The term “cognition” can be roughly translated as knowing what is going on out there and responding appropriately. The neocortex can be considered as the ultimate organ of cognition but no cortical area makes sense on its own. Every student of neuropsychology knows that specialized regions of the neocortex handle different sensory (input) or motor (output) modalities. The occipital lobes handle vision, and the temporal lobes, sound. The sensory cortex on the front edge of the parietal lobe receives and interprets all body sensations. The adjacent frontal lobe motor cortex outputs body movements.
Vernon Mountcastle summarized the work done to elucidate the cellular organization of the neocortex dominated by vertical columns of neurons usually 6 layers deep. Cortical columns are formed by the linkage of minicolumns by common inputs and short range horizontal connections. The number of minicolumns per column varies between 50 and 80. Long-range, intracortical projections link columns with similar functional properties. Columns vary between 300 and 500 µm in transverse diameter, and do not differ significantly in size between brains that vary in size over three orders of magnitude.
Cortical expansion during evolution involved increases in surface area with little change in thickness. There is a consistent columnar structure all over the brain. Different inputs, outputs and variations in interconnectivity are the basis of cortical specialization. Neurotransmitters also vary and reflect old specializations.
The main question – what do cortical columns actually do? - remains unanswered. In a simplified schemata, inputs come from the thalamus, other cortical columns, cerebellum, and outputs go to the thalamus, other cortical columns and to a complex array of effector systems in all parts of the brain. There are dense, recursive, looping networks and more focused input and output networks that connect the neocortex to the rest of the body and the outside world.
We can think of local cortical processors that do their own thing, but at the same time coordinate their activities with other local processors. In addition there is an overall context within which the processors must work together. The contextual field (CF) involves connections that link processors within and between cortical regions.
The direct input of sensory information enters receptive fields (RF). RFs and CFs interact to guide learning and information processing. Cortical computation requires flexible evaluation of relations between input signals by specialized processors whose activity is coordinated among regions by contextual connections.
Phillips and Singer suggested some important distinctions. For example, the organization of cognition into sub-systems is not based on recognizing differences in the information processing operations that they perform, but only differences in the information they send and receive. Some cognitive functions require special information processing capabilities such as episodic memory and working memory; intentional representation, creative aspects of language and strategic planning.
They stated that: “Cognitive functions that are central to human mental life depend on cortical activity and may not arise in a simple way from capabilities that are common to cortex in general, however, because (i) intentional representation and language, are not characteristic of mammals but are restricted to a few; (ii) in contrast to skills, episodic memories cannot be acquired in the absence of the hippocampus, and may require special computational capabilities and (iii) the ability to dynamically create more than one level of grouping within the same set of units, such as ((AB)(CD)), involves special computational problems. We have argued that networks of local processors can discover relevant information in diverse data sets… the distinction between representation and referent is critical. The relation between representation and referent can be iconic, symbolic, or both, and the relationship is asymmetrical. "