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

CHAPTER 2: THE WORLD AS VIEWED THROUGH A GIS

2.7 Conclusions

The core idea behind GIS is that it attempts to represent features that are, or were, on the earth's surface. The first stage in doing this is to take real-world themes that are relevant to the research and abstract them as layers with each theme being represented by a separate layer. Features on each layer are represented using a combination of spatial and attribute data with each layer's spatial data consisting of one of four types of graphic primitives: points, lines, polygons or pixels. The resulting data are therefore highly abstracted from the real world. Where the data can be effectively represented using this model, GIS provides a powerful research tool because of its unique ability to handle location and attribute simultaneously. However, this structure fundamentally restricts the way that GIS can be used and where features cannot be effectively represented in this way the use of GIS may not be appropriate.

The implementation of most of what has been described in this chapter is performed automatically by modern GIS software and is well hidden from the user, but it is important to understand the way in which GIS represents the world. Many datasets simply do not fit this model comfortably. If they don't, it is unlikely to be worth investing much effort in GIS. Where they do, this model represents a powerful way of integrating space into an analysis or exploration of a particular location, be it an area of only a few square metres, a whole country, or even the whole world. In addition to the primitive method of representing space, many other issues also need to be borne in mind. These include scale, accuracy and uncertainty. The GIS research community expends large amounts of energy worrying that GIS users will not understand these limitations. They should, however, be familiar to historians as they are the limitations of the original source plus the additional limitations created by converting the source into digital form. The problem is that for spatially referenced data these are more complicated than they may appear at first sight. The issues involved will be discussed in more detail in subsequent chapters.