The Trap and the Rabbit

The primary purpose of language is to fit the reality of what you are attempting to describe. In science and clinical practice, language should be elegant: maximally explanatory, yet minimally discursive. It should say as much as necessary, and no more. When we oversimplify in the name of convenience, we often throw out the baby with the bathwater.

Consider how traditional neurofeedback speaks about the brain. For decades, the field has relied on the static language of isolated “frequencies”. But isolated frequencies make sense only if you assume a non-varying, stationary signal.

When you tune a household FM radio to 102.3, you must land squarely on 102.3 to lock onto the carrier wave. Turn the dial to 103, and the music dissolves into static.

The living human brain is nothing like an FM receiver. Nobody measures static, stationary frequencies inside living brain tissue. The central nervous system is composed of complex networks of nonlinear biological oscillators. They ping, they strobe, and they burst in dynamic, ever-shifting rhythms.

Standard Fourier analysis assumes an infinite time domain. Even short-time Fourier transforms, which chop data into brief intervals, disrupt frequency-based accuracy during the calculation.

Because cortical activity is non-stationary and constantly evolving, time and frequency cannot be separated. This is why dynamical processing relies on time-frequency analysis—utilizing transforms such as Choi-Williams, Wigner-Ville, and Gabor models.

Instead of freezing the brain into an artificial frequency bin, time-frequency transforms respect the natural movement of cortical oscillators as they unfold across time. The model fits the reality, rather than forcing the living brain to fit the snare.