It is now notorious that the production of inscriptions in the Roman Empire was not constant over time, but rose over the first and second centuries A.D. and fell in the third. Ramsay MacMullen pointed this out more than five years ago, with conclusions more cautionary than explanatory: ‘history is not being written in the right way’, he said, for historians have deduced Rome's decline from evidence that–since it appears only epigraphically–has merely disappeared for its own reasons, or have sought general explanations of decline in theories political, economic, or even demographic in nature, none of which can, in turn, explain the disappearance of epigraphy itself. Why this epigraphic habit rose and fell MacMullen left open to question, although he did postulate control by a ‘sense of audience’. The purpose of this paper is to propose that this ‘sense of audience’ was not generalized or generic, but depended on a belief in the value of romanization, of which (as noted but not explained by MacMullen's article) the epigraphic habit is also a rough indicator. Epitaphs constitute the bulk of all provincial inscriptions and in form and number are (generally speaking) the consequence of a provincial imitation of characteristically Roman practices, an imitation that depended on the belief that Roman legal status and style were important, and that may indeed have ultimately depended, at least in North Africa, on the acquisition or prior possession of that status. Such status-based motivations for erecting an epitaph help to explain not only the chronological distribution of epitaphs but also the differences in the type and distribution of epitaphs in the western and eastern halves of the empire. They will be used here moreover to suggest an explanation for the epigraphic habit as a whole.