Nunez's description of the brain as a medium capable of wave propagation has provided some fundamental insights into its dynamics. This approach soon reaches the descriptive limits of the brain as a physical system, however. We point out some biological constraints which differentiate the brain from physical systems and we elaborate on its consequences for future research.

The theory of local/global neocortical EEG dynamics responds to newly emerging conceptualizations in neuroscience. An extended application of the model to event-related EEG activity composed of distinctive global and local functional epochs with presumably different timing is proposed.

Nunez is to be applauded for putting forward a theoretical brain model. In order to improve any model it needs to be experimentally testable. The model presented in the target article suffers from insufficient clarity as to how new experimental designs could be derived. This is a consequence of neglecting the purpose of the brain, which is to produce effective and adaptive behavior. It might be possible to overcome this drawback by including Hebb-based modeling.

Hierarchical systems such as cortical tissue will indeed exhibit dynamics that are scale-dependent, but Nunez's conjecture that theoretically independent transfer functions for each level can be specified and then combined ignores the fact that, in cortex, the same excitatory neuronal populations will contribute to both local and global dynamics.

This commentary focuses on how the large-scale cortical dynamics described in Nunez's target article are related to various phenomena at different scales, both spatial and temporal, in particular, how the brain dynamics measured with EEG could relate to (i) experience and mental state, (ii) neuromodulatory effects, and (iii) spontaneous firing and autogenerated electromagnetic effects.

Among the metaphors used in the target article are “musical instruments,” “water waves,” and other types of mechanical oscillators. The corresponding equations have inertial properties and lead to standing waves that depend on boundary conditions. Other, physiologically relevant quantities like refractory times are not contained in the mechanical oscillator model but occur naturally, for instance, in biological forest fire metaphors.

We used Nunez's physiologically based dynamic theory of EEG to make predictions about a clinical population of children with Attention Deficit Hyperactivity Disorder (ADHD) known to have neuronanatomical abnormalities. Analysis of high-density EEG data (long-range coherence) showed expected age-related differences and surprising regional specificity that is consistent with some of the literature in this clinical area.

This commentary discusses three features of the general theoretical framework proposed by Nunez: (1) Functional concepts, such as computation and control, are not foundational. (2) A mismatch between the concept of subcortical input and EEG output is problematic for the input/output operator concept of cortical dynamics. (3) The concept of brain state is relatively static.

Studies have shown that as MRI T2 relaxation time lengthens there is a shift toward more unbound or “free-water” and less partitioning of the protein/lipid molecules per unit volume. A shift toward less water partitioning or lengthened MRI T2 relaxation time is linearly related to reduced high frequency EEG amplitude, reduced short distance EEG coherence, increased long distance EEG coherence, and reduced cognitive functioning (Thatcher et al. 1998a; 1998b).

Metaphors, particularly the implicit ones, constrain imagination. If we think of the brain as a collection of centers of cognitive activations, lighting up on demand, then this becomes all we can imagine. By thinking of the cortex as propagating its functional work through physical waves, Nunez offers us a new, rich model for distributed representation. Now let's add real anatomy.

We have performed a set of experiments that correlate EEG spectral parameters with cognitive functions. The tasks (visual perception, supramodal object recognition, short-term memory) were chosen so that the cortical area involved extended over different length scales. The extent of the cognitive neuronal assemblies correlated inversely with the frequency where EEG synchronization was found. This provides a further relation between experiment and the theory put forward in the Nunez target article.

The development of theories of global cortical dynamics, using linear wave theory, owes much to the pioneering work of Nunez. His work leads to clear predictions on relations of brain size, axonal conduction velocity, and the frequencies of the cerebral rhythms. These predictions do not appear to be fulfilled, but their falsification constrains the range of parameters applicable in further formulations.

EEG and synaptic action fields provide experimental and theoretical entry points into brain complexity. Such entry is distinguished from the core system of cell assemblies assumed to underlie cognitive processing. The global theory of synaptic action predicts several new properties of EEG, providing limited penetration into brain complexity.

Joseph apparently does not understand the main purpose of my target article and how different it is from any purpose underlying his work. In addition, most of the neurological ideas of the target article for which he claims unacknowledged priority are not original to him, but instead predate the work of both of us.

In a recent BBS target article, MacNeilage (1998) presents what he claims to be the only theory that can account for the evolution of language. However, major portions of his target article basically repeat and in many respects are identical to the theories of language evolution and development first proposed and detailed by the present commentator.