Monochromatic Sums and Products over $\mathbb{Q}$

Technical Summary: Monochromatic Sums and Products over $\mathbb{Q}$

Problem Statement
The paper addresses a central open problem in Ramsey theory concerning the partition regularity of mixed additive and multiplicative patterns. While Hindman's Theorem (1974) guarantees that any finite coloring of the natural numbers $\mathbb{N}$ contains an infinite sequence whose finite subset sums are monochromatic, and the multiplicative analogue holds similarly, the simultaneous requirement for both sums and products is significantly more restrictive.

Hindman (1979) demonstrated that no infinite sequence exists in $\mathbb{N}$ such that all its pairwise sums and pairwise products are monochromatic. Consequently, the natural finite generalization—Hindman's Conjecture (Conjecture 1.1)—posits that for any $n \ge 2$, any finite coloring of $\mathbb{N}$ contains $n$ numbers $\{x_1, \dots, x_n\}$ such that all non-empty subset sums $\sum_{i \in S} x_i$ and subset products $\prod_{i \in S} x_i$ are monochromatic. This conjecture remains open for $\mathbb{N}$, even for the simplest case $n=2$ (the pattern $\{x, y, x+y, xy\}$).

The paper focuses on the weaker version of this conjecture over the rational numbers $\mathbb{Q}$ (Conjecture 1.2), which was previously settled only for $n=2$ by Bowen and Sabok (2023). The authors aim to prove the statement for arbitrary $n$ over $\mathbb{Q}$.

Methodology
The proof relies on a constructive, algorithmic approach that iteratively updates variables to satisfy increasingly complex monochromatic conditions. The core technical machinery involves:

1. Polynomial van der Waerden Theorem: The authors heavily utilize the multidimensional polynomial van der Waerden theorem (Bergelson and Leibman). Specifically, they apply a version over $\mathbb{Q}^\ell$ (Theorem 2.4) which guarantees the existence of a vector $u$ and an integer $n$ such that $u + \sum c_i p_i(n) v_i$ are monochromatic, where $p_i$ are integer-valued polynomials with $p_i(0)=0$ and $c_i$ are rational coefficients of bounded "size."
2. Auxiliary Product Colorings: To manage the interaction between sums and products, the authors define auxiliary colorings $\chi_M$ that encode the colors of scaled and shifted versions of the variables. This allows the application of van der Waerden-type results to control the colors of complex polynomial expressions.
3. Iterative Updates (Shifts and Scalings): The proof is structured as an algorithm that maintains a sequence of variables $x_n, \dots, x_1$. The authors define two specific types of operations:
   • Shifts: Operations of the form $x_i \leftarrow x_i + f(x_{i-1}, \dots, x_1)$.
   • Scalings: Operations of the form $x_i \leftarrow k x_i$.
     These updates are performed to ensure that specific polynomial combinations (e.g., $P(x_i \prod_{j \in T} x_j; \prod_{j \in S} x_j)$) remain monochromatic relative to their base terms.
4. Size Parameters: A technical parameter $M$ (representing the maximum size of coefficients and denominators) is introduced. The proof proceeds by choosing a sufficiently large initial $M_0$ and iteratively reducing it to $M_1, M_2, \dots$ as updates are performed, ensuring that the conditions hold for the final required size.

Key Contributions and Results
The primary result is Theorem 1.3, which states:

For any $n \ge 2$, if $\mathbb{Q}$ is colored in finitely many colors, there exist non-zero $x_1, \dots, x_n$ such that all numbers $\sum_{i \in S} x_i$ and $\prod_{i \in S} x_i$ (for non-empty $S \subseteq [n]$) are the same color.

The paper establishes this through a main lemma (Proposition 4.1) which proves a stronger, more general structural property. This lemma asserts the existence of variables $x_n, \dots, x_1$ such that for any $i$ and any subset $T \subseteq [i-1]$, the color of $P(x_i \prod_{j \in T} x_j; \prod_{j \in S} x_j)$ matches the color of $P(x_i \prod_{j \in T} x_j)$ for all "good" polynomials $P$ with bounded size, where $S$ ranges over subsets of $[i-1]$ strictly greater than $T$.

The proof of the main theorem (Section 5) synthesizes this lemma with:

• Ramsey's Theorem: To extract a subsequence where the color of products depends only on the degree (number of factors).
• Folkman's Theorem: Applied to an auxiliary coloring of indices to find a set of indices whose subset sums correspond to the required degrees.
• Final Construction: Combining these to construct the final set $\{X_1, \dots, X_r\}$ where both sums and products are monochromatic.

Significance and Distinctions
The authors highlight several distinctions between their work and the recent $n=2$ result by Bowen and Sabok:

• Generality: This work resolves the problem for arbitrary $n$, whereas previous results were limited to $n=2$.
• Methodology: Unlike Bowen and Sabok, this proof does not rely on Moreira's result as a "black box." Instead, it shares a spirit with the author's earlier proof of Moreira's result, utilizing the polynomial van der Waerden theorem directly.
• Explicit Bounds: The proof yields explicit bounds on the size of the solution, a feature not present in the Bowen-Sabok proof.
• Limitations over $\mathbb{N}$: The authors explicitly note that their proof technique relies on the ability to perform "shifts" involving division (e.g., $b/c$), which is generally impossible in $\mathbb{N}$. This explains why the result is established for $\mathbb{Q}$ but the conjecture for $\mathbb{N}$ remains open. The paper suggests that resolving the $\mathbb{N}$ case likely requires new ideas, potentially involving higher-degree polynomials or different structural constraints, as hinted at in Conjecture 4.1 of a previous work [1].

The paper concludes by discussing potential extensions to other patterns (e.g., involving squares or distinct products) and posing open questions regarding partition regularity over polynomial semi-rings and patterns involving negative exponents.