Curves and Spaces

Geometry is one of the oldest branches of mathematics, and for most of human history it meant one thing: Euclidean geometry. The flat, ruler-and-compass world described by Euclid around 300 BCE was so successful and so intuitive that it dominated mathematical thinking for over two thousand years. It described the shapes of fields, the angles of buildings, and the paths of light. For practical purposes on a human scale, it worked beautifully. But it was always, in a deep sense, an approximation of a more complex reality.

The cracks in the Euclidean picture began to show when mathematicians started asking what would happen if Euclid's famous parallel postulate were simply not true. In the early nineteenth century, mathematicians including Gauss, Bolyai, and Lobachevsky independently discovered that perfectly consistent geometries could exist where parallel lines behave differently. These non-Euclidean geometries were not just mathematical curiosities. They turned out to describe real physical spaces, and they opened the door to an entirely new way of thinking about shape and space.

Riemann took this revolution further in his landmark 1854 lecture, On the Hypotheses Which Lie at the Foundations of Geometry. He proposed a general framework for geometry in any number of dimensions, where the notion of distance could vary from point to point. This was a profound conceptual leap. Instead of assuming that space has a fixed, uniform structure, Riemann allowed the geometry itself to be a dynamic, variable thing. That idea would later become the mathematical backbone of Einstein's general relativity.^[1]

Today, geometry matters far beyond physics. In computer graphics, curved surfaces must be represented and rendered accurately. In robotics, the configuration space of a robot arm is a curved manifold, and planning efficient motions requires geometric tools. In machine learning, data often lives on curved, high-dimensional spaces, and algorithms that respect that curvature outperform those that ignore it. Geometry is not an abstract luxury. It is a practical necessity in a wide range of modern fields.

Understanding geometry at this deeper level also sharpens mathematical thinking more broadly. Working with curved spaces forces you to question assumptions that feel obvious in flat space. What does it mean for two vectors to be parallel if they live at different points? What is the shortest path between two locations on a curved surface? These questions have precise, beautiful answers, and finding them builds the kind of rigorous, flexible thinking that is valuable across all of mathematics and science.

Perhaps most importantly, geometry connects the abstract and the concrete in a way that few other areas of mathematics do. A formula for curvature is not just a symbol on a page. It tells you something real about the shape of a space, something you can sometimes see and always reason about. That connection between calculation and meaning is one of the most satisfying aspects of Riemannian geometry, and it runs through every chapter of this book.

To appreciate what Riemannian geometry does, it helps to start with what you already know and then watch it transform. In flat Euclidean space, distance is measured by the Pythagorean theorem. In two dimensions, the distance between two nearby points separated by small amounts dx and dy is given by ds² = dx² + dy². This formula is simple, uniform, and the same everywhere in the plane. The geometry does not change depending on where you are.

Now imagine moving to the surface of a sphere. Suddenly, the same formula no longer works. The distance between two nearby points depends on where on the sphere you are, because the sphere curves differently near the poles than near the equator. To measure distances correctly, you need a more general formula, one that can vary from point to point. That generalized distance formula is encoded in the metric tensor, the central object of Riemannian geometry. It is a matrix-valued function that tells you, at each point of a space, how to measure distances and angles in the directions available there.

The transition from flat to curved also changes what counts as a straight line. In flat space, straight lines are the shortest paths between points, and they are also the paths that do not turn. On a curved surface, these two properties still define a special class of curves, but those curves are no longer straight in the Euclidean sense. On a sphere, the shortest paths between points are arcs of great circles, like the flight paths of airplanes traveling between continents. These curves are called geodesics, and they play the role of straight lines in curved geometry.

Curvature itself is the measure of how much a space deviates from being flat. A flat piece of paper has zero curvature. A sphere has positive curvature, which is why the angles of a triangle drawn on a sphere add up to more than 180 degrees. A saddle-shaped surface has negative curvature, where triangle angles sum to less than 180 degrees. These are not just visual descriptions. They correspond to precise mathematical quantities that can be computed from the metric tensor.

A simple way to test curvature: draw a triangle on a surface by connecting three points with geodesics (shortest paths). If the angles sum to exactly 180°, the surface is flat. More than 180° means positive curvature (like a sphere). Less than 180° means negative curvature (like a saddle). This angle-sum test is a direct, measurable consequence of the underlying geometry.

One of the most important insights in Riemannian geometry is that curvature is an intrinsic property of a space. This means it can be detected by measurements made entirely within the space, without any reference to a surrounding higher-dimensional environment. A two-dimensional creature living on the surface of a sphere could, in principle, discover that their world is curved by measuring triangles, without ever being able to step off the surface and look at it from outside. This intrinsic perspective is what makes Riemannian geometry so powerful and so general.

The journey from flat to curved spaces is also a journey in abstraction. Euclidean geometry lives in a specific, concrete setting. Riemannian geometry works in any smooth space of any dimension, as long as you can define a metric. This generality is not just mathematical elegance for its own sake. It is what allows the same framework to describe the two-dimensional surface of a donut, the four-dimensional spacetime of general relativity, and the infinite-dimensional spaces that appear in some areas of mathematical physics. The abstraction is the point.

As you move through this book, you will build up the tools needed to work in this general setting. Each new concept will be introduced with concrete examples first, so that the abstraction always has a foothold in something familiar. The sphere, the flat plane, and the hyperbolic plane will serve as recurring examples throughout, giving you a set of test cases to check your understanding against as the theory develops.

This book covers the core ideas of Riemannian geometry in a logical sequence, starting from the most basic building blocks and working up to sophisticated concepts and real-world applications. Here is a map of the journey ahead, so you know what to expect and how each piece fits into the larger picture.

Chapter 1: Foundations. The book begins with manifolds, the fundamental objects of study in differential geometry. A manifold is a space that looks locally like flat Euclidean space, even if its global shape is curved. The surface of a sphere is a manifold. So is the configuration space of a mechanical system. You will learn what makes a space a manifold, how smooth maps between manifolds work, and how to define tangent vectors at a point. These ideas form the scaffolding on which everything else is built.

Chapter 2: The Metric Tensor. Once you have a manifold, you need a way to measure distances and angles. That is the job of the metric tensor. This chapter introduces the metric as a smoothly varying inner product on the tangent spaces of a manifold. You will see how to compute lengths of curves, angles between vectors, and areas of regions using the metric. Concrete examples include the standard metric on the sphere, the flat metric on Euclidean space, and the hyperbolic metric on the upper half-plane.

Chapter 3: Connections and Parallel Transport. On a curved surface, moving a vector from one point to another is not straightforward. The notion of a connection makes this precise by defining what it means to differentiate a vector field along a curve. Parallel transport is the process of moving a vector along a curve while keeping it as constant as possible. You will learn about the Levi-Civita connection, the unique connection that is compatible with the metric and has no torsion, and see how it gives rise to covariant derivatives.

Chapter 4: Geodesics. Geodesics are the curves that generalize straight lines to curved spaces. They are the paths of shortest distance (at least locally) and the paths along which a freely moving particle travels in the absence of external forces. This chapter derives the geodesic equations, works through examples on the sphere and other surfaces, and explores key properties like the exponential map and geodesic completeness.

Chapter 5 and 6: Curvature. These two chapters form the heart of the book. Chapter 5 introduces the Riemann curvature tensor, the fundamental object that measures how much a space curves, and derives from it the sectional, Ricci, and scalar curvatures. Chapter 6 puts these ideas to work with explicit calculations on familiar spaces, including the sphere, hyperbolic space, and product manifolds. By the end of these chapters, you will be able to compute curvature from a given metric and interpret what the result means geometrically.

Chapter 7: Special Spaces. Some of the most important Riemannian manifolds are those with constant curvature or high degrees of symmetry. This chapter studies Euclidean space, the sphere, and hyperbolic space as the three model geometries of constant curvature. It also introduces symmetric spaces and homogeneous spaces, which arise naturally in Lie group theory and have beautiful geometric properties. Isometries, the distance-preserving maps of a Riemannian manifold, are explored in detail.

Chapter 8: Applications. The final chapter connects the theory to the real world. You will see how Riemannian geometry underlies Einstein's general relativity, where the metric tensor of spacetime encodes the gravitational field. You will also explore applications in optimization on manifolds, a growing area with direct relevance to machine learning and data science. Other applications include shape analysis, medical imaging, and network geometry. These examples show that Riemannian geometry is not just beautiful mathematics. It is also genuinely useful.

Throughout all of these chapters, the emphasis is on building understanding rather than just presenting results. Proofs are included where they illuminate the ideas, and skipped or sketched where they would obscure them. Worked examples appear regularly, and each chapter ends with a summary of the key ideas. The goal is always to leave you with a clear, confident grasp of the material, not just a collection of formulas.

This book is designed to be read sequentially. Each chapter builds on the ones before it, and the concepts introduced early on are used repeatedly in later chapters. If you skip ahead, you may find yourself missing the definitions or intuitions that make a later argument make sense. That said, if you already have a solid background in manifolds and smooth maps, you can move through Chapter 1 quickly and spend more time on the metric tensor and curvature chapters, where the specifically Riemannian content begins.

The most effective way to engage with this material is to work actively rather than passively. When a definition is introduced, pause and make sure you can state it in your own words. When an example is worked through, try to reproduce the key steps yourself before reading the solution. When a formula appears, ask yourself what each symbol means and what the formula is actually computing. Mathematics is not a spectator sport, and Riemannian geometry in particular rewards the reader who engages with it hands-on.

Here are some specific strategies that will help you get the most out of each chapter:

• Draw pictures. Geometric intuition is central to this subject. Whenever a new concept is introduced, try to sketch what it looks like on a simple surface like the sphere or the plane. Even rough sketches help anchor abstract ideas.
• Work through the examples. Every worked example in this book is chosen to illustrate a key idea. Do not just read them. Cover the solution and try the calculation yourself first, then compare your approach to the one presented.
• Check special cases. When you encounter a general formula, plug in the flat Euclidean metric and verify that you recover the familiar result. This is a powerful way to build confidence and catch errors.
• Connect concepts across chapters. The metric tensor, connection, geodesics, and curvature are all deeply related. As you progress, look for the threads that connect them. For example, the Levi-Civita connection is defined using the metric, and geodesics are defined using the connection. Seeing these links makes the theory feel coherent rather than fragmented.
• Return to earlier sections. It is normal to find that a concept from Chapter 2 makes more sense after you have worked through Chapter 4. Do not hesitate to go back and reread earlier material with fresh eyes.

The mathematical notation in this book follows standard conventions in differential geometry. Indices are used extensively, particularly in the Einstein summation convention, where repeated upper and lower indices are summed over. For example, g_{ij} dx^i dx^j means summing over all values of i and j. This notation will be introduced carefully when it first appears, and a glossary at the back of the book collects all the key terms and symbols for easy reference.

If you find a particular section difficult, the best approach is to keep moving forward rather than getting stuck. Sometimes a concept that seems opaque in isolation becomes clear once you see how it is used in the next section. You can always return to a difficult passage after gaining more context. Riemannian geometry has a way of revealing its logic gradually, and patience is genuinely rewarded.

For those who want to go deeper, each chapter points toward more advanced topics and references. The Further Reading section at the end of the book lists classic texts and modern resources for continuing your study. Some of the most rewarding next steps include the study of Riemannian comparison theorems, the geometry of Lie groups, and the role of curvature in global topology through results like the Gauss-Bonnet theorem and the sphere theorem.^[2]

Finally, it is worth saying something about the spirit in which this book is written. Riemannian geometry is a subject of genuine beauty. The way that a single object, the metric tensor, gives rise to distances, angles, geodesics, and curvature is deeply satisfying. The fact that this abstract mathematical structure turns out to describe the physical universe at its largest scales is astonishing. The aim throughout this book is to share that sense of wonder alongside the technical content, so that the mathematics feels alive rather than mechanical. Geometry, at its best, is a way of seeing. This book is an invitation to see more clearly.