Geoff Cunfer’s “Scaling the Dust Bowl” won’t teach you how to create a GIS or draw a map (it presents the results of his research, not how it unfolded step by step). But it is an excellent demonstration of the interest of maps from data exploration. Maps are sometimes used to show something that numbers would better demonstrate. Here, on the contrary, specific spatial patterns matter. In fact, what we like even more in this paper is the way it weaves together the quantitative and spatial analysis with a thorough, really “humanistic” discussion of different sources on the Dust Bowl, what they show and what they hide. The author’s results contradict received knowledge: he does not stop there but investigates where this knowledge came from and how to make sense of it.
I also like Galanaud et al.’s “Historical Epidemics Cartography”, addressing medieval plagues. We don’t cite enough medieval history, and it’s an opportunity to do it. Besides, this is also a paper that, however quantitative, takes its sources seriously and discusses them in detail. It also changes scales in all senses of the word: see how Fig. 3 zooms in on one street, showing the map along with the original source. This is only possible because the authors have performed a statistical analysis before and projected it on their map. The paper is also general careful as to what can and cannot be learned from each source and each anaysis. But however much detail is given on many things, it is still not here that you will learn to devise a map or GIS. You will only learn about the preconditions (sources and how to deal with them) and interpretation – which is already something!
If you want to discover all the steps required for the building of a historical GIS (and reflect about whether it’s worth the effort in your case), you can turn to “The Valley of the Shadow” project, or more precisely, the website “The Differences Slavery Made.” This is a really impressive project, especially as it ended more than a decade ago, but the digital archive is still fully functional – a rare case. It is a detailed, partly quantified and spatialized, comparison of two US counties in the 1860s. The two counties were situated on each side of the North-South divide. There have been more classical publications from this project, but this website (and others) offer a different type of narrative, with more emphasis on methods and evidence. You should really have a tour. To come back to our main point here: you will find here a detailed discussion of all the steps involved in creating a historical GIS. This was written in 1999-2001, so you certainly won’t use the exact same tools; but you’ll get a good sense of all the required steps.