Notes on locally order-polynomially complete lattices (Q2577724)

Let \((L,\vee,\wedge)\) be a lattice. A \(k\)-ary function \(p: L^k\to L\) (where \(k\) is a positive integer) is called a polynomial if \(p\) can be represented by a term containing only the lattice operations \(\vee\), \(\wedge\), function variables \(x_1,\dots,x_k\), and constants from \(L\). Furthermore, a function \(f: L^k\to L\) is called a local polynomial if, for every finite subset \(M\subseteq L^k\), there exists a polynomial \(p\) such that \(f\) and \(p\) agree on \(M\). The lattice \((L,\vee,\wedge)\) is called (locally) order polynomially complete if all order-preserving finitary functions on \(L\) are (local) polynomial functions. There are two crucial papers by \textit{R. Wille} on this topic [Arch. Math. 28, 557--560 (1977; Zbl 0357.06006) and Houston J. Math. 3, 593--597 (1977; Zbl 0397.06007)]. The present paper contains several supplements to Wille's results, in particular to those in the second paper. For example, it is shown that a lattice \((L,\vee,\wedge)\) is locally order-polynomially complete if and only if it is tolerance simple (which means that every reflexive, symmetric and compatible binary relation on \(L\) is trivial).