Method or Madness? History and its Disparate Approaches

Conference Abstracts

Session 1, 9.45 - 11.00

Abductive Reasoning, A-Life, and the Historian's Craft: One Scenario for the future of History and Computing
Dr John Bonnett (Tier II Canada Research Chair in Digital Humanities, Department of History, Brock University)

In any scenario devoted to the future of history and computing, it is important to refer first to the content and analytical methods historians are likely to employ. In the scenario presented here, this paper will offer three propositions. One, scholars in future will devote renewed attention to a class of historical events for which there is no clear presumptive cause, a domain in which some scholars points to environmental constraint, and others to social action to explain their occurrence. Second, historians will employ a method of reasoning known as abduction to assess the relative plausibility of competing explanations. Unlike deduction, which starts with a given hypothesis, abduction starts with an event, and searches for a plausible hypothesis. Possible world reasoning provides one means to employ abduction. In this method, candidate causes are individually tested to determine their capacity to produce an outcome mirroring the one produced by the historical record. Third, this paper suggests historians will turn to possible world reasoning due to the capacity of computer agent-based modeling to support its exercise. Archaeologists have already successfully applied this method to re-examine a long-standing problem: the reason for the disappearance of the Anasazi from the Long House Valley, Arizona in 1300.

Multiple Recapture Methods and Historical Sources: New Methods, Models and Applications
Dr Gidon Cohen (Politics Division, School of Arts & Social Sciences University of Northumbria (University of Durham from September 2005))

Multiple Recapture methods are widely used in the natural and social sciences to study the size and dynamics of hard to measure populations. The central idea is that patterns of encounters with individuals from hard to measure populations can be used to give insights into the population dynamics even with limited data. This paper will report the results of an ESRC funded project using recapture methods to look at an historical population - activists in the largest constituency Labour Party in the post-war years. In contrast to previous historians using the method, the paper will highlight the limitations of 'closed' recapture methods due to the sensitivity of the method to violations of underlying assumptions of the model. However, the paper will highlight a wide range of more robust 'open' recapture models, which can be used to examine patterns of recruitment to, and departures from, historical populations. Thus we suggest that given the partial and incomplete nature of historical sources 'open' recapture methods have a very wide application to historical studies.

Reading the Readers: Modelling Complex Humanities Processes to Build Assistive Computational Tools
Dr Melissa Terras (School of Library, Archive and Information Studies, University College London)

The ink and stylus tablets discovered at the Roman Fort of Vindolanda are a unique resource for scholars of Ancient history. However, the stylus tablets have proved particularly difficult to read. This paper describes the initial stages in the development of a computer system designed to aid historians in the reading of the stylus tablets. A detailed investigation was undertaken, using Knowledge Elicitation techniques borrowed from Artificial Intelligence, Cognitive Psychology, and Computational Linguistics, to elicit the processes experts use whilst reading an ancient text. The resulting model was used as the basis of a computer architecture to construct a system which intakes images of the tablets and outputs plausible interpretations of the documents. It is demonstrated that using Knowledge Elicitation techniques can further the understanding of complex processes in the humanities, and that these techniques can provide an underlying structure for the basis of a computer system which replicates that process. As such it provides significant insight into how experts work in the field of Ancient history and papyrology, whilst providing the means to develop tools to assist them in their complex task.

Session 2, 11.15 - 12.30

The project 'The Health of the Cecils (c.1550-c.1660)' funded by the Wellcome Trust, based at Royal Holloway, University of London, examines the health-seeking behaviours of an elite family and their correspondence network in early modern England. The immense number of letters, bills, accounts and other archive material for the project led the principle researcher (a historian) to decide on the use of Computer Aided Qualitative Data Analysis (CAQDAS). Two social scientists were recruited to be part of a research team to consider the opportunities and challenges of using the NVivo software in the analysis of historical data. In both a practical and a conceptual sense, NVivo is a powerful research tool which allows the researcher to analyse a vast amount of material with relative ease and speed. First, NVivo enables the researcher to mark certain concepts and topics as well as to establish her own; second, it offers the opportunity to create conceptual and empirical coding hierarchies and correlations between concepts; third, it allows for the combination of qualitative and quantitative data analysis; finally, it makes it possible to keep a record of the development of emerging thoughts in the process of reaching a greater understanding of the collection. This paper will consider the ways in which the discrete skills of the researchers have been brought together to use NVivo as an effective means of studying a branch of early modern medical history and it will also demonstrate how the software has facilitated the generation of new, and sometimes unexpected findings.

Statistical Analysis of Historical Conflicts: The ICB Project
Glen Segell (Institute of Security Policy)

Imagine a computerised database of 440 international crises, 32 protracted conflicts, and 970 crisis actors from the end of World War I through 2002. The data of the International Crises Behaviour project is available formatted for SPSS for Windows and tab-delimited text for the intrepid historian to use statistical analysis to determine the nature of international politics. Indeed over the course of the twentieth century, the onset of an interstate military-security crisis represents one of the most frequent forms of hostile interaction between adversarial states yet there is still little systematic knowledge about crisis perceptions and the decision making style of such key actors. The aim of this paper is to detail the ICB Project shedding light on a pervasive phenomenon of world politics through computerised statistical analysis of historical events.

Computational 'Polichart' Cartography for Visualization of Historical GIS Patterns and Processes
Prof Claudio Cioffi-Revilla (Center for Social Complexity, George Mason University)

Conventional cartography for visualizing historical and geographically distributed information (e.g., thematic maps or historical atlases of, say, poverty, conflict, economic development, or other social data of historical interest) is limited by several longstanding problems, including arbitrary projections, information cluttering, territorial distortions, and other issues. Misconceptions and inferential errors can also result from these problems. 'Polichart' visualization - a new computational methodology inspired by social cartography and computational tools-solves some of these problems by improving the topology of the base grid onto which historical data are plotted, and by adopting a set of simple albeit graphically efficient conventions. The resulting visualizations maintain several critical features (N-S-E-W orientation, territorial contiguities, and others), while eliminating some of the worse problems with conventional thematic maps of historical processes. This paper describes this new approach in support of computational historical analysis and provides some examples with specific policharts applied to historical patterns of conflict and peace. Examples include the historical evolution of nuclear proliferation potentials in various world regions, state failure potentials that may provide historical insights, and other macro-historical processes. Extensions to smaller geographic units (e.g., county or provincial data) and risk analysis (e.g., visualization of hazard rates in event history analysis) are also discussed.

Session 3, 2.00 - 3.15

"The Medieval Settlement at Onley, Northamptonshire: an evaluation of the process of formation and desertion of the medieval settlement"
Grenville Hatton

The deserted mediaeval settlement at Onley, Northamptonshire (OS map reference SP 520715) lies close between the route of the M45 motorway and the site of HM Rye Hill Prison, on a flat plain between the upland of the Rugby plateau to the west and the commanding eminence of Barby hill to the east. In 1966, Allison Beresford and Hurst commented '... a number of features combine to make it one of the most important in the county, despite the unfortunate lack of documentation.' In spite of their implied suggestion that more work should be done on the settlement, in fact remarkably little progress has been made since that early publication. The main purpose of this paper, therefore, is to attempt to develop a basic chronology for the site, and to try to understand and explain the background causal forces behind the desertion of the settlement aided by a variety of graphical software tools?

The author then suggests that the desertion of the settlement may have been the tip of a larger iceberg. Yeoman inter-relationships between neighbouring villages in the surrounding area (derived from a study of wills and inventories), along with other factors based both on documentary research and fieldwork, point to the possible evolution of a "logistics chain" for breeding and rearing, pasturing, shearing and finally selling meat on the hoof, which spread across a wide tract of neighbouring country. Yeoman initiative may have prompted the initial planning of the network, but seigneurial influence was needed to create the conditions in which it would flourish and grow.

Weblogs and podcasts provide the potential for capturing for historical analysis a great deal of rich material about everyday life and social interactions as well as shifts in intellectual thought. Journalists, academics, and politicians are keeping blogs as well as high school students. Extracting relevant insights from these sources presents a wide range of problems. Issues of authorship arise; for example, "ghostblogging" is popular among politicians with limited time to devote. Matters concerning how to judge the significance of blogs are emerging; simple blog "visits" or podcast downloads may not be as revealing a measure as trackbacks, blogrolls, and data from websites that handle RSS (Really Simple Syndication) feeds for the public. The massive amounts of data that are produced in blogging and the aural format of podcasting also provide technical obstacles for some sensitive forms of coding and analysis.

After a brief overview of weblog and podcast technology, this presentation describes how various researchers are currently capturing weblogs and podcasts for analysis, and explores various approaches to gleaning historical significance from them. The presentation will conclude with some speculation about the interactions between weblogs and podcasts and other forms of historical documentation. Will blogs and podcasts (often produced without substantial editing and revision) provide a rich source of personal reflections about an age, or, as some scholars assert, a massive amount of vain and often factually-flawed babble?

Can hypertext be used to introduce new forms of pedagogy for history scholars?
Andy White (Centre for Media Research, School of Media and Performing Arts, University of Ulster Coleraine)

This paper will apply Roland Barthes's and Jacques Derrida's theories on non-linear reading ('writerly' rather than 'readerly') and inter-textuality respectively to a study of the uses of hypertext in the teaching of history. Both theorists argued for a form of interactive readership long before the invention of computer-based hypertext. The use of electronic non-linear narratives promotes an 'associative' form of learning, rather than the more traditional form of pedagogy offered by linear monographs. But to what extent is this new form of pedagogy to be encouraged? George Landow argues that cognition is 'associative' in character and therefore hypertext pedagogy is the best method of learning. Others, like Miall, are sceptical of these claims about the nature of cognition, but are supportive of a form of pedagogy that can work alongside, not eclipse, traditional linear forms.

In this paper, the theoretical debate about hypertext pedagogy will be grounded in concrete examples of online archives of historical sources. In particular, there will be an exploration of the way in which this theoretical debate impacts on the design of digital archives and a discussion on the interrogative strategies of users.

Session 4, 3.30 - 4.45

Data Warehousing and Multi-Media Historical Data Repositories
J. Delve (History of Computing Group, School of Computing, University of Portsmouth) and R. G. Healey (Dept of Geography/School of Computing, University of Portsmouth)

Data warehousing technology was originally developed in the commercial sector to utilise transaction processing data for business intelligence purposes. Typical data warehouses now extend into the terabyte range, but in general they are restricted to standard tabular data resources, specially optimised in terms of structure and indexing to facilitate on-line analytical processing (OLAP) type queries.

Originally, most data warehouse tools were special-purpose software packages distinct from widely used relational-type database management systems. More recently, however, relational software vendors have integrated data warehousing functionality into their 'universal server' database technologies. This has opened up possibilities for utilising the analytical power of data warehousing structures and query tools in conjunction with other kinds of multi-media data types that are of interest for historical research and historic preservation. These could include archival photographs or illustrations, large bodies of textual materials and map-based data.

This paper examines possible strategies for 'adding value' to multi-media data resources by deploying this combination of data warehousing and multi-media database technologies. Both design and implementation issues are considered in the context of an operational data warehouse (originally created as a research and teaching demonstrator), and other potential application areas. The demonstrator relates to the 19th century anthracite coal mining industry of Pennsylvania. It uses data in a variety of forms, including tabular data on employment, output and capacity utilisation, textual data derived from newspaper sources and Mine Inspectors Reports, and a large digital archive of scanned historical mine photographs.

Computer-aided Construction of a Dictionary: Expected and Unexpected Benefits
Nancy Cox and Karin Dannehl (The Dictionary Project, University of Wolverhampton)

Nine sets of documents, selected from contrasting source types, transcribed and digitised, form the basis for the construction of a 'Dictionary of Traded Goods and Commodities 1550-1820'. Each set of documents has been captured in a stand-alone database, with each item of text stored in a separate text field. A search facility permits the analysis of the fields and the construction of the dictionary: the headwords and their definitions, and the variant spellings found to have existed.

• Inter-relationships between commodities and/or their descriptors;
• Constructive tensions between words and things;
• Emotive and social connotations of words and things.

The package has been used by PhD students from the University of Wolverhampton and elsewhere, and by researchers on topics varying from textiles and glass to German toys and Jews' harps. In its finished form it will be useful to students at many levels, not only as an historical Dictionary, but also for investigations into topics like the spread of consumer goods, the nature of inventions and innovations, and the genre of promotional literature.

Back to top