Coufignall

Louis Couffignal

Louis Pierre Couffignal (1902-1966), a French Mathematician and Cybernetics pioneer, was born in Monflanquin, Fâcheries, on March 16th, 1902, in a humble family: his father Guillaume Couffignal (then 44 years old), was a chief cantonnier, his mother, Marie Deroux-Couffignal (31 years old), was a tailor.
After attending a primary school at Monflanquin, in 1913 Louis enrolled in the Collège et au Lycée de Villeneuve sur Lot, where he received a bachelor degree in Latin, sciences and mathematics. After graduation he remained in the college as a part-time lecturer, later on continued his work as a lecturer in other colleges. Couffignal was almost 20 years director of Laboratoire de Calcul Mecanique at the Institut Blaise Pascal. Couffignal died on 4th of July, 1966.

Couffignal was a pupil of the prominent French mathematician Maurice d'Ocagne (1862-1938), who instilled him with his own passion for calculating machines.

Couffignal published several articles and sent several notes to the French Academy of Science regarding calculating machines. The first note was in 1930, and is devoted to a new calculating machine. In 1932, he organized in the Collège de France a conference on the “calculating machines, their principle and their future”. In 1933 he published a monograph for calculating machines. In 1936 appears his decisive note on “the use of the binary notation in the calculating machine”. In 1938, Couffignal published an article, important for cybernetics, where it defines the machines as “a whole of inanimate or partially animated beings or even exceptionally animated able to replace the man” and further: “Since the machine is made for the man, it is field of the mechanical analysis to acquire an overall picture of the various activities where the man was or could be replaced by the machine, and to establish laws of substitutions”. In the same 1938, he becomes Ph. D. thanks to his thesis “The mechanical analysis, application to the calculating machines and the celestial mechanics”, which poses the principles of the electromechanical binary all-purpose computer. In his thesis he described two calculating machines: a decimal calculator, and a binary electromechanical program-controlled calculator. There is also another frenchman, who devised binary-based calculating machine even before Couffignal—Raymond Valtat (Valtat filed a patent in 1936 for a mechanical calculator founded on the conversion of decimal input into binary prior to calculation).

Couffignal apparently had every intention of building this machine in association with the French computers manufacturer Logabax, but presumably because of the WWII never did so. After the war however, he designed and Logabax manufactured in 1952 the first French electronic digital computer (it contained 2000 tubes). Bellow is a diagram of the machine (from the translation of Couffignal's thesis in the book of Brian Randell The Origins of Digital Computers, Birkhäuser, 1982):

Coufignall scheme

Look what wrote in conclusion of the description of his machine Couffignal:
Without entering into very great detail, we believe we have shown that it is possible to construct a calculating machine, able, without any intervention by an operator, to execute a sequence of calculations, to store the intermediate results, to read mechanically a function table and to print all the numbers, recorded in its registers; we think that in view of its capacity the machine may be considered to be of great simplicity.
It is worth noting, that the connections between different parts of the machine are all electrical; the arrangement of the latter does not clash with any of the geometric or kinematics constrains which one meets with in a purely mechanical device, where all movements are caused or guided by physical contact: the number of each different type of component is therefore theoretically unlimited and in practice very high.