It seems like a truism to say that any building should conform to human scale. After all, the main purpose of almost all buildings is to house human activity, so it seems to follow that anyone designing a building should think carefully about the size, proportions, and functionality of the human form. And indeed, even the quickest scrutiny of the history of architecture reveals that many of the most hallowed principles of design spring from this simple truth. Every student of Western architecture must learn something of Vitruvius’ masterwork, De Architectura, which, written around 15 BC, is the earliest surviving treatise describing the most important principles of classical architecture. There, Vitruvius emphasized the importance of proportion and harmony in design, and he argued that good buildings should mirror what he described as the most exquisitely perfect form that was known—that of the human body.
Almost everyone has seen images of Leonardo da Vinci’s Vitruvian Man and, as the title of the illustration suggests, Leonardo was making reference to the same set of ideas: that there is a kind of mathematical perfection to the human body, an almost sacred set of proportions that we find mirrored everywhere in nature.
It isn’t hard to find this relationship imbued in a great deal of architecture, and sometimes very explicitly, as in the Caryatids at the Acropolis in Athens, constructed long before Vitruvius’ time and further evidence of the perceived importance of the proportions of the human form to good architecture.
In the more modern canon, the same emphasis on human proportions and architecture was picked up by what might seem to be an unlikely source. The famed Swiss-French architect Le Corbusier, perhaps all too often associated with some of the more egregious failures of modernist architecture and urban development like Pruitt-Igoe, was also preoccupied with the relationship between the human size and shape and architectural design. In his book Modulor, Le Corbusier built on the earlier arguments of Vitruvius to describe a set of ideal architectural proportions based on the human body and on certain mathematical relationships found in abundance in nature (such as the golden ratio). [Strange side note here—Le Corb’s original plan for Modulor was based on the average height of a French man—5 feet 9 inches—but was later adjusted to a height of six feet because “"in English detective novels, the good-looking men, such as policemen, are always six feet tall!"].
Of all the body parts that might be mirrored, simulated or otherwise represented in architectural forms, one might expect the face to have pride of place. There’s a great deal of evidence that our brains are exquisitely well-tuned to faces and their expressions. Not only are there brain areas that are largely specialized for the recognition of faces and facial expressions, but the operation of these areas seems to proceed with extraordinary speed and dexterity. Probably in part as a consequence of the pre-eminence of facial processing systems in the human brain, most of us are subject to occasional instances of a phenomenon called pareidolia in which we see compelling illusions of faces in non-face objects. Most recently, witness the furor over the evident appearance of a human face in a flock of birds flying over New York. Given our predilection for faces and our tendency to pareidolic fancies, it makes perfect sense to me that one way to elicit an emotional response to a piece of architecture might be to tap into the specialties of a brain tuned in this way to notice and respond to signals from the face. There are some great examples of house faces on the Internet that might help to convince you of the possibilities.
What if there was a tool that we could use to measure objectively the likelihood that a building façade might generate a pareidolic response? What if that tool, designed to find faces in houses, could also measure their expressions? Michael Ostwald and colleagues at the University of Newcastle have begun to develop such a program. Computer software that can recognize faces is nothing new these days. When you upload a photograph to Facebook, for example, you’ll notice that the site is easily able to identify faces in the photograph and to encourage you to tag them. Software that measures facial expression is also becoming increasingly common. Companies geared mostly to the use of facial expression analysis for marketing research are beginning to spring up and many of them can do a good job of describing the expression on an image of a human face in terms of a small set of primary emotions. Ostwald’s team, using software that they developed themselves, tested the idea that we could measure the pariedolic potential of buildings by aiming their cameras at two different archetypical buildings. One of them was Le Corbusier’s Villa Savoye on the outskirts of Paris and the other was Frank Lloyd Wright’s Robie House in Chicago. The two houses, and the architectural philosophies of the architects who designed them, could not be more different.
As you can see from the images, Villa Savoye, with its clean, boxy geometry, absence of ornamentation, and flat roof is the epitome of modernity. Wright’s house, with its pitched roof, and warm, interesting textures looks more inviting and friendly.
So what did the image processing software have to say? Perhaps not surprisingly, given that this was a first attempt, the results were not completely clear-cut but they were tantalizing. For one thing, the software found many face-like images in both houses, but there were strong effects of both time-of-day and season. Thinking about your own experiences with architecture, this might not be surprising. The same house that might appear welcoming on a spring morning might seem forbidding and perhaps even a bit frightening on a cold, winter night. But what about differences between the houses? Here the differences were small but interesting. Both houses read as being angry, but the Wright house was less angry than the Le Corbusier house. The Wright house was also a little bit happier than the Le Corbusier house, but the differences overall were quite small.
I think the real importance of the findings of Ostwald’s team is that they give us a new and empirically based way of thinking about the connections between emotions and architecture. Everyone knows in their bones that the appearance of the places of their lives affects how they feel. Some houses do seem to make us feel sad, even before we set foot inside them, while others seem to call out to us from the curb with warm invitations to approach. What these researchers are suggesting is that these kinds of reflexive responses to building facades may come about in part by co-opting ancient brain circuitry designed to help us manage our emotional and social lives. It’s not hard to draw a direct line between such findings and the thoughts of architectural theorists both ancient and modern who have espoused the importance of a fundamental link between the forms of buildings and the forms of the beings who must inhabit them. Wittingly or not, every time an architect helps to place a new building on the street, they are contributing to the emotional lives of everyone who comes into contact with that building, either for better or for worse.
Follow me on Twitter
Like my Facebook page
Check out my new book