Nickspeare - Generate your Shakespearean pseudonym

Nickspeare is a compact computational-literary project for generating Shakespearean-flavoured nicknames from traceable textual evidence. Instead of asking a language model merely to “sound Shakespearean”, it turns literary atmosphere into explicit rules, reproducible transformations and inspectable provenance.

Its central conceit is the movement from Eastcheap to Agincourt: tavern vocabulary associated with Falstaff and his circle is fused with royal or martial language drawn from Henry V. Generated names are treated as new coinages, while their ingredients remain tied to identifiable textual sources.

The project is modular. The lexicon defines editorial categories such as tavern, royal and archaic vocabulary. The corpus layer parses Folger Shakespeare XML, excludes stage directions and speaker labels, and trains character- and word-level Markov models. The combinators implement rogue-to-king fusion, quotation splicing, archaic coinage and Markov generation. The phonetic layer performs conservative orthographic blending. Provenance records source plays, speakers, line labels, XML identifiers, blend candidates, source hashes, dates and transformation traces. The generator coordinates these parts into a reproducible pipeline.

The principal challenge during vibe coding was preventing literary plausibility from becoming fabricated scholarship. Nickspeare therefore separates editorial interpretation from corpus evidence, attested ingredients from invented outputs, and historical dates from decorative numbers. Its scores and speaker groups are presented as design heuristics and editorial lenses, not objective linguistic truths.

The browser interface follows the same principle: it evokes an Early Modern composing room, but openly displays seeds, scores, probabilities, source IDs and exportable provenance.

Nickspeare is therefore less a novelty generator than an experiment in disciplined vibe coding: use generative models freely for invention, but make deterministic code, versioned data, tests and provenance responsible for factual claims.