Shakespeare's World – I thought this would be simple but
Recorded: May 30, 2026, 9 p.m.
| Original | Summarized |
Shakespeare’s World – I thought this would be simple but … – KnowWhere
Strategy and LocationHomeBlogAboutAbout KnowWhereAbout meMap GalleryVibe CodingMaking MapsOld Web Maps that I MadeThe Good, the Bad and the UglyHistoric Map LinksContact Search Search …
Strategy and Location Search Search …HomeBlogAboutAbout KnowWhereAbout meMap GalleryVibe CodingMaking MapsOld Web Maps that I MadeThe Good, the Bad and the UglyHistoric Map LinksContactHome » GIScussions » Vibe Coding » Shakespeare’s World – I thought this would be simple but … Published May 18, 2026Last week I had a call with Professor Gavin Hollis who is writing about Shakespeare’s use of maps and coining the term mapp’ry – you can read a bit more about our conversation here. That conversation prompted me to think about Shakespeare’s references to places in his plays and what that might tell me about his understanding of the world at the end of the 16th century. I thought this would be quite simple but of course the devil is in the detail!I started by downloading the Project Gutenberg complete works text file and then with a lot of help from Claude used spaCy NER (Named Entity Recognition) to extract candidate place names — produced 578 candidates requiring manual review, I manually reviewed and approved 288 places, added countries and then geocoded them using the OpenCage API and finally made manual coordinate fixes for ancient/mythological places (Ilium, Barbary, Corioles, Belmont).Once I had a list of places referred to in Shakespeare’s plays I needed to extract the quotes with play names and act/scene references. This was challenging to say the least, some place names are also character names (particularly in the English historical plays), I needed to ignore dramatise personae sections and also distinguish scene settings from quotes. I ended up with 2,685 quotes and 153 scene settings across 288 places and 38 plays.Having built a number of maps with MapLibre and Claude’s help I thought the map build would be easy but I had the neat idea to use a quill symbol as a map marker – hours of wasted effort! I don’t really understand why this would not work, when I decided to scrap the quill and use a standard circle symbol it just worked straight away. Later on I changed to a teardrop symbol with no problem. The rest of the map build was relatively straight forward although striving for very good (forget perfection) burnt some time. For this map I wanted a Shakespearian feel so I used Stamen’s Watercolour tiles via Stadia Maps, I think they look really nice, I added a black and white option as well.As I tested, I kept discovering glitches in the data which I had to work through with a combination of python scripts, courtesy of Claude, and manual edits which were easier than solving edge cases in a script. The funniest of errors was Maidenhead – spaCy identified it as a place with 14 references but when I looked at the quotes they were all Shakespeare referring to virginity rather than a place!I am getting better at this stuff but this map was much tougher than I had expected, mainly because of the data. I am pleased with the end result and I think it works pretty well. I particularly like the feature to search for a place or a play, if you select a play the map filters just the places mentioned in that play and zooms to its extents, you can then explore a sample of the quotes mentioning a place. I am sure you will find some humorous mistakes in place and quote extraction, send them to me and I will try to fix.View Shakespeare’s WorldYou may also like Published January 15, 2026 Published March 19, 2026 Published January 13, 2026 Published February 9, 2026 Search Search …CategoriesBusiness Powered by WP – Designed with the Customizr theme |
The process detailed in the article outlines an endeavor to map Shakespeare's references to geographical locations to infer his understanding of the world in the late sixteenth century, prompted by a conversation with Professor Gavin Hollis regarding Shakespeare's use of maps and the term mappery. The methodology involved several complex steps to transition textual references into geocoded map data. Initially, the project involved downloading the complete works text file from Project Gutenberg, followed by the application of spaCy Named Entity Recognition to extract potential place names, which yielded 578 candidates requiring subsequent manual review and approval; countries were also added to this list. These names were then geocoded using the OpenCage API, necessitating manual coordinate fixes for mythological or ancient locations such as Ilium, Barbary, Corioles, and Belmont. To facilitate deeper analysis, the process expanded to extracting textual data, requiring careful scrutiny to distinguish place names from character names found within the plays, particularly in historical works, which necessitated filtering out dramatise personae sections. This phase resulted in the extraction of 2,685 quotes and 153 scene settings associated with the 288 identified places and 38 plays. The subsequent visualization utilized MapLibre alongside assistance from Claude. The mapping itself involved experimentation with symbolic representation, initially testing a quill symbol before settling on a standard circle and later a teardrop symbol, while utilizing Stamen’s Watercolour tiles via Stadia Maps to achieve an aesthetically pleasing, Shakespearian visual style. The execution was fraught with data challenges, including the discovery of inconsistencies that required manual remediation alongside programmatic correction. A notable error occurred when spaCy identified Maidenhead as a place with fourteen references, when in reality, the references related to Shakespeare's discussions of virginity rather than a geographical location. Resolving these edge cases and glitches required the use of Python scripts, augmented by Claude, in conjunction with manual edits, demonstrating that data quality was the most significant hurdle. Ultimately, the resulting map features an interactive search capability, allowing users to filter the map based on a specific play, focusing the view on mentioned locations and displaying associated quotes. The experience underscores the difficulty of accurately handling historical textual data to produce precise geospatial representations. |