In recent years, the field of historical semantics has experienced notable advancements. Despite the emergence of promising methodologies, a standardized procedure has not yet been established in the research of the history of concepts. A significant methodological challenge is the seamless and transparent integration of analyzing large datasets (distant reading) with the traditional workflows of conceptual historical research (close reading). Additionally, the effective capture of various word senses (polysemy) and tracking their change over time using computational methods remains a complex task. This article presents the tool Sense Clustering over Time (SCoT), specifically designed to address these challenges in conceptual historical research. SCoT employs the method of Word Sense Induction (WSI) to facilitate the semi-automatic detection and visual representation of the historical semantics of conceptual words into sense clusters. Through an open-access web interface, users can: (1) analyze the diachronic development of sense clusters within extensive text corpora, (2) explore the linguistic contexts driving these changes, and (3) identify and compile relevant references in the text corpus for further study. The article explores the hermeneutic processes involved in the analysis of large historical text corpora utilizing a WSI approach and discusses epistemic challenges associated with using SCoT as a research tool in the digital history of concepts.