Navigating Towards Better Maps

     

Bing map, circa 2011
Accurate but useless

I've been thinking about how to improve on modern maps and charts (in case you're wondering about the difference, maps are for land and charts are for oceans). What prompted this was the realization that today's charts, created by armies of very smart people working for large corporations and governments with the aid of huge computer clusters, are strictly worse than the ancient charts created by people who did not even have slide rules.



Take Anaximander's map, arguably the first comprehensive map of the known world. Who is Anaximander, you might ask. He's the philosopher who invented the concept of "infinity." I always thought that whoever invented zero was cool (try doing multiplication with Roman numerals if you disagree), but recently realized that whoever invented infinity must, by definition, be infinitely cooler. Well, Anaximander is that guy, and he's amazing. Xeno, with his paradox about the racing hare and turtle, gets a lot of credit for infinity-related issues, but Anaximander preceded him by 200 years, introduced the concept, gave it a name, and didn't get sidetracked by some stupid paradox involving talking animals. Not content with inventing the idea of infinity, A-man looked at fossils of his time and speculated that men evolved from fish. Before the Tea Partiers converge here and start ranting against evolution, keep in mind that A-man could not have denounced all this "reasoning-based" reasoning and simply followed creationism, intelligent design, and Christianity, because he lived 600 years before Christ was born. And as if doing all this back then wasn't enough of an accomplishment, he then predicted that there would be an earthquake, that the citizens of an entire city should sleep out in the country with their weapons, that they would otherwise be killed by the quake. And there actually was an earthquake that destroyed the said city. Let's just say A-man knows his stuff.

Mappa Mundi, circa 1300
Chock full of useful information

Anaximander drew one of the earliest known maps of the world. It did not survive, but it's a safe bet that it was full of inaccuracies. The landmass must have been all out of proportion. Major features were likely missing, and because he combined maps from different sources without common reference points, he likely duplicated some rivers and mountains. It didn't help that he believed the world to be shaped like a cylinder (actually, this was a major accomplishment at the time, as it allowed celestial objects to dip below the horizon. Previous models had the ancient world floating in a basin with water, but what contained the basin?). In any case, you can bet that A-man's data was not collected with millimeter radar mounted on satellites, not registered against the World Geodetic System (WGS-84) to capture the imperfections of the sphere that earth exhibits, and nor was it processed by armies of cartographers and brilliant software developers. He did it all by himself, with ink on parchment or deerskin under candlelight.

But we know that A-man's map was immensely useful and had tremendous impact, because people still talk about it to this day. And the reason behind its success is simple: it had a fucking purpose and he knew what that purpose was. He drew the map to help further trade between Greek city states. He could distort areas that he wanted to make appear closer to each other. He could distort regions if he were running out of parchment and wanted to keep everything on the same sheet. He could center it any way he liked, and he chose to center it on Milet, Turkey, his birthplace and a frequent stopover point on traderoutes of the time. His map, just like the maps drawn for the next 2000 years, must have contained tips for the traveler, as well as useful drawings of the wildlife one might encounter. In short, it must have been incredibly inaccurate, but had impact because it was engineered to suit the needs of its audience.

Contrast this with today's maps. Accuracy has become the sole, the only, the exclusive goal for cartographers. We've become such slaves to accurate rendering that we've forgotten that maps are drawn for a purpose. And if you keep in mind that maps are not evaluated by some abstract Ministry of Truthful and Consistent Drawings, but actually used by people with a purpose, it's easy to see that almost all current maps fall far short of actually helping the users accomplish their mission.

Greenland towering over every  continent

First, let's start with the projection. We can probably agree that a basic purpose of any map is to give a sense of the relative sizes and positions of geographic regions. Yet the Mercator projection used for most maps is absolutely horrendous. The farther you are from the equator, the more distorted things get, and no normal human being without severe astigmatism can compensate for the weird torque that a Mercator projection applies to the globe.  For instance, most people think that Greenland is a huge island. But look at the evidence. If it were a huge island, it would undoubtedly contain lots of precious resources, and we wouldn't let a two bit country with an army of only 22000 people (I'm looking at you Denmark) claim the whole place. Any established soccer hooligan can round up more than 22000 people on a good weekend, with a reserve force of 30000 if it's a three-day holiday weekend, and another 50000 if Liverpool is involved. You don't need nukes or even bombs to kill off 22000 people; you can just redirect the urban runoff from any developing nation and they'd drown. If there was a giant landmass bigger than the US and Africa (as it seems from a Mercator map), and it had only 22000 European soldiers defending it while they spent their leaves in coffee houses that openly served hash, it's a safe bet that the "Islamo-terrorists" that the Danish right wing press is so worried about, and keeps on goading into action with cartoons, would have overrun the place. Clearly, Greenland can't be that big, but if you look at a modern map, it looks bigger than the biggest continent. Cartographers will claim that the map is accurate and your perception is wrong. This is precisely the problem: your perception is the only thing that matters, and cartographers have lost track of their purpose.

Cartographers have an unhealthy
fixation on long cylindrical objects
such as radio antennas and towers

Second, let's look at what cartographers put on maps. Nautical charts are the worst offenders here, as they typically feature only depths and leave out most details involving terrain. To be fair, they also include features like radio towers because cartographers think that you could somehow tell apart the tower's puny blinking red light from all the nearby Coca-Cola signs, and thus believe them to be useful for navigation. I guess they would be, if your goal was to get to a radio tower. I want to get to places. I want to learn what I would see as I look around. I want to find the most happening place, with the most interesting people. I want to know what people are like in a town. And guess what: I can handle imprecise information. I'm human, I was bred to deal with uncertainty and imprecision. Give me a sense, even if it's biased, of what it's like to experience a location. If you don't know, put a serpent there or whatever, just like the old-timer cartographers, so I know you don't know. But don't fill up all available surfaces with satellite pics, your catalog of all poles higher than 10 meters (why do you have such a catalog anyway?), or low-res photographs fat American tourists took at various spots. This is really a problem of ranking relevant information, and you guys are living in the age of Alta Vista and Webcrawler. Help me with my destination, not the route. Help me with ranked, relevant content; don't just display any and all geographical content that matches my search. Who makes the best sobrasado in Mallorca? Where can I order a paella for one person? I really want to know (and incidentally, why can't the Spanish industrial apparatus manufacture a paella pan smaller than for two people?) A map is a medium that should be uncluttered, and maybe even blank, by default, and what you choose to put on it should depend on what the user is trying to accomplish.

Who could possibly care?

Third, look at the way cartographers display their data. These guys plot tons of data every chance they get. They just can't resist the appeal of data bling. Look how meticulously they plot depths throughout the globe. Yet most people could not give a rat's ass about depths greater than the depth of their keel. Since pleasure boat keels are under 2m, the details of the subsea surface up to a few meters deeply matters to people, and the rest, none at all. There is a huge difference between a bay that is mostly 2m with occasional rocks up to 1m and one that is a solid shallow 1m. Whereas it's just not important if, out on the high seas, depth goes from 3876 to 4124 meters. Ok, there are a few exceptions: if you're fishing, then the location of the continental shelf and seamounts could be important. Also, if you're looking for a burial ground for Osama, then you need to find a really deep spot. Those are the only two exceptions, and one of them is no longer relevant. Notice that oceanographers and submarine captains are not among the exceptional cases. If you're commanding a submarine, I hope you splurged for the extra charts package at the rental counter, so you'll have good charts that cover everything all the way up to crush depth. And if you're an oceanographer, you'd better be using different maps than the ones I use for navigation -- that's like an astronomer trying to use my toy stereoscope from third grade. If cartographers had not lost their sense of purpose and audience, they'd have learnt to use a log scale to present lots of detail for the shallow areas. Because that's what normal people need and want. Stop plotting irrelevant data like a first-year grad student.

If this is not part
of your route calculation,
you're doing it wrong

Fourth, maps are used for travel over time. The world changes over time. Why can't the map take the passage of time into account? I would think that helping someone get from point A to point B is a pretty important part of the reason why people use charts. So here's a trick question from left field: What's the fastest path from point A to point B? I know all you budding software developers who are fresh out of CS100 are itching to use the PlotLine function. And I know all you seasoned developers are reaching for the PlotGreatCircle function. But because a substantial percent of people who ask this question are going to be sailing, the right answer depends on an extended weather forecast. For a sailboat, you have to simulate the boat's trajectory over time and take the weather at the boat's position into account to determine the route. This is pretty difficult for a human, as you have to integrate the boat's speed to track its position and simultaneously query a meteorology database for the conditions at that location. This is precisely the kind of thing for which we invented computers, so get to it already! If you think this is too niche a request, how about just changing the satellite pictures to reflect seasons? Those super-clear pictures taken on a crisp, cloud-free day in October are no good when trying to figure out if the hills are going to be colorful in June. And this actually matters to people who want to figure out how a place will look when they visit.

What is this? I don't even

Finally, why is the default screen centered on Wichita, Kansas? I've never been there, and I plan to, with a bit of luck, die without ever having to visit. Even the people who live in Wichita, Kansas couldn't care less about Wichita, Kansas, or else they'd agree to pay a bit more in taxes and make it a better place. Put the darned center on top of someplace I could possibly care about (hint: red states are definitely not in that set). Even Milet, Turkey would be a better spot.

Some might say that I'm being unreasonable, that these feature requests are too difficult or costly to implement. To them, I ask: is it really much harder for you to add these features to Bing or Google using your 3.0+ GHz 8-core desktop than it was for Anaximander to invent infinity, theorize about evolution, successfully predict an earthquake, and draw the first world map?