hide
Free keywords:
-
Abstract:
When the eyes are closed the visual cortex generates high-amplitude oscillations in the alpha frequency range (8 to 13 Hz) which can be recorded on an electroencephalogram (EEG). This low input state is thought to reflect cortical idling. Here I propose that in fact the opposite is correct: large oscillation amplitudes reflect a state of alertness, that is high sensitivity in the visual system, preparing it to respond to weak stimuli. Opening the eyes reduces the amplitudes of the alpha oscillations (Berger effect). This is at least partly due to light adaptation indicating a loss of sensitivity. This relatively slow (compared to 10 Hz alpha frequency) light adaptation with a time constant of some 300 ms is the nonlinear component in the alpha EEG.
Brain oscillations are commonly thought to be generated by (nonlinear) oscillators. I propose that - as occasionally suggested - alpha waves basically are the outcome of a linear process such as band-pass filtering. I demonstrate that the superposition law holds (which defines linearity of a system): evoked potentials are just superimposed on on-going alpha waves. The phase of the latter is not affected. The results are in agreement with the view that alpha waves and the 10 Hz components of visually evoked potentials are generated by the same linear mechanism.
The linearity of the alpha waves generating mechanism implies that visual input cannot modify the phase of on- going alpha waves. If the same is also true for gamma oscillations, theories on brain oscillations which imply phase resetting by visual input (temporal correlation hypothesis) are questioned.