The book aims at studying the history of the mathematization computerization of the language sciences, in the wake of the first mathematization of the 1930s. Two stages can be distinguished: machine translation in the 1950s, devised as a cold war technology without linguistics, and computerized corpora in the 1990s with the unprecedented development of computers. The study rests on archives and adopts a comparative approach (USA, ex-USSR, Great Britain and France) in order to determine how the language sciences have integrated the computerization, depending on various linguistic and cultural traditions.