Oxford Seminar in the History of Alchemy and Chemistry
Chair: Rob Iliffe (Oxford)
Vojtěch Kaše (University of West Bohemia, Plzeň)
and
Sarah Lang (Max Planck Institute, Berlin)
Tracing the Histories of Early Modern Conceptual Ecosystems: Remote Sensing Methods for the Archaeology of Alchemical Knowledge
Early modern alchemy formed a dense “conceptual ecosystem”, intervening precise technical vocabulary with rich polysemous terms. In this paper, we propose a “remote-sensing” computational methodology to conduct an archaeology-like inquiry of the knowledge embedded in these concepts. Drawing on recent advances in methods for word sense disambiguation and semantic change detection, we analyze large-scale corpus of Latin texts by means of the so-called contextual word embeddings, derived from pretrained deep learning models. We have to be aware that not only words change meanings over time (semasiology), but also the concepts behind them tend to evolve, requiring novel lexical realizations (onomasiology). Our methodological tools allow researchers to navigate efficiently between these two levels of analysis at scale. To demonstrate the feasibility of our approach, we turn to Martin Ruland the Younger’s Lexicon Alchemiae (1612), and offer a distributional semantic analysis of the history of usage of the terms included in this dictionary across a vast corpus of older Latin literature.
Guillermo Restrepo (Max Planck Institute, Leipzig)
Computational History of Chemistry: How Big Data Illuminates Macrohistorical Trends and Microhistorical Events
The computational history of chemistry is grounded in the rich digital archive of chemical practice—spanning its material, social, and semiotic dimensions. From 1800 to the present, this data reveals a remarkably stable exponential expansion of the chemical space: the number of new chemicals doubles approximately every 16 years, a trend resilient even through global conflicts such as the world wars.
At the macrolevel, we identify three distinct regimes in discovery, each characterised by progressively lower variability—indicating a long-term regularisation of chemical innovation. These regimes are separated by two pivotal transitions: one around 1860, associated with the rise of molecular structural theory, and another in the 1980s, linked to a major technological shift enabling the synthesis of increasingly complex compounds.
At the microlevel, we uncover the fixed-substrate approach: chemists routinely combine no more than two substances to generate new chemicals, with one of them typically being a well-established compound. This pattern reflects a systematic, incremental strategy underlying discovery. From a more recent perspective, our analysis of the geopolitics of chemical discovery reveals that China’s dominant role—evident from the early 2000s—extends beyond production to knowledge generation. This is particularly clear in rare-earth chemistry, where material access and scientific innovation are deeply intertwined.
In this sense, the computational history of chemistry allows us to perceive the discipline’s historical pulse: a vast, steady pulse shaped by the daily contingencies of millions of practitioners working across diverse social, political, and economic contexts. It enables us to move from this broad, macroscopic pulse to the fine-grained details of individual chemists, their choices, collaborations, and the material conditions of their work.
Convenors: Ellen Hausner (Oxford), Sergei Zotov (Warburg), and Jo Hedesan (Oxford)
