Why do we like some images more than others?

We often know what art we like, but can't describe in words why. 

There are groups of scientists who try to understand our aesthetic preferences. Their literature provide two types of information: trends and interpretations. Trends describe with tests and graphs things we often already know, like that yellow colours make us feel warm. Interpretations attepmt to provide reasoning for the trends - we relate yellow colours to the sun, which provides warmth.

Two talks I've listened to recently on Youtube are by a psychologist and neurologist who research how people respond to works of art. I've summarized - or in some cases taken a sound-bites of -  their presentations as a list of concepts:

Stephen Palmer (UC Berkeley)

• Find his lab's publications here.

• Composition can be split into spatial and colour composition.

• Scientific knowledge is still catching up to people's aesthetic sense. General rules of composition work well, even if we don't fully understand why.

• Spatial - Front facing objects are preferred by the viewer to be in the center of an image (center bias). Objects facing right are preferred to exist on the left of the image and vise versa (inward bias). Implied motion overrides the direction in which the object faces.

• The mind generally likes objects to be at their expected vertical elevation, e.g. flying birds above the center of the frame (inward bias), canonical (naturally familiar) perspective and at a canonical size (small butterflies and large elephants). 

• Colour - There are hues and saturations of colour we naturally find attractive or repulsive. This is related to objects that we strongly associate with that colour.

Vilayanur Ramachandran (UC San Diego)

• Find his group's publications here.

• The brain is the interface of art and science.

• Art is not about realism, but exaggerations and distortions that provoke a response. 

• A picture of reality is not stimulating. Something more is needed to produce art.

• We appreciate an amplification of a subject's unique or defining features. This is the concept of 'peak shift'. Caution here - over-amplification results in caricature, which is rejected as beyond realistic by the viewer.

• Critical information should be isolated - what is it about the subject that defines it? Line? Texture? Clutter should be removed.

• Abstraction gives the viewer a pleasing problem to solve - what is the subject?

• Metaphor and symbolism add layers of meaning that keep the viewer engaged.