Sustainability of The Treasury of Lives

This is a guest post by Alex Gardner and Catherine Tsuji

Background

On June 3, 2023, the Treasury of Lives participated in an online forum at Digital Orientalist on the topic of “sustainability.” As this topic has been at the core of our work over the last seventeen years, we were grateful for the opportunity to share our experiences and learn from others in the field. While financial sustainability is a key concern for every digital project, data management is equally vital. 

The Treasury of Lives (TOL) is an online biographical encyclopedia of Tibet and associated cultural regions. In continuous development since 2007, it was initially a project of the Shelley & Donald Rubin Foundation and became an independent nonprofit in 2016. TOL has published over 1,400 biographies together with associated data such as geography, tradition, and social roles, as well as annotated images and documents. After relying for fifteen years on a relational database, TOL is working on transitioning to Linked Open Data using a custom-built ontology and a populated knowledge graph. This will not only ensure greater capacity for preservation and expansion but also provide increased access to data and the underlying model.

TOL was developed alongside the Buddhist Digital Resource Center (BDRC, formerly the Tibetan Buddhist Resource Center). Person and related data for incarnation lineages (i.e., the Dalai Lama, Karmapa, Panchen Lama etc.) and institutional roles (i.e., monastery abbots, government officials) was imported from BDRC to form the basis of the TOL relational database, which uses MySQL with a content management system (CMS) based on CodeIgniter, an open source PHP framework for building websites. 

Originally, the project was conceived as a resource for information on specific “lineages” (meaning traditions) of Tibetan Buddhism through biographies, and in fact the early URL for the site was tibetanlineages.com. Early biographies were short narrative biographies, designed to spur further research but by no means comprehensive. Over time, the essays became increasingly complex and more metadata was collected, and it also became necessary to extract data from narrative content to create new or expand existing taxonomies. This time-consuming process is ongoing, as new narratives are mined and linked to data internally and externally as new categories arise. 

The limitations of relying exclusively on BDRC’s data became apparent over time. A major constraint arose when authors began contributing biographies for which there was no BDRC person record, which the CMS was simply not designed to accommodate. Similarly, the CMS was not designed to accommodate geographic records without an equivalent BDRC record. Although the backend database created unique TOL identifiers for its unique data, these were not visible in the CMS, where all data is managed and edited. Although TOL editors had the ability to create BDRC person records in the BDRC database, as new unique TOL records began to rapidly accumulate we were unable to keep pace. In 2020 we revised our CMS to allow us to develop our own data and edit our own IDs to accommodate a growing corpus of life stories expanding beyond pre-modern religious masters and including information on Tibetan Buddhist sites outside of the Himalayan region. We remain grateful to BDRC Librarians Lobsang Shastri, Karma Gongde and Chungdak Nangpa, who have reviewed and created new records based on TOL content, offering their insights and feedback whenever needed. 

Planning for Sustainability

In 2018 TOL received an 18-month planning grant from the National Endowment of the Humanities (NEH) that enabled us to look toward sustainability of the resource. Around this time we implemented peer-review of all submissions, in recognition that the site had become an authoritative academic resource. With NEH funding we were able to engage experts from a variety of disciplines to think about the future of the resource and what we needed to ensure its long-term survival. This project came at a crucial time, as we were just beginning to operate as an independent nonprofit organization while also assessing our mission and determining future expansion. In the process of developing a storage and preservation plan we were able to better articulate our scope and mission, refine the handling of backups and logs, streamline our editorial workflow, identify scholarly repositories for long-term preservation, and review technical dependencies such as our reliance on BDRC data and external APIs such as Mapbox.

We continued to assess what expansion and growth should look like. While we remained dedicated to the core mission of publishing biographies of Tibetan Buddhist teachers, at the same time new types of data and biographies were being added. We had to consider whether we would ever add biographies of Western explorers and scholars of Tibet. We also had to consider expanding our map as new biographies detailed activity that occurred outside of the Tibetan cultural region. In short, we had to define our parameters as we continued to grow. Establishing and periodically updating Style Guides and Standards documents establish our approaches and allow potential authors to familiarize themselves with editorial considerations and expectations.

In addition to establishing guidelines and processes that support long term sustainability, financial sustainability is a constant consideration. Although the resource is open-access and content is freely available without a paywall from anywhere in the world, naturally there are costs to operate, and our independence required that we raise our funding entirely on our own. Government and foundation grants, individual donations, and a subscription program geared toward universities are our main sources of income. The latter, which asks academic institutions to recognize their responsibility in supporting digital humanities resources on which their faculty and students rely, offers subscribers a layer of specialized tools useful to students and faculty. 

Current Redevelopment

A subsequent three-year implementation grant from the NEH (2020–2023) has supported a significant expansion of TOL and is focused around three goals: 

1) increasing content with the addition of 100 biographies of 20th-century Tibetan and Himalayan individuals and related data

2) creating an ontology that models people and places of the Tibetan cultural region

3) replacing the existing relational database with a triple store in order to facilitate a Linked Open Data model that will increase internal data inference and foster collaboration with external resources

In October 2023, we completed the first goal, the publication of one hundred peer-reviewed biographies. These add considerable depth and variety to the Treasury of Lives by opening the door into recent history. They include some of the twentieth century’s most important religious leaders, as well as military figures, chieftains, translators, doctors, estate managers, hidden yogis, and contemporary scholars. They document how religious and political authority is distributed within families as well as across institutional structures, and illustrate the individual experiences of the politically complex twentieth century. Of these one hundred biographies, nineteen are of women, admittedly a small percentage but one that meaningfully contributes to the Treasury of Lives’ ambition to increase the number of women on the site. Many of these new biographies are of lamas who were foundational teachers of Tibetan Buddhism in North America and Europe. All benefit from extensive linking to new and existing Treasury biographies and place records, as well as to external data at the Buddhist Digital Resource Center.

All of this new content has informed the development of our in-progress ontology designed to show relationships between people and places in the Tibetan cultural region. For example, as we began to gather data on family relationships we needed to create new ways to model these sometimes complex relationships, such as the marriage between Dasang Damdul and three sisters of the Tsarong family (Pema Dolkar, Rinchen Dolma, and Tseten Dolkar) and the widow (Rigdzin Chodron) of the family heir, Samdrub Tsering. Another challenge has been modeling incarnation lines that split into multiple branches, such as the many reincarnations of the famous nineteenth-century lama Jamgon Kongtrul. 

As the ontology is being developed, we are working on migrating our relational data to RDF (Resource Description Framework), the data format used for Linked Data. As RDF has strict requirements for formatting and effort is being spent ensuring that languages, scripts, dates are properly formatted so that they can be processed. Data for places and their types needs to be property categorized according to the new ontology before conversion to RDF, another time- consuming process. One major challenge is that while all this backend cleanup is ongoing, new content is being added at a rapid pace as new biographies are submitted. (To date 139 authors have published biographies on TOL.) 

Throughout this process, we have been extremely fortunate to work with ontologist Rebecca Younes, who is writing the ontology under the auspices of Semantic Arts, Inc., an information systems consulting firm that specializes in semantic modeling, knowledge graphs, and data-centric enterprise architecture. We are grateful to Tenzin Dickie, former editor at ToL and later communications manager at BDRC, a long-term contributor and advisor whose involvement has driven meaningful content expansions over the course of the last decade. Our work has also been informed by the pioneering efforts of BDRC and its lead developer Élie Roux, who has been a constant source of information and support as we transition to a new database. (We highly recommend reading his article, “Adopting RDF at a large humanities digital library: lessons from a challenging success”, for a more technical discussion of the transition to RDF.) One of the greatest benefits of this project will be more seamless data sharing between TOL and BDRC.

Ultimately, we hope that the transition to a semantic knowledge graph with a new data model will mean that in the long term the data itself, and the relationships between people, places, kinship groups, and social roles it expresses, will be easy to access, preserve and replicate.

______

Catherine Tsuji has been an Editor at The Treasury of Lives since 2014.

Alexander Gardner is the Director of Treasury of Lives. He completed his PhD at the University of Michigan and is the author of The Life of Jamgon Kongtrul the Great (Shambhala, 2019).