Borel-Weil Theorem

The Borel-Weil theorem provides a concrete realization of the irreducible representations of a compact semisimple Lie group as spaces of holomorphic sections of line bundles over the flag variety. Named after Armand Borel and André Weil, who developed it in the early 1950s, the theorem bridges representation theory, algebraic geometry, and complex manifold theory in a way that makes abstract harmonic analysis visually and computationally tractable.

The construction begins with a compact semisimple Lie group G and a maximal torus T. The flag variety G/T can be identified with the space of Borel subgroups of the complexification of G, and it carries a family of holomorphic line bundles L_λ indexed by dominant integral weights λ. The Borel-Weil theorem states that the space of holomorphic sections H⁰(G/T, L_λ) is either zero (if λ is not dominant) or an irreducible representation of G with highest weight λ (if λ is dominant). Every finite-dimensional irreducible representation of G arises in this way, and the correspondence between dominant weights and line bundles is bijective.

The theorem is the compact-group analog of the orbit method: the flag variety is a symplectic manifold (in fact, a coadjoint orbit), the line bundles are the prequantum bundles of geometric quantization, and the spaces of sections are the quantum Hilbert spaces. This geometric picture extends to the Bott-Borel-Weil theorem, which uses sheaf cohomology in higher degrees to realize representations that do not appear as global sections.

The Borel-Weil theorem is often presented as a tool for constructing representations. It is more than that. It is a proof that representation theory is not separate from geometry — that the algebraic problem of finding all symmetries of a group is identical to the geometric problem of finding all holomorphic line bundles over its flag variety. This identification is not a convenience; it is a hint that the division between algebra and geometry is itself an artifact of how we organize knowledge, not a feature of the mathematics.