other. They must maintain a minimum distance, like people at a party who instinctively keep their personal space bubble.

If you try to pack too many people of similar height into a tiny corner of the room (which is what a simple arithmetic progression tries to do), they will bump into each other. The "personal space" rule (the gap principle) forces them to spread out.

3. The "Spherical Code" (The Packing Problem)

The author uses a concept from geometry called Spherical Codes. Imagine you are trying to stick as many magnets as possible onto the surface of a basketball. If the magnets repel each other (like the points on the curve), you can only fit a limited number of them before they push each other off.

The paper shows that the "ladder" (the arithmetic progression) forces the points to crowd into a tiny, narrow region of the mathematical space. But the "magnets" (the curve points) repel each other.

• Result: You can't fit many magnets in that tiny region. The number of points is strictly limited.

4. The Main Result

The paper proves a specific formula:
If you have an elliptic curve with a certain "rank" (a measure of its complexity, let's call it $r$), and you find a ladder (GAP) that holds a significant chunk of the curve's points, the total number of points you can find is roughly proportional to $r$.

In simple terms: The more complex the curve is, the more points it can have, but even then, they can't form a long, neat ladder. The length of the ladder is capped.

5. Why Does This Matter? (The Applications)

The paper doesn't just stop at ladders. It uses a famous theorem from "Additive Combinatorics" (Freiman's Theorem) to say:

• If a group of numbers has a "small sumset" (meaning if you add any two numbers in the group, the results don't explode into a huge variety of new numbers; they stay clustered), then that group must look like a ladder.
• Since the curve points hate ladders, they also hate having "small sumsets."

The Takeaway:
The x-coordinates of rational points on an elliptic curve are "additively rigid." They are chaotic and scattered in a way that prevents them from forming neat, predictable patterns. If you see a pattern, it's a sign that the set of points is actually very small.

Summary Metaphor

Imagine trying to arrange a flock of birds (the elliptic curve points) into a perfect, straight line (the arithmetic progression).

• The birds have a natural instinct to fly in a chaotic, swirling V-formation (the curve's geometry).
• If you try to force them into a straight line, they will only stay there if the line is very short.
• If the line gets too long, the birds' natural spacing rules (the gap principle) force them to break the line and scatter.
• The paper calculates exactly how long that line can be before the birds scatter, and the answer depends on how many "leaders" (the rank) the flock has.

In a nutshell: Elliptic curve points are too "socially distant" to stand in a neat, long queue. They prefer to spread out, and this paper proves mathematically just how much they can spread out before the queue breaks.

Technical Summary

Here is a detailed technical summary of the paper "Additive Rigidity for x-Coordinates of Rational Points on Elliptic Curves" by Seokhyun Choi.

1. Problem Statement

The paper investigates the interaction between two distinct algebraic structures:

1. The Mordell–Weil Group: The group of rational points $E(\mathbb{Q})$ on an elliptic curve, which is a finitely generated abelian group.
2. Additive Combinatorics: The additive structure of the rational numbers $\mathbb{Q}$, specifically sets with strong additive properties like arithmetic progressions (APs) and generalized arithmetic progressions (GAPs).

Core Question: Can the $x$-coordinates of a large set of rational points on an elliptic curve lie within a set possessing strong additive structure (e.g., a GAP)?
Context: This generalizes Bremner's Conjecture (1.1), which posits that the length of an arithmetic progression formed by $x$-coordinates of rational points on an elliptic curve of rank $r$ is bounded by $A \cdot r$. The author seeks to prove a higher-dimensional, density-based version of this conjecture.

2. Methodology

The proof relies on a "tension" between the additive structure of $\mathbb{Q}$ and the geometric constraints of the Mordell–Weil lattice. The methodology proceeds in three main stages:

A. Geometric Framework: The Mordell–Weil Lattice

• The free part of $E(\mathbb{Q})$ is identified with $\mathbb{R}^r$ via the canonical height $\hat{h}$.
• Rational points are viewed as vectors in a Euclidean space.
• Key Insight: If a large collection of points has comparable canonical heights, they must be separated by a definite angle in this lattice to avoid arithmetic degeneracies. This transforms the problem into bounding the size of a spherical code (a set of unit vectors with a minimum angular separation).

B. Gap Principles (Diophantine Estimates)

The author establishes "gap principles" (Theorems 3.4–3.6) which assert that rational points with comparable heights cannot be too close in the lattice unless specific arithmetic conditions are met.

• Mechanism: The proof analyzes the height of the sum of two points, $h(P+Q)$, relative to $h(P)$ and $h(Q)$.
• Case Splitting: The argument distinguishes between points with large $x$-coordinates and small $x$-coordinates.
  • For large $x$, the height of the sum is controlled by the arithmetic of the denominators.
  • For small $x$, different Diophantine estimates are used.
• Result: These estimates imply that if points satisfy certain additive constraints (related to their denominators), their canonical heights force them to be angularly separated.

C. Extraction from Additive Sets

A critical technical hurdle is that rational points in a GAP may share common denominators, potentially violating the "coprime" conditions required for the gap principles.

• Extraction Lemma (Section 4): The author proves that any dense subset ($\rho$-proportion) of a GAP contains a large subset of elements where the greatest common divisor (GCD) of the numerator and the common denominator is small (specifically $\le s^\delta$).
• This allows the application of the gap principles to a significant portion of the original set.

3. Key Contributions and Results

Main Theorem (Theorem 1.2)

Let $E/\mathbb{Q}$ be an elliptic curve of Mordell–Weil rank $r$. Let $d \ge 1$ and $\rho > 0$. There exists a constant $A(E, d, \rho)$ such that if a finite set of rational points $P \subseteq E(\mathbb{Q})$ satisfies:

1. The set of $x$-coordinates $x(P)$ is contained in a $d$-dimensional Generalized Arithmetic Progression (GAP) $G$.
2. $|x(P)| \ge \rho |G|$ (i.e., the points occupy a positive proportion of the GAP).

Then:
$$|P| \le A(E, d, \rho) \cdot r$$

Significance: This shows that the additive structure of $\mathbb{Q}$ and the group structure of $E(\mathbb{Q})$ are fundamentally incompatible. A large number of points cannot "hide" inside a low-dimensional additive structure.

Dependence on Lang's Conjecture

The constant $A(E, d, \rho)$ generally depends on the curve $E$. However, the author notes that assuming Lang's Conjecture (which provides a uniform lower bound for the canonical height of non-torsion points in terms of the curve's invariants), the constant $A$ can be chosen to depend only on $d$ and $\rho$, removing the dependence on the specific curve $E$.

Corollaries

1. Bremner's Conjecture: The result confirms Bremner's conjecture for $d=1$ (standard APs) and extends it to higher dimensions.
2. Small Sumsets (Freiman's Theorem): If the set of $x$-coordinates $S$ has small doubling ($|S+S| \le K|S|$), then $|P| \le A(E, K)r$. This applies to sets with small difference sets and small tripling as well.
3. Additive Energy: If the set of $x$-coordinates has high additive energy (many solutions to $x_1+x_2=x_3+x_4$), the size of the set is bounded linearly by the rank.

4. Technical Significance

• Synthesis of Fields: The paper successfully bridges Diophantine geometry (canonical heights, Mordell–Weil lattices) and Additive Combinatorics (GAPs, Freiman's theorem, spherical codes).
• Quantitative Rigidity: Unlike previous results that often relied on specific families of curves or integral points, this provides a general rigidity theorem for rational points based on the rank.
• Spherical Code Bounds: The use of bounds for spherical codes (Theorems 2.1 and 2.2) to limit the number of points with uniform angular separation is a novel application in this context, providing a quantitative cap on the density of points in the lattice.
• Extraction Mechanism: The "Extraction Lemma" (Corollary 4.5) is a significant technical contribution, showing how to extract "primitive" elements from dense subsets of rational GAPs to apply Diophantine estimates.

5. Conclusion

The paper establishes that rational points on elliptic curves exhibit additive rigidity: they cannot form large, dense subsets within generalized arithmetic progressions unless the rank of the curve is sufficiently large to accommodate them. This result strengthens the understanding of the distribution of rational points and provides a powerful tool for bounding the size of point sets with specific additive properties. The work suggests that the geometry of the Mordell–Weil lattice acts as a strong barrier against the formation of structured additive patterns among $x$-coordinates.