Completely representable lattices

Abstract: It is known that a lattice is representable as a ring of sets iff the lattice is distributive. CRL is the class of bounded distributive lattices (DLs) which have representations preserving arbitrary joins and meets. jCRL is the class of DLs which have representations preserving arbitrary joins, mCRL is the class of DLs which have representations preserving arbitrary meets, and biCRL is defined to be the intersection of jCRL and mCRL. We prove CRL is a strict subset of biCRL which is a strict subset of both jCRL and mCRL. Let L be a DL. Then L is in mCRL iff L has a distinguishing set of complete, prime filters. Similarly, L is in jCRL iff L has a distinguishing set of completely prime filters, and L is in CRL iff L has a distinguishing set of complete, completely prime filters. Each of the classes above is shown to be pseudo-elementary and hence closed under ultraproducts. The class CRL is not closed under elementary equivalence, hence it is not elementary.

Review text

The aim of the paper under review is to investigate the possibility of a complete representation for the class of bounded distributive lattices. This problem is more complex than the case of Boolean algebras, since in the absence of Boolean complementation, a representation of a distributive lattice may be complete with respect to one of the lattice operations but not with respect to the other. It is proved that the class \textbf{CRL} of completely representable bounded distributive lattices is not elementary. However, the authors show that it is possible to characterize the complete representability of a lattice in terms of the existence of certain prime filters. Using this characterization, an alternative proof of the identification of the completely representable Boolean algebras with the atomic ones is provided. It is also shown that \textbf{CRL} and the class of bounded distributive lattices that have representations respecting either or both arbitrary infima and suprema are pseudo-elementary, and thus closed under ultraproducts. I think the article contains valuable results and may be a starting point for other studies on this subject.

Recommendations

Cites work

Cited in

(16)

This page was built for publication: Completely representable lattices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q422347)