Bhakti Virtual Archive (BHAVA)
Summer 2018 Seed Grant Funding Report
Report submitted by Jon Keune and Gil Ben-Herut
Summary of seed grant activity
The summer seed grant from DH@MSU enabled us to create a set of static image mockups that were integrated into a basic website that demonstrates the search capabilities of the Bhakti Virtual Archive (BHAVA) that we are working to build: http://www.regionalbhakti.org/bhava/walkthrough/. We contracted with Addis Enterprises (AE) of Lansing, who consulted with us about what we were looking for in a user interface, pointed out issues that we will need to address when it comes time to code the project, and designed twelve images in Photoshop that illustrate what we plan for BHAVA’s functionality. We included this HTML walkthrough in our proposal to the National Endowment of the Humanities for a Humanities Collections and Reference Resources (HCRR) grant at the Foundations level.
BHAVA project overview
The Bhakti Virtual Archive (BHAVA) will be a freely accessible online platform for exploring scholarship on diverse regional devotional (bhakti) traditions in South Asia. The ultimate goal of the BHAVA project is to have an expertly curated database of research on South Asian devotional traditions that accelerates research and enhances teaching about South Asian history, literature, philosophy, language, and culture. We have been slowly developing BHAVA for the past three years and have recognized the importance of getting an external grant to support more intensive work, to make it visible to university-level metrics. To carry out this project, we are seeking support from the NEH HCRR at the Foundations level to create 1) a faceted classification system with 2) cataloging protocols, 3) a robust database that integrates linked open data, and 4) an online search interface. After this infrastructure is built, we will submit a proposal for an HCRR Implementation grant to populate the database with bibliographic records that are compiled and cataloged by thirteen experts in diverse Indic languages and regions.
The rise of regional bhakti traditions, starting in the 7th century CE, heralded major transformations across South Asia. Bhakti traditions became known for promoting emotional experience of the divine, communal participation in ritual worship, and an inclusive vision that everyone was welcome to serve the traditions’ preferred deities. These regionally based and locally organized groups deeply impacted South Asian society, philosophy, and literature. As part of cultural studies’ turn toward the popular and non-elite, over the past thirty years, scholars of South Asia have paid greater attention to the regional-linguistic contexts of bhakti traditions. Research on bhakti traditions encapsulates, in many ways, the study of popular, non-elite culture in South Asia. This approach required scholars to invest great time in learning regional languages (some of which are rarely taught outside India) and becoming familiar with regional particularities. This research trend heightened appreciation of the distinct local characters of these traditions, thereby enriching our understanding of social and cultural history. But it also had an unfortunate side effect: compartmentalizing knowledge into region- and language-specific silos. So, a new challenge arose from this approach’s own success: how to coordinate multiple, rapidly growing, largely independent bases of knowledge about a phenomenon that crossed languages and regions.
Two hurdles stand in the way. The first is conceptual. Given the diverse languages and regions involved in the study of bhakti traditions and literatures, an overarching rubric is needed to help people navigate this vast area of knowledge to address big-picture questions about the field as a whole. Identifying broad patterns across regions and languages requires a level of comparative analysis that is difficult to carry out now, with current silos of knowledge. The second hurdle is logistical and pertains to the lack of a solid platform (beyond accidental conversations at conferences) to encourage knowledge sharing among scholars of various languages, regions, and periods. Today, someone studying one particular tradition has no central resource to identify relevant bhakti-related materials in other languages and regions. This problem is exacerbated by the fact that most research on bhakti is published in journal articles and edited volumes, which are not cataloged by the Library of Congress intensively like monographs, leaving them less visible and well connected in the digital network of humanistic knowledge. Consequently, bhakti scholarship (with all its implications about historical, literary, and social knowledge) has tended to proceed by focusing on a partial understanding of the history and significance of a given tradition while overlooking its development, cross-fertilization, and interaction with traditions elsewhere. BHAVA will provide a way to link region- and language-specific silos of knowledge together, so that people within and outside academia can find connections and deeper understandings of Indian culture and history.
The bibliographic holdings of BHAVA will refer mainly to secondary scholarship in Western and Indic languages about bhakti traditions, as well as to key primary texts, in translation and in original languages. By integrating linked data to interface with external catalogs and repositories (WorldCat, Archive.org, Transliteral.org, etc.) and other South Asia-related digital resources (University of Chicago’s Digital South Asia Library, the South Asia Open Archive, the PANDiT Project, and Sahapedia.org) the project will be well situated for further possible content expansion (images, video, geographic data, e-texts, and manuscripts) and continued growth, after the Foundations and Implementation phases of BHAVA’s development are complete.
Mockups and walkthrough
Although we had roughly envisioned the search features that we wanted BHAVA to have, working with AE to create the interface images brought these hopes into clearer focus. AE asked us to refer them to examples of existing websites that had some of the features we wanted. We knew that BHAVA’s search interface would be based on users selecting subjects from our faceted classification table that will be used to catalog each bibliographic reference in BHAVA. Designing this aspect of the search (what eventually become the Standard Search Mode in the mockups) was just a matter of deciding on layout. What we didn’t anticipate was that in the process of looking at other search interfaces online, we were inspired by one site (American Merchant Marine Veterans Oral History Project at http://seamenschurch-archives.org/sci-ammv/tagcloud) to imagine using a word cloud as an alternative interface that would visually represent the relative frequency of subjects in BHAVA’s holdings. Based on users’ selections (single or multiple) in one facet’s word cloud, other facets could then be represented to aid further searching. Not only would this be visually engaging, but it would build a level of interactivity into the search activity itself; the world clouds would help users anticipate the number of discoverable items within their selected parameters. This eventually became the Word Cloud Browse mode of the search in the mockups. Although describing the Word Cloud Browse feature is difficult in a narrative, it is easily understandable when viewed in the HTML walkthrough.
We provided AE with some potential search terms and imagined results numbers (educated guesses), so that they could create images that would lead users through a few potential searches in the BHAVA interface. In the HTML walkthrough, we also wrote short commentaries on each of the twelve image mockups.
Challenges
A challenge that we faced immediately after receiving the seed grant was that we hadn’t anticipated how limited our hiring options would be because of the MSU accounting office. We soon discovered anyone we wanted to hire to do the mockup work would have to be vetted by MSU Purchasing—a process that could take several weeks. This would not have allowed us to have the mockups ready before the NEH HCRR submission deadline. We were eventually directed toward a handful of design and coding services that MSU had pre-vetted, whose rates were higher than we could have found with freelancers. We are pleased with AE’s work, be we had anticipated being able to afford more work with the seed grant and a couple weeks went by as we sought answers from MSU administrators how to do this kind of contracting.
The mockup creation process left us with several questions about design that we will need to address in the future. 1) In the Standard Search Mode, the terms of nearly all the facets can be displayed vertically at the same time on the search interface. But the Topic facet will have too many terms and must be represented with a pop-up balloon that users will have to click to open. If additional terms get added to other facets eventually, we may encounter a similar problem with them. 2) We haven’t yet found an elegant way to incorporate Boolean operators into the search modes on a search interface screen that is already fairly full. By default, we would assume that items selected within a facet would be governed by OR, while items selected across facets would be governed by AND. We’ll need to think more about this. 3) We will also need to think more about enabling users to search on details of bibliographic references themselves (author name, publication year, etc.). Currently, this kind of search is possible only through the Keyword field. We always planned that BHAVA’s strength would come from the faceted classification table that will be used to catalog each bibliographic record. This level of detailed cataloging will lead users to make connections that they could not have made before. But perhaps users may miss searching on conventional fields like author name. We will think more about this, including how those fields could be laid out on the screen. 4) We will need to consider further how using Linked Open Data in the search interface will enable us to link to external resources meaningfully and in ways that can be coded affordably. 5) As AE created the mockups, they built in a “Preview” feature on the results screen, so that a user could hover over an individual bibliographic item in the results list and see some details pop up; this potentially makes the full view of a single bibliographic item redundant.
