Architecture that respects the statistics of natural scenes is more comfortable for people to look at and there is scientific evidence to support this statement.
Our visual systems evolved to look at scenes from nature. Natural scenes have a characteristic structure called scale invariance. Imagine a map of the coast. The number of wiggles in the outline of the coast remains the same whatever the scale of the map you are using. If you double the size of the image, it has the same amount of detail. Double it again, and again the detail is the same. That’s scale invariance. The brain has evolved to process images with scale invariance and it can do so very efficiently using only a small number of neurons at any one time. The brain has more difficulty processing stripes (which are such a common feature of modern architecture) because they are not scale invariant. The brain becomes less efficient, and it uses more oxygen. We can measure the oxygen usage with neuroimaging and we find that the patterns of stripes that people find uncomfortable to look at use more oxygen.
Some people, those who suffer migraine, find the patterns particularly unpleasant, and these people have an abnormally large oxygenation of the brain when they look at them. In fact, the larger the number of symptoms of visual discomfort the greater the oxygenation of the brain.
So both in terms of the differences between images and the differences between people, images that are unpleasant induce a large oxygenation of the visual cortex of the brain. We can perhaps think of the discomfort as protective or homeostatic, in much the same way as most pain is – it acts to protect the organism. Perhaps the discomfort protects us from too great an oxygen usage. After all, the brain consumes 15-20% of the energy used by the body, so we need to reduce the energy consumption when we can.
Now we said that stripes are unnatural, and uncomfortable and cannot be processed efficiently by the brain; they use more energy. Stripes are everywhere in the modern urban environment.
It turns out we can use a computer to predict how uncomfortable images are to look at just by measuring mathematically how much the image departs from scale invariance. We can do this for a wide variety of images. The algorithm is described in a recent paper by Olivier Penacchio and myself and will shortly appear in the journal Vision Research.
We examined the achitectural design of the exterior of residential buildings and demonstrated mathematically that over the last century design has become progressively more and more striped.
Why are stripes such a common feature of design?
Designers like designs that are striking on the eye, perhaps this is one reason why.
Perhaps it is also because it is simpler and cheaper to build things using repetitive elements. But we need to think again about stripes, and avoid them when we can, making our urban environment visually more like that in nature. This will make design more comfortable on the eye for neurological reasons that we are just now beginning to discover.