🔢 mathematics

More on the Boolean Prime Ideal Theorem

This paper establishes the consistency of Zermelo–Fraenkel set theory augmented with the Axiom of Dependent Choices, the non-existence of Vitali sets, and a substantial fragment of the Boolean Prime Ideal Theorem.

Original authors: Jacob Kowalczyk, Jindrich Zapletal

Published 2026-07-21

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Jacob Kowalczyk, Jindrich Zapletal

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 trying to build a house, but you have a strict rule: you can't use a specific, magical hammer (the Axiom of Choice) that lets you instantly pick the perfect brick from an infinite pile. Without this hammer, some construction projects become impossible to finish. Mathematicians have long wondered if there is a smaller, more modest tool—a "Boolean Prime Ideal Theorem"—that is strong enough to finish most of the house but weak enough to avoid creating a specific, chaotic mess known as a "Vitali set."

To understand the mess, think of the real number line as an endless, perfectly smooth road. A Vitali set is a bizarre collection of spots on this road where, if you slide the whole collection left or right by any rational distance (like a fraction of a mile), you get a completely new set of spots that never overlaps with the original. It's like having a pattern of footprints that, when you shuffle them, never touch the old ones, yet somehow cover the whole road. This creates a mathematical paradox where you can't measure the size of the collection at all. For decades, mathematicians have asked: Can we have our cake and eat it too? Can we have a powerful enough tool to organize our math (the Boolean Prime Ideal Theorem) without accidentally creating these unmeasurable, chaotic footprints?

This paper, written by Jacob Kowalczyk and Jindřich Zapletal, dives deep into this question. They don't just guess; they build a brand-new mathematical universe to test their ideas. Their main finding is a "yes, but": they prove it is consistent that you can have a very large, powerful version of the Boolean Prime Ideal Theorem (enough to solve many complex coloring and ordering problems) while still ensuring that no Vitali sets exist. However, they also show that their method has limits; it doesn't solve the entire problem for every possible mathematical scenario, leaving a few big questions open for future explorers.

The Great Math Puzzle

In the world of set theory, mathematicians are like architects trying to organize infinite collections of things. One of their biggest tools is the Axiom of Choice, which basically says, "If you have a bunch of boxes, you can pick one item from each box, even if there are infinite boxes." This tool is incredibly powerful, but it's also a bit of a bully. When you use it, you can create these weird, unmeasurable Vitali sets that break the rules of geometry and measurement.

To avoid the chaos, some mathematicians try to use a weaker tool called the Axiom of Dependent Choices (DC). This is like saying, "You can pick the next item based on the one you just picked," which is enough for most everyday math but stops short of the wild power of the full Axiom of Choice. The big question has been: If we use this weaker tool (DC), does the Boolean Prime Ideal Theorem (BPI)—a slightly less powerful version of the Axiom of Choice that says "every logical puzzle has a solution"—force us to create those chaotic Vitali sets anyway?

For a long time, no one knew. It was like asking if a specific type of glue would hold a bridge together without causing the bridge to spontaneously turn into jelly.

The Authors' Solution: Building a New World

Kowalczyk and Zapletal decided to stop guessing and start building. They constructed a special mathematical "sandbox" (a model of set theory) to see what happens. They started with a standard universe of math and then used a technique called forcing. Think of forcing as a way to gently nudge the universe, adding new pieces to the puzzle without breaking the existing rules.

Their goal was to add a massive amount of the Boolean Prime Ideal Theorem to their universe. They wanted to add enough of it to solve difficult problems, like:

The Coloring Problem: Imagine a giant map where every region is connected to many others. Can you color every region with a limited number of colors so that no two touching regions have the same color? The BPI helps prove you can do this for certain complex maps.
The Ordering Problem: Can you line up a messy pile of items in a perfect row from first to last? The BPI helps with this too.

The authors found a way to add these solutions using a very specific, carefully controlled method they call Boolean balanced forcing. They proved that if they use this specific method, they can successfully add the power to solve these problems without accidentally creating a Vitali set.

The "Balanced" Trick

How did they avoid the chaos? They used a concept they call "balance." Imagine you are trying to balance a stack of plates. If you add a new plate, you have to make sure it doesn't tip the whole stack over. In their math, they showed that for certain types of logical theories (the rules of the game), you can add a solution (a "completion") in a way that is perfectly balanced.

They proved that if the theory is "balanced," the new solutions you add won't create the unmeasurable Vitali sets. They identified a huge class of theories that are balanced, including:

Theories about extending ideals (rules about which groups of numbers are "small" enough to ignore).
Theories about coloring graphs (the map problem mentioned earlier) on certain types of spaces.
Theories about linearizing orders (lining things up).

By combining all these balanced theories into one giant "super-theory" and forcing with it, they created a universe where:

The Axiom of Dependent Choices holds (the rules are safe).
There are no Vitali sets (no chaotic, unmeasurable footprints).
A large, powerful chunk of the Boolean Prime Ideal Theorem holds (the bridge is strong, and the maps are colorable).

What They Didn't Solve

While this is a huge step forward, the authors are honest about what they didn't do. They didn't prove that the entire Boolean Prime Ideal Theorem is safe from creating Vitali sets. They only proved it for a "large fragment" of it.

They leave us with a few open questions, like:

Can we go even further? Is it possible to have the entire Boolean Prime Ideal Theorem without Vitali sets?
What about specific, tricky cases, like ordering all the subsets of the real numbers?

They also showed that not all theories are balanced. Some specific types of mathematical rules are too "unbalanced" to be added without risking the creation of a Vitali set. This suggests that the line between "safe" and "chaotic" is very thin and depends heavily on the specific structure of the problem.

The Takeaway

In simple terms, Kowalczyk and Zapletal have shown that you can have your cake and eat it too, but only for a specific, very large slice of the cake. They proved that it is possible to have a math world that is powerful enough to solve complex ordering and coloring problems, yet safe enough to avoid the most notorious paradoxes of measurement. They didn't solve the whole mystery, but they built a sturdy bridge across a wide chasm, showing us exactly where the safe ground is and where the jelly might still be waiting.

Technical Summary: "More on the Boolean Prime Ideal Theorem"

Problem Statement
The Boolean Prime Ideal Theorem (BPI) is a well-known consequence of the Axiom of Choice (AC) asserting that every Boolean algebra carries an ultrafilter. While it is established that BPI does not imply AC in the absence of the Axiom of Dependent Choices (DC), the relationship between BPI and AC in the presence of DC remains a central open question in choiceless set theory. Specifically, it is unknown whether ZF + DC + BPI implies the existence of a Vitali set (a subset of the real line containing exactly one element from each rational translate).

The paper addresses Conjecture 1.1: Is it consistent with ZF + DC that the Boolean Prime Ideal Theorem holds while no Vitali sets exist? The authors aim to demonstrate that a "large fragment" of BPI is consistent with the nonexistence of Vitali sets, assuming the existence of an inaccessible cardinal.

Methodology
The authors employ a framework based on compactly balanced forcing and the geometric axiomatization of the Solovay model. The methodology proceeds through the following steps:

Geometric Axiomatization: The paper utilizes the Solovay model (constructed from a model of ZFC with an inaccessible cardinal) as a base. This model is axiomatized geometrically, focusing on properties of sets definable from ordinals (OD) and their behavior under forcing extensions.
Balanced Theories: The core technical innovation is the definition of balanced propositional theories. A theory $T$ is "balanced" if, for any two sets of ordinals $y_0, y_1$ that are independent over a base set $x$ (denoted $y_0 \perp_x y_1$ ), and any formulas $\phi_0, \phi_1$ in the language of $T$ definable from $x \cup y_0$ and $x \cup y_1$ respectively, if $T \vdash \phi_0 \to \phi_1$ , there exists an interpolant $\psi$ definable from $x$ such that $T \vdash \phi_0 \to \psi$ and $T \vdash \psi \to \phi_1$ .
Boolean Balanced Forcing: The authors associate a poset $P_T$ with any consistent propositional theory $T$ . If $T$ is balanced, $P_T$ is a "Boolean balanced" forcing. The paper proves that such forcings preserve DC and, crucially, do not add Vitali sets.
Preservation Theorems: A key preservation theorem (Theorem 5.1) is established, showing that Boolean balanced forcings cannot add colorings to certain hypergraphs (specifically those invariant under shifts in abelian Polish groups) that are countably chromatic in ZFC. This directly implies the nonexistence of Vitali sets in the resulting models.

Key Contributions and Results

Characterization of Balanced Theories: The paper provides criteria for determining when a propositional theory is balanced.
- Ultrafilter Theories: Theorem 3.1 establishes that the ultrafilter theory of an ideal $I$ on a well-orderable set $A$ is balanced if $I$ is generated by a well-orderable collection of sets. This covers examples like the ideal of finite sets on $\omega$ and the ideal of nonstationary subsets on $\omega_1$ .
- $K_\sigma$ Theories: Theorem 3.8 and Theorem 3.11 provide a powerful criterion for theories on locally compact Polish spaces. They show that any consistent $K_\sigma$ theory can be extended to a balanced theory on a larger space. This implies that every $F_\sigma$ -ideal on $\omega$ can be extended to a maximal ideal via a balanced forcing.
- Linearization Theories: Theorem 3.14 gives a sufficient condition for the linearization theory of a partial order to be balanced, involving a specific function $F$ mapping elements to well-orderable sets.
- Products: Theorem 3.16 proves that the union of balanced theories in disjoint languages remains balanced.
Imbalanced Theories: Section 4 presents examples of theories that are not balanced, including the ultrafilter theory of the ideal of sets with finite vertical sections on $\omega \times \omega$ , and certain analytic P-ideals. This delineates the boundaries of the method.
Main Consistency Result (Theorem 1.2): Assuming the existence of an inaccessible cardinal, the authors construct an inner model of a generic extension satisfying:
1. ZF + DC;
2. No Vitali sets exist;
3. Every consistent $K_\sigma$ propositional theory on a locally compact Polish space of atomic formulas has a consistent completion.

This result confirms that a significant fragment of BPI (encompassing the extension of $F_\sigma$ -ideals and the de Bruijn–Erdős theorem for $K_\sigma$ graphs) is consistent with the nonexistence of Vitali sets.

Significance and Limitations
The paper claims to provide a "streamlined presentation" of compactly balanced forcing, offering a partial affirmative answer to Conjecture 1.1. It demonstrates that the nonexistence of Vitali sets is compatible with a robust fragment of the Boolean Prime Ideal Theorem within the context of ZF + DC.

However, the authors are modest regarding the scope of their solution. They explicitly state that their methods "can hardly be used to provide a complete affirmative solution" to the full conjecture (i.e., the full BPI). The paper leaves open several questions, including:

Whether ZF + DC + "The power set of reals can be linearly ordered" is consistent with no Vitali sets.
Whether ZF + DC + "There is a maximal ideal extending the asymptotic density zero ideal on $\omega$ " is consistent with no Vitali sets.
A full characterization of which ideals on $\omega$ yield balanced ultrafilter theories in the Solovay model.

The work establishes a technical foundation for analyzing the interplay between choice principles and regularity properties (like the nonexistence of Vitali sets) but stops short of resolving the equivalence of BPI and AC in ZF + DC.