The only access we have to the outside world is through neural activity. A central question in neuroscience is:​ How do the patterns of neural activity which correspond to perceptions arise, and how do they represent features of the outside world? On the one hand, a central challenge for a sensory system is to form stable internal neural representations of the invariant properties of the outside world in the presence of noise and self-induced signal variations. On the other hand, it is also possible th at useful information about our relationship to the outside world is present in these signal variations.

We addressed these questions using the vibrissa (whisker) system in awake, freely behaving rats. We trained rats to run on a linear track so th at their vibrissae brushed against tactile stimuli placed on the sides of the track. We used high-speed, high resolution videography to measure the motion of the base of the vibrissae, which is the input to the sensory system, and simultaneously recorded neural activity in the area of primary somatosensory cortex responsive to vibrissa motion. By having access to all three parts of the sensory system (the external tactile stimulus, the vibrissa motion, and the internal neural activity), we were able to determine how information was represented in each part of the pathway.

We found that the motion of the base of the vibrissae contained information related both to the features of the textures and to the animal’s own motion as it ran down the track. We were able to classify trials as belonging to one of four texture classes with significantly greater than chance performance using a single scalar measure of the vibrissa deflection power spectrum. We also found that multi-unit spiking activity in the cortex was positively correlated with the same measure of vibrissa deflection, and th at trials could be classified using the neural activity. Our results suggest th at cortical cells encode the integrated power of the vibrissa deflection angular velocity, and th at this quantity is also a good measure of tactile information.