The purpose of this research is to examine the phenomenon of aestheticization in architecture as a reflection of architectural narcissism, where aesthetic appeal, primarily promoted through images, takes precedence over ethical and spatial considerations. The study focuses on the two defined modalities of architectural dualism: architectural concept (staged modality associated with an image) and the space of everyday life (the authentic modality associated with reality), within the broader context of contemporary aestheticization.
The research adopts a multidisciplinary methodology combining theoretical analysis of aestheticization and imagery in architecture, survey data from focus groups and results from a numerical experiment conducted using deep machine learning with a neural network.
The findings highlight the dominance of staged modality (associated with image) over authentic modality (associated with spatial experience) in contemporary architectural discourse. The results of the neural network experiment support the notion that architecture is increasingly perceived and evaluated through visual representation rather than direct spatial experience.
This research acknowledges limitations regarding the use of deep convolutional neural networks for visual classification in architecture, especially their reliance on training data quality and quantity. The complexity of visuals of everyday life challenges both machines and humans, revealing blurred boundaries between staged and authentic spaces. Despite high accuracy, the margin of error suggests the need for larger, more diverse datasets.
The study highlights the potential of digital tools in architectural education and practice, enhancing critical analysis and interpretation of architectural imagery, while also encouraging further exploration of automated methodologies in design research. These digital tools enable efficient organization and analysis of architectural visual data, helping architects understand the relationship between staged and authentic spaces.
This study makes a distinctive contribution by integrating architectural theory with machine learning analysis to critically examine the growing reliance on aesthetics and imagery in contemporary architectural practice. It underscores the importance of reframing architecture as a dynamic, lived experience rather than merely a visual and representational artefact.
