Sorgenfrey plane

In the study of topology, our intuition is often shaped by the familiar properties of Euclidean space. However, to truly understand the boundaries of topological theorems, mathematicians construct 'pathological' spaces that challenge these assumptions. The Sorgenfrey plane is one of the most famous and instructive of these constructions—a seemingly simple variation on the Cartesian plane that harbors a wealth of counter-intuitive properties.

This article addresses the gap between our standard geometric intuition and the rigorous definitions of topological properties like normality, compactness, and metrizability. By dissecting the Sorgenfrey plane, we reveal why certain 'obvious' properties do not always hold, especially when dealing with product topologies.

We will embark on a journey into this peculiar landscape. The "Principles and Mechanisms" section will delve into its fundamental construction from half-open intervals and uncover the strange behavior of its anti-diagonal subset. Following this, the "Applications and Interdisciplinary Connections" section will demonstrate its crucial role as a powerful counterexample that shatters common geometric shapes and disproves plausible-sounding conjectures in topology.

In the introduction, we hinted that the Sorgenfrey plane is a strange and wonderful beast, a topological space that defies many of our intuitions built from the familiar world of Euclidean geometry. Now, it is time to venture into this peculiar landscape and understand its inner workings. Like a physicist taking apart a watch, we will examine its gears and springs to see what makes it tick. Our journey will reveal not just a collection of odd properties, but a beautiful, logical chain of consequences stemming from a single, simple change in our definition of "openness."

Our story begins not with the plane, but with a line. In the standard topology of the real number line, the fundamental building blocks are open intervals $(a, b)$. They are democratic, treating their endpoints with impartiality—neither $a$ nor $b$ is included. The ​​Sorgenfrey line​​, $\mathbb{R}_l$, is born from a small act of rebellion. It declares that its fundamental open sets will be ​​half-open intervals​​ of the form $[a, b)$, including the left endpoint but excluding the right.

What happens when we build a plane from this new line? The ​​Sorgenfrey plane​​, $\mathbb{R}_l \times \mathbb{R}_l$, is the Cartesian product of two Sorgenfrey lines. Its basic open sets, the "bricks" from which all other open sets are constructed, are rectangles of the form $[a, b) \times [c, d)$.

Imagine what this looks like. Unlike a standard open rectangle $(a, b) \times (c, d)$, which is a boundary-less region, a Sorgenfrey basis rectangle includes its bottom edge and its left edge, but not its top or right edges. Crucially, it includes its bottom-left corner point $(a, c)$. Every basic open set is "anchored" at this corner and extends exclusively "northeast" from there. This seemingly minor detail—this bias for the bottom-left—is the source of all the plane's bizarre and fascinating behaviors.

At first glance, some things might still look familiar. For instance, if we consider the first quadrant, $A = \{(x,y) \mid x > 0, y > 0 \}$, its interior and closure in the Sorgenfrey plane are exactly the same as in the standard Euclidean plane. But this is a deceptive calm before the storm. The true nature of this space is revealed when we examine a very special subset.

Let's turn our attention to the line defined by the equation $y = -x$. We'll call it the ​​anti-diagonal​​, $L = \{(x, -x) \mid x \in \mathbb{R}\}$. In the familiar Euclidean plane, this is just an ordinary, unremarkable line. In the Sorgenfrey plane, it is the star of the show.

First, is this line open or closed? Let's check if its complement, the set of all points not on the line, is open. A point $(x, y)$ is not on the line if $x+y \neq 0$. If $x+y > 0$, we can draw a little Sorgenfrey rectangle $[x, x+\epsilon) \times [y, y+\delta)$ around it. For any point $(u, v)$ in this rectangle, we have $u \ge x$ and $v \ge y$, so $u+v \ge x+y > 0$. The rectangle entirely avoids the line $L$. A similar argument works if $x+y < 0$. This means the complement of $L$ is open, which in turn means the anti-diagonal $L$ is a ​​closed set​​ in the Sorgenfrey plane. So far, so good.

Now for the first great surprise. What does the topology of the Sorgenfrey plane look like from the perspective of the line L? We are asking about the subspace topology on $L$. Let's pick an arbitrary point $p = (x_0, -x_0)$ on the line. Can we find an open set in the Sorgenfrey plane that intersects $L$ only at this single point?

Consider the Sorgenfrey basis rectangle $B = [x_0, x_0+1) \times [-x_0, -x_0+1)$. The point $p$ is certainly in this rectangle. Now, let's look for any other points of $L$ that might be in $B$. A point $(x, -x)$ is in $B$ if and only if $x_0 \le x < x_0+1$ and $-x_0 \le -x < -x_0+1$. The second inequality, when we multiply by $-1$, becomes $x_0 \ge x > x_0-1$. So, for a point to be in the intersection, its $x$-coordinate must satisfy both $x_0 \le x$ and $x \le x_0$. The only number that can do that is $x_0$ itself!

This means the big open rectangle $B$ from the plane touches the line $L$ at exactly one point: our chosen point $p$. In the subspace topology of $L$, the set $\{p\}$ is equal to $B \cap L$, the intersection of an open set with $L$. By definition, this means the single-point set $\{p\}$ is an open set in $L$. Since we can do this for any point on the line, every single point of the anti-diagonal is an isolated open set. The line $L$ has the ​​discrete topology​​.

Think about what this means. We have a set that looks like a continuous line, but topologically, it's just an uncountable collection of disconnected, isolated points. It's as if each point on the line is in its own universe, utterly separate from its neighbors. This is a profound break from our Euclidean intuition, and it's the key that unlocks the rest of the Sorgenfrey plane's secrets.

This single discovery—that the closed anti-diagonal $L$ is a discrete subspace—triggers a domino rally of astonishing consequences.

First, the Sorgenfrey plane is ​​not compact​​. A compact space is one where any open cover has a finite subcover—a kind of topological finiteness. In a compact space, every closed subset must also be compact. We've established that the anti-diagonal $L$ is a closed subset of the plane. But is $L$ compact? As an infinite discrete space, its open cover consisting of all its singleton points, $\{\{(x, -x)\} \mid x \in \mathbb{R}\}$, has no finite subcover. Thus, $L$ is not compact. Since the Sorgenfrey plane contains a non-compact closed subset, the plane itself cannot be compact.

Now for the most celebrated property. A topological space is called ​​normal​​ if any two disjoint closed sets can be enclosed in two disjoint open "bubbles". This sounds like a very reasonable property; a sheet of paper is normal, as is the 3D space we live in. We can always draw a line between two separate closed shapes. The Sorgenfrey plane, however, is ​​not normal​​.

To see why, we use our strange anti-diagonal $L$. Let's split it into two sets:

• $A = \{(q, -q) \mid q \in \mathbb{Q}\}$, the points on $L$ with rational coordinates.
• $B = \{(r, -r) \mid r \in \mathbb{R}\setminus\mathbb{Q}\}$, the points on $L$ with irrational coordinates.

Since $L$ is a closed set with the discrete topology, any subset of $L$ is also closed in the Sorgenfrey plane. Thus, $A$ and $B$ are two disjoint closed sets. If the plane were normal, we should be able to find disjoint open sets $U$ and $V$ such that $A \subseteq U$ and $B \subseteq V$.

But this is impossible. Imagine trying to build the open set $U$ to cover all the rational points in $A$. For each point $(q, -q)$, we must place a Sorgenfrey rectangle $[q, q+\epsilon) \times [-q, -q+\delta)$ around it. Remember, these rectangles always point northeast. Now, the rational numbers and irrational numbers are infinitely interwoven. No matter how small you make your rectangles around the rational points, because they extend to the right, they will inevitably intrude into the space right next to an irrational point. Any open set $U$ that successfully covers all of the dense set $A$ will invariably "spill over" and intersect any open set $V$ designed to cover $B$. There is no way to build these two open bubbles without them touching.

This makes the Sorgenfrey plane the classic counterexample to the statement that the product of two normal spaces must be normal. The Sorgenfrey line $\mathbb{R}_l$ is a normal space, but when you multiply it by itself, this desirable property is destroyed.

However, the Sorgenfrey plane is not a complete topological anarchist. It is, in fact, a ​​regular space​​. A space is regular if you can separate a point from a disjoint closed set. This is a weaker condition than normality. The reason the Sorgenfrey plane is regular is quite elegant. The basis sets $[a, b)$ of the Sorgenfrey line are not just open; they are also closed (they are ​​clopen​​). This means the basis rectangles of the plane are also clopen. Any T1 space (which the Sorgenfrey plane is) that has a basis of clopen sets is guaranteed to be regular. So, the Sorgenfrey plane occupies a specific, well-defined place in the hierarchy of spaces: it is regular, but not normal.

We have discovered a space that is regular and Hausdorff, but not compact and not normal. What is the ultimate implication of this? It means the Sorgenfrey plane is ​​not metrizable​​.

A space is metrizable if its topology can be generated by a distance function, or metric. The familiar Euclidean distance is the metric that gives rise to the standard topology on $\mathbb{R}^2$. The Urysohn Metrization Theorem gives us conditions under which a space is metrizable. A necessary condition derived from this and other theorems is that any metrizable space must be normal.

Since we have proven, with our clever anti-diagonal argument, that the Sorgenfrey plane is not normal, it cannot be metrizable. There is no conceivable distance function $d((x_1, y_1), (x_2, y_2))$ that could produce these peculiar half-open rectangles as its fundamental "open balls."

And so, our journey ends. By starting with a simple twist on the definition of an open interval, we constructed a geometric world that looks familiar on the surface but contains a hidden, discrete line. This line, in turn, unravels the plane's compactness and normality, ultimately revealing a space whose topology is so strange that it cannot be described by any notion of distance. The Sorgenfrey plane is more than a mathematical curiosity; it is a powerful illustration of the depth and subtlety of topology, a reminder that even the simplest rules can generate worlds of unexpected complexity and beauty.

Now that we have acquainted ourselves with the peculiar rules of the Sorgenfrey plane, you might be asking a perfectly reasonable question: Why? Why would mathematicians concoct such a seemingly bizarre and unnatural space? We don't live on a plane where you can only step forward and to the right, so what is the point of studying it?

The answer is the same reason a physicist might create a near-perfect vacuum or temperatures near absolute zero. To truly understand a law of nature—or a mathematical principle—you must push it to its limits. You must see where it breaks down. The Sorgenfrey plane, this strange world built from half-open intervals, is a topologist's laboratory. It is a place where our comfortable intuitions, forged in the familiar Euclidean world, are put to the ultimate test. By exploring what goes wrong in the Sorgenfrey plane, we gain a much deeper appreciation for the hidden structures that make our own world so beautifully coherent. It is a source of profound counterexamples that have shaped the foundations of modern topology.

Let's begin our tour by taking a few familiar geometric objects and placing them into the Sorgenfrey plane to see what happens. The results are often startling.

Consider the simplest of shapes: a straight line. What happens to the main diagonal line, $L_1 = \{(x,x) \mid x \in \mathbb{R}\}$, and the anti-diagonal, $L_2 = \{(x,-x) \mid x \in \mathbb{R}\}$? On the standard plane, they are identical—you can rotate one to get the other. Here, they could not be more different. The main diagonal inherits a topology that makes it, essentially, another Sorgenfrey line. But the anti-diagonal is another story entirely. If you try to take a small neighborhood around a point $(x_0, -x_0)$ on the anti-diagonal, like the Sorgenfrey rectangle $[x_0, x_0+\varepsilon) \times [-x_0, -x_0+\varepsilon)$, you find that the only point from the anti-diagonal it contains is $(x_0, -x_0)$ itself! Every single point on the anti-diagonal is isolated from every other. The line has been shattered into an uncountable dust of disconnected points. This immediately reveals a fundamental truth about the Sorgenfrey plane: it is not isotropic. Direction matters.

This "shattering" effect is not unique to the anti-diagonal. Let's take the unit circle, $S^1$. In our world, it is the very definition of a smooth, connected loop. But when viewed as a subspace of the Sorgenfrey plane, it too begins to fracture. For any point on the circle in the first quadrant, like $(\frac{1}{\sqrt{2}}, \frac{1}{\sqrt{2}})$, we can draw a tiny Sorgenfrey box starting at that point and extending up and to the right. Because the circle curves "down and away," this box will contain no other points of the circle. These points become isolated. The familiar connected circle is revealed to be a much finer, stranger object, full of isolated points.

The story gets even more curious with a parabola, like $P = \{(x, x^2)\}$. The right half of the parabola (where $x \ge 0$) behaves much like the Sorgenfrey line. But the left half (where $x < 0$) is completely different. For any point $(x_0, x_0^2)$ with $x_0 < 0$, a small Sorgenfrey box starting at that point will miss all other points on the parabola, because the curve immediately goes "down" while the box goes "up". So, the left half of the parabola shatters into isolated points. The parabola becomes a topological chimera: one part a continuous (in the Sorgenfrey sense) curve, the other part a discrete collection of points.

If simple shapes are so profoundly altered, can one even "move" in this space? In topology, movement is captured by the idea of an arc or a path. In the Euclidean plane, we can draw a continuous path from any point to any other. In the Sorgenfrey plane, this is impossible. Any attempt to draw a path from one point to another inevitably fails. The strict "forward and right" nature of the open sets prevents any continuous back-and-forth motion required for a path. The shocking conclusion is that the only arcwise connected subsets are the individual points themselves. The Sorgenfrey plane is a universe of frozen moments, with no way to travel between them. This phenomenon extends even to complex shapes like the topologist's sine curve, a classic example of a connected but not path-connected space. In the Sorgenfrey plane, it is rendered totally disconnected.

The most significant role of the Sorgenfrey plane in mathematics is as a powerful counterexample. It is a graveyard for many plausible-sounding conjectures. It teaches us that many properties we take for granted are not automatically preserved when we combine spaces.

The most famous of these is the property of ​​normality​​. A normal space is one where any two disjoint closed sets can be cleanly separated by disjoint open "neighborhoods." This is a very desirable property, crucial for constructing many important functions. The Sorgenfrey line, $\mathbb{R}_l$, is normal. You might naturally assume, then, that the product of two Sorgenfrey lines—our Sorgenfrey plane—would also be normal.

This is false. The Sorgenfrey plane is the canonical example of a space that is not normal. The proof is subtle, but it relies on the properties of that strange anti-diagonal we discovered earlier. It constitutes a closed set, and so does the "rational" anti-diagonal, and these two disjoint closed sets cannot be separated by open sets. So, the property of normality is not "productive"—it is not preserved when taking products. Curiously, while the whole space is not normal, many of its subspaces are. A horizontal line is normal, and so is the discrete anti-diagonal. Yet, an open subset like the "open" unit square is homeomorphic to the entire plane and is therefore also not normal.

The Sorgenfrey plane demolishes other "obvious" theorems as well.

• A space is a ​​Baire space​​ if the intersection of any countable collection of dense open sets is still dense. Think of it as a space that isn't "too thin" or "full of holes." Complete metric spaces like the real line are Baire spaces. The Sorgenfrey line is also a Baire space. But their product, the Sorgenfrey plane, is not. Again, a "good" property is lost in the product.
• A space is ​​Lindelöf​​ if every open cover has a countable subcover. This is a crucial "smallness" condition. The Sorgenfrey line is Lindelöf. The Sorgenfrey plane is not.

In each case, the Sorgenfrey plane serves as a vital warning sign to mathematicians: be careful with your assumptions! The preservation of a property under an operation like the Cartesian product is not guaranteed and must be proven.

With all its strange properties, where does the Sorgenfrey plane fit in the grand "zoo" of topological spaces? Can it be seen as a sub-world of some other, perhaps even stranger, space?

Consider the plane with the ​​lexicographical order topology​​, $(\mathbb{R} \times \mathbb{R})_{lex}$, where points are ordered like words in a dictionary. This space is also famously non-metrizable and possesses its own bizarre features, like being connected but having its paths constantly "broken" by uncountable gaps. One might wonder if our Sorgenfrey plane could be "embedded" into this ordered plane—that is, if it is topologically equivalent to a subspace of it.

The answer is a definitive no, and the reason is a beautiful piece of topological detective work. The Sorgenfrey plane has two seemingly contradictory properties: it is ​​separable​​ (it contains a countable dense subset, namely $\mathbb{Q} \times \mathbb{Q}$), but it also contains an ​​uncountable discrete subspace​​ (our old friend, the anti-diagonal). It turns out that this combination is impossible for any subspace of the lexicographically ordered plane. Any separable subspace of a linearly ordered space cannot contain an uncountable number of isolated points. The Sorgenfrey plane fails this test spectacularly. It is, in this specific sense, a fundamentally different kind of beast from the ordered plane.

From shattering familiar shapes to demolishing fundamental theorems, the Sorgenfrey plane is more than a mere curiosity. It is an indispensable tool. It sharpens our understanding, refines our theorems, and reveals the profound and often surprising beauty hidden in the abstract world of topology. It reminds us that even in the purest of mathematics, progress is often made by exploring the exceptions, the pathologies, and the places where our intuition runs aground.