The theory of generalized functions realized as continuous linear functionals on spaces of test functions (e.g., C_c^∞ or the Schwartz space), providing a rigorous setting for singular objects like the Dirac delta and for distributional differentiation used in PDEs and Fourier analysis.

Principle

Extend classical differentiation and linear operations to objects that act on test functions by duality: distributions are elements of the topological dual of a test-function space, with operations defined by continuity and dual pairing rather than pointwise values.

Demonstration

The Dirac delta δ is the distribution that maps a test function φ to φ(0); the distributional derivative of the Heaviside step H satisfies H' = δ; principal value 1/x defines a tempered distribution used in singular integral formulas and Fourier transforms.

Misapplication

Treating distributions as pointwise-defined functions or assuming arbitrary products of distributions exist (for example, multiplying δ by itself without a chosen regularization) leads to contradictions and ill-defined expressions.

Consequence

A robust toolkit for solving linear PDEs with singular sources, defining fundamental solutions and Green's functions, extending the Fourier transform to tempered distributions, and analyzing singularities via wavefront sets.

Reversal

Restricting attention to classical functions, measures, or L^p spaces eliminates singular linear functionals and the ability to represent point sources or perform distributional differentiation; some problems become unsolvable in the classical framework.

Boundary

Applies to linear continuous functionals on specified test spaces; results depend on the chosen test space (compactly supported smooth, Schwartz, etc.); nonlinear operations and arbitrary products require additional structure (Colombeau algebras, microlocal analysis).

Semantic Tension

The term 'generalized function' overlaps with measures, hyperfunctions, and ultradistributions; distributions are distinct by their construction as duals of particular test spaces and by their continuity properties.

Synthesis

Distribution Theory unifies singular objects and classical functions into a dual-space framework that preserves linear operations and differentiation via continuity on test functions, enabling rigorous manipulation of sources and singularities in analysis and PDEs.