I would take a different perspective that puts as "higher forces" things like:
-- what helps thinking about things in general, about problems, and resolving them (epistemological concerns, which include the whole environment as intrinsic to "langauge")
-- representational matchups to what we are trying to model and create dynamic inference processes for (mathematical concerns -- this is why "mathematics" is a plural, in real math you invent maths when needed ...)
-- orthogonal axes for many areas, including meaning and optimizations, including definitions and meta definitions, debugging, reformulation, etc. (pragmatic concerns for eventually winding up with workable artifacts)
