In this work we put forward an algorithm for the mechanical verification of an extension of Martin-Löf's theory of types with dependent record types and subtyping. We first give a concise description of that theory and motivate its use for the formalization of algebraic constructions. Then we concentrate on the informal explanation and specification of a proof checker that we have implemented. The logical heart of this proof checker is a type checking algorithm for the forms of judgement of a particular formulation of the extended theory which incorporates a notion of parameter. The algorithm has been proven sound with respect to the latter calculus. We include a discussion on that proof in the present work.

Researchers have recently proposed that for certain applications it is advantageous to use functional languages whose type systems are based upon linear logic: so-called linear functional languages. In this paper we develop reasoning techniques for programs in a linear functional language, linPCF, based on their operational behaviour. The principal theorem of this paper is to show that contextual equivalence of linPCF programs can be characterised coinductively. This characterisation provides a tractable method for reasoning about contextual equivalence, and is used in three ways:

[bull ] A number of useful contextual equivalences between linPCF programs is given.

[bull ] A notion of type isomorphism with respect to contextual equivalence, called operational isomorphism, is given. In particular the types !ϕ[otimes ]!ψ and !(ϕ&ψ) are proved to be operationally isomorphic.

[bull ] A translation of non-strict PCF into linPCF is shown to be adequate, but not fully abstract, with respect to contextual equivalence.

A program derivation is said to be polytypic if some of its parameters are data types. Often these data types are container types, whose elements store data. Polytypic program derivations necessitate a general, non-inductive definition of ‘container (data) type’. Here we propose such a definition: a container type is a relator that has membership. It is shown how this definition implies various other properties that are shared by all container types. In particular, all container types have a unique strength, and all natural transformations between container types are strong.