About the Wellespring Project
The Wellespring Project is an experimental digital-humanities initiative that explores the lyrical, cultural, and conceptual worlds of singer-songwriter Jesse Welles. Combining methods from data science, digital humanities, and cultural studies, the project seeks to uncover and visualize the relationships between lyrics, themes, people, places, historical references, and emotional patterns in Welles' music.
At its heart, Wellespring is a study of creativity as a network: of how songs connect to each other, to their cultural moments, and to the ideas and emotions that flow through them. It is both a research project and a teaching laboratory, designed to introduce students and collaborators to the powerful possibilities of knowledge-graph thinking, semantic modeling, and human-driven data interpretation. Our work will utilize offline non-LLM natural language processing (NLP) tools that have been around long before online LLM AI platforms but NO Jesse Welles material will be processed by LLMs. Alongside our analysis of Welles and his reception network, we will critically engage the rising generative LLM AI movement by examining ethical concerns and the ways in which it fails and/or succeeds at similar processes on non-Welles datasets already within these platforms; this will be in order to inform a critical outlook on the contemporary push for LLM expansion, alongside deep dives into serious ethical concerns around generative AI and its theft and abuse of artists' creative outputs.
What We're Building
The project centers on a Neo4j graph database, in which each lyric line, song, theme, concept, and cultural reference becomes a node connected by meaningful relationships, such as inspired by, mentions, contrasts with, or develops into. Using locally-hosted Jupyter notebooks and offline natural-language processing tools, we will process select data alongside human analysis of patterns across Jesse Welles' body of work, tracing motifs of identity, struggle, love, and transcendence.
The resulting data model will serve as the foundation for:
- Interactive visualizations that map the interconnected world of Welles' songs.
- Analytical tools and documentation that guide others in exploring or extending the model.
- Pedagogical resources for students learning digital analysis, data ethics, and humanistic interpretation.
Why "Wellespring"?
The name reflects both a source and a flow: the idea that creativity, like water, springs from deep wells of experience and emotion. Jesse Welles' music draws from these wells, blending personal reflection with broader social and philosophical commentary. Our goal is to make visible those underlying patterns and relationships. To turn songs into a landscape of meanings that listeners can explore.
Who We Are
The Wellespring Project was founded by Dr. Darrell J. Rohl, Associate Professor of Archaeology, History, and Digital Humanities at Calvin University. It brings together students, scholars, and technologists interested in how modern digital tools and data-driven methods can illuminate artistic and cultural expression without breaching ethical boundaries as has been done in the development of LLMs and derivative generative AI. The project's initial collaborators include participants in Calvin's Fall 2025 Digital Humanities course, but I hope that it will soon expand to include external partners in data visualization and cultural analytics.
Our Aims
- To model and visualize connections between lyrical ideas, cultural influences, and human emotions.
- To document and share open, reproducible methods for graph-based humanities research.
- To engage both academic and public audiences in thinking about music, meaning, and data.
- To critically explore the limitations and ethical concerns of LLMs and other generative AI tools.
- To serve as a living digital laboratory where students can contribute, learn, and innovate.
Explore Further
This site serves as both a public showcase and a technical documentation hub. Here, visitors can:
- Learn about the project's data model and analytical tools
- Explore interactive graph visualizations of lyrical networks
- Follow research updates, blog posts, and student insights
- Access documentation and tutorials for working with Neo4j and Colab
- Critically engage the much-touted "promise" of generative AI tools by exploring critiques and ethical concerns about their use and abuse of artists and their creative outputs.
Whether you are a digital humanist, a computer scientist, or a fan (or critic!) of Jesse Welles, you are invited to dive in, explore the data, and discover the many stories hidden within the songs.