Manfred's Column, November 2005: Post and pop-out: About the neuroplasticity of cortical maps in postal workers


November 2005

Post and pop-out: About the neuroplasticity of cortical maps in postal workers

Maps in the cortex
The cortex of the human brain is organised into map-like areas and each “map” represents certain functions. For example, a part of the brain known as the somatosensory cortex is like a body map for sensory sensations. Touching the skin of the thumb activates a group of neurons in the brain and, within this map for sensory representations, these represent the sensation of touch on the thumb. If one now touches the skin of the index finger, a group of neurons in the cortex is activated and these are close to the group of neurons activated by the touch of the thumb. If the skin of the middle finger is then touched, a group of neurons in the cortex is activated and these are close to the group of neurons activated by the touch of the index finger. Thus the touching of body parts that are close to each other activates brain areas that are close to each other, and the touching of body parts is represented by the activation of a map-like pattern of groups of neurons in the brain.

As indicated above, this part of the brain representing touch is called the somatosensory cortex. The body map for somatosensory perceptions is often illustrated by a caricature termed the “homunculus”, which literally means “little man”. It shows a man with face and hands grossly enlarged compared to the torso and proximal limbs. This caricature therefore represents the neurons in the somatosensory cortex that are activated by touching of certain body parts. Since the hands or the lips are extraordinarily important for humans, they require more brain circuitry to govern them, and consequently take up more space in the somatosensory map.

It has been generally known for more than half a century that the cortex is able to form such map-like representations. As was demonstrated in the example given above, such a map-like representation derives from the fact that, a signal (e.g. touching the thumb) activates a neuron or a group of neurons in the brain (e.g. in the somatosensory cortex). The activation of this group of neurons thus represents the signal from outside and is thus called a “cortical representation” of this signal (e.g. of the touching of the thumb). Such cortical representations have a map-like form, meaning they are ordered in a certain manner: frequent input signals cover more space, i.e. activate more neurons, than rare input signals and the representations of similar signals are closer to each other.

These two principles of order – frequency and similarity – are of a very general nature. As computer simulations show, they can be put down to the principles of neuronal connection, which are implemented in the whole neocortex (therefore also called the isocortex and making up about 96% of the whole cortex in humans) 1. This is why it is proposed that, besides the brain areas, which are known to be map-like in structure (especially the primary and secondary sensory areas), map-like brain areas also exist for higher cognitive functions, such as object recognition.

New empirical data from functional imaging also supports this notion. The existence of maps of higher cognitive functions are difficult to prove, because their input signals do not stem directly from the outside world, but rather from other lower level cortical areas. Thus the principles for ordering higher function map-like structures are difficult to determine. However, there have been some findings that hint at such representations.

If representations of higher cognitive functions are also stored in a map-like form, these maps must also depend on experience and thus might change through experience. Generally, the dependency on experience of neuronal representations is called neuroplasticity.

Cognitive maps
Recently, Polk and Farah 2 provided a very nice example of the flexibility of cognitive maps. Investigations with functional imagining show inter alia that our ability to deal with digits and our ability to deal with letters are represented in different cortical areas. This makes sense, as we are dealing only with letters when we are reading and only with digits when we are calculating and it is rare that we see and process letters and digits in a totally disordered way. Taking into consideration this separation of input (either letters or digits), one would conclude that the neuronal representations of letters and digits are also separated.

This conclusion, however, may not hold true for everyone. Someone who regularly processes letters and digits simultaneously and in a disordered way has, one would assume, a representation of both letters and digits in the same map instead of in two different maps. In order to investigate this assumption, Polk and Farah investigated Canadian postal workers.

In Canada, the zip code consists of digits and letters, e.g. M5T 2S8. Thus Canadian postal workers, who are sorting mail every day, process letters and digits simultaneously for many hours a day. It could then be surmised that these mail sorters do not have different cortical representations for letters and digits, but a more or less homogenous map for letters and digits.

In order to prove this hypothesis, Polk and Farah used the perceptual “pop-out” effect. This effect refers to the fact that for most people, a letter is easier to find when presented in an array of digits compared to when it is presented in an array of letters (Fig.1). This is because the letter “pops out” more clearly when it is surrounded by digits and pops out less when it is surrounded by letters. For digits it is the reverse.

The cortical segregation of letter and digit recognition, i.e. letter and digit recognition being represented in separate cortical maps, accounts for this pop-out effect. The explanation of this perceptual “pop-out” effect is based on interference effects within the cortical map for letters, which makes recognition more difficult when one letter has to be found among other letters. This means that the representations of letter distracters, which are in the same cortical region as the representation of the letter target, interact with and interfere with the target representation.

Conversely, if a letter has to be recognised among digits, these interference effects do not occur, since the digit distracters are represented elsewhere in the brain, leading to a quick, undisturbed recognition of the letter. In this case, the letter pops out of its surrounding field and thus is recognised more quickly, and the pop-out effect shows that the corresponding cortical representations are separated from each other. Only if the pop-out effect did not exist could it be concluded that the cortical representations are not separated from each other.

Figure 1
Visual “pop-out” effect. Upper panel: a letter (B) is easier to recognise within an array of digits than a digit (9) within an array of digits. Lower panel: a digit (9) is easier to recognise within an array of letters than a letter (B) within an array of letters.

Sorting letters or not
The Canadian postal workers were separated in two groups: one group sorted letters, the other did not. Other than this, there were no differences between the two groups.

Someone who sorts letters for several hours a day in Canada processes letters and digits simultaneously during this time. This should lead to a representation of letters and digits in the brain where, unlike the usual representations, letters and digits are not represented in separate maps. Such a representation is in line with the hypothesis that there should be a reduced or non-existent pop-out effect in the Canadian mail sorters. This means it should take the Canadian mail sorters longer to find a letter among digits or a digit among letters compared to the postal workers who do not sort letters and thus are not used to processing digits and letters simultaneously.

This is exactly what was proven by Polk and Farah through their investigation of recognition times. As described above in Fig.1, Polk and Farah showed arrays of letters among which a digit or a letter, respectively, was hidden. The postal workers’ task was to find the letter or the digit among the array of letters. When they had found the required letter or digit, they had to press a button. The time that it took them to find the letter or the digit, i.e. their reaction time, indicated the time needed for the recognition of a letter or a digit within an array of letters.

The postal workers who did not sort mail showed the pop-out effect, i.e. they almost immediately found the required digit among the letters. In contrast, the mail sorters did not show the pop out effect. That means that they were slower in recognising a digit among an array of letters compared to the postal workers who did not sort mail. The reason for this is probably that letters and digits are not represented in different cortical maps in the brains of Canadian mail sorters, but in the same one, which causes interference effects during recognition. Thus, with an extraordinary example, this study shows that higher cortical functions, such as number or letter recognition, are stored and represented in cortical maps in an experience-dependent way.

This finding is also supported by a series of studies that investigated the functional differences between musicians and non-musicians, or bilingually educated people compared to people who first learned their mother tongue and then a second language in adulthood. Kim and colleagues found, for example, that second languages acquired in adulthood are spatially separated from the native language in the brain. However, when acquired during the early language acquisition stage of development, native and second languages tended to be represented in common frontal brain areas 3.

Further studies of this kind are expected and they should clarify how the human cortex “draws” and administrates maps in an experience-dependent manner.


1. Spitzer, M., (1996), Geist im Netz, Spektrum Akademischer Verlag, Heidelberg.
2. Polk, T.A. and M.J. Farah (1998), “The Neural Development and Organization of Letter Recognition: Evidence from Functional Neuroimaging, Computational Modeling, and Behavioral Studies”, Proceedings of the National Academy of Science of the U.S.A., Vol. 95, No. 3, February 3, pp. 847-852.
3. Kim, K.H., et al. (1997) “Distinct Cortical Areas Associated with Native and Second Languages”, Nature, Vol. 388, No. 6638, July 10, pp. 171-174.

(Translation of Spitzer, Manfred (2000),  "Post und pop-out: Zur Neuroplastizität kortikaler Karten bei Postbeamten in Spitzer" in Geist, Gehirn & Nervenheilkunde: Grenzgänge zwischen Neurobiologie, Psychopathologie und Gesellschaft, Schattauer, Stuttgart, pp. 18-21)