Does the T/G/P/M family split have a stated domain?

The friction. v4 §4 carried a four-family split — T / G / P / M — forward from v3 §§1–6 as part of the place axis. Its own scope line, added in v4, admits the flaw: "the four families have no stated domain. What space they partition is not answered anywhere."

Why it is a probe and not a coordinate. A coordinate is a structural axis things get placed at; it classifies without absorbing. A partition with no stated domain classifies nothing — there is no space it cuts into four. So T/G/P/M is a candidate coordinate that never earned one: it looks like an axis (four labelled slots) but names no territory. The RUNG-AUDIT split kept the place axis as a genuine coordinate and evicted T/G/P/M here, out of v5, rather than delete the idea outright — the four names may yet be given a domain.

The question, stated so it can be answered or killed. What space do T, G, P, and M partition? Concretely: (a) is there a set of objects such that every object is in exactly one of the four, and (b) is the assignment forced by structure rather than by taste? If (a) and (b) hold, T/G/P/M is a coordinate and returns to the vocabulary with its domain stated. If they cannot be met, it is a mnemonic, not an axis, and stays out.

Stakes. Whether the founding four-family vocabulary is a coordinate or a habit. The place axis survives either way; this is only about the four families riding on it.