Convex elements in lattice–ordered rings and modules

To expose the results in this paper recall some notions which are not so common. For any subset \(X\) in an (additively written) partially ordered group (po-group) \(A\) denote by \(X_+\) its positive cone, i.e. \(X_+= \{x\in X\mid x\geq 0\}\). If there holds the condition \(A_+ R_+\subseteq A_+\) for a right module \(A_R\) over a po-ring \(R\), then \(A\) is called a po-module. If \(F\) is a commutative and unital po-ring with \(1\in F_+\), then a po-algebra is a po-ring \(R\) which is an \(F\)-algebra and a po-\(F\)-module. If a po-module \(A\) (a po-ring \(R\)) happends to be a lattice, then it is called an \(l\)-module (\(l\)-ring). An \(l\)-module \(A\) is called an \(f\)-module if \(a\wedge b=0\) in \(A\) and \(r\in R_+\) imply \(ar\wedge b= 0\). A po-ring \(R\) is called an \(f\)-ring if both \({_RR}\) and \(R_R\) are \(f\)-modules. In this paper, the author is concerned with the characterization of convex elements in po-modules and po-rings. Here, a positive element \(y\) in a po-module \(A\) is called convex if for any \(x\in A\), \(0\leq x\leq y\), there holds \(x= yr\) for some positive \(x\in R\). A po-module \(A\) is called convex if all its positive elements are convex. A po-ring \(R\) is called right-convex if the module \(R_R\) is convex. The author starts with the observation that in \textit{M. Rieffel's} characterization of real semigroup (group) algebras [Pac. J. Math. 16, 347-363 (1966; Zbl 0142.26103)] convex elements appear as semigroup (group) elements. Here Rieffel's results are extended to the general setting of locally finite semigroups. In more detail, for a semigroup \(S\), the group \(\mathbb{R}_+\) (of positive reals) and a 2-cocyle \(f: S^2\to\mathbb{R}_+\), denote by \(\mathbb{R}^f[S]\) the twisted semigroup \(l\)-algebra defined as the \(f\)-module \(\bigoplus_{s\in S}\mathbb{R}\overline s\) together with multiplication on it given by \(\overline s\overline t= f(s,t)\overline{st}\). The algebra \(\mathbb{R}^f[S]\) is locally finite iff \(S\) is locally finite. The author proves (Theorem 2) that for a locally finite semigroup \(S\) the following facts hold: (a) If the 2-cocycle \(f\) satisfies the condition \(f(s^n, s^m)= 1\) for all \(s\in S\) and \(m,n\in\mathbb{N}\), then \(f\) is fundamental; this last means that the conditions \((*)\) \(\forall s,t\in S\), \(f(s,t)= f(t,s)\) and \((**)\) \(st= ts\Rightarrow f(s,t)= 1\) hold; (b) The group \(Z^2(S,\mathbb{R}_+)\) of all cocycles coincides with the direct product of the subgroup \(H\) of fundamental 2-cocycles and the group \(B^2(S,\mathbb{R}_+)\) of 2-coboundaries (c) If \(S\) is commutative then the second cohomology group \(H^2(S,\mathbb{R}_+)= Z^2(S,\mathbb{R}_+)/ B^2(S,\mathbb{R}_+)\) of \(S\) with coefficients in \(\mathbb{R}_+\) is trivial. The author proves also (Theorem 3) that a locally finite commutative \(l\)-algebra \(R\) is a semigroup \(l\)-algebra iff \(R\) is finitely valued, Archimedean and with its set of special elements multiplicatively closed. Here, a value of a given element \(x\) in an \(l\)-group \(A\) is a convex \(l\)-subgroup that is maximal with respect to the property of not containing \(x\); single-valued positive elements in \(A\) are called special. Among the results in Section 2, we mention those on Ore rings. It is proved (Theorem 8) that for a prime \(l\)-ideal in a right convex unital \(f\)-ring \(R\) the set \(\Sigma= R\backslash P\) is a right Ore set in \(R\) and \(R\) itself is a right Ore ring; here, right convexity of \(R\) means the \(R_R\) is convex. A similar result holds for a right \(p\)-convex \(F\)-algebra \(R\) having bounded inversion; this last means that each element in \(R\) that exceeds \(1\) \((\geq 0)\) is both left and right convex. Here, right \(p\)-convexity of an element \(y\in R_+\) means that for \(0\leq x\leq y\) in \(R\) there exist \(r\in R_+\) and a polynomial \(0\neq p(t)\in F_+[t]\) so that \(xp(y)= yr\). The author extends to the noncommutative setting \textit{J. Martinez} and \textit{S. Woodward's} characterization [Commun. Algebra 20, 2975-2989 (1992; Zbl 0766.06018)] of Prüfer \(f\)-rings with bounded inversion, and gives a new, direct proof of it. Further, denote \(|y|= y\vee (-y)\). For a subset \(X\) in an \(l\)-module \(A\) the set \(p(X)= \{a\in A\mid|a|\wedge|x|= 0\) for all \(x\in X\}\) is called the polar of \(X\). The author proves (Theorem 12) that for a semiprime \(f\)-ring \(R\) with a.c.c. on polars to be right convex is equivalent for \(R\) to have an essential right ideal that is a convex po-module. If \(R\) is totally ordered then \(R\) is right convex iff it has a nontrivial nonsingular convex right po-module.

Recommendations

Cites work

Cited in

(6)

This page was built for publication: Convex elements in lattice–ordered rings and modules

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4361104)