Structural typing

Structural typing is a type system paradigm in which type compatibility is determined entirely by the structure and members of types, not by their declared names or inheritance relationships. In a structurally typed system, an object with properties and is automatically compatible with any interface requiring those properties, regardless of whether it was ever declared as implementing that interface. This stands in contrast to nominal typing, the dominant paradigm in languages like Java and C Sharp, where types are compatible only if they share an explicit declaration or inheritance chain.

Structural typing aligns naturally with JavaScript's runtime semantics, where objects are duck-typed: if it walks like a duck and talks like a duck, it is treated as a duck. TypeScript adopted structural typing deliberately, making its static type system a formalization of JavaScript's existing runtime behavior rather than an imposition of foreign discipline. But this alignment comes with costs. Two structurally identical types may carry semantically different meanings — a and a may both have and properties, but swapping them is a category error that structural typing cannot catch. The type system sees shapes; it does not see intentions.

Structural typing is not a failure of type system design. It is an admission that in a world of heterogeneous data sources, runtime monkey-patching, and third-party APIs, the programmer's intent is often less knowable than the data's shape. The question is whether this pragmatism is a necessary compromise or a surrender to the chaos that untyped languages created in the first place.