Goertzel (1994a) emphasizes that the activation of a cell assembly does not necessarily imply periodic neural activity. He advocates more complex models in which one cell assembly may represent different cognitive entities depending on its mode of activation. We fully agree that more complex theories must be developed that avoid various simplifying assumptions implicit in our model. However, we consider it a good research strategy to start with simple models and to proceed to more complex ones when limitations of the simple models become evident.
2. The model adopted in the target article implies that each cognitive entity has its own cell assembly. In particular, each word is assumed to have a cortical representation in a specific transcortical assembly. While at first this position may appear to be very simple, it turns out that if additional phenomena are considered, it leads to more complex assumptions. For example, there are sentences in which one word occurs several times. This is true for the first sentence of this paragraph (where the article "the" is used twice) and also for more extreme examples, such as the sentence "the spaces between lost and and and and and found are too large". In a brain model of language postulating that each word is represented in one assembly, additional assumptions must be made in order to allow for a representation of such sentences. If there is only one assembly representing an individual word, different states of activity of this assembly may be assumed to correspond to the number of occurrences of this word in a particular sentence. This is, in principle, very similar to the proposal by Goertzel (1994a; 1994b) that different activity states of the same assembly may correspond to different cognitive states.
3. Goertzel's (1994a; 1994b) view that the dynamics of active cell assemblies are characterized by strange attractors is certainly a possible starting point for further investigation of brain dynamics. We agree that there is some evidence for chaotic-like phenomena in EEG dynamics. However, there are considerable theoretical and practical problems if one analyzes EEG recordings with methods from "chaos" science. At this point, we see no demonstration that the dynamics of neuronal networks can be described by a strange attractor. However, there is some indication based on estimations for the fractal dimension, Lyapunov spectra, and similar measures that neuronal dynamics do demonstrate strange attractor behavior (Skarda & Freeman, 1987; Elbert et al., 1994). Furthermore, we wish to emphasize that there is little evidence that a larger neuronal system may primarily exhibit deterministic low-dimensional chaos. Estimations of the deterministicity reveal considerable random behavior, even in such states as coma, where dimensionality is lowest (Muehlnickel et al., 1994). Thus, the assumption that cell assemblies may produce low-dimensional chaos currently remains a hypothetical one (for further discussion, see Lutzenberger, Preissl & Pulvermueller, 1994).
4. What can we learn from this discussion? We believe that one conclusion is very obvious: At this point, strong assumptions are necessary for making progress. Little is known about the networks in the brain necessary for reasoning or for using language. Without theorizing, there is no perspective for uncovering their function. Thus, any theory of cognitive brain processes needs speculative ingredients. One may agree to such theorizing as long as it stimulates empirical research. At this point, theoretical considerations do not help us to determine whether cell assemblies exhibit rhythmic activity or whether the sequence of their activity states is best described by strange attractors. In our target article, we preferred to investigate what appeared to us to be the simplest possibility. Why not try simple models first and look how far they take us? However, it is perhaps best to explore simple as well as more complex possibilities. When limitations of simple models become apparent, it is advantageous to have more complex ones they can be replaced with.
Elbert, T., Ray, W.J., Kowalik, Z.J., Skinner, J.E., Graf, K.E. & Birbaumer, N. (1994) Chaos and physiology: deterministic chaos in excitable cell assemblies. Physiol. Rev. 74:1-47.
Goertzel, B. (1994a) Periodic Brain Responses and Beyond. Commentary on Pulvermueller et al. on Brain-Rhythms. PSYCOLOQUY 5(51) brain-rhythms.3.goertzel.
Goertzel, B. (1994b) Chaotic logic: language, thought and reality from the perspective of complex systems science. New York: Plenum.
Lutzenberger, W., Preissl, H. & Pulvermueller, F. (1994) Fractal dimension of EEG time series and underying brain processes. submitted.
Muehlnickel, W., Rendtorff, N., Kowalik, Z.J., Rockstroh, B., Miltner, W. & Elbert, T. (1994) Testing the Determinism of EEG and MEG. Integrative Physiological and Behavioral Science (in press).
Pulvermueller, F., Preissl, H., Eulitz, C., Pantev, C., Lutzenberger, W., Elbert, T. & Birbaumer, N. (1994) Brain Rhythms, Cell Assemblies and Cognition: Evidence from the Processing of Words and Pseudowords. PSYCOLOQUY 5(48) brain-rhythms.1.pulvermueller.
Skarda, C.A. & Freeman, W.J. (1987) How brains make chaos in order to make sense of the world. Behav. Brain Sci. 10:161-195.