Yesterday, reports that Google Maps had stopped obscuring satellite imagery of sensitive Russian military facilities spread like wildfire across Twitter. Only there was no official announcement from Google saying they’d done so, and while Ukrainian Twitter was seriously running with it, I wanted to see some confirmation from the mapping side. In the event, an update to Ars Technica’s story says that Google hadn’t stopped blurring the imagery—the imagery hadn’t been blurred in the first place. “A Google spokesperson told Ars that the company hasn’t changed anything with regard to blurring out sensitive sites in Russia, so perhaps none of us were looking closely until now.”
Meanwhile, the South China Morning Post reports that “iPhone and Apple Watch users in China can no longer see their geographic coordinates and elevation on the Compass app, according to Chinese media reports and user comments. However, information including bearings and general location are still available.”
And according to a report in The Information (paywall) that was summarized by John Gruber, back in 2014 or 2015 the Chinese State Bureau of Surveying and Mapping required Apple Maps to make the disputed Diaoyu (Senkaku) Islands appear large even when zoomed out, and made the Apple Watch’s Chinese release contingent on that request—to which Apple acquiesced.
Google has agreed to Parks Australia’s request that user photos taken from the summit of Uluru (formerly known as Ayers Rock) be removed from Street View; climbing Uluru, which is owned by and sacred to the Pitjantjatjara people, has been prohibited since 2019. ABC Australia, CNN. As of this writing a couple of images are still visible. Aerial coverage is unaffected. [Boing Boing]
As part of their investigation into China’s practice of detaining Uighur and other Muslim minorities in Xinjiang, Buzzfeed News journalists compared blanked-out areas in Baidu Maps with uncensored imagery from Google Earth and satellite data providers, and, after sorting through some 50,000 possible locations using custom web tools, built a database of some “428 locations in Xinjiang bearing the hallmarks of prisons and detention centers.” This article explains the methodology.
Blurring or removing map data to prevent people from seeing something important or sensitive is a pretty loud signal that there’s something important or sensitive to see there. Some five million Baidu Maps tiles were masked in Xinjiang alone—there’s a lot the Chinese government considers sensitive—which made the unmasking considerably harder. But not impossible.
A blog post from the National Library of Wales explores how sensitive military and industrial sites were omitted from the published versions of Ordnance Survey maps.
The removal of military installations from OS maps was at its height in the 19th century and the World Wars, but throughout the Cold War and beyond, many sensitive sites were left off the maps entirely. It took the public availability of high-resolution satellite imagery at the turn of the 21st century to render this type of censorship largely ineffective, although labels are still omitted in some cases.
The Ordnance Survey did survey and map sensitive sites, but those maps were military-only. The differences between these military maps and the public maps make for a number of interesting comparisons: see the post for examples.
I hadn’t realized that Tavish Scott’s amendment preventing Scottish maps from displaying Shetland in an inset map actually passed. Section 17 of the Scottish Parliament’s Islands (Scotland) Act 2018 requires maps of Scotland produced by Scottish public institutions to display Shetland “in a manner that accurately and proportionately represents their geographical location in relation to the rest of Scotland.” The Act passed the Scottish Parliament last May and received Royal Assent in July. Now that the provision in question has come into force, the media, which always likes a weird map story, is seized of the issue all over again: there were news stories last week from BBC News, CBC Radio’s As It Happens, and NPR.
Basically, Shetlanders are delighted and cartographers are horrified: maps of Scotland will perforce be less detailed to accomodate all the empty ocean. In practice I suspect little will change: a loophole in paragraph 17(2)(b) enables a public authority to sidestep the requirement if they can justify it. I imagine that justification will be coming up a lot in maps that people actually use, leaving only illustrative and symbolic maps affected by the law. And, of course, private mapmakers and mapmakers not under the purview of the Scottish government (which I imagine includes the Ordnance Survey) will not be affected by this law.
Meanwhile, Maps Mania’s Keir Clarke gives us Unboxing the Shetlands, a tool to place mainland Scotland in an inset map instead.
The kerfuffle about Shetland being relegated to inset maps (Ed Parsons has taken to calling this “Insetgate”) is not quite done. Kenneth Field shares his thoughts in a post titled “In Praise of Insets,” in which he calls Scottish politician Tavish Scott’s proposal to ban the use of inset maps to portray Shetland as “utter nonsense” and goes on to defend their use more generally.
Insets are not just used to move geographically awkward places. They are commonly used to create larger scale versions of the map for smaller, yet more densely populated places. Often they are positioned over sparsely populated land to use space wisely. I’m guessing Scott would have an objection to an inset that, to his mind, would exaggerate the geographical importance of Glasgow compared to Shetland. Yet … in population terms it’s a place of massively greater importance so one could argue it deserves greater relative visual prominence on the map. Many maps are about people, not geography.
Writing on the Royal Scottish Geographic Society’s blog, Bruce Gittings challenges the notion that putting Shetland in an inset box is a map error:
It is plainly not: it is a cartographic compromise. And there are always implications to a compromise. To include the Northern Isles in their actual geographical location, separated from the mainland by almost 100 miles of water, would reduce the scale at which the country can be displayed by around 40%.
That means Scotland’s smaller Council Areas (e.g. Dundee) effectively disappear, reduced from any kind of area to an insignificant point, or major features such as the Firths of Tay and Forth lost under text-labels for Dundee and Edinburgh. We are left having to put the Central Belt in a zoom-box because of the loss of detail in areas where most people live, or having to use two sheets of paper rather than one for maps of Scotland. […]
The circumstance of Shetland-in-a-box (and indeed Orkney-in-a-box-too) is a feature of maps intended to display our entire country with a reasonable level of detail.
Previously: Don’t Put Shetland in a Box.
Shetland’s representative to the Scottish Parliament has moved an amendment to proposed legislation that would require public authorities to portray Shetland “accurately and proportionately” in Scottish maps: BBC News, iNews, The Scotsman. Because Shetland is so far to the northeast of the island of Great Britain, it’s usually shown in an inset map; this move would, it seems, prohibit this, and presumably require Scottish maps to show vast tracts of ocean (as above). [NLS Maps]
Self-driving cars require insanely detailed maps in order to function. But, as The Drive’s Stephen Edelstein writes, “The Chinese government is blocking foreign companies from mapping its roads in great detail, according to a Financial Times report. The restrictions, which reportedly do not apply to Chinese firms, are being instituted in the name of national security. China is concerned about spying.” Mapping, geotagging, geographic surveys—all of these have been subject to Chinese government restrictions for many years (recall the trouble Google Maps has had operating in China), so this is more of an additional data point than an actual surprise. [Boing Boing/
Previously: The Business of Making Maps for Self-Driving Cars.
Czech Railways (České dráhy) have pulled its upcoming annual diary from circulation because it includes a sensitive map of Europe, the Lidové nivony reports (in Czech; Google Translate). The map, created by Kartografie Praha, shows Crimea, Nagorno-Karabakh, Abkhazia and South Ossetia as disputed regions and marks the territory held by the Islamic State in Syria and Iraq. Apparently afraid of offending ambassadors and business partners, the railways is holding some 5,000 copies of the diary in a warehouse. [Maps on the Web]
This is not the first time a Czech publisher has gotten into trouble over a contested map. (I wonder if it’s the same publisher.)
A Czech publisher has managed to get itself entangled in the dispute over how to map Israel and Palestine, with a school atlas that showed Jerusalem as the capital of Israel (which Palestinians dispute). The Palestinian ambassador protested; the Czech education ministry relented—which enraged the Israelis, until the Czech education ministry reversed itself again. This is one of those situations where a neutral map is impossible: each option pisses off the other side. As Google found out about Crimea, it isn’t always enough to show the “right” map to the right people.
The Guardian continues to track the issue of Palestine’s absence from Google Maps. In a long essay that is definitely worth your time, Petter Hellström links the issue with the long history of colonial maps that omitted the indigenous populations that settlers would soon displace.
Because Palestine, after all, has been removed. It is there on old paper maps, of the Holy Land, of the Roman and Ottoman empires, of the British mandate. Yet in our digital age, a search on Google Maps for Israel produces a map without Palestine. It displays Israeli urban centres down to a few thousand inhabitants, and even marks Ma’ale Adumin, an Israeli settlement on the occupied West Bank. At the same time it shows no Palestinian place-names or urban centres, not even major ones like Gaza City, Khan Yunis or Nablus. The dotted, inconsistent borders of the occupied territories leave the impression that they are not claimed or administered by anyone. […]
Historians of cartography have long studied the practices and consequences of cartographic omission. In a landmark study, “New England cartography and the Native Americans”, published posthumously in 1994, the British historian of cartography J. B. Harley analysed seventeenth-century maps to follow the progressive replacement of the Native Americans with European settlers. In Harley’s analysis, the maps were something more than historical records of that process. Because they made the colonists visible at the expense of the indigenous population, they were also instruments of colonial legitimisation.
Many colonial mapmakers preferred to leave the areas of predominantly indigenous presence blank, rather than to reproduce an indigenous geography; one example is Herman Moll’s 1729 map of New England and the adjacent colonies, seen above. The traces of indigenous presence, past and present, were gradually removed from the maps as the colonists pushed west. The apparent emptiness helped to justify the settlers’ sense that they had discovered a virgin territory, promised to them by Providence. The pattern was the same in all areas of colonial activity, including Australia and Africa.
Previously: Google, Palestine, and the Unbiased Map.
As the Disputed Territories site, which catalogues how Google manages various contested borders, points out, “Google’s maps of disputed territories differ depending on who’s looking at them.” As we’ve seen recently with regard to Crimea, that doesn’t always keep Google out of trouble. An online petition asking Google to label Palestine on Google Maps has garnered more than 300,000 signatures since March. The petitioners accuse Google of removing Palestine at Israel’s insistence; but, as the Guardian reports, “the truth is, it was never labelled by Google in the first place.” (The West Bank and Gaza Strip had their labels removed by a bug; Google’s restoring them.)
In a follow-up piece for the Guardian, Leigh Alexander writes:
The swiftness of the backlash, though, is not just about the wish for justice on behalf of an occupied people, but about the belief—now punctured—that our technology is neutral, that it presents an unbiased, infallible version of the world. […]
While it might seem imperialistic for Google to decide how the US should see the rest of the world, perhaps it would be equally troubling to see the company wade into global verdicts on the righteousness of every international occupation. That it allows its sketch of the geopolitical climate to reflect the perspective of who is viewing it, rather than impose the prevailing popular opinion in the west, may not be neutral or unbiased, but it is probably the most fair.
Writing for The Wire, Sumandro Chattapadhyay and Adya Garg discuss the recent Indian draft bill that proposes fines and jail terms for publishing a map that shows the “incorrect” Indian borders. They provide some background, setting out the government’s past history of trying to regulate maps of India, and point out some flaws in the proposal:
The regulatory measures proposed by the bill do not only cause worry but also bewilderment. Take for example Section 3 that states that ‘no person shall acquire geospatial imagery or data including value addition of any part of India’ without being expressly given permission for the same or being vetted by the nodal agency set up by the Bill. If implemented strictly, this may mean that you will have to ask for permission and/or security vetting before dropping a pin on the map and sharing your coordinates with your friend or a taxi service. Both involve creating/acquiring geospatial information, and potentially adding value to the map/taxi service as well.
Let’s take an even more bizarre hypothetical situation—the Security Vetting Agency being asked to go through the entire geospatial data chest of Google everyday (or as soon as it is updated) and it taking up to ‘ three months from the date of receipt’ of the data to complete this checking so that Google Maps can tell you how crowded a particular street was three months ago.