Is there a class of neoclassical compounds, and if so is it productive?

Discussions of English word formation include a category of neoclassical compounds to deal with formations such as geology and telephone. This type is claimed to show behavior that makes it clearly distinct from both affixation and compounding. A close examination of actual words suggests that this picture is oversimplified. Rather than there being a discrete class of neoclassical compounds, there is a range of types of lexical enrichment, and every category has a fuzzy boundary. Accordingly, it is suggested here that it may be useful to view English words as being formed within a conceptual space defined by three dimensions: a simplex-compound dimension, a native-foreign dimension and an abbreviated-nonabbreviated dimension. Neoclassical compounding is a label given to one small section within this three-dimensional space, but actual words diverge from the prototype considerably