Content uploaded by Hans Herzog
Author content
All content in this area was uploaded by Hans Herzog on Mar 11, 2019
Content may be subject to copyright.
SHORT COMMUNICATION
Neural correlates of religious experience
Nina P. Azari,
1,2
Janpeter Nickel,
1
Gilbert Wunderlich,
1,3
Michael Niedeggen,
4
Harald Hefter,
1
Lutz Tellmann,
3
Hans Herzog,
3
Petra Stoerig,
4
Dieter Birnbacher
5
and Ru
Èdiger J. Seitz
1
1
Department of Neurology, University Hospital Du
Èsseldorf, Germany
2
The Iliff School of Theology/University of Denver, 2201 S, University Blvd, Denver, Colorado 80210, USA
3
Institute of Medicine, Research Center, D-52407 Ju
Èlich, Germany
4
Department of Physiological Psychology, Heinrich-Heine-University, D-40225 Du
Èsseldorf, Germany
5
Philosophical Institute, Heinrich-Heine-University, D-40225 Du
Èsseldorf, Germany
Keywords: brain, human, neuroimaging, PET, religion
Abstract
The commonsense view of religious experience is that it is a preconceptual, immediate affective event. Work in philosophy and
psychology, however, suggest that religious experience is an attributional cognitive phenomenon. Here the neural correlates of a
religious experience are investigated using functional neuroimaging. During religious recitation, self-identi®ed religious subjects
activated a frontal±parietal circuit, composed of the dorsolateral prefrontal, dorsomedial frontal and medial parietal cortex. Prior
studies indicate that these areas play a profound role in sustaining re¯exive evaluation of thought. Thus, religious experience
may be a cognitive process which, nonetheless, feels immediate.
Introduction
For more than a century, the nature of religious experience has been a
topic of considerable scholarship and debate, yet virtually nothing is
known about its biological foundations. The commonsense view of
religious experience is that it is a preconceptual, immediate affective
event (James, 1902; Schleiermacher, 1958. It should be noted that
James in The Principles of Psychology (1890) was reluctant to
contrast thought and feeling). Work in the philosophy of mind and
psychology of religion have challenged this view, suggesting that
religious experience may be a cognitive attributional phenomenon
(Proudfoot & Shaver, 1975; Shaver, 1975), constituted not simply by
the felt immediacy, but by a causal claim regarding a religious source
for the personal experience. Therein resides the authority and `force'
of the experience (Proudfoot, 1985), rendering religious experience a
unique and persistent personal event.
We studied a group of self-identi®ed religious subjects, who
attributed their religious experience to a biblical psalm, in order to
explore for the ®rst time using functional neuroimaging the brain
areas involved in religious experience. While the view that religious
experience is a preconceptual feeling would predict the activation of
limbic brain areas engaged by emotion, attribution theory would
predict brain areas mediating reasoning to be activated. We show
that, during religious recitation, religious subjects activated a frontal-
parietal circuit composed of the dorsolateral prefrontal, dorsomedial
frontal and medial parietal cortex, suggesting that religious experi-
ence may be a cognitive process.
Methods
Subjects
Twelve right-handed, healthy, native-German-speaking adults par-
ticipated in this study. All had given informed consent to their
participation. Six were self-identi®ed as religious (2F, 4M; age,
31 63 years), and six were self-identi®ed as non-religious (2F, 4M;
age 26 63 years). Values are given as mean 6SD throughout. The
two groups of subjects were matched on education level (2±5 years
university). The religious subjects were members of a `Free
Evangelical Fundamentalist Community' in Germany, all of whom
reported having had a documented conversion experience, and
interpreted biblical text literally as the word of God. For the
conversion experience and, hence, for their personal religious
experience, the ®rst verse of biblical Psalm 23 was essential. The
religious subjects were teachers at a private secondary school (run by
their same faith community), having been selected for their posts on
the basis of rigorous faith-based criteria. In this regard, they were a
religiously homogeneous group, having religious experience in a
personal reproducible manner which could be standardized for the
purpose of this study. The non-religious subjects were students at the
University of Du
Èsseldorf, who were studying various subjects in the
natural sciences. Interaction with the subjects during all study phases
was conducted in German by native German speakers. The religious
and non-religious subjects did not differ signi®cantly on imaginability
and verbal traits (Imaginability: religious subjects, 27 67; non-
religious subjects, 25 67, group differences P> 0.5; Verbal:
religious subjects, 27 68; non-religious subjects, 26 66, group
differences P> 0.6) (Paivio & Harshman, 1983; this questionaire was
translated into German for the purposes of this study), nor did they
differ signi®cantly on personality or life satisfaction measures
(P> 0.1, groups different) (Fahrenberg et al., 1984; Eysenck &
Eysenck (1975).
Correspondence: Dr Nina P. Azari,
2
The Iliff School of Theology, as above.
E-mail: nazari@du.edu
Received 12 December 2000, revised 8 February 2001, accepted 13 February
2001
European Journal of Neuroscience, Vol. 13, pp. 1649±1652, 2001 ãFederation of European Neuroscience Societies
Stimuli
The texts used for the different tasks conditions were: `religious': the
®rst verse of biblical Psalm 23 (Die Bibel, 1975); `happy': a well-
known German children's nursery rhyme (Lange & Stephan, 1984);
and `neutral': instructions on using a phone card from the Du
Èsseldorf
telephone book (Deutsche Telekom, 1996). The texts for all stimuli
were matched on length, and the rhyme was specially chosen because
it was familiar to all subjects and was not associated with music. All
subjects were able to recite from memory both the psalm and rhyme
at the time of the PET session. According to their responses in
prestudy interviews, the religious subjects regarded the induction of
repeated, transient religious states in a single scanning session as
antithetical to religious experience (and disrespectful to their faith).
Thus, they found it acceptable (and were asked) to induce in
themselves and then sustain for the duration of a given scanning
session a unique religious state.
PET scanning
Positron-emission tomography (PET) imaging was approved by the
Ethics Committee of Heinrich-Heine-University Du
Èsseldorf
(Application #1301). The PET-regional cerebral blood ¯ow (rCBF)
data were acquired as described elsewhere (Herzog et al., 1996;
Duncan et al., 2000). For this study we used a 24-ring ECAT
EXACT-HR camera (resolution 4 mm in-plane, 9 mm FWHM, slice
distance 2.4 mm). The session for each subject included six scans
separated by 10 min. For each scan, 555 MBq of
15
O butanol were
injected into the right brachial vein, ¯ushed with saline. PET
scanning began at the moment of the injection and lasted for 60 s.
The 40 s of dynamically recorded head uptake were used for
calculation of the rCBF data. PET image slices were reconstructed
using a Hanning ®lter to an effective image resolution of 9 mm
(FWHM) with a slice distance of 2.425 mm. Image analysis was done
using SPM97d (http://www.®l.ion.ucl.ac.uk/spm).
Subjects were PET-scanned in six conditions, inclusive of `reli-
gious' and appropriate control conditions: reading silently or reciting
(eyes covered) biblical Psalm 23 (1, `religious-read'; 2, `religious-
recite'); reading silently or reciting (eyes covered) the children's
nursery rhyme (3, `happy-read'; 4, `happy-recite'); reading silently the
set of instructions (5, `neutral-read'); and while lying quietly (eyes
covered) (6, `rest'). In the recite conditions (eyes covered), subjects
started the task at injection time. In the read conditions, presentation of
the stimuli were on a ¯at-screen monitor; horizontal display extent was
»12 degrees. For the 30-s preinjection engagement period of the task,
the ®rst half of the text was displayed on the screen and the subject read
the text aloud. At the time of the injection and the start of the scanning,
the second half of the text was displayed on the screen and the subject
read the text silently. These conditions ensured that the subject was
fully engaged in the task by the start of the scan, and that all subjects
were engaged in the same stimulus material during a given task
condition. Task order was pseudo-randomized (`rest' was always ®rst,
and for the religious and happy conditions, `read' was always before
`recite'). The felt-quality of each target state was assessed with the
Positive Affect Negative Affect Scale (PANAS) (Watson et al., 1988)
before and after each PET scan.
Results
During recitation of this psalm, the religious subjects successfully
attained the religious state (Table 1). And, as predicted from previous
TABLE 1. Self-assessment ratings for the target state in religious and non-religious subjects after the PET session
Happy-
read
Happy-
recite
Religious-
read
Religious-
recite
Neutral-
read
Religious subjects
Achieve 5.8 61.9 5.7 61.4 6.3 61.4 8.5 60.8* ² 8.17 62.6*
Sustain 6.7 61.8 6.2 61.2 6.7 61.9 8.3 61.6* ² 8.17 62.6*
Non-religious subjects
Achieve 6.8 61.3 8.7 60.8* ² 4.2 62.2 5.3 62.6 9.0 60.6*
Sustain 6.8 61.2 8.2 61.2* ² 4.8 62.5 5.0 63.0 8.8 61.0*
Values are mean 6SD for the target state (scale of 1±10, max is 10) in religious and non-religious subjects. *P< 0.05, paired t-test, different from other task
conditions;
²
P< 0.05, two-tailed, independent t-tests, groups different. Note that only the religious subjects achieved the religious state, and only the non-religious
subjects achieved the happy state; both groups achieved the neutral state.
FIG. 1. Signi®cant activations for the contrast `religious-recite' vs. `rest' in
religious subjects, rendered onto canonical T1-weighted image of SPM97d
(P< 0.001, uncorrected for multiple comparisons) (see also Table 2).
Shown are the left, dorsal and right view of the brain. Scans for each
subject were realigned and spatially normalized onto the PET template, and
smoothed using an isotropic Gaussian kernel with FWHM set at 20 mm.
The SPM grey matter threshold was set to its default value. For task
comparisons, an ANCOVA (analysis of covariance) model was ®tted to the
data for each voxel.
1650 N. P. Azari et al.
ã2001 Federation of European Neuroscience Societies, European Journal of Neuroscience, 13, 1649±1652
studies in psychology of religion, demonstrating that religious
attributions are commonly directed toward countering or improving
negative or `crisis' situations (Spilka et al., 1992), the religious
subjects showed a trend toward decreased negative affect during the
religious state [PANAS scores for religious subjects in the `religious-
recite' condition: negative postscan (14.0 64.5) < negative prescan
(15.8 66.1), P< 0.09. No other task condition showed changed
PANAS values in these subjects, and the non-religious subjects
showed no changes in PANAS values (P> 0.5)]. Note that religious
experience must be described from the perspective of the subject
(Proudfoot, 1985), because the subject's interpretation thereof is
central to (constitutive of) the experience. Moreover, there is no
acceptable, subject-independent de®nition of religion or religious
(Sharpe, 1983). Thus, attainment of a target state can be determined
solely on the basis of subjective self-assessment. The subject
achieved and sustained the target state if and only if the subject
reported that this was the case (and not, if the reverse).
The PET images showed a speci®c, signi®cant, activation of the
right dorsolateral prefrontal cortex in the religious subjects during the
religious state as compared with non-religious subjects. Also as
compared to a happy state and a neutral read condition these areas
were observed (Fig. 1, Table 2). During the religious state, the
religious subjects showed few additional activations, including the
dorsomedial frontal cortex and the right precuneus. These activations
involved areas that are engaged in cognitive processes, while limbic
areas (including the amygdala and orbitofrontal cortex) activated by
emotional stimuli (Morris et al., 1996; Paradiso et al., 1997) did not
show rCBF changes. As expected, however, the non-religious
subjects showed a speci®c left amygdala activation during the
happy state (x= ±20, y= ±8, z = ±22; P< 0.001). Therefore, our
results strongly support the view that religious experience is a
cognitive attributional phenomenon.
Discussion
Recent neuropsychological and functional imaging studies suggest
that the prefrontal cortex holds representations of knowledge
structured in the form of cognitive schemas (Partiot et al., 1995).
Cognitive schemas are mental representations containing organized
prior knowledge about speci®c domains, inclusive of speci®cations of
the causal relations among the attributes therein (Taylor & Crocker,
1981). Religious attributions are made in accordance with religious
schemas, which consist in organized knowledge about religion and
religious issues, and include reinforced structures for inferring
religiously related causality of experienced events (Spilka &
McIntosh, 1995). Prefrontally localized schemas manage such
memories stored as representations in posterior brain areas
(Grafman & Hendler, 1991), and the dorsolateral prefrontal cortex
is critical for memory retrieval and conscious monitoring of thought
(Fletcher et al., 1998; McIntosh et al., 1999; Duncan et al., 2000;
Gallagher et al., 2000). The medial parietal area activated in the
religious subjects was the precuneus, which has strong anatomical
connections to prefrontal cortex (Goldman-Rakic, 1988; Petrides &
Pandya, 1984), and plays a key role in visual memory (Fletcher et al.,
1995). The dorsomedial frontal area we observed corresponds to the
presupplementary motor area (pre-SMA) (Rizzolatti et al., 1998).
Pre-SMA receives strong inputs from the prefrontal cortex, as well as
from the cerebellum (Luppino et al., 1993), the latter of which now is
understood to participate in a wide variety of both cognitive and
noncognitive processes (Allen et al., 1997). Converging evidence
from both animal and human studies indicates that pre-SMA is
important in automatically controlling the readiness for action
processed by lateral parieto-frontal circuits (Deiber et al., 1991;
Tanji et al., 1996), and sustaining the preparation to act based upon
the current contents of working memory (Jahanshahi et al., 1995;
Haxby et al., 2000).
Thus, we suggest that religious experience may be a cognitive
process, mediated by a pre-established neural circuit, involving
dorsolateral prefrontal, dorsomedial frontal and medial parietal
cortex. Because religious attributions are made in anomalous or
ambiguous situations, when a person does not know what to expect or
what to do, yet actively and persistently seeks a solution, a persistent,
internally generated `readiness' emerges, which subsequently serves
to re-activate the religious schema in the presence of salient religious
cues (Proudfoot, 1985). For the religious subjects of this study,
biblical Psalm 23 was the salient religious cue. There is substantial
evidence from the psychology of religion to suggest that people are
`prepared' for religious experiences (Spilka et al., 1996). This
`readiness' is probably mediated by the dorsomedial frontal cortex,
leading to the commonly reported felt immediacy of religious
experience. The experience, however, becomes religious when the
subject has consciously identi®ed it as consistent with the subject's
own religious schema (Proudfoot, 1985). This cognitive process most
probably involves the dorsolateral prefrontal and medial parietal
cortex (Fletcher et al., 1995; Partiot et al., 1995). Consequently,
subsequent self-initiated religious actions will reinforce the subject's
personal religious schema (Proudfoot, 1985; Spilka & McIntosh,
1995).
Notably, the religious experience studied here did not involve
activation of limbic areas, although the religious subjects were
TABLE 2. Peak activations for `religious-recite' in religious subjects
Brain region
Coordinates Z-scores
xyzVs. `rest' Vs. non-religious Vs. `happy-recite'
R dorsolateral prefrontal (BA 9) 24 50 34 3.74** 3.29** 3.78**
L dorsolateral prefrontal (BA 9) ±28 36 32 3.62** ± ±
dorsomedial frontal (pre-SMA) (BA 6) 7 10 68 3.58** 2.74* 3.32**
R medial parietal (precuneus) (BA 7) 14 ±82 46 3.57** ± 3.32*
L cerebellum ±2 ±48 ±20 3.31** ± 2.99*
The Z-scores are shown for the comparisons `religious-recite' vs. `rest' in religious subjects (vs. `rest'); religious vs. non-religious subjects in `religious-recite'
(vs. non-religious); and `religious-recite' in religious subjects vs. `happy-recite' in non-religious subjects, i.e. group3task interaction effect, (vs. `happy-recite') at
**P< 0.001, *P< 0.01, uncorrected. Coordinates (x,y,zin space of Montreal Neurological Institute (MNI) template) and selection of foci according to
conventions of SPM97d. Brain regions (approximate Brodmann areas ± BAs) estimated from Talairach & Tournoux, 1988), after adjustment as appropriate for
differences between MNI and Talairach coordinates (http://www.mrc cbu.cam.ac.uk/imaging/mnispace.html).
Brain acitivity and religious experience 1651
ã2001 Federation of European Neuroscience Societies, European Journal of Neuroscience, 13, 1649±1652
positively affected by the recital of Psalm 23. Thus, the religious
experience was not an emotional experience, nor an arousal
comparable to that of the happy state which we observed in the
non-religious subjects. Admittedly we used highly standardized
stimuli to identify a neurobiological correlate of the concept of
`religious experience' in this initial study. A challenge for future
work will be to explore transient religious states and the evolution of
other varieties of religious experience. It may turn out that `religious
experience' can be divided eventually into a variety of subprocesses,
as has been, for example, the case with memory.
Acknowledgements
Supported by grants to N.P. Azari from the Alexander von Humboldt
Foundation and the Deutscher Akademischer Austausch Dienst (DAAD), and
SFB 194 of the Deutsche Forschungsgemeinschaft to R.J.S. and H.H.
Abbreviations
PANAS, Positive Affect Negative Affect Scale; PET, positron emission
tomography; pre-SMA, presupplementary motor area; rCBF, regional cerebral
blood ¯ow.
References
Allen, G., Buxton, R.B., Wong, E.C. & Courchesne, E. (1997) Attentional
activation of the cerebellum independent of motor movement. Science,275,
1940±1943.
Deiber, M., ±P., Passingham, R.E., Colebatch, J.G., Friston, K.J., Nixon, P.D.
& Frackowiak, R.S.J. (1991) Cortical areas and the selection of movement:
a study with positron emission tomography. Exp. Brain Res.,84, 393±402.
Deutsche Telekom, A.G. (1996) Telefonbuch (1996/97) fu
Èr Du
Èsseldorf.
Dienstleistungen rund ums Telefon: T-Card. Deutsche Telekom AG,
Germany, p. 3.
Die Bibel (1975) Der gute Hirte, revised text [oder die ganze Heilige Schrift
des alten und neuen Testaments nach der U
Èbersetzung Martin Luthers].
Deutsche Bibelstiftung, Stuttgart, pp. 581±519.
Duncan, J., Seitz, R.J., Kolodny, J., Bor, D., Herzog, H., Ahmed, A., Newell,
F.N. & Emslie, H. (2000) A neural basis for general intelligence. Science,
289, 457±460.
Eysenck, H.J. & Eysenck, S.B.G. (1975) Manual of the Eysenck Personality
Questionaire. [Hodder & Stoughton, London. (EPQ-R, German translation:
Ruch, W. (1999).] Z Differentielle Diagnostische Psychologie,20, 1±24.
Fahrenberg, J., Hampel, R. & Selg, H. (1984). Das Freiburger
Persoenlichkeitsinventar FPI, 4 Rev. Au¯. Hogrefe, Go
Èttingen.
Fletcher, P.C., Frith, C.D., Baker, S.C., Shallice, T., Frackowiak, R.S.J. &
Dolan, R.J. (1995) The mind's eye ± Precuneus activation in memory-
related imagery. Neuroimage,2, 195±200.
Fletcher, P.C., Shallice, T., Frith, C.D., Frackowiak, R.S.J. & Dolan, R.J.
(1998) The functional roles of the prefrontal cortex in episodic memory. II.
Retrieval. Brain,121, 1249±1256.
Gallagher, H.L., Happe, F., Brunswick, N., Fletcher, P.C. & Frith, C.D. (2000)
Reading the mind in cartoons and stories: an fMRI study of `theory of mind'
in verbal and non-verbal tasks. Neuropsychologia,38, 11±21.
Goldman-Rakic, P.S. (1988) Topography of cognition: parallel distributed
networks in primate association cortex. Annu. Rev. Neurosci.,11, 137±156.
Grafman, J. & Hendler, J. (1991) Planning and the brain. Brain Behav. Sci.,
14, 563±564.
Haxby, J.V., Petit, L., Ungerleider, L.G. & Courtney, S.M. (2000)
Distinguishing the functional roles of multiple regions on distributed
neural systems for visual working memory. Neuroimage,11, 380±391.
Herzog, H., Seitz, R.J., Tellmann, L. & Mu
Èller-Ga
Èrtner, H.-W. (1996)
Quantitation of regional cerebral blood ¯ow using an autoradiographic-
dynamic approach in positron emission tomography. J. Cereb. Blood Flow
Metab.,16, 645±649.
Jahanshahi, M., Jenkins, I.H., Brown, R.G., Marsden, D., Passingham, R.E. &
Brooks, D.J. (1995) Distinguishing the functional roles of multiple regions
on distributed neural systems for visual working memory. Brain,118,
913±933.
James, W. (1902) Varieties of Religious Experience. Longs, Greens, New
York.
Lange, S. & Stephan, W. (1984) Das ist der Daumen. In Lange, S. & Stephan,
W. (eds), Das Grosse Buch der Kinderreime. Diogenes-Verlag, Zu
Èrich,
p. 38.
Luppino, G., Matelli, M., Camarda, R. & Rizzolatti, G. (1993) Corticocortical
connections of area F3 (SMA-proper) and area F6 (pre-SMA) in the
macaque monkey. J. Comp. Neurol.,338, 114±140.
McIntosh, A.R., Rajah, M.N. & Lobaugh, N.J. (1999) Interactions of
prefrontal cortex in relation to awareness in sensory learning. Science,
284, 1531±1533.
Morris, J.S., Frith, C.D., Perrett, D.I., Rowland, D., Young, A.W., Calder, A.J.
& Dolan, R.J. (1996) A differential neural response to fearful and happy
facial expressions. Nature,383, 812±815.
Paivio, A. & Harshman, R. (1983) Factor analysis of a questionaire on imagery
and verbal habits and skill. Can. J. Psychol,37, 461±483.
Paradiso, S., Robinson, R.G., Andreasen, N.C., Downhill, J.E., Davidson, R.J.,
Kirchner, P.T., Watkins, L., Ponto, L. & Hichwa, R.D. (1997) Emotional
activation of limbic circuitry in elderly normal subjects in a PET study. Am.
J. Psychiatry,154, 384±389.
Partiot, A., Grafman, J., Sadato, N., Wachs, J. & Hallett, M. (1995) Brain
activation during the generation of non-emotional and emotional plans.
Neuroreport,6, 1269±1272.
Petrides, M. & Pandya, D.N. (1984) Projections to the frontal cortex from the
posterior parietal region in the rhesus monkey. J. Comp Neurol.,228,
105±116.
Proudfoot, W. (1985). Religious Experience. University of California Press,
Berkeley.
Proudfoot, W. & Shaver, K.G. (1975) Attribution theory and the psychology of
religion. J. Scienti®c Study Religion,14, 317±330.
Rizzolatti, G., Luppino, G. & Matelli, M. (1998) The organization of the
cortical motor system: new concepts. Electroencephalogr. Clin.
Neurophysiol.,106, 283±296.
Schleiermacher, F.D.E. (1958). On Religion: Speeches to its Cultured
Despisers. [translation Oman, J.]. Harper & Row, New York.
Sharpe, E.J. (1983). Understanding Religion. St. Martin's Press, Inc., New
York, pp. 33±48.
Shaver, K.G. (1975) An Introduction to Attribution Theory. Winthrop,
Cambridge, MA.
Spilka, B., Brown, G.A. & Cassidy, S.A. (1992) The structure of religious
mystical experience in relation to pre- and post-experience lifestyles. Int. J.
Psychol. Religion,2, 241±257.
Spilka, B., Ladd, K.L., McIntosh, D.N. & Milmoe, S. (1996) The content of
religious experience: The roles of expectancy and desirability. Int. J.
Psychol. Religion,6, 95±105.
Spilka, B. & McIntosh, D.N. (1995) Attribution Theory and Religious
Experience. In Hood, R.W. (ed.), Handbook of Religious Experience.
Religious Education Press, Birmingham, Alabama, pp. 421±445.
Talairach, J. & Tournoux, P. (1988). Co-Planar Atlas of the Human Brain.
Thieme, Stuttgart.
Tanji, J., Shima, K. & Mushiake, H. (1996) Multiple cortical motor areas and
temporal sequencing of movements. Cogn. Brain Res.,5, 117±122.
Taylor, S.E. & Crocker, J. (1981) Schematic bases of social processing. In
Higgins, E.T., Herman, C.P. & Zanna, M.P. (eds), Social Cognition.
Erlbaum, Hillsdale, NJ.
Watson, D., Clark, L.A. & Tellegren, A. (1988) Development and validation
of brief measures of positive and negative affect: The PANAS scales. J.
Personality Social Psychol.,54, 1063±1070.
1652 N. P. Azari et al.
ã2001 Federation of European Neuroscience Societies, European Journal of Neuroscience, 13, 1649±1652
