The Death of Cartography and the Rise of the Machines
This article explores how technology has influenced map making and how cartography requirements have changed, and what skill sets one needs to make the modern map.
Cartography: Back in the mid 1980’ in geology departments hand drawn maps were given to the drafting department, who meticulously drafted the maps onto transparency, copies were made and one spent weeks colouring them in with pencil crayons. Five years later the computers and A0 printers came; the drafting department started shrinking, and today has disappeared. Being replaced by head office GIS-geologists that fulfill a coordinating role of data “QA/QC”, which is more of an IT role. Most industry professionals generate their own maps and technologies like GIS are nothing more than a tool of the trade. If one need a flood map made, get a hydrologist who use the tools of the trade.
Cartography had gone from trained drafters/cartographers to professionals from all walks of life. Producing detailed and artistic maps with perfectly designed outlay has fallen by the wayside. On local television they used to give detailed weather forecasts with isobars and standard 18 character weather station symbols. Today one is shown an icon graphically depicting percentage cloud, sun cover and animated rainfall. Watching CNN the other day they came up with a breaking story of a 7.8 earthquake on the Iranian Pakistan border and showed worst map I have ever seen, a big red dot in a desert area (satellite image) with no identifying borders or towns locations or legends. It is rapid quick information exchange; map is put together in less than 5 minutes.
In the company board room company directors and decision makers want to explore and question the information by panning, zooming, adding and removing overlays. They don’t want a north arrow or scale bar, but a distance and area measuring tool. The A0 printers now stand idle. There is no time for the finer cartographic touch-up’s. When it comes to the cartographic design of maps I apply the Pareto (80/20) principle, gets the map to 80% correct using only 20% of the time and further effort is not warranted. Fiddling around with colour contrasts is redundant when every 5 minutes it is let’s try this overlay or add this and remove this!
Coding: A bit more than a year ago while completing my masters in GIS & RS I was on the lookout for GIS opportunities. I came up with a very long list for example while attending a course on designing sediment catchment dams. One need to refer to about six data tables including a rainfall intensity contour map to get all the parameters for the calculation, which are then processed in a spreadsheet, how easy would that be on a web-GIS. All these small industry specific online GIS applications that are being developed are such that a clerk can operate them. Where does that leave the GIS analysis (non programmer)?, as the piggy in the middle trying to catch the odd ball bouncing past. Is there any wonder why the vast majority of advertised GIS positions require some coding ability?
Software is going through another evolution. Even Microsoft whose software is based on easy to use GUI (Graphic user interface) has now gone back to command prompt and scripting methods of configuration. For example if one want to add a name to an outlook server it is done though a PowerShell prompt, because all configuration functionality of the software can’t be put into single GUI window. The same applies to a modern GIS where there are too many options to put them all into menus. I often use ‘Jscript’ in ArcView to customise label selection and format the output.
It is a fallacy to think because a person can do computer coding they aren’t creative enough to make good maps, look at the computer gaming industry real-time interactive war game where the limits of spatial interaction and graphics are pushed to higher levels. It was after all computer gamers that started Google Earth, which has revolutionised GIS. Gone is the fancy locality map as everyone has interactively explored the project area on ‘Google Earth’.
Do Programmers make better cartographers? Yes for the following reasons:
- Better able to utilise the full potential of software and add customised features;
- Can produce dynamic online maps where the end-user has control over which layers they view, and can change colours and formatting of objects to their personal likes;
- Embed data tables give the user more information,
- Digital online maps have much better formatting and contrasting options ( e.g. http://maps.stamen.com/ )
- Automated methods generate the maps quicker getting the information to the user when it is needed.
I have a strong opinion that spending the time and pain that is required to learn to code will reap long-term career benefits and likely become a mandatory future requirement for GIS professionals. Likewise I encourage graduate and students that come through the workplace to learn to touch-type; when I see them using the mouse excessively and only typing with 2 fingers. I get a lot of negative response like ‘I am not a secretary’, but if one is going to spend the next +40 years of your working career behind a keyboard why not learn to utilise the system to its full potential and get the touch-typing and scripting skills.
Conclusion: There is a fundamental shift in software design requiring uses to interact with software through command prompts and scripting, and online dynamic maps is now what industry requires. Formal Geomatic courses should teach both cartography and programming, as both are essential skills, and being able to code does help one produce better cartographic products.
“It is a fallacy to think because a person can do computer coding they aren’t creative enough to make good maps”. AGREED!!! But we must also concede that just because someone can write GIS code doesn’t necessarily mean that they can make meaningful maps!
Not everyone can create good maps on the fly, even with the best of software. There still is an art form to it that can make a meaningful map, at the right scale with the right context. I’ve had way too many Managers who ask me why they cannot create a map of the entire Province or State, show the 500,000 points of interest they want, show the water layers and put it on a postage size plot or PDF. Then they say, well, isn’t it digital data, just shrink it down…. They just don’t get it!
I disagree with your assertion. Programmers can produce graphics, not maps. If I give you some base maps at 1:20,000 for UTM Zone 10 NAD83 datum and other land use maps at 1:2,000,000 in Lambert Conformal projection NAD27 datum, a programmer could mash them together, but could the progammer tell you the limitations of the maps, and the problems with NAD conversion in Canada? I don’t think so.
Anybody with a hammer and saw can build a house, but I’d rather buy my house by someone that has been trained in carpentry.
Well put Karl. I once saw a map put together in ArcGIS by someone with business expertise, but not cartographic expertise. The north arrow wasn’t even pointing north.
Being a true cartographer myself I have to agree with Andrew on many points. That’s why I had decided for myself that sticking to cartography and even GIS alone will probably leave me with few career options if I really want to go far. Here in Germany you will nowadays often find job ads looking for people called “Geo-Informatiker”. An Informatiker in Germany is what’s called a computer scientist in English language and obviously the geo stands for a deeper understanding of earth sciences like cartography or geography and obviously GIS. Many universities here have now come up with degrees in Geoinformatik and therefore dropping plain cartography or geography. This were the trend is going and this is where I agree to Andrew.
Where I disagree is that coders are the better map makers. They may be for many purpose driven applications that create maps which are needed this minute but will be gone tomorrow. I strongly doubt that anyone who is not a cartographer (or from a closely related science) would be able to produce any of the maps published in an Oxford atlas of the world, edit a swiss topo map or make any of the beautiful maps that National Geographic in the US publishes. These maps are made by true artists who not only are artists but also have to know all of the science that cartography is based on. Real cartography is becoming a niche product that is true indeed, but it will never die.
I have just seen too many bad examples of maps where people with no cartographic skills produce maps which may even lead to false conclusion. The classic error is that GIS data from largely different scales are put together or put on a map in a scale where they simply should not be used any more. Other poor examples are often seen in online or GIS maps where placeames appear just because the system reads them off the database but that are not verified by a human any more.
Teaching “geo” people to code is a good start but also not the ultimate solution to me. You should learn about algorithms before you even think about coding. The world of geo and IT are growing together very fast which leaves us with the dilemma that a “geo” is not a computer scientist (and therefore doesn’t really understand computers) and the computer scientist is not a geo (and therefore doesn’t really understand map making or geography). My conclusion to this was to add a bachelor’s degree in computer sciences to my studies and I can really recommend this to others as I think it is a good combination to really understand and work with the best of both worlds. This may also not be the ultimate solution for anyone but it is probably not a bad idea.
I bake bread at home but I don’t say I am a baker, I grow a garden but do not consider myself a farmer, hem my own pants but do not think I am a tailor, I use a GPS, does that mean I am a surveyor? I am a programmer so does that mean I am a cartographer too? I use ArcMap sometimes so I guess I must know GIS too then 🙂
Angela, If you are producing graphic output using ArcMap you are doing cartography. Skill and competence as a cartographer is another issue. If you grow produce you are farming the land that is available to you!!
One of the first ‘GIS’ maps ever made was by a medical doctor, John Snow 1854 (http://en.wikipedia.org/wiki/John_Snow_(physician)), one does not need to be anything to make a useful map.
The point being made is that if a cartographer learns to code/script they would be ‘richer’ for it and the skill would assist in producing better output.
Many have touched on the purpose of maps and GIS. To me the role of GIS is to make complex data ‘digestable’ (easier to understand and read). That is where the “ART OF MAP MAKING” is critical to present the “ANALYSIS” in a clear and digestable form for the whole audience. Not all people can read maps and not all people can read tables of data and associate the table with the real world. That comes down to the “ART” of an artist that used to be know as a cartographer. An artist with coding capacity and a scientific/analytical mind is set to flourish from working in GIS.
Yes there are many interactive maps and decision making maps that are harder to interpret that they could be. But that is life and the current status is still better than static maps or delays in information. Long live ‘Live Data’.
There is no thing more contsant than change itself. We all have to change or become stagnant in our careers. If your mind is that of a cartographer then you are under the time squeeze. If you are a manager with too big an ego to take the time for the best decision your failures will take affect on others as well as your career.
The solution is “horses for courses”. Meaning there never was and never will be a single mapping solution that fits the needs of every body. It is up to the audience to go somewhere else, or the “Artist” to “Sell” their story and managers to “hear” what is being said.
Cartography is dead. Long live cartography!
The debate about GIS vs. cartography is not likely going to go away soon – so I won’t mention it again, promise. The answer to “what is cartography”, however, is changing and your article kind of misses artography actually is. Your description of maps created by drafting departments in the mid 1980s gave me a good chuckle – but that wasn’t ever cartography – it’s called digitizing (and it’s a perfect task for coders to automate). I have to agree with Karl – just knowing the difference between a Chardonney and a Cabernet doesn’t make you a sommelier (Karl, I couldn’t resist the wine reference).
I’ll put my neck out here, but I think Andrew’s statement “Death of Cartography” really refers to the digital re-birth of “map-making” or the “wilting of cartographic ivory towers” or “Dr. Stange Maps – how I learned to live without a north arrow”.
The article is a little distracting in that it uses basic examples of how a talking head on CNN might portray an event via a nonsensical map, how sets of data layers are used for decision-support in a fast-changing environment and why the Pareto principle is king in today’s map-making environment to support an argument that ‘GIS coders’ make better ‘cartographers’. In fact, these points illustrate perfectly the missing distinction between analysis and the presentation of the analysis.
In support of Andrew’s point, a person with the ability to manipulate data (especially big data) programatically will have the advantage of producing products that can be analysed, customized, discussed, integrated, migrated, etc. to suit client needs. However, Andrew’s thesis becomes quite fragile when it comes to properly presenting the analysis – the ability to write good code doesn’t necessarily come with an inate ability to present/portray information in a meaninful and useful way (a quick search will deliver buckets of fantastic / horrible examples).
BTW, the USGS (an early adopter of GIS and it’s related technologies) is one of the world’s biggest consumers/producers of spatial and geological data and, by far, one of the most technologically advanced implementors of automated mapping and data delivery systems. If you were to look into how their map production has changed over the years, you’d be quite surprised by the roles the “QA/QC-ing” “head office GIS-geologist” had within the Survey, and still do, to this day.
Maybe you concluding statement should read “being able to code does help one produce maps quicker.” 😉
Cheers.
I do agree with Andrew on some points but do not on others.
I believe the difference now is based on the use of the information.
In the past the cartographer determined the content of the map largely. This is no longer the case. The content of the map is largely determined by the business user, that could be government, business etc
Maps are no longer the final product. The map is a component used to solve a business problem.
As such if the business problem is answered with a cartographically poor map, the end user does not care.
A great example is navigation systems, the user wants to find the easiest, quickest way to get from point A to B. If they are happy with a voice telling them when to turn etc and it achieves the result, then the problem has been solved. The map of the screen is only there as a guide and often they are very poor maps.
As an fresh Geomatics Engineer, amidst so much of experienced giant I have a simple query. If I have changed my mind to add some programming skills to my brain, which language would be better? Java, Python, . Net languages or any other?