Introduction
A 2009 issue of Digital Humanities Quarterly addressed issues surrounding the completion of digital humanities projects. This is of great interest to those of us who engage in the development of large-scale humanities computing projects, for unlike printed articles and books, which are essentially done once they have been published (unless one is fortunate enough to get a second printing), digital projects are potentially much more fluid and transient. Indeed, this flexibility is one of the digital mediums greatest strengths since new developments in scholarship can be readily integrated. It is also one of the greatest challenges facing digital humanists, for if changes can be so easily incorporated is a work ever truly completed? In one of the articles from that issue of Digital Humanities Quarterly Susan Brown et al navigated a middle ground around this issue of completion by suggesting that it is “valuable ... to consider particular tasks and stages done, even as the capacities of digital media push against a sense of finality” (Brown, 2009). Yet this approach tends to presuppose a linear project development path – that a project is necessarily and constantly evolving from one iteration to another, pushing toward some final, unreachable goal.
This commitment to a linear development path may in part stem from familiarity; as Michael Best has proposed, there is a tendency among academics “generally to be safe, to progress through accepted and understood channels” (Best, 2006). As a result, we tend to fall back into the academic paradigm in which we were trained by developing our digital projects along the same line as our written work – by producing drafts which we then rework and edit until ready for publication/launch. However, this is not necessarily the only path that a digital project will take, as the very nature of project design in the digital humanities can allow for a lateral development path. In particular, the use of modular design in humanities computing allows us to reconstitute the infrastructure from previously developed projects for different purposes, thereby reclaiming and revitalizing parts of “completed” projects for other initiatives that use newer technologies and methodologies or that seek to achieve a different outcome. As such, lateral project development does not require any changes or alterations to the original project. Indeed, these secondary or sibling projects (describing the link between the projects as a “parent-child” relationship suggests a linear development path from which I want to diverge) can be initiated even as the original project is still being developed.
This paper draws on my experience working with the University of New Brunswick (UNB) Electronic Text Centre (ETC) to digitise the papers of prominent New Brunswick Loyalist Edward Winslow to address issues relating to the development of laterally developed projects. It argues that there are many reasons to develop projects laterally. First, it can be financially prudent to develop projects laterally; given the difficulties in securing grant money for projects, lateral development can assist digital humanists in building new and more robust projects on the shoulders of previously completed work. Second, lateral project development can facilitate greater production in the digital humanities, as laterally developed projects need not be undertaken by the researcher or institution that developed the original. Indeed, once the core elements of a project have been published online any other researcher can build upon this previously completed work. However, this presents a bit of a problem, as to fully realise the potential of lateral project development we will have to do away with a proprietary mindset with regard to our digital projects and be willing to share already completed project elements and modules with other researchers.
Modular design
Before examining lateral project development, it is useful to give a brief overview of modular design principles, for the sharing of project modules is what facilitates collaborative lateral project development. At the most basic level, modular design is the breaking up of complex systems or projects into smaller and more manageable chunks or modules. These modules are often developed in isolation; when all of a project's various modules are completed they are stitched together to effect the finished project. This subdivision of development tasks into a variety of modules is a design principle widely used in the corporate sector.
There are a number of reasons that we, as digital humanists, should adopt a modular design philosophy in our projects. First, the fact that much of the corporate sector has adopted this design philosophy suggests that it can improve corporate profit margins. With funding for our projects increasingly at a premium, there is much to be said for leveraging the most “bang for our buck” when developing projects. Second, in many respects we already design our work modularly – for example, the projects developed around the Edward Winslow Family Papers detailed below utilised a modular design philosophy. Finally, the use of modular design facilitates lateral project development. The use of modular design in computer programming allows for the possibility of moving from sequential computing – where one process must be completed before another can be begun – to parallel computing – where these processes can be run simultaneously and concurrently. This same principle holds true for humanities computing projects. Rather than developing our projects linearly and sequentially – beginning one project and seeing it through to completion before beginning another – we can instead develop our projects concurrently and laterally.
The Edward Winslow Family Papers
Housed at the UNB University Archives, the Edward Winslow Family Papers are the largest extant collection of Loyalist documents in the world, consisting of 38 volumes of correspondence, accounts and ledgers, letter books, diaries, and notebooks with a chronology that ranges from 1695 to 1866. All told, the collection contains over 3,600 separate documents in excess of 11,000 pages, with the bulk of the collection consisting of the correspondences between Edward Winslow Jr. and his friends. An incredibly rich historical resource, they remain perhaps the single most important collection of personal papers describing the Loyalist experience in the wake of the American Revolution and one of the most valuable collections held by UNB.
Edward Winslow is of particular interest to scholars for two reasons. First, as Muster Master General during the Revolution he was an integral part of British colonial society who either saw firsthand or read detailed accounts of the horrors of the war. Even more importantly, following the conflict’s resolution he joined the Loyalist Diaspora and became a key player in the political and social life of post-Loyalist Nova Scotia. In this capacity he assisted in the partition of New Brunswick from Nova Scotia, served in a variety of positions in the new government, and was a vital part of New Brunswick high society until his death in 1815 at age 70. Second, Winslow was a prolific and gifted writer who left behind an incredibly engaging, humorous, and rich series of correspondence through which the eighteenth century comes alive on the page. His papers are a veritable treasure trove for scholars of colonial British North America.
As a result of the collection’s historical value, the UNB ETC was quite keen to digitise the papers. Since its founding in 1996, the ETC has been at the Canadian forefront in the use of emerging technologies to promote and deliver scholarly research, and it continues to publish to standards and best-practices electronic journals, images, and special collection texts such as the Chadwyck-Healey poetry database (Burk and Charlong, 2006). This commitment to the use of electronic resources for the promotion of scholarly communication led the ETC in 2000 to cooperate with the University Archives in undertaking the ambitious task of digitizing all 38 volumes of Edward Winslow's papers. The ETC's goals for the website were rather modest: some background papers on the collection's preservation and provenance would be presented, along with a brief introduction to the history of both Edward Winslow and the Loyalist Diaspora and a tribute to noted UNB Loyalist scholar Ann Gorman Condon. Not only would this project digitally preserve this valuable collection, it would greatly increase scholarly access to a valuable, yet underutilised, source through web publication. The format chosen for online presentation would be that of an image archive, featuring a searchable database containing images of every letter in the collection. Search functionality would be facilitated through document metadata, which would include a digitised version of an index of the Winslow Papers compiled in the 1980s by amateur Fredericton historian Barry Grant. However, what the archive lacked in depth would be made up for with breadth, as every item from the Winslow papers – in excess of 10,000 pages – would be presented in the searchable archive.
The digitisation of the Edward Winslow Family Papers was begun by the ETC in 2000. From the outset a modular design philosophy was used. For example, the digital imaging of all the archival documents can be considered a module. So too can the development of the metadata for each of the imaged documents. Other modules include the construction of the database that holds this metadata and links it to the relevant image files, and the design of the web interface that allows the collection to be presented online. Each of these modules was completed separately and then stitched together at the end in order to complete the project. The use of modular design in the development of the Edward Winslow Family Papers image archive allows for greater long-term project viability. As opposed to projects that are interdependent – that is, projects produced as one heterogeneous whole rather than as a series of interlinked modules – modular projects allow us to remove and replace one part of the whole without having to redo the entire project. For example, should new advances in digital imaging techniques render the digital surrogate JPEG images for the Edward Winslow Family Papers obsolete, new image files can be created from the lossless TIFF files of each document. These new image files could then be re-integrated into the website with very little alteration to other project modules..
In a similar manner, completed modules can be taken and integrated into other modularly developed projects. For example, the images captured for the Edward Winslow Family Papers image archive and the metadata that describes these images can be repurposed in another virtual archive. The image and metadata modules could thus be the same in both projects, but the end result would be different due to the variety of the modules that surround the images and metadata. This is precisely the approach that was taken when the UNB ETC undertook the lateral development of a more robust image archive based on Winslow’s papers.
Lateral development and the Winslow Papers
Even as the ETC was completing the imaging for the image archive website, the limitations of the project were becoming clear. While the value of putting the entire collection online was never questioned, it seemed clear that the digital humanities allowed for much more to be done with the Winslow papers than the static preservation of searchable document images. The catalyst for exploring the digital humanities via the Winslow Papers was Dr. Margaret Conrad, who arrived at UNB in 2002 as the Canada Research Chair in Atlantic Canada Studies. One of her chair's mandates was to increase access to scholarly material on Canada's Atlantic Region (the provinces of New Brunswick, Nova Scotia, Prince Edward Island, and Newfoundland and Labrador).
In cooperation with the ETC, Conrad began the development of a portal website designed to provide scholars of Atlantic Canada with "one stop shopping" for resources pertaining to the region. Launched in 2003, the Atlantic Canada Portal (http://atlanticportal.hil.unb.ca) hosts a searchable bibliography of works published on the region, an EPrint repository, and a wide selection of annotated links to websites on Atlantic Canada. In addition, the portal hosts the Atlantic Canada Virtual Archives (ACVA), a digital archive of historical documents pertaining to the Atlantic Provinces. Currently, the ACVA hosts five archival collections, two of which were developed laterally by drawing extensively on the infrastructure developed for the Winslow image archive. These two sites present the stories of the early years of the Loyalist migration to British North America and the lifeways of Loyalist Women. The ongoing use of images and metadata from the image archive in the development of more robust ACVA digital history initiatives illustrates the relationship between modular design and lateral project development.
The first project chosen for inclusion in the ACVA was based on the Edward Winslow Papers and told the story of the early years of Loyalist settlement in Nova Scotia and the newly formed New Brunswick. Later, the Edward Winslow papers were mined for all letters to or from the Winslow sisters in order to reveal the day to day lives of elite Loyalist women. By developing these projects laterally – that is, by drawing on much of the project infrastructure from the image archive – the ETC was better able to secure outside funding through a grant from the Canadian Culture Online Program (CCOP) of the federal Canadian Heritage department. Moreover, since all funding for the imaging of the papers came from the library's internal budget, we were able to list the earlier costs associated with digitisation as an in kind contribution on grant applications. The importance of this should not be overlooked for, as William Kretzshmar has suggested, granting agencies are reluctant to fund projects that have previously received money and are considered finished (Kretzshmar, 2009).
Differences between the Winslow and ACVA websites
Although both websites were developed using the same images and metadata, significant variation exists to differentiate the foundational Edward Winslow Family Papers image archive and the two laterally developed ACVA projects. At the most basic level, the amount of background information pertaining to the Loyalists is significantly greater on the two latter projects. While the original Edward Winslow Family Papers image archive features a brief introduction to the collection, a series of brief essays designed to show site visitors thematic “trails” that can be followed in the collection, and a paper on the Loyalists penned by the late Loyalist scholar Ann Gorman Condon, the ACVA websites feature a number of reprints of peer reviewed articles, a number of contextual pieces specifically written to situate the material contained on the website into the broader historiography, and a detailed bibliography of suggestions for further reading. In short, the two ACVA sites are much fuller introductions to themes in Loyalist historiography.
However, there are two more significant markers of difference between the Edward Winslow Papers image archive and the ACVA websites. First, the foundational image archive takes a broad approach to digitisation and has placed all of Winslow’s papers online for the world to see. The ACVA sites, on the other hand, adopt a more thematic approach. The first website focuses on the early years of the Loyalist arrival in Nova Scotia, when the partition of the colony and the creation of New Brunswick occurred, while the second site examines correspondences from Loyalist women in order to provide a gendered analysis of the Loyalist experience. Second, and of greater importance for digital humanists, is the inclusion in the ACVA websites of searchable transcriptions of the document’s text. All documents contained in the ACVA have been transcribed, encoded, and presented alongside the document images as a complement to the original material. Although for the professional historian the searchability and features of the ACVA websites may be trumped by the breadth of coverage contained in the original Edward Winslow Family Papers image archive, for most casual visitors the ACVA sites provide a much more thorough and interesting introduction to the lives and the lived experiences of the Loyalists.
In many ways, the digitisation of the text on the ACVA websites was facilitated by the more limited thematic approach to the source material. Recognising the difficulty in transcribing and encoding all 38 volumes of Winslow's papers – indeed, the impossibility of transcribing and encoding 38 volumes given the project time-frame dictated by the CCOP grant that funded the project – the initial ACVA project focused on only the three volumes pertaining to the years 1783-85. After all, these three momentous years encompassed the Loyalists' voyage from New York to Shelburne, Nova Scotia, their difficulties integrating with Nova Scotia's landed elite, their subsequent voyage to the St. John River Valley, and the founding of the new colony of New Brunswick. All told, 260 of the approximately 450 letters written to or from Edward Winslow during this three year period are included in the ACVA Edward Winslow website. The Loyalist women site, developed later, is smaller still, and contains 62 documents with a chronology spanning from 1783 until 1828.
Since the imaging of the documents had already been completed for the Winslow image archive, the teams developing the ACVA Winslow and Loyalist Women sites were able to focus their attention on transcribing and encoding the document text. This task was made somewhat easier because of the presence of W.O. Raymond's 1901 Winslow Papers, A.D., 1776-1826, which featured transcriptions of many of the collection's more important letters; only those letters that had not already been treated by Raymond needed to be transcribed (Raymond, 1901). Documents were transcribed and encoded according to applicable industry standards for primary source texts. Person and place names were encoded, as well as any structural or textual elements that may have been missed. In addition, project encoders encoded archaic or abbreviated language. This allows site visitors who are viewing document transcriptions to toggle between two display options. This means that although the actual document images present on each site may be the same,
Figure 1:
Figure 2:
Figure 3:
Figure 4:
Another factor that differentiates the Winslow image archive from its laterally developed ACVA counterparts is the presence of a learning resources section on the latter projects. While an educational section was a prerequisite for the funding that supported ACVA's development, it is also a means of introducing students to the processes of conducting historical research through the digital humanities. Lesson plans are available for teachers of students in Grades six, seven and eight that make extensive use of the archival holdings and familiarise students not only with such concepts as primary and secondary sources, but with how to use these sources to understand the daily lives of New Brunswick's Loyalist settlers. There are also learning resources on each site designed to make learning about the past more interesting. For example, using Flash animation the ACVA 1783-1785 website allows visitors to write with a virtual quill pen which occasionally has to be dipped in an inkwell, much as Edward Winslow would have had to,
Figure 5:
Figure 6:
Figure 7:
Sharing our work: lateral project development writ large
The utility of using modular design to facilitate lateral project development should not be seen merely as a means of maximizing funding opportunities in the digital humanities by reconstituting previous work. It also has broader implications for the development of digital humanities projects. For example, by focusing on core modules of digital project development – images and metadata – one can effectively set the stage for later efforts to be developed laterally from this initial undertaking. Indeed, this focus on images and metadata – the foundational elements for any virtual repository – allows for greater opportunities to develop new and innovative projects later; once the original project is available online, anyone, from any location with an internet connection, can examine the artefacts presented on the project website. Of greater importance, anyone can take these core elements of the project and use them as the foundation of their own digital humanities initiatives.
Our work as digital humanists need not be confined by our institutional affiliation or geographic location. Projects building on our work can be undertaken around the world. Moreover, such sharing of core elements might invite more collaborative work between institutions. This is of vital importance, at least for those of us engaged in the digital humanities in Canada. The Canadian Social Science and Humanities Research Council has recently revamped its funding programs and implemented grants for digital humanists specifically geared toward collaborative endeavors. Both the Partnership Grants and the Partnership Development Grants set up by SSHRC require inter-institutional collaboration in order to qualify for funding (SSHRC, 2011). In the future perhaps projects could be undertaken with the UNB ETC that utilise the Winslow papers in which the digitisation work already completed for the broader image archive can be listed as an in kind contribution. Yet this exercise in the democratisation of our work is dependent on freely and openly sharing the core elements – images and metadata – of work that has already been completed.
Indeed, the importance and the practicality of sharing digital facsimiles of archival holding has recently become the topic of online debate. The listserv H-France recently had a lengthy string of posts by French historians that discussed the desirability and feasibility of sharing digital images of documents collected by scholars on research trips (H-France, 2011). On the whole, the sentiments of the discussion participants seemed to be mixed. While most recognised the utility of sharing the archival images, some scholars displayed a reluctance to do so (or to commit to do so) until they had used those sources in their own scholarly endeavors. Although this reluctance to share archival material might be expected to arise in a discussion among scholars who have a vested interest in using those sources in their own scholarly pursuits – publish or perish – it is hoped that among digital humanists – who would likely be more concerned with the process by which the archival images are placed online and the ways that they could be digitally manipulated than with the actual content of the archival images – a consensus regarding the sharing of digitised resources can be easily reached. If the digital humanities are to be a truly open source initiative, we must move past simply using freely available software in order to complete our projects and accept the utility of allowing colleagues from around the world to build on our project architecture – to use our image and metadata modules – in order to laterally develop projects that will ideally continue to expand the envelope of what can be accomplished when we meld computers with the humanities.
The sharing of our digitised material is not without its potential problems. For example, concerns that allowing others free access to digitised documents created for other projects might lead to people exploiting these digital images for profit in some fashion are understandable. However, these reservations could potentially be alleviated through the application of a Creative Commons License that specified that the digital work can only be reconstituted for non-profit scholarly activities. Despite this, the potential of sharing the foundational elements of our work to facilitate lateral project development and encourage a greater variety of digitisation efforts far outweighs any concerns over unscrupulous profiteering off of this work.
Conclusion
It is hoped that the two ACVA websites detailed above provide a model for future laterally-developed digital humanities projects. Since all documents in the collection had already been imaged for the Winslow image archive, and since much of the infrastructure of the two later ACVA projects can easily be replicated, funding is required primarily to transcribe and encode selected documents and to develop related site content, such as historical context and learning resources specific to the themes or letters being explored. Thus, the older, sibling Winslow image archive can remain the foundation for subsequent and more robust initiatives that can be undertaken laterally at either UNB or at another institution elsewhere around the globe. As Susan Brown contends, “the first iteration often merely begins to tap the potential of the projects data architecture and potential for interface development” (Brown, 2009). This truthfulness of this sentiment is exposed – and hopefully will continue to be revealed – when applied to UNB’s experience creating the Winslow image archive and to the development of its laterally developed sibling projects.