Bibliometrics seeks to measure epistemic change—how scientific knowledge evolves—by analyzing publication patterns and citations. While established methods identify emerging topics and disruptive papers, current approaches face conceptual and methodological constraints. Many recent indicators like Innovativeness or Disruptiveness attempt to quantify epistemic change from individual papers, but this approach implicitly assumes change is a discrete event rather than a process, that change is fully captured in a single paper, and that generic measures can describe all types of change.
The field also depends fundamentally on bibliographic metadata, which may not fully reflect a paper's actual content, raising questions about whether bibliometric signals validly capture epistemic change. This special collection brings together papers from workshops addressing these challenges: how to conceptualize and measure epistemic change, how to establish validation benchmarks, what metadata signals epistemic change, how well existing indicators perform, and whether large language models analyzing full texts can better identify epistemic shifts in publications.