Discrete, place-defined macrocolumns in somatosensory cortex: Lessons for modular organization of the cerebral cortex

Abstract

In 1957 Mountcastle introduced the concept of the cortical column as the vertical processing unit of the cerebral cortex. This idea, the “columnar hypothesis,” was based on the then prevailing view that the cortex is most richly interconnected in its vertical dimension (Lorente de No 1949) and on Mountcastle’s demonstration in single-unit recording experiments in cat (and later monkey; Powell and Mountcastle 1959) primary somatosensory cortex (SI) that neurons in ~0.5�mm wide vertical columns are activated by peripheral stimuli of the same submodality and have similar receptive fields (RFs). Mountcastle (1957) and Powell and Mountcastle (1959) also showed that their cortical columns � later named “macrocolumns” to distinguish them from single-cell-wide “minicolumns” (Mountcastle 1978) can be separated from each other by abrupt boundaries, on the opposite sides of which neurons respond to stimuli of different submodalities and/or have prominently different RFs.