Should spatialCoverageDescription be required?

[ptgolden]

Some entries in the dataset do not currently have a value for spatialCoverageDescription, which (if I understand correctly) is the spatial coverage explicitly defined within the source (ie not added by the curator)

[atomrab]

We used this to describe spatial coverage that couldn't be immediately translated into a URI for a modern administrative division. If the source explicitly said "Israel" or "France" or "Sicily" or anything else resolvable against a modern gazetteer, we put that information here (like "Mesopotamia" or "Northern Etruria") to indicate that we had to do some parsing to get our spatial coverage. If it's more consistent to have this for all entries, all the blanks should just repeat the spatial coverage value.

[rybesh]

Isn't it possible that the source says nothing explicit about spatial coverage but that it can be inferred (e.g. given the topic/focus of the source)? In which case it's possible that there is no spatialCoverageDescription but there are spatialCoverage entities chosen by the curator.

[atomrab]

Yes -- for instance, a book on "Iron Age France" or "the Bronze Age Aegean". This is how we did it in the spreadsheet. So I guess that undermines my statement that you can copy spatialCoverage to spatialCoverageDescription, because the author may never have explicitly said "and I'm talking about France".

[ptgolden]

Here's my proposed answer to my provocatively-posed question, combining both of your comments

1. A spatial coverage description should be the actual description included in the source
2. The value of spatial coverage is a curatorially-derived group of linked data entities that have associated shapes.
3. If the source did not include a description, but its spatial coverage is obvious, the spatial coverage description field will remain blank, but the spatial coverage will be appropriately filled

(2) applies even when (1) can be mapped to a modern administrative division. So, if the source says a time period covers "Spain", the spatial coverage will still have an entry for "Spain" in geonames (or dbpedia, or whatever).

This is similar to our approach with start/stop dates: a textual label directly from the source, and derived structured data that estimates the natural language for use in a computer interface.

[atomrab]

Yes, I think this is more or less what we've already been doing, but
without the clear and explicit rationale. Thanks!

On Thu, Mar 12, 2015 at 1:00 PM, Patrick Golden notifications@github.com
wrote:

Here's my proposed answer to my provacatively-posed question, combining
both of your comments

1. A spatial coverage description should be the actual description
   included in the source
2. The value of spatial coverage is a curatorially-derived group of
   linked data entities that have associated shapes.
3. If the source did not include a description, but its spatial
   coverage is obvious, the spatial coverage description field will remain
   blank, but the spatial coverage will be appropriately filled

(2) applies even when (1) can be mapped to a modern nation state. So, if
the source says a time period covers "Spain", the spatial coverage will
still have an entry for "Spain" in geonames (or dbpedia, or whatever).

This is similar to our approach with start/stop dates: a textual label
directly from the source, and derived structured data that estimates the
natural language for use in a computer interface.

—
Reply to this email directly or view it on GitHub
#18 (comment).