Texture Perception: Tactile

Perception of Roughness

Modern work on roughness perception began with a series of studies by Susan Lederman and her colleagues employing special textures called gratings. A grating is a texture made of a series of closely spaced ridges, separated by deep grooves. Subjects moved their fingertips across the gratings and gave magnitude estimates of their roughness. Grating roughness was found to depend strongly on the width of the grooves, but to be virtually independent of the width of the ridges. This asymmetry was accounted for by a model in which roughness is proportional to irregularities that are imposed on the skin as the surface presses against it, for example, the amount of skin that bulges into the surface's grooves. Mechanical forces created within the skin by these local deformations are presumably what stimulate receptors and lead to our perceptions. Roughness is only marginally affected by the speed of movement of such surfaces, supporting the idea that spatial, rather than temporal, information is being used to create this aspect of perception.

The physiological basis of roughness judgments was explored by Kenneth Johnson and his colleagues. They recorded the responses of single receptors in the fingerpads of anesthetized monkeys as textured surfaces (arrays of bumps) were moved across them. It was found that roughness was predicted not by the firing rate of any one receptor, but by variations in firing among neighboring receptors. This was especially true in the case of a type of receptor that gives a sustained response to mechanical forces and is therefore called a slowly adapting mechanoreceptor. By the time messages from groups of these receptors reach the somatosensory cortex, they combine to influence certain cortical neurons in such a way that the neuron's firing rate reflects the bumpiness of the surface. Activity of these texture-detecting neurons may be the way the brain creates sensations of roughness.

This explanation of roughness perception, in terms of the neural response to a spatial pattern impressed on the skin, accounts for most data on the roughness of surfaces with texture elements larger than about 0.1 millimeter (mm). With smaller texture elements packed closely together, localized skin displacement is negligible, and elements cannot be resolved; yet roughness is still experienced and varies systematically with the geometry (perhaps better called microgeometry) of the surface. Clearly a different mechanism for coding roughness must be used for fine textures.

...