Séminaire – Histoire des sciences, histoire du texte

Diagrammatic dimensions of textual elements

• Edgar Lejeune (Centre Alexandre Koyré, EHESS)
  What is the difference between plotting a graph by hand and generating it automatically ? Epistemological consequences of two approaches of vizualisations in the humanities (France, 1960-1980)
  Résumé :
  From the early 1960s onwards, two competing models of data analysis were developed in France for the humanities and social sciences. On the one hand, mathematician Jean-Paul Benzécri (1932-2019) and his team worked on a series of computer-assisted mathematical procedures, such as the famous correspondence analyses. On the other hand, cartographer and semiologist Jacques Bertin (1918-2010) developed with his colleagues data analysis methods that relied on the manual manipulation of cardboard files or reorderable matrices made of plastic dominoes.
  Behind these different material and epistemological approaches are two opposing visions of how to produce visualizations for data analysis. In Benzécri’s approach, visualizations are automatically generated by computer. In doing so, the structure of the dataset ’appears’ to the researcher in the form of a printer output. The researcher must then decipher a visualization derived from a mathematical procedure. In contrast, the methods developed by Bertin require no mathematical knowledge. Instead, the researcher handles physical artifacts representing the data, with the aim of revealing the structure of the dataset step by step, guided by his visual perception.
  How do these two regimes of visualization production affect the epistemology of researchers ? What are the theoretical discourses associated with these two approaches ? And to what extent a comparison between them shed light on the epistemological consequences of automatic generation of scientific visualizations ? We will address these questions on the basis of a few case studies and an extensive theoretical literature produced by Jacques Bertin and his colleagues.
• Scott Trigg (SYRTE, Observatoire de Paris)
  Manuscript Diagrams as Tools of Reasoning in Islamicate Astronomy
  Résumé
  The tradition of Islamicate astronomy known as ʿilm al-hayʿa, or “the science of the configuration [of the orbs],” aimed to produce a cosmology based on uniformly-rotating physical orbs, while at the same time preserving the predictive accuracy of models inherited from Ptolemy’s Almagest. Indeed, many works in this tradition identified “doubts,” and unresolved “problems” that arose when considering the Almagest models in terms of physical bodies, and offered innovative solutions. The manuscript sources contain many sophisticated diagrams, but until now little attention has been paid to analyzing the processes through which these diagrams served as tools of reasoning. In this paper I explore the intersection between the materiality of diagrams, their epistemic meanings, and the mental and physical actions required on the part of the viewer in order for a diagram to be “read” correctly. In particular, I argue for the emergence of new types of diagrams that adapted and transformed the “visual vocabulary” and conventions of the existing tradition of mathematical astronomy in order to support a different type of geometric reasoning.
• Wang Xiaofei (IHNS, Beijing, & SPHERE)
  How the notes from Fourier’s course at the Ecole Polytechnique were produced
  Résumé :
  The manuscripts Ms. 1852 and Ms.2044 are two individuals’ notes deriving from Joseph Fourier’s course of analysis at the Ecole Polytechnique. Each manuscript contains the notes of the lectures Fourier gave to different classes of students at the school during the year 1796. This talk will focus on the process of the production of these notes. Through an analysis of the textual features of the two manuscripts, it aims to clarify their factual relationship with Fourier’s teaching.

Lieu : Salle Rothko 412B bâtiment Condorcet