Chapter 49: Exterior Algebra and Grassmannians¶
Prerequisites: Multilinear Algebra (Ch21) · Determinants (Ch03) · Subspaces and Bases (Ch04)
Chapter Outline: From Multilinearity to Antisymmetry → Definition and Axioms of the Exterior Product (\(\wedge\)) → Exterior Algebra (Grassmann Algebra) \(\Lambda(V)\) → Basis Construction and Dimension \(\binom{n}{k}\) → The Exterior Origin of Determinants → Decomposable Tensors and their Correspondence to Subspaces → Algebraic Description of the Grassmannian Manifold \(Gr(k, V)\) → Plücker Coordinates and Plücker Relations → Applications: Differential Forms, Electromagnetic Field Tensors, Subspace Clustering, and Projective Geometry
Extension: Exterior algebra is the natural language for describing "oriented volumes"; by introducing antisymmetry, it treats subspaces themselves as algebraic objects. It is the core of modern differential geometry, general relativity, and network flow analysis in theoretical computer science.
In classical algebra, we study combinations of vectors. In Exterior Algebra, we elevate our perspective to "parallelepipeds spanned by vectors." By introducing the Exterior Product (wedge product) \(\wedge\), we can handle areas, volumes, and their generalizations in a purely algebraic manner. This theory ultimately leads to the Grassmannian—the manifold of all \(k\)-dimensional subspaces—providing the ultimate framework for "subspace operations" in operator spaces.
49.1 The Exterior Product and Algebra¶
Definition 49.1 (Exterior Product)
The exterior product \(v \wedge w\) of vectors \(v, w \in V\) satisfies: 1. Bilinearity: \((av_1+bv_2) \wedge w = a(v_1 \wedge w) + b(v_2 \wedge w)\). 2. Antisymmetry: \(v \wedge w = -w \wedge v\). 3. Nilpotency: \(v \wedge v = 0\).
Exterior Algebra \(\Lambda(V)\)
The direct sum of all \(k\)-th exterior powers \(\Lambda^k(V)\) forms the exterior algebra. Its total dimension is \(2^n\) for an \(n\)-dimensional space.
49.2 Geometric Meaning and Subspaces¶
Theorem 49.1 (Linear Independence Criterion)
The set \(\{v_1, \ldots, v_k\}\) is linearly independent iff \(v_1 \wedge v_2 \wedge \cdots \wedge v_k \neq 0\). Geometric Intuition: A non-zero \(k\)-th order wedge product represents a "subspace fragment" with a specific orientation and \(k\)-dimensional volume.
49.3 Grassmannians and Plücker Coordinates¶
Definition 49.2 (Grassmannian)
\(Gr(k, V)\) is the set of all \(k\)-dimensional subspaces of \(V\). Each \(k\)-dimensional subspace can be uniquely represented (up to a scalar) by the exterior product of its basis vectors \(v_1 \wedge \cdots \wedge v_k\) (a decomposable tensor).
Plücker Coordinates
Expanding \(v_1 \wedge \cdots \wedge v_k\) in the basis of \(\Lambda^k(V)\) yields coefficients called Plücker Coordinates. They satisfy a set of quadratic equations known as Plücker Relations.
Exercises¶
1. [Basics] Calculate \((e_1 + e_2) \wedge (e_1 - e_2)\).
Solution
Steps: 1. Apply distributivity: \(e_1 \wedge e_1 - e_1 \wedge e_2 + e_2 \wedge e_1 - e_2 \wedge e_2\). 2. Apply nilpotency: \(e_1 \wedge e_1 = 0\) and \(e_2 \wedge e_2 = 0\). 3. Apply antisymmetry: \(e_2 \wedge e_1 = -e_1 \wedge e_2\). 4. Substitute: \(0 - e_1 \wedge e_2 - e_1 \wedge e_2 - 0 = -2 e_1 \wedge e_2\). Conclusion: The result is \(-2 e_1 \wedge e_2\).
2. [Dimension] If \(\dim V = 4\), find the dimension of \(\Lambda^2(V)\) and list the basis.
Solution
Calculation: 1. Dimension is \(\binom{4}{2} = \frac{4 \times 3}{2 \times 1} = 6\). 2. Let the basis of \(V\) be \(\{e_1, e_2, e_3, e_4\}\). Basis: \(\{e_1 \wedge e_2, e_1 \wedge e_3, e_1 \wedge e_4, e_2 \wedge e_3, e_2 \wedge e_4, e_3 \wedge e_4\}\).
3. [Determinant] Prove that the determinant of an \(n \times n\) matrix is the wedge product of its columns.
Solution
Algebraic Mapping: 1. Consider \(v_1 \wedge v_2 \wedge \cdots \wedge v_n\). 2. Substitute \(v_j = \sum a_{ij} e_i\) and expand using antisymmetry. 3. Terms are non-zero only when the indices are a permutation of \(\{1, \ldots, n\}\). 4. The coefficient of each term is the sign of the permutation \(\operatorname{sgn}(\sigma)\). Conclusion: \(v_1 \wedge \cdots \wedge v_n = \det(A) (e_1 \wedge \cdots \wedge e_n)\). This reveals the most fundamental definition of the determinant: it is the scaling factor of the top-level exterior power.
4. [Decomposability] Determine if \(v = e_1 \wedge e_2 + e_3 \wedge e_4\) is decomposable (i.e., can be written as \(u \wedge w\)).
Solution
Plücker Criterion: 1. in 4D, a 2nd-order exterior element is decomposable iff \(v \wedge v = 0\). 2. Compute \(v \wedge v = (e_1 \wedge e_2 + e_3 \wedge e_4) \wedge (e_1 \wedge e_2 + e_3 \wedge e_4)\). 3. Expand: \(e_1 \wedge e_2 \wedge e_1 \wedge e_2 + e_1 \wedge e_2 \wedge e_3 \wedge e_4 + e_3 \wedge e_4 \wedge e_1 \wedge e_2 + e_3 \wedge e_4 \wedge e_3 \wedge e_4\). 4. Repeated terms are zero, leaving \(2 e_1 \wedge e_2 \wedge e_3 \wedge e_4\). Conclusion: Since \(v \wedge v \neq 0\), the element is not decomposable. It does not represent a single 2D subspace but a superposition of them.
5. [Property] Prove: If \(v_1, \ldots, v_k\) are linearly dependent, then \(v_1 \wedge \cdots \wedge v_k = 0\).
Solution
Proof: 1. If dependent, at least one vector is a combination of others, say \(v_1 = \sum_{i=2}^k c_i v_i\). 2. Substitute into the wedge product: \((\sum c_i v_i) \wedge v_2 \wedge \cdots \wedge v_k\). 3. Every term in the sum contains a repeated vector (e.g., \(c_2 v_2 \wedge v_2 \wedge \cdots\)). 4. By nilpotency, every term is zero.
6. [Grassmannian] To what geometric object does \(Gr(1, V)\) correspond?
Solution
Conclusion: The Projective Space \(P(V)\). \(Gr(1, V)\) represents all lines through the origin. Geometrically, this is the definition of a projective space.
7. [Plücker Relation] Write the single Plücker relation for \(Gr(2, 4)\).
Solution
Formula: Let the coordinates be \(p_{ij}\). \(p_{12}p_{34} - p_{13}p_{24} + p_{14}p_{23} = 0\). This is the algebraic equation identifying which 6D vectors represent 2D subspaces.
8. [Duality] What is the relationship between \(\Lambda^k(V)\) and \(\Lambda^{n-k}(V)\)?
Solution
Conclusion: They are isomorphic. Reasoning: Their dimensions are equal \(\binom{n}{k} = \binom{n}{n-k}\). In a space with an inner product, this correspondence is given by the Hodge Star operator (\(\star\)).
9. [Basics] In \(\mathbb{R}^3\), how does \(v \wedge w\) relate to the cross product \(v \times w\)?
Solution
Connection: \(v \wedge w\) is a 2nd-order tensor (representing an oriented area element) in \(\Lambda^2(\mathbb{R}^3)\). \(v \times w\) is a vector in \(\mathbb{R}^3\). Through the Hodge dual in 3D, area elements map uniquely to their normal vectors. Thus, the cross product is the dual manifestation of the wedge product in 3D.
10. [Application] Briefly state the application of exterior algebra in electromagnetism.
Solution
In relativistic electrodynamics, the electric and magnetic fields are unified into a single 2nd-order antisymmetric tensor (the Faraday tensor \(F\)), which is an element of the exterior algebra of 4D spacetime. Maxwell's equations can be written as \(dF = 0\) and \(d{\star F} = J\), beautifully demonstrating the natural advantage of exterior algebra in describing flux and circulation.
Chapter Summary¶
Exterior algebra is the algebraic culmination of geometric intuition:
- Algebraization of Subspaces: Through the wedge product, we transform abstract "subspaces" into concrete "algebraic elements," realizing the jump from studying points to studying fragments of space.
- Operatorization of Volume: The origin and properties of determinants are most thoroughly explained within the exterior algebra framework, proving that antisymmetry is the algebraic essence of multi-dimensional measure theory.
- Foundations of Manifolds: The Grassmannian manifold and its Plücker coordinates provide fine-grained local characterization tools for modern geometry and topology, serving as a vital link between pure algebra and modern physics (e.g., string theory).