🔢 mathematics

Fields where torsion forms decompose

This paper proves that over a real field which is a transcendence degree 1 extension of a hereditarily pythagorean base field, every torsion quadratic form decomposes into an orthogonal sum of 2-dimensional torsion forms, a result derived from a broader analysis of weakly isotropic forms over henselian valued fields and function fields in one variable.

Original authors: M. Archita, Karim Johannes Becher

Published 2026-05-14

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: M. Archita, Karim Johannes Becher

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 master architect working with a very specific type of building block: quadratic forms. In the world of mathematics, these are like complex structures built from numbers. Sometimes, these structures are "torsion forms." Think of a torsion form as a building that, if you stack enough copies of it on top of each other, eventually collapses into a flat, useless pile (mathematically, it becomes "zero" in a specific ring).

The big question the authors, M. Archita and Karim Johannes Becher, are asking is: Can we always take these complex, collapsing buildings and break them down into simple, two-story "binary" blocks?

They call this property being "strongly balanced." It's like asking, "Can every complicated, wobbly tower be taken apart and rebuilt using only simple 2x2 Lego bricks?"

The Problem

In some mathematical worlds (fields), the answer is "yes." In others, the answer is "no."

The "No" Case: The authors mention a famous example over a field with two variables (like $Q(X, Y)$ ). There, they found a specific torsion form that is so stubbornly complex that it cannot be broken down into simple 2-dimensional pieces. It's like a tower made of a weird, unbreakable alloy that refuses to be simplified.
The Goal: They want to find out exactly which types of mathematical worlds (base fields) guarantee that every torsion form can be simplified into these 2-dimensional blocks.

The Solution: The "Hereditarily Pythagorean" World

The authors prove a major theorem: If you start with a base field that is "hereditarily Pythagorean," then every torsion form in any 1-step extension of that field (like adding one new variable, $X$ ) will be strongly balanced.

What does "Hereditarily Pythagorean" mean?
Let's use an analogy:

A Pythagorean field is a world where if you take any two "squares" (like $3^2$ and $4^2$ ) and add them together, the result is always another perfect square. It's a world of perfect harmony where sums of squares never break the rules.
A Hereditarily Pythagorean field is a world where not only do the current rules hold, but if you build any new, finite extension of this world, the rules still hold. It's a world of perfect, unbreakable stability.

The authors say: "If your starting world is this perfectly stable, hereditarily Pythagorean type, then no matter how you stretch it out by adding one new variable, you will never encounter a 'stubborn' torsion form. They will all break down nicely into 2-dimensional pieces."

How They Solved It (The Detective Work)

The authors didn't just guess; they used a sophisticated detective toolkit involving valuations and local-global principles.

The Microscope (Valuations): Imagine looking at a mathematical field through a microscope called a "valuation." This lets you zoom in on the "residue" or the core essence of the field. The authors proved that if you can solve the puzzle in the "residue" (the zoomed-in view), you can often solve it for the whole field.
The Local-Global Principle: This is a famous mathematical idea: "If something works everywhere locally (in every small neighborhood), does it work globally (for the whole thing)?"
- The authors extended this principle to handle infinite-dimensional spaces (very large, complex structures).
- They showed that if a form is "weakly isotropic" (a technical way of saying it has a hidden weakness or collapse point) in every local view, it must be weakly isotropic globally.
The Chain Reaction: They connected the stability of the base field (hereditarily Pythagorean) to the behavior of these forms. They proved that in these stable worlds, the "weakness" of these forms is always detectable in a way that forces them to decompose into those simple 2-dimensional blocks.

The Big Takeaway

The paper settles a long-standing question for a specific, important class of fields.

Before: We knew some fields allowed simplification, and some didn't. We didn't know if all extensions of a "Pythagorean" number field would behave nicely.
Now: We know that if you start with a hereditarily Pythagorean field, you are safe. Any field you build on top of it (with just one extra variable) will have the property that all torsion forms are strongly balanced.

In short: In these mathematically "perfectly stable" worlds, complexity can always be reduced to simple, two-dimensional building blocks. The "stubborn" towers that refuse to break down simply do not exist there.

Technical Summary: Fields Where Torsion Forms Decompose

Problem Statement
The paper investigates the decomposition properties of torsion quadratic forms over real fields. Specifically, it addresses the question of whether every torsion form over a real field $F$ is "strongly balanced," meaning it is isometric to an orthogonal sum of binary (2-dimensional) torsion forms. While it is known that over nonreal fields, all regular quadratic forms are torsion and decompose into 1-dimensional forms, the situation over real fields is more complex. Previous work by Arason and Pfister established that over the rational function field $\mathbb{Q}(X, Y)$ , there exist torsion forms that are not strongly balanced, implying that the property does not hold universally for all real fields.

The central problem is to identify classes of base fields $K$ such that for any real field extension $F/K$ of transcendence degree 1, every torsion form over $F$ is strongly balanced. The authors specifically target the case where $K$ is a hereditarily pythagorean field (a real field where every finite real extension is pythagorean, i.e., every sum of two squares is a square). This confirms a conjecture by the authors in a previous work [2, Conjecture 4.4].

Methodology
The authors employ a combination of valuation theory, the theory of quadratic forms, and local-global principles. The methodology proceeds through several key steps:

Invariant Analysis: The study utilizes the invariant $w(K)$ , defined as the supremum of the dimensions of minimal weakly isotropic forms over $K$ . The condition that every torsion form is strongly balanced is equivalent to $w(K) \le 2$ (Corollary 2.5).
Valuation-Theoretic Characterization: The authors utilize Bröcker's characterization of hereditarily pythagorean fields, which relates them to fields admitting a henselian valuation with specific residue field properties.
Extension of Local-Global Principles: A significant methodological component is the extension of Mehmeti's local-global principle (originally for finite-dimensional forms over complete rank-1 valued fields) to infinite-dimensional quadratic spaces. This is formalized in Theorem 4.1, which establishes that for an anisotropic quadratic space of dimension $\ge 3$ over a function field $F/K$ with a henselian rank-1 base valuation, there exists an extension valuation where the form remains anisotropic.
Inductive Argument on Valuation Rank: The proof for the main theorem involves an inductive argument on the rank of the valuation. The authors decompose valuations of finite rank into compositions of rank-1 valuations and utilize the properties of henselian valuations to relate the invariant $w(F)$ of the function field to the invariant of the residue field.

Key Contributions and Results

Main Theorem: The paper proves that if $K$ is a hereditarily pythagorean field and $F/K$ is a real field extension of transcendence degree 1, then every torsion form over $F$ is strongly balanced. This is established via Theorem 5.6, which shows that the invariant $ew(K) = \sup \{w(F) \mid F/K \text{ is a function field in one variable}\}$ equals 2 for hereditarily pythagorean $K$ .
Generalization of Local-Global Principles: Theorem 4.1 provides a local-global principle for function fields in one variable over henselian valued fields that applies to quadratic spaces of arbitrary (possibly infinite) dimension. This result is noted as being of independent interest.
Relationship between $w$ -invariant and Valuations: The paper establishes that for a henselian valuation $v$ on $K$ with $\text{char}(K_v) \neq 2$ , the invariant $ew(K)$ is equal to $ew(K_v)$ (Theorem 5.4). This allows the reduction of the problem over the function field to the residue field.
Refinement of Known Bounds: The authors clarify that while hereditarily pythagorean fields ensure $w(F) \le 2$ , this cannot be derived solely from bounds on the Pythagoras number $p(F)$ , as $p(F)$ can be up to 5 in these cases (citing [4, Theorem 1.6]), whereas Corollary 2.3 would require $p(F) \le 2$ to directly imply $w(F) \le 2$ .

Significance
The paper resolves the specific question of whether torsion forms decompose into binary forms over real extensions of transcendence degree 1 for the class of hereditarily pythagorean base fields. By confirming Conjecture 4.4 from [2], the work delineates a clear boundary for the "strongly balanced" property.

The authors explicitly state that their proof follows the path of [4] but requires the extension to infinite-dimensional spaces to handle the specific valuation-theoretic structure of hereditarily pythagorean fields. The significance lies not in discovering new examples of non-decomposable forms (which are known for $\mathbb{Q}(X,Y)$ ), but in proving that such counterexamples do not exist over function fields of transcendence degree 1 when the base field is hereditarily pythagorean. The paper concludes by listing open questions regarding the existence of fields with $w(K) > 3$ and the precise relationship between $w(K)$ and the Pythagoras number $p(K)$ , maintaining a modest stance on the broader landscape of these invariants.

The Problem

The Solution: The "Hereditarily Pythagorean" World

How They Solved It (The Detective Work)

The Big Takeaway

Technical Summary: Fields Where Torsion Forms Decompose

More like this