Mapping the Military Past: A Personal Journey
Remember when I said that historians do a lot of work that rarely makes it to publication? What follows is a visually-illustrated history of my experience with mapping history. EMEMH that is.
Way back in the late 1990s, I took a course as a History grad student in Cartography. Even though the professor was horrible, it was one of the best courses I took because he taught us how to use a relatively new technology called a scanner – technically it was a digitizing tablet, because software back then couldn’t handle large raster images. You taped a paper copy of a map to the digitizing tablet, and used the digitizer ‘puck’ (calibrated with a wire grid marked out on the tablet) to digitize points and lines on the map. The resulting line and point positions would be imported into a CAD program (Computer-Aided Design, used by architects and engineers the world over), and you could then add other features to put whatever you wanted on the map. I think one of the assignments was to create a graduated circle map of Australia’s population.
While I had little interest in Australia per se, that cartography course opened up the possibility to make maps all by myself. I recall a friend of mine talking about making his maps by cutting out place names from a printed sheet of paper and pasting them onto a hand-traced map before photocopying the whole thing – remember tracing paper? That sounded like too much work for me (little did I know…), and I’m not artistically inclined enough to draw my own map like something you see in a fantasy novel, so the more I can have the computer do the work, the better. I purchased AutoCAD (thank you, educational discount) and over the remaining years of my graduate career slowly learned how to create my own maps. Here is the tale of my journey from cartographic novice to cartographic dilettante, may you glean some wisdom from it and share your own.
Over the years I’ve learned four things that weren’t covered in the cartography course, but are important for historian cartographer-wannabes to know from the start.
- Most fundamentally, creating your own maps from scratch is critically important because of copyright issues. You can pull features from other maps (especially base maps) since no one can copyright the geographical features or the projections (most projections?); the basic geographical information such as the terrain and locations is common factual knowledge, e.g. you don’t have to ask Rand McNally for permission to place Paris on the Seine River. But a map creator’s particular combination of information and design is itself copyrightable, so you can’t just copy a map from another book and publish it without copyright clearance, although I’ve seen people do that for their dissertations. Contemporary maps‘ copyrights are likely held by the archives the photographs came from (or the possessor, or the photographer…), and they will undoubtedly require payment for reproduction and especially for publication – at the least, your publishers will want you to find out if they will get in trouble for publishing it. In a few cases, devious cartographers have even created copyright traps, fake data that would prove that their map had been copied.
But if you make your own map, it’s yours. Historical information, such as when someone died or when and where an event took place, is common knowledge (i.e. they are ‘facts’) and as such are as copyrightable as where Paris is located, although specific reporting of those facts (say, a newspaper’s account of a battle) is copyrightable. So if you take information from another map and incorporate it into your own, the most that you need to do is cite that other map if it has unique information. Most of the data that a historian will map out is probably drawn from textual sources anyway, so the standard citation rules apply. The flexibility from making your own maps extends to when you publish a map in a book or article: you can still modify the map however you want and it’s still yours to use elsewhere without asking your publisher for copyright approval. Even more usefully, if you created the map yourself, you can add to it however you want whenever you want. Find a new source that describes garrison strengths in a theater? Add a layer for graduated circles to illustrate these dispositions relative to army movements. It’s just that easy. I’ve even tweaked maps to publish in different places, as I try out new ways of visualizing the information I want to see.
Warning: I’m no lawyer, so get some real legal advice about copyright if it’s important to your situation – although everybody says that copyright law is byzantine, and getting more so with every law and court decision. The Internet has made a mess of copyright, making it difficult to determine where certain images come from. To mention a recent example I just came across: anybody know what book this cool map is in? Undoubtedly it’s in some Spanish-language history atlas, which prompts us to remember to look for maps in the books of the countries we are studying, i.e. a Spanish map of the Spanish theater in the WSS will probably be pretty detailed, since it’s their own history. (See my new Citation page for copyright issues on this blog, and how to use my images in particular).
- As for more concrete lessons, I quickly learned that there’s a fair learning curve involved with graphics software. It’s amazing how much time it takes to digitize base maps when you don’t really know exactly what you’re doing (same thing happened with my database experience, come to think of it…). But once you learn, it’s incredibly powerful.
- More recent versions of illustration software (that you usually can’t afford) always seem to include tools to do exactly what you need to do but can’t in your version, e.g. placing text along a curve. And you may have difficulty using the same version of the software over the years as you upgrade from one OS to another.
- Most practically, I learned that it’s particularly difficult to locate historical modifications when you’re dealing with hand-input maps – even if you have maps that show changing rivers and borders, they are particularly problematic to locate on your base map when the two maps have different projections or scales, and especially when dealing with regions without landmarks such as coastlines. Some geographical software can convert between coordinate systems, but that’s a bit more sophisticated than my abilities allow.
Learning these lessons the hard way, I persevered, and an early result of my playing around with maps was on display in my 2000 Ramillies article (completed in 1998 I believe), a version of which is reproduced here:
The map is extremely rudimentary (that was the only font my version of AutoCAD had at the time), but I think it gets the point across – even though the Allies captured multiple fortresses in the Spanish Netherlands in the wake of Ramillies, there were still plenty more French garrisons awaiting them. Despite its Map-making 101 look, it was more cartography than most other history articles had at the time, and military history in particular really requires maps. I wasn’t in a position to pay a cartographer to make a map for me and was probably among the most computer literate of my fellow grad students, so I was happy with the results, if still aware of how far I had to go.
In constant quest for historical maps, I spent the late 1990s and early 2000s searching around on that new invention called the World Wide Web for more. One of the earliest things I found was Centennia (I think an earlier DOS version on floppy that I had was called Millennia), a program that illustrated border changes in Europe over several millennia. It’s still around in fact, and if you need a program that shows boundary changes over time in movie fashion, it’s for you. A slightly later purchase of mine was a French CD-ROM (remember those?) of the Cassini maps, Carte de Cassini. It was a stand-alone program (required the CD in the drive to run) that allowed you to zoom in on parts of the Cassini maps, and you could even print out a small square of the maps. But, as has happened too often, small niche software never seems to work once you upgrade your computer to a new OS, though I think I still have the Cassini CD laying around here somewhere. These niche software packages were fine as far as they went, but they lacked flexibility and, most importantly, an ability to manipulate and modify the maps themselves.
So I returned to AutoCAD. While it could make maps, the results were rudimentary and there was practically no ability to spruce them up – almost as if construction companies don’t need cutesy icons on their building plans. After I graduated from Ohio State, my old educational version of AutoCAD expired and I no longer worked at a school that offered a cheap educational license for engineering software, so I needed to figure out what to get next. Dedicated map programs such as Centennia and Carte de Cassini existed, but they were for consumption only, not creation. I even toyed around a bit with the line-art maps including with the diagramming program Visio, but it didn’t seem very powerful or intuitive to me; it’s since been bought by Microsoft. Then I found the Holy Grail: Adobe Illustrator. I’d used drawing programs like MacDraw to create a rudimentary map of directions to my house way back in high school, but Illustrator is in a class all its own: an über-powerful vector-based drawing program that lets you create incredibly-detailed illustrations using both raster and vector data. Because it’s über-powerful, it’s also über-expensive and über-complicated to learn. But the former is taken care of if you have an institutional license, and maybe some funds from your school. And for the latter, that’s why there are computer books and classes – I took advantage during my postdoc at George Mason to even take a one-day course on it, although I remember getting lost about half-way through the day. But like anything else, it’s best to learn it by camping out in the local bookstore’s computer books section (when they still had brick-and-mortar stores) and start by learning how to do what you want to do first. Don’t worry, there will be plenty of time later on to realize there is always an easier way to do it. You’ll never learn all of Illustrator features, but you only need a small bit to do a lot. So slowly I learned all about paths and pens and layers and fills and strokes and palettes.
I could now control the Illustrator beast, at least enough to draw things. Then I had to decide whether I wanted to continue with the old fashioned digitize-and-edit process of Cartography 101, or try something a bit more polished. I explored an easy but expensive route, finding a company that sold Illustrator files (extension .ai) of European countries, files that you can open up in Illustrator and manipulate to your heart’s content. Some even offer royalty-free versions. As the following screenshot illustrates however, they’re more useful for a modern historian than an early modern one (remember that you can turn off any of the objects’ layers: the highways, the waterway names…).
This would have required making a lot of alterations, but there are also a few companies that will sell Illustrator files of historical maps. Here’s an example of European boundaries c. 1700 from Euratlas:
Again, nice information, but I just didn’t feel like I wanted to tweak all that data – and there actually isn’t that much town information in Belgium, for example, so you’d still have to add a lot of the info yourself. Most of the work went (I assume) into determining the boundaries, but I have no idea where the map’s border information came from (so I’d end up wasting time verifying it). More importantly, the important information on my map would not be the precise boundary locations, but the information I added. Plus I don’t think it’s royalty free, so those drawbacks eliminated it as a possibility for me to base my own maps off of.
So I was left striking out on my own yet again. I still had the old AutoCAD map files, so I took the route of least effort and simply imported them into Illustrator; in retrospect it might have been better to start with a totally clean slate. But even if you lack pre-existing files, current versions of software like Illustrator are powerful enough that you can eliminate the digitizer tablet entirely and just scan in maps with a regular flatbed scanner. You import the full image into the software and then trace whatever parts of the image you want; newer versions even have a Live Trace feature that will automate converting imported raster images into paths and points. The key is the use of layers – you scan in a raster image (made up of pixels) and import it in one layer, then create a new blank layer on top of it (like a transparency overlay) and trace whichever features you want on this layer as vectors, i.e. lines and points based off of mathematical formulas. Hide or delete the lower layer and now you have the beginning of a real map. You can create an infinite number of layers, adding (or hiding) whichever features you might wish to add, and can combine features from various maps, putting each on its own layer. Add in the ability to use precise grid coordinates and mathematically manipulate the vectors, and the result is a very powerful piece of software. Transferring my AutoCAD maps into Illustrator and editing them, I created the two dozen maps that appear in my Vauban under Siege book. Here is one of the better theater maps which shows my early Illustrator handiwork:
I based the idea for this map off of one of the maps in the old West Point Atlas. Nothing to fear though, since you can’t copyright the cartographic idea of graduated shading to distinguish differently-valued areas – though who knows anymore, given the way the Copyright Office, Congress and the Supreme Court work these days. And, I should note, some of the West Point map’s borders were just plain wrong, so there was definitely a need for a better map, and I was just the person to create it. For those of you who took a careful look at the map, bonus points if you found vestigial clues that this map was an import from AutoCAD. Check out the font used for the scale. So as you can see, my cartographic skills were improving, but I was still overwhelmed by all the details you have to focus on, all while trying to finish the book by the deadline. These details include shifting from a broad look at the map as a whole to focusing on each line and point, choosing fonts and spacing text, developing a consistent yet coherent symbolism that will maximize the amount of information displayed, creating and placing symbols, converting color maps to black-and-white (and changing your symbolism as a result), as well as paying attention to which layers display on top of which other layers (organizing layers is quite challenging when you have a few hundred layers and sublayers). Note too that I forgot to mention on this map that those lines are based off of each year’s winter quarter garrisons; in fact, the lower right key is entirely missing here. A start, but not quite satisfactory.
Ever since Cartography 101 I had heard about GIS, or geographical information systems that were computerized databases that allowed one to easily display any subset of the underlying geospatial data on a map. Moving to Connecticut, I discovered that I could get a free license to ArcGIS, the GIS used by serious geographers, criminalists, environmental engineers and spatial analysts everywhere (MapInfo is a similar GIS). GIS are based on an underlying database, and since I am familiar with database design (having created my own note-taking database), I installed a copy. I was really hoping that I could take advantage of its geographical precision – the big difference between a GIS and an illustration program like Illustrator is that GIS usually comes with data that is attached to geographical coordinates (e.g. latitude and longitude), whereas a map traced in Illustrator is only as good as your tracing precision and is limited to the projection and scale used by the original source map you traced. So I acquired a European dataset for ArcGIS, and here is one example, a choropleth of Europe’s population density.
The software itself is easy to use if you know how databases and layers work. GIS has the advantage of precision and pre-existing datasets, but of course these datasets are usually current, and require database-knowledge to alter them. If I was doing serious geographical work I would use ArcGIS to take advantage of its statistical abilities, e.g. have it compare Area 1 to Area 2, or display and calculate the average population of all towns within 2 miles of a waterway… But I must admit that what stopped me cold was not only the fact that ArcGIS is purely for mapping (unlike Illlustrator’s ability to draw anything), but also frankly the map of Europe itself. I’m just not a big fan of this projection, particularly since my interest lies more with the compressed northern latitudes rather than the Mediterranean. It’s even uglier if you zoom in; I guess you can’t avoid aesthetics sometimes.
Another, much more recent possibility for making your own data-added maps is Google Maps and Google Earth. You can create a ‘mashup’ using Google Maps as the base layer. One historical example is the David Rumsey Collection’s overlay of the Cassini maps, here. This is a relatively simple matter of attaching your data to points on the Google geospatial database, and useful for online display. But as with the other options mentioned above, the resulting maps are limited to online display (i.e. not publishable and copyrighted as well), and you are largely stuck with current representations of the geography being mapped. In other words, I consider it a toy, maybe worthwhile for an undergrad course in digital history, but not serious scholarship.
In short, after looking at a variety of options, most of what I found was either copyrighted and therefore of limited use, or locked down within a propriety piece of software, or it lacked the detail that I needed to talk about operations in a country as small as, say, Belgium. So I’ve stuck with the more flexible Illustrator, and slowly built up a repertoire of skills that fit my needs. It’s taken me years to really get it down (admittedly working on it off-and-on, forgetting much of what I’d once learned), but now I can make visualizations that I’m actually proud of. You’ve seen a few of my timecharts already (check the Graphics tag), and I now use this timechart format to take notes, or at least to develop a chronology and timeline for various wars that I use for my classes.
Below is another use of Illustrator, a map that I recently resurrected from AutoCAD and cleaned up. As a grad student I’d spent months working on campaign narratives of the Spanish and Italian theaters in the WSS, but they were filed away (almost forgotten) when my dissertation went in a different direction. I just recently had reason to dig them out during my sabbatical, and was fortunate to find an old backup CD with the original AutoCAD files on it. Here is one result that I’ve been playing around with for the past couple of days:
It’s not perfect, e.g. I don’t have any info on the Bourbon advance into southwest Portugal, and I don’t list the sources. As someone with absolutely no artistic or graphic design training, I’m also still working on the nuances of map-making, i.e. providing just the right amount of information and combining line weights and fonts and spacing to create an aesthetically-pleasing whole. I’m also still trying to standardize my various symbols between the maps and timecharts. Nevertheless, even at this stage I’m surprised at how professional it looks; of course it helps to have color – yet another point in favor of online publication. But IMHO this is one of the best operational maps for EMEMH that I’ve ever seen, and I created it! The map above is unique in its detailed narration of early modern military operations, and far more informative than others that I’ve seen, including those from the West Point atlas, which I’ve never really liked (though maybe it has particular advantages for military practitioners). But as you’ve seen from my timecharts and database, I tend to go for high information density (as Tufte would say) which require some attention. I think the map works mostly because I’ve tried in particular to avoid the pitfall made by most history books that do take visualization seriously, including most historical atlases, including the generally-good Cambridge Illustrated Atlas of Warfare and the Cassell/Smithsonian series on warfare (particularly the works by John Childs and Tom Arnold). These works provide professional-looking graphics which provide useful information, but they too often suffer from trying to do too much in a single map, or, in the case of publishers like Osprey, go for the cool eye-candy factor with 3D views that confuse more than clarify. For operational military history, the most common problem is slapping a whole war’s worth of army movements onto a single map, with confusing and cluttered results. Obviously it’s an editorial decision regarding how much space to dedicate to any given topic, but too often several different themes are placed on the same map, or a decade worth of campaigning ends up looking like a plate of multicolor spaghetti (spoor, as James Wood colorfully described the army trails). My map, however, is intended as a stand-alone map that focuses on the operations, with the text boxes providing a bit of narrative to help contextualize the actions displayed through the symbols. It isn’t necessary to read 10 pages of text for the map to make sense on its own, which makes it very useful for future review purposes, which is a key function of all my historical visualizations. And I think it presents a lot of information without being too busy: it’s not just a physical map (although it includes elevation above 1000 meters in gray); it’s not just a map of army movements including the commanders and where/when each side is on the attack or defensive (in the future I’ll try to size the arrows proportional to the strengths of the armies); it’s not just a map indicating which places were besieged by whom and with what result (I have a separate version with graduated circles indicating the lengths of the sieges); it’s not just a map of Allied-controlled territory. It’s all of these, as well as an attempt to narrate the phases of the campaign with several text boxes. I’m not sure if the text boxes work or not, but it’s the most obvious option and the only one I’ve tried thus far. The other problem, not fully resolved here despite my use of Roman numerals, is how to show sequence – animation or small multiples are the best bet for the challenge of indicating chronology in only two dimensions, but they’re also in the future. So there’s still work to be done, but I’m extremely pleased with the results thus far.
I should, however, be more humble and admit that three scholars in particular deserve credit for spurring my interest in the graphical display of history, and their best expository maps/charts are certainly among the best I’ve seen: those French Annales historians and particularly Fernand Braudel’s The Identity of France as well as his History and Capitalism series; the graphs in Geoffrey Parker’s The Army of Flanders and the Spanish Road; and the most imaginative of them all, James Wood’s wide range of maps, charts and graphs in The King’s Army.
As mentioned, a whole ‘nother level of historical mapping would be animation. Back in grad school my wife and I briefly toyed around with Adobe Flash, which can create very sophisticated animations, as well as run the interactive graphics on many websites. I have a current version of Flash, but am hesitant to dive back into it, not only because of the learning curve, but because who knows if Steve Jobs’ final legacy will be to kill off Flash altogether. Undoubtedly the teachers among you probably know of an easier and more accessible way to animate in the classroom – use Powerpoint’s Advanced Animation feature. The general problem with any form of animation, however, is that not only editing the results but even displaying them are intimately tied to the specific equipment/software used for creating it – there is no single-screen capture or physical printout of an animation like there can be for a map created in Illustrator. Thus it is even more susceptible to problems with obsolete versions and out-of-date compatibility with new systems and new software. Text files (and to a lesser extent image files) can be easily transmitted from cassette tape (Timex Sinclair 1000!) to floppy drive (5.25″ or 3.5″?) to Zip disk to CD-ROM to DVD to flash drive to who-knows-what-next, but I have no idea how easy it will be to convert an animation file in software X from one version to another, or from software X to application Y, short of recording it on analog film and keeping a projector handy. Look at what the iPad is doing to Flash. The past 30 years of computer history is full of companies and their software using proprietary formats, going out of business, or otherwise becoming unreadable. The trend toward cloud-based computing where you don’t even own the software or control the data will only make our data even more dependent on the corporations that control them. If you are really concerned about longevity, make sure your work gets published in a real book that will survive the coming zombiepocalypse.
But enough of the doom and gloom. If you are interested in creating your own maps and have awhile to play around with software, consider using a vector drawing/illustration program like Illustrator or one of its cheaper cousins. Maybe you’re already marking up maps and the like in Powerpoint; if so, consider graduating to a higher level of drawing software. If you are computer literate, you’ll be surprised at how easy it is to learn just enough to make a rudimentary map, and in the process you’ll think of all kinds of ideas to try out. More importantly, you’ll be able to actually create those ideas. I have a whole boatload full of ideas that I’ll be implementing over the next several years, and it’s all because of software like Illustrator. Visualize information when you first read it and you’ll remember it better, make it easier for yourself to recall it later on, and help others see what you’re talking about.
- How do you use maps for your study of EMEMH?
- What is the best map of military operations that you’ve seen and why?
- Do you have any suggestions for my Spanish operations map, or maps of military operations in general? (Future post will discuss in more detail)
- Tufte, Edward. The Visual Display of Quantitative Information. Cheshire, CT: Graphics Press, 1983. This is the first of his many classic works on visualizing information. He can be a bit of a Nazi with his aesthetics at times, but his books are definitely worth checking out. He is also the one who created the idea of sparklines, which everyone should be using whenever possible, a feature that is now available in Microsoft Excel 2010. Tufte is also a proponent of small multiples, which I’ll be using a lot for my WSS work.
- Monmonier, Mark. Mapping It Out. Expository Cartography for the Humanities and Social Sciences. Chicago: University of Chicago Press, 1993. An excellent book describing how to combine maps and statistics. Taking a course on Quantitative Methods in History around the same time as my course on Cartography was definitely useful.
- Knowles, Anne Kelley, ed. Past Time, Past Place. GIS for History. ESRI Press, 2002. An early collection of case studies of GIS applied to history. Published by ESRI, the makers of ArcGIS.
- An early (c. 2002) page on my website discussing maps for history. http://www.jostwald.com/Research/Maps.shtml
- Tucker, Sara W. “Map Power: Using Computers to Make and Teach History.” http://www.washburn.edu/cas/history/stucker/AHA2002.html. A useful introduction to creating basic computerized maps using a graphics program (raster images) like Photoshop. Illustration programs use vector objects, and so are much more powerful for most map needs.
Note: If you want to use any of the graphics I have created and posted for your own use, consult the Citation page first.