When participants listen to a rhythm produced by a human performer, they are able to “follow” the rhythm even though it can contain relatively large fluctuations in periodicity. This is due to the fact that autocorrelated temporal fluctuations contain relevant information for beat expectation: the occurrence of the upcoming beat is predicted by the occurrences of the preceding beats1. The Dynamic Attending Theory2 proposes that the entrainment to nearly-periodic rhythms emerges from the dynamics of interaction between neural systems acting as “internal oscillators”. The internal oscillators are coupled by their anatomical connexions and, therefore, are able to mutually adjust their synchronization, even after the perturbation of one of the oscillators. Importantly, the internal oscillators generate a repetitive time window within which the system is expecting to receive a stimulation. The auditory events of an autocorrelated rhythm match those expectational windows better than those on an non-correlated, random, rhythm. Therefore, the autocorrelated rhythm would reinforce the neural oscillations, yielding a better neural entrainment than a randomly fluctuating rhythm. In this study, we recorded the EEG activity elicited by auditory beats with autocorrelated fluctuations versus non-correlated fluctuations, in order to provide a direct electrophysiological measure of dynamic attending in healthy human participants.
Chemin, B., & Mouraux, A. (2017). TRACKING TIME VARYING ACOUSTIC RHYTHM: an EEG frequency-tagging approach of dynamic attending. Auditory Cognitive Neuroscience Meeting, Max Planck Institute, Leipzig, Germany. https://hdl.handle.net/2078.5/268320