Overview and Background

This chapter is an introduction to the process of adding your own digital content to a map. The process of adding this digital content is called digitizing, and the digital content that you create is called vector data.

Vector data is a type of data within a GIS. Another form of data in GIS is the raster. A vector is digital data that the user creates. Users can create vector data in the form of points, lines, or polygons. GIS users have traditionally created vector data in the form of a shapefile, although more recent formats include the Geodatabases and GeoPackages. The key characteristic of GIS is that it assigns geographic coordinates to this shapefile. By contrast, a graphics software program can also create points, lines, and polygons, but it will not assign them a geographic location.

In cartography, a feature is any representation of a real-world object on a map. Features could include the lines of a historical railway, the patrol routes of an army, the buildings in a town, or the forests, fields, and bodies of water in a nation. Features can be places you know of through personal experience, but, in the humanities, this knowledge is more often what you learn from primary-source research.

The examples in this chapter focus on the historical maps of eastern Prince Edward Island (PEI) that we added to our GIS in Chapter 3: Georeferencing. We will digitize historical boundaries and other features of the town of Georgetown, PEI, using the maps from Chapter 3. No matter if you completed Stream A or Stream B in that lesson, you can still complete this step and start adding your own content to a historical map.

When Is It Best to Digitize?

Using a GIS to annotate a map is relatively straightforward, but it is difficult enough that we might want to pause and ask when it makes sense to use digitization and when you should simply use other graphic design programs. Essentially, you should digitize in a GIS any time you want to add your own digital content to a map, have your content remain interactive, and retain the ability to remove unwanted content. In other words, do this when you need your content to remain graphically dynamic, geographically scalable, and analytically robust. For example, if you simply want to label a map, you can do so with graphic design programs provided that you can find an open-access basemap that suits your needs. However, if you were to label dozens of features, it would be difficult to change the text on those labels all at once. With a GIS, it is possible to alter, concatenate, or even translate the attributes of the content you created, and those attributes may be used to generate dynamic labels and other graphics. Moreover, these graphics are scalable in a GIS. Many scholars need to make multiple versions of their maps at different geographic scales, or create inset maps, or even create scalable and interactive web maps. In a graphic design program, this would require resizing and repositioning those dozens of labels for each output. In a GIS, labels can be set to appear only at certain scales, and their font size will change dynamically as the scholar or the audience zooms in and out on the map. Finally, with dynamic features you retain the ability to show only the features you want to show. It is often difficult to simplify labels and other features using graphic design applications. In a GIS you simply turn off the layers or filter your features (as we did in Chapter 2) so that only certain features are visible on your final map.

Perhaps the most important reason for creating your own content in a GIS is to enhance the analytical characteristics of your research. A GIS is, after all, a spatially enabled database as well as a cartographic tool. As you gather information on your research subjects, you may want to simply map their locations, but you likely also want to analyze the significance of those locations. Why do the spatial characteristics of these subjects matter? What do they tell you about their relationships to each other and to other aspects of the built and natural environments? For instance, the density of those dozens of features you created is often about more than simply the difficulty of fitting the labels onto your map. If those features represent people living in a crowded city, or sharing limited resources in the country, then density mattered to them in important ways. The proximity of your human features to pollutants, threats, infrastructure, and many other features mattered, as well. So also did the time it took people to travel to work or across the country. These are only some of the many spatial questions you can ask of your content once you add them to a GIS. So, if you’re ready to give it a try, we hope this chapter will help.

When Is It Best to Use Georeferenced Maps to Digitize?

Not all digital humanities projects require the use of georeferenced historical maps. In fact, you can “digitize” new features in any map, even without historical maps to guide you. The most basic form of digitization involves identifying the location of places that you want featured in your project and clicking on the map to add a digital vector, such as a point, line, or polygon to represent that location. We will do this in the first part of this chapter. In its most basic form, digitizing can be done “by eye” without the help of georeferenced maps that show the location of the features. If the feature still exists today, you can even select from vectors and reuse them (with permission from the content creator, of course). However, when mapping highly detailed features, say a series of historical houses in an urban area, you will quickly find that having a georeferenced map will help with digitization in at least two ways. First, the process of mapping “by eye” will become confusing when dealing with a large number of features. Second, many features have changed substantially over time, and, in the cases of urban homes, many were removed or remodeled so that the modern GIS features are no longer useful as a reference point.

In other cases, historical maps will not be particularly helpful at all. For instance, if you wanted to create a map of the five largest communities in PEI in 1921, you would access the 1921 Census of Canada, record the names of the largest communities (using one of the tables), verify if those names and communities have remained the same, and proceed to make a map showing a single point on each of the five communities. The historical maps of Prince Edward Island would be most helpful if the communities had changed their names, or, less frequently, their locations! Georgetown, PEI, is now part of Three Rivers, but its location has not changed. However, as the town continues to grow and blend into others in the Three Rivers municipality researchers may have to eventually consult historical maps to learn where the town of Georgetown was located as the Census understood it in 1921. At the time we wrote this textbook (2021), Google Maps and other common ways to search the location of towns still showed where Georgetown exists within the larger Three Rivers municipality.

In other types of digital humanities projects, historical maps and modern GIS tools must be used in conjunction with other research. For instance, if you wanted to create a fictional map showing the geographic places that were most likely intended by an author like L.M. Montgomery in her novel, Anne of Green Gables, you would have to compile a list of communities mentioned in the novel, and then another list with the most likely geographic place name. Other scholars have debated this in their works, and it would be helpful to have maps of PEI both from the time that book was published (1908) and more recently. Still others have mapped sites related to Montgomery’s own life. Depending on the scale of such projects, and the level of detail used by the author and more recent scholars, it could help to have historical maps that have been georeferenced and overlaid on a modern GIS map.

Example: The Town That Time Forgot?

One of the few urban maps of the Island helps us understand maritime urban development in this period and a good deal about the social life of settler economies. It also helps uncover a historical mystery that has puzzled local heritage communities for 100 years.

As you will recall from Chapter 3: Georeferencing, Georgetown is the capital of Kings County, so selected for its deep water harbour and other transportation features. It became a commercial and shipbuilding capital as well, and, in 1880, it was enjoying what Graeme Wynn calls the industrial ascendancy in the Maritime provinces. That’s also when things turned sour.

The map in Meacham‘s Atlas from 1880 depicts Georgetown at close to its height. The town boasted a bustling merchant quarter, a manufacturing district including shipbuilding and foundries, and a service sector marked by the new PEI railway’s eastern terminal, the county’s highest court, and its highest high churches, too. And therein lies the mystery. Many of these structures disappeared in the twentieth century, leaving behind several abandoned homes and a hollowed out square. At the beginning of the twentieth century, the town’s population was about 1,200 people. Today it is around 555. What was once a centre and a connector has become a margin and, in some ways, the end of the line. Montague has become to Kings County what Summerside is to Prince County (except all of Three Rivers is much smaller than even Summerside).

Through maps, we will be able to see Georgetown’s stagnation between 1880 and 1991, which resulted in little new construction or development. We will also see how Georgetown lost much of its commercial activity over the same period, including the losses of a ferry service and a shipbuilding company.

Two Streams

This chapter is divided into two streams. In Stream A, we will digitize features on the 1991 National Topographic Series (NTS) map of Georgetown. We will identify what remained of the once bustling town in 1991, a century after its golden age.

If you made it to the end of Chapter 3: Georeferencing, you will also be able to complete Stream B of this lesson. In Stream B, we will digitize historical content from the earlier and much more detailed historical map of Georgetown, the one found in the 1880 Meacham’s Atlas of PEI.

By digitizing features from both 1880 and 1991, we can visually demonstrate how Georgetown declined over this period.