Hatcher 1.2.11: Mapping Torus Proof Simplified

Hey guys, welcome back to Plastik Magazine! Today, we're diving deep into the fascinating world of Algebraic Topology, specifically tackling a rather intricate proof from Hatcher's "Algebraic Topology" – problem 1.2.11. This one deals with the mapping torus, a concept that might sound a bit abstract at first, but trust me, it's super cool once you get the hang of it. We're going to break down the proof step-by-step, making sure you all understand what's going on, especially when $X$ is the wedge sum of two circles, S^1igvee S^1. So, grab your thinking caps, and let's get this done!

What Exactly is a Mapping Torus?

Before we even look at the proof, we gotta nail down what a mapping torus is. Imagine you have a space $X$, and you have a map $f$ that takes points in $X$ and squishes them somewhere else within $X$ (i.e., $f: X o X$). Now, picture $X$ as a shape. We take a copy of $X$ and stretch it out into an interval, $I = [0, 1]$. This gives us $X imes I$. Think of this as a stack of copies of $X$, where each copy is slightly deformed by $f$ as you move up the stack. The mapping torus, denoted as $T_f$, is formed by taking this stacked space $X imes I$ and then gluing the bottom layer ($X imes 0$) to the top layer ($X imes 1$) in a specific way. The rule for gluing is that a point $(x, 0)$ on the bottom is identified with the point $(f(x), 1)$ on the top. It's like bending the interval $I$ into a circle and attaching the ends of the stack, but with a twist dictated by the map $f$. So, essentially, the mapping torus captures how the map $f$ deforms the space $X$ when you 'wrap it up'.

Now, Hatcher's problem 1.2.11 focuses on a specific case: when X = S^1 igvee S^1. This space is like two circles joined at a single point – imagine the number '8' or a pair of spectacles. It's a pretty fundamental building block in topology. Understanding the mapping torus of a map on this specific space is crucial because it helps us understand the fundamental group of more complex spaces. The fundamental group, $\pi_1(Y)$, tells us about the 'loops' in a space $Y$. The mapping torus construction gives us a powerful way to construct new spaces whose fundamental groups are related to the original space's fundamental group and the map $f$. It's all about how $f$ 'twists' the loops of $X$ when forming the torus. So, when we talk about $T_f$ for X = S^1 igvee S^1, we're looking at a space formed by stacking and gluing copies of this 'figure 8' shape, with the gluing process determined by how $f$ maps the 'figure 8' onto itself. This structure has profound implications for the resulting space's topology and its fundamental group.

The Core Statement of Hatcher 1.2.11

Alright, let's get to the nitty-gritty of what Hatcher 1.2.11 actually asks us to prove. The problem states that if $X$ is a path-connected space and $f: X o X$ is a map, then the fundamental group of the mapping torus $T_f$ is related to the fundamental group of $X$ and the induced map on the fundamental group, $f_*: \pi_1(X) o \pi_1(X)$. Specifically, the problem asks to show that $T_f$ is homotopy equivalent to the quotient space $X imes I / \{(x,0) \sim (f(x),1) \mid x \\in X \}$. This might seem a bit redundant at first glance, as this is precisely the definition of the mapping torus. However, the real juice of the problem often lies in understanding the implications of this construction, especially concerning the fundamental group. The proof usually involves showing that the inclusion map $i: X imes \{0 \} o X imes I$ induces an isomorphism on the fundamental group when considering the resulting mapping torus. Or, more generally, it deals with calculating $\pi_1(T_f)$ using the Seifert-van Kampen theorem. The theorem allows us to compute the fundamental group of a space formed by the union of two smaller spaces. In the case of a mapping torus, we can view it as a space formed by 'attaching' a 3-cell (or higher dimensional cells) to $X$ via the map $f$. The essence of Hatcher 1.2.11 is to provide a rigorous way to compute the fundamental group of $T_f$. It essentially states that $\pi_1(T_f)$ can be understood as a semidirect product of $\pi_1(X)$ and the integers $\mathbb{Z}$, where the action of $\mathbb{Z}$ on $\pi_1(X)$ is given by the induced map $f_*$. This relationship is super important because it connects the algebraic structure of the fundamental group of the original space with the dynamics of the map $f$. It tells us that the 'twisting' introduced by $f$ as we go around the 'cylinder' of the mapping torus has a direct algebraic manifestation in the structure of its fundamental group. The problem often involves showing that $T_f$ deformation retracts onto a space whose fundamental group is easier to compute, or it uses the Seifert-van Kampen theorem applied to a clever decomposition of $T_f$. The crucial takeaway is that the algebraic structure of $\pi_1(T_f)$ is not just a simple product but a semidirect product, reflecting the non-trivial action of $f_*$ on $\pi_1(X)$.

Proof Strategy: Decomposing the Mapping Torus

So, how do we actually prove this? The standard strategy for problems like Hatcher 1.2.11, especially when dealing with the fundamental group, often involves the Seifert-van Kampen theorem. This theorem is your best friend when you want to compute the fundamental group of a space that can be written as the union of two path-connected subspaces, say $Y = A \cup B$, where the intersection $A \cap B$ is also path-connected. The theorem gives you the fundamental group of $Y$ in terms of the fundamental groups of $A$, $B$, and $A \cap B$, and the maps induced by inclusions. For the mapping torus $T_f$, we can cleverly decompose it. Think of $X imes I$ before we do the gluing. We can split the interval $I$ into two halves, say $I_1 = [0, 1/2]$ and $I_2 = [1/2, 1]$. Then $X imes I = (X imes I_1) \cup (X imes I_2)$. Let $A = X imes I_1$ and $B = X imes I_2$. The intersection $A \cap B = X imes \{1/2 \}$ is just a copy of $X$. Now, we need to consider the effect of the gluing at the ends. When we identify $(x,0)$ with $(f(x),1)$, this identification happens after we've potentially decomposed the space. A more refined approach for the mapping torus $T_f$ is to consider it as being formed by taking $X imes I$ and then identifying the ends. We can view $T_f$ as the union of two pieces that overlap. Let's take $X imes [0, 1-\epsilon]$ and $X imes [\epsilon, 1]$ for some small $\epsilon > 0$. Their union covers $X imes I$, and their intersection is $X imes [\epsilon, 1-\epsilon]$. When we impose the identification $(x,0) \sim (f(x),1)$, things get a bit more complex. A cleaner way is to think of $T_f$ as being constructed from $X$ itself. We can view $T_f$ as $X imes S^1$, but with a twist. Consider $X imes [0,1]$ with the $(x,0) \sim (f(x),1)$ identification. We can pick a point $x_0 \\in X$. The map $f$ induces $f_*: \pi_1(X, x_0) o \pi_1(X, f(x_0))$. Let's consider a point $p = (x_0, 0)$ in $X imes I$. In $T_f$, this point is identified with $(f(x_0), 1)$. The path $\gamma(t) = (x_0, t)$ for $t \in [0,1]$ becomes a loop in $T_f$ based at the image of $x_0$. The crucial idea is often to show that $T_f$ has a retractile structure. We can often find a subspace of $T_f$ that deformation retracts onto $X$, and then analyze how the 'cylinder' part is attached. Another common strategy involves using the fact that the inclusion $j: X o T_f$ defined by $j(x) = (x, t)$ for any fixed $t$ (say $t=1/2$) is not a homotopy equivalence, but $T_f$ can be thought of as 'attaching' a copy of $X$ 'along' $f$. The Seifert-van Kampen theorem is applied by decomposing $T_f$ into two overlapping path-connected sets. For instance, consider $T_f$ as $X imes [0, 1] / (x, 0) \sim (f(x), 1)$. Let $U = X imes [0, 1)$ and $V = X imes (0, 1]$. The union $U \cup V$ covers $T_f$. The intersection $U \cap V = X imes (0, 1)$. The application of Seifert-van Kampen theorem requires careful handling of the base points and the induced homomorphisms on the fundamental groups. The map $f$ plays a vital role in how the fundamental groups of these pieces combine.

The Case: X = S^1 igvee S^1

Now, let's get specific and talk about the case where X = S^1 igvee S^1. This is where things get really interesting for Hatcher 1.2.11. Remember, S^1 igvee S^1 looks like the number '8'. Let's call the two circles $C_1$ and $C_2$, and let the wedge point be $x_0$. The fundamental group \pi_1(S^1 igvee S^1, x_0) is the free group on two generators, say $a$ and $b$, corresponding to loops that go once around $C_1$ and $C_2$ respectively. So, \pi_1(S^1 igvee S^1) \\cong F(a, b). Now, we have a map f: S^1 igvee S^1 o S^1 igvee S^1. This map $f$ induces a homomorphism $f_*: F(a, b) o F(a, b)$. This homomorphism $f_*$ is determined by where $f$ sends the generators $a$ and $b$. For instance, $f(a)$ could be a loop that goes around $C_1$ some number of times and $C_2$ some other number of times, and similarly for $f(b)$. The problem is about understanding the fundamental group of the mapping torus $T_f$. According to the general result for mapping tori, $\pi_1(T_f)$ is the semidirect product $F(a, b) times_{\phi} \mathbb{Z}$, where $\phi$ is an automorphism of $F(a, b)$ determined by $f_*$ and the fact that the interval $I$ is being 'wrapped around'.

To be more precise, let's think about the structure of $T_f$. We are identifying $(x, 0)$ with $(f(x), 1)$ for x \\in S^1 igvee S^1. A loop in $T_f$ can be thought of as a path in $X imes I$ starting and ending at the same point, possibly crossing the boundary $X imes 0$ and $X imes 1$. Crucially, the path $(x, t)$ for $t \\in [0, 1]$ becomes a loop in $T_f$. This loop's 'homotopy class' is related to the element $a$ or $b$ if it's just around $C_1$ or $C_2$. However, as we traverse the 'cylinder' part of $T_f$, we are essentially moving through different 'copies' of $X$ that are deformed by $f$. The structure of $\pi_1(T_f)$ is derived using the Seifert-van Kampen theorem. We decompose $T_f$ into two overlapping pieces. For instance, let X_0 = S^1 igvee S^1. We can consider $T_f$ as being built from $X_0 imes [0, 1]$ and attaching $X_0$ along $f$. A common way to apply Seifert-van Kampen is to consider $T_f$ as a union of $A = X imes [0, 1]$ (with one end slightly shrunk to a point) and $B = X$ (the 'second copy' of $X$ that gets glued). A more rigorous decomposition would be to take $U = X imes [0, 1)$ and $V = X imes (0, 1]$. The intersection $U \cap V = X imes (0, 1)$. The fundamental group of $U$ is essentially $\pi_1(X)$, and the fundamental group of $V$ is also $\pi_1(X)$. The tricky part is the intersection and how the map $f$ affects the fundamental group. The inclusion of the intersection into $U$ and $V$ induces homomorphisms. The fundamental group of $T_f$ will be the fundamental group of $X$, with an additional generator representing the loop obtained by traversing the 'cylinder' part once, and this generator interacts with the generators of $\pi_1(X)$ via $f_*$. The key insight is that the map $f$ dictates how loops in one 'layer' of the torus are transformed into loops in the next 'layer'. This transformation is precisely what the semidirect product captures. The automorphism $\phi$ of $F(a,b)$ describes how $f$ transforms the loops $a$ and $b$ after one full turn around the torus cylinder. So, for X = S^1 igvee S^1, the fundamental group $\pi_1(T_f)$ will be the free group $F(a,b)$ combined with an action of $\mathbb{Z}$ generated by $f_*$. This algebraic structure precisely encodes the topology of the mapping torus built from this 'figure 8' space.

Implications and Further Study

The result of Hatcher 1.2.11, especially the calculation of the fundamental group of the mapping torus, has profound implications in topology. It allows us to construct spaces with very specific and often complex fundamental groups starting from simpler spaces and maps. For instance, if $f$ is the identity map on $X$, then $T_{id}$ is just $X imes S^1$. In this case, $f_*$ is the identity homomorphism, and the semidirect product becomes a direct product: $\pi_1(X imes S^1) \\cong \pi_1(X) imes \mathbb{Z}$. This is a fundamental result in algebraic topology. If $X = S^1$, then $T_f$ is the mapping torus of a map $f: S^1 o S^1$. Such maps are classified by their degree, $d = f_*(1) \in \mathbb{Z}$. The resulting mapping torus is $S^1 imes_f S^1$. Its fundamental group is $\mathbb{Z} times_d \mathbb{Z}$, which is often called the $(d,1)$-torus knot group. If $d=1$, it's $\mathbb{Z} imes \mathbb{Z}$, the fundamental group of the standard torus $S^1 imes S^1$. If $d=0$, it's $\mathbb{Z}$, as $f$ is homotopic to a constant map. For X = S^1 igvee S^1, as we saw, the fundamental group is $F(a,b) times_{\phi} \mathbb{Z}$. The specific structure of $\phi$ (which depends on $f_*$) determines the precise nature of the resulting group. For example, if $f$ is homotopic to the identity, then $\phi$ is the identity automorphism, and $\pi_1(T_f) \cong F(a,b) imes \mathbb{Z}$. This means that $T_{id}$ for X=S^1 igvee S^1 is homotopy equivalent to (S^1 igvee S^1) imes S^1. This result is powerful because it connects the world of maps between spaces to the algebraic structures of their fundamental groups. It provides a framework for understanding how 'twisting' or 'deformation' in a map $f$ translates into a change in the group structure via the semidirect product. Further study could involve exploring specific examples of maps $f$ on S^1 igvee S^1 and explicitly computing the resulting fundamental group. You could also investigate higher-dimensional analogues or different ways of constructing torus-like spaces. The mapping torus construction is also fundamental in the study of dynamical systems on manifolds and knot theory. So, while Hatcher 1.2.11 might seem like a challenging proof, mastering it opens doors to many exciting areas in mathematics. Keep practicing, guys, and don't hesitate to revisit the Seifert-van Kampen theorem – it's the key! Happy topologizing!