Quadratic Extension

From the moment we first encounter an equation we cannot solve, like $x+1=0$ with only positive numbers, mathematics has taught us to expand our world. The need to solve equations like $x^2-2=0$ within the rational numbers forces us to create a new, larger field containing numbers like $\sqrt{2}$. The simplest and most foundational of these expansions are quadratic extensions, which represent the first step beyond our familiar numerical territory. But what does it mean to build these new algebraic worlds, and what hidden structures do they possess?

This article delves into the core of quadratic extensions. It addresses the fundamental question of how these structures are defined by the base fields they extend and what properties emerge from their creation. You will gain a clear understanding of their elegant symmetries and the rules governing their construction.

We will begin in "Principles and Mechanisms" by exploring the definition of a quadratic extension and its degree. We will uncover its inherent symmetries through the lens of Galois theory and investigate what happens when we stack these simple blocks, revealing both predictable outcomes and surprising complexities. We will also examine how the arithmetic of prime numbers changes within these new fields. Following this, the "Applications and Interdisciplinary Connections" chapter will demonstrate the profound impact of this theory, showing how it solves ancient Greek construction problems and serves as a cornerstone for modern concepts in algebraic number theory, from cyclotomic fields to the local-global principle.

In our journey to understand the structure of numbers, we often find ourselves needing to expand our world. We start with whole numbers, but soon we need fractions. We have positive numbers, but the simple equation $x+1=0$ forces us to invent negative numbers. What happens when we encounter an equation like $x^2 - 2 = 0$? Within the realm of rational numbers, there is no solution. So, we do what we always do: we invent a new number, which we call $\sqrt{2}$, and we create a new, larger world, a ​​field extension​​, that contains it. The simplest and most fundamental of these expansions are the ​​quadratic extensions​​, which are, in a sense, the first step beyond our familiar territory. But what does it truly mean to take this step?

Imagine a number system, a ​​field​​, which we can call $F$. This is our starting point, our "home ground." It could be the rational numbers $\mathbb{Q}$, the real numbers $\mathbb{R}$, or even a finite world like the integers modulo 5, $\mathbb{Z}_5$. A quadratic extension is what we get when we are forced to solve a quadratic equation, like $ax^2+bx+c=0$, whose solutions are not in our home field $F$.

Let's consider the seemingly innocent equation $x^2 + 1 = 0$. You might immediately think of the imaginary unit $i = \sqrt{-1}$. But whether we actually need to invent a new number depends entirely on our home field.

If our home field is the rational numbers $\mathbb{Q}$, there is no fraction you can square to get $-1$. The polynomial $x^2+1$ is ​​irreducible​​—it cannot be factored into simpler polynomials with rational coefficients. To solve the equation, we must append a new element, let's call it $\alpha$, satisfying $\alpha^2 = -1$. Our new world, $\mathbb{Q}(\alpha)$, consists of all numbers of the form $a + b\alpha$, where $a$ and $b$ are rational. This new field is "twice as big" as $\mathbb{Q}$ in a specific sense: every element is described by two rational numbers ($a$ and $b$), much like a point on a plane is described by two coordinates. We say the ​​degree of the extension​​ is 2.

The same is true if we start with the real numbers $\mathbb{R}$. No real number squares to $-1$. So, we append a root of $x^2+1=0$, which we famously call $i$, and we get the field of complex numbers, $\mathbb{C}$. Again, the degree of the extension $[\mathbb{C}:\mathbb{R}]$ is 2.

But what if our home field is the finite world of integers modulo 5, $\mathbb{Z}_5 = \{0, 1, 2, 3, 4\}$? Let's check: $0^2 \equiv 0$, $1^2 \equiv 1$, $2^2 \equiv 4$, $3^2 \equiv 9 \equiv 4$, and $4^2 \equiv 16 \equiv 1$. Notice that $4 \equiv -1 \pmod{5}$. So, in the world of $\mathbb{Z}_5$, the equation $x^2+1=0$ does have solutions, namely $x=2$ and $x=3$. The polynomial $x^2+1$ is not irreducible; it factors as $(x-2)(x-3)$. We don't need to invent anything! The field "extended" with a root of $x^2+1$ is just $\mathbb{Z}_5$ itself. The extension has degree 1.

This simple example reveals the core principle: a quadratic extension arises from an irreducible quadratic polynomial. The decision of whether an extension is truly quadratic is a local affair, entirely dependent on the properties of the base field.

Quadratic extensions are not just simple; they are exceptionally well-behaved. They possess a property called ​​normality​​. In simple terms, an extension is normal if, whenever it contains one root of a polynomial from the base field, it must contain all the roots of that polynomial.

For any quadratic extension $F(\sqrt{d})$, the minimal polynomial is $x^2 - d = 0$. If we have one root, $\sqrt{d}$, the other root is $-\sqrt{d}$. Since our new world is made of elements like $a+b\sqrt{d}$, the other root, $-\sqrt{d} = 0 + (-1)\sqrt{d}$, is automatically included. This is always true for degree-2 extensions.

This might seem obvious, but it's a special property. Consider the extension $\mathbb{Q}(\sqrt[3]{2})$. The minimal polynomial for $\sqrt[3]{2}$ is $x^3 - 2 = 0$. The three roots of this equation are $\sqrt[3]{2}$, $\sqrt[3]{2}\omega$, and $\sqrt[3]{2}\omega^2$, where $\omega$ is a complex cube root of unity. The field $\mathbb{Q}(\sqrt[3]{2})$ contains only real numbers, so it's missing the two complex roots. It is not a normal extension. It's as if we've invited one member of a family to a party but left its siblings out in the cold. Quadratic extensions are always polite enough to invite the whole family.

This "completeness" is tied to a deeper concept: symmetry. The set of all transformations of an extension that preserve the base field's structure forms a group, the ​​Galois group​​. Think of it as the set of all possible "shuffles" of the new numbers that leave the old numbers untouched. For a quadratic extension like $\mathbb{Q}(\sqrt{d})$ over $\mathbb{Q}$, what shuffles are possible?

An element is of the form $a+b\sqrt{d}$. Any symmetry must leave $a$ and $b$ alone, since they are rational. The only thing it can possibly change is $\sqrt{d}$. Since a symmetry must send roots of a polynomial to other roots of the same polynomial, it must send $\sqrt{d}$ to either $\sqrt{d}$ or $-\sqrt{d}$.

1. The ​​identity​​ symmetry: $a+b\sqrt{d} \mapsto a+b\sqrt{d}$. This does nothing.
2. The ​​conjugation​​ symmetry: $a+b\sqrt{d} \mapsto a-b\sqrt{d}$.

That's it! There are only two symmetries. If you apply the conjugation twice, you get back to where you started: $a-b\sqrt{d} \mapsto a-(-b\sqrt{d}) = a+b\sqrt{d}$. This two-element group, where the only non-trivial element flips things back and forth, is the simplest non-trivial group in the universe: the ​​cyclic group of order 2​​, $C_2$. The algebraic degree of the extension, 2, is perfectly mirrored by the size of its symmetry group. This beautiful correspondence is the heart of Galois theory.

If quadratic extensions are the fundamental building blocks, what happens when we stack them?

Let's start with $\mathbb{Q}$ and append $\sqrt{2}$ to get $K = \mathbb{Q}(\sqrt{2})$. Then, let's append $\sqrt{3}$ to get $L = \mathbb{Q}(\sqrt{2}, \sqrt{3})$. We've performed two consecutive quadratic extensions. The final field $L$ contains numbers like $a + b\sqrt{2} + c\sqrt{3} + d\sqrt{6}$. The degree of this extension over $\mathbb{Q}$ is 4. What about its symmetries? The Galois group has four elements, corresponding to the independent choices of flipping the signs of $\sqrt{2}$ and $\sqrt{3}$. This group is the ​​Klein four-group​​, $C_2 \times C_2$. It’s like having two independent light switches.

You might think that stacking well-behaved normal extensions will always produce a larger, well-behaved normal extension. But here, mathematics throws us a wonderful curveball.

Let's again start with $F=\mathbb{Q}$ and build $K = \mathbb{Q}(\sqrt{2})$, which is a normal extension. Now, consider the element $\alpha = \sqrt{1+\sqrt{2}}$. This number is not in $K$. Let's build a new field $L=K(\alpha) = \mathbb{Q}(\sqrt{1+\sqrt{2}})$. The extension $L/K$ is a degree-2 extension, so it is also normal. We have a tower of normal extensions: $F \subset K \subset L$.

But is the total extension $L/\mathbb{Q}$ normal? To find out, we need the minimal polynomial of $\alpha$ over $\mathbb{Q}$. From $\alpha^2 = 1+\sqrt{2}$, we get $(\alpha^2-1)^2 = 2$, which expands to $\alpha^4 - 2\alpha^2 - 1 = 0$. The four roots of this equation are $\pm\sqrt{1+\sqrt{2}}$ and $\pm\sqrt{1-\sqrt{2}}$. Our field $L$ contains $\alpha = \sqrt{1+\sqrt{2}}$, which is a real number. But look at the other roots: $1-\sqrt{2}$ is negative, so $\sqrt{1-\sqrt{2}}$ is a complex number! Our field $L$ is entirely contained within the real numbers and cannot possibly contain these complex roots. Since it doesn't contain all the siblings, the extension $L/\mathbb{Q}$ is ​​not normal​​. This is a profound lesson: local tidiness does not guarantee global order. The way blocks are stacked matters just as much as the blocks themselves.

Expanding a number field doesn't just add new numbers; it fundamentally alters the arithmetic landscape. Consider the prime numbers in our home field $\mathbb{Z}$: 2, 3, 5, 7, 11, ... They are the indivisible atoms of arithmetic. What happens to them when we move to the ring of integers of a quadratic extension, say $\mathbb{Q}(\sqrt{-77})$?

The fate of a prime $p$ in this new world is intimately tied to the number we used to create it, $D=-77$. A prime from the old world can face one of three fates in the new:

1. It can ​​remain inert​​, staying prime in the new world as well.
2. It can ​​split​​, breaking apart into a product of two distinct new prime elements.
3. It can ​​ramify​​, becoming the square of a new prime element. Think of it like a single root splitting into a double root. This is a special case, a point of singularity in the arithmetic.

The great discovery of 19th-century mathematicians was that there is a simple rule governing this behavior. A prime $p$ ramifies in $\mathbb{Q}(\sqrt{D})$ if and only if it divides the discriminant of the field (with a special rule for the prime 2). For $\mathbb{Q}(\sqrt{-77})$, where $d=-77 \equiv 3 \pmod 4$, the discriminant is $4d = -308 = -4 \times 7 \times 11$. The primes that ramify are precisely 2, 7, and 11. All other primes will either split or remain inert, based on another elegant rule. This connection is breathtaking: the seemingly abstract algebraic choice of $d$ acts as a master blueprint, dictating the very fabric of factorization and primality in the new universe we've constructed.

This principle extends far beyond quadratic fields and forms the basis of algebraic number theory, connecting the structure of field extensions to the deepest properties of numbers. In some sense, the "interesting" arithmetic of an extension happens at the primes that ramify, those that divide the discriminant.

We've seen that quadratic extensions are simple, symmetric, and foundational. But their importance goes even deeper. They represent a fundamental barrier in the structure of all fields.

A field is called ​​algebraically closed​​ if every polynomial equation with coefficients in the field has a solution within that same field. The complex numbers $\mathbb{C}$ are the most famous example. The rational numbers $\mathbb{Q}$ and real numbers $\mathbb{R}$ are not.

The Artin-Schreier theorem provides a stunning characterization of this property. It states that if a field $K$ is not algebraically closed, but its algebraic closure $\bar{K}$ is a finite extension of it, then the degree of that extension, $[\bar{K}:K]$, must be exactly 2!

Let that sink in. A field cannot be "three steps away" or "five steps away" from being algebraically complete in a finite sense. If you are not at the summit of algebraic completion, but you are a finite distance from it, you are precisely one quadratic extension away. The only way to finitely extend a field to reach algebraic closure is via a degree-2 step.

The quintessential example is the real numbers $\mathbb{R}$. They are not algebraically closed (e.g., $x^2+1=0$ has no solution). Their algebraic closure is the complex numbers $\mathbb{C} = \mathbb{R}(i)$. And, of course, $[\mathbb{C}:\mathbb{R}]=2$. A field like $\mathbb{R}$, whose only finite extension leads to algebraic closure, is called a ​​real closed field​​.

This places quadratic extensions on a unique pedestal. They are not just one type of extension among many. They are, in a profound sense, the final and only bridge between certain number systems and the universe of complete algebraic solvability. From the simple act of solving $x^2-d=0$, we have uncovered a story of symmetry, structure, and a deep, universal principle governing the very nature of fields.

After our journey through the algebraic heartland of quadratic extensions, one might be tempted to ask, "What is all this for?" It is a fair question. The abstract beauty of constructing new number fields by adjoining square roots is a reward in itself for the mathematically inclined, but the true wonder of this concept lies in its astonishing reach. Like a master key, the theory of quadratic extensions unlocks profound connections between seemingly disparate realms of thought—from the geometric constructions of the ancient Greeks to the deepest streams of modern number theory. It is not merely a topic within algebra; it is a powerful lens through which we can perceive a hidden unity in the mathematical landscape.

Our story begins in a world of pure geometry, with the tools of the ancient Greek mathematicians: an unmarked straightedge and a compass. With these, they could construct lines, circles, and an impressive array of geometric figures. But they also hit frustrating limits. They could not, for example, trisect an arbitrary angle, or "square the circle." For two millennia, these remained open challenges, until the 19th century when the language of algebra finally provided the answer.

The breakthrough came from re-framing the question. What numbers can be constructed? If we start with a segment of length 1, what other lengths can we create? We can add and subtract lengths, and using similar triangles, we can multiply and divide them. This means we can construct any length corresponding to a rational number, $\mathbb{Q}$. The compass, however, adds a new power: it can draw circles, whose equations are quadratic. The intersection of lines and circles leads to solutions of quadratic equations, which means we can construct lengths like $\sqrt{2}$.

This is the crucial link. A number is constructible if and only if it belongs to a field that can be reached from the rational numbers through a finite sequence of quadratic extensions. Each application of the compass to find a new, non-rational point corresponds to taking a square root—a single step up a tower of fields.

Consider the number $\alpha = \sqrt{5+\sqrt{5}}$. At first glance, it looks complicated. But from an algebraic perspective, its construction is laid bare. We begin in the familiar field $\mathbb{Q}$. We cannot construct $\alpha$ directly. First, we need the length $\sqrt{5}$. This requires one quadratic step, taking us to the field $\mathbb{Q}(\sqrt{5})$. Now, working within this new, larger field, the number $5+\sqrt{5}$ is available to us. To get $\alpha$, we need to take its square root. This is our second quadratic step, which lifts us to the field $\mathbb{Q}(\sqrt{5+\sqrt{5}})$. The elegant algebraic tower, $\mathbb{Q} \subset \mathbb{Q}(\sqrt{5}) \subset \mathbb{Q}(\sqrt{5+\sqrt{5}})$, is a precise blueprint for a geometric construction. Each step of degree 2 represents a specific, achievable geometric operation.

This principle not only tells us what is possible, but also what is impossible. The reason for the Greeks' failures was that problems like trisecting the angle or doubling the cube algebraically require solving cubic equations, which cannot, in general, be reduced to a sequence of quadratic ones.

The connection runs even deeper. For a certain class of irreducible quartics, whether a real root is constructible depends on a subtle property of a related cubic equation, called the resolvent cubic. A real root of the quartic is constructible only if this associated cubic has a rational root, which ensures that the symmetries of the quartic equation are simple enough to be resolved by a sequence of square roots.

To truly appreciate the boundaries of this algebraic world, we can ask a whimsical question: what if we tried to "construct" functions instead of numbers? Is a function like $f(x) = \sqrt{\sin(x)}$ constructible over the base field of rational functions, $\mathbb{Q}(x)$? The rules are the same: can we obtain it through a tower of quadratic extensions? The answer is a definitive no. The reason is profound. Every function in such a tower must be algebraic over $\mathbb{Q}(x)$. But the function $\sin(x)$ is not—it is transcendental, satisfying no polynomial equation with rational function coefficients. It belongs to a different universe of functions entirely. This powerful negative result clarifies the domain of our theory: constructibility is a story written in the language of algebra.

Leaving the visual world of geometry, we now turn inward, to the architecture of the number system itself. Here, quadratic extensions are not just a tool; they are fundamental, structural components. One of the most important families of fields in number theory are the ​​cyclotomic fields​​, $\mathbb{Q}(\zeta_n)$, generated by the $n$-th roots of unity ($\zeta_n = \exp(2\pi i/n)$). These fields are the bedrock of much of modern algebra and number theory.

And where do our simple quadratic extensions fit into this grand picture? They are found nestled inside. For example, the cyclotomic field $\mathbb{Q}(\zeta_9)$ is a degree-6 extension of $\mathbb{Q}$. A careful look reveals that it contains exactly one quadratic subfield: $\mathbb{Q}(\sqrt{-3})$. This is no coincidence. The monumental ​​Kronecker-Weber theorem​​ states that any finite abelian extension of $\mathbb{Q}$ (an extension whose Galois group is commutative) must be a subfield of some cyclotomic field. Since all quadratic extensions are abelian (their Galois group has only two elements), this theorem tells us that every single quadratic field $\mathbb{Q}(\sqrt{d})$ lives inside a field generated by roots of unity. They form the simplest, foundational layer in the vast hierarchy of abelian number fields.

The 20th century brought a revolutionary new way of thinking about numbers, spearheaded by the ​​local-global principle​​. The idea is to understand a problem over a "global" field like $\mathbb{Q}$ by first analyzing it over all its "local" completions: the real numbers $\mathbb{R}$ and the $p$-adic numbers $\mathbb{Q}_p$ for every prime $p$.

Quadratic extensions provide the perfect stage to see this principle in action. Suppose we want to solve a Diophantine equation like $a^2 - 5b^2 = -4$ in rational numbers. This is equivalent to asking if the number $-4$ is the "norm" of some element in the field $\mathbb{Q}(\sqrt{5})$. The ​​Hasse Norm Theorem​​ gives us a powerful strategy. Instead of tackling the difficult global problem head-on, we check if $-4$ is a local norm everywhere. Is it a norm in $\mathbb{Q}_2(\sqrt{5})$? Yes. Is it a norm in $\mathbb{Q}_5(\sqrt{5})$? Yes. In $\mathbb{R}$? Yes. And so on for all primes. For a quadratic (and more generally, any cyclic) extension, if the answer is "yes" at every single local place, the theorem guarantees that a global solution must exist. This principle converts a single, often intractable, problem into an infinite series of simpler, mechanical checks.

This local-global viewpoint leads to one of the crowning achievements of number theory: ​​class field theory​​. This theory reveals a breathtakingly deep connection between the internal arithmetic of a number field $K$ and the kinds of abelian extensions it can have. For an imaginary quadratic field like $K = \mathbb{Q}(\sqrt{-5})$, its integers do not have unique factorization. The extent of this failure is measured by its "ideal class group," which for $K$ has size 2. Class field theory predicts that this number, 2, means that $K$ must admit a unique "unramified" quadratic extension—an extension that is arithmetically "nice" in every local sense. And what is this prophesied field? It is $L = K(\sqrt{-1}) = K(\sqrt{5})$. The internal structure of one quadratic field dictates the existence and nature of another.

This entire tapestry of connections finds its most elegant expression in a perfect dictionary between quadratic fields and certain functions called ​​primitive quadratic Dirichlet characters​​. Every quadratic extension $K/\mathbb{Q}$ corresponds to exactly one such character $\chi$. This character is a complete blueprint for the field's arithmetic. The "conductor" of the character, an integer $N$, tells you precisely which primes "ramify" (behave in a complicated way) in the field $K$. The value of the character at any other prime $p$, either $\chi(p)=1$ or $\chi(p)=-1$, tells you exactly how that prime will factor in $K$. All the complex behavior of primes in quadratic extensions is perfectly encoded in these remarkably simple functions.

From the geometry of lines and circles to the symphony of class field theory, the concept of a quadratic extension proves itself to be an idea of enduring power and beauty. It is the first step beyond the rational world, and in taking it, we discover that it forms the structural backbone of a vast, interconnected mathematical universe.