Mei Yang¹ (), Claudia Damiano¹, Paul Gauvreau², Dirk B. Walther¹; ¹Department of Psychology, University of Toronto, ²Department of Civil Engineering, University of Toronto

People readily recognize London when they see Tower Bridge, San Francisco by the Golden Gate Bridge, and Sydney when they see the Sydney Harbour Bridge. These bridges have emerged as iconic landmarks that shape their city’s skyline by virtue of their aesthetic qualities. Bridges can also elicit a neutral aesthetic response and, sometimes, can be regarded as downright ugly. Bridges are designed as public infrastructure, which often shape their surroundings for centuries. Nonetheless, little is known about what shapes the aesthetic appeal of bridges. Here we explore how aesthetic judgements of bridges relate to engineering and design features. Our dataset comprises 318 images of 118 bridges from around the world, rated by 200 participants for aesthetic pleasure, interest, complexity, and safety. A civil engineering team annotated each image for type, depth, material, apparent age, and aesthetic premium. Using Factorial Analysis of Mixed Data (FAMD), we found two significant dimensions. The first “aesthetics” dimension shows strong correlations among aesthetic, complexity, and interest ratings and is connected to bridge type. The second “safety” dimension relates subjective ratings of safety to bridge age and material. Analysis of visual features of bridges, using the Mid-Level Vision (MLV) Toolbox, shows that contour length is a predictor of both bridge type and the aesthetic, complexity, and interest ratings. For example, truss bridges, made up of several interconnected beams, are represented by many short contours and are generally rated as more complex, interesting, and pleasing. On the other hand, the visually simple slab and girder bridges are often represented by a few long contours and are rated as uninteresting and not aesthetically pleasing. Our study offers the first attempt to systematically collect and analyze subjective ratings of bridge aesthetics, paving the way for empirically supported decisions for the design of bridges and, potentially, other public infrastructure projects.

Acknowledgements: This work was supported by NSERC Discovery Grant (RGPIN-2020-05778) to PG and an NSERC Discovery Grant (RGPIN-2020-04097) as well as a SSHRC Insight Grant (435-2023-0015) to DBW.