We’ll roughly follow the Elephant and Seely. The following is all entirely standard and here only to fix notation and terminology.
A signature is a collection of sorts, together with function symbols and relation symbols , where is a finite list of sorts, also called a type. We’ll denote types by too.
For each sort there should be an inexhaustible supply of fresh variables and bound variables of sort . A context is a finite list of sorted variables, or equivalently a single typed variable .
A term is either a variable, a tuple of terms, or an expression of the form where is a function symbol and a term. The free variables of a term are precisely the variables that appear in it. Each term lives in a context, which must contain every variable in the term, perhaps together with ‘dummy’ variables that don’t. We write to indicate that is the context of , and to denote the obvious substitution. The type of a term-in-context is determined in the obvious way: for variables, , while a tuple if each , and if and .
The formulas of regular logic are built up from the atomic formulas (where the type of is the type of ) and (where and are terms of type ) using (true), (conjunction) and , where if is a formula then so is . The free variables of a formula are defined in the usual way. A context.
The substitution of the term for the variable is defined as usual.
We’ll use the usual natural-deduction rules — conjunction and truth are governed by
existentials by
where on the right is not free in , and equality by
We also have to add the Frobenius axiom , where is not in .
Proofs are considered up to the rewrites given in Seely and Prawitz, together with those making the Frobenius axiom the two-sided inverse to the usual proof of the converse (see e.g. Elephant p. 832).
Models of regular theories are regular fibrations.
If is a regular fibration, then an interpretation of a signature assigns
to each sort an object of
to each list of sorts the product
to each function symbol a morphism in
to each relation symbol an object in (the fibre of over )
Given an interpretation of a signature, terms and formulas over it are given the following meanings:
.
.
, where is the pairing of the meanings of the .
, where is the diagonal.
and are interpreted by the finite products in , of course, and by the left adjoints to product projections .
An interpretation satisfies a sequent , written , if there is a vertical morphism over where is the sort of .
An interpretation is a model of a theory, written , if it satisfies each sequent contained in the theory.
If a theory proves a sequent , then for any model of , .
…