According to Gibson, a system or an occurrence can be studied efficiently when they fail to function properly. For example, if an owner tries to fix a car themselves, they can identify the different components of the car and study their functions. The owner can work out which component caused the car to break down. On the other hand, it is difficult to examine something if it is in perfect working order. Similarly, fMRIs are used to identify different parts of the brain. If a sighted person opens their eyes, their occipital lobe lights up in an fMRI. However, the occipital lobe does not light up in the fMRI of a blind person.

On the contrary, studying something which is completely “coming apart” can lead to misunderstanding. Let us use the example of the car again. If a car was completely stripped of its parts, the owner would not be able to identify the importance and function of each part. Similarly, if the fMRI of a dead person was studied, no parts of the brain would light up. Therefore, the doctors would not be able to study what each part does.

In order to examine something efficiently, some parts of it must work and other parts must not. Without any background knowledge of the functions of different parts of a system, it would be near impossible to work out what each part does.