A place in history: a guide to using GIS in historical research

CHAPTER 6: VISUALISATION FROM GIS

6.2 Mapping and cartography in historical research

The map is a powerful way of presenting the information held within spatially referenced data to an audience. Cartography is an academic discipline in its own right with a long history. It is both a science and an art. From a scientific perspective its role is to present features on the earth's surface to an audience in an accurate and objective manner. From an artistic perspective its role is to present this information in a way that is both communicative and pleasing to the eye. These two roles are sometimes contradictory, and it requires skilled use of cartographic principles to balance these two objectives.

Most GIS software packages make producing basic maps easy as it is a core part of their functionality. This means that almost as soon as the data are in a GIS format, the researchers are able to explore such data through maps. The maps can be refined and re-drawn multiple times as part of the research process, giving the researcher the ability to gain a thorough understanding of the spatial patterns the data contain. At the end of the research process production of maps for publication either on paper, or more recently electronically on the Internet or CD-ROMs, becomes a relatively trivial process.

This means that historians wanting to use GIS need to learn the basics of cartography, so that the maps they create and interpret lead to improved understanding rather than misleading or causing confusion. In this chapter it is only possible to explain briefly a few basic rules about how good-quality maps can be produced. Many good guides to cartography are available, and the bibliography lists some of them.

A map can be regarded as a simplified abstraction of the world which presents complex information about one or more phenomena in an understandable manner; and which is also a valid picture of the underlying data. To do this effectively it is important to follow a number of general rules:

1. The map should contain as much detail as is necessary but not so much that the pattern becomes obscured, cluttered or over-complicated.
2. A map should stand on its own and be understandable without referring to the accompanying text. To this end it needs a title, a legend and representation of scale. The legend should explain all symbols and shading used on the map.
3. The method of symbolisation used should be appropriate for the data being represented. GIS usually simplifies this. Points are usually represented by point symbols, polygons are represented using choropleth maps, and continuous surfaces are often represented using isolines (for example, contours to show relief).
4. The symbols and shading used should be as self-explanatory as possible to minimise the amount of times that the user needs to refer to the legend. For example, water features should be coloured blue.
   
5. If a shading scheme is used to represent a hierarchy, the features at the bottom of the hierarchy should be shaded in the lightest colours, those at the top in the heaviest, and there should be a clear and self-apparent progression up the hierarchy.
6. Where a continuous variable, such as the unemployment rate, is sub-divided into discrete classes, care should be taken in both the choice of the number of classes and in how they are defined. In general, for grey-scale maps no more than four or five classes should be used. Where colour is available this may be increased if necessary but never to more than ten. The intervals should not be arbitrarily chosen but should rely on some characteristics of the data. Examples include putting equal numbers of observations in each class; using equal intervals (for example if the range of a variable is from 0 to 20 and four classes are required the breaks would fall at 5, 10, and 15); using the mean and a standard deviation either side of the mean, and so on. The choice depends on the frequency distribution of the data, with heavily skewed datasets being among the most difficult to represent. Evans (1977) provides a detailed discussion of this.