2005 Gruber Neuroscience The Science

Sound localization is a computational process that requires the central nervous system to measure specific auditory cues and then associate particular cue values with appropriate locations in space. The neural basis of sound localization was elucidated in a brilliant series of experiments conducted over the course of ten years by Masakazu Konishi of the California Institute of Technology and Eric Knudsen of Stanford University.

One of the brain’s important design principles is that the external sensory world is represented in the form of topographic maps of space. For senses such as vision and touch, map formation within the central nervous system is relatively (!) straightforward; primary receptors in the sensory epithelium (the retina or the skin) give rise to axonal projections that replicate centrally the spatial representation that pre-exists in the periphery.

The inner ear, however, does not register the location of auditory stimuli, but rather their frequency, intensity and timing. For decades, therefore, the neural mechanisms by which the brain enables auditory localization appeared deeply mysterious.

Conventional wisdom held it to be unlikely that the auditory system uses a map-like representation of space, but Konishi and Knudsen thought differently. Guided by behavioral observations on the sound localization ability of barn owls and starting their search centrally rather than in the peripheral nervous system, Konishi and Knudsen made the startling discovery that a map of auditory space exists in the midbrain of the barn owl.

In that map, each neuron responds maximally only to sounds coming from a restricted region of space, and these neurons are arranged in an orderly sequence that recreates the topography of sounds in the external world.

This discovery was fundamentally important for two reasons. First, it demonstrated that novel spatial maps can be synthesized within the central nervous system based on primary cues that are encoded in the periphery. Second, it shows that maps are a primary mechanism used by the brain to represent and process sensory information.

Previously, it had been possible to argue that topographic maps (e.g. of visual space and the skin surface) were simply developmental adaptations whose major purpose was to get the brain wired up properly. This view was no longer tenable after Konishi and Knudsen revealed that the brain goes to great lengths to synthesize a map even when a template does not exist in the periphery.

After their initial collaborative studies, Konishi and Knudsen capitalized on their primary discovery in very different ways in their independent and remarkably productive laboratories.

Konishi and his junior colleagues conducted a series of experiments to identify the computational principles and neural mechanisms underlying synthesis of the auditory map, while Knudsen and his junior colleagues undertook a set of studies to elucidate the plastic mechanisms by which the auditory map is kept in spatial register with the visual map of space in the tectum (the dorsal part of the midbrain).

A number of important principles that emerged during study of auditory localization in the owl continue to exert a substantial impact on neuroscience today. Importantly, the basic neural design principles that were discovered in the barn owl are now known to generally apply to many other species including primates, although frequently in a less elaborate manner.

By William T. Newsome and Allison J. Doupe