9.4: Neuropsychological approach

Investigating the brain - a way to resolve the imagery debate?

Visual imagery was investigated by psychological studies relying solely on behavioural experiments until the late 1980s. By that time, research on the brain by electrophysiological measurements such as the event-related potential (ERP) and brain-imaging techniques (fMRI, PET) became possible. It was therefore hoped that neurological evidence how the brain responds to visual imagery would help to resolve the imagery debate.

We will see that many results from neuroscience support the theory that imagery and perception are closely connected and share the same physiological mechanisms. Nevertheless the contradictory phenomena of double dissociations between imagery and perception shows that the overlap is not perfect. A theory that tries to take into account all the neuropsychological results and gives an explanation for the dissociations will therefore be presented in the end of this section.

Support for shared physiological mechanisms of imagery and perception

Brain imaging experiments in the 1990s confirmed the results which previous electrophysiological measurements had already made. Therein brain activity of participants was measured, using either PET or fMRI, both when they were creating visual images and when they were not creating images. These experiments showed that imagery creates activity in the striate cortex which is, being the primary visual receiving area, also active during visual perception. Figure 8 (not included yet due to copyright issues) shows how activity in the striate cortex increased both when a person perceived an object (“stimulus on”) and when the person created a visual image of it (“imagined stimulus”). Although the striate cortex has not become activated by imagery in all brain-imaging studies, most results indicate that it is activated when participants are asked to create detailed images.

Another approach to understand imagery has been made by studies of people with brain damage in order to determine if both imagery and perception are affected in the same way. Often, patients with perceptual problems also have problems in creating images like in the case of people having both lost the ability to see colour and to create colours through imagery. Another example is that of a patient with unilateral neglect, which is due to damage to the parietal lobes and causes that the patient ignores objects in one half of his visual field. By asking the patient to imagine himself standing at a place that is familiar to him and to describe the things he is seeing, it was found out that he did not only neglect the left side of his perceptions but also the left side of his mental images, as he could only name objects that were on the right hand side of his mental image.

The idea that mental imagery and perception share physiological mechanisms is thus supported by both brain imaging experiments with normal participants and effects of brain damage like in patients with unilateral neglect. However, also contradictory results have been observed, indicating that the underlying mechanisms of perception and imagery cannot be identical.

Double dissociation between imagery and perception

A double dissociation exists when a single dissociation (one function is present another is absent) can be demonstrated in one person and the complementary type of single dissociation can be demonstrated in another person. Regarding imagery and perception a double dissociation has been observed as there are both patients with normal perception but impaired imagery and patients with impaired perception but normal imagery. Accordingly, one patient with damage to his occipital and parietal lobes was able to recognise objects and draw accurate pictures of objects placed before him, but was unable to draw pictures from memory, which requires imagery. Contrary, another patient suffering from visual agnosia was unable to identify pictures of objects even though he could recognise parts of them. For example, he did not recognise a picture of an asparagus but labelled it as “rose twig with thorns”. On the other hand, he was able to draw very detailed pictures from memory which is a task depending on imagery.

As double dissociation usually suggests that two functions rely on different brain regions or physiological mechanisms, the described examples imply that imagery and perception do not share exactly the same physiological mechanisms. This of course conflicts with the evidence from brain imaging measurements and other cases of patients with brain damage mentioned above that showed a close connection between imagery and perception.

Interpretation of the neuropsychological results

A possible explanation for the paradox that on the one hand there is great evidence for parallels between perception and imagery but on the other hand the observed double dissociation conflicts with these results goes as follows. Mechanisms of imagery and perception overlap only partially so that the mechanisms responsible for imagery are mainly located in higher visual centres and the mechanisms underlying perception are located at both lower and higher centres (Figure 9, not included yet due to copyright issues). Accordingly, perception is thought to constitute a bottom-up-process that starts with an image in the retina and involves processing in the retina, the Lateral Geniculate Nucleus, the striate cortex and higher cortical areas. In contrast, imagery is said to start as a top-down process, as its activity is generated in higher visual centres without any actual stimulus, that is without an image on the retina. This theory provides explanations for both the patient with impaired perception but normal imagery and the patient with normal perception but impaired imagery. In the first case, the patient’s perceptual problems could be explained by damage to early processing in the cortex and his ability to still create images by the intactness of higher areas of the brain. Similarly, in the latter case, the patient's impaired imagery could be caused by damage to higher-level areas whereas the lower centres could still be intact. Even though this explanation fits several cases it does not fit all cases. Consequently, further research has to accomplish the task of developing an explanation that is able to explain the relation between perception and imagery sufficiently.