A method for the construction of complete congruences on lattices of pseudovarieties (Q1380059): Difference between revisions

A pseudovariety of semigroups is a class of finite semigroups that is closed under finite direct products and taking divisors (morphic images of subsemigroups). Let \(\mathcal S\) denote the pseudovariety of all finite semigroups. A ``divisor system'' for \(\mathcal S\) selects for each \(S\in{\mathcal S}\) a set \({\mathcal D}(S)\) of divisors of \(S\), such that: (D1) if \(S\in{\mathcal D}(T)\) for some \(T\in{\mathcal S}\), then \(S\in{\mathcal D}(S)\); (D2) each member of \({\mathcal D}(S\times T)\) divides a product \(S'\times T'\), for some \(S'\in{\mathcal D}(S)\), \(T'\in{\mathcal D}(T)\); (D3) if \(S\) divides \(T\) then each member of \({\mathcal D}(S)\) divides a member of \({\mathcal D}(T)\). Let \({\mathcal D}(S)\) be the union of all the members of the divisor system. Then the relation \(D\), defined on the lattice \({\mathcal L}({\mathcal S})\) of pseudovarieties of \(\mathcal S\) by \({\mathcal U}D{\mathcal V}\) if and only if \({\mathcal U}\cap{\mathcal D}({\mathcal S})={\mathcal V}\cap{\mathcal D}({\mathcal S})\), is a complete congruence. For \({\mathcal U}\in{\mathcal L}({\mathcal S})\), the maximum element in its congruence class is the pseudovariety comprising the finite semigroups \(S\) for which \({\mathcal D}(S)\subseteq{\mathcal U}\); the minimum element is also simply described. The concept of divisor system generalizes that of ``radical congruence system'', studied by the author, \textit{T. E. Hall}, \textit{N. R. Reilly} and \textit{S. Zhang} [in Int. J. Algebra Comput. 7, No. 4, 433-455 (1997; Zbl 0885.20036)]. Complete congruences on lattices of pseudovarieties (and varieties) have played fundamental roles in the study of such lattices. The outstanding example was the use by L. Polák of such congruences to find fundamentals results on lattices of varieties of completely regular semigroups. Along with an analysis of the basic properties of divisor systems and their joins, the author enumerates a range of examples, old and new. Some important ones include: (1) the set of all local submonoids of the semigroup; (2) its idempotent-generated subsemigroup; (3) the set of all its regular principal factors. The first two examples motivated much of the early development of these ideas (e.g. by \textit{M. Petrich} and \textit{N. R. Reilly} [J. Aust. Math. Soc., Ser A 49, No. 1, 1-23 (1990; Zbl 0708.20019)]). The general theory holds true for algebras in general. Interesting examples are given for pseudovarieties of non-semigroup type.