Induced Homomorphism

In the fascinating field of algebraic topology, mathematicians seek to understand the fundamental properties of shapes by translating them into the language of algebra. This approach allows for rigorous analysis of concepts that are often hard to grasp through geometric intuition alone. The central challenge, however, is building a reliable dictionary for this translation. How can we be sure that the algebraic representation, or "shadow," accurately reflects the original topological space and the transformations it undergoes? This is the gap filled by the concept of the induced homomorphism.

This article explores the induced homomorphism, a foundational tool that acts as a bridge between the world of continuous spaces and the world of discrete algebraic structures. Over the following sections, you will gain a deep understanding of this elegant mechanism.

• ​​Principles and Mechanisms​​ will introduce the core idea of the induced homomorphism using the fundamental group as our primary example. We will define it, explore its crucial rules of functoriality, and uncover its immediate and profound consequences for classifying spaces.
• ​​Applications and Interdisciplinary Connections​​ will showcase the concept in action. We will see how it provides an "algebraic veto" to prove topological impossibilities and how its influence extends far beyond topology, appearing in abstract algebra, category theory, and even modern physics.

By the end, you will see how this single, powerful idea allows us to study the shadow to understand the substance, revealing hidden structures and connections across diverse scientific landscapes.

Imagine you are trying to understand a complex, perhaps even four-dimensional, object. You can't hold it, you can't see it all at once, but you can study its shadow. This shadow might be a simpler, two-dimensional projection, but it captures essential features of the original object. If you rotate the object, the shadow changes in a predictable way. Algebraic topology operates on a similar, beautiful principle. It's a craft of creating "algebraic shadows"—like groups—from complex topological spaces, and the induced homomorphism is the mechanism that tells us precisely how these shadows transform when we manipulate the original spaces.

So, what is this magical shadow projector? Let's get a feel for it. Suppose you have two topological spaces, let's call them $X$ and $Y$. Think of $X$ as a rubber sheet in the shape of a donut (a torus, $T^2$) and $Y$ as a simple sphere ($S^2$). Now, imagine you have a continuous map $f: X \to Y$, which is just a rule for stretching, squishing, or wrapping the donut onto the surface of the sphere without tearing it.

The "algebraic shadow" we'll use is the ​​fundamental group​​, denoted $\pi_1(X, x_0)$. In essence, this group catalogs all the different kinds of loops you can draw on the space $X$ starting and ending at a base point $x_0$. Two loops are considered the "same" if you can continuously deform one into the other without breaking the loop or lifting it off the surface.

The induced homomorphism, written as $f_*$, is the bridge connecting the algebra of $X$ to the algebra of $Y$. It’s an astonishingly simple and natural idea. If you have a loop $\gamma$ in $X$, the map $f$ takes every point on that loop and places it somewhere in $Y$. The result is a new loop, $f \circ \gamma$, in $Y$. The induced homomorphism $f_*$ is simply the function that takes the class of the loop $\gamma$ in $\pi_1(X, x_0)$ and gives you the class of the new loop $f \circ \gamma$ in $\pi_1(Y, y_0)$.

What happens if our map $f$ is ridiculously simple? Suppose $f$ is a ​​constant map​​: it takes every single point in our space $X$ and sends it to one single point, $y_0$, in $Y$. Any loop in $X$, no matter how wild and complicated, gets squashed down to the single point $y_0$ when we apply $f$. The resulting "loop" in $Y$ is just the loop that stays put at $y_0$. In any fundamental group, the class of a constant loop is the identity element—the "do nothing" element. This means a constant map induces the ​​trivial homomorphism​​, a map that sends every loop class from the original space to the identity element in the target space.

This process of inducing homomorphisms isn't just a neat trick; it follows a rigid and beautiful set of rules. This set of rules is so important it has a fancy name: ​​functoriality​​. It's the logical backbone that makes this whole enterprise so powerful. There are just two rules you need to know.

​​Rule 1: The Identity Principle​​

What is the simplest possible continuous map? The identity map, $\text{id}_X$, which takes every point in a space $X$ and maps it to itself. It changes nothing. What should the induced homomorphism $(\text{id}_X)_*$ do? If our shadow analogy holds, it should also change nothing. And it does! If you take a loop $\gamma$ and apply the identity map to it, you just get the same loop $\gamma$ back. Therefore, the induced homomorphism $(\text{id}_X)_*$ is the identity homomorphism on the fundamental group. It maps every loop class to itself. Doing nothing to the space does nothing to its algebraic shadow.

​​Rule 2: The Composition Principle​​

Now for the master rule. Suppose you have three spaces, $X$, $Y$, and $Z$, and two maps: $f: X \to Y$ and $g: Y \to Z$. You can create a composite map, $g \circ f$, by first applying $f$ and then applying $g$. This takes you directly from $X$ to $Z$.

Each of these maps creates an induced homomorphism:

• $f_*: \pi_1(X) \to \pi_1(Y)$
• $g_*: \pi_1(Y) \to \pi_1(Z)$
• $(g \circ f)_*: \pi_1(X) \to \pi_1(Z)$

The composition principle states that the algebraic shadows behave exactly as you'd hope: the homomorphism induced by the composite map is the same as the composition of the individual induced homomorphisms. In symbols, this is the elegant and powerful statement: $(g \circ f)_* = g_* \circ f_*$ Notice the order is preserved. First applying $f$ then $g$ in the world of spaces corresponds to first applying $f_*$ then $g_*$ in the world of groups. This rule is the key that unlocks almost everything else. It guarantees that our algebraic translation is consistent. It's a promise that the structure of how spaces relate to one another is faithfully mirrored in how their algebraic shadows relate.

With these two simple rules, we can deduce profound truths about topology. This is where the magic happens.

Let's start with a classic puzzle. Suppose you have a map $r: X \to X$ that is its own inverse, meaning doing it twice gets you back to where you started. A reflection is a good example. So, $r \circ r = \text{id}_X$. What does this tell us about its induced map $r_*$? Applying the composition rule, we get $(r \circ r)_* = r_* \circ r_*$. And since $r \circ r$ is the identity map, its induced map must be the identity homomorphism by Rule 1. So, without knowing anything else about the space or the map, we instantly know that $r_* \circ r_* = \text{id}_{\pi_1(X)}$. The algebraic property of the map is perfectly reflected in its induced homomorphism.

Now for the crown jewel. When are two spaces "the same" in topology? When they are ​​homeomorphic​​. This means there's a continuous map $f: X \to Y$ that has a continuous inverse $f^{-1}: Y \to X$. The compositions $f^{-1} \circ f = \text{id}_X$ and $f \circ f^{-1} = \text{id}_Y$. Let's translate this into algebra using our rules!

• From $f^{-1} \circ f = \text{id}_X$, we get $(f^{-1})_* \circ f_* = (\text{id}_X)_* = \text{id}_{\pi_1(X)}$.
• From $f \circ f^{-1} = \text{id}_Y$, we get $f_* \circ (f^{-1})_* = (\text{id}_Y)_* = \text{id}_{\pi_1(Y)}$.

This shows that the homomorphism $f_*$ has an inverse, namely $(f^{-1})_*$. In group theory, a homomorphism with an inverse is called an ​​isomorphism​​. So we've just proven a fundamental theorem: if two spaces are homeomorphic, their fundamental groups must be isomorphic!. They have the "same" algebraic shadow. This is how we can prove two spaces are different: if their fundamental groups are not isomorphic, they cannot be homeomorphic. For example, the circle $S^1$ has the group of integers $\mathbb{Z}$ as its shadow, while the sphere $S^2$ has the trivial group $\{e\}$. Since $\mathbb{Z}$ and $\{e\}$ are not isomorphic, we know a sphere can't be deformed into a circle.

The rules give us power even in less perfect situations. Consider a subspace $A$ sitting inside a larger space $X$. A ​​retraction​​ is a map $r: X \to A$ that squishes $X$ down onto $A$ but leaves points already in $A$ untouched. If we let $i: A \to X$ be the simple inclusion map, the definition of a retraction means that $r \circ i = \text{id}_A$. Let's apply our rules: $(r \circ i)_* = r_* \circ i_* = (\text{id}_A)_* = \text{id}_{\pi_1(A)}$. The existence of this relationship in the world of groups immediately tells us that the homomorphism $r_*$ must be ​​surjective​​ (it hits every element in the target group) and $i_*$ must be ​​injective​​ (it doesn't collapse any distinct elements). This gives us a powerful tool. For instance, we can prove that you cannot retract a solid disk ($D^2$) onto its boundary circle ($S^1$), because this would imply a surjective homomorphism from the trivial group $\pi_1(D^2) = \{e\}$ to the infinite group $\pi_1(S^1) = \mathbb{Z}$, which is impossible.

There's one more layer of sophistication. What if two maps, $f$ and $g$, from $X$ to $Y$ are not identical, but one can be continuously deformed into the other? We say such maps are ​​homotopic​​. The theory tells us that the algebraic shadow cannot see this continuous blurring. If $f$ is homotopic to $g$, then their induced homomorphisms are identical: $f_* = g_*$.

A beautiful consequence of this is when a map is ​​nullhomotopic​​, meaning it can be deformed into a constant map. Since it's homotopic to a constant map, its induced homomorphism must be the same as the one induced by the constant map. And as we saw earlier, that is the zero (trivial) homomorphism. Any map that can be "shrunk to a point" algebraically annihilates all the loops in the original space.

The translation from topology to algebra is powerful, but it's not always a direct, one-to-one dictionary of properties. A common mistake is to assume that if a map $f$ has a certain property (like being one-to-one or onto), then the induced map $f_*$ must have the same property. Nature is more subtle and interesting than that.

Consider an ​​injective​​ (one-to-one) map. Take a rubber band ($S^1$) and place it inside a solid disk ($D^2$). This inclusion map is clearly injective. The fundamental group of the circle, $\pi_1(S^1)$, is $\mathbb{Z}$, generated by the loop that goes once around. The fundamental group of the disk, $\pi_1(D^2)$, is trivial, $\{e\}$, because any loop in a disk can be shrunk to a point. The induced map $f_*: \mathbb{Z} \to \{e\}$ takes the non-trivial loop on the circle and maps it to a loop in the disk, which is trivial. So, the entire non-trivial group $\mathbb{Z}$ is mapped to the identity element. The homomorphism $f_*$ is anything but injective!.

Likewise, consider a ​​surjective​​ (onto) map. The map $f(t) = \exp(i2\pi t)$ takes the entire real line $\mathbb{R}$ and wraps it infinitely around the unit circle $S^1$. This map is clearly surjective. But what about the induced homomorphism? The space $\mathbb{R}$ is contractible, so its fundamental group is trivial, $\pi_1(\mathbb{R}) = \{e\}$. The circle's group is $\pi_1(S^1) = \mathbb{Z}$. The induced map $f_*: \{e\} \to \mathbb{Z}$ can only map the single identity element of its domain to the identity element of its codomain. Its image is just $\{0\}$, which is a far cry from the entire group $\mathbb{Z}$. So, the induced map is not surjective at all.

These "cautionary tales" are not failures of the theory; they are its most profound lessons. They reveal the deep and sometimes non-intuitive ways that topology and algebra are connected. The induced homomorphism doesn't just copy properties blindly; it reveals the consequences of a topological action in the algebraic realm. It's a tool for seeing the unseen, for understanding the structure of shapes in a language of pure logic.

Having understood the principles of induced homomorphisms, we are now ready to embark on a journey to see them in action. If the previous section was about learning the rules of a new game, this section is about playing it—and discovering that its reach extends far beyond the board we started on. The true power and beauty of a mathematical concept are revealed not in its definition, but in its applications. The induced homomorphism is a master key, unlocking insights across topology, algebra, and even physics by acting as a kind of magical translator.

Imagine you are trying to understand a complex, squishy, three-dimensional object. It's difficult to get a grip on. What if you could shine a light on it and study its two-dimensional shadow? The shadow is simpler, flatter, and easier to analyze. You lose some information—the shadow of a sphere and a disk might look the same—but you also preserve crucial features. The induced homomorphism is our "light source." It projects the rich, often infinitely complex, world of topological spaces onto the more rigid, discrete, and countable world of algebraic groups. By studying the "shadow" (the algebraic structure), we can deduce profound truths about the object itself (the topological space).

The magic of this translator lies in its consistency. It obeys a simple set of rules, which mathematicians call functoriality. The two most important rules are: first, mapping a space to itself with the identity map induces the identity homomorphism on the algebraic side. Second, and more powerfully, if you apply one map and then another ($g \circ f$), the induced homomorphism is the same as composing the two individual induced homomorphisms ($g_* \circ f_*$). This second rule is a godsend; it allows us to analyze a complex process by breaking it down into a sequence of simpler steps, a strategy that is the bread and butter of science.

Perhaps the most dramatic application of induced homomorphisms is in proving that something is impossible. You might have an intuition that you can't, say, continuously map a doughnut's surface onto a sphere without tearing it, but how do you prove such a thing? This is where our translator provides an algebraic veto.

Consider the fundamental group of a torus (the surface of a doughnut), $\pi_1(T^2)$, which is isomorphic to $\mathbb{Z} \times \mathbb{Z}$. A key feature of this group is that it is abelian—its elements commute. This is an algebraic reflection of a geometric fact: on a torus, you can trace a path around the length, then the width, and you'll end up at the same point as if you'd gone around the width, then the length. The paths commute. Now, consider the wedge sum of two circles, $S^1 \vee S^1$ (two circles joined at a single point). Its fundamental group, $\mathbb{Z} * \mathbb{Z}$, is famously non-abelian. Traversing the first circle and then the second is a fundamentally different loop from traversing the second and then the first.

Now, let's ask: can we find a continuous map from the torus to the wedge sum that wraps the torus's two principal loops onto the two circles of the wedge? Our intuition might be fuzzy, but algebra gives a clear "no". If such a map existed, its induced homomorphism would have to map the commuting generators of $\pi_1(T^2)$ to the non-commuting generators of $\pi_1(S^1 \vee S^1)$. But a homomorphism must preserve the structure of the original group. It must map a commuting relationship to another commuting relationship. Since the target elements don't commute, the translator throws up an error. No such homomorphism can exist, and therefore, no such continuous map can exist. The algebraic shadow forbids it.

This same principle allows us to answer other fundamental topological questions. For example, is the real projective plane, $\mathbb{R}P^2$, contractible? That is, can we continuously shrink it to a single point? We can prove it cannot be. The identity map on $\mathbb{R}P^2$ induces the identity homomorphism on its fundamental group, $\pi_1(\mathbb{R}P^2) \cong \mathbb{Z}_2$. A constant map (shrinking the whole space to a point) must induce the trivial homomorphism, which sends everything to the identity element. Since the identity and trivial homomorphisms are different for the group $\mathbb{Z}_2$, the identity map cannot be continuously deformed into the constant map. Again, a simple algebraic calculation provides a definitive answer to a subtle geometric question.

Beyond proving impossibilities, induced homomorphisms are powerful tools for classification and analysis. They help us understand the very nature of maps and the spaces they connect.

When we consider a simple geometric projection, like the map $p_1$ from a torus $T^2 = S^1 \times S^1$ to its first circle factor $S^1$, the induced homomorphism does exactly what you'd expect. It takes the algebraic product $\mathbb{Z} \times \mathbb{Z}$ and projects it onto its first factor, $\mathbb{Z}$, via the map $(m,n) \mapsto m$. The algebra perfectly mirrors the geometry, giving us confidence that our "shadow" is a faithful one.

Sometimes, the shadow reveals a surprising hidden simplicity. Consider the antipodal map on a circle, $f(z) = -z$. This map flips every point to its opposite. Geometrically, this feels like a significant transformation. Yet, when we compute the induced homomorphism on the fundamental group $\pi_1(S^1)$, we find it's just the identity map—the same homomorphism induced by doing nothing at all! This tells us that, from the perspective of a loop, the antipodal map is homotopic to the identity. You can continuously deform one into the other (imagine rotating the circle by $\pi$ radians). The induced homomorphism cuts through the superficial geometry to reveal a deeper topological equivalence.

This tool becomes even more powerful when dealing with covering spaces. A covering space is like an "unwrapped" version of a space. For example, the real line $\mathbb{R}$ is the universal covering space of the circle $S^1$; you can think of it as coiling the infinite line into a circle. A fundamental question is: given a map into a space, can we "lift" it to a map into its covering space? The answer, provided by the lifting criterion, depends entirely on the induced homomorphism. A map $f: Y \to X$ can be lifted to the universal cover of $X$ if and only if its induced homomorphism $f_*: \pi_1(Y) \to \pi_1(X)$ is trivial—that is, if it "kills" all the loops in $Y$. The algebraic shadow of the map must be zero. This criterion is central to the theory of covering spaces and has profound consequences, including a deep connection between the triviality of the induced homomorphism and whether the original map is nullhomotopic (deformable to a constant map).

The concept of an induced map is so fundamental that it appears all over mathematics, far beyond its origins in topology. It is a recurring pattern, a testament to the unified structure of mathematical thought.

In abstract algebra, if you have a homomorphism between two rings, say $\phi: R \to S$, does this tell you anything about their groups of units (the elements with multiplicative inverses)? Yes. The ring homomorphism naturally induces a group homomorphism between their groups of units, $\phi|_{R^\times}: R^\times \to S^\times$. And just as in topology, this induced map doesn't necessarily inherit all the properties of the original. A surjective ring homomorphism does not always induce a surjective homomorphism on the units, as the simple example of the projection from the integers $\mathbb{Z}$ to the integers modulo 5, $\mathbb{Z}_5$, demonstrates. This parallel is not a coincidence; it reflects a deep structural principle captured by the language of category theory.

In this more abstract realm, we study functors like $\text{Hom}(A, -)$, which are "machines" that take groups and homomorphisms as inputs and produce new groups and homomorphisms as outputs. We can then ask about the properties of these machines. For instance, does the $\text{Hom}(A, -)$ machine turn surjective maps into surjective maps? The answer is "not always," and finding a counterexample is a problem of induced maps. The powerful Five Lemma is another star player in this world. It's a theorem about diagrams of induced maps, which allows us to deduce that a map is an isomorphism if the maps around it are, providing a powerful logical tool for deducing properties in complex nested structures, such as relative homology groups.

Perhaps one of the most stunning interdisciplinary applications is in the world of Lie groups and modern physics. Lie groups are spaces that are also groups, such as the group of all rotations in three dimensions, $SO(3)$. They are the mathematical language of symmetry. These spaces are typically curved and non-linear, making calculations difficult. However, any homomorphism between Lie groups induces a simple linear map between their corresponding Lie algebras—the tangent spaces at the identity. This is a monumental simplification. For example, the determinant is a complicated multiplicative homomorphism from the group of invertible matrices $GL(n, \mathbb{R})$ to the non-zero real numbers $\mathbb{R}^*$. Its induced Lie algebra homomorphism is simply the trace, a straightforward sum of diagonal elements!. A non-linear, multiplicative problem is translated into a linear, additive one. This principle is the cornerstone of how physicists study the symmetries of the universe, from the rotation of a rigid body to the fundamental forces of nature described by the Standard Model.

From proving that a space cannot be shrunk, to understanding the symmetries of physical law, the induced homomorphism is our faithful guide. It is a simple idea with immense consequences, a testament to the interconnectedness of mathematics and a beautiful example of how by studying the shadow, we can learn to see the object in a whole new light.