Reading in the Brain by Stanislas Dehaene (2009)
Summary and review of the above book
INTRODUCTION: Dehaene describes how a specialised area of the temporal cortex processes the written word, and then projects to other brain regions that deal with the sound and with the meaning of words. This specialist area of the visual cortex originally evolved for some other function, so modern literate humans may have suffered some offsetting cognitive loss.
The brains of literate humans contain specialised structures for dealing with written words. One problem in understanding how we read is invariance. Whatever the type of font, size of letters or colour of the print, our interpretation of the word does not change. It is suggested that some neurons may have the capacity to identify particular letters whatever their form. In the primary visual cortex, different forms of words and letters may be processed separately, but these are merged in the higher visual cortex.
The so-called ‘letterbox of the brain’ is situated within the ventral visual area in the lateral occipito-temporal sulcus. This is an area through which all information about written words goes, before it is distributed to a variety of other cortical areas. Moreover, this ‘letterbox’ brain region is located in the same area of the brain, whatever the language of the subjects, including East Asian languages, with their visually very different texts. Dehaene on reading in the brain.
This ‘letterbox’ area is stimulated only by written words and not by other visual patterns, and also not by the sound of words. The brain is seen as dividing its visual work into categories, with each one performed by a different patch of the cortex. So the reading area of the cortex has an orientation to only deal with written letters. This area analyses visual inputs for the existence of letters, and then projects to other cortical areas for the extraction of sound and meaning. The ‘letterbox’ area is very specific in being interested in letters and not interested in how they are presented. So it does not give a different response for a word presented in upper of lower cases of in different types or heights of font. Within the ‘letterbox’ area some parts respond to individual letters and some to whole words.
This ‘letterbox’ for recognising written material is located between an area near to the midline of the brain that recognises landscapes and buildings and a lateral area in the fusiform gyrus that recognises faces. Thus the ventral surface of the brain contains areas tuned to recognising particular types of feature. However, the specialisation of such areas overlaps without rigidly demarcated boundaries. Specialisation, in fact, seems to be at the level of neurons rather than brain areas, with neurons of different specialisations intermingled with one another. Many dispersed neurons are seen as working to eventually create meaning in specialised areas of the temporal lobe. Many of the preferred shapes of the inferior temporal neurons respond to shapes that resemble letters, symbols or sometimes Chinese characters, and these are collectively here referred to as proto-letters. These are suggested to have evolved because they helped in identifying objects by being invariant to many changes in images. The inferior temporal cortex is good at detecting correlations between images, and this makes it possible to detect things such as the association between letters regardless of case or font.
The propagation of a written word inside the brain has recently been tracked by researchers. Activation started in the occipital lobe which contains the primary visual cortex, becomes lateralised to the left hemisphere after 170 ms, then after 250 ms there is a sudden burst of activity across parts of the temporal lobe; after 300 ms the activity moves beyond the temporal lobe into areas such as the anterior insula and Broca’s area.
Visual analysis is only the first step in reading. It is followed by extraction of meaning and sound patterns, which are performed simultaneously in different cortical areas. After going through the ‘letterbox’ area, nerve fibres project the processing of this visual area towards other regions of the brain. Visual information is shuttled further along the temporal lobe by a major pathway, the inferior longitudinal bundle. Projections are thought to go from the ‘letterbox’ in the left visual cortex to cortical areas in both hemispheres of the brain.
Separate networks appear to be involved in processing sound and meaning. A region known as the planum temporal within the temporal lobe is involved in matching letters and sounds. The left-middle and ventral regions of the left temporal lobe are thought to be specialised in processing meaning. Regions near the front of the temporal lobe appear to concentrate on the meanings of words that are combined into sentences. The function of such regions are assumed to depend on information stored in memory. It is suggested that these areas act as convergence zones with information from the rest of the brain converging to be processed in these regions.
The lateral temporal region appears to be subdivided into regions dealing with the meaning of particular categories of words, such as for instance faces, animals or vegetables, with these regions in turn drawing on information in particular areas of the brain, such as the parietal for numbers, V4 for colours and V5 for motion and Brodmann area 10 for beliefs.
The cost of literacy?
In evolution, there is a concept called ‘exaptation’, by which something that evolved for one purpose gets taken over for another use; something like this looks to have happened with the ‘letterbox’ area of the brain. Literacy is seen as changing the brain with an increase in left brain engagement, while at the same time there is an increase in activity between the two hemispheres, with a thickening of the corpus callosum. It is suggested that areas used for reading, something that humans have only done in the last few thousand years, was originally evolved to be used for something else, so that the ability to read may be paid for by some offsetting cognitive loss. Thus reading ability could reduce the cortical space available for other activities. Studies of hunt-gatherers suggest that we may have lost something in terms of our ability to read signs in the natural world.