The arrival of digital technology in the field of architectural practice, and particularly in the study and documentation of the built environment, has led to major upheavals over the last three decades. Through a historical perspective and an extended definition of architectural surveying, this research aims to discuss and compare the cognitive dimension inherent in its practice with the technological challenges it faces. On this basis, one observation can be made: the digital prints produced by the application of the most recent acquisition methods (photogrammetry & lasergrammetry) express a high degree of completeness (visual and metric) with the artefacts surveyed, but also produce large volumes of data. This ‘information overload’ does little or nothing to reinforce architectural representation in its role as a vehicle of knowledge. Faced with this lack of intelligibility, the aim of this thesis is to provide an epistemological response that allows us to consider this need for intelligibility without turning our backs on the richness of the digitisations produced. We propose here an original low-level (non-interpretative) approach, in which the meaning of the elements (‘semantic’ and ‘geometric’ structure) is derived from the study of morphological similarities observed within a corpus of shapes. Statistical analysis and the use of various morphological descriptors have made it possible to formalise signatures that are characteristic of the corpus studied; these are no more and no less than the formal expression of a given number of geometric attributes. At a deeper level, our approach provides a methodological solution for processing massive data. By relying on the accumulation of data, this work offers a complementary analytical response to ‘traditional’ high-level (interpretative) approaches where the characterisation of shape is based on pre-structured knowledge of the domain. While one of the challenges is to apply geometric analysis methods to massive data, the other is to compare high-level and low-level observations through a comparative study. Through this research, our aim is to refine our understanding of stylistic propagation in time and space. In order to test this approach, 31 columns from the cloister of the Abbey of Saint-Michel-de-Cuxa were subjected to in-depth morphological analysis. Key words: Architectural representation, survey, digital, architectural heritage, shape analysis, semantic description.
References
Lo Buglio, David. 2016. Caractérisation de formes architecturales. Une approche expérimentale intégrant complexité et intelligibilité des représentations numériques. [Thèse de doctorat]. Bruxelles : Université libre de Bruxelles.
De Luca, Livio. 2006. Relevé et multi-représentations du patrimoine architectural : définition d’une approche de reconstruction 3D d’édifices. [Thèse de doctorat]. Paris : HESAM Université. https://www.theses.fr/2006ENAM0009.