When sources disagree: borders and place names in Google Earth and Maps
Friday, December 4, 2009
Collecting and sharing the most accurate information about place names and borders is a tough task that every map maker faces. The first sources are the nations themselves, but when neighboring countries claim overlapping territories and conflicting place names, even showing the dispute on a map may be prohibited by local law. We continue to work hard on these issues, and thought it would be worth sharing our general approach on this blog.
We want to be transparent about the principles we follow in designing our mapping products, particularly as they apply to disputed regions. Last year, for example, we explained how we determine the names for bodies of water in Google Earth. For each difficult case, we gather a cross-functional group of Googlers including software engineers, product managers, GIS specialists, policy analysts, and geopolitical researchers. This process benefits from the local knowledge and experience of Googlers around the world.
We follow a hierarchy of values to inform our depictions of geopolitically sensitive regions:
Google's mission: In all cases we work to represent the "ground truth" as accurately and neutrally as we can, in consistency with Google's mission to organize the world's information and make it universally accessible and useful. We work to provide as much discoverable information as possible so that users can make their own judgments about geopolitical disputes. That can mean providing multiple claim lines (e.g. the Syrian and Israeli lines in the Golan Heights), multiple names (e.g. two names separated by a slash: "Londonderry / Derry"), or clickable political annotations with short descriptions of the issues (e.g. the annotation for "Arunachal Pradesh," currently in Google Earth only; see blog post about disputed seas).

Authoritative references: While no single authority has all the answers, when deciding how to depict sensitive place names and borders we use guidance from data providers that most accurately describe borders in treaties and other authoritative standards bodies like the United Nations, ISO and the FIPS. We look for the references that are the most universally recognized for each individual case. For example, in the case of "Myanmar (Burma)" ISO and FIPS each use a different name, so we include both to provide a more complete reference for our users.
Local expectations: We work to localize the user experience while striving to keep all points of view easily discoverable in our products. Google Maps has launched on 32 region domains (e.g. maps.google.ca for Canada) and Google Earth is now available in 41 languages. Each domain and language user population is most familiar with a slightly different set of place names. For example, for the "Yellow Sea" or "West Sea," Chinese speaking users are conversant with the label Huáng Hǎi or 黄海 (Mandarin), while Korean users are used to the label Sŏ Hae or 서해 (Hangul). Carefully considering Google's mission, guidance from authoritative references, local laws and local market expectations, we strive to provide tools that help our users explore and learn about their world, and to the extent allowed by local law, includes all points of view where there are conflicting claims.
Sometimes these factors compete with one another. For example, is localizing a place name inconsistent with Google's mission? What happens when an authoritative references does not seem to represent the truth on the ground? What about when local user expectations don't match international convention, or when local laws prohibit acknowledging regional conflicts? These are questions we continue to think through in our efforts to provide comprehensive, authoritative, free, and, most importantly, useful products for our users.
Local expectations: We work to localize the user experience while striving to keep all points of view easily discoverable in our products. Google Maps has launched on 32 region domains (e.g. maps.google.ca for Canada) and Google Earth is now available in 41 languages. Each domain and language user population is most familiar with a slightly different set of place names. For example, for the "Yellow Sea" or "West Sea," Chinese speaking users are conversant with the label Huáng Hǎi or 黄海 (Mandarin), while Korean users are used to the label Sŏ Hae or 서해 (Hangul). Carefully considering Google's mission, guidance from authoritative references, local laws and local market expectations, we strive to provide tools that help our users explore and learn about their world, and to the extent allowed by local law, includes all points of view where there are conflicting claims.
Sometimes these factors compete with one another. For example, is localizing a place name inconsistent with Google's mission? What happens when an authoritative references does not seem to represent the truth on the ground? What about when local user expectations don't match international convention, or when local laws prohibit acknowledging regional conflicts? These are questions we continue to think through in our efforts to provide comprehensive, authoritative, free, and, most importantly, useful products for our users.
I checked map of India in different versions like one on maps.google.com other on maps.google.co.in and one on Chinese version why Indian state Arunanchal Pradesh is in China for Chinese version? and why Indian and US version of maps are different in border
ReplyDeleteआशु, did you read the part of the post that says "or when local laws prohibit acknowledging regional conflicts."?
ReplyDeleteInteresting that Google's using its various domain names to improve localization of place names. Perhaps Google should get behind the campaign to get ICANN to allow the .cym TLD, so that there can be a version that uses spellings of Welsh placenames that respect modern orthographic conventions (rather than Anglo-Norman exonyms) in Wales and adjoining parts of Shropshire and Herefordshire.
ReplyDeleteIf Google professes that they try to reflect the local views on disputed territories and also maintain ground realities, maps.google.co.in should reflect Arunachal Pradesh as a part of India where elections were held recently. If a concession has been made on the chinese version of Google maps, I am sure reflecting a reality and a local view on the indian version is not even much of a concession
ReplyDeleteHaving recently published a book on place-names, documenting the origins of the names in a small New England town, I can understand the difficulties of keeping up with what is accurate and acceptable to locals. Fortunately, I could get around the dilemma with a sub-title: facts, folklore, fibs. However, we still have the difficulty of two towns disagreeing on the spelling of a road: Rucum or Rucuum and no information on its name origin.
ReplyDeleteI just read the "we're improving borders" posting by Charlie Hale, Google Geo Policy Analyst, and I want to ask why it was necessary to remove borders around large enclaves such as Büsingen, Germany and Campione d'Italia recently? That was an unannounced step backwards that caused me to swear off using Google Maps because of the creation of even more erroneous borders than were there before.
ReplyDeleteOn the topic of "borders with little resolution, there's never been a reason to merge or disregard geo border data into compromise borders graphics. We're big boys and girls, and if multiple sets of data show variances, the maps legends could have simply explained the multiple sets of source materials were at odds with one another. Better that than creation of border graphics that look like Rand McNally World Atlas mediocrity.
lnadybal@exclave.info
I just read the "we're improving oborders" posting by Charlie Hale, Geo Policy Analyst, and I want to ask why it was necessary to remove borders around large enclaves such as Büsingen, Germany and Campione d'Italia recently? That was an unnanounced step backwards that caused me to swear off using Google Maps because of the creation of even more erroneous borders than were there before.
ReplyDeleteOn the topic of "borders with little resolution, there's never been a reason to merge or disregard geo border data into compromise borders graphics. We're big boys and girls, and if multiple sets of data show variances, the maps legends could have simply explained the multiple sets of source materials were at odds with one another. Better that than creation of border graphics that look like Rand McNally World Atlas mediocrity.
As a follow on to the message I just posed about enclaves that went missing a few months ago (about which there's been no border dispute, name discrepancies or local prohibitions to pander to, Google has never made even a remote attempt to show the 106 enclaves of India in Bangladesh or the 92 Bangladesh enclaves in India.
ReplyDeleteLN
Maps of these are hard to find, but they exist. It's almost impossible for Google to avoid knowing of them, for a search of news articles on Google turns up many, many stories of gun battles on their borders, proposed treaties to trade exclaves, etc., treaties providing access leases over the other nation's territory, etc.
I propose Google also try to draw military armistice borders in places like Nagorno Karabach (and the enclaves within it), as well as borders of the Han River Neutral Zones between the UN Armistice Commission and North and South Korea, for which no mention is made.
LN