Terrabyte Incognita
April 2, 2014 9:46 AM   Subscribe

Africa Might Not Look Like You Think It Does
There is no such thing as an objective map. This was true of cave paintings, Roman tapestries, and colonialists' charts of Africa. It is also true of Google Maps.
posted by infini (58 comments total) 18 users marked this as a favorite
 


Center the equator. Down with Mercator!
posted by Phredward at 9:52 AM on April 2, 2014


There is no such thing as an objective map.

With regard to size? Sure, there is. It's called a 'globe'.
posted by Mitrovarr at 9:54 AM on April 2, 2014 [20 favorites]


I was under the (mistaken?) impression that this sort of thing is one of the reasons why OSM exists?

I mean, not that it's impartial either, inasmuch as nothing is.
posted by aramaic at 9:58 AM on April 2, 2014 [1 favorite]


I thought that google earth was pretty accurate as far as sizes of continents, etc. A virtual globe, if you will.
posted by bbqturtle at 10:00 AM on April 2, 2014


I don't often do the "obligatory xkcd" thing, but I like this one.
posted by Holy Zarquon's Singing Fish at 10:00 AM on April 2, 2014 [23 favorites]


yeah, this was begging for that xkcd.
posted by k5.user at 10:04 AM on April 2, 2014


"With regard to size? Sure, there is. It's called a 'globe'."

A globe is still not the territory.
posted by Blasdelb at 10:04 AM on April 2, 2014


It took a lot of reading, a lot of the history of mapmaking, a lot of the pitfalls previous cartographers have had to overcome, a lot on the problems with orthographic projections, and a lot on the orientation of the Earth as presented on maps, to get to the meat of the argument, the point where the article reveals:

"It is telling that some townships in South Africa are just blank spaces on the map," says Brotton. "In Google's corporate model, certain communities just aren't of any interest.

That is the only concrete example of Google's biases offered in the article.

I myself am wary of Google at this point (GOOGLE READER NEVAR FORGET), but even here in Brunswick, Georgia there are details that haven't made it into Google Maps. Google could do better, yes, but I haven't seen anything to suggest that these are willful omissions, as opposed to details they haven't filled in yet. Yes, African townships are not less important than Brunswick, Georgia. But you have to start somewhere, and end somewhere else, and there may just be more smartphone users in Brunswick, GA than in African townships.
posted by JHarris at 10:05 AM on April 2, 2014 [13 favorites]


Down with Mercator!

As a cartographer I hear this all the time, and yes Mercator is imperfect, but SO ARE ALL MAPS. They're all skewed. That's how projections work. When making a map you use the projection that best preserves what's most important: area, shape, distance, direction. Down with Mercator when there's a better projection for what you're trying to convey.

Also: "Mapmakers have always claimed objectivity," he says, "and cartographers always imagine they're creating maps from some omniscient Godlike position. When it comes to Google Maps, however, the reality is that they're being produced on the west coast of America."

No, we don't do either. Cartography is inherently subjective. What you choose to put on the map says just as much as what you choose to leave off. And cartographers ≠ Google Maps.
posted by troika at 10:06 AM on April 2, 2014 [22 favorites]


Even Mercator has a purpose. It's for sailing. (I know this point is in the article, but my link elaborates on it.)
If you draw a straight line between two points on a map created using the Mercator projection, that line represents the direction you need to sail to travel between the two points. This type of route is called a rhumb line or loxodrome. It is not the shortest route, but if you keep the direction of your ship constant with respect to north then you will stay on course and arrive at your destination.
posted by desjardins at 10:16 AM on April 2, 2014 [11 favorites]


"That is the only concrete example of Google's biases offered in the article."
When using google maps to look at cities in Africa like Goma for example, in the DRC next to Rawanda, where a million people live there are only four landmarks, a highway and a handfull of streets named. The automatically drawn streets go through people's houses and the only label on the airport is the one drawn on the runway. Contrast that with San Francisco and you do indeed have a bias that is very much worth noting and considering depending on the context.
posted by Blasdelb at 10:17 AM on April 2, 2014 [4 favorites]


Stating or implying that Google chose to not incorporate easily available and properly formatted cartography data that exists for townships in South Africa

....for any reason other than the simple fact that data simply doesn't exist...

is just the worst sort of false flag tumblr gotcha game grarr.

We are talking about a company that has gone way out of their way to make our existing maps, and cartography data better....sometimes too enthusiastically.

Just over the last four years, the google maps data I see driving around rural areas of Mexico have had massively improved, despite terrible cellular service. They used to be completely erroneous or more commonly, just completely blank for the majority of areas, even when driving through major cities. Now they're nearly as good as they are on the other side of the border. It really is a work in progress. Google maps wasn't that great when it first came out here in the states, either, but they make it better by using data as soon as it's available, sometimes even going so far as to create it where none exists.
posted by hobo gitano de queretaro at 10:19 AM on April 2, 2014 [23 favorites]


Blasdelb, I don't think what you are saying is true in scale, but is in spirit. Turn off the satellite imagery and look at the "map" layer. I see much much more than 4 landmarks noted.
posted by LoopyG at 10:19 AM on April 2, 2014 [2 favorites]


A globe is still not the territory.

Problem solved!
posted by anonymisc at 10:29 AM on April 2, 2014 [1 favorite]


Maps are lies sold to us by the government and big cartography. I prefer getting to where I need to go the way my ancestors did - angrily driving around at random until I accidentally stumble upon the place I was heading.
posted by Joey Michaels at 10:31 AM on April 2, 2014 [21 favorites]


I am totally putting "predilection for doodling odd creatures" on my resume.
posted by desjardins at 10:31 AM on April 2, 2014


Contrast that with San Francisco

IME, contrast pretty much any city in the world with San Francisco on Google Maps and San Francisco will be more detailed or more up to date :)

Silicon Valley is the new England-at-longitude-0.0
posted by anonymisc at 10:37 AM on April 2, 2014 [3 favorites]


Yes, African townships are not less important than Brunswick, Georgia. But you have to start somewhere, and end somewhere else, and there may just be more smartphone users in Brunswick, GA than in African townships.

Given the intensely rapid adoption of wireless telephony throughout Africa + the massive populations of the townships (Soweto alone has four million people), I'm going to categorically disagree.
posted by psoas at 10:39 AM on April 2, 2014 [5 favorites]


I wish the article expanded the premise that Google's map is culturally specific. It makes sense that an American company would have access to much more data about cities in the US and Europe and similar data may not be available for Africa or the Middle East.

But what data is available for those locations? Google emphasizes roads and established businesses or attractions. If you culled data from residents of Nairobi or a village in Turkmenistan, to where would you be pointed? The results may be much more subjective (I like the imagery of a map from a fisherwoman), but then again so are all maps. A Maine lobsterman knows a lot more about the sea than Google can offer, and probably doesn't care much about the star rating of nearby restaurants.
posted by Turkey Glue at 10:40 AM on April 2, 2014 [1 favorite]


To contrast the linked google map for Goma in the Congo, let's go 120 miles due south to Bujumbura, Burundi.
posted by ursus_comiter at 10:47 AM on April 2, 2014 [4 favorites]


"As was the case a century ago, it is still just a small group of Western individuals with specific ideas of the world that have the resources to map the world"

As noted above, this is no longer quite the case thanks to projects like OpenStreetMap and WikiMapia. Locals absolutely now have the ability to populate maps with details they find relevant.
posted by any portmanteau in a storm at 10:49 AM on April 2, 2014 [2 favorites]


Was just going to post that West Wing clip. Brilliant.
posted by Fizz at 10:52 AM on April 2, 2014


Some of you may be interested in Ushahidi, a non-profit organization that makes open source software for community mapping, and the concept of PPGIS (Public Participation Geographic Information Systems). Both of these fall under the umbrella of neogeography.
posted by desjardins at 10:53 AM on April 2, 2014 [2 favorites]


When using google maps to look at cities in Africa like Goma for example, in the DRC next to Rawanda, where a million people live there are only four landmarks, a highway and a handfull of streets named. The automatically drawn streets go through people's houses and the only label on the airport is the one drawn on the runway. Contrast that with San Francisco and you do indeed have a bias that is very much worth noting and considering depending on the context.

And yet, it's the best and most current map of the city I can find. If you actually view the city in map form rather than satellite image and zoom so that the landmarks are shown, there are a large number. The only other comparable map is OpenStreetMap, which does a better job with naming roads and includes the figure-ground details of buildings, but has far fewer landmarks. (In my experience, in places with less developed infrastructure landmarks are more useful for navigation than street names.) I looked at one section at random, the area SW of the BDGL roundabout, between Kanyamuhanga and Mount Goma boulevards and the lake, and OSM identifies 8 landmarks where Google Maps identifies all 8 of those and 17 more; Alanis-ironically, it's the only map where you can find the land survey bureau. Other maps I can find are much less useful and focus on the volcanic disaster 12 years ago.

Is Goma mapped by Google as well as San Francisco? No. It's only better mapped by Google than by anybody else in history.
posted by Homeboy Trouble at 10:54 AM on April 2, 2014 [14 favorites]


I was just puttering around the Congo on Earth, and I guess the thing that struck me is that none, or very little, of the content was local, or would seem to be useful or of interest to local people. It all had a very imperialistic feel, what little there was.
posted by maxwelton at 10:57 AM on April 2, 2014


> I was just puttering around the Congo on Earth

I read this very poetically until I realized what you meant.
posted by scose at 11:06 AM on April 2, 2014 [1 favorite]


In fact, even at the 1885 Berlin Conference where, legend has it, Europe's colonial powers each drew lines across a map of Africa and coloured in their territories with their imperial hue-of-choice, the colonialists weren't really sure what those areas contained.

I was noticing that an early 20th century map reproduced on the cover of a blank book I bought seems to have the colors of colonized territories keyed to match those of the countries that colonized them. Image of Stanford's General Map of the World from 1920.
posted by larrybob at 11:17 AM on April 2, 2014


On Exactitude in Science
posted by mumimor at 11:40 AM on April 2, 2014 [1 favorite]


Google maps is, first and foremost, a navigation tool. You want directions from A to B, and it draws you a map with a purple line telling you which way to go.

For this, it should have a conformal projection, because it's more important that local shapes are accurate than it is that global shapes are accurate. At a complex intersection in Tromsø, you don't want to depict a 75° turn with an 145° angle or vice-versa.

And because we talk about directions in terms of compass directions, it's important that the latitude and longitude lines are straight.

I believe the only projection that does both these things is Mercator, though it is true that Google chose put North at the top of the map. I do accept that North-is-up may have some tricky cultural implications, but it's hard to make a strong case that it is less accurate than South-is-up, in Africa or elsewhere. And 2/3 of Africa is North of the equator anyway.
posted by aubilenon at 11:42 AM on April 2, 2014 [4 favorites]


Zoom in close enough on any US city and you'll get the outline of individual houses. Zoom in on a Canadian city and you'll get outlines for commercial and industrial buildings, and multi-level residential, but not houses. This isn't because of an anti-Canada bias, but because Google just doesn't have the data.

Similarly, in Canada, every little body of water is on the map, unlike the US. In the US, the terrain data is good enough that you can see elevation changes block by block, but no so in Canada.
posted by Pruitt-Igoe at 12:11 PM on April 2, 2014 [3 favorites]


When I coded a CD-ROM atlas, my US publisher wanted me to make a version for the Chinese market where Taiwan and Tibet were just marked as part of China. I told them to fuck off.
posted by w0mbat at 12:30 PM on April 2, 2014 [6 favorites]


As far as the detailed data of labeling places, buildings, streets, etc. then yes one could argue there are flaws in Google Maps, but like others here I feel that's just gaps in information, not actual bias or intention.

When it comes to the subject of projections, though, Google Maps in my view has all other projections beat. Zoom in enough and it's flat, as is reasonable for a local area map, but zoom out enough and it becomes a globe. It's the best of both worlds.
posted by dnash at 12:51 PM on April 2, 2014


I was just puttering around the Congo on Earth, and I guess the thing that struck me is that none, or very little, of the content was local, or would seem to be useful or of interest to local people.

I wonder if this is for the same reason that content about America on Weibo probably isn't always relevant to American locals. Does anyone know if there are big African maps sites that are not on American critics' radar, like I hear Orkut was big in India at least for a while? I could see either - wireless penetration argues Africans might very well be demanding maps, lower overall economic development argues they might not have made them yet. Kinda feels White Man's Burden-y that there doesnt seem to be any prospect of African made Africa Maps discussed.

Think I'll have a look at Baidu's map coverage later.
posted by save alive nothing that breatheth at 1:06 PM on April 2, 2014


So the part where they discussed putting South at the top of maps kind of blew my mind. My whole concept of direction is tied to North="up" and knowing which way is West vs. East on a map because it spells "we" from left to right.
posted by emjaybee at 1:07 PM on April 2, 2014 [1 favorite]


Google Maps is useful for some things, but useless for others. AFAIK they don't mark most railways, so it's not all that useful for learning about them. The lack of waterway information is also an issue.
posted by Monday, stony Monday at 1:44 PM on April 2, 2014


Organization for Cartographers for Social Equality

Heh, the first map they show isn't the Eurpean imperialist map that they claim it is, it's actually that insane American map (that used to be?) used in US schools that (unlike the European maps) chops up dozens of countries (instead of ocean) at the sides, drops the equator way below the center (deleting the lower latitudes and most of Antarctica off the bottom of the map to achieve this!), and shifts the dateline and the 0 longitude to bizarre arbitrary places, all so that US schoolkids don't need to learn geography but can simply point to the center of the world and say "that's us!".

Maybe it is the most imperialist map ever, but it ain't European imperialism! :-)
posted by anonymisc at 1:52 PM on April 2, 2014


"Bias" doesn't necessarily mean "evil scheme". It simply states a trend. The reason may well be data availability, but that doesn't change the fact that Google Maps, objectively speaking, has a huge bias in coverage towards North America and away from Africa (for example).
posted by threeants at 2:12 PM on April 2, 2014


"Mapmakers have always claimed objectivity," he says, "and cartographers always imagine they're creating maps from some omniscient Godlike position.["]

This interests me. Have postmodernism and critical theory really not made any headway into today's cartography field? I actually kind of wonder who cartographers are, in terms of professional background-- as far as I'm aware universities don't have much in the way of cartography programs. Did they begin as geographers; GIS technicians; graphic designers; artists?
posted by threeants at 2:19 PM on April 2, 2014


"why have a map library in the age of Google Maps?" – I love demonstrating the answer to this question on a daily basis as part of my job. On my phone or else I would post much more in this and the recent NYPL thread...
posted by avocet at 2:23 PM on April 2, 2014 [3 favorites]


I began as a graphic designer, then made a lateral step into GIS, which led to undergrad/grad research in cultural/political geography.
posted by avocet at 2:25 PM on April 2, 2014 [1 favorite]


paging betafilter to the thread!
posted by avocet at 2:25 PM on April 2, 2014


For a beautiful treatment of the philosophical and cultural questions underlying the theory and practice of cartography (some of which are raised above), I recommend David Turnbull's Maps are Territories, Science is an Atlas
posted by narcotizingdysfunction at 2:32 PM on April 2, 2014


"Mapmakers have always claimed objectivity," he says, "and cartographers always imagine they're creating maps from some omniscient Godlike position.["]

This interests me. Have postmodernism and critical theory really not made any headway into today's cartography field?


Yeah, I think the premises you quote are wrong. Just rhetorical hot air. The larger point might or might not have substance, but grandiose statements like that are weak. I'm sure that if we searched hard enough, we could find mapmakers so clueless about their craft (perhaps a schoolkid drawing their version of not-Middle-Earth for their epic fantasy trilogy which they will only get a few pages into writing?), but all the mapmaking I've ever been exposed to has been extremely focused on who the audience might be, which information might interest them, how this should influence how information might be presented, how much of that information can be obtained, at what cost, what accuracy, etc etc.

Just uttering the words "London Tube Map" makes claims of mapmakers chasing objectivity go wobbly. Objective pertaining to what?
posted by anonymisc at 2:44 PM on April 2, 2014 [2 favorites]


Came for CJ; not disappointed.
posted by The Bellman at 3:13 PM on April 2, 2014


I actually kind of wonder who cartographers are, in terms of professional background-- as far as I'm aware universities don't have much in the way of cartography programs. Did they begin as geographers; GIS technicians; graphic designers; artists?
GIS librarian-who-makes-maps here. I was a geography major with an eye for design who ended up working on historical GIS (largely digitization) projects during my time in school. These days, most of my cartography skills go toward helping researchers create figures for their articles, as well as a smattering of web mapping projects. Our university doesn't teach cartography per se, much to my disappointment, so I end up covering the very basics for students in library workshops instead. Making digital maps is easier than ever these days, but design choices matter a whole bunch when it comes to conveying information.
posted by betafilter at 3:35 PM on April 2, 2014 [2 favorites]


I actually kind of wonder who cartographers are, in terms of professional background-- as far as I'm aware universities don't have much in the way of cartography programs. Did they begin as geographers; GIS technicians; graphic designers; artists?

GIS Specialist/Cartographer here, I work in international development and occasionally write about maps on the internet. My degree is in Geography with a minor in GIS, and I began as a person that loves maps, when I found out there was a career in it I never looked back. It's awesome, I love my job, it's fun and challenging.

You're right, though, universities don't have much cartography. They don't have much geography, either. None of the ivies have had a geography program for decades, for example. It's barely taught in schools, and when it is it rarely goes beyond memorizing capitals.* No Child Left Behind is terrible for many reasons, but one that you don't hear so often is that of the 12 'core education' areas, 11 were funded. Geography, the 12th, got nothing. President Obama's update to NCLB didn't mention geography at all. The subject is increasingly bundled with social studies and inevitably dropped in favor of standardized testing prep or. No wonder folks don't know how big Africa is! No one's taught them to question the map!

*a common response to "I studied geography" is "Oh, so you know all the capitals?" :-|
posted by troika at 5:34 PM on April 2, 2014 [2 favorites]


I'd love to get into cartography, but like troika above said, I never learned geography beyond memorizing capitals and memorizing Middle Eastern countries in 7th grade (which I am somehow still pretty decent at knowing, so if someone on the street ever tries to trick me on one of those shows I AM PREPARED). My girlfriend is studying community development and is taking GIS classes this term and has a cartography class, which is really neat. I think I'd like to do something like that, but the urban planning stuff here sounds like it's over saturated.
posted by gucci mane at 6:19 PM on April 2, 2014


The fact that our maps typically put North at the top is a mere convention.

I dunno. Just as Mercator's map is useful for navigation, having a map with North at the top seems intuitive in a hemisphere where there's a nice, nearly-fixed star in view right above the pole that you can look "up" to when holding a map.
posted by RonButNotStupid at 6:23 PM on April 2, 2014 [1 favorite]


The True Size of Africa.

Visual comparisons of land areas are much more memorable to me than global "grand overviews".

After 20 years of internet, printed atlases remain generally much more useful for many purposes. The promise of software and computers have a long ways to go to realize their potential superiority. Alas, knowing where the local MacD store is seems to dominate the popular field.
posted by Twang at 9:38 PM on April 2, 2014 [1 favorite]


So the part where they discussed putting South at the top of maps kind of blew my mind. My whole concept of direction is tied to North="up" and knowing which way is West vs. East on a map because it spells "we" from left to right.

I know, right? Put south at the top? Ew!
posted by sodium lights the horizon at 11:07 PM on April 2, 2014


I've always loved the Goode Homolosine and wondered why more people didn't love its orange peely goodness. It seems like a good balance of sizes and visually reminds you that we're on sphere.

That xkcd comic always makes me want a Waterman Butterfly projection... My study doesn't have enough wall space.
posted by sodium lights the horizon at 11:11 PM on April 2, 2014


University of Wisconsin-Madison has a solid geography program.

Here's a loooong list of university cartography labs. Mostly in the US, but not all.
posted by desjardins at 8:30 AM on April 3, 2014


Fun fact: betafilter and I became friends in undergrad when she caught me reading MeTa in the hostel lounge on a field course we were on together. :)

We then worked together on a semester-long mapping project, and we found ourselves watching the infamous Miss Teen South Carolina video quite a bit. Back then I found the "well maybe they don't have maps" line pretty ridiculous, but rewatching it in 2014 the comment has a totally different and troubling resonance. What kind of perspective does the personalized cartography and increasingly large-scale uses of Google Maps offer at a regional or global scale? What kind of topologies and geographic knowledge are constructed when paper maps of varying scales and themes don't see the same kind of permanence in the realm of everyday life as the dynamic digital geospatial experience unfolds? I dunno, growing up with maps in the car and the globe as my favourite childhood object prompted such a curiosity for learning about these other places on the map. I'm not a parent and I don't have any young people in my life to ask, but do kids these days find the same fun on a phone or in Google Earth? Thinking about Miss Teen South Carolina's remarks and interpreting "maybe they don't have maps" as "maybe they only have a map" made that amusing video a little dark.

But wow, introducing younger people to our map collection is my favourite part of my job – the unexpected insights they pull from historical maps and aerial imagery blow my mind.
posted by avocet at 9:49 AM on April 3, 2014 [2 favorites]


do kids these days find the same fun on a phone or in Google Earth?

I don't know about kids, but adults do. I do. And exploring in Google Earth is its own hobby these days.
posted by anonymisc at 2:39 PM on April 4, 2014


Just as Mercator's map is useful for navigation, having a map with North at the top seems intuitive in a hemisphere where there's a nice, nearly-fixed star in view right above the pole that you can look "up" to when holding a map.

Similar to how it would be intuitive to have South at the top in a hemisphere where there's a nice, nearly-fixed constellation in view that you can look up to when holding a map.
posted by mosessis at 7:10 AM on April 8, 2014






« Older Okay, but, street slang is an increasingly valid...   |   Saudi Arabia declares atheists and political... Newer »


This thread has been archived and is closed to new comments