Tuesday, September 10, 2019

The Saints of Europe

I was so fascinated by Dirk Kloosterboer's fascinating investigation into the spatial distribution of roads named 'Holleweg' (or 'Holloway') around the world that I decided to try my own investigation into the spatial distribution of place-names. For my analysis I wanted to explore the density and distribution of towns and cities in Europe named for Christian saints.

As a work of data visualization my finished map should be taken with a piece of salt. Soon after starting my investigation I realized my lack of language skills was seriously going to impair my research into towns named for saints across Europe. However despite the fact that my map is next to useless I still think the process of making it is quite interesting.

If you want to view the finished map then you can find it here - The Saints Of Europe.

All the data from my map comes from OpenStreetMap. To get the data I used Overpass Turbo. I queried Overpass Turbo to find towns and cities which included the words 'St', 'Saint', 'San', 'Santo', 'Santa' and 'Sankt'. You can see how the query is formed in the example below:
// gather results
node[place=town] ["name"~"Santo "]({{bbox}});
node[place=city] ["name"~"Santa "]({{bbox}});
node[place=city] ["name"~"Santo "]({{bbox}});
node[place=town] ["name"~"Santa "]({{bbox}});
node[place=city] ["name"~"San "]({{bbox}});
node[place=town] ["name"~"San "]({{bbox}});
// print results
out body;
out skel qt;

Rather than search for specified countries I restricted my search to rough bounding boxes around the UK, France, Spain, Germany and Italy. I wanted to restrict my search to these countries but I was also interested in capturing any towns and cities named for saints in areas closely bordering these countries as well. Therefore rather than querying these countries individually by name I used a bounding box for searching each country. These bounding boxes overlapped other countries to capture towns in other countries near the country border.

I carried out five separate searches in Overpass Turbo (one each for the UK, France, Spain, Germany and Italy. I downloaded the results of each search as a GeoJSON file. I then combined the five separate files into one GeoJSON file using geojson.io.

Once I had one GeoJSON file containing all the towns and cities named for saints I saved it as a tileset in Mapbox Studio. The data downloaded from Overpass Turbo also included the population of every town and city named for a saint. When you load a tileset into the Mapbox Studio style editor you can style how the data appears on the map. I decided to show each town as circles with the radius of the circles determined by the population size of the town or city. Therefore on my finished map the size of the circles represents the size of the population.

Although I wouldn't recommend my map as an effective data visualization I do find it interesting that the Protestant countries of Germany and the UK appear to have far fewer towns named for saints than the Catholic countries of Italy, France and Spain. Germany in particular seems to have very few towns named for saints.


Anonymous said...

In the official German Statistics, Statistisches Bundesamt, Gemeindeverzeichnis, https://www.destatis.de/DE/Themen/Laender-Regionen/Regionales/Gemeindeverzeichnis/Administrativ/Archiv/GVAuszugQ/AuszugGV3QAktuell.xlsx?__blob=publicationFile
there are 33 independant cities/villages in Germany containing Sankt or St. in there name.
The problem that your are finding only 6, is probably due to the fact the the place node is not very common in Germany.
With the overpass query
// gather results
relation["boundary"="administrative"][name~"Sankt "][admin_level=8][area.searchArea);
relation["boundary"="administrative"][name~"St\. "][admin_level=8](area.searchArea);
// print results
out center;
you will find all of those.
Removing [admin_level=8] will produce 82 including suburbs etc.

rmikke said...

Here, have a list of Polish localities named after saints. Only independent location (not parts of some town).