Right Cosets

In the study of abstract algebra, groups offer a powerful framework for understanding symmetry and structure. While subgroups represent self-contained worlds within a larger group, a fundamental question remains: how does their presence impose order on the group as a whole? This knowledge gap is bridged by the elegant concept of cosets, which allow us to dissect a group and reveal its large-scale architecture. This article explores this important concept, starting with its core principles and mechanisms. You will learn how cosets are constructed, the "all or nothing" rule that governs their interactions, and the critical distinction between left and right cosets that gives rise to the special status of normal subgroups. From there, we will explore the applications and interdisciplinary connections of cosets, demonstrating their utility in fields ranging from chemistry and quantum physics to number theory. Our journey begins now, by uncovering the fundamental mechanisms behind this powerful algebraic tool.

In our journey into the world of groups, we've become familiar with the idea of a ​​subgroup​​—a smaller, self-contained universe living inside a larger one. A subgroup $H$ is a bit like a club within a larger society $G$; its members can interact amongst themselves (using the group's operation) and they'll never produce an outcome that isn't also a member of the club. But what about the relationship between the club members and the outsiders? How does the existence of this special substructure $H$ influence the larger structure of $G$?

This is where the wonderfully elegant concept of a ​​coset​​ comes into play. It’s a beautifully simple idea that will allow us to slice up a group into perfectly uniform pieces, revealing a hidden, large-scale architecture that was invisible before.

Let’s start with a simple thought experiment. Imagine you have your subgroup $H$. Now, pick any element $g$ from the larger group $G$ that might not be in $H$. What happens if we take this element $g$ and multiply it by every single element in $H$? Let's say we multiply from the right. We aren’t just creating one new element; we are creating a whole new set of elements. We call this new set the ​​right coset​​ of $H$ determined by $g$, and we write it as $Hg$. Formally, it's the set $Hg = \{hg \mid h \in H\}$.

Let’s make this concrete. Consider the group $G$ of all polynomials with real coefficients, where the "multiplication" is actually just standard addition. Let our subgroup $H$ be the set of all polynomials that look like $ax^4 + c$. Now, let's pick an element not in $H$, say $g(x) = 2x^2 + 7$. The right coset $H + g(x)$ (we use + because that's our group operation) is the set of all polynomials you get by adding an element of $H$ to $g(x)$. Any such polynomial will look like $(ax^4 + c) + (2x^2 + 7) = ax^4 + 2x^2 + (c+7)$. Since $c$ can be any real number, $d = c+7$ can also be any real number. So the coset is the set of all polynomials of the form $ax^4 + 2x^2 + d$. We've essentially taken the entire "family" of polynomials in $H$ and shifted it, creating a new, parallel family.

This shifting idea works even in more complicated, non-commutative settings. Consider the dihedral group $D_4$, the group of symmetries of a square, with its rotations $r$ and reflections $s$. Let's take the very simple subgroup $H = \{e, s\}$, containing the identity and one reflection. If we want to find the right coset $Hr^2$, we just multiply the elements of $H$ by $r^2$ on the right: $er^2 = r^2$ and $sr^2$. The resulting coset is the two-element set $\{r^2, sr^2\}$. Again, we've taken the subgroup $H$ and "translated" it within the group by $r^2$.

So we can create these new sets, these cosets. Let’s make another one. If we have $Ha$ and we make a new coset $Hb$, what can we say about them? You might guess that they could overlap a little, sharing some elements but not all. But here, the universe of groups presents us with a stunningly rigid piece of architecture. For any two right cosets $Ha$ and $Hb$ of a subgroup $H$, they are either ​​perfectly identical​​ or ​​completely disjoint​​—they have no elements in common. There is no middle ground.

This "all or nothing" principle is one of the most fundamental properties of cosets. It means that the collection of all distinct right cosets of a subgroup $H$ forms a ​​partition​​ of the entire group $G$. They slice $G$ up perfectly, like a mosaic, with no gaps and no overlapping tiles. Every element of $G$ lives in exactly one coset.

When are two cosets $Ha$ and $Hb$ identical? The test is beautifully simple: $Ha = Hb$ if and only if the element $ab^{-1}$ is a member of the original subgroup $H$. Think of it as a secret password. To see if $a$ and $b$ generate the same "shifted copy" of $H$, you see if the combination $ab^{-1}$ "gets you back into" the original club $H$. If it doesn't, then $Ha$ and $Hb$ are total strangers, with no common members at all.

For example, in $D_4$ with the subgroup $H = \{e, s\}$, let's look at the cosets $C_1 = Hr$ and $C_2 = Hr^2$. Are they the same or different? We check the element $ab^{-1}$, which is $r(r^2)^{-1} = r \cdot r^2 = r^3$. Is $r^3$ in $H=\{e, s\}$? No. Therefore, the cosets $Hr$ and $Hr^2$ are completely disjoint. The size of their union is simply the sum of their individual sizes, which is $2+2=4$.

So far, we've been very "right-handed," always multiplying by our element $g$ on the right side ($Hg$). What if we were left-handed? We could just as easily define a ​​left coset​​, $gH = \{gh \mid h \in H\}$.

In some nice, orderly groups like our polynomial example where the operation is commutative, the distinction is pointless: $h+g$ is always the same as $g+h$. But in a non-commutative group, like the symmetries of a triangle or square, the order of multiplication matters immensely. Multiplying from the left might be a very different operation from multiplying from the right.

Let's look at the group $D_3$ (symmetries of an equilateral triangle) with the subgroup $H=\{e, s\}$ and the element $g=r$. The left coset is $rH = \{r \cdot e, r \cdot s\} = \{r, rs\}$. The right coset is $Hr = \{e \cdot r, s \cdot r\} = \{r, sr\}$. In this group, the rules state that $sr = r^2s$, which is not the same as $rs$. So the left coset $\{r, rs\}$ and the right coset $\{r, sr\}$ are different sets! They share the element $r$, but their other elements, $rs$ and $sr$, are distinct.

This is a profound realization. A subgroup can partition the group in one way from the left, and a completely different way from the right. It’s as if the subgroup casts two different "shadows" on the larger group, depending on which side the light (our element $g$) is coming from. For some subgroups, these two shadow patterns are wildly different.

With left and right cosets creating potentially different partitions, things might seem messy. But beneath this apparent chaos lies another layer of beautiful, hidden symmetry. Even if the partitions are different, it turns out there is always the same number of left cosets as there are right cosets. This common number is called the ​​index​​ of the subgroup $H$ in $G$, written $[G:H]$.

How can we be sure? There is an elegant, almost magical, one-to-one correspondence between the set of left cosets and the set of right cosets. This correspondence is given by the map that sends a left coset $gH$ to the right coset $Hg^{-1}$. This map provides a perfect pairing; for every left coset, there is a unique corresponding right coset, and vice-versa. So, while the individual pieces of the two partitions might not match, the total number of pieces is always identical.

This naturally leads us to a crucial question: when do the two worlds, the left and the right, align? When are the left and right shadows identical for any choice of $g$? That is, for which subgroups $H$ is it true that for every single element $g$ in $G$, the left coset $gH$ is exactly the same set as the right coset $Hg$?

Such subgroups are incredibly special, and they are called ​​normal subgroups​​. They are "symmetrically" embedded within the larger group. For most subgroups in a non-abelian group, this property fails. But when it holds, it unlocks a whole new level of structure.

A wonderful example of this alignment comes from a simple counting argument. It turns out that any subgroup $H$ whose index $[G:H]$ is 2 is automatically a normal subgroup. The logic is inescapable. If the index is 2, there are only two left cosets: $H$ itself and "everything else," which we can write as $G \setminus H$. There are also only two right cosets: $H$ and $G \setminus H$. There's no other way to slice the group into two pieces. Therefore, the collection of left cosets must be identical to the collection of right cosets. For any $g$ not in $H$, the left coset $gH$ must be $G \setminus H$, and the right coset $Hg$ must also be $G \setminus H$. Thus, $gH=Hg$ for all $g$.

You might be asking, why all this fuss about whether left and right cosets are the same? It might seem like a technical curiosity, but it is, in fact, the key to one of the most powerful ideas in all of algebra: the construction of new groups from old ones.

The grand prize is this: if (and only if) a subgroup $H$ is normal, then the set of its cosets can itself be turned into a new, smaller group, called the ​​quotient group​​. The "elements" of this new group are the cosets themselves, and the group operation is defined by set multiplication.

Let's see what goes wrong when a subgroup is not normal. Consider the subgroup $H = \{e, s\}$ in $D_4$, which we've seen is not normal. Let's take the right coset $C = Hr = \{r, sr\}$ and try to "multiply" it by itself: $C^2 = C \cdot C$. We calculate all the possible products of an element from the first $C$ with an element from the second $C$. When we do this, we get the set $\{r^2, s, sr^2, e\}$. This new set has four elements. But a coset of $H$ must have the same size as $H$, which is two. Our result, $C^2$, is not a coset of $H$. The operation is not "closed." The system collapses; we cannot form a group out of these cosets.

However, if $H$ were normal, the product of any two cosets, $(Ha)(Hb)$, would miraculously simplify to become another single coset, $H(ab)$. The cosets would behave like respectable group elements, and their collection would form a new group that often reveals the essential structure of $G$ in a much simpler form.

And so, the seemingly minor distinction between multiplying on the left and multiplying on the right has blossomed into a profound structural principle. It is the gatekeeper that determines whether we can distill the essence of a large group into a smaller, more manageable one, a path that leads to some of the deepest and most beautiful results in modern mathematics.

Having explored the foundational machinery of right cosets—how they are constructed and the algebraic laws they obey—we might be tempted to file them away as a neat, but perhaps slightly sterile, piece of abstract mathematics. But that would be like learning the principles of lens grinding and never looking through the telescope. The real magic begins when we turn this new instrument upon the world, both the abstract world of mathematics and the physical world around us. Cosets are not merely an exercise in partitioning sets; they are a powerful lens for revealing hidden structure, classifying complexity, and forging surprising connections between seemingly disparate fields.

Before we venture into other disciplines, let's first point our lens inward, at the very groups from which cosets are born. Here, cosets act as a powerful organizational tool, sorting the chaotic jumble of group elements into beautifully ordered families.

Imagine the group of all invertible $2 \times 2$ matrices, $GL_2(\mathbb{R})$. This is a vast, infinite collection of transformations. Now, suppose we are interested in a particular property: the determinant. Let's say we want to understand all matrices whose determinant is exactly 5. This seems like a hopelessly complex set. And yet, coset theory tells us something astonishing: this entire, infinite family of matrices is nothing more than a single right coset of the subgroup of matrices with determinant 1, the special linear group $SL_2(\mathbb{R})$. Every matrix with determinant 5 can be obtained by taking some matrix with determinant 1 and multiplying it on the right by a single, fixed representative—for instance, by the simple diagonal matrix $\begin{pmatrix} 5 & 0 \\ 0 & 1 \end{pmatrix}$. The coset structure neatly packages all elements sharing a common "determinant property" into one object.

This is a general principle. Whenever a group property can be described by a homomorphism (like the determinant function), the elements that map to the same value form a single coset. Consider the permutations of $n$ objects, the symmetric group $S_n$. We know that every permutation is either "even" or "odd". The even permutations form a subgroup, the alternating group $A_n$. And what about the odd ones? They are not a subgroup, but they are not a random assortment either. The set of all odd permutations is, in fact, the one and only other coset of $A_n$ in $S_n$. The entire universe of permutations is cleanly and fundamentally split into two halves: the subgroup $A_n$ and its traveling companion, the coset of odd permutations.

This talk of "left" and "right" cosets may seem like pedantic hair-splitting. Does it really matter which side we multiply on? The answer is a resounding "yes," and the implications are profound. In some special, highly symmetric situations, the partition of a group into left cosets is identical to its partition into right cosets. That is, for any element $g$, the set $gH$ is identical to the set $Hg$. Subgroups $H$ with this property are called ​​normal​​, and they are the royalty of group theory.

Consider the symmetries of a regular decagon, the group $D_{10}$. It contains 10 rotations and 10 reflections. The 10 rotations form a subgroup $R$. If you pick a reflection, say $s$, and look at the left coset $sR$ and the right coset $Rs$, you find they are exactly the same set of 10 reflections. This "two-sidedness" means the structure of the rotations is respected by all other symmetries in the group. They form a self-contained, stable unit.

What's truly remarkable is that sometimes we can know a subgroup is normal without checking every coset. Here is a beautiful, simple piece of reasoning: if a subgroup $H$ makes up exactly half of the parent group $G$ (in technical terms, its index is 2), it must be normal. Why? A group $G$ is partitioned by the left cosets of $H$. If there are only two, they must be $H$ itself and "everything else," the set $G \setminus H$. The same is true for the right cosets: they must be $H$ and $G \setminus H$. So for any element $g$ that isn't in $H$, the left coset $gH$ and the right coset $Hg$ must both be equal to that same "everything else" compartment. They must be equal to each other!. This elegant argument instantly explains why the subgroup of even permutations $A_n$ (which is half of $S_n$) and the subgroup of rotations in $D_{10}$ (which is half of $D_{10}$) are normal.

But what happens when this symmetry breaks? When left and right are not the same, we often find interesting phenomena. Consider a non-abelian group $G$ endowed with a "right-invariant" measure $\mu$, a way of assigning "size" or "volume" to subsets that doesn't change when we shift a set to the right. If we know a subgroup $H$ is a "measurable" set, then any of its right cosets $Hg$ are guaranteed to be measurable and to have the same "size" as $H$. However, the same guarantee does not apply to its left cosets $gH$. While they are also measurable, their "size" or "volume" may be different from that of $H$. This distinction, born from pure algebra, has real consequences in the analytical theory of groups, forming a cornerstone of the theory of Haar measure, which is essential for doing calculus and probability on continuous groups. The difference between left and right is not a trifle; it's a fundamental feature of the group's geometric landscape.

The utility of cosets extends far beyond the borders of group theory. They appear as essential tools in chemistry, physics, number theory, and computer science, providing a language to describe complex systems.

​​Chemistry and Physics:​​ In the world of quantum mechanics and physical chemistry, symmetry is king. The set of symmetries of a molecule (like rotations and reflections that leave it looking the same) forms a mathematical group, a "point group." These groups dictate the molecule's properties: its vibrational modes, how it absorbs light, and the shapes of its electron orbitals. Cosets provide a way to decompose these complex symmetry groups. For instance, in analyzing the tetrahedral methane molecule, one might decompose its rotational symmetry group $T$ by one of its subgroups of 3-fold rotations. Properties relevant to spectroscopy are often constant across the elements of a single coset. By summing these properties over cosets, chemists and physicists can simplify enormously complex quantum calculations, breaking a large problem into smaller, manageable blocks defined by the coset structure.

​​Quantum Information:​​ The rise of quantum computing has opened up new applications for classical algebraic structures. Imagine a quantum system whose states are indexed by the elements of a group $G$. Suppose we are only interested in a part of this system, described by a subgroup $H$. A "quantum channel" can be constructed to project information from the large system $G$ onto the smaller system $H$. One natural way to build such a channel is through a right coset decomposition. The cosets partition the full state space, and the channel's action can be thought of as identifying which coset a state is in, and then determining its "local position" within that coset. The coset structure provides a blueprint for how information is processed or controllably lost, a key concept in building robust quantum algorithms and error-correcting codes.

​​Number Theory:​​ Perhaps most surprising is the role cosets play in the abstract and ancient field of number theory. Modern number theory is deeply intertwined with the study of "modular forms," which are highly symmetric functions on the complex plane. To understand these functions, mathematicians study "Hecke operators," which act on the space of modular forms. And what are these mysterious operators? At their heart, they are built from cosets. Specifically, the fundamental Hecke operator $T_p$ is defined by a double coset of a matrix group, which is then decomposed into a collection of $p+1$ distinct right cosets. The structure and number of these cosets dictate the operator's properties, which in turn reveal profound truths about prime numbers and Diophantine equations. The solution to Fermat's Last Theorem, for instance, relied on this deep and intricate theory. It is a stunning realization that the simple act of partitioning a group, which we learn in an introductory algebra course, is a crucial gear in the mechanism that solved one of the oldest problems in mathematics.

From classifying matrices to understanding molecular vibrations, from building quantum channels to unlocking the secrets of prime numbers, the concept of a coset proves itself to be far more than an abstract curiosity. It is a unifying thread, a testament to the power of a simple mathematical idea to impose order on complexity and to reveal the hidden consonances that link the diverse fields of scientific inquiry.