Groep-Foncke, Marianne

Water quality was a hot topic in the early modern city of Haarlem (Holland, United Provinces of the Netherlands). This is hardly surprising, because the city earned its prosperity from two rivaling trades: brewing - which needs pure water to obtain a potable product - and the notoriously polluting cloth industry. In addition, the city had to cope with the impact of a growing population and was prone to flooding by brackish water from a nearby sea-arm. We can consider water of the right quality as a common-pool resource for the inhabitants of Haarlem. The stakeholders had to make arrangements and clashed frequently with each other while doing so. In this paper, we explore the interplay between stakeholders with divergent institutional backgrounds, who each in their own way took responsibility to control water quality: provisional collective action groups of bleachers, the established institutions for collective action of the brewers’ guild (local) and waterboards (supra-local), and the municipal government that represented the communality of inhabitants. How did they address the sustainability challenge at their doorstep, what was at stake for each of them, how did their problems interrelate, and to what extent did they succeed guarding their precious resource?

The use of Large Language Models (LLMs) in tandem with the Institutional Grammar is a promising way to speed up the process of breaking down and categorizing institutional accounts, advancing the analysis of institutional arrangements. However, the application of LLMs comes with challenges of its own, especially when working with data - such as historical datasets - that are divergent from the data that the model was trained on. Applying the open-source Phi-3-mini model, this paper assesses how commoners historically managed environmental resources through self-governance, with a focus on the rules they developed over time to address environmental limitations. By applying grammatical and content-based analysis, we will assess the environmental literacy embedded in these regulations and evaluate their effectiveness in translating that literacy into actionable governance. At the same time, we will highlight the opportunities, pitfalls, and limitations to be reckoned with when applying LLMs to historical data.