Introduction
The presentation will address the problem of how to display uncertain historical data in a web-based Geographical Information System (GIS). In recent years the use of internet GIS has allowed the general public to access large volumes of spatial data. Although GIS has been applied in specific areas of academic research, the technology has largely remained the preserve of specialists, as it can be difficult to use, requiring training and experience (Boonstra 2009). Yet the application of GIS to humanities research in general, and to the interpretation of historical data in particular, has the potential to develop and answer innovative research questions (Bodenhamer, Corrigan & Harris 2010).
The ‘Mapping the Jewish Communities of the Byzantine Empire’ project will use Internet GIS technology to disseminate data that outline the history of these communities. The project is based at the Centre for Advanced Religious and Theological Studies, Faculty of Divinity, University of Cambridge and is funded by the European Research Council. Textual, epigraphic and archaeological data allow the project to establish the locations, dates and other attributes of Jewish communities. Prior to our work these data were scattered, fragmentary and difficult to access. The project incorporates varied data in an Internet GIS which provides an efficient method of dissemination through dynamic, interactive maps. The GIS will feature an easy-to-use interface making data accessible through a standard web browser. The project team hopes that this accessibility will promote the inclusion of the Byzantine Jewish communities in the study of Byzantine, Jewish and minority history. A further aim is that medieval historians with little experience of GIS will become familiar with the technology and its potential through use of the website.
The spatial perspective that the project offers will allow research into Jewish communities to develop in new directions. For example, the project’s GIS could be used to aid interpretation of the involvement of Jews in trade, the effect of political change on their distribution and lives, and the factors that influenced the development of separate Jewish residential quarters. The focus of this presentation will be the role of Jews in the Byzantine economy. A robust understanding of geography is important for interpreting trade and therefore the perspective that GIS offers makes it a valuable aid to understanding historical economies. Characteristics of Jewish communities such as their location and date are crucial to evaluating their interaction with the wider economy and are fundamental to the structure of the project’s database. Yet these data are beset by uncertainty, causing significant problems with their representation and interpretation in GIS.
Uncertain Data in GIS
The presentation of historical data using GIS is not straightforward as the technology has developed in the context of empirical research. Symbols used in any form of mapping can convey an unwarranted air of reliability by masking differences between features represented in a single category. The original sources on which maps are based are rarely revisited by scholars in the same way that references to texts are. Maps have a veneer of objectivity making them easily manipulated to suit particular interpretations (Monmonier 1996). Similar problems affect GIS, and the technology has, since the late-twentieth century, been criticised for being over-reliant on a positivist interpretative framework. Historical or other cultural data are often uncertain and therefore the application of GIS in these areas requires a different approach.
Uncertainty can be divided into three categories. First are factors that are inherent in defining the object of study; second are problems with sources of information and their interpretation; and third are issues with the representation and reduction of these data to feature in a database (Plewe 2002). Uncertainty can arise in each of these areas where characteristics are, for example, contested, ambiguous or unreliable. These problems are particularly acute in interpreting historical data from medieval and earlier periods (Bartley & Campbell 1997). To be a useful resource for historical research, our website must incorporate and communicate the complexity of these data in a format that is straightforward enough to be understood quickly and by non-specialists. This presentation will examine uncertainty in the location and date of Jewish communities as these attributes are fundamental to the structure of the project’s database.
Representing Uncertainty Through Transparency
The project has developed a system of symbols that communicates uncertainty in dating and locating Jewish communities. Degree of uncertainty is represented by variation in the transparency of symbols. To give an example, the date of a person’s birth can be known with varying degrees of certainty. Where one source may provide the date to the exact day, another might only give the year. Dates may also be reached by comparison or interpretation. If the symbol that represents a piece of data in the project GIS is more transparent, the user is made aware that there is less certainty regarding its date. Uncertain dates will be displayed as periods with reference to a time series. The system of symbols will communicate the certainty with which the existence of a Jewish community at a particular date is known. This can allows users of the website to examine the Jewish presence within the wider framework of Byzantine economic history with an appropriate level of confidence.
The system will also use transparency to communicate uncertainty regarding where Jewish communities were located. The location of a Jewish community can be questioned due to, for example, ambiguous or imprecise references in sources as well as differences in their interpretation. The degree to which a symbol is transparent will be determined by the number of possible locations. The siting of Jewish communities was influenced by and was an influence on their role in the economy. Preliminary analysis supports the view that some Jewish communities resided in particular locations for economic reasons. Internet GIS can help historians gain a better understanding of these relationships providing that uncertainties in the data are communicated.
To explicate the representation of uncertainty, the website must communicate the sources and types of uncertainty to the user. Information on uncertainty and the transparency of symbols will be made accessible through clicking on symbols. The sources used to produce the GIS will be communicated through metadata. The symbols will also allow access to other attribute data provided in the sources including religious sect, occupation and name. These attributes provide further structure to the database and form the basis of searches.
Summary and Implications
The historical study of medieval economies would benefit from greater integration with geography, particularly to develop a detailed understanding of the spatial character of trade. Our project’s GIS can help to develop this understanding by disseminating historical data whilst being sensitive to the needs of medieval historians. It is hoped that the system used for communicating uncertainty in our project’s Internet GIS will be taken up by future historical projects that deal with uncertain data.
References
Bartley, K., and B. M. S. Campbell (1997). Inquisitiones Post Mortem, GIS, and the Creation of a Land-use Map of Medieval England. Transactions in GIS 2: 333-46.
Bodenhamer, D., J. Corrigan, and T. Harris, eds. (2010). The Spatial Humanities: GIS and the Future of Humanities Scholarship. Indiana University Press, Indiana.
Boonstra, O. (2009). Barriers Between Historical GIS and Historical Scholarship. In D. Bodenhamer and P. Ell (eds.), International Journal of Humanities and Arts Computing 3. Edinburg: Edinburgh UP, pp. 3-7.
Monmonier, M. S. (1991). How to Lie with Maps. London: U of Chicago P.
Plewe, G. (2002). The Nature of Uncertainty in Historical Geographic Information. Transactions in GIS 6: 431-56.
