Janelle Jenstad on Urban Gazetteer Design

Karl Grossner
Pelagios
Published in
3 min readSep 19, 2020

--

On September 9th, the Pelagios Network Gazetteer Activity group hosted an online talk by Dr. Janelle Jenstad of the University of Victoria, titled “Urban Gazetteer Design: Principles for Disambiguating Places and Toponyms.” Dr. Jenstad shared experiences and lessons learned directing development of the Map of Early Modern London project she initiated in 2009, and her talk was followed by an hour or so of fruitful discussion.

Some notes about questions and points raised in the discussion appear below the video and PDF links, and there are links to relevant projects below that.

The video recording here begins with slide 6, a minute or so in. The full PDF slide deck is available below.

PDF slides

Questions and discussion notes

NOTE: Questions and answers are paraphrased and not direct quotes

Q: To what extent are the design principles for disambiguating Places and Toponyms specific to *urban* gazetteers, as opposed to gazetteers in general?
A: I suspect that the need to disambiguate increases as place becomes smaller and more granular. Repeated names would be more common.

Q: How do you deal with places that change? Both the city itself but also spaces within the city, especially districts, buildings, roads, rivers?
A: In MoEML, in many cases this is handled with multiple records for the same location
Comment: Would temporal scoping help? E.g. in Linked Places format, temporal scoping is possible for places as a whole, for geometries, place types, and for relations to other places.

Q: To what extent do you think it is the case that Space-Time <SameAs> Place-Time? E.g. “This was the Thames, but now the city has been built over it”: space becomes different places over time.
A: Space-Time is not <SameAs> Place-Time

Q: Have you faced the issue of verticality within places? that is, do different levels within the same building need to be disambiguated and how to do that?
A: A problem with GIS is it captures x, y, but not a good job of capturing z. London now 14 feet higher than in medieval times. Capture “Feet above sea level?”
Comment: Perhaps a “level filter” for buildings would be useful

Q: Would you be interested in turning MoEML also into a kind of ontology/controlled vocabulary of London place names that other projects could link to and share? (like the pleiades gazetteer for pelagios/peripleo).
A: Absolutely. Note that place data from MoEML is already a gazetteer option within Pelagios’ Recogito annotation platform. Also, MoEML has a published controlled vocabulary, developed since 2005.
Comment: MoEML as linked data is used for the Reading Early Medicine database

Comment thread: Indigenous names for places are often absent in official records; some places are apparently unnamed, but are in fact named
[Bill Pascoe]: Gazetteer of Historical Australian Placenames “hoping to enable decolonisation of the gazetteer by enabling user contributions.”
CF: Reuben Rose-Redwood’s scholarship on toponymy and indigenous placenames (and pseudo-indigenous toponyms like “the Salish Sea”)
CF: the work of John E. Bishop who has done quite valuable research on this question with the ‘James Bay’ Cree in North Central Turtle Island.

Q: Is there a correlation between particular place types and particular problem types? Is it possible to frame toward building best practices?
A: This talk in part frames issues this way; see the slides on challenges of ‘platial’ characteristics

Relevant urban historical project links

--

--