The Born Digital Graduate: Multiple Representations of and within Digital Humanities PhD Theses

The Born Digital Graduate: Multiple Representations of and within Digital Humanities PhD Theses
Webb, Sharon, Department of History, National University of Ireland, Maynooth, sharon.webb@nuim.ie
Teehan, Aja, An Foras Feasa Research Institute, National University of Ireland, Maynooth, aja.teehan@nuim.ie
Keating, John, An Foras Feasa Research Institute, National University of Ireland, Maynooth, john.keating@nuim.ie
Introduction
This paper describes the methodology used in the creation of digital chapters and subsequent recreation of digital entities or objects derived, modified, transformed and visualised from XML encoded scholarship. It considers the changing function of traditional printed theses and how the use of technologies affects the representation and functions of graduate digital scholarship.

This paper is based upon the working methodologies of two PhD theses. Specifically, Webb’s thesis examines the creation of factlets and subsequent visualisation of factoids, which inform not only the source information and encoding but also the development and completion of historical research outputs. These outputs, supported by XML, XQuery and factlets, demonstrate the use of digital technology as an essential feature of humanities research and its methodologies. Teehan’s thesis reflects upon current digital representation models for pre-existing sources relevant to humanities research. Focusing on transactional, or functional, documents, it proposes a methodology for contextually modeling and XML-encoding those resources, using established software engineering and computer science paradigms such as Use Case analysis and UML modeling, which foreground the User. Both theses examine procedures and strategies for conducting humanities research using digital tools and applications. Thus, this paper is central to a reflective and reflexive process resulting from, and in, the critical self-evaluation of the theses and their outputs.

Traditionally, research outputs codified as chapters or sections can be seen as the final manifestation of a PhD thesis and reflect the use of print or static technology. The functionality of these outputs varies according to different headings and ranges from literature reviews, general narrative and concept generation, to the development of structured arguments based on theory and source material, to the provision of essential referencing and bibliographic material. These functions are referred to as “generic characteristics of academic discourse” (Mingwei, 2010) in linguistic structural analysis. Chapter functionality represents and reflects the original research statement and provides the means to convey and articulate traditional scholarship within the medium of print. The use of XML, and XSLT, along with the provision of software libraries, creates a framework to add dynamic functionality to an otherwise static text. “Generic characteristics” (Mingwei, 2010) are encoded, which enable the use of the described framework.

This approach reflects the innate capability of the digital medium to layer extra functionality over the restricted functionality of printed works. Rather than creating just a single representation of scholarly output, the use of XSLT and software libraries generates and encourages a reflexive process between text, argument, narrative and source material.

These methods change reader and user activity - one user may be a reader while another may have access to an interactive environment. Different user roles and environments transform the user from a passive participant to an active one. The realisation of various use cases enables the user to do more than just read the text and this activity realises the importance of data reusability.

Figure 1 outlines the process involved in creating multiple representations of digital scholarship and will be used to detail the various stages involved in creating new digital objects based on specified use cases.

Figure 1: The stages of creating new research outputs, various chapters are defined by various use cases.

Full Size Image

Stage 1 - Defining the Use Cases, Creating the Model, Encoding the Source
Text is innately encoded with semantics and functionality and each chapter or section in a piece of scholarship establishes or conveys various essential processes in the life of a text. These processes consist of deduction, concept development, narrative, consequence, etc., and it is these “lexical relations” (Eggins, 2004) within a text which develop specific research statements. Other bibliographic properties of a section are concrete rather than abstract and provide essential functionality e.g. references, footnotes, paragraphs and titles.

Despite these multiple perspectives, transformation of a born-digital text (a thesis) into both the print and digital media relies upon the existence of a single, defining text-model. Figure 1 shows the process involved in creating new research objects. The first stage makes possible all subsequent processes; creating a unifying model allows the generation of XML schema and subsequent XML encoding in order to manifest the new research objects (the various chapters in a research thesis). The model is driven by specific Use Cases such as the production of both a static printed version of the text, and an interactive digital version.

The source chapters are encoded at the final stages of the research process, rather than during the writing process. The model considers both presentation properties (chapter, paragraph, section), which allows for transformations specifically for presentation purposes, and semantic properties which encode the “textual semantics” (Eggins, 2004) of the text, its logical class (Teehan, 2010). This approach makes the text reusable and ensures “a single lexical can function very differently” (Landow, 2006) in different environments.

The model is translated into a schema which allows us to mark up the content of the scholarship, including narrative which in historical research pertains to ‘logical’ rather than ‘ideological’ content. We view narrative as the logical information contained in the text that contributes to a narrative of the past (Coffin, 2002). The encoding of dynamic narrative and data supports the creation of new research outputs as non-linear components derived from the text.

Stages 2 and 3 - Realising Use Cases
Stage 2 and 3 are the realisation of the various Use Cases. The XSLT transforms and software libraries are templates from which different text from different sources can be modified and transformed, in effect creating a suite of tools.

These various macros are supporting tools for manifestations of a text. Our encoded texts depict the various functions embedded in standard print theses, but augment those capabilities for these born-digital theses. Here, two specific Use Cases address (i) the creation of a dynamic bibliographic referencing model, and (ii) the context-dependent presentation of boundary objects.

A referencing model in XSLT can automatically create a dynamic bibliography for a chapter with features including “intertextual links” (Samraj, 2008) between the text and source material. Software libraries can be used to support the innate variability of a boundary object, which is defined as an object with user dependent functionality and meaning (Thomas). Thus, depending on the User’s activity and perspective, the presentation of the boundary object will change; for instance, a table diagram, static in the print version, may become interactive within a digital context.

These low-level Use Cases support our higher level one; dynamic creation of static or interactive versions of a base text-model. The print model transforms the original text to a print ready text, and can account for various institutional templates. Embedding references to the various primary sources used in the XML encoding instructs an XSLT to create a hypertext of linked resources and creates “intertextual links” (Samraj, 2008) and boundary objects for user interaction between the narrative and various digital objects within the digital medium. Figure 2

Figure 2: Text encoding of this proposal and XSLT transforms

Full Size Image

Conclusion
This paper demonstrates the process and production of support tools for digital scholarship, and how the creation of appropriate templates can make manifest various representations of Digital Humanities PhD theses from a single model. The Use Cases are reliant on the ability of the encoding and the schema to encapsulate both the functions of the text and the various transformations and software libraries. Figure 2 demonstrates the interactions between the encoded text, the transformations and the outputs.

Current research students in Digital Humanities constitute a newly ‘born digital’ generation, the nature of whose outputs differs markedly from earlier generations. Reflections on this changing process should also include an analysis of new methods and techniques to create dynamic scholarship. The encoding of the final phase in a PhD thesis allows scholarship to be reused, modified, visualised and transformed, allowing for greater distribution and accessibility of digital scholarship. Thus the dissertation, in its multiple representations, can not only remain central to the discipline of Digital Humanities but shape its future development.

References:
Mingwei, Zhao Yajun Jiang “Dissertation acknowledgements: Generic structure and linguistic features, ” Chinese journal of applied linguistics, 33 (1) 2010 94-109

Eggins, Suzanne “An introduction to systemic functional linguistics, ” Continuum, 2004

Teehan, Aja John G. Keating “'A digital edition of a Spanish 18th- century account book, Part 1: User-driven digitisation', ” Jahrbuch für Computer-Philologie, 10 2010 169-18 (link)

Landow, George P Hypertext 3.0, Critical theory and new media in an era of globalization, The John Hopkins University Press 2006

Coffin, Caroline “Constructing and giving value to the past: an investigation into secondary school history, ” . Genre and institutions, social processes in the workplace and school, Frances Christie J.R. Martin London 2002

Samraj, Betty “A discourse analysis of master’s theses across disciplines with a focus on introductions, ” Journal of English for academic purposes, 7 2008 55-67

Thomas, Robyn Sargent, Leisa D Hardy, Cynthia Power and participation in the production of boundary objects (link)