Lower Limit Topology and the Sorgenfrey Line

In our everyday experience with numbers, nearness is a symmetric concept. A point is surrounded by a bubble of other points on all sides. This intuitive idea forms the basis of the standard topology on the real line, the foundation for calculus and physics. But what happens if we break this symmetry? What if a neighborhood only includes a point and those immediately to its right, like standing on a ledge? This single change creates the lower limit topology and its most famous example, the Sorgenfrey line, a space with deeply counter-intuitive yet logically consistent rules. This article addresses the knowledge gap between our standard geometric intuition and the bizarre world of non-standard topologies. By exploring this fascinating space, we can test the limits of familiar mathematical concepts and gain a deeper appreciation for the assumptions that underpin them. Across the following chapters, you will learn the fundamental rules that govern this strange world and uncover its surprising connections to other fields. The "Principles and Mechanisms" chapter will deconstruct the Sorgenfrey line, revealing its shattered structure and peculiar behaviors. Following that, "Applications and Interdisciplinary Connections" will demonstrate how this abstract space serves as a powerful diagnostic tool in analysis, geometry, and beyond.

Imagine the familiar number line, stretching infinitely in both directions. We usually think of a "neighborhood" around a point, say the number $x$, as a small open bubble surrounding it, like the interval $(x-\delta, x+\delta)$. No matter how small you make that bubble, you can always find points to the left and to the right of $x$. This simple idea is the heart of the "standard topology" on the real line, the one we use every day in calculus and physics. It's symmetric, comfortable, and intuitive.

But what if we were to define a neighborhood differently? What if, when standing at a point $x$, your neighborhood only included the point you're on and the points immediately to your right? This is the central idea of the ​​lower limit topology​​, which gives us a bizarre and fascinating space called the ​​Sorgenfrey line​​.

In the world of the Sorgenfrey line, the fundamental building blocks of space are not the familiar open intervals $(a, b)$, but rather ​​half-open intervals​​ of the form $[a, b)$, which include the left endpoint $a$ but exclude the right endpoint $b$.

Think of it this way: a neighborhood of a point $x$ is no longer a symmetric bubble. Instead, it's like you're standing on the very edge of a cliff or a ledge. A basic neighborhood of $x$ is any set of the form $[x, x+\epsilon)$ for some tiny positive number $\epsilon$. You can see all the points for a short distance in front of you (to the right), but there is nothing behind you (to the left) that is considered "near" in the same way. The point $x$ is the absolute "lower limit" of its own neighborhood. This one simple, asymmetric change in our definition of "nearness" unravels the familiar fabric of the number line and weaves it into something entirely new.

How does this new rule change our notion of "open sets"? An open set is simply any set that can be built by gluing together our basic building blocks. A remarkable thing happens: every open set from our old, familiar world is still an open set in this new one. Why? Take any standard open interval $(a, b)$. We can write it as a union of our new half-open building blocks:

Since any standard open set is just a collection of open intervals, and each of those can be built from Sorgenfrey intervals, all the old open sets are also new open sets. In the language of topology, we say that the lower limit topology $\mathcal{T}_l$ is ​​finer​​ than the standard topology $\mathcal{T}_u$; it contains more open sets.

But this "fineness" comes at a cost. The set $[0, 1)$, for instance, is a basic open set in the Sorgenfrey line, but it is certainly not open in the standard topology (the point $0$ has no symmetric bubble around it that stays inside the set). This difference means the two spaces are not topologically equivalent. Specifically, the identity map $id(x)=x$ from the standard line to the Sorgenfrey line is not continuous (the preimage of the Sorgenfrey-open set $[0,1)$ is $[0,1)$, which is not open in the standard topology). Therefore, it cannot be a homeomorphism. The Sorgenfrey line has a richer, more complex structure, and this newfound complexity leads to some deeply counter-intuitive behaviors.

With our sense of nearness warped, how do points "stick" to sets, and how do sequences "arrive" at their destinations?

Getting Close, But Not Too Close

Let's look at the open interval $A = (0, 1)$. In the standard world, the points that are "infinitesimally close" to this set—its limit points—are all the points inside it, plus the two endpoints, $0$ and $1$. The closure is the closed interval $[0, 1]$.

On the Sorgenfrey line, things are different. The point $1$ is no longer "close" to the set $(0, 1)$. We can find a neighborhood of $1$, for example $[1, 2)$, that is completely disjoint from $A$. The point $1$ can place a "wall" at its feet that separates it from all points to its left. However, the point $0$ is still a limit point. Any neighborhood of $0$, like $[0, \epsilon)$, immediately steps into the set $(0, 1)$. It has no wall to protect it. Therefore, the closure of $(0, 1)$ in the Sorgenfrey line is the set $[0, 1)$. The topology has surgically removed one endpoint from the closure!

A One-Sided Welcome

This asymmetry has profound consequences for the convergence of sequences. Consider a sequence approaching a point from below, like $x_n = -1/n$, which in our usual intuition converges to $0$. On the Sorgenfrey line, this sequence goes nowhere! To converge to $0$, the terms of the sequence would eventually have to land inside every neighborhood of $0$. But we can pick the neighborhood $U = [0, 1)$. Every single term of the sequence $x_n = -1/n$ is negative, so not a single one ever enters $U$. The sequence is held back by the "ledge" at $0$ and can never arrive.

Now, consider a sequence approaching a point from above, like $x_n = 2 + 1/n^2$. This sequence does converge to $2$. For any neighborhood of $2$, say $[2, 2+\epsilon)$, all the terms $x_n$ are greater than $2$. We just have to wait until $n$ is large enough that $1/n^2 < \epsilon$, and from that point on, all terms of the sequence will be inside the neighborhood. The sequence is "welcomed" because it is approaching from the "correct" side.

Even in this strange world, one comfort remains: if a sequence does manage to converge, its limit is unique. This is because the Sorgenfrey line is a ​​Hausdorff space​​, a property essential for any world where measurements should yield unambiguous results. We can always find disjoint neighborhoods for any two distinct points, preventing a sequence from being confused about its destination.

Zooming out from the behavior of individual points, the global picture of the Sorgenfrey line is even more alien. The familiar, connected number line is gone, replaced by something fragmented and unwieldy.

A Universe of Points

Is the Sorgenfrey line connected? Can you trace a path from one point to another without any breaks? The answer is a resounding no. In fact, the space is ​​totally disconnected​​. For any two distinct points $x$ and $y$, say with $x < y$, we can shatter the space right between them. Consider the set $S = (-\infty, y)$. This set is a union of basis elements like $[a, y)$ for all $a < y$, so it's open. But its complement, $[y, \infty)$, is also open, since it can be written as the union of all sets $[z, z+1)$ for $z \ge y$.

Since we have found a set $S$ that is both open and closed (a "clopen" set), it acts as a perfect separator. Any subset of the line that contains points both inside and outside $S$ is, by definition, disconnected. This reasoning can be applied to any two points, which means the only connected subsets of the Sorgenfrey line are the individual points themselves!. The entire line has been pulverized into a dust of disconnected points, each one its own isolated island.

Covering an Unruly Infinity

Another hallmark of well-behaved spaces is compactness—the idea that you can take any infinite collection of open patches that covers the space and always find a finite number of those patches that still do the job. The interval $[0, 1]$ in the standard topology is compact. The Sorgenfrey line is not.

To see why, consider the following collection of open sets:

This is an open cover of the entire Sorgenfrey line; any real number $x$ is contained in $[n, \infty)$ where $n$ is the integer part of $x$. But can you pick a finite number of these sets and still cover the whole line? No. If you pick any finite subcollection, there will be a smallest integer among their starting points, let's call it $m$. The union of your finite sets will just be $[m, \infty)$, which fails to cover any number less than $m$. The line stretches off to negative infinity in a way that our finite collection of patches can never tame.

Why is the Sorgenfrey line so strange? Is there a single, underlying reason for this collection of bizarre properties? The answer lies in a deeper investigation of its structure, and it connects to one of the great theorems of topology.

Countable Dust, Uncountable Bricks

First, a deceptive bit of familiarity: the rational numbers $\mathbb{Q}$ are still dense in the Sorgenfrey line. Any basic open set $[a, b)$ contains the standard interval $(a, b)$, which is guaranteed to contain a rational number. So, you can always find a rational number "near" any real number, even with our new definition of near. This means the space is ​​separable​​—it has a countable dense subset.

This might lead you to believe the space is somehow "small" or "manageable." But here lies the paradox. The Sorgenfrey line is ​​not second-countable​​. A space is second-countable if you can build its entire topology from a countable collection of basic building blocks. For the standard topology, the set of all intervals $(q_1, q_2)$ with rational endpoints works. It’s a countable set of "bricks". For the Sorgenfrey line, no such countable set of bricks exists. To see why, think about the open sets of the form $[x, x+1)$. For the topology to be valid, any basis $\mathcal{B}$ must contain some basis element $B_x$ such that $x \in B_x \subseteq [x, x+1)$. The only way this can happen is if $B_x$ has $x$ as its left endpoint. Therefore, for every distinct real number $x$, there must be a distinct basis element $B_x$ starting at $x$. Since there are uncountably many real numbers, any basis for the Sorgenfrey line must be uncountable!. We need an uncountable supply of bricks to build this house.

The Recipe for Reality

This failure to be second-countable is the key to everything. In mathematics, we often want to know if a space is ​​metrizable​​—that is, if its topology could have been generated by a distance function, a metric. Metrizable spaces are the ones that feel most like our physical world; they are exceptionally well-behaved.

The ​​Urysohn Metrization Theorem​​ gives us the recipe for metrizability. It states that a space is metrizable if and only if it is ​​Regular, Hausdorff, and Second-Countable​​. The Sorgenfrey line actually satisfies the first two conditions. It's Hausdorff, as we saw. It is also a ​​Normal​​ space (which is stronger than regular), meaning any two disjoint closed sets can be separated by disjoint open neighborhoods. This is a very "nice" property to have.

So, the Sorgenfrey line follows much of the recipe. It is a T4 space (Normal and Hausdorff). Why, then, isn't it metrizable? The Urysohn Metrization Theorem points to the culprit with unerring precision: it is because the Sorgenfrey line is not second-countable. This one missing ingredient is the source of all its pathologies—its shattered connectedness, its obstinate sequences, its non-compactness. It is a world built from an uncountable number of ledges, a space so fine-grained that it resists our attempts to measure it with any ordinary ruler. And it stands as a beautiful testament to the fact that in mathematics, changing just one simple rule can create an entirely new universe.

Now that we have acquainted ourselves with the peculiar rules of the lower limit topology, you might be asking a very fair question: why bother? Is the Sorgenfrey line, this world where intervals are half-open, merely a mathematical curiosity—a strange beast kept in a topological zoo for theoreticians to poke at? The answer, perhaps surprisingly, is a resounding no. The true value of the Sorgenfrey line and its relatives is not in describing the physical world we see, but in helping us understand the logical world we build. It serves as a laboratory, a testing ground where we can push our familiar ideas of space, continuity, and dimension to their limits. By seeing where our intuition breaks, we learn what holds it together.

Let's begin our exploration by looking at familiar sets of numbers through the new lens of the lower limit topology. What happens when we restrict our Sorgenfrey world to just the integers, $\mathbb{Z}$? In the standard topology, the integers are a scattered collection of isolated points, but the space itself is not discrete because the single-point sets $\{n\}$ are not open. On the Sorgenfrey line, however, a wonderful simplification occurs. For any integer $n$, the basic open set $[n, n+1)$ contains exactly one integer: $n$ itself! This means every singleton set $\{n\}$ is an open set. The subspace topology on the integers is the discrete topology, where every point lives in its own private, open bubble. This simple observation is our first clue that the lower limit topology has a powerful "separating" quality.

What about a denser set, like the rational numbers, $\mathbb{Q}$? Here, things get more intricate. A set like $[q, q+\epsilon) \cap \mathbb{Q}$ for a rational $q$ is no longer just a single point; by the density of the rationals, it's packed with infinitely many other rationals. So, the subspace is not discrete. Yet, it possesses a remarkable property that its parent space, the Sorgenfrey line $\mathbb{R}_l$, lacks. While the Sorgenfrey line requires an uncountable number of basis elements to describe its topology (one for each starting point $x \in \mathbb{R}$), its rational subspace is "second-countable." We can form a perfectly good basis for the topology on $\mathbb{Q}$ using only intervals of the form $[q_1, q_2) \cap \mathbb{Q}$, where both endpoints $q_1$ and $q_2$ are rational. Since there are only a countable number of such pairs, we have found a countable basis. This is a beautiful paradox: the enormous, uncountable Sorgenfrey line contains a countable, dense "skeleton" that is, in a topological sense, much simpler than the whole.

Emboldened by our findings on the line, let's venture into a higher dimension. What kind of world is the Sorgenfrey plane, $\mathbb{R}_l \times \mathbb{R}_l$, whose basic open sets are "bottom-left-inclusive" rectangles $[a, b) \times [c, d)$?

At first glance, some things might seem familiar. If you consider a straight line within this plane, like a horizontal line $y = c_0$ or even the main diagonal $y=x$, the topology you find on that line is just the good old Sorgenfrey line topology you started with. It seems as though the structure is behaving consistently.

But this is a deceptive calm before a topological storm. Let's ask a simple question: can we draw a path from one point to another in the Sorgenfrey plane? In our familiar Euclidean plane, of course we can. But in the Sorgenfrey plane, the answer is a shocking no. Any continuous path from the unit interval $[0,1]$ into the Sorgenfrey plane must be a constant function—it cannot move at all! The space is not just disconnected; it's totally disconnected. Why? The basis rectangles $[a,b) \times [c,d)$ are not just open sets; they are also closed sets. The entire plane is shattered into a collection of "clopen" (closed and open) pieces, a fine dust of points that cannot be continuously bridged. The continuous image of a connected set like $[0,1]$ must be connected, but the only connected subsets of the Sorgenfrey plane are single points. Thus, no path can ever get from a point $P$ to a different point $Q$. This bizarre property makes the Sorgenfrey plane an invaluable counterexample, reminding us that path-connectedness is not a given, but a precious property of our Euclidean intuition.

The Sorgenfrey line's greatest service is perhaps as a diagnostic tool, revealing hidden structures and assumptions in other branches of mathematics.

​​Calculus and Linear Algebra:​​ Consider a function you know well from linear algebra, the determinant of a $2 \times 2$ matrix. As a map from the space of matrices (which we can think of as $\mathbb{R}^4$) to the real numbers, this function is beautifully continuous and smooth. Its graph is a saddle-shaped surface. Now, let's ask: is this map still continuous if we change the topology on the target space from the standard real line to the Sorgenfrey line $\mathbb{R}_l$? The answer is a surprising and universal "no." The determinant map is not continuous at any matrix. The reason is profound. For a function to be continuous into $\mathbb{R}_l$, every point in its domain must be a local minimum with respect to its image. That is, for a matrix $A$, there must be a small neighborhood of $A$ where all matrices have a determinant greater than or equal to $\det(A)$. But the determinant function almost never has local minima; it's a saddle, meaning you can always move in some direction to make the determinant smaller. The Sorgenfrey line's topology has revealed a deep geometric property of the determinant function that was always there, but invisible in the standard topology.

​​Measure Theory:​​ Let's turn to the foundations of probability and integration. A topology on $\mathbb{R}$ generates a collection of "measurable" sets called a $\sigma$-algebra. Since the Sorgenfrey topology $\mathcal{T}_l$ has far more open sets than the standard topology $\mathcal{T}_u$, one would instinctively assume it must generate a much larger, more complex $\sigma$-algebra. But this is not the case. In one of the most elegant surprises in the field, it turns out that both topologies generate the exact same Borel $\sigma$-algebra. How can this be? The key is that the "new" open sets of the Sorgenfrey line, the intervals $[a, b)$, can already be constructed using the tools of the standard $\sigma$-algebra (for example, $[a, b) = [a, \infty) \cap (-\infty, b)$). While topologically distinct, the two systems are identical from the perspective of what can be measured. This teaches us a crucial lesson: a finer topology does not necessarily lead to a finer measure theory.

​​Advanced Topology:​​ Finally, the Sorgenfrey line serves as a key character in some of the most important theorems of general topology. It is a "normal" space, meaning any two disjoint closed sets can be separated by disjoint open sets. This property is the key that unlocks the powerful ​​Tietze Extension Theorem​​. This theorem guarantees that any continuous real-valued function defined on a closed subset of a normal space can be continuously extended to the entire space. For example, the standard Cantor set is closed in $\mathbb{R}_l$, so any continuous function you can imagine on the Cantor set is guaranteed to have a continuous extension to the whole Sorgenfrey line. At the same time, the Sorgenfrey line provides cautionary tales. While it is a perfectly nice Hausdorff space, its one-point compactification is not. The point at infinity cannot be separated from any other point, which means the compactified space is not even Hausdorff, and therefore cannot be described by any metric.

From a simple tweak of an interval, we have uncovered a universe of behavior that challenges, refines, and ultimately deepens our understanding of the mathematical world. The Sorgenfrey line is more than a counterexample; it is a teacher, demonstrating with beautiful clarity the hidden connections and subtle dependencies that form the true foundation of analysis and geometry.