Mathematics
Lie Groups and Their Lie Algebras
Quick fact
The Lie algebra of a Lie group is a vector space that can be seen as the 'linearization' of the group around its identity, and the exponential map lets you 'exponentiate' its elements to recover the full group locally.
Why this is interesting
You've seen symmetries like rotations of a sphere. But what if the symmetry can be applied continuously, in infinitely many small steps? How do we capture the 'tiny' rotations that build up to any rotation?
Read the full explanation
Understanding Lie Groups and Their Lie Algebras
Think of a globe. You can rotate it by any angle about any axis. These rotations form a continuous family: you can perform a rotation by a tiny amount, then another tiny amount, and so on. Such a continuous family of symmetries is called a Lie group. It has two structures at once: it is a group (algebraic) and a smooth manifold (geometric). Now, zoom in near the 'do nothing' rotation, the identity. Infinitesimally, a tiny rotation is essentially described by a vector indicating the axis and an infinitesimal angle. These vectors form a vector space called the tangent space at the identity. This space, plus a way to combine vectors (the bracket), is the Lie algebra. The Lie algebra captures all the 'infinitesimal' information, and from it you can reconstruct the group (near the identity) using the exponential map.
A deeper explanation
The reason this works is a deep interplay between smooth and algebraic structure. For a Lie group G, the tangent space at the identity, denoted g, is a vector space. The group operation (multiplication) induces a bracket on g: for X and Y in g, one defines [X,Y] as the derivative at the identity of the commutator of flows generated by X and Y. This bracket is bilinear, antisymmetric, and satisfies the Jacobi identity, making g a Lie algebra. The exponential map exp: g → G, defined by following the flow of a left-invariant vector field for time 1, locally reconstructs the group structure: multiplication in G is encoded in g via the Baker-Campbell-Hausdorff formula. This linearization is powerful because many problems in the group (which is nonlinear) become linear algebra in the Lie algebra. For example, the classification of Lie groups often reduces to classifying Lie algebras. Moreover, representations of the Lie algebra can often be exponentiated to representations of the group (when the group is simply connected). Thus, the Lie algebra serves as the 'shadow' of the group, capturing its essential structure.