Image of a Homomorphism

In the vast landscape of abstract algebra, we often seek to understand complex structures by comparing them. A homomorphism is a map that connects two such structures while respecting their inherent rules of operation. But what does the result of this mapping—the "shadow" it casts—look like? This article addresses the fundamental question of what information is preserved, and what is lost, in this projection. We will explore the concept of the ​​image of a homomorphism​​, a seemingly simple idea that holds the key to understanding deep structural truths. The following chapters will guide you through this concept, first by establishing its foundational principles and mechanisms, such as why the image is always a subgroup and its elegant relationship with the kernel. Subsequently, we will journey through its diverse applications, uncovering how this single algebraic idea provides a unifying lens to explore secrets in fields ranging from the geometric world of topology to the discrete realm of number theory.

Imagine you are standing in a large, intricate clock tower, filled with gears of all shapes and sizes, all turning and interacting in a complex, three-dimensional dance. Now, imagine a single light source casts a shadow of this entire mechanism onto a flat wall. This shadow is a two-dimensional projection of the three-dimensional reality. It doesn't capture everything—you've lost the depth, the layering, the way gears mesh from front to back. Yet, the shadow is not just a random blob. It moves. It has a structure of its own. It preserves certain truths about the original machine: the relative speeds of some gears, the overall rhythm, the shape of the outer components. This shadow is a beautiful analogy for the mathematical concept of an ​​image of a homomorphism​​.

A homomorphism is a map, a function, that connects two algebraic structures (like groups or rings) in a way that respects their inherent operations. The ​​image​​ of this map is simply the set of all the points in the destination structure that are "hit" by the map. It's the "shadow" cast by the source structure onto the codomain.

Let's start with the most familiar structure: the integers, $(\mathbb{Z}, +)$. Suppose we define a homomorphism $\phi$ from the integers to the integers by the simple rule $\phi(n) = 6n$. What does the image, the set of all possible outputs, look like? It's the set of all integers that are multiples of 6—..., -12, -6, 0, 6, 12, ...—which we denote as $6\mathbb{Z}$. Our original group was the entire number line of integers, and its "shadow" under this map is a sparser, more spread-out version, yet still an infinite, orderly lattice.

This idea isn't confined to numbers. Consider the world of polynomials, which can be added and multiplied. Let's define a map $f$ that takes any polynomial $p(x)$ and maps it to a new polynomial $(x^2 - 9)p(x)$. The image of this map is the set of all polynomials that are multiples of $(x^2 - 9)$. A polynomial is a multiple of $(x^2 - 9) = (x-3)(x+3)$ if and only if it evaluates to zero at both $x=3$ and $x=-3$. So, the image is precisely the set of all polynomials that have roots at $3$ and $-3$. Here, the "shadow" isn't defined by a simple arithmetic pattern, but by a shared geometric property—passing through specific points on a graph. The image captures a fundamental structural characteristic.

Here is where something remarkable happens. The image isn't just a random collection of elements. The shadow cast by a group is itself a group. This is not an accident; it's a direct consequence of the homomorphism's "structure-preserving" nature.

Why must this be true? Let's take any two elements, say $y_1$ and $y_2$, from the image of a homomorphism $\phi: G \to H$. Because they are in the image, they must be the "shadow" of some elements from the original group $G$. So, there exist $x_1$ and $x_2$ in $G$ such that $\phi(x_1) = y_1$ and $\phi(x_2) = y_2$.

What happens if we try to combine $y_1$ and $y_2$ using the operation in $H$? Let's say we compute $y_1$ times the inverse of $y_2$. Because $\phi$ is a homomorphism, it respects the group operations, including inverses. This means $\phi(x_2^{-1}) = (\phi(x_2))^{-1} = y_2^{-1}$. So, we get:

$y_1 y_2^{-1} = \phi(x_1) \phi(x_2^{-1})$

And now for the magic trick: because $\phi$ preserves the operation, we can combine the elements before mapping them:

$\phi(x_1) \phi(x_2^{-1}) = \phi(x_1 x_2^{-1})$

Since $G$ is a group, $x_1 x_2^{-1}$ is guaranteed to be some element within $G$. Therefore, its image, $\phi(x_1 x_2^{-1})$, must be in the image of $\phi$. We have just shown that for any $y_1, y_2$ in the image, the combination $y_1 y_2^{-1}$ is also in the image. This is a classic test (the subgroup criterion) which proves that the image, $\text{Im}(\phi)$, is a ​​subgroup​​ of the codomain $H$. The shadow has a life of its own, obeying the same rules as the world it was projected from.

Now, what if the original group is very simple? What if it's a ​​cyclic group​​, one that can be generated by repeatedly applying the group operation to a single element? For instance, the integers $\mathbb{Z}$ are generated by $1$ (by adding it to itself) and the integers modulo $n$, $\mathbb{Z}_n$, are generated by $[1]_n$.

It turns out that the homomorphic image of a cyclic group is always cyclic. Even better, the entire image is generated by the image of the original generator. If $G = \langle g \rangle$, then its image is simply $\text{Im}(\phi) = \langle \phi(g) \rangle$. This is an incredibly powerful idea. To understand the entire shadow, we only need to see where one single point lands!

Let's see this in action. Consider a map from the integers $\mathbb{Z}$ to the integers modulo 18, $\mathbb{Z}_{18}$, defined by where the generator $1 \in \mathbb{Z}$ lands: $\phi(1) = 6$. The entire image is now determined. It's the cyclic subgroup generated by $6$ in $\mathbb{Z}_{18}$. What are the multiples of $6$ modulo $18$? We have $6 \cdot 1 = 6$, $6 \cdot 2 = 12$, $6 \cdot 3 = 18 \equiv 0$, and then the pattern repeats. So, the image is the set $\{0, 6, 12\}$. The infinite group of integers has cast a tiny, three-element shadow.

This principle holds even when the destination is more exotic. Take a map from the cyclic group $\mathbb{Z}_{12}$ to the group of symmetries of a hexagon, $D_6$. Let's say the map is defined by sending the generator $1 \in \mathbb{Z}_{12}$ to a $120^\circ$ rotation, $r^2 \in D_6$. The image will be the cyclic subgroup generated by $r^2$. The powers of $r^2$ are $(r^2)^1 = r^2$, $(r^2)^2 = r^4$, and $(r^2)^3 = r^6 = e$ (the identity). So, the image is $\{e, r^2, r^4\}$. Again, a 12-element group casts a 3-element shadow, and we knew its complete structure just by looking at the image of the generator. This leads to a profound general statement: the possible homomorphic images of a cyclic group like $\mathbb{Z}_{12}$ are, up to isomorphism, precisely the cyclic groups whose order divides 12.

We've seen that the image is a structured shadow of the original group. But how much information is lost in this projection? The information lost is captured by another fundamental concept: the ​​kernel​​ of the homomorphism, written $\ker(\phi)$. The kernel is the set of all elements in the source group that are "squashed" down to the identity element in the target group. In our projector analogy, the kernel is everything on the film that completely blocks the light.

The relationship between what's projected (the image) and what's lost (the kernel) is not arbitrary. It is governed by one of the most elegant and powerful theorems in all of algebra: the ​​First Isomorphism Theorem​​. It states that the image of a homomorphism is structurally identical (isomorphic) to the source group divided by its kernel. In symbols:

$\text{Im}(\phi) \cong G / \ker(\phi)$

This theorem is a cosmic balancing act. It says that the size and structure of the shadow are perfectly determined by the size and structure of the original object and the part that was blocked. The size of the quotient group $G / \ker(\phi)$ is $|G| / |\ker(\phi)|$, so the theorem implies a beautiful formula relating the orders of these finite groups: $|\text{Im}(\phi)| = |G| / |\ker(\phi)|$.

Let's check this with a concrete case. Consider a homomorphism $\phi: \mathbb{Z}_{20} \to \mathbb{Z}_{15}$ defined by $\phi([x]_{20}) = [6x]_{15}$. The kernel consists of elements $[x]_{20}$ such that $6x$ is a multiple of $15$. A little arithmetic shows this happens when $x$ is a multiple of $5$. So, the kernel is $K = \{[0], [5], [10], [15]\}$ in $\mathbb{Z}_{20}$, and its size is $|K|=4$. The image is the subgroup of $\mathbb{Z}_{15}$ generated by $[6]_{15}$, which is $I = \{[0], [6], [12], [3], [9]\}$, with size $|I|=5$. Does the theorem hold? Yes! $|I| = 5$, and $|G|/|K| = 20/4 = 5$. The law is satisfied. The structure of the 20-element source group, after collapsing the 4-element kernel, perfectly matches the 5-element image.

This law is so powerful it allows us to make predictions. For any homomorphism from $\mathbb{Z}_{48}$ to $\mathbb{Z}_{60}$, the order of the image must be a common divisor of 48 and 60, and thus a divisor of $\gcd(48, 60)=12$. The structure of the source and target groups places rigid constraints on what kind of shadows are possible.

The final question is, what is this all good for? We care about homomorphic images because they inherit some, but not all, of the essential properties of their parent group. By studying what is preserved and what is lost, we gain a deeper understanding of the properties themselves.

Some deep structural properties are always passed down to a homomorphic image. For example, if a group is ​​solvable​​ (meaning it can be broken down into a series of abelian groups), then any of its homomorphic images must also be solvable. This is a profound "hereditary" trait.

However, many other properties are lost in the projection.

• A ​​non-abelian​​ group (where order of operation matters, like the symmetry group $S_3$) can have an abelian image (like $\mathbb{Z}_2$). It's like a complex 3D machine casting a simple 1D shadow.
• The ​​order​​ of the group is usually not preserved; as we've seen, the image is often smaller.
• Even more subtle properties can disappear. The ring of integers $\mathbb{Z}$ is a "Principal Ideal Domain" (PID), a very well-behaved ring. Yet its homomorphic image $\mathbb{Z}_6$ (under the map $a \mapsto a \pmod 6$) is not even an "integral domain" because it has zero divisors ($2 \cdot 3 = 0$), a property that PIDs are forbidden from having. Niceness can be lost in projection.

Perhaps the most beautiful illustration of this principle comes from studying homomorphisms into abelian groups. If we take a highly non-abelian group like the symmetric group $S_5$ (the 120 symmetries of five objects) and map it to any abelian group, something amazing happens. The image can only have order 1 or 2. Why? Because forcing the image to be abelian means we must collapse the entire "non-abelian character" of $S_5$ into the kernel. This non-abelian core of $S_5$ (its commutator subgroup, $A_5$) is huge, containing 60 elements. By the First Isomorphism Theorem, the image can have size at most $|S_5|/|A_5| = 120/60 = 2$. Projecting the intricate structure of $S_5$ onto an "abelian screen" flattens almost all of its complexity, leaving behind at most a simple two-point shadow.

In the end, the image of a homomorphism is more than just a subset. It is a faithful, if simplified, reflection of its source, governed by profound and elegant laws that connect the projected with the lost. By studying these shadows, we learn not only about the object casting them, but about the very nature of light and projection—the fundamental principles of structure itself.

We have spent some time understanding the machinery of homomorphisms and their images. On the surface, it might seem like a rather abstract affair—a game of mapping one algebraic structure to another and seeing what lands where. But this is where the magic truly begins. To a physicist, a mathematician, or a computer scientist, the image of a homomorphism is not just a resulting set; it is a shadow, a projection, a compressed summary of a complex object. And by studying this shadow, we can often deduce profound truths about the object casting it—sometimes, truths that are difficult to see when looking at the object in its full, bewildering complexity.

Let us embark on a journey through different scientific landscapes to see how this one idea—the homomorphic image—appears again and again, a golden thread weaving together seemingly disparate fields.

Our first stop is the fascinating world of topology, the study of shapes and spaces. Imagine drawing a loop on the surface of a donut. You can draw a loop that goes around the hole, one that goes through the hole, or one that can be shrunk to a point. The collection of all these distinct types of loops, with a clever way of "multiplying" them, forms a group called the fundamental group, $\pi_1$. It’s an algebraic fingerprint of the space's "holey-ness."

Now, what happens when we map one space into another? A continuous map, say from a circle $S^1$ into some other space $Y$, induces a homomorphism between their fundamental groups. The fundamental group of the circle, $\pi_1(S^1)$, is simply the group of integers, $\mathbb{Z}$, where each integer corresponds to how many times a loop winds around the circle. What can we say about the image of this homomorphism? A beautiful theorem from group theory tells us that any homomorphic image of a cyclic group must itself be cyclic. Since $\mathbb{Z}$ is the quintessential cyclic group, the image of any loop from a circle into another space must form a cyclic subgroup within that space's fundamental group. The circle acts as a fundamental "probe," and the image of its homomorphism tells us what kind of simple, repeating path it has traced in the target space.

We can see this principle at work with wonderful clarity. Consider the torus, or donut surface, $T^2 = S^1 \times S^1$. Its fundamental group is $\mathbb{Z} \oplus \mathbb{Z}$, representing the two independent ways you can loop around it. If we create a map that simply projects the entire torus onto one of its constituent circles, $f: T^2 \to S^1$, this corresponds to a homomorphism $f_*: \mathbb{Z} \oplus \mathbb{Z} \to \mathbb{Z}$. What is the image? It is simply $\mathbb{Z}$. The homomorphism has algebraically "forgotten" one of the directions, and its image is the structure that remains.

This connection culminates in one of algebraic topology's most powerful tools: the lifting criterion. Imagine you have a map from a space $Y$ to a base space $B$, and you know that $B$ is being "covered" by another space $E$ (like a parking garage with multiple levels covering the same ground plan). The question is: can you "lift" your original map from $Y$ into the covering space $E$ in a consistent way? This geometric puzzle seems horribly complicated. Yet, the answer is breathtakingly simple and purely algebraic. A lift exists if and only if the image of the homomorphism induced by your map, $f_*(\pi_1(Y))$, is a subgroup of the image of the homomorphism induced by the covering map, $p_*(\pi_1(E))$. The entire geometric problem dissolves into a simple check of subgroup inclusion between two algebraic shadows.

Leaving the world of shapes, we can turn this tool inward to dissect algebraic structures themselves. The image of a homomorphism acts like a lens, simplifying a complex group or ring to reveal its essential properties.

Consider the unitary group $U(n)$, the set of $n \times n$ complex matrices that are fundamental to quantum mechanics. These are vast, continuous objects. But we can define a homomorphism from $U(n)$ to the group of non-zero complex numbers, $\mathbb{C}^*$, using a familiar function: the determinant. What is the image of this map? It’s not all of $\mathbb{C}^*$. A key property of unitary matrices is that the absolute value of their determinant is always $1$. The image is precisely $U(1)$, the group of complex numbers on the unit circle. The determinant homomorphism projects the entire, high-dimensional structure of $U(n)$ onto this simple circle, capturing a crucial invariant—a conserved quantity, if you will—that all its elements share.

Sometimes, the image tells a story through its sheer simplicity. The alternating group $A_5$, the group of even permutations of five objects, is famous for being "simple." This has a technical meaning: it has no non-trivial normal subgroups. What happens if we try to map $A_5$ to an abelian (commutative) group, the simplest kind of group there is? The result is startling: the image is always the trivial group, containing just the identity element. Any attempt to project the intricate, non-commutative structure of $A_5$ onto a commutative world causes it to collapse into a single point. This tells us that $A_5$ is fundamentally and irreducibly complex; it cannot be "approximated" by any simpler, commutative structure. This very fact lies at the heart of why there is no general formula for the roots of a fifth-degree polynomial, a puzzle that tormented mathematicians for centuries.

The study of numbers is another fertile ground for our concept. Let's start with the integers, $\mathbb{Z}$. A classic result, the Chinese Remainder Theorem, can be elegantly rephrased in the language of homomorphisms. Consider the map from the integers to pairs of integers modulo $m$ and $n$, defined by $\phi(x) = (x \pmod m, x \pmod n)$. The image of this homomorphism consists of all the pairs $(a, b)$ that can be formed by some integer $x$. Using the First Isomorphism Theorem, we find this image is isomorphic to the ring $\mathbb{Z}_{\text{lcm}(m, n)}$. This tells us precisely which combinations are possible and reveals that the structure of simultaneous congruences is governed by the least common multiple—a fact essential for algorithms in cryptography and computer science.

What about more exotic numbers? The number $e$ is transcendental, meaning it is not the root of any non-zero polynomial with rational coefficients. Let's define an "evaluation homomorphism" that takes a polynomial from $\mathbb{Q}[x]$ and plugs in $e$, mapping it to a real number. What does the set of all such resulting numbers—the image—look like? Since no non-zero polynomial becomes zero when we plug in $e$, the kernel of this map is trivial. This means the homomorphism is an isomorphism onto its image. The set of numbers you can make from $e$ with rational coefficients, $\mathbb{Q}[e]$, has a structure identical to the ring of polynomials $\mathbb{Q}[x]$ itself. The image reveals that the world built upon a transcendental number is just as rich and complex as the world of abstract polynomials.

The image can also expose subtle differences between number systems. In algebraic number theory, we study fields like $\mathbb{Q}(\sqrt{d})$. The "units" in the ring of integers of such a field form a multiplicative group. The field norm, which for an element $a+b\sqrt{d}$ is $a^2 - db^2$, provides a homomorphism from this group of units to the simple group $\{-1, 1\}$. For the field $\mathbb{Q}(\sqrt{3})$, the norm of any unit is always $1$. The image of the norm homomorphism is just the trivial group $\{1\}$. But for the field $\mathbb{Q}(\sqrt{5})$, we can find a unit whose norm is $-1$, such as the golden ratio conjugate $\frac{1-\sqrt{5}}{2}$ has norm $-1$. So the image is the full group $\{-1, 1\}$. The image of this simple homomorphism acts as a litmus test, revealing a deep structural difference in the arithmetic of these two fields that is far from obvious at first glance.

The power of the homomorphic image extends even further, into the higher realms of abstract algebra and the very foundations of computer science. In the theory of modules (a generalization of vector spaces), certain abelian groups are called "divisible"—for any element $x$ and any integer $n$, you can always find a $y$ such that $ny=x$. The group of rational numbers, $\mathbb{Q}$, is a prime example. It turns out that this property of divisibility is preserved by homomorphisms. The homomorphic image of any divisible group is always another divisible group. So, the structure of "being infinitely divisible" is something that the shadow faithfully retains from the original object.

Perhaps the most surprising application comes from theoretical computer science. In interactive proofs, a powerful "Prover" (Merlin) tries to convince a skeptical "Verifier" (Arthur) of a mathematical truth. Consider the Graph Non-Isomorphism problem: proving two graphs $G_0$ and $G_1$ are not the same. A standard protocol involves Arthur taking one of the graphs at random, scrambling its vertices (creating an isomorphic copy), and sending it to Merlin, who must guess which one it came from. If the graphs are truly isomorphic, Merlin learns nothing and can only guess with $0.5$ probability.

Now, consider a flawed hypothetical protocol where Arthur, instead of sending an isomorphic copy, sends a random homomorphic image of the graph (e.g., by contracting sets of vertices). One might think this is just as secure. It is not. The protocol fails catastrophically because, even if $G_0$ and $G_1$ are isomorphic, the set of possible homomorphic images you can get from each can be structurally different. Applying the same partitioning rule to two differently labeled (but isomorphic) graphs can produce non-isomorphic images. This means the "shadow" cast by the homomorphism leaks information about the original graph's labeling. A clever Merlin can analyze the structure of the homomorphic image he receives and determine, with probability greater than $0.5$, which graph it originated from. This beautiful failure teaches us a profound lesson: being structurally identical (isomorphic) does not mean your projections onto a simpler world will be indistinguishable.

From the loops on a donut to the security of cryptographic protocols, the image of a homomorphism is a concept of extraordinary power and reach. It allows us to simplify, to probe, to classify, and to understand. By studying the shadows, we learn about the light, and about the beautiful, intricate objects that stand in its way.