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

CHAPTER 1: GIS AND ITS USES IN HISTORICAL RESEARCH

1.4 Problems with GIS

It is important to note at an early stage that there are also serious limitations to GIS. These fall into four main classes: problems to do with the GIS data model; problems to do with the data themselves; problems with the academic paradigm; and practical problems.

Spatial data consist of one of four types of graphic primitive, namely: points; lines; polygons, or pixels. Where the data have precisely defined locations that realistically represent the features to be modelled, GIS is a powerful tool. Other data cannot be adequately represented spatially in this manner. This may be because the data do not fit the four types of graphical primitives well, or because the data are imprecise, a problem that GIS cannot cope with easily. For example, GIS is well suited to modelling hospitals and census districts in the manner described above, but is not well suited to representing the catchment areas for the hospitals where these are poorly defined and overlap heavily with surrounding catchments.

Secondly, the data themselves can also cause problems. Much historical data will be taken from historical maps which may not be accurate, and the representation of features from these maps in the GIS at best will only be as accurate as the original source. In reality they are likely to be worse, as new errors are added when the data are captured (or transcribed, to use the historical term). Many of the clues about the accuracy of the original source will be lost when the data are captured. An obvious example is that if a feature is represented by a crude, hand-drawn, thick line on a map we may question its accuracy. In the GIS it will simply appear as a digital line like any other. Less obvious, but at least as important, is the scale of the source map: a map is only ever accurate within the limitations of its scale. In a GIS, however, we are able to zoom in hard or to integrate data taken from maps with very different scales. This demands more from the data than the original map or maps were designed to accommodate and may lead to inaccuracy, error and misunderstanding. Although historians will be familiar with issues associated with the accuracy of transcriptions, GIS is particularly demanding of the accuracy of its source data, as will be described in section 3.4.

Thirdly, the academic origins of GIS were located within technological advances in the earth sciences. Its role in academic geography has yet to be fully established, and history trails some way behind this. Through the 1990s there was considerable debate in geography about whether GIS offered a cohesive, scientific framework that could re-unite the subject, or whether it was a return to a naively positivist agenda. From the historian's point of view, GIS offers new tools, new techniques and new approaches. These approaches must be used critically and should complement traditional ideas, approaches and concerns.

The final set of limitations on GIS is practical. GIS software is expensive and may be difficult to use. GIS hardware has fallen in price over recent years but can still be expensive, and GIS data are often financially expensive to buy and capturing them yourself is costly in time as well as money. People with GIS training are often very employable and thus expensive. As a result, entering into GIS is often more costly than originally anticipated and should be done with care.