Abstract
Involutivity is the algebraic property that guarantees solutions to an analytic and torsion-free exterior differential system or partial differential equation via the Cartan-Kähler theorem. Guillemin normal form establishes that the prolonged symbol of an involutive system admits a commutativity property on certain subspaces of the prolonged tableau. This article examines Guillemin normal form in detail, aiming at a more systematic approach to classifying involutive systems. The main result is an explicit quadratic condition for involutivity of the type suggested but not completed in Chapter IV, §5 of the book Exterior Differential Systems by Bryant, Chern, Gardner, Goldschmidt, and Griffiths. This condition enhances Guillemin normal form and characterizes involutive tableaux.
Introduction
In the study of Exterior Differential Systems (EDS) and overdetermined partial differential equations (PDEs), the property of involutivity acts as a "holy grail." When a system is involutive, the Cartan-Kähler theorem ensures the existence of analytic solutions. Despite its importance, checking for involutivity has historically been a difficult task. While "Cartan’s test" offers a method based on the dimensions of prolonged spaces, it provides limited insight into the internal structure of the equations themselves.
This paper addresses a long-standing gap in the literature—a "missing link" suggested decades ago in the foundational text Exterior Differential Systems (the "Big Green Book"). By deriving an explicit quadratic condition that characterizes involutive tableaux, the author transforms an abstract existence property into a concrete, computable matrix test.
Methods: Endovolutivity and the Quadratic Condition
The author utilizes tools from linear algebra and homological algebra, focusing on tableaux—vector spaces of matrices representing the first-order relations of a differential system. A central innovation is the introduction of endovolutivity. This term describes a tableau expressed in a "generic basis" where symbol maps act as endomorphisms on specific subspaces.
The core of the research involves a rigorous inductive proof. By analyzing the "2-form condition"—the algebraic counterpart to the requirement that second derivatives commute—the author tracks how independent generators are organized within the matrix representation. This leads to the derivation of specific quadratic equations involving commutator-like products of symbol matrices ($B^\lambda_i B^\mu_j - B^\lambda_j B^\mu_i = 0$).
Key Insights: From Dimensions to Commutativity
A major conceptual shift presented in the paper is the realization that involutivity is not merely an abstract property of dimensions. Instead, it is a literal requirement that certain matrix-valued components of the system must commute.
The paper strengthens the classic Guillemin normal form, showing it is a consequence of a much stricter, bidirectional quadratic condition. By choosing a "generic" or "preferred" basis (the endovolutive basis), the underlying algebraic structure of the PDE becomes visible, moving beyond the coordinate-free abstractions that sometimes obscure practical implementation.
Key Results
The primary contribution of this work is Theorem 3.2, which provides the explicit quadratic involutivity criteria. It establishes that a tableau is involutive if and only if a specific set of quadratic relations between its symbol coefficients holds. Supporting this is Lemma 3.1, which proves that every involutive tableau admits a basis that makes it endovolutive, bridging the gap between abstract existence and practical representation.
Furthermore, these results enhance computational feasibility. By reducing involutivity to a set of quadratic equations, the search for new types of solvable PDEs can be automated. To this end, the author has developed an open-source SageMath package called Symbol to implement these algorithms.
Limitations and Future Work
While the paper provides a powerful new tool, some challenges remain. The quadratic condition relies on the system being in a "generic basis." Although the paper proves such a basis always exists for involutive systems, finding it for an arbitrary system of equations is non-trivial. Additionally, while the variety of involutive tableaux is now better identified, the total dimension of the space of all involutive tableaux with fixed Cartan characters remains an open question. Future research may also explore the behavior of these systems under non-Borel linear transformations.
Summary updated