Describing and understanding an ancient language like Inscriptional Tamil is an enormous challenge in many ways. The most trivial among the difficulties is the relation between the text (form) and meaning. Since the 19th century, there has been a regular paradigm shift in linguistic theory and in the study of languages (from philology to structuralism, generative, functional, etc.). This paradigm shift has provided us with a handful of analytical tools, enabling us to view the discourse (written and spoken) from a new perspective; it has also shown us that texts are more than strings of sentences. These developments, among other things, have put greater emphasis on the social, cognitive, communicative and cultural aspects of language. In this paper, we try to examine the interaction between text architecture, information structure and linguistic devices used. We formulate some hypotheses on the relationship between pragmatics, semantic content (meaning) and structure (syntax), which is crucial in interpreting and extracting information from epigraphic texts. The analysis presented here is based exclusively on a corpus of naṭukal “Hero stone” inscriptions dating between 400 and 650 CE.