J. L. Lagrange's changing approach to the foundations of the calculus of variations (Q1059621)

The name calculus of variations was coined by Euler in response to Lagrange's early researches. While there have been several surveys of Lagrange's main results in this calculus, this paper provides a detailed study of his approach to the foundations of the calculus of variations, in which his changing derivation of the Euler-Lagrange equations is a central element. The development of Lagrange's approach to the calculus of variations falls into two distinct periods: 1. From his first letters to Euler in the mid-1750s to the publication of his Méchanique analitique in 1788, 2. the time of the appearance of his didactic works on differential, integral and variational calculus at the end of the eighteenth and the beginning of the nineteenth century. In the author's view, the passage from the first to the second stage may be characterized as a transition from a context of discovery to a context of justification, reflected in the two types of audience; 1. small groups of geometers in the forefront of research, 2. larger groups of students eager to learn from an eminent practitioner. The first stage opens in August 1755 when Lagrange sent to Euler an algorithm for solving the problems of Euler's treatise Methodus inveniendi without any appeal to geometry. In the early correspondence a change of approach from a non- parametric to a parametric presentation of the variational method may be traced. The parametric method was published in a paper presented to the Turin Academy in 1760, while a second paper presented in the same year contained an application of the variational techniques to the principle of least action in dynamics. Lagrange published his new foundations in two treatises, Théorie des fonctions analytiques and Leçons sur le calcul des fonctions. In these he used a theory of Taylor-series to provide a basis for the calculus avoiding infinitesimals. A sense for Lagrange's approach, the author claims, as well as an understanding of the immediate background of his work on integrability and calculus of variations, may be obtained from a description of the notation introduced in these treatises. In the Leçons he separates the investigation of integrability from his treatment of the calculus of variations, expands the presentation of each and discusses the connection between them. The author concludes with a critical evaluation of the work of Lagrange on these subjects.