The sensed presence is the experience of another consciousness or sentient being within close proximity.1 These phenomena, which involve attribution of personal cognition to another consciousness, are prevalent in all cultures.2 The sensed presence has been hypothesized to be the transient awareness of the right hemispheric equivalent of the left hemispheric sense of self.3 In double-blind studies, asymmetric activity within the temporoparietal region—associated with intrinsic stimulation or externally applied, physiologically patterned, weak magnetic fields—has been associated with reports by healthy volunteers of a sensed presence and the feelings that thoughts did not originate from the subject’s own mind.4 Two case reports with patients, each associated with a powerful sensed presence, showed increased perfusion within the frontoparietal or frontotemporal regions.5,6 These data suggest that the brain correlates of the sensed presence may be similar in healthy and patient populations, and may differ only in subjective intensity, duration, and the degree to which the phenomenology dominates the person’s experiences and beliefs. This would be typical of all behaviors.

Frith7 and others have suggested that sensing a presence outside of the self, a common correlate of Schneiderian or first-rank positive symptoms of schizophrenia, may arise from misattribution of one’s own action or thoughts to an external agent. The cerebral patterns associated with the capacity to discern neurocognitive processes as originating from within rather than outside the self ("ego-alien") involve the frontal-temporoparietal regions.8 Reduction of the normal coherence between electrical activity between the (self-monitoring) prefrontal lobes and the temporal lobe speech-related areas in schizophrenia patients has been hypothesized as the source of misattribution of self-generated thoughts to an external source.9 In fact, imaging studies have shown that activation within the inferior parietal lobe resulted in attributing self-action to another.10,11

A recent imaging study by Franck et al.12 showed the severity of first-rank symptoms was correlated positively with rCBF (O-PET) in Brodmann areas 7 and 40 of the parietal cortices. According to Taber and Hurley,13 these more medially activated areas involved the white matter contributing to the caudal corpus callosum. Alterations in white matter associated with the prefrontal and temporal lobes have been reported in the brains of schizophrenia patients whose symptoms were most severe. Roll et al.14 found that the cerebral SPECT of Shawn Harribance, a person who has exhibited a lifelong capacity to accurately discern the personal history of others by mechanisms not known to date, displayed increased perfusion within the right medial parietal lobe during the discernment. He attributed the information to "other entities" that he described as "angels." If misattribution of self-generated thoughts are due primarily to anisotropic activity between the inferior parietal and prefrontal regions, then this discrepancy of ongoing function should be evident in healthy volunteers even with the relatively less precise quantitative electroencephalography (QEEG).

To test this hypothesis, 12 university students (six men, six women), 20 to 25 years of age, volunteered (with informed consent) as subjects for a relaxation study (no reference to the sensed presence) after a single 5-minute recruitment request within the class. We reasoned that if the effect size between intracerebral increase in electroencephalographic power and the experiences of a sensed presence were either causal or strongly correlated, the association should be statistically significant (p<0.05) even with this small sample.

Participants were exposed continuously for 5 minutes each to six successive magnetic field patterns. Between each 5-minute period there was a period of no field for 30 seconds. Six 20-second samples of electroencephalographic paper and electronic records were obtained simultaneously for each participant during the middle of each 5-minute exposure period. During this 30-minute period the subject sat in a comfortable armchair. The hardcopy was employed to verify that the 20-second samples for quantitative analyses were not conspicuously contaminated by movement or machine artifacts. A blindfold and disposable earplugs minimized the effects of ambient stimuli. Monopolar (referenced to ear) recordings were obtained with a Grass model 8—16 machine from the left and right frontal (F7, F8), temporal (T3, T4), parietal (P3, P4) and occipital (O1, O2) positions in accordance with the 10/20 system of electrode placement. The participants were instructed to relax but to avoid drowsiness. Visual assessment of the paper records showed no evidence of Stage II spindles for any of the subjects; all records were also artifact free.

The signals for each 20-second segment were extracted by Labview software with a sampling rate of 1 kHz. In order to obtain the frequencies, fast Fourier analyses software filtered the raw data and generated power spectra for delta (1 Hz—3.9 Hz), theta (4 Hz—7.9 Hz), low alpha (8.0 Hz—10.4 Hz), high alpha (10.5 Hz—13.0 Hz), low beta (13.1 Hz—20 Hz) and gamma (35 Hz—45 Hz). On the basis of our theoretical approach to consciousness15 as being strongly correlated with second and third derivatives sequestered within narrow bands of theta activity,16 relative power was also obtained for six successive 1 Hz increments between 4.0 Hz and 9.0 Hz.

During the 30-minute period the six different configurations of weak (1 microT) physiologically patterned (frequency-modulated and burst-firing) magnetic fields were successively generated around (circumcerebral) the head through eight pairs of small solenoids separated by 45 degrees. They were arranged in the horizontal plane just above the ears.17—19 The counterclockwise and clockwise rotational velocities and derivatives of the change (acceleration) have been described previously18,19 and primarily involved durations of 10 msec to 20 msec at the area of stimulation with either accelerations or negative accelerations of +2 or −2 msec. There is no evidence of induction artifacts within the EEG electrodes with these configurations.

These experimental conditions are often associated with the experience of a sensed presence and other mystical states that complement or accompany the attribution of experiences to nonself sources. In addition, our older, qualitative EEG (bipolar) measures have generally shown brief appearances (about 2—10 seconds) of irregular theta activity over the right temporal lobe during these experiences. Subjectively, the "location" is usually attributed to the upper left peripheral visual field for most subjects.

Relative scores for power within each of the different frequency bands and for each of the 1-Hz increments within the theta range were calculated for each of the eight channels for the two hemispheres and four lobes by dividing the activity exhibited during the field presentations by the baseline (no field). In order to discern intracerebral effects (each brain as its control), z scores were calculated by subtracting the raw power during each of the field configurations by the mean of all six configurations and dividing by the standard deviation (SD) for these values.

After the 30-minute exposure, each participant completed a 20-item exit questionnaire4,20 that reflected the type of experiences often reported in this context. The response (0=never, 1=once, 2=many times) to the item "I experienced a presence" served as the dependent variable. Spearman rho and Pearson r correlations were completed between this response and the z scores and relative scores for power for each of the six lobes and each of the six configurations of magnetic field presentations. Because there was no appreciable difference in the strength of the correlations between the configurations, the relative values and z-scores for the power scores for the six samples recorded for the six configurations were averaged.

The only statistically significant (p<0.05) and conspicuous correlation between the reported experience of a sensed presence and electroencephalographic power involved the right parietal and right frontal lobes within the theta band, and not within the delta, alpha, beta, or gamma bands. The Spearman rho values ranged between 0.78 and 0.86. The strength of the correlations for the right temporal lobe within each 1 Hz increments of the theta range was between 0.63 and 0.70, but did not achieve our a priori criterion for statistical significance. All other correlations for the other cerebral areas (right occipital, all four left hemispheric measures) were not statistically significant. Specific analyses with each of the 1-Hz increments within the theta range indicated that both relative scores and z scores for power over the right parietal lobe were positively correlated (0.83, 0.77, respectively) with the report of a sensed presence only within the 4.0 Hz to 4.9 Hz increment. On the other hand, the only statistically significant increment associated with the sense of a presence for theta activity over the right prefrontal region was for the 7.0 Hz to 7.9 Hz band for both the relative and z-score measures of power (ρ=0.77, ρ=0.78, respectively). The effect was conspicuous.

The subjective experiences for those participants who reported a sensed presence were similar to those reported by others exposed to these procedures. The detailed experiences have included feeling another self, sensing a deceased person known to the participant, experiencing a human-like gray face with lights around it (upper left peripheral field), and experiencing "someone" standing nearby.

These results are convergent with similar results inferred from several studies using various imaging methods while individuals experienced a sensed presence or the sensation of another proximal sentient being.21 Our results also support the recent report by Taber and Hurley13 that activation within the parietal lobe is associated with the attribution of actions to another. In our study the increase in power within the theta range over the right prefrontal and parietal regions was reflected within different discrete frequency bands. Such discrepancies would be consistent with the diminished coherence of the electrical activity within these two regions during attribution of experience to others.

It may be relevant that the "beat frequency," the difference between the two peaks, would be 3 Hz to 4 Hz; this was the dominant paroxysmal discharge over the temporal lobes observed during a powerful experience of "God’s presence" by a meditating woman22 in our laboratory many years ago. The inference of origin from the montage for this bipolar measurement was the right temporal lobe. A moderate correlation (ρ=0.49) between the amount of theta activity over the temporal lobe and the report of a sensed presence as well as greater right hemisphericity was reported for healthy volunteers (n=26) by Munro and Persinger.23

These results are also consistent with the older clinical reports by Slater and Beard,24 who described a schizophrenia-like psychosis associated with temporal lobe epilepsy where paranoid ideation was dominated by religious themes. Sudden religious conversions and the feeling of powerful presences of God in patients with temporal lobe epilepsy (frequently with right hemispheric foci) were also reported by Dewhurst and Beard.25 In healthy volunteers (n=28) the numbers of spontaneous spikes over the temporal (bipolar T3, T4) lobes but not over the occipital lobes were moderately (Spearman ρ=0.67; Pearson r=0.76) correlated with psychometric indices of strength of religious beliefs.26

The apparent contradiction between temporal lobe versus parietal lobe anomalies as common correlates of the sensed presence or the right hemispheric equivalent to the sense of self1,20 may reflect the type of misattribution. If the hypoperfusion data of Hillis et al.27 are generalizable to nonstroke patients, then misattributions that are allocentric (on the left side of individual stimuli, regardless of location with respect to the viewer) would involve the right superior temporal gyrus, while misattributions that are egocentric (the left side of the participant) would involve the right angular gyrus. The subtle distinctions in the complexity of the associated subjective experiences may require more sophisticated psychometric testing. In addition, recent fMRI data have shown that the functional source for the classic left-side neglect associated with right parietal lobe injuries involves the right superior temporal gyrus.28 This may facilitate the explanation for why the sensed presence can be associated with parietal21 and temporal lobe29 anomalies.

It is important to realize that "remembering" or "reiterating" mystical experiences attributed to God or other sources, such as reported by Beauregard and Paquette30 for Carmelite nuns, should involve different patterns of cerebral activity than spontaneous occurrences. The difference would be comparable to the patterns of activation during the experience of a novel stimulus compared to the later reconstruction (memory) of that experience.31 There is now epidemiological evidence that the experiences of sensed presence may be facilitated by geophysical transients32 within many environments; these patterns and hallucinations of "presences" can be reproduced directly within the laboratory by stimulating the right temporoparietal lobes.33

One of the limitations of this study was the small sample size. However, the experiences of these participants were similar to those reported in several other experiments,4,20 some of which were completed under double-blind conditions. In addition, the strongest effect sizes for the association between these experiences and EEG activity involved regions similar to those reported by other experimenters employing different measurements. Although EEG measurements are considered not useful for localization by some researchers, this issue may be less critical if experiences are correlated with electromagnetic matrices involving large cerebral volumes.

This research was supported by Grant 71/00 from the BIAL Fellowship Program (Portugal).