Corpus: 8 paraphrastic snippets (4 Godot-like dialog passages; 4 Notes-like monologues) with id, work, year, voice, text.
Processing: tokenization; metrics: token count, type-token ratio, naive sentiment (demo lexicon).
Themes: lexicon-based hits for outward (quest) and inward (introspection).
Extra Feature: dialogicity measure (dialog vs monologue) as a proxy for outward vs inward orientation.
Visualisation: per-work charts, theme hits, top keywords, and a global co-occurrence network.