The third panel of the Fourth Triptych asks which dependences explain, and when an embedded observer thereby understands. Paper 16 specified individual difference-making; Paper 17 specified which structures stay invariant across admissible differences. This paper takes those outputs as separate, mode-specific inputs - type-level and local difference-making and productive mediation for the causal mode, lawlike standing for the nomological mode, and structural invariance or constitutive constraint for the structural mode - and asks which of them answer a declared contrastive why-question. The signature commitment is the denial of a global narrator: no recounter outside the history explains it by survey, and no external standard fixes which contrast is relevant. World-side contrast-answering dependences may obtain unobserved, but every explanatory representation, act, attribution, and claim to understanding is history-internal and record-mediated. The paper builds two parallel tracks joined by one bridge: an explanation specification leads to a typed explanatory verdict and explanatory entitlement, while an uptake specification leads to an understanding profile and entitlement to claim understanding. The two tracks cross only where explanatory entitlement becomes an input to understanding entitlement. The verdict is a typed package rather than a scalar, and understanding is adjudicated on its own profile and entitlement, never as support for attributing an explanation. The central rejections are that explanatory success is not truth, understanding is not explanation, and explanation is not record compression. The results are conditional on a declared history-internal specification. The paper provides a discipline for adjudicating explanatory claims and for bounding inference to the best explanation through a status-valued non-self-undermining condition; it does not derive the uniquely correct explanations, impose a unique explanatory hierarchy, or provide a psychology of why understanding feels illuminating.
Tomoyuki Uchida (Wed,) studied this question.