Clarify your criteria for "country".#218

Country is used inconsistently throughout the site. I understand that this is a somewhat nebulous term. However, since Atlas Obscura has a passport feature, are we to assume that a country is a political designation (as used for entry with a passport or visa)? Or a cultural or geographic designation? Here England and Scotalnd are two separate countries (even though both are part of the UK in your passport section), while the island of Cyprus–politically divided between the Republic of Cyprus and the Turkish Republic of Northern Cyprus–falls under the heading of one (Cyprus). Palestine for its part is not listed separately from Israel (although there is admittedly no consensus on this point). And is it fair to classify Hong Kong as a country?

Thanks for this suggestion. True, we have long struggled with the definition of country in organizing the Destination Guides and data for each place. We use the Google geocoding service to pull and verify location data, so our database tends to reflect what comes back in the “Country” field from Google for a given place/coordinates.

If we are able to re-build the way geographical data is stored on the back-end or switch geocoding services we will certainly take this feedback into account.

Related to #157

@Nicola but is the Passport feature dependent on external geocoding too? If it is entirely built by AO however, I’m still seeing separate stamps for United Kingdom and Scotland, yet none for England.

This goes beyond the topic of how AO classifies countries, but to my mind a problem with geocoding, google maps, search engines, related spacial tools, and classification systems of this kind, is that they become eqivalent in their limitations to analogue systems (guidebooks, phone directories, road maps, atlases, and so on). As older intrepid travelers know, if you can’t find it on a map or guidebook, it doesn’t exist, and conversely, it can’t exist if it doesn’t appear in a map or guidebook. Even when it does. So, the problem becomes a recursive one. Especially since truly obscure sites are sometimes conspicuous in their absence. The modern equivalent is the internet, geocoding, google maps, and so on. If you can’t find it on the internet, it doesn’t exist. Except that it does. Or said another way, the absense of proof is not proof of absense–a point driven home to me recently when searching for a Mexican town that I’m intimately familiar with that I had a challenging time locating on the internet (but can find readily on many Mexican maps). These tools don’t represent the fact, but rather become the fact (as in the case of defining “countries”). Singh makes some interesting points here. You might find it worth reading: (Singh, S.K. Evaluating two freely available geocoding tools for geographical inconsistencies and geocoding errors. Open geospatial data, softw. stand. 2, 11 (2017). https://doi.org/10.1186/s40965-017-0026-3

@artandarchaeology Yes, thanks for this detailed and thoughtful response. I look forward to checking out that paper. We at AO do not want to be the arbiter of country definition, but we will keep this feedback in mind as we improve our language, interface, and features on the site that refer to countries.

@linkogecko Yes the passport feature does depend on the geocoding. Essentially, it grabs all the places you’ve marked as “Been There” and compares what is in the Country field to our database of Destinations that are also classified as “Country” by the geocoder.

We could consider a hardcoded dictionary of countries that then looks up if any of the places you’ve marked as “Been There” appear in the list, but then we’d have to maintain a dictionary of countries. We also don’t have plans to update the Passport feature right now, as we believe other feature are higher priority.