Lecture 9: Paths, Homotopy of Paths, and the Fundamental Group

← Back to Course Home

KSM3E02: Algebraic Topology I — Fall 2026   |   Tuesday, September 8, 2026 (12:00–1:00 PM)   |   Instructor: Pranav Haridas

Having developed the foundations of CW complexes, cell attachments, and homotopy equivalence in Lectures 1–8, we now turn to our first major algebraic invariant: the fundamental group $\pi_1(X, x_0)$. Introduced by Henri Poincaré in 1895, the fundamental group captures 1-dimensional "holes" and loops in a topological space by assigning to each pointed space a group whose elements are homotopy classes of loops based at $x_0$.

In this lecture, we rigorously define paths, path homotopy (homotopy relative to endpoints), path composition, and verify in detail that path homotopy classes of loops form a group under concatenation.

1. Paths and Homotopy Relative to Endpoints

Let $X$ be a topological space, and let $I = [0,1]$ denote the standard closed unit interval equipped with the subspace topology from $\mathbf{R}$.

Definition 1 (Paths and Loops).

  1. A path in $X$ from $x_0$ to $x_1$ is a continuous map $\gamma: I \to X$ such that $\gamma(0) = x_0$ (the initial point) and $\gamma(1) = x_1$ (the terminal point).
  2. If $\gamma(0) = \gamma(1) = x_0$, we say that $\gamma$ is a loop based at $x_0$. The point $x_0$ is called the basepoint.

To form an algebraic structure, we cannot simply compose arbitrary continuous curves directly; rather, we consider paths up to a deformation that leaves their endpoints fixed.

Definition 2 (Path Homotopy).

Let $\gamma_0, \gamma_1: I \to X$ be two paths in $X$ having the same initial point $x_0$ and the same terminal point $x_1$. A path homotopy between $\gamma_0$ and $\gamma_1$ is a continuous map $H: I \times I \to X$ satisfying:

  • $H(s, 0) = \gamma_0(s)$ and $H(s, 1) = \gamma_1(s)$ for all $s \in I$,
  • $H(0, t) = x_0$ for all $t \in I$ (initial endpoint remains fixed),
  • $H(1, t) = x_1$ for all $t \in I$ (terminal endpoint remains fixed).

If such a map $H$ exists, we say that $\gamma_0$ and $\gamma_1$ are path homotopic, denoted $\gamma_0 \simeq_p \gamma_1$ (or simply $\gamma_0 \simeq \gamma_1$ when the context is clear).

Proposition 1 (Equivalence Relation).

Path homotopy $\simeq_p$ is an equivalence relation on the set of paths in $X$ with fixed initial point $x_0$ and terminal point $x_1$.

Proof

Reflexivity: For any path $\gamma$, the constant homotopy $H(s, t) = \gamma(s)$ is continuous, fixes endpoints, and shows $\gamma \simeq_p \gamma$.

Symmetry: If $H(s, t)$ is a path homotopy from $\gamma_0$ to $\gamma_1$, then $H'(s, t) = H(s, 1-t)$ is continuous, satisfies $H'(s, 0) = \gamma_1(s)$, $H'(s, 1) = \gamma_0(s)$, and leaves the endpoints fixed. Thus $\gamma_1 \simeq_p \gamma_0$.

Transitivity: If $H_1$ is a path homotopy from $\gamma_0$ to $\gamma_1$ and $H_2$ is a path homotopy from $\gamma_1$ to $\gamma_2$, define $H: I \times I \to X$ by: $$ H(s, t) = \begin{cases} H_1(s, 2t), & 0 \le t \le 1/2 \\ H_2(s, 2t-1), & 1/2 \le t \le 1 \end{cases} $$ By the Pasting Lemma, since $H_1(s, 1) = \gamma_1(s) = H_2(s, 0)$, $H$ is continuous, and it clearly preserves the endpoints for all $t$. Hence $\gamma_0 \simeq_p \gamma_2$.

The equivalence class of a path $\gamma$ under path homotopy is denoted by $[\gamma]$.

2. Path Composition and Reparametrization

Given two paths where the first ends where the second begins, we can traverse them in succession at twice the speed.

Definition 3 (Path Concatenation / Product).

Let $f, g: I \to X$ be paths such that $f(1) = g(0)$. The product path $f * g: I \to X$ is defined by:

$$ (f * g)(s) = \begin{cases} f(2s), & 0 \le s \le \frac{1}{2} \\ g(2s - 1), & \frac{1}{2} \le s \le 1 \end{cases} $$

Since $f(2 \cdot \frac{1}{2}) = f(1) = g(0) = g(2 \cdot \frac{1}{2} - 1)$, the Pasting Lemma guarantees that $f * g$ is continuous, and is a path from $f(0)$ to $g(1)$.

Lemma 1 (Well-Definedness on Homotopy Classes).

If $f_0 \simeq_p f_1$ and $g_0 \simeq_p g_1$ with $f_0(1) = g_0(0)$, then $f_0 * g_0 \simeq_p f_1 * g_1$.

Proof

Let $F: I \times I \to X$ be a path homotopy from $f_0$ to $f_1$, and let $G: I \times I \to X$ be a path homotopy from $g_0$ to $g_1$. Define $H: I \times I \to X$ by:

$$ H(s, t) = \begin{cases} F(2s, t), & 0 \le s \le 1/2 \\ G(2s-1, t), & 1/2 \le s \le 1 \end{cases} $$

At $s = 1/2$, $F(1, t) = f_0(1) = g_0(0) = G(0, t)$ for all $t$. By the Pasting Lemma, $H$ is continuous. Furthermore, $H(0, t) = F(0, t) = f_0(0)$ and $H(1, t) = G(1, t) = g_0(1)$, so $H$ fixes both endpoints throughout. Hence $f_0 * g_0 \simeq_p f_1 * g_1$.

Consequently, we define the product on path homotopy classes by $[f] * [g] = [f * g]$.

3. The Group Structure of $\pi_1(X, x_0)$

Let $c_{x_0}: I \to X$ denote the constant path $c_{x_0}(s) = x_0$, and for any path $f$, let $\bar{f}(s) = f(1-s)$ denote the reverse path.

Theorem 1 (Fundamental Properties of Path Product).

The operation $*$ on path homotopy classes satisfies:

  1. Associativity: If $f * (g * h)$ is defined, then $([f] * [g]) * [h] = [f] * ([g] * [h])$.
  2. Identity: If $f$ is a path from $x_0$ to $x_1$, then $[c_{x_0}] * [f] = [f] = [f] * [c_{x_1}]$.
  3. Inverse: For any path $f$ from $x_0$ to $x_1$, $[f] * [\bar{f}] = [c_{x_0}]$ and $[\bar{f}] * [f] = [c_{x_1}]$.
Proof Details (Reparametrizations)

General Reparametrization Lemma: If $\phi: I \to I$ is a continuous map such that $\phi(0) = 0$ and $\phi(1) = 1$, then for any path $f$, $f \circ \phi \simeq_p f$. The homotopy is simply $H(s, t) = f((1-t)\phi(s) + ts)$.

Associativity: The paths $(f * g) * h$ and $f * (g * h)$ are reparametrizations of one another:

$$ ((f * g) * h)(s) = \begin{cases} f(4s), & 0 \le s \le 1/4 \\ g(4s-1), & 1/4 \le s \le 1/2 \\ h(2s-1), & 1/2 \le s \le 1 \end{cases} \quad\text{and}\quad (f * (g * h))(s) = \begin{cases} f(2s), & 0 \le s \le 1/2 \\ g(4s-2), & 1/2 \le s \le 3/4 \\ h(4s-3), & 3/4 \le s \le 1 \end{cases} $$

The piecewise linear map taking $[0, 1/4] \to [0, 1/2]$, $[1/4, 1/2] \to [1/2, 3/4]$, and $[1/2, 1] \to [3/4, 1]$ gives the desired boundary-preserving homotopy.

Identity: The path $c_{x_0} * f$ stays at $x_0$ for $s \in [0, 1/2]$ and then traverses $f(2s-1)$. Contracting the idle interval $[0, 1/2]$ down to $0$ via a linear homotopy gives $c_{x_0} * f \simeq_p f$.

Inverse: For $f * \bar{f}$, we travel along $f$ and then retrace backwards. At stage $t \in [0, 1]$, define a path that traverses $f$ up to $f(1-t)$ and then immediately reverses back along $f$. At $t=0$, this is $f * \bar{f}$; at $t=1$, this is the constant path $c_{x_0}$.

Definition 4 (The Fundamental Group).

Let $X$ be a topological space and $x_0 \in X$. The set of path homotopy classes of loops based at $x_0$, equipped with the operation $[f][g] = [f * g]$, is a group called the fundamental group of $X$ with basepoint $x_0$, denoted:

$$ \pi_1(X, x_0) $$

4. Convex Spaces and Simply Connected Spaces

Example 1 (Convex Subsets of $\mathbf{R}^n$).

Let $C \subseteq \mathbf{R}^n$ be a convex set, and $x_0 \in C$. Any loop $f: I \to C$ based at $x_0$ is path homotopic to the constant loop $c_{x_0}$ via the straight-line homotopy:

$$ H(s, t) = (1-t)f(s) + tx_0. $$

Since $C$ is convex, $H(s, t) \in C$ for all $(s, t) \in I \times I$. Moreover, $H(0, t) = (1-t)x_0 + tx_0 = x_0$ and $H(1, t) = x_0$, so the basepoint remains fixed. Therefore, every loop is homotopic to the constant loop, which implies:

$$ \pi_1(C, x_0) = \{0\} \quad (\text{the trivial group}). $$

Definition 5 (Simply Connected Space).

A topological space $X$ is said to be simply connected if:

  1. $X$ is path-connected, and
  2. $\pi_1(X, x_0)$ is the trivial group for some (and hence any) basepoint $x_0 \in X$.
← Back to Schedule & Course Overview Next: Lecture 10 (Induced Homomorphisms & Change of Basepoint) →