Ken Field from ESRI is working on a book based around 100 maps made from the same data set. Learn about how to think cartographically and the kind of strategies required for experimenting and successfully making maps with impact.
(light electronic music) - So now it's time to bring in Ken Field who is Senior Product Engineer at Esri, and Ken spends his time thinking about, writing about, talking about, and of course, making maps. So Ken, thank you very much for joining me here today. - Hi Bill, yeah good to join you. - So Ken, our audience is extremely broad, we have business people, executives, non-profit and NGOs, we have HR staff, people from all demographics, roles, titles, et cetera, some of whom are probably students of information design, all of whom have used maps to get from point A to point B which is probably their primary sort of conceptualization of what maps are for.
We also all of course have GPS in our pockets and maybe know less about maps than we should because our maps just talk to us and tell us what to do. So maybe we can just start off on a very general subject here and just ask you to share with us, what do you think is the most important thing our audience needs to understand about maps and cartography in this context, of data visualization, data storytelling, et cetera? - Yeah sure, I mean I think it goes back to maps are one of the earliest forms of visual communication.
We go back to hieroglyphics, stone tablets, stick charts, and various other forms of representing the world around us that people have always used in order to navigate or to understand the world around us. And I don't think that's really changed. We go back thousands of years or 10 minutes, people are using maps to do all sorts of things. The old adage of a picture is worth 1,000 words I think is probably relevant because there's nothing really better than actually looking at a picture to try to unravel something about what it is you're trying to understand.
With that said, I think that the picture is not that effective unless it is well composed and the person reading it has an ability to decipher it and understand it. So I think there's challenges for cartography and for visualization more generally in codifying these images and allowing people a better way of trying to understand what it is they're looking at in a simpler, easier fashion. - Thank you.
So we do see maps out in the universe all the time as you said and we're constantly seeing them online, we're using them, et cetera. What would you say are some of your pet peeves about maps that are out there in the universe? - I don't have any, everything's good. - I am sure that's not true. - So, I think, I think in any profession, okay, you have what might be seen as experts and it doesn't matter whether we're talking about making maps, law, medicine, whatever it is.
We have certain professional level of expertise. And those sort of professionals have, to some extent, gone through an education, a training, life experience in working with whatever it is that they're doing, and they've come to a position where they understand quite well best practices, they understand quite well how to effectively use maps and the signage on a map for instance, and they've come to understand what works well and what necessarily doesn't work well.
And the problem, or my biggest pet peeve, is we've almost been too successful because mapping has become so widely used and so widely constructed by almost anyone, but possibly we haven't necessarily articulated what works well and what doesn't work so well, so I see a lot of maps that are constructed by people with very good intention, but where I can perhaps see some very common pitfalls that you wouldn't necessarily know unless you study cartography, you've gone to a few classes maybe, you listen to somebody with a little more expertise.
So for instance, and I don't want to particularly pick on a particular group, but the scientific visualization community are renowned for using less than optimal colors on their maps to illustrate some pattern of natural phenomenon. We might commonly call these rainbow colors, the sort of traditional ROY G BIV colors of the rainbow to show perhaps empirical data from low to high.
And whilst they're very pretty and they're very attractive and people naturally gravitate to colorful maps, cognitively they do not work as effectively as another color scheme might because it's very difficult to know where on the scale red is in relation to yellow or to green, which is higher, which is lower? What do these colors represent in terms of data values on a linear scale. So it's not that, there's very little that's right or wrong in cartography, it's simply how do, how would a cartography apply best practices and do we see those best practices reflected in what people more generally are doing? And I think that to me is a pet peeve, we need to close that gap and better educate the lay cartographer to actually implement some of these best practices a little earlier in their cartographic career than perhaps just relying on defaults that people have used for decades that really aren't that optimal.
So that's just one example. - That's a good example, and color is hard. Certainly the rainbow map is a great example. I wonder if another example, I was going to ask you about projections. For those of you who don't know, the earth is a sphere or pretty close to a sphere and when you create a flat representation of it you have to use a projection, which essentially represents that three dimensional spiritual thing on a flat plane. And there are a lot of different map projections out there, most of us end up creating maps using the Mercator Projection and I'm sure Ken is upset right now even hearing that a little bit, not that he doesn't know it already.
So Ken, I guess I was going to ask you, will there ever, will the questions of projections ever be settled? Is the equal area projection like the savior of humanity or is it just a constant thing in flux that we'll always be debating and it depends on the context of what you're trying to accomplish? - Right, I think you're asking a very typical question and it's a great question we always get. What's the best projection? There isn't an answer because different projections suit different purposes and that is the point of a map projection.
You know, the interest thing is my mathematical background isn't particularly strong, so I'm maybe the last person to talk in detail about projections to you, but I look at projections in an applied sense. It's a matter of choosing a projection that fits a particular purpose. So you pick on Mercator for instance as a projection. Now this might surprise you but there's nothing wrong with the Mercator projection or the web equivalent, absolutely nothing wrong at all.
It's a perfectly useful projection, and if I was in a small boat navigating the ocean I would want that projection because it preserves angles and it allows me to navigate rom one place to another. So that is what we might call the special property of a projection. So it's not necessarily the name of a projection and we can't say that Mercator is good or bad, it's the property that the projection holds that allows you to use it for a particular purpose. So if we look at Mercator and its use for a lot of web mapping and data visualizations, it's used as a base map because it's the default in many many software systems and without going into detail there's perfectly good reasons why Web Mercator became the default, because it's computationally a well organized way of organizing the basic tiles of a map on a computer screen.
But if we're creating data visualizations of population information, and that might be median income across the world, it might be something to do with poverty, it might be whatever, unfortunately the Web Mercator projection has not got an equal area property. It isn't equal area. That's why you see the classic example of Greenland being massively distorted in the northern latitudes and also the southern latitudes are massively distorted.
And so it's not optimal, because if you place population information across the top as a thematic layer, either it's a choropleth map or a dot density or something, that massive exaggeration of size, north and south or increasingly north and south of the equator has problems cognitively, because human beings are going to look at that map and they see big as more important, that's just the way we naturally look at things. And certainly maps. So you unintentionally get given a story that exaggerates certain countries over others, and by implication the data that is displayed in those countries over others.
So you would choose an equal area projection or a projection with an equal area property to better display thematic data. Now fortunately that's becoming a lot simpler, many many many different online mapping packages and mechanisms are allowing you now to switch the projections. So again, it comes back to the point I made earlier, it's more about not just saying to people well, go with the default, Web Mercator, but perhaps pick something that is better suited to the map and the story that you want to tell.
I think you asked about Equal Earth Projection. It's not the savior, but it's a really good projection. There's a lot of value to it and it's a great option if you need an equal area map to hang your own data visualizations onto, yeah. - Yeah, yeah so you bring up a really good point which is the answer to every data visualization question ever, it depends, right? - Right. And if there was a right answer, then there's no point in any of this any more, you just follow a recipe, right? - That's right, you just press a button and it just happens.
- Right, but that's not the case, there are always alternatives, options, slightly better, slightly worse. - Depending on what you're trying to do and say. - Yeah. - Yeah, so that's actually a great segue to talking about your new project, which I saw you talk about at Tapestry Conference. And I know you're writing a book about it. For the audience, Ken has created, or is in the process of creating, I don't know if you've finished them yet, but 100 maps based on the 2016 presidential election in the United States.
And so each map is a different approach, you have a choropleth, you have cartograms, you have all these different variations of map styles for mapping that thematic data. And each one tells a very different story, and so it's a great example of that it depends conversation. If you're trying to say this, this map approach works. If you're trying to say this, this map approach works. So Ken, maybe tell us a little bit about that project and about the book and when you expect it to come out and anything else you'd like to share about it. - Sure, so as a Brit I moved to the states in 2011.
And if you move to another country, it's kind of beholden on you to understand a little bit something of the country and what's going on. So I started to look at the differences between the political systems in the states as opposed to where I'm from in the UK, and it seems to have gone awry lately in both countries, but that's another matter. But it got me to really look at the data and the stories that you can tell with the data.
So I actually started this with the Obama Romney 2012 election data and just started mapping it. And tried to go beyond the traditional blue state red state maps that we very very often see. And see what we can experiment with and how far we can go with making the map. And I've long held this belief, and I used to teach students this, that very very rarely is a map wrong. And very very rarely could you look at a map and say that is right.
So there's no right or wrong, there's kind of gray areas, there's better or worse. And of course, those better or worse maps can be used to fit a particular narrative, tell a particular story, suit a particular base, if we're talking about political maps. And they can if you take the natural extension to that, be used for persuasion. And if we look historically, maps have been used for propagandist purposes, if we want to take it to its natural sort of conclusion.
So I began to just create these maps, and then with the 2016 election I updated a gallery of maps that I had prepared and I made maybe 30 different maps from the fairly traditional as you said, the choropleth, proportional symbols, dot density, very very long established techniques. And the purpose of doing that was to demonstrate how you could employ that technique with that data to create an effective map.
But then you start to get into the more interesting techniques, you mention cartograms. They've got not quite as long a history as let's say the choropleth, but nevertheless, there's good value to be had in certain cartograms. And they perhaps suit different purposes. They're very attention grabbing, you could put them on the front page of something and people would stop to look at it, whereas a choropleth is perhaps less visually arresting.
It doesn't capture the same attention that a cartogram might. But you could argue that maybe it's more easily read by a greater proportion of the population. So different maps for different purposes here. And what you find when you start creating these maps is you can create a continuum, where on one end of the continuum you might have a map that is ostensibly a very blue looking map. And then on the other end, a map that is very red looking. And of course, you might use these different maps in different ways, and I do recall in the aftermath of Trump moving into the White House, there was a very famous photograph of a staffer putting a new map on the wall.
And it was a choropleth and it showed the percentage vote for either Republican or Democrat party, and it was a very very red map. And there was a lot of frustration I read online about this map, saying oh it's terrible, it's wrong. No it wasn't wrong. If I was a Republican president that's exactly the map I would put on the wall in my office because it speaks to my victory. And if Clinton had won the map would have been different based on the same data.
So I've been working on these maps after relative success of this first book that I wrote that was published in the middle of last year. I pitched another idea at the people that work at Esri Press, the company that I work for. And they thought it was a go, so it's going to be a book of 100 maps. It's actually not a new idea either, very little in cartography is actually new.
It was an idea that Jacques Bertin actually utilized in a very famous book called "Semiology of Graphics" where he took a data set of different land use types and created 100 different ways to visualize it, again as a way to reflect the rich variety of cartographic work and approaches and how you can work with the data and tell different stories. Sop that's where we're at, currently about 1/3rd of the way writing it through.
Probably about half to 2/3rds of the maps are finished. Hopefully it will be out before the end of the year, but I wouldn't place a huge bet on that, it's just a question of finishing the work, getting the maps done, getting the writing done and going through all of the editorial processes and fact checking. And it's important to get this right. And then really launch it as a way to educate people.
There's no right map, there's no wrong map, it's a very interesting data set, it causes people to be very passionate one way or another and hopefully it will show people the variety and almost the palette of maps that they can choose from when they're thinking about telling their own stories. - That's great. Maybe just in closing, if you could just share, and this is a tough question maybe, or maybe it's really easy, what's the most important thing happening in cartography today would you say? - I think it's the way in which mapping has become democratized, you don't have to go back too far when most of the maps that we see are drawn by, let's say national mapping agencies.
And others of similar ilk. But everyone's making a map now, absolutely everybody. You can make a map just by placing a pin on a web map, Google Maps or something. You've made your own map, you've added something. And I think that's the important thing that we have to recognize, that the mapping is no longer owned by major societies, professional learning societies, national mapping agencies. And sure they've still got a very important part to play.
But mapping can be done by everyone. It always has been, but perhaps the balance has shifted and it's more so now. And with that comes great responsibility because the key thing to make sure your map does is to inform and clarify something. You don't have to necessarily oversimplify, but to clarify something that you're trying to inform somebody else about. And that's important. The other major thing I think that is part of the same democratization is, you used to buy a map and it was a sheet, a map sheet.
It may be a reference sheet or something. And naturally, the map had edges. And the center of the map was sometimes arbitrary, because it was based on a network, a grid that people use to create a map series. Well you're now at the center of the map. The map is based around you on mobile phones, our cell phones. We walk around and you're constantly in the middle of the map, so it's now the case that we have almost an edgeless map, and you are now central focus, you are now the object on the map around which everything else is positioned and referenced.
And if we take that logically, some of our smartphone maps actually provide us with personalized information related to the sort of things that we do, the places we go, the restaurants we like, the hotel bookings we've got. So maps are becoming more personalized. They're becoming less detached from us as individuals and more personalized with more personalized content around us. So I think it's really exciting, there's lots of really interesting things happening in cartography, and they're largely based around this shift from a traditional more objective way of mapping to a perhaps more personalized form of cartography.
- I know that I love maps, I really do, and I now do spend my time in maps. If I'm traveling, Google Maps is my open app on my phone 20% of the day sometimes. So it's a great topic, I really appreciate you coming here to talk to us about it today. - Sure. - We only touched the surface of this topic but it's a great foundational conversation I think to get our audience thinking about cartography and mapping and hopefully we'll all be able to bring this back as a topic in the future and go a little bit more in depth.
But Ken, thank you very much for sharing your expertise with our audience. - Yeah no problem, it's an absolute pleasure. Happy mapping.