Geometry

Geometry is the branch of mathematics concerned with the properties of space, shape, size, and the relative position of figures. Its origins lie in practical problems of measurement — land surveying, architecture, astronomy — but its modern form is far more general: geometry studies the structures that remain invariant under specified transformations, regardless of whether those structures are spatial in any intuitive sense.

The classical Euclidean geometry, codified in Euclid's Elements around 300 BCE, treated geometry as the study of idealized spatial objects (points, lines, planes) subject to axioms about incidence, congruence, and parallelism. For two millennia, Euclidean geometry was considered the necessary structure of physical space. The discovery of non-Euclidean geometries in the 19th century — by Gauss, Bolyai, Lobachevsky, and Riemann — destroyed this necessity. Space, it turned out, could be curved, parallel lines could meet, and the sum of angles in a triangle could differ from 180 degrees. The question of which geometry describes the actual universe became empirical, not a priori.

This transformation had profound consequences. Einstein's general theory of relativity (1915) identified gravity with the curvature of spacetime, making Riemannian geometry the mathematical language of cosmology. The question "what is the geometry of the universe?" became a question for astrophysicists, not philosophers.

Modern geometry is not a single subject but a family of related disciplines:

• Differential geometry studies smooth manifolds and their curvature, with applications in physics, engineering, and computer graphics.
• Algebraic geometry studies the solutions of polynomial equations, connecting algebra and geometry through the correspondence between geometric objects and algebraic structures. This was the framework Grothendieck used to reformulate the foundations of the field, introducing the concept of a scheme and the machinery of sheaf theory and topoi.
• Topology studies properties preserved under continuous deformation — connectivity, holes, dimension — and has become the language of quantum field theory and data science.
• Discrete and computational geometry studies geometric algorithms, convexity, and combinatorial structures, with applications in robotics, computer vision, and machine learning.

The unifying thread is not spatial intuition but structural invariance. Geometry asks: what properties of a structure remain unchanged when the structure is transformed in specified ways? This abstraction makes geometry applicable far beyond its origins. The "geometric" in "geometric deep learning" refers not to physical space but to the metric structure of data manifolds. The "geometry" of a neural network's loss landscape is a question about curvature and topology, not about rulers and compasses.

Geometry is therefore not merely a branch of mathematics. It is a mode of thought — the study of what persists through change — and this mode has proven indispensable across science.