🔢 mathematics

Minkowski shapes of pure number fields

This paper establishes a discrete–archimedean factorization for the Minkowski shapes of pure number fields, revealing a dichotomy where the shape serves as a complete invariant for odd degrees but only determines the core field for even degrees, while also providing explicit discriminant formulas and characterizing the geometric locus of these shapes within shape space.

Original authors: Khai-Hoan Nguyen-Dang

Published 2026-06-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Khai-Hoan Nguyen-Dang

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 an architect trying to understand the "shape" of a building. In the world of numbers, specifically Pure Number Fields, the building is constructed from a specific type of mathematical block. These fields are created by taking a number $a$ and finding a root of the equation $x^n = a$ . Think of $a$ as the blueprint and $n$ as the number of floors.

This paper, written by Khai-Hoan Nguyen-Dang, is like a master surveyor who has discovered a hidden rulebook for how these buildings look from the outside (their "Minkowski shape").

Here is the breakdown of the findings using simple analogies:

1. The Building Blocks: Pure Number Fields

Imagine you have a special type of Lego set. You pick a number $a$ (like 2, 3, or 10) and a size $n$ (like 3, 4, or 5). You build a tower where the top piece is the $n$ -th root of $a$ .

The Challenge: Mathematicians want to know: "If I look at the shape of this tower, can I tell exactly which blueprint ( $a$ ) I started with?"
The "Shape": The paper studies the "Minkowski shape." Imagine taking the building, removing the ground floor (which is just a straight line), and looking at the remaining structure's geometry. This shape is a unique fingerprint of the number field.

2. The Big Discovery: The "Two-Part" Recipe

The paper's main breakthrough is a Discrete-Archimedean Factorization. This is a fancy way of saying the shape of the building is made of two distinct ingredients mixed together, like a recipe:

Ingredient A (The Continuous Part): This comes from the size of the number $a$ . It's like the "scale" or "volume" of the building. The paper shows this part is always a simple, diagonal grid (like a perfectly aligned set of shelves).
Ingredient B (The Discrete Part): This comes from the "fine print" of the number theory—the specific way the number $a$ is built from prime numbers. This part acts like a "glue" or a "twist" that rearranges the shelves.

The Analogy: Imagine you have a stack of books (the shape).

The Continuous part tells you how thick the books are.
The Discrete part tells you how the books are stacked (tilted, shifted, or rotated).
The paper proves that for these specific "Pure" fields, you can always separate the thickness from the stacking pattern. The shape is just the stacking pattern applied to the thickness.

3. The Odd vs. Even Mystery

The paper finds a dramatic difference between buildings with an odd number of floors ( $n=3, 5, 7...$ ) and even number of floors ( $n=4, 6, 8...$ ).

Odd Floors (The Perfect Detective): If the building has an odd number of floors, the shape is a complete ID card. If you see the shape, you can instantly know exactly which blueprint ( $a$ ) was used. There is no confusion. The shape tells you everything.
Even Floors (The Twin Confusion): If the building has an even number of floors, the shape is a bit trickier. It can tell you the "core" of the building, but it cannot always tell you the sign (positive or negative) of the blueprint.
- Example: A building made from $x^4 = 10$ and a building made from $x^4 = -10$ can look identical from the outside (same shape), even though they are built from different blueprints. The shape forgets the sign in even degrees.
- However, if you know the building is "square-free" (a specific type of clean blueprint), the shape tells you the size of the number, and the only thing left to guess is whether it was positive or negative.

4. The "Leaf" on the Tree

Mathematicians imagine a giant garden called "Shape Space," where every possible shape of a number field is a point.

Generic Fields: For most random number fields, the shapes are scattered all over the garden, filling up the space like fog.
Pure Fields: This paper shows that Pure Number Fields are not scattered. They are all stuck on very specific, thin leaves (like flat sheets of paper) floating in the garden.
- These leaves are "Rational Diagonal Leaves."
- The paper proves that no matter how big the number $a$ gets, the shape always stays on one of these specific sheets. It never wanders off into the open space.

5. The Two Controls: Ratio vs. Product

The paper also explains how the shape changes as you change the blueprint $a$ . It separates the variables into two types:

Ratio Variables: These control the shape (the geometry). If you change the blueprint in a way that keeps the ratio of parts the same, the shape stays on the same leaf.
Product Variables: These control the discriminant (a measure of the building's complexity or "roughness"). If you change the blueprint in a way that changes the product of parts, the building gets "rougher" or "smoother," but it stays on the same leaf.

The Takeaway: The shape moves along the leaf based on ratios, while the complexity grows based on products. They are independent controls.

Summary

In simple terms, this paper says:

Pure Number Fields have a very rigid, predictable structure.
Their "shape" can be broken down into a simple size part and a complex stacking part.
If the field has an odd degree, the shape is a perfect ID card. If it has an even degree, the shape is a perfect ID card except it can't tell you if the number is positive or negative.
All these shapes live on specific, thin "leaves" in the mathematical universe, not scattered randomly.
The paper provides the exact formulas to calculate these shapes and the "roughness" (discriminant) of the fields, separating the geometry from the arithmetic.

The author essentially mapped out the "neighborhood" where these specific number fields live, showing they are much more orderly and constrained than previously thought.

Technical Summary: Minkowski Shapes of Pure Number Fields

Problem Statement
The paper investigates the "Minkowski shape" of pure number fields $K_a = \mathbb{Q}(\theta)$ where $\theta^n = a$ . While the discriminant of a number field records the covolume of its Minkowski lattice and the ramified primes, the Minkowski shape is a finer archimedean invariant. It is defined as the point in the shape space $S_{n-1} = \text{GL}_{n-1}(\mathbb{Z}) \backslash \text{GL}_{n-1}(\mathbb{R}) / \text{GO}_{n-1}(\mathbb{R})$ obtained from the trace-zero lattice of the ring of integers, modulo integral basis changes, orthogonal similitudes, and positive scaling.

The central question is how the shapes of pure fields distribute within the ambient moduli space $S_{n-1}$ . While generic families of number fields (e.g., $S_n$ -families) exhibit equidistribution in the full shape space, "thin" families often concentrate on smaller loci. This paper seeks to explicitly characterize the geometric support and arithmetic structure of the shape locus for pure fields, specifically determining whether the shape serves as a complete invariant for the field and how the shape relates to the discriminant.

Methodology
The author employs a combination of explicit integral basis theory, linear algebra over rational fields, and geometric analysis of symmetric spaces.

Strong Decomposition and Normalized Monomials: The parameter $a$ is decomposed into a "strong decomposition" $a = \epsilon \prod_{j=1}^{n-1} a_j^j$ , where $a_j$ are pairwise coprime squarefree integers. Using this, the author defines normalized monomials $e_m = \theta^m / C_m(a)$ , where $C_m(a)$ are specific rational factors derived from the strong decomposition.
Integral Basis Theorem: Under Hypothesis (H) (a local condition at primes dividing $n$ : for every $p|n$ , either $v_p(a)=0$ or $p \nmid v_p(a)$ ), the paper utilizes the explicit integral basis theorem of Jakhar–Khanduja–Sangwan. This provides a normalized integral basis for $\mathcal{O}_{K_a}$ involving denominators supported only on primes dividing $n$ .
Discrete–Archimedean Factorization: The core methodological innovation is separating the shape into two components:
- An archimedean diagonal part derived from the normalized monomials, which are shown to be pairwise orthogonal under the Minkowski inner product.
- A discrete rational part arising from the transition matrix between the normalized monomial basis and the true integral basis.
Geometric Support Analysis: The paper analyzes the image of the pure family in shape space by identifying it with "rational diagonal leaves" $T_C$ , defined as the orbits of rational diagonal matrices under the action of $\text{GL}_{n-1}(\mathbb{Z})$ and $\text{GO}_{n-1}(\mathbb{R})$ .

Key Contributions and Results

1. Discrete–Archimedean Factorization (Theorem 1.1)
Under Hypothesis (H), the Minkowski shape of $K_a$ admits a factorization:
$\text{sh}(K_a) = \left[ C(a)^T \text{diag}(s_1(a), \dots, s_{n-1}(a)) C(a) \right]$
where:

$s_m(a) = |a|^{2m/n} / C_m(a)^2$ are explicit normalized monomial terms forming a diagonal matrix.
$C(a) \in \text{GL}_{n-1}(\mathbb{Q})$ is an upper-triangular matrix arising from the normalized integral basis.
The class $\Xi(a) = [C(a)] \in \text{GL}_{n-1}(\mathbb{Q}) / U_{n-1}^+(\mathbb{Z})$ is a well-defined discrete invariant (unipotent class) independent of the specific choice of normalized integral basis.

2. Arithmetic Consequences: Parity Dichotomy (Theorem 1.2)
The factorization leads to a sharp distinction between odd and even degrees regarding the shape as an isomorphism invariant:

Odd Degree ( $n$ odd): The shape is a complete invariant. If $\text{sh}(K_a) = \text{sh}(K_b)$ , then $K_a \cong K_b$ . This is because the diagonal ratios $s_i/s_j$ generate the field $\mathbb{Q}(|a|^{2/n})$ , which recovers the full field when $n$ is odd.
Even Degree ( $n=2r$ ): The shape determines the core field $\mathbb{Q}(|a|^{1/r})$ but not the sign of $a$ . Specifically, $\text{sh}(K_a) = \text{sh}(K_{-a})$ for infinitely many squarefree $a$ , yet $K_a \not\cong K_{-a}$ (they have different signatures). On the squarefree subfamily, the shape determines the field up to the sign ambiguity ( $K_a \cong K_b$ or $K_a \cong K_{-b}$ ).

3. Discriminant Factorization (Theorem 1.3)
The paper derives explicit formulas for the absolute discriminant $|\text{disc}(K_a)|$ . It shows that the discriminant factors into a periodic part depending only on the residue class of $a$ modulo $M(n) = n \cdot \text{rad}(n)$ , and a part governed by the strong decomposition variables. Specifically, after removing the periodic factor, the discriminant growth is controlled by "divisor-products" $B_d(a) = \prod_{d|j} a_j$ , reflecting a divisor-lattice structure.

4. Geometric Support and Leafwise Structure (Theorem 1.4)
The paper establishes that the pure family lies on a very thin subset of the shape space:

Rational Diagonal Leaves: Every admissible pure-field shape lies on a rational diagonal leaf $T_C$ for some $C \in \text{GL}_{n-1}(\mathbb{Q})$ .
Finite Support under (H): While the unconditional support is a countable union of closed leaves, under Hypothesis (H), the support is contained in a finite union of such leaves for a fixed degree $n$ .
Separation of Variables: On a fixed normalized stratum (fixed $\Xi(a)$ ) and fixed discrete label $\delta(a)$ , the shape depends only on ratio variables $\rho_j = a_j / a_{n-j}$ (for $\gcd(j,n)=1$ ), whereas the discriminant depends only on product variables $P_j = a_j a_{n-j}$ . This separation implies that the natural counting problem for pure-field shapes is "leafwise" rather than ambient.

Significance and Claims
The paper claims to provide a concrete model for the theory of thin families in shape space. Its significance lies in:

Structural Clarity: It explicitly separates the finite-place arithmetic data (which selects the supporting rational leaf) from the archimedean data (which moves the shape within that leaf).
Rigidity vs. Ambiguity: It precisely quantifies the rigidity of the shape invariant, proving completeness in odd degrees and characterizing the exact nature of the sign ambiguity in even degrees.
Geometric Framework: It identifies the "rational diagonal leaves" as the fundamental geometric objects supporting pure families, showing that these leaves are closed, rational, and of dimension $n-2$ .
Foundation for Counting: By separating ratio and product variables, the paper sets the stage for future counting and equidistribution theorems on these specific leaves, noting that the current work establishes the geometric support and structural decomposition but does not yet supply the counting theorem itself.

The author explicitly notes that the results rely on Hypothesis (H) for the finite leaf support and the explicit integral basis, and identifies removing this hypothesis and intrinsic reformulation of the invariants as the primary open problems.