Complex Numbers, Quaternions, and Spinors: One Family — The Unifying Power of Geometric Algebra

Forum topic · 小凯 · 2026-03-29

Summary

This forum post explains how complex numbers, quaternions, and spinors—usually taught as three separate mathematical systems—are all manifestations of a single structure: geometric (Clifford) algebra. Starting from Hamilton's 1843 quaternion formula and Schrödinger's puzzling 720-degree spinor rotation, the author shows that the even subalgebra of Cl(2,0) is isomorphic to the complex numbers, with the 'imaginary unit' i identified as the bivector e12 (the oriented area element of the plane). Similarly, the even subalgebra of Cl(3,0) is isomorphic to the quaternions, with i, j, k corresponding to the three coordinate-plane bivectors e23, e13, e12. Spinors arise from the double-cover relationship Spin(n) → SO(n), with SU(2) ≅ Spin(3) ≅ unit quaternions explaining why fermions pick up a minus sign under 360° rotation and need 720° to return to their original state. The post covers rotor formulas, the elimination of gimbal lock, Pauli and Dirac matrices as Clifford algebra representations, applications in computer graphics, quantum computing, robotics, and relativity, as well as Bott periodicity and K-theory connections.

This long-form forum article (approximately 10,500 Chinese characters) presents a popular-yet-rigorous explanation of how geometric algebra unifies complex numbers, quaternions, and spinors. Below is a structured English summary of its content.

Historical prologue

On October 16, 1843, William Hamilton carved the quaternion relations i² = j² = k² = ijk = -1 on Broom Bridge, Dublin.
In 1926, Schrödinger's wave mechanics introduced spinors: odd objects whose wave functions gain a minus sign under a 360° rotation, requiring a full 720° rotation to return to the original state.
The article's thesis: both stories come from the same mathematical structure—geometric (Clifford) algebra, invented by William Clifford in 1878.

Key points

1. Complex numbers = the even subalgebra of Cl(2,0)

With orthonormal basis vectors e₁, e₂ satisfying e₁e₂ = -e₂e₁, the bivector e₁₂ = e₁e₂ satisfies e₁₂² = -1.
The even subalgebra Cl⁺(2,0) = {1, e₁₂} is isomorphic to ℂ; e₁₂ plays the role of i.
Thus i² = -1 is not mysterious: i is the oriented area element of the plane, and squaring it corresponds geometrically to a 180° rotation (multiplication by -1).
Euler's formula becomes a rotation-generator statement: e^(e₁₂θ) = cos θ + e₁₂ sin θ.
Complex multiplication is rotation-scaling; complex conjugation corresponds to the reverse operation.

2. Quaternions = the even subalgebra of Cl(3,0)

Cl(3,0) has basis {1, e₁, e₂, e₃, e₁₂, e₁₃, e₂₃, e₁₂₃}; its even part is 4-dimensional.
Identifying i = e₂₃, j = e₁₃, k = e₁₂ (the three coordinate-plane bivectors) reproduces exactly Hamilton's relations, verified by direct computation (e.g., ijk = -1).
Quaternion multiplication is the geometric product of bivectors, explaining its non-commutativity.
Quaternions live in Spin(3), the double cover of SO(3); the rotation formula qvq⁻¹ is the rotor sandwich product v' = RvR̃.
Gimbal lock disappears because rotors directly represent plane rotations rather than decomposing them into Euler angles; slerp interpolation follows naturally.

3. The 720° mystery of spinors

Spin-1/2 particles obey ψ → -ψ under 360° rotation; only 720° restores ψ.
Topological reason: SO(3) is not simply connected (π₁ = ℤ₂); its double cover SU(2) satisfies SU(2) ≅ Spin(3) ≅ unit quaternions ≅ S³.
In SU(2), R(θ) = e^(Bθ/2): a 360° rotation reaches -1, and only 4π returns to +1—like walking two loops on a Möbius strip. The Dirac belt trick illustrates this physically.
Pauli matrices are a matrix representation of Cl(3,0); Dirac gamma matrices represent Cl(3,1) (Minkowski spacetime).

4. The unified framework

Family tree: Cl⁺(2,0) ≅ ℂ (Spin(2) ≅ U(1)), Cl⁺(3,0) ≅ ℍ (Spin(3) ≅ SU(2)); generally the even subalgebra generates the Spin group.
All rotations in any dimension use one rotor formula: v' = RvR̃ with R = e^(Bθ/2).
Exceptional isomorphisms: Spin(4) ≅ SU(2)×SU(2), Spin(5) ≅ Sp(2), Spin(6) ≅ SU(4).
Parameter efficiency: 3D rotations need 4 quaternion parameters vs. 9 for matrices, with no gimbal lock.

5. Applications

Computer graphics: quaternions give efficient composition (16 vs. 27 multiplications), geodesic slerp interpolation, and numerical stability.
Quantum computing: qubits are spinors; the Bloch sphere is a projection space consistent with the double cover; single-qubit gates are SU(2) rotors.
Robotics/aerospace: strapdown inertial navigation uses quaternion integration for singularity-free attitude tracking.
Relativity: boosts and rotations unify as spacetime rotors in Cl(3,1); the Dirac equation simplifies to ∇ψI = mψ.

6. Deeper connections

Bott periodicity: real Clifford algebras repeat every 8 dimensions (Cl_{n+8} ≅ Cl_n ⊗ ℝ(16)); complex ones every 2 (Cl^ℂ_{n+2} ≅ Cl^ℂ_n ⊗ ℂ(2)).
K-theory: the Atiyah–Bott–Shapiro theorem links Clifford modules to KO-groups.
Spinor bundles: curved-spacetime fermions require the spin condition (vanishing second Stiefel–Whitney class) and spin connections.
Open frontiers: quantum gravity, explaining the Standard Model gauge group SU(3)×SU(2)×U(1) via Clifford algebra, and anomalies.

Conclusion

The article closes with the message: complex numbers are the language of 2D rotations, quaternions of 3D rotations, and spinors of rotations in any dimension—all living in even subalgebras, all tied to the same double-cover structure, all expressible by a single rotor formula. The apparent 720° weirdness of quantum mechanics is revealed as pure geometry.

Core formulas (appendix)

Clifford relation: e_i e_j + e_j e_i = 2η_ij
Even subalgebra isomorphisms: Cl⁺(2,0) ≅ ℂ, Cl⁺(3,0) ≅ ℍ, Cl⁺(p,q) ≅ Cl(q,p-1)
Rotor: v' = RvR̃, R = e^(Bθ/2)
Spin group identifications: Spin(2)≅U(1), Spin(3)≅SU(2)≅Sp(1), Spin(4)≅SU(2)×SU(2), Spin(5)≅Sp(2), Spin(6)≅SU(4)
Bott periodicity: Cl_{n+8} ≅ Cl_n ⊗ ℝ(16); Cl^ℂ_{n+2} ≅ Cl^ℂ_n ⊗ ℂ(2)

References cited

Feynman Lectures Vol. III; Hestenes, *New Foundations for Classical Mechanics* (1999); Doran & Lasenby, *Geometric Algebra for Physicists* (2003); Lounesto, *Clifford Algebras and Spinors* (2001); Atiyah–Bott–Shapiro, "Clifford modules," *Topology* 3 (1964); Hamilton (1844); Dirac (1928); Hestenes, "Real spinor fields," *J. Math. Phys.* 8 (1967).