The six-layered neocortical columnar microcircuit, implemented most extensively in the human brain, supports a huge variety of sensory, cognitive and motor functions in mammals ranging from mouse to man. As discussed in Vernon Mountcastle’s introduction to this special issue, anatomical and functional investigations of radial structure in neocortex are converging on a definition of cortical columns; however, the secret behind the incredible flexibility and computational power of these structures has remained elusive. New insights are now emerging from several different areas of research, as cortical anatomists, physiologists, modellers and theoreticians join forces in attempting to decipher computation in cortical columns. This special issue of Cerebral Cortex is intended to provide an overview of the current efforts, in fields ranging from anatomy and synaptic physiology to neural imaging and theoretical modelling.

The special issue on ‘Computation in Cortical Columns’ was inspired by a workshop we organized in December 2000 at the Neural Information Processing Systems 2000 (NIPS*2000) meeting in Breckenridge, Colorado. Thanks to the generous support of the NIPS Foundation and the Sloan Foundation, that workshop brought together more than 25 leading experimental and theoretical neuroscientists to speak with an interdisciplinary audience of conference attendees. This special issue addresses in depth many of the questions that were debated in the NIPS*2000 workshop: Does a common denominator, a repeating micro-circuit element, exist across the neocortex? To what extent does a functionally defined column reflect a distinct anatomical entity? What are the computational roles of the different cell and synapse types, and of the feedforward and recurrent circuitry? What response properties are similar and different among neurons within a column? Does the column function as a discrete computational unit? What components of the cortical microcircuit, and what rules of columnar organization, are sufficient to generate observed feature selectivity in sensory cortex? Are there common themes for columnar function across different cortical areas? What are the computational advantages of the recurrently connected, multilayered structure of cortical columns?

We hope that this special issue of Cerebral Cortex will inspire further discussion about these and other questions, and further interdisciplinary efforts to understand computation in cortical columns.