Klimesch (1994) proposes that reduced spectral power in various frequency bands (the gamma-band included) is an indicator of increasing task difficulty. He argues against the possibility that the enhancement of gamma power could be an indicator of activation of a cortical cell assembly. Our response is the following: (1) Klimesch needs to spell out his model in terms of biological mechanisms. The result will probably be one within the cell assembly framework. (2) In recent experiments, focal enhancement of gamma-band power has been recorded in the EEG, whereby the local distribution is a critical factor which discriminates it clearly from the more unspecific type of distributions commonly found in the alpha or beta range.
1. In his comment on our target article on brain rhythms (Pulvermueller et al., 1994), Klimesch (1994) makes the point that an account of EEG phenomena should refer to inhibitory processes. We wish to emphasize that in our target article two possible functions of inhibitory mechanisms are specified. First, nonspecific inhibitory mechanisms are assumed to play a critical role in the regulation of cortical activity; second, local inhibitory processes are discussed as one possible source of rhythmic activity patterns (see Figure 3 of our target article). It may be useful to base the present discussion on some aspects of the brain theory underlying the proposed model.
2. According to the brain-theoretical view adopted, the human cortex is a huge associative memory in which neurons that are frequently active at the same time strengthen their connections and develop into an assembly, a functional unit (Hebb, 1949). In a network including several assemblies, a regulation mechanism must be assumed. If not, it would be hard to explain why activation of one assembly will not automatically "ignite" neighboring assemblies, and eventually the whole cortical network (Braitenberg, 1978; Braitenberg & Schuez, 1991; Palm, 1982; Palm, 1990). Klimesch (1994) argues that an inhibitory mechanism must "distinguish between the relevant and the irrelevant part of the network" (paragraph 3). This is not necessarily so. In fact, the inhibitory mechanism may well be nonspecific, that is, if one transcortical assembly ignites, all others will receive inhibition. It has been proposed that the striatum, a network with powerful lateral inhibition, may serve this function (Miller & Wickens, 1991; Wickens, 1990). According to Elbert and Rockstroh (1987; Birbaumer et al., 1990; Rockstroh et al., 1989), EEG alpha activity is caused by such nonspecific regulation processes that guarantee the cortical equilibrium of activity. We believe that the problems Klimesch addresses regarding specific inhibition do not arise in this brain-theoretical approach (while they may be present in connectionists' models).
3. The following two results from our EEG and MEG experiments call for an explanation. (1) In distinct brain regions, gamma-band power is reduced after presentation of pseudowords, while it is not reduced after word presentation. (2) The presentation of pseudowords leads to a much stronger late negativity compared to words. In order to explain (1), we assume that word presentation leads to an ignition of a strongly coupled cortical network in which activity circulates, while pseudoword presentation leads to uncorrelated activity in loosely coupled neurons of different assemblies. In addition, continuous cell assembly ignitions must be assumed during the baseline. Klimesch is absolutely correct in pointing out that this "baseline assumption" is a stong one. However, it is easy to design experiments to test this assumption (see below). We explain (2) by assuming that only one cell assembly ignites after presentation of a meaningful word, while no full activation, no ignition of one single network takes place after pseudoword presentation, but activity is present in several assemblies. The underlying logic is the following: in-part activation of, say, ten assemblies may imply more neural activity compared to full activation of only one assembly. However, Klimesch's view offers an alternative explanation: it may well be that (2) goes back to inhibitory processes. If one assembly ignites (word presentation), the threshold regulation mechanism may lead to general reduction of activity (smaller negativities). If several assemblies are only slightly pre-activated, the inhibitory mechanism may not be switched on (larger negativities). Thus, Klimesch is correct in pointing out that additional explanations can be created related to inhibitory processes.
4. Klimesch proposes what he considers an alternative explanation of our results on differential gamma-band responses to words and pseudowords. He draws attention to the well-known findings that power of the alpha (and sometimes also the beta) band decreases with increasing task difficulty. By extension, reduction of gamma-band power may also be a correlate of more difficult tasks. We have one principle problem regarding this proposal: The correlation of psychological phenomena (e.g., task difficulty) and physiological phenomena (e.g., gamma depression) is only the very first step in a psychophysiological explanation. We have to ask: WHY gamma depression should show up in very complex tasks? Here, we are in need of a physiological explanation. The one we proposed in the target article is the following: certain complex activities (those during which different cognitive entities are simultaneously present in one's mind) may induce non synchronized activity in several cell assemblies, causing a reduction of gamma-band power in the EEG or MEG. In contrast, simple tasks may lead to activation of only one assembly and, therefore, to well-synchronized activity in the gamma-band of the EEG/MEG signal. This becomes very clear in the case of words and pseudowords. During processing of a pseudoword, the subject may be simultaneously reminded of several words and the induced processes of lexical search (pre-activation of several assemblies) may be much more complex compared to simple lexical access after word presentation (ignition of one assembly). We feel that if one tries to elaborate Klimesch's proposal in biological terms, the resulting model will include the cell assembly concept.
5. Klimesch presents theoretical considerations suggesting that "cortical activation processes... lead to... a suppression of band power" (paragraph 6). This implies that enhancement of gamma-band power should never be observed in an experiment where cortical activation processes are induced. However, local enhancement of gamma-band power has already been reported (Pfurtscheller & Neuper, 1992; Pfurtscheller et al., 1993; 1994). The cell assembly model predicts that gamma depression or increased gamma power should occur, depending on what processes take place in the baseline. If cell assembly ignition is excluded during the baseline by the experimental paradigm, gamma enhancement instead of gamma depression should be observed. Moreover, gamma enhancement may be area-specific if cell assemblies have distinct cortical topographies. In a recent experiment, we put these assumptions to a test using visual stimuli. (For further details of this experiment, see Lutzenberger, Pulvermueller, Elbert & Birbaumer (1994).) Slowly moving regular bars were compared to irregular patterns of bars appearing at randomly selected sites. During the baseline, only the irregular pattern of bars was visible. In the critical conditions, regularly moving bars appeared either in the upper or lower visual field. Surprisingly, enhancement of gamma-band activity (relative to the baseline) was found when regular patterns appeared. The cortical topography of gamma-band enhancement changed with visual field presentation of regular visual patterns. These results are consistent with similar experiments in animals using single cell and multi-unit recordings (Engel et al., 1992; Singer, 1994). During presentation of the "meaningful" stimulus, specific neural populations join a rhythmic activity patter. The topography of the cortical population being activated changes with stimulus properties. We believe that there is increasing support for the view that gamma-power can be enhanced when cortical populations are activated (see e.g. Pantev, Elbert & Luetkenhoener, 1994).
Birbaumer, N., Elbert, T., Canavan, A.G.M. & Rockstroh, B. (1990) Slow Potentials of the Cerebral Cortex and Behavior. Physiol. Rev. 70:1-41.
Braitenberg, V. (1978) Cell assemblies in the cerebral cortex. In Theoretical approaches to complex systems. (Lecture notes in biomathematics, vol. 21). R. Heim & G. Palm, eds. Berlin: Springer Verlag, pp. 171-188
Braitenberg, V. & Schuez, A. (1991) Anatomy of the cortex. Statistics and geometry Berlin: Springer Verlag.
Elbert, T. & Rockstroh, B. (1987) Threshold regulation-a key tothe understanding of the combined dynamics of EEG and event related potentials. Journal of Psychophysiology 4:317-333.
Engel, A.K., Koenig, P., Kreiter, A.K., Schillen, T.B. & Singer, W. (1992) Temporal coding in the visual cortex: new vistas on integration in the nervous system. Trends in Neurosciences 15:218- 226.
Hebb, D.O. (1949) The organization of behavior. A neuropsychological theory New York: John Wiley.
Klimesch, W. (1994) The ignition of cortical cell assemblies: some arguments against the assumption of a selective increase in gamma band power. Psycoloquy 5(58).
Lutzenberger, W., Pulvermueller, F., Elbert, T. & Birbaumer, N. (1994) Local 40-Hz activity in human cortex induced by visual stimulation. Neuroscience Letters (in press).
Miller, R. & Wickens, J.R. (1991) Corticostriatal cell assemblies in selective attention and in representation of predictable and controllable events: a general statement of corticostriatal interplay and the role of striatal dopamine. Concepts in Neuroscience 2:65-95.
Palm, G. (1982) Neural assemblies Berlin: Springer Verlag.
Palm, G. (1990) Cell assemblies as a guideline for brain research. Concepts in Neuroscience 1:133-147.
Pantev C., Elbert T., Luetkenhoener B. (eds.) (1994) Oscillatory Event-Related Brain Dyamics. New York, London, Plenum Publishing Corp.
Pfurtscheller, G., Flotzinger, D. & Neuper, C. (1994) Differentiation between finger, toe and tongue movement in man based on 40 Hz EEG. Electroencephalography and Clinical Neurophysiology (in press).
Pfurtscheller, G. & Neuper, C. (1992) Simultaneous EEG 10 Hz desynchronization and 40 Hz synchronization during finger movements. NeuroReport 3:1057-1060.
Pfurtscheller, G., Neuper, C. & Kalcher, J. (1993) 40-Hz oscillations during motor behavior in man. Neuroscience Letters 164:179-182.
Pulvermueller, F., Preissl, H., Eulitz, C., Pantev, C., Lutzenberger, W., Elbert, T. & Birbaumer, N. (1994a) Brain Rhythms, Cell Assemblies and Cognition: Evidence from the Processing of Words and Pseudowords. PSYCOLOQUY 5(48) brain-rhythms.1.pulvermueller.
Rockstroh, B., Elbert, T., Canavan, A., Lutzenberger, W., & Birbaumer, N. (1989) Slow cortical potentials and behaviour Baltimore: Urban & Schwarzenberg.
Singer, W. (1994). Putative functions of temporal correlations in neocortical processing. In Large scale neuronal theories of the brain. C. Koch & J. Davis, eds. Boston, MA: MIT Press.
Wickens, J. (1990) A theory of the mammalian striatum Dunedin: Doctoral thesis, University of Otago.