Introduction to Algebraic Topology

Foundations of Algebraic Topology What is Algebraic Topology? Algebraic topology is a mathematical field that bridges two seemingly different worlds: topology (the study of shapes and spaces) and algebra (the study of abstract structures like groups and rings). The central idea is remarkably elegant: rather than trying to study spaces directly, we attach algebraic structures to them. These structures capture essential geometric information in a form that's easier to work with algebraically. The Core Problem Algebraic Topology Solves At its heart, topology asks: which properties of a space are preserved when we deform it continuously? For example, if you have a coffee mug and a doughnut, you could smoothly reshape one into the other without tearing or gluing—they're "topologically the same" even though they look different. But a coffee mug and a sphere are fundamentally different: no amount of continuous deformation can turn one into the other. The challenge is: how do we prove that two spaces are genuinely different? Algebraic topology answers this by creating algebraic invariants—properties that don't change under continuous deformation. If two spaces have different invariants, they cannot be continuously deformable into each other. Topological Spaces (Necessary Background) A topological space is a set of points equipped with a notion of "closeness." More formally, it specifies which subsets are "open," which determines which points can be continuously deformed into one another. You don't need to memorize the formal definition; just think of a topological space as a geometric shape where we can talk about continuous deformations and which points are "near" each other. Key Spaces We'll Encounter Three fundamental examples appear repeatedly in algebraic topology: The circle ($S^1$): A one-dimensional closed loop, like the boundary of a disk. The sphere ($S^2$): The surface of a ball, a two-dimensional surface in three-dimensional space. The torus ($T^2$): The surface of a doughnut—it has a hole through the middle. Invariants: Detecting Topological Differences An invariant is a property assigned to a space that remains the same whenever we apply a homeomorphism (a continuous deformation that can be undone by another continuous deformation). Think of invariants as "fingerprints" for spaces. Two spaces that are topologically different will have different invariants, allowing us to distinguish them. Two spaces with the same invariants might be the same, or they might just be indistinguishable using that particular invariant. The fundamental group and homology groups are the two main invariants we'll study. Both assign algebraic structures to spaces in a way that respects continuous deformations. Fundamental Group What is the Fundamental Group? The fundamental group is the first and most intuitive algebraic invariant in algebraic topology. It captures how many fundamentally different ways loops can wrap around a space. Definition and Construction To define the fundamental group of a space $X$, we first choose a base point $x0$ in the space. The fundamental group $\pi1(X, x0)$ consists of equivalence classes of loops based at $x0$. A loop is simply a continuous path that starts and ends at $x0$. Two loops are considered equivalent (or homotopic) if one can be continuously deformed into the other while always keeping the starting and ending point fixed at $x0$. Think of it like moving a rubber band around a surface—you can stretch and wiggle it, but you can't break it or move its fixed endpoint. The Group Structure Here's the key insight: these equivalence classes form a group! The group operation is concatenation: to combine two loops $\alpha$ and $\beta$, follow $\alpha$ from $x0$, then immediately follow $\beta$ from $x0$ (imagine speeding up your path traversal so you finish both loops in the same amount of time). This respects the equivalence relation, so concatenation is well-defined on equivalence classes. The identity element is the trivial loop that stays at $x0$. The inverse of a loop is the same loop traversed in reverse. This operation is associative, so we genuinely have a group. Fundamental Group of the Circle The fundamental group of the circle is $\pi1(S^1) = \mathbb{Z}$, the group of integers under addition. What does this mean? It means every loop around the circle can be classified by a single integer: how many times it winds around the circle. A loop that winds clockwise twice is different from a loop that winds counterclockwise three times (which would be classified as $-3$). The trivial loop has winding number $0$. This complete classification is remarkably clean: the fundamental group exactly counts the winding number. Fundamental Group of the Sphere In contrast, the fundamental group of the sphere is $\pi1(S^2) = \{e\}$, the trivial group containing only the identity element. This means every loop on a sphere can be continuously shrunk to a point. Intuitively, this is because the sphere has no "hole" for a loop to wrap around. Interpretation: Detecting Holes Here's the conceptual takeaway: a non-trivial fundamental group signals the presence of a "hole" in the space that loops cannot avoid. Circle: has a hole in the middle → loops can wrap around it → non-trivial $\pi1$ Sphere: has no hole → all loops can be contracted to a point → trivial $\pi1$ For the torus (a doughnut), you can imagine loops wrapping around the hole in the middle or loops wrapping around the tube. This gives $\pi1(T^2) = \mathbb{Z} \times \mathbb{Z}$, two independent winding numbers. Homology and Cohomology While the fundamental group captures information about loops, homology groups provide a more systematic way to detect "holes" at every dimension. Building Homology: Simplices and Chain Complexes Before we can compute homology, we need to break our space into simple pieces. Simplices: The Building Blocks A simplex is the simplest possible geometric object in each dimension: 0-simplex: A single point 1-simplex: A line segment connecting two points 2-simplex: A filled triangle 3-simplex: A filled tetrahedron And so on in higher dimensions By combining simplices—gluing them together along their boundaries—we can construct any space we want. This is called a simplicial complex. For example, the surface of a sphere can be built from many small triangles glued together. Chain Complexes: Recording Relationships Once we've built our space from simplices, we record the relationships between them using chains and boundaries. A $n$-chain is a formal sum of $n$-simplices (like $\sigma1 + \sigma2 - \sigma3$, where we can use integer coefficients). For each $n$-simplex, we define its boundary: the collection of $(n-1)$-simplices that form its surface. For example, the boundary of a triangle is the three line segments forming its edges. We arrange these chains and boundary maps into what's called a chain complex, which keeps track of how simplices fit together. Homology Groups: Measuring Holes The $n$-th homology group $Hn(X)$ measures $n$-dimensional holes in a space. It's defined as the quotient: $$Hn(X) = \frac{\text{n-cycles}}{\text{n-boundaries}}$$ Don't let the notation intimidate you. Here's the intuition: An $n$-cycle is a chain whose boundary is zero—it's "closed" with no loose ends An $n$-boundary is something that forms the boundary of an $(n+1)$-chain An $n$-hole is detected by an $n$-cycle that is not a boundary of any $(n+1)$-chain. Intuitively, it's something closed that doesn't bound a higher-dimensional region. Interpreting Each Homology Group $H0(X)$: Counts the connected components of $X$. If your space is a single connected piece, $H0(X) = \mathbb{Z}$. If it's three disconnected pieces, $H0(X) = \mathbb{Z}^3$. $H1(X)$: Detects one-dimensional holes. A non-trivial $H1$ means there are loops that cannot be filled in by a surface. For instance, the circle has $H1(S^1) = \mathbb{Z}$ because any loop around it cannot be filled. $H2(X)$: Detects two-dimensional voids—regions that are "surrounded" by the space. A sphere has $H2(S^2) = \mathbb{Z}$ because the sphere surrounds its interior. Higher homology: $Hn$ detects $n$-dimensional holes analogously. Example: The Torus The torus $T^2$ (a doughnut surface) is a beautiful example: $H0(T^2) = \mathbb{Z}$: It's connected $H1(T^2) = \mathbb{Z} \oplus \mathbb{Z}$: Two independent one-dimensional holes—one loop can wrap around the central hole, another around the tube $H2(T^2) = \mathbb{Z}$: The torus itself forms a two-dimensional surface that encloses an empty region (the "hole" in the doughnut) Notice how homology captures geometric intuition: the two generators of $H1$ correspond exactly to the two independent ways you can wrap a loop around a torus! <extrainfo> Cohomology (Brief Overview) Cohomology is a dual version of homology. Where homology is built by looking at chains and their boundaries, cohomology reverses the direction of arrows and works with cochains and coboundaries. While homology and cohomology give the same group structures for nice spaces (by the Universal Coefficient Theorem), cohomology has additional structure: it's a graded ring, not just a graded abelian group. This extra algebraic structure sometimes makes cohomology more convenient for advanced applications. For the purposes of most introductory algebraic topology courses, you can think of homology and cohomology as essentially interchangeable invariants—both are functorial and both distinguish spaces that are topologically different. </extrainfo> Functoriality and Induced Maps One of the most powerful aspects of algebraic topology is that its invariants are functorial: they preserve the structure of continuous maps between spaces. How Maps Induce Homomorphisms Suppose we have a continuous map $f: X \to Y$ between two spaces. This map doesn't just passively sit there—it actively induces homomorphisms on the algebraic invariants. Induced Homomorphism on Fundamental Groups If $f: X \to Y$ is continuous and $x0 \in X$, $f(x0) \in Y$, then $f$ induces a group homomorphism: $$f: \pi1(X, x0) \to \pi1(Y, f(x0))$$ defined by taking a loop in $X$ and composing it with $f$ to get a loop in $Y$. This respects the group operation (concatenation) automatically. Induced Homomorphisms on Homology Similarly, every continuous map $f: X \to Y$ induces homomorphisms: $$f: Hn(X) \to Hn(Y)$$ on each homology group. Why This Matters This functoriality is remarkable because it means we're not just looking at spaces in isolation. We can compare spaces via the maps between them. If two spaces are different, we can often prove it by showing that no map can induce certain homomorphisms, or by comparing invariants. <extrainfo> Advanced Tools: Brief Mentions The Classification of Surfaces One of the crown jewels of algebraic topology is the classification of surfaces: using fundamental groups and homology, mathematicians can completely enumerate all compact two-dimensional surfaces. There are infinitely many, but they fall into clear families (orientable vs. non-orientable), and each is determined by its genus (roughly, the number of holes). The Mayer-Vietoris Sequence The Mayer-Vietoris sequence is a powerful computational tool that says: if you decompose a space $X$ as the union of two overlapping subspaces $U$ and $V$, then the homology of $X$ can be computed from the homology of $U$, $V$, and their intersection $U \cap V$. This reduces difficult global computations to easier local ones. These tools appear in more advanced courses and applications, but understanding them requires comfort with the foundations we've covered here. </extrainfo>