A solution to the controversy between rate and temporal coding.

The apparent irregularity of in vivo spike sequences could be read as being derived either irregularly in time from a constant rate or regularly from a fluctuating rate. We recently proposed the Empirical Bayes method that selects one out of the alternative interpretations (J. Phys. A: Math. Gen. 2005; 38:L531-L537). In this paper, we demonstrate that this method may also give a solution to the controversy between rate and temporal coding. For a given single sequence of spikes, the method may indicate two potential hypothetical rate processes, nearly constant and highly fluctuating, and selects one according to the statistical plausibility. As the rate fluctuation of a spike sequence is increased, the Bayesian interpretation exhibits a discontinuous transition from one to the other. In this way, a sequence of spikes can be distinctly categorized as either rate or temporal coding, according to the estimated degree of rate fluctuation.