Friday, January 19, 2018

How To Lie With Maps

My wife gave me the book by Monmonier for Christmas - she likes to try and find things not on my list if she can. It's a solid book, one that you pick up to get a Cartography 101 course. There were things I knew about maps just from frequent use that I didn't know the names for, and close definitions usually give you quick, improved understanding of what you learned informally. So I'm sailing through the first 40 pages happily, even though there aren't many of the fascinating anecdotes one hopes to get from a book with such a title. Mildly challenging, head-nodding "yep...yep..." sort of stuff.

On page 41 he shows me something I have never noticed, but instantly recognise as a way to lie with maps that can be powerful: the ranges in the legend. Such as this one, which I came across a day later. (Click to enlarge)

By Bill Rankin — Citynoise (talk • contribs) - Own work, CC BY-SA 3.0, Link

I almost missed this, following my little information trail after Grim mentioned the general topic. Just as I was about to click off I checked the population density legend and saw, by golly, exactly what Monmonier had written about. Those are curious adjoining ranges in there.
Less than 150
Over 7500

I don't say they are wrong or deceitful.  They may be the cartographer's best compromise to illustrate density.  For all I know, those ranges may be a common convention used worldwide. The downward track in the later ranges, 200%...167%...150% seems sensible, as do the Less than/Over cutoffs. Other numbers might have told the story as well, but these do the job. If "less than 250" had been chosen for the first color, then the second would be 300% higher, fitting the trend above. Perhaps Rankin though it important to add that extra level of distinction at the lowest level. One of Monmonier's repeated points is that all maps lie, because they have to suppress some information in order to highlight other information.

Yet I had unconsciously assumed the legend marked some regular interval.  Bad assumption, even with honorable, skilled cartographers. Cue Batman: "Imagine what this weapon could do if it fell into the wrong hands, Robin."

Additional notes: The only other map with NE at the top is the Appalachian Trail map, which covers an expanded version of this area. The 45-degree difference really does show the terrain and city connections differently. Another way to "lie" with a map, by telling a more important truth.


Grim said...

Program a map to display frequency of data exchange, every thousand megabytes a single pixel on a very large screen. Manhattan and Atlanta burn solid white. Then they start to pulse, the rate of traffic threatening to overload your simulation. Your map is about to go nova. Cool it down. Up your scale. Each pixel a million megabytes. At a hundred million mega bytes per second, you begin to make out certain blocks in midtown Manhattan, outlines of hundred-year-old industrial parks ringing the old core of Atlanta...

William Gibson, Neuromancer.

james said...

Yep. Guide the eye by removing the clutter. Adjust the end of the scale to display what's interesting.
If you want to focus on the population centers, you don't want distracting shades of small and smaller out in the countryside, so tweak the log scale. If you want to focus on the countryside, shift to a linear scale and let anything town or city be hrair.

Sam L. said...

James! Last word: hrair? What that means?

james said...

JMSmith said...

I'm a regular reader and very infrequent commenter who happens to be an academic geographer. In fact, I completed my PhD in the geography program at Syracuse University, where Mark Monmonier was on the faculty. Last I knew, he was still an active emeritus professor. I knew him, although I never took one of his classes. Everyone admired him for his very high productivity, which ran to about one book a year, but graduate students also joked about the sign that was almost always posted on his office door. This read: "Not Now, Please!" The big professional association for academic geographers in this country is called the AAG, and sometime in the 1970s they decided to run their annual meeting by open submissions. That means anyone who pays the registration fee can present a paper, no matter what. Mark Monmonier once told me that he was the only person to have his abstract rejected since open submissions became the rule. His rejected abstract stated that he was going to do an impersonation of Liberaci, and asked the Program Committee to be sure that the room be equipped with a grand piano and candelabra. I have no independent confirmation of this story, but Mark doing such a thing would be entirely in character (whereas telling me a tall tale would not).

The title would not have worked so well at selling books, but I think it would be more accurate to call HTLWM "How Not to be Deceived by Maps." Basically this comes down to understanding that reading a map is harder than it looks, and that cartographers often take advantage of the fact that most people think that reading a map is easy. Learning to check the data interval on a choropleth map, such as the one you reproduce, is a step in the right direction. There are often very good reasons to use an irregular data interval. There are also underhanded reasons. It is always worth knowing that the map in your hands was drawn by a deceptive cartographer, because knowing this make the map into valuable evidence of a different sort.

In any case, thanks for your blogging and keep up the good work!

Assistant Village Idiot said...

I remember you commenting before, Jonathan, on the post about the geographic feature in NH that used to be called "Nigger Nose."
It was a great comment, BTW, as is this one. Thanks for the anecdotes about Monmonier.