Skip to content

SunoikisisDC Summer 2022 Session 4

Jump to bottom
Monica Berti edited this page May 21, 2022 · 14 revisions

Sunoikisis Digital Classics, Summer 2022

Session 4: ToposText and Wikidata

Thursday May 19, 2022, starting at 17:15 CEST (for 90 minutes)

Convenors: Brady Kiesling (ToposText)

Youtube link: https://youtu.be/e79Yl6VMUE8

Slides

Outline

ToposText began as a tool imagined for a specific, limited purpose. Travels in the Greek landscape could and should be enriched with on-the-spot access to the ancient texts that gave that landscape its centrality to a specific intellectual tradition. The Perseus Project at Tufts had compiled a critical mass of Greek and Latin literature in English translations, clean digital texts, under a Creative Commons license. The Pleiades Project had made freely downloadable a huge database of ancient place names with coordinate pairs derived from the Barrington Atlas. My insight in 2012 was that those two datasets could be mashed together to feed ancient texts to users in small, potentially relevant doses. They should be together in an app, one that would work in places with no internet connection. And suddenly, mobile phones had enough memory for this to be possible.

The first step was to make ancient place mentions findable through adequate [i.e., NOT obsessively perfect] indexing. The second was to make the ancient places they described findable as well. In practice, this meant using Google Earth and other resources to geolocate ancient sites with an accuracy that would allow Google Maps to navigate to them successfully. A multi-year treasure-hunt began, still ongoing.

In the process of remedying its major flaws, ToposText has evolved into a more flexible and ambitious tool, a website students can consult to locate obscure ancient sources and quickly test historical hypotheses using primary sources. It aspires to be a source of Wikidata-annotated texts for more advanced analytic work as well.

I was not aware when I began of Digital Humanities as an established discipline with rapidly evolving tools and daunting lists of best practices. Metadata meant nothing to me. Intellectual property legalisms seemed petty at best. If I known what first-year DH students know, I never would have been arrogant enough to attempt ToposText. On the other hand, basic knowledge of DH tools and rules would have saved me hundreds of hours and dozens of false starts.

In describing ToposText and its awkward childhood in our session, I hope to highlight a manageable set of basic software tools, data-manipulation skills, and attitudes needed to transform your body of curated data into a digital tool that will help teach us something.

Seminar readings

ToposText Project

Other resources

https://topostext.org/

Wikidata query examples

Exercise

I.

  1. Download Google Earth Pro unless you already have GIS software you are at ease with.
  2. Download the ToposText Places by place type kmz file from https://topostext.org/TT-downloads.
  3. Open this file in Google Earth and zoom down to your chosen location to check what if anything ToposText has to report on ancient places near you. If ToposText is wrong, write a blistering note to that effect.
  4. Look up an ancient place in a good reference work (Wikipedia or better). Compare any cited ancient references to that place against the list of quotations in ToposText. Identify an ancient reference that ToposText doesn't have, and prepare to share it in the chat along with the ToposText place URL.

II.

  1. Use the Wikidata SPARQL query service [sample queries] to download in your preferred format (csv works well) all the notable monuments OR archaeological sites OR famous artworks, that have Wikidata IDs within bicycling distance of, or within the NUTS level 3 (Greek regional unit, French Department, German Kreis) administrative district of, your home location or favorite place intellectually.
  2. Import this file into a spreadsheet (check for UTF8 encoding) and into Google Earth Pro or your favorite GIS program, and save the result at a kml file (or kmz if you use custom icons). [if objects in museums, use the museum coordinates].
  3. Compare the ToposText and Wikidata maps. If a Wikidata item from above matches closely with a ToposText place, add the ToposText ID (123456ABcd in the URL) to the Wikidata item [Property P8068 "ToposText Place ID"].

III.

  1. Identify a scholarly or at least thorough compendium of the same category of entities in your chosen location, ideally in a well-structured open-access website with robust-seeming URLs and WGS84 coordinate pairs, but a published book if appropriate. Pleiades is a good start for ancient places, but you will want to drill deeper afterwards.
  2. Compile a spreadsheet of six notable-seeming objects that could be in Wikidata but are not, with columns for as many as possible of the Wikidata properties that seem appropriate. Make sure you have either a respectable reference URL or else a reference work that already has its own Wikidata identity [or create one for the reference if you are bold].
  3. Map those objects as well as you can on Google Earth or other GIS, and compare that map against your Wikidata map output to ensure no duplication.

IV.

  1. Create a Wikidata user ID for yourself.
  2. Create a new Wikidata item by hand for your favorite object.
  3. IF YOU FEEL BRAVE ENOUGH after our session is over, use QuickStatements to upload the remaining new items.
Clone this wiki locally