Geometry of the Lawson-Scott coordinatization of MV-algebras
For any countable MV-algebra A the Lawson-Scott coordinatization process picks some unital dimension group (G,u) such that A coincides with the unit interval [0,u] of (G,u), then picks some Bratteli diagram B of (G,u), and finally constructs the inverse semigroup I(B) having the property that A is isomorphic to the MV-algebra of principal ideals of I(B). One may specialize this construction as follows:
\begin{enumerate}[{\em (i)}]
\item let (G,u) be the uniquely determined unital lattice ordered abelian group corresponding to
A via the categorical equivalence Γ\/ between MV-algebras and unital lattice ordered abelian groups;
\item then let B=B(A) be the uniquely determined direct system of simplicial groups, all with the same unit u,
and unit preserving monotone homomorphisms, sitting inside (G,u).
\end{enumerate}
By Marra ultrasimplicial theorem, lim, U B(A) and (G,u) are isomorphic as unital lattice ordered abelian groups.
Via Elliott classification and its K_0-theoretic refinements, the AF-algebra E(A) given by the direct system B(A) satisfies the identity K_(E(A)) = (G,u).
The Murray-von Neumann order of projections of E(A) is a lattice. To illustrate the geometry of this special coordinatization process we will exemplify steps (i)-(ii) in the
all-important case A when is free, i.e., (by McNaughton theorem), A consists of all continuous
piecewise linear continuous functions f:[0,1]^n\rightarrow [0,1], each linear piece of f having integer coefficients.
Our variant of the Lawson-Scott coordinatization process draws from over 65 years of MV-algebraic theory, including the McNaughton representation of free MV-algebras (1951), Chang completeness theorem MV = HSP([0,1]) (1959), \Gamma functor theory (1986),
and the theory of MV-algebraic Schauder bases and their underlying regular/unimodular triangulations of rational polyhedra in euclidean space (1988?).