Friday 8 March 2013

The fallacy of new cartography

Perceived wisdom has it that the web is the new medium for just about anything, maps included. The map-making process is in a constant state of reinvention. It always has been. Even in recent history, professional cartography blamed GIS in large part for decimating their industry; then GIS improved their wares and making decent maps became possible. Now, some would have it that GIS seems to be under threat (attack?) by tech-savvy web developers. The map is everything. It's front and centre. In this blog I want to offer an opinion on how I think cartography itself fits into the scheme of things.

Javier de la Torre gave a presentation at the SIGLibre 2013 conference in Girona this week that prompted my thinking; ideas that have been swirling for a while and which have surfaced in previous blogs in parts. I didn't see it but his slides are online.  I've given similar presentations on the democratization of cartography before as have many others.  It's becoming a familiar story at many geo-conferences...and the story goes that new tools taking over the mantle from GIS and that they are allowing for much better cartography.

While evolution is important I find myself disagreeing with some of the rhetoric when it comes to what cartography is apparently becoming...

There's no doubt we're witnessing a profound technological revolution but does this necessarily improve maps and map-making. No. Whatever technology has underpinned map-making we have always seen people harness it superbly.  These examples become widely appreciated and lauded. I can point to some stunning maps made using GIS (and previous technologies). Likewise there are some terrific maps being made by people using the current crop of web mapping tools. So why the claim of a new cartography? Well we're dealing with a revolution in who is making the maps for sure, but it's not democratization, it's a new geo-silo. Many new map-makers are new to cartography (but they love geo and maps like we all do); are discovering it through experimentation; and pushing out their work in a medium that can be consumed instantly by huge numbers of people. Many maps made by GIS never got that exposure so people are unaware of them...but that doesn't mean they were never made.

Are all "cartographers" old-skool; unwilling to move with the times? No. Of course there will be people who refuse to move beyond their beloved version of Freehand (for instance) but there are many more who do shift their skillset and take on board new technologies.  So it's a little invidious for new mapmakers to claim they are a revolution distinct from "cartographers".  I compare it a little to a first year student claiming that their work is novel because "they couldn't find any references". What that really means is they didn't look hard enough, in the right places; or didn't know enough about what came before to make an informed judgement about its value to their work. Knowing the history of what you are doing is important. It breeds respect not only amongst the community but also for the work you are doing. We are all cartographers. Some newer than others but that's the only real difference.

Are the examples we see in web mapping that much better than what went before? No. There are new, disruptive ways that are being harnessed to visualize data it's true (some of which are really fantastic!) but as a proportion of the web maps out there, the number that are truly well designed is low. Some of them are beautiful and have a strong aesthetic which is why they capture attention and wow us...but then I could point to similar maps made using copperplate, Rotring or GIS in the same way. Map art has a strong pull factor. What is truly different about web maps to what went before is twofold...the map is no longer one map; it's many maps. A different map designed for each web mapping scale that has to work at that scale and also transform well between scales. Second, interaction now allows  a number of important possibilities from popups to moving map detail that are really only just beginning to be used effectively. New technology demands a different approach to map design but it's nothing new to cartography. Good design takes advantage; bad design makes mistakes. We didn't have to wait for web mapping to come along for Harry Beck et al. to have the ability to create a great map. All that's happening now is we're waiting to see who the current Beck's of the mapping world are going to be; what their innovative work is; and how it will shape our cartographic thinking moving forward.

But let's be clear about the way in which we narrate the shift we're seeing. Web technologies are moving rapidly and the pace of transformation for map-making is phenomenal.  But pointing to great current web map examples as evidence that they support great cartography while at the same time pointing to pretty poor maps made using GIS; or awful geoportals; or god-forbid a "print" map does not stack up. Paper or's just different. Similarly, many of the so-called 'new ideas' for visualizing data are not, in fact, new.  Yes, they are now much more easily implemented and the result more rapidly achieved but let's leave it at that. It's a new technological implementation, not a new cartography.

And what of the idea it's all free? Well it is, to a point. CartoDB is a point. MapBox is a point. ArcGIS Online is a point. They you pay. You pay because actually, free is never free. Hosted cloud services are not free. Servers are not free. People's time is generally not free. Sure, the geocommunity has driven all sorts of open agendas forward (a good thing) to create an empowering open community but behind it all is something, somewhere, that has to be paid for. We mustn't forget that else we'll become far too complacent. Having access to rich data sets is possibly the biggest boom of the moment. It's true that data is now more readily available but really all that has done in a cartographic sense has shortened the path to the map.  It doesn't obviate the often time-consuming effort required to reformat the data and get it map-ready.

How do you deploy the new map making tools? You code. Which means you need to know how to code. Which means if you want to create great work, just like knowing a GIS inside out to produce a great map using GIS, you need to be damn good at coding. It's a new map-making method. Well, actually, it isn't. I created a map using code in 1989 before desktop mapping GUIs arrived. Is coding for everyone? No. It's not. It excites some and causes palpitations for others. I happen to believe it's important, particularly if you are entering the industry but it's not how everyone wants to make a map. It's not in everyone's skillset. My skillset in this area is not as good as some of my new map-making colleagues but I'm willing to learn new ways so that my cartography harnesses the best of what is available to me [ web maps have to use the tools my employer produces; it's part of my job...others use the tools their employer's just different!]

So is new cartography simply about scraping some novel data from somewhere, being able to master web design (HTML, JavaScript, CSS) and then deploy your coding skills in your choice of web mapping tool to create your masterpiece? People point to interaction, user experience, data visualization, analysis and methodology as being differentiating. I'm not so sure. Every map I have ever made, be it on the web or in print has always sought to provide a way for people to interact either physically or emotionally. Even print mapping allows interaction...there's something magical about seeing people touch a map you've made as they explore it. User experience is also independent of medium. Different user experiences are required it's true and maybe this is 'new'; but conceptually, user experience is something that every map-maker should keep in mind. Data visualization? Well, call it what you like but that's cartography when it comes to making choices about generalizing and symbolizing your data to display it in map form. Yes, the web affords new and additional methods but is an interactive pie chart in a web map popup really that different to Florence Nightingale's coxcombs? All maps require sound analysis of the data to make it drive the map properly. So the point here is that the new medium is simply that...a new medium and yes it affords different ways to implement the craft of map-making. I don't doubt the ability of people to harness this new technology superbly...just as others before harnessed what was at their disposal but I don't believe that it fundamentally defines a new cartography per se...just a new cartographic approach.

 Slide 65 of Javier's presentation bothered me...look at the last line:

If I am interpreting this correctly he is saying that cartographers should not necessarily be the people developing the map-making tools or having to know the full technology stack...leaving that to tech folks while they get on with story telling. I disagree. Some of my previous blogs have bemoaned that too may new map makers seem to ignore basic cartography in their work. Their work might be technically impressive but actually, the map sucks. The converse is also true...if cartographer's are reticent to understand modern methods then they can't possibly hope to take advantage of what their is to offer. That said, being a master of both is tough so the best approach is through collaboration. Join forces. Divide and conquer. Making a map using neo-tools is fine if that's your weapon of choice but don't dismiss the value that someone with a cartographic background can bring to your work and vice versa. And from a community that often criticises outmoded geo-silos, if this sort of view is allowed to persist then all that's happening is a new type of silo is developing; one that actively derides cartography, GIS and all that came before it and laughs at those who choose not to code to make their map.

Finally, I liked today's blog How to go viral, every time which effectively said that to make a web page go viral you need content and tactics but, most of all you need luck. Currently, there are a good number of mapping startups and organisations that have lucked-out. They're riding the wave but for how long? You have to constantly reinvent to innovate. Back to my earlier point...maps made on the web have lucked out compared to many predecessors made using different technology simply because of the massive increase in visibility they can muster. Maps made by people or organisations currently riding the wave get a head start because they're the current hot thing. What is the shame here is that there are many great web maps that do not go viral and, possibly of more concern, there are many that go viral that truly do not deserve to. The right map, built right still needs a huge amount of luck to go viral. It's a shame that so much crap gets lucky!

Innovation is vital for cartography. Evolution, rather than revolution is, in my view, the better road ahead. I feel that some would have it that the web has been a disruptive technology for cartography. To some extent this is undoubtedly true. Google Maps has only been around 8 years and look how much has changed since then. On the other hand, I feel that too many have sought to draw a line under cartography and tried to fashion something new from scratch. Don't throw the baby out with the bathwater.

Some asked me to provide a more positive slant on web mapping. I don't think I'm being harsh about it...I actually embrace it and it's been a while since I even contemplated making a map on paper for those that are interested. Change is happening and I'm happy to ride the wave because it's an exciting time for mapping. Bring it on!

Also, some want good examples of mapping. Along with my colleague Damien Demaj, we did some work at the end of 2011 that resulted in two papers in The Cartographic Journal. The latter of these pointed to what we believe are 39 of the best maps ever made; in 13 separate categories that we discuss in design terms. You can see the list over at the ICA Commission for Map Design blog. In truth there are more paper maps than web but we're dealing with a vast repository of maps and web mapping is so new the examples simply cannot compete with what's gone before in many categories. They have a category in their own right but as Ferris Beuller once famously said "Life moves pretty fast. if you don't stop and look around once in a while, you could miss it".  Will the list change...well, yes. Classics are classics but in time some of what we are seeing today will become classics and might possibly replace others as demonstrating the very pinnacle of the craft of map-making.


  1. That whole "we've democratized cartography" phrase always bugged me. As though people haven't *always* made maps until we came along to lift them into enlightenment. C.f. my rantings here in a seminar some years ago:

  2. Thanks Daniel. You make a good point. Has cartography been democratized? What's actually happened is that the "new religion" has just found a means to engage with their flock on an industrial scale through the ministry of the web. Do people have an equal say in cartography? I'm not so sure. They are being fed the new cartography. It's one religion; one form of cartographic governance. I just would like to see its ideology be a little more representative.

  3. Your premise is that technology does not necessarily create better maps is accurate but not profound. Giving someone a pen and compass did not make them a cartographer originally either. One could point out that technology may assist in better cartography and maps through user interfaces, feedback and good defaults. But users are still free to do good and bad things.

    I do completely disagree that these maps are not different. The interfaces we now have available require new thinking and methodology to design, interface, experience, and interactivity. Making a paper map slippy is not interesting or different. Making the map queryable, animated, multi-perspective, and bi-directional is much different than traditional cartographic techniques.

    These new maps should leverage cartographic techniques as well as it leverages general psychology, physiology, and cognition. We should not merely say "everything new is old" but instead realize there is an amazing opportunity to accomplish new and valuable capabilities if we can step out of our restricted past and reconsider from basic principles.


  4. Thanks Andrew. I'm not trying to be profound; merely cutting through some of the spin to offer a realistic counterpoint.
    I agree with you and I guess we’re just approaching this from two sides of the same coin. Technological change is inevitable; evolution is good. Developments in cartographic method and practice are to be welcomed...and I do. What I am saying boils down to a very simple thing...maps have been made for thousands of years. Web mapping has been around for maybe 10-15 years and the pace of change in the last 2-3 years has been enormous. It is, though, immature; there is much work to do and a lot of the current rhetoric seems to suggest web mapping is already capable of everything and more.

    Web maps certainly have a number of characteristics that mean they are different to other map types but it's the extent to which this actually defines a 'new cartography' that I think is sometimes overplayed and acts to create a perceived divide between old-skool approaches and new approaches. The web as a medium creates advantages and disadvantages for mapping and understanding how these are best utilized is vital. The web has been a hugely disruptive technology for many aspects of life but I would like to think of it as being transformative and of benefit to cartography; not as something that drives its own agenda forward at the cost of ignoring what has gone before. There will be ways in which web mapping might develop that we've not even thought of yet. This is exciting. I am not suggesting we become restricted by the past but I don't necessarily agree that starting from basic principles is the right approach. Re-imagining and evolving can move the technology forward but there is much to be learnt from cartographic history that, if implemented well, would short-cut to a web-based cartographic environment capable of supporting high-end mapping as well as the quick and dirty maps that we see a lot of.

    Here's a simple acid test...there are still many things I can do in making my map using other (more mature) technologies that I cannot similarly do in a web environment. For all the 'new stuff' I can do (which is in the main impressive) there are still basic cartographic needs that are not I really have to compromise how I expect the finished product to look and function. That suggests to me that the development of the technology has been to make it work at a basic level and not necessarily work to support good quality map-making. It's a subtle difference of course.

    I’m not proposing a restricted view of the's about moving with the times and bringing along what works; and seeing how we can improve it. Currently, a lot of what I see has taken 2 steps backwards to go 1 step forwards. Any time there is a major shift in the technology that underpins cartography the same is true so it's nothing new and I have no doubt whatsoever that in 5 years time I will be able to make the map look and function how I want as well as be able to properly implement all sorts of new features. I want my mapping to improve as new technology is developed...not to have to compromise by producing sub-par designs because the tech doesn't yet offer what I need. Of course, the counter to this is that many more people have access to tools to get them to where they want to be in making their maps. That's truly empowering. Whether they make 'good maps' is another debate. I'd like to see the tech develop in a way that supports both possibilities.

    I'm hugely positive about change...but also hugely frustrated that my mapping is currently restricted as cartography works its way through its latest reincarnation. As I said before though, this is a truly amazing time to be involved in mapping and there are some incredible opportunities ahead if we tackle them well.

  5. Generally I think that worry over divides and democratization and all that is kind of obsolete. On one hand I do think there is something to interactive web cartography being "new" but on the other hand I don't think we're really talking about that these days now that the initial wild phase has settled down.

    There was a time not so long ago when web cartography went in crazy and perhaps ill-advised directions trying to do something new, but in the past couple of years it has started to circle back to a reality where it knows that it fundamentally grows from all the cartography that came before. A simple example is how we're now breaking free of the Mercator projection in web maps. We may have taken two steps backward, but lately the steps are forward. The limitations are being overcome one by one. Sure, bad maps can still get a lot of play easily, but they don't set the cartographic course so much these days.

    The current charge for "old-skool" types with respect to the "new" cartographers is not to point out that the cartography isn't new but rather to help them actually understand the knowledge and history behind it. We don't want people doing things right based solely on precedent; we want them doing things right because they understand what's right. It's good if people know that Mercator isn't appropriate for choropleth maps, but it's better if they understand why. It's good if people use ColorBrewer color schemes, but it's better if they understand why one works better than another. And so on.

    This is all said as someone who is more or less old-skool but working entirely in the web medium—and on record as an occasional grouch about the modern cartography world. Thanks for all the thought provocation!

  6. Andy, thanks for contributing. I don't find it obsolete because while there are some that "get it" there are far more that don't. What we need to do is encourage those that don't to feel that there might be value in understanding a little about cartography

    You're right to point to some great work in bringing more useful projections to the web. This is to be encouraged but I wonder how many general purpose map-makers will make good choices given a choice? They didn't always do so before the web so what will change things in this medium I wonder?

    My entire approach is to try and wave a flag of help and encourage us all to work together to develop the mapping tools of tomorrow. It has to be wanted to be received though. Knowing why something is right is the ultimate goal but maybe there's a middle ground. Wouldn't it be useful to populate tools with such good defaults that even beginner map-makers are likely to make better choices off the bat? That would be great.