The mouse and the rectangle
I have always been fascinated by pavlovian conditioning because it shows how little control animals really have over their own behavior. The most famous example of this is the experiment Ivan Pavlov performed on his dogs in the 1890's. Dogs normally salivate at the smell and sight of meat, but Pavlov noticed his dogs began to salivate when they noticed the lab technician who fed them would enter the room. These “psychic salivations” didn’t make sense to him.
Eventually, he performed a beautifully simple experiment to describe the behavior. He started to ring a bell every time his dogs were fed. After a few days, the dogs began salivating at the sound of his bell alone. He showed that a physical connection can be created between a stimulus (the bell sound) and the following behavior (the dog's salivation) with no cognitive interaction.
Almost a hundred years later, the neurologist V.S. Ramachandran performed an experiment on mice that took Pavlov's research further. He took a mouse and placed two shapes in front of it: a perfect square and a rectangle. The mouse would randomly choose one of the shapes to move toward. When it moved toward the rectangle, it was rewarded. After a while, obviously, the mouse began to prefer the rectangle. It associated the positive reward with the rectangle and then began to prefer it over the square. These results are all very simple.
Left: Ramachandran suggests the artist who made this statue of Parvati used peak shift to exaggerate the feminine aspects of the woman to make it more aesthetically pleasing. In a way, the more rectangular rectangle may have become more aesthetically pleasing to the mouse.
But when the mouse was presented with a third option, a more elongated and more rectangular rectangle, it preferred this shape over the normal rectangle. At first this behavior might seem illogical. Why would the mouse prefer another shape when it knows there will be a reward if it goes toward the normal rectangle? The answer is that the mouse did not associate that particular rectangle with the reward. It associated the "rectangularity" with the reward. When the mouse was presented with an exaggerated, more rectangular rectangle, its reaction was likewise exaggerated.
This behavior is called peak shift, and it has vast implications for the understanding of how the human brain works. •