On 2017 May 31, Lydia Maniatis commented:

"In summary, we have shown a nonlinear response to a compound of gratings (plaid) that does not arise purely from contrast normalization between spatial frequency channels."

There is zero evidence and zero rationale for the existence of "spatial frequency channels" in the visual system. A spatial frequency analysis of the retinal stimulation would not contribute to the derivation of the percept, but only make the job more difficult-to-impossible. The notion has never been tested, and it's not clear how it could be tested. Simple data derived from carefully constrained experimental conditions (for which the assumptions can be made to seem to fit) are simply interpreted as though these untestable/untested/false assumptions were true, and it were only a matter of fine-tuning "models" via data-fitting. Similarly, I could posit other invisible forces as being responsible, and interpret accordingly.

The notion that EEG's correlated with particular percepts can be directly (theoretically) correlated with the activity of particular populations of neurons in the brain is absurd, and, at any rate, has not been tested, and cannot be tested in the foreseeable future. The many reasons why it is absurd have been discussed by me in various pppr comments and by Teller (1984).

On 2017 May 31, Lydia Maniatis commented:

"In summary, we have shown a nonlinear response to a compound of gratings (plaid) that does not arise purely from contrast normalization between spatial frequency channels."

There is zero evidence and zero rationale for the existence of "spatial frequency channels" in the visual system. A spatial frequency analysis of the retinal stimulation would not contribute to the derivation of the percept, but only make the job more difficult-to-impossible. The notion has never been tested, and it's not clear how it could be tested. Simple data derived from carefully constrained experimental conditions (for which the assumptions can be made to seem to fit) are simply interpreted as though these untestable/untested/false assumptions were true, and it were only a matter of fine-tuning "models" via data-fitting. Similarly, I could posit other invisible forces as being responsible, and interpret accordingly.

The notion that EEG's correlated with particular percepts can be directly (theoretically) correlated with the activity of particular populations of neurons in the brain is absurd, and, at any rate, has not been tested, and cannot be tested in the foreseeable future. The many reasons why it is absurd have been discussed by me in various pppr comments and by Teller (1984).