EEG Correlates of Stages of Meditative Quiescence: a Pilot Study
ABSTRACT: This pilot study investigates EEG correlates of successive stages of concentration meditation. During the course of a 30 minute meditation session, an experienced meditator subjectively observed stages of meditative depth, and reported the presence of a particular stage of meditation using simple finger positions. Meanwhile, EEG data were measured at 19 locations on the scalp. The EEG data collected during each reported stage of meditation were independently analyzed. A comparison of the frequency spectra at successive stages of the meditation session show positive correlation between subjectively reported meditative depth and the strength of 7-8 Hz theta peaks at all locations, with the strongest signals present in the central locations. Only in the deepest stages do these peaks emerge in the frontal locations.
Technology developed over the course of the past 50 years has enabled scientists to measure and analyze subtle electrical signals on the human scalp. These electroencephalograph (EEG) signals are manifestations of the electrical activity of neurons in the brain, and their characteristics have been correlated with subjective experience. Lindsley (1952) examined the frequency spectrum of EEG signals of normal adults and identified four distinct bands that are correlated with various subjective states of consciousness:
Lindsley’s work, however, did not account for differences between EEG measurements at various locations on the scalp. Jasper (1958) introduced the standardized system for electrode placement in EEG research that is still in use today. There are seven frontal positions (FP1, FP2, F3, F4, F7, F8, Fz), three central positions (C3, C4, Cz), four temporal positions (T3, T4, T5, T6), three parietal positions (P3, P4, Pz), and two occipital positions (O1, O2). There are also two ear-lobe positions (A1, A2) used as electrical references for the other electrodes. Even-numbered positions correspond to the right side of the head, odd-numbered positions correspond to the left side of the head, and positions on the left-right plane of symmetry are denoted by the letter “z”. (This system is illustrated in each of the various figures shown later in this article.)
The relationship between EEG measurements and meditation was first investigated in the 1950s and 1960s. Although the differences between various meditative practices must be taken into account, these studies generally suggest that increased alpha and theta activity is strongly correlated with most meditative states of mind. Of particular relevance to the present paper is the study of advanced zazen meditators by Hirai (1974). Four EEG stages were identified in meditation sessions: (1) initial appearance of alpha, (2) increased alpha, (3) decrease in alpha, (4) increased rhythmical theta (bursts of 6-7 Hz primarily in central locations). The presence of these distinct phases of the objective EEG data suggest that there may be distinct subjective states that correspond to these phases during the meditation session. It is the purpose of this pilot study to experimentally explore this question.
Experimental Design and First-Person Description
In order to study the correlations between subjective states and EEG data during the course of a meditation session, a method is needed to identify and record subjective qualities of mind so that they can be correlated with EEG data. The experiment thus requires a subject who is skilled at reliably identifying distinct subjective qualities of mind during meditation. Although a very advanced meditator would be an ideal subject, for reasons of convenience the subject in this pilot study was myself. The particular meditation I practiced for this experiment is a common form of concentration/stabilization meditation that takes the breath as the object. This is a meditation I have practiced on-and-off for about 13 years, including intense practice during several retreats. It is practiced with the eyes open.
The experiment also requires a technique that allows the subject to communicate introspective data during the meditation without interfering significantly with the qualities of mind being reported. After experimenting with various communication schemes, it was found that the subjective experience of mental disruptions were minimized by using a simple communication technique that keeps physical movements to a minimum. The technique used for the present study uses the fingers of one hand to indicate a successive stage in the meditation session. A sequence of identifiable subjective stages were identified and practiced during a meditation session every day for several weeks prior to the experiment. For example, the meditation session begins with no fingers extended; after “settling” the mind and body and resting the mind on an object of meditation the next phase of the meditation is indicated by extending one finger (which remains extended). When a deeper sense of mental and physical relaxation and stability is experienced, an additional finger is extended. During practice, typically three to four fingers were extended by the end of a 30 minute session. The experiment was performed in the morning of Wednesday 15 March, 2000 in a small room containing research-grade EEG equipment. An EEG cap was placed on my head, and 19 electrodes were all connected. They all had impedances below 5 kilo-Ohms with reference to my ear lobes. As is usual in my morning meditation practice, I sat in a half-lotus on my zafu pillow. An experimental assistant (David) was present in the room to operate and monitor the EEG equipment. He also had the task of recording my finger signals during the experiment, as well as their corresponding EEG epoch numbers.
For a three minute period just prior to the formal meditation session, a “baseline” measurement was made. In order to avoid artifacts due to eye movements, it was necessary for me to hold the gaze of my eyes fixed as much as possible. This had the undesired effect of beginning to induce a slight meditative state of mind. This draws attention to the fact that even a “normal” state of mind is in some sense conditioned by various factors. In any case, the “baseline” at least provided a measurement of a state of mind prior to an extended period of meditation.
The meditation session proper then started as I started my digital watch timer for 30 minutes and rang my gong. In the preliminary phase (phase “zero”), my attention was intentionally scanned throughout my entire body, and any tensions were released. I also took two or three deep breaths. After about a minute or two of these standard preliminaries, I then began the initial phase of the concentration meditation, which was indicated by moving my right forefinger from the curled position to the extended position. In this initial phase my practice was to place the attention on the breath and continue to relax the mind and body. After a few minutes, there was a sense of slight deepening of relaxation of mind and body, and my middle finger on my right hand was also extended to signal phase two. The fingers remained in an extended position as long as I sensed that the degree of depth in the meditation was at a certain subjective level. After a few more minutes, the meditation deepened slightly more and I extended my ring finger to signal phase three. Most of the remaining meditation session was experienced in phase three.
It should be noted that there were loud sounds of a jackhammer and other construction sounds during the meditation session (the experimental room was located on the 6th floor of a building in downtown San Francisco). These distracted the attention from the breath. Rather than resisting these distractions, however, my attitude was to be open to them and accept them into awareness. Nevertheless, these sounds did tend to keep the mind and body at a certain level of stimulation and seemed to prevent relaxation from progressing beyond a certain point, as compared to other meditations in other more quiet circumstances. It should also be noted that, near the end of phase three, physical discomforts due to minor back strain and the pressure of the EEG cap on the head were distractions from the breath. As with the jackhammer, my attitude toward these was to open up to them with acceptance rather than resistance. Near the end of the 30 minute session, my state of consciousness shifted for a minute or two into a significantly deeper level that was accompanied by an enhanced sense of stillness of both mind and body, and an increased subtlety of awareness. At this point, the physical discomforts seemed to pale in comparison to the increased intensity of awareness. There was also an increased sense of disengagement with the visual field for a few moments. This state then seemed to gradually shift back toward phase three, but the meditation session then came to a close as the watch timer rang. Other than the discomfort near the end of phase three, the cap was not experienced as interfering with the meditation.
Regarding the finger signals, because I had practiced these signals every day or two for two weeks prior to the experiment, and since they required little effort or involvement, they did not seem to interfere much at all with the meditation.
Third-Person Results and Analysis
Two days after the experiment was performed, the EEG data were artifacted to remove distortions due to muscle movements (e.g., eye lid blinking and eye movement). The artifacted data for each phase of the meditation was then separately analyzed.
The preliminary “phase zero” stage of the meditation includes some small 12 Hz alpha peaks at most locations. Somewhat stronger 11-13 Hz alpha peaks are present in the rear-right region at T6, P4, and O2. There is also a somewhat strong 8-9 Hz alpha peak at T4. It is interesting that even at this preliminary stage of the meditation beta activity does not predominate, and already some slightly enhanced alpha is present.
The most noticeable difference between phases 0 and 1 is that phase 1 has significant 7 Hz theta peaks at C3, Cz, P3, Pz, as well as increased 8 Hz energy at C4 and P4. The 11-13 Hz alpha peaks seen in phase 0 are less prominent in phase 1. Thus, the spectra peak in the central region seems to shift from high frequency alpha to high frequency theta when the meditation begins.
Phase 2 of the meditation has an EEG signature very similar to that of phase 1. The 7 Hz theta activity present in phase 1 is slightly more enhanced in phase 2, with the peaks being narrower and more well-defined. The 11-13 Hz alpha peaks are still present, although slightly reduced.
Phase 3 shows extremely well-defined and narrow energy peaks in the 7-8 Hz theta band, located most prominently at P3, Pz, P4, C3, Cz, C4, but also quite clearly present at T3, T4, T5, T6, O1, O2, as well as at F3, Fz, F4. Smaller 7-8 Hz theta peaks are also present at FP1 and FP2. These peaks at F3, Fz, F4, FP1 and FP2 were not clearly discernable in earlier phases of the meditation. The theta activity in this phase of the meditation increases in its intensity and begins to spread to the frontal regions.
Phase 4 exhibits the same theta peaks as phase 3, with the exception that the peaks at P3, Pz, and P4 begin to broaden slightly. Phase 4 also has unusually strong beta activity isolated at T5. Because this beta activity is so strong and localized, it may indicate muscle activity.
Conclusions and Suggestions for Further Inquiry
The EEG data, combined with the introspective data, suggest that my meditative states are positively correlated with the presence of 7-8 Hz theta peaks. In the initial phases of meditation, these theta peaks are most prominent at C3, Cz, C4, P3, Pz, and P4. In deeper meditative states these peaks increase in strength and definiteness, and similar peaks begin to appear at other locations as well. These results are consistent with Hirai (1974), who also observed narrow 7 Hz theta activity associated with deeper stages of a meditation session.
A Biofeedback Protocol
This pilot study suggests a biofeedback protocol that might be useful in cultivating meditative states of consciousness. Providing real time feedback of enhanced 7-8 Hz activity in the frontal regions may assist in the cultivation of the deeper meditative states. It is important to remember that the EEG measurements and subjective states may not be in one-to-one correspondence. For example, two or more distinct subjective states may have the same or very similar EEG correlates. Thus, EEG biofeedback should be used as a supplement to, not a replacement for, guidance from a qualified meditation instructor.
Possible Extensions and Refinements
It might be possible and interesting to develop a more refined map of subjective states and attempt to correlate these with more subtle aspects of EEG data. For example, the Tibetan Buddhist tradition has precisely defined nine very specific stages in meditative quiescence (also called calm abiding). Expert meditative practitioners could use simple finger signals to communicate these stages of the meditation, and the results could be correlated with EEG data. (These experts may be able to signal just prior to shifting into the next stage, eliminating interference effects.) If strong correlations are observed, and if the same correlations hold true for all individuals, they may be useful in developing biofeedback protocols for training new meditators. Another application would be for the meditation instructor to use the data as supplementary information when evaluating students. The Nine Mental States/Stages of Meditative Quiescence are as follows:
- Placement. The object can be ascertained for brief moments with little continuity.
- Continual Placement. Brief periods of continuity on object.
- Patch-like Placement. Longer periods of continuity on object.
- Close Placement. The object is never forgotten. Gross excitement vanishes.
- Subduing. Continuity is maintained, but gross laxity is sometimes present.
- Pacifying. Gross laxity vanishes. Subtle laxity and excitement present.
- Fully Pacifying. Immediate application of antidotes to subtle laxity and excitement.
- Single-Pointed Application. With effort, no laxity or excitement arises.
- Meditative Equipoise. Effortless single-pointed attention. No antidotes required.
This technique involves an interplay between laxity and excitement. This suggests another possibility, which would be to use a two-dimensional map rather than a linear map of the subjective states. Fingers of the right hand could be used to indicate degrees of excitement, while fingers of the left hand could be used to indicate degrees of laxity. For an expert meditator, introspecting and reporting these two qualities of mind should be feasible without interfering with the meditation. For example, the laxity and excitement could be independently reported, each with three degrees. There are a total of six fingers are used (e.g., three fingers 1R, 2R, 3R on the right hand and three fingers 1L, 2L, 3L on the left hand), providing a map of nine states of mind, as shown in the table below.
Jasper, H. H. (1958), The ten-twenty electrode system of the International Federation. Electroencephalography and Clinical Neurophysiology, 10, 371-375.
Thomas J. McFarlane
Edited and revised for the web March 2004