In the first part of the talk I will recall the duality approach to language recognition. To start with, I will explain the following simple fact. The elements of the syntactic monoid of a regular language $L$ over a finite alphabet $A$ are in one to one correspondence with the atoms of the finite sub-Boolean algebra of $P(A^*)$ generated by the quotients of $L$. This correspondence can be seen as an instance of Stone duality for Boolean algebras, and has lead to a topological notion of recognition for non-regular languages, the so called Boolean spaces with internal monoids.

A fundamental tool in studying the connection between algebraic recognisers, say classes of monoids, and fragments of logics on words is the availability of constructions on monoids which mirror the action of quantifiers, such as block products or other kinds of semidirect products. In the second part of the talk I will discuss generalisations of these techniques beyond the case of regular languages and present a general recipe for obtaining constructions on the topological recognisers introduced above that correspond to operations on languages possibly specified by transducers.

This talk is based on joint work with Mai Gehrke and Luca Reggio.