🔢 mathematics

Locatedness, Convexity, and Integrability in $\mathbb{R}^N$

This paper provides an improved and detailed constructive proof demonstrating that the support of a Lebesgue integrable complemented set in $\mathbb{R}^N$ with positive measure, which is both bounded and convex, is necessarily totally bounded and located.

Original authors: Douglas S Bridges

Published 2026-07-20

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Douglas S Bridges

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

The Shape of the Unknown

Imagine you are a cartographer trying to draw a map of a mysterious island. In the world of mathematics, this island is a shape called a "set," and the ocean around it is the vast space of numbers we call $\mathbb{R}^N$ . Sometimes, these shapes are messy, jagged, or have holes that make them impossible to measure or pin down. But other times, they are smooth, solid, and well-behaved. This paper lives in a special corner of math called "constructive analysis." Think of this not as a place where mathematicians just say, "It must exist somewhere," but a workshop where they insist on actually building the thing, step-by-step, to prove it's real.

The main tools in this workshop are measure (a way to calculate the "size" or volume of a shape, like how much water fits in a bucket), convexity (a property where if you pick any two points inside a shape, the straight line connecting them stays entirely inside the shape—like a smooth marble, not a starfish), and locatedness (the ability to say exactly how far any point in the universe is from the edge of your shape). Why does this matter? Because in the constructive world, you can't just assume a shape has a boundary you can find. You have to prove you can get close to it. If a shape is "located," it means it's a reliable, usable object in the mathematical toolkit. If it's not, it's like a ghost: you know it's there, but you can't touch it or measure your distance to it.

The Paper's Discovery

In this note, Douglas S. Bridges tackles a specific puzzle about these shapes in a space with $N$ dimensions. He wants to prove a corrected version of a previous idea: If you have a shape $S$ that is "Lebesgue integrable" (meaning you can calculate its size without getting stuck), has a positive size (it's not empty), and its core part $S_1$ is both bounded (it doesn't stretch to infinity) and convex (it's a smooth, solid blob), then $S_1$ is not just a vague concept—it is totally bounded and located. In plain English, this means the shape is compact enough that you can cover it with a finite number of tiny dots, and you can always calculate the distance from any point in the universe to the shape.

To get there, Bridges builds a logical ladder using a few clever tricks. First, he shows that if you take a flat slice of your $N$ -dimensional space (like a sheet of paper in a 3D room), that slice has zero volume. This seems obvious, but in the strict rules of constructive math, you have to prove you can actually build that slice and show it has no "thickness." He then proves that if you have a shape with a positive size, it must contain enough "spine" to hold up the entire space—specifically, it must contain $N$ independent directions, like the $x$ , $y$ , and $z$ axes in a 3D room.

The real magic happens when he combines these directions. He shows that if you have a convex shape with a positive size, it must have a "core" or interior that is so dense with points that you can find a small ball of points inside it. Once you have this solid core, he uses a geometric trick involving a "ball" (a perfect sphere) to show that the interior of the shape is "uniformly dense" in the whole shape. Imagine the shape is a sponge; this lemma proves that no matter where you poke the sponge, you can always find a tiny, solid ball of sponge material nearby.

Finally, he uses this density to prove the main result. He argues that because the shape is so full of points, you can't have a part of it that is "too far" from a finite collection of points. If you try to find a point in the shape that is far away from your map of dots, the math forces a contradiction: the "empty" space between your dots and the shape would have to be so big that it would eat up the entire volume of the shape, which we already know is impossible. Therefore, the shape must be "totally bounded" (you can cover it with a finite number of dots) and "located" (you can always measure the distance to it).

The paper doesn't just suggest this might be true; it provides a rigorous, step-by-step constructive proof. It rules out the possibility that a bounded, convex, positive-size shape could be "unlocated" or impossible to approximate. By following the logic of the lemmas, the conclusion stands firm: such shapes are reliable, measurable, and fully within our grasp.

Technical Summary: Locatedness, Convexity, and Integrability in $\mathbb{R}^N$

Problem Statement
This note addresses a specific problem within constructive analysis, specifically Errett Bishop's framework. The primary objective is to provide an improved and detailed constructive proof for a corrected version of a theorem previously stated in [2, Theorem 2]. The theorem concerns the properties of a Lebesgue integrable complemented set $S$ in $\mathbb{R}^N$ (where $N$ is a fixed positive integer). Specifically, if $S$ has positive Lebesgue measure ( $\mu(S) > 0$ ) and its support $S_1$ is both bounded and convex, the goal is to demonstrate that $S_1$ is totally bounded and, consequently, located in $\mathbb{R}^N$ . The paper assumes the reader has familiarity with the constructive integration theory outlined in Chapter 6 of [1].

Methodology
The proof strategy relies on a sequence of lemmas that establish the geometric and measure-theoretic properties of convex sets and subspaces within the constructive framework. The methodology proceeds as follows:

Measure of Subspaces: The author first establishes that any $n$ -dimensional subspace $V$ of $\mathbb{R}^N$ (where $1 \le n < N$ ) is a Lebesgue integrable complemented set with measure zero (Lemma 1). This is achieved by approximating $V$ with a countable union of rectangles of measure zero.
Intersection Properties: It is shown that the intersection of a Lebesgue integrable complemented set $A$ with a subspace $V$ of lower dimension retains measure zero (Lemma 2).
Existence of Bases: A critical step involves proving that if a Lebesgue integrable complemented set $A$ has positive measure, its support $A_1$ must contain a basis for $\mathbb{R}^N$ (Lemma 3). This is proven inductively by showing that one can always find a point in $A_1$ that is linearly independent of a previously constructed set of vectors, utilizing the fact that the intersection with the span of previous vectors has measure zero.
Geometric Density: The paper utilizes and improves upon existing lemmas regarding convex sets. Lemma 4 provides a geometric condition for balls within convex sets. Lemma 5 proves that for a convex set $C$ with an inhabited interior, the interior $C^\circ$ is uniformly dense in $C$ . This improves upon the proof found in [2, Lemma 9].
Interior of Convex Hulls: Lemma 6 demonstrates that the average of linearly independent vectors (specifically $\frac{1}{N+1}\sum x_k$ ) lies in the interior of the convex hull of those vectors and the origin.

Key Contributions and Results
The central result, denoted as $(*)$ , is the constructive proof that a bounded, convex, Lebesgue integrable complemented set $S$ with positive measure has a support $S_1$ that is totally bounded and located.

The proof of $(*)$ synthesizes the preceding lemmas:

By translating $S_1$ to contain the origin, Lemma 3 ensures $S_1$ contains $N$ linearly independent vectors.
Lemma 6 guarantees that the centroid of these vectors lies in the interior of $S_1$ .
Lemma 5 establishes that the interior of $S_1$ is uniformly dense in $S_1$ .
Using the definition of uniform density, the author constructs a compact set $K$ approximating $S_1$ with a specific measure constraint.
By analyzing the distance between an arbitrary point $x \in S_1$ and the compact set $K$ , the author derives a contradiction if the distance is too large, thereby proving the existence of a finite $\epsilon$ -approximation for any $\epsilon$ . This confirms that $S_1$ is totally bounded.

Significance and Claims
The paper claims to offer an "improved, detailed presentation" of the constructive proof for the theorem regarding the locatedness of such sets. The significance lies in the rigorous constructive verification of properties that are often taken for granted in classical analysis. The author explicitly notes that the proofs of Lemma 5 and Lemma 6 serve to improve upon the arguments found in [2, Lemmas 9 and 2]. The work does not propose new applications or future implications but focuses strictly on refining the logical structure and detail of the existing constructive proof within the context of Bishop's constructive mathematics.

The Shape of the Unknown

The Paper's Discovery

Technical Summary: Locatedness, Convexity, and Integrability in RN\mathbb{R}^NRN

More like this

Technical Summary: Locatedness, Convexity, and Integrability in $\mathbb{R}^N$