Endomorphism monoids in small varieties of bands

From MaRDI portal





A monoid is representable in a class \(\mathfrak K\) of algebras if it is isomorphic to \(\hbox{End}(A)\) for some \(A\in{\mathfrak K}\). The class \({\mathfrak K}\) is said to be monoid universal if all monoids are representable in \({\mathfrak K}\) and finite monoid universal if every finite monoid is isomorphic to the endomorphism algebra of a finite member of \(\mathfrak K\). A variety \({\mathfrak K}\) is universal if the category of all graphs and compatible mappings can be fully embedded into \({\mathfrak K}\). If there is such an embedding which maps finite graphs into finite algebras then \({\mathfrak K}\) is finite-to-finite universal. All known monoid universal varieties are also universal, but it is not known whether this is true in general. In this paper the algebras in question are bands, which admit all constant maps as endomorphisms, so no variety of bands is monoid universal. However, the addition of two or three nullary operations can make even rather small varieties of bands universal. The main results of the paper are: Theorem 1. Neither the variety of rectangular bands nor the variety of semilattices can be made universal by the addition of any number of nullary operations. Theorem 2. A variety \({\mathfrak V}\) of bands with two nullary operations added is universal if and only if \({\mathfrak V}\) contains either the variety of semilattices of left zero semigroups \(([xyx=xy])\) or the variety of semilattices of right zero semilattices \(([xyx=yx]\)). Theorem 3. A variety \({\mathfrak V}\) of bands with three nullary operations added is universal if and only if the variety of semilattices \(([xy=yx])\) is a proper subvariety of \({\mathfrak V}\). (The variety defined within the variety of all bands by the single law \(u=v\) is denoted by \([u=v]\).) The techniques used in the paper actually prove rather stronger results than those stated above. These are summed up in the final Corollary: For a variety \({\mathfrak V}\) of bands with \(k\) nullary operations the following are equivalent: (a) \({\mathfrak V}\) is finite-to-finite universal; (b) \({\mathfrak V}\) is universal; (c) \({\mathfrak V}\) is monoid universal; (d) \({\mathfrak V}\) is finite monoid universal; (e) there exists a finite nontrivial group \(G\) and an algebra \(A\in {\mathfrak V}\) with \(\hbox{End}(A)\simeq G\); (f) either \(k\geq 2\) and \([xyx=xy]\subseteq {\mathfrak V}\) or \(k\geq 2\) and \([xyx=yx]\subseteq{\mathfrak V}\) or \(k\geq 3\) and \([xy=yx]\subset{\mathfrak V}\).











This page was built for publication: Endomorphism monoids in small varieties of bands

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3984099)