Mapp’ry: Building an Interactive Map of Shakespeare’s World

Inspired by the concept of 'mapp’ry,' the author created an interactive map visualizing every geographical location mentioned in Shakespeare's plays. The process involved using AI to extract data from the complete works, followed by extensive manual cleaning to resolve ambiguities between places and characters. The final product allows users to explore the Bard's world through a searchable, thematic interface featuring thousands of quotes.

Key Points

• The project utilized AI and NLP tools to extract 288 unique locations and over 2,600 quotes from Shakespeare's complete works.
• Data cleaning was the most challenging aspect, requiring manual intervention to separate geographical locations from character names and linguistic puns.
• The map uses the OpenCage API for geocoding and features a specialized 'Watercolour' visual style to match the historical theme.
• The interactive tool allows users to filter by specific plays, automatically zooming to the relevant geographic extents of that work.
• The author highlights the iterative nature of the build, involving Python scripts for data glitches and manual fixes for ancient or mythological places.

Sentiment

The overall sentiment is mixed-positive: commenters like the idea and call the map interesting, but they are more engaged by its methodological limits than by simple praise. HN broadly accepts the article's point that the project is harder than it first appears, and the strongest disagreements focus on whether the automated workflow is reliable enough without deeper human scholarship and historical context.

In Agreement

• The map is interesting and useful as an exploratory way to see the geographical range of Shakespeare's references.
• The project illustrates how Shakespeare's references can change a reader's sense of the world known to him and his audience.
• Some far-flung references, including places such as Ethiopia, India, and the Indies, are clear enough to support the map's broader premise.
• The map's limitations are understandable because the task involves complex literary language, historical geography, and metaphor rather than simple place lookup.

Opposed

• LLM-assisted extraction appears to miss important references, with Bermuda in The Tempest raised as a notable example.
• Modern geocoding can misrepresent historical meaning when broad labels such as America, Asia, or Russia are placed at contemporary centroids rather than where Shakespeare's audience likely imagined them.
• A traditional concordance or subject index may provide a more reliable foundation than automated entity extraction alone.
• The map would be more useful with contextual notes about what contemporary audiences would have known or associated with each place.
• Generated maps can become misleading when the maker lacks enough domain knowledge to verify the geography and source material.