THL Toolbox > Fieldwork > Digital Ethnography
Contributor(s): David Germano
These are guidelines for final production of publications that draw upon fieldwork and historical research on places and communities.
The first step upon return is to create an online Worksite dedicated to the project in question. THL's worksites provide a standard set of tools for communication, archiving resources, and collaborative authoring. A worksite provides a central online site where materials can be stored, plans can be maintained and revised, emails to the working group archived, and much more. Contact THL (thdl virginia.edu - put the @ sign where the space is) to request such a worksite be created for your project.
Fieldwork usually involves notes taken in the fields, which are often done in bound notebooks.
Field notes are perhaps best stored in Wikis, where they can be easily privately archived, or even published, but be easily modified by the scholar. If there is no time or desire to enter the data as text in the Wikis, they can also be easily PDFed, and stored in the Resources section of the Worksite and organized by subdirectors as makes sense.
Ultimately, of course, any scholarly project will involve substantial essays that formally address various topics.
Formal essays are best marked up in XML for formal publication. XML is a way of preparing the text for Web publication which enables powerful searching, flexible display, and much else. XML has various ways of being implemented using "DTDs" or "schema" which are essentially toolkits that define the types of things you want to mark up - say a title, or personal names, and so forth. THL uses the international standard for humanities text, which is called TEI, or Text Encoding Initiative. THL tends to use Oxygen as its commercial XML Editor of choice, since UVa provides a site license for local use. However, there is a variety of commercial and free alternatives.
The problem with XML is workflow, since most scholars do not have XML editors and are only familiar with the use of word processors, such as Microsoft Word. Thus THL has created a clumsy but viable system for scholars to use Microsoft word in a particular way to create word processing documents that THL staff can efficiently transform into XML documents. It basically works by THL staff having taken TEI markup, and defined "styles" in Word which correspond to individual markup items. Thus the scholarly only needs to learn how to apply styles in Word, and then THL staff use a Visual Basic macro to convert those styles into XML markup.
The actual end presentation of such XML-marked up essays is done through "XSLT stylesheets" which interpret the markup into a visual presentation for the web (i.e. make titles appear bold face, and so forth). The best example of this in THL is in the Journal of the International Association for Tibetan Studies (www.jiats.org). There you can see the basic features of the THL essay publication system, which includes hyperlinked table of contents, hyperlinked footnotes, and interactive glossary of terminology.
On the horizon, THL is now working on the dynamic connection of specific terms in the essays to reference resources, such as our Dictionary, Place Dictionary, Bibliographies, and more. The basic idea is that wherever, say, a place name occurs, it would be automatically linked to the entry for that place in the Place Dictionary. In addition, we are working on integrating the essay publication system with the media management system so that media with their metadata (captions, dates, etc.) are automatically pulled in for attractive and integrated display within the essay at the point where they are cited. The image, or audio-video, should appear within the essay, whether than the user being pulled away to some other web page devoted to images, while ussers should be able to easily proceed from the image, or audio-video to both consulting the full descriptive data about the media object, and to request that other such media objects (with the same theme, or from the same place, etc.) to shown to them.
Fieldwork usually involves at least photographic documentation of places, people, and the landscape, and often involves audio and video recordigns.
We are currently working on a Media Management System (using Ruby on Rails) which provide an online editorial interface and end user publication system for cataloging and storing images, audio recording, video recordings, immersive objects. This allows also for classifying and indexing the media objects according to place, project, collections, subject, and keywords.
On the horizon, we are also dealing with advanced issues relating to transcription of audio-video and its searching. This starts with transcribing and translating, but also includes time-coding and annotation. On the user end, it involves searching and synchronized play back of media and transcription, as well as automated creation of subtitled movies, and the ability to search the transcripts and then select segments of recordings that you combined together into a new user-composed recording.
In the past THL has used its own custom made software for transcription, but we are looking into whether we might instead use Elan or Transana, developed at the Max Planck Institute and U Wisconsin system respectively.
Of course all academic work involves the citation of sources, such as journal articles, books, and web sites. In formal publications, one wants to be able to enter a source once and then repeatedly draw upon that, as well as enable users to consult bibliographies overall as well as cite single sources.
THL currently has a general bibliographical management system for such resources, including links to the PDFs, etc. of a reprint. It is a PHP-MYSQL system called Scout Portal Toolkit. It allows for detailed cataloging, linking to PDFs, and thematic classification which can then be easily browsed.
On the horizon, we plan to integrate this functionality into the Ruby on Rails Multimedia Management System.
In addition, THL has a very elaborate, custom-made bibliographical system for Tibetan literature which function as “thematic research collections” focused on specific bodies of literature and designed to use catalogs as front ends for reproduction of editions, translations and scholarship for that literature.
A key element of all fieldwork is documenting places, whether drawing maps or creating inventories of features and their interrelationships. Components of this include documenting place names in different forms, locating place names with latitude/longitude, documenting the relationship between features, creating a typology of types of features, and creating various thematic maps.
THL has created a complex Place Dictionary for documenting features, their names, location, characteristics, and relationships. This includes output to a GIS-based interactive map based on their latitude and longitude values. The Gazetteer is stored and edited online using Ruby on Rails, but is delivered using XML output and XSLT stylesheets.
GIS Mapping is currently being dealt with using Map Server, but we are exploring other such services. We have also used Flash to create more specialized and labor-intensive interactive maps. We have also created static PDFs from GIS software.
On the horizon, we need to implement functionality in Map Server that will allow users to click on a toponyms therein, and access information and associated resources for that feature. We have done much work at using animation to show historical change using maps, but would like to explore this in the near future.
Documentation: Places & Geography.
Documenting temporal change in communities, and especially broader historical changes, involves documenting events and creating timelines.
THL has created an offline File Maker Pro database and later an online but only partially implemented Django database for documenting events. The database allows for their naming, description, typing, details on the date according to different calendrical systems, and the citation of course information. We plan to replace this with a fully functional Ruby on Rails database.
MIT has offered an excellent tool named Simile for visualizing timelines, which we would like to explore use.
The Institute of Advanced Technology in the Humanities is currently working with Kurtis Schaeffer's Mapping the Dalai Lamas project to create a resource for documenting time as well, which THL may use.
Documentation: Temporal Issues.
In structuring your Web site's presentation, you should keep in mind the utility of searching and browsing.
In conducting searches of all media (including of course text), what keywords and other parameters would you like to be able to use to constrain queries?