In the second half of the 19th century, the French architect Viollet-le-Duc published his Dictionnaire raisonné de l’Architecture Française. In this work he proposes, for the first time, a rational theory about gothic structures based on the existence of “active” and “passive” elements: vaults are made of passive webs supported by active ribs; pinnacles contribute, in an active way, to the stability of the buttresses; flying buttresses transfer the load from the vault, etc. Until 1900 Viollet-le-Duc’s approach to the structural behaviour of gothic constructions was a dogma unanimously accepted by most architects, archaeologists and historians, but throughout the following two decades it became the object of harsh criticism. Discussion about the way gothic structures worked reached its climax at the end of the 1920’s, when the engineer Victor Sabouret published his first article against Viollet-le-Duc’s rationalist ideas, entitled Les voûtes d'arêtes nervurées. Rôle simplement décoratif des nervures. He specifically focused his discourse on the behaviour of vaults: on the decorative as opposed to structural or constructive function of the ribs. Although his arguments are mistaken and inaccurate in some aspects, he had an enormous influence over subsequent generations, and a large number of studies in the field were published throughout the years following this publication. Authors of such studies include Marcel Aubert, Henri Focillon and Henri Masson, whose publications revealed their disagreement with both theories, as well as Pol Abraham, who was in complete opposition to Viollet-le-Duc. In this paper, the Limit Analysis of the modern theory of masonry structures, formulated by J. Heyman in the 1960’s, is used to evaluate the accuracy and suitability of Sabouret’s criticisms to the rationalist theories.