Research projects in the humanities and social sciences increasingly generate substantial digital datasets, yet for the most part, universities lack the institutional repositories, mandates or long-term infrastructures required to preserve these resources once project funding ends. As a result, datasets of high cultural and linguistic value remain vulnerable to loss, even when they have been created through decades of public investment. This paper examines this challenge through my long-term involvement in a number of research programmes across the island of Ireland: the Northern Ireland Place-Name Project (NIPNP), OS200 and the Logainm Project (along with The Placenames Branch) in the Republic of Ireland who have joint responsibility for maintaining place-name infrastructure. The wider policy context shapes all of these projects, yet the central issue remains the same. Without an institutional repository or a defined preservation pathway, university-hosted datasets face significant risks to their sustainability. The NIPNP database contains almost forty years of linguistic and historical research funded primarily by government departments. Although it benefitted in its early stages from technical support from government periods of intermittent funding resulted in fluctuations in database management and maintenance. Today, the database resides at QUB, but only as long as the project continues to secure external funding and maintain an active PI, neither of which can be guaranteed in today's fiscal climate. The university does not have a remit or infrastructure to guarantee this (or any) database’s preservation beyond the life of the project. This creates recurrent vulnerability datasets that are relied upon by researchers, public bodies and the wider community. Further, NIPNP and its digital content must also operate within a policy environment in Northern Ireland that is currently shifting, as new responsibilities for Irish place-name provision are expected to arise under the Identity and Language Act, yet remain undefined. The OS200 project provides an interesting comparison by showing what becomes possible when long-term sustainability is considered at the outset. OS200 (a cross-border initiative) was designed with a clear preservation pathway, and its dataset will be deposited with the Digital Repository of Ireland. This avoids the vulnerabilities associated with university-only hosting and demonstrates how planned alignment with trusted national infrastructures can secure long-term access. More can also be learned by looking at a similar dataset in the Republic of Ireland, where long-term preservation is supported through a stable partnership between Fiontar at DCU and a dedicated civil service team in the Placenames Branch. This arrangement is supported by clear statutory responsibilities under language legislation, which ensures continuity, expertise and ongoing investment. Crucially, it places the dataset within a shared national infrastructure rather than within the constraints of a short-term research project. By placing these cases alongside one another, the paper argues that sustainability cannot depend on individual researchers, PIs, short-term funding or temporary hosting within universities. It must instead be supported through deliberate planning, cross-sector partnerships, government departments and through access to stable national infrastructures. The paper concludes by identifying lessons for the future of NIPNP and similar projects, and suggests practical approaches for ensuring that publicly funded digital resources remain accessible, preserved and usable for decades to come.
Frances Kane (2026) studied this question.