"All Things Are Possible": Towards A Sociological Explanation of Pentecostal Miracles and Healings

Sociology of Religion 2011, 72:4 456-482
doi:10.1093/socrel/srr019
Advance Access Publication 20 April 2011
“All Things Are Possible”: Towards a

Sociological Explanation of Pentecostal
Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Miracles and Healings
Jörg Stolz*†
Observatoire des religions en Suisse, University of Lausanne
Pentecostal miracles and healings have often been described and interpreted, but rarely explained in
their sociological workings. As former research implies, actual biomedical effects of Pentecostal
healings are possible (the so-called placebo effect), but quite limited. In Pentecostal healing services,
however, very impressive miracles and healings are routinely produced: paralytics arise from wheel-
chairs, cancerous ulcers disappear, legs grow, cavities are mysteriously filled, and the deaf suddenly
hear. Drawing on a case study and qualitative interviews, this paper offers a sociological, mecha-
nism-based, explanatory scheme for the observed phenomena. It is argued that a number of “social
techniques” (e.g., suggestion, rhythm, music), context factors (e.g., audience size and beliefs),
and causal mechanisms (e.g., probability, latency, selection, and editing effects) are combined in
an ingenious way in order to produce miracles and healings.
Key words: healing; Pentecostalism; charismatic movement; anthropology of religion; rational
choice theory; theory.
Pentecostalism is one of the fastest growing religions worldwide (Burgess

2006:xiii) and one of its main attractions is miracles and healings. In
Pentecostal healing services, paralytics arise from wheelchairs, stiff knees
become flexible, cancerous ulcers disappear, and headaches vanish. Speakers
receive “words of knowledge” and can foretell healings that are taking place in
*Direct correspondence to Jörg Stolz, University of Lausanne, Observatoire des religions en

Suisse, Bâtiment Vidy, CH, 1015 Lausanne, Switzerland. E-mail: [email protected].
†
I thank Barbara Dellwo, Anaı̈d Lindemann, and Edmée Ballif for their help with data collection
and important discussions. I thank Régis Déricquebourg, Olivier Favre, Nicole Durisch Gauthier,
Philippe Gilbert, Denise Hafner Stolz, Christophe Monnot, Roberto Motta, Detlef Pollack, Jeanne
Rey, Ingrid Storm, David Voas, and three anonymous reviewers for their valuable comments.
I thank Christine Rhone for correcting the English version of the paper. The usual disclaimers apply.
# The Author 2011. Published by Oxford University Press on behalf of the Association
for the Sociology of Religion. All rights reserved. For permissions, please e-mail:
[email protected].
456
SOCIOLOGICAL EXPLANATION OF PENTECOSTAL MIRACLES AND HEALINGS 457
the audience. The atmosphere is one of miracles and healings, a time in
which—as the speaker tells us—“all things are possible.” Anthropologists have
often taken a positive view of alternative healing and have created a rich,
mainly descriptive and interpretive literature. However, they have not yet

explained with what “social techniques” Pentecostals routinely produce impres-
sive miracles and healings. Critical scholars and doctors, on the other hand,
have mainly taken a negative view of Pentecostal healings. While they have
pointed to some techniques used, they have great trouble explaining why
Pentecostal healing workshops exist in the first place, given that the biomedi-
cal effects are slight or nonexistent.
What seem to be missing from the literature are attempts by explanatory,
nonnormative sociology. This article therefore asks, first: by what social techni-
ques are miracles and healings in Pentecostal healing workshops produced?
Second: why do these workshops continue to persist, even though people
might learn over time that these healings and miracles are often greatly exag-
gerated? In asking these questions, I try to abstain from value judgments and to
argue neither for nor against Pentecostal healing. Assuming a special version of
“methodological atheism,” I do not wish to argue that “miracles are impossi-
ble,” but try—as I would with any other social phenomenon—to explain as
much as possible with sociological tools.
To exemplify my claims, I draw on a case study of a healing workshop that
took place during an international healing conference in French-speaking
Switzerland as well as on in-depth interviews with participants and the
speaker. I argue that a number of “social techniques” (e.g., suggestion, rhythm,
music), context factors (e.g., audience size and beliefs), and causal mechanisms
(e.g., probability, latency, selection, and editing effects) are combined in order
to produce miracles and healings. I show that the model explains a variety of
Pentecostal healing phenomena.
ACCOUNTING FOR MIRACULOUS PREDICTIONS

AND SUPERNATURAL HEALINGS
Pentecostal Approach
Before going into social scientific explanations, we should turn to the way
Pentecostals account for miracles and healings. Healings and miracles are
highly valued in Pentecostalism and the charismatic movement (Anderson
2002). They belong to the various “gifts” that are given by God to the faithful.
In the Pentecostal view, both miraculous “words of knowledge” (supernatural
knowledge of facts one could not have known about) and divine healing are
undeniable facts. Not only are they rooted in scripture, but also they are visibly
occurring on a regular basis both in healing services and in everyday life.
In the Pentecostal view, illnesses of all kinds are not God’s work but due
to “evil forces,” “sin,” or “the devil.” Divine healing is made possible because
458 SOCIOLOGY OF RELIGION
Christ has already suffered for humanity’s sins on the cross. It can heal any
illness whatsoever, be it a small ailment (e.g., a headache), a mental problem
(such as depression), or very serious physical maladies (e.g., cancer, AIDS). It
follows that healing is never just physical, but by definition physical, mental,

and spiritual simultaneously. Furthermore, healing is always the work of God
and the Holy Spirit, never of the person praying for healing (Warrington
2006:235). Although most Pentecostals nowadays do not reject biomedicine,
they see divine healings as clearly more pure and powerful (Andrews 2003:17).
Having looked at the Pentecostal explanation, what does social scientific
research make of miracles and healings?
Anthropological Approach
In the social sciences, it is especially anthropologists1 and, sometimes,
qualitative sociologists who have researched Pentecostal and related healing
phenomena. Their approach has mostly been descriptive and interpretive, thus
often refraining from causal explanations. Anthropologists readily admit that
Pentecostal (or other “alternative”) healing often does not have great “physi-
cal” effects measured by biomedical standards (Csordas 1988:136).
However, alternative healing may cure their “selves.” This is accomplished
through reinterpreting the patients’ illness in the framework of a given symbol
system (“symbolic manipulation”) (Csordas 2002). By conducting a ritual, alter-
native healers convince the patient of a change of meaning of his illness. This
may have different beneficial effects. First, meaningless pain is transformed into
a manageable burden (Csordas 1988:122; Geertz 1993:105). Second, patients
may be empowered. Before, they might have felt helpless; now they feel able to
transform and overcome their malady (McGuire 1987:366). Third, patients may
be led to change their habitus and lifestyle (e.g., stop drinking) (Laurent
2001:333). Fourth, patients may be integrated into the religious group, provid-
ing them with new social capital. All of these results may have additional
(though often limited) psychophysical effects, reducing or even eliminating the
malady. The latter may especially be the case for problems that were partly or
wholly psychosomatic in the first place (Moerman 1979).
When critics say that healers produce “only a placebo effect,” these anthro-
pologists answer that, precisely, the placebo effect shows that humans are a
socio-psycho-physical entity in which the symbolic may have an effect on the
physical (Dow 1986:63; McGuire 1991:228; Moerman 1979:62).
Anthropologists have generally shown a rather positive attitude towards alter-
native healing practices and Pentecostal healing in particular. They insist on
the idea that the culturally “dominant” biomedicine should not be used to
1
I am aware that there are various types of anthropology concerned with healing. My
claim is only that some highly respected scholars in anthropology and qualitative sociology
hold the views on healing described in the paragraph.
judge “alternative” medicine. Rather, biomedical and alternative healing
systems are incompatible paradigms, which should both have the right to exist
(Moerman 1979:60; Rossi 2009). For example, McGuire (1991:5) writes:

A . . . notion that must . . . be set aside is the assumption that the medical reality, as promul-
gated by the dominant health specialists in this culture, is necessarily the “true” reality. From
a sociological perspective, this medical definition of reality must be seen as one among many
competing conceptions of illness, its causes, and treatment. Medical reality, too, is socially
constructed.
While the anthropological approach has produced many important insights,

anthropologists—due to their focus on descriptive and interpretive methods
and legitimizing impetus—have missed asking some crucial explanatory ques-
tions. Specifically, they have not attempted to show which “social techniques”
Pentecostal healers and their audiences employ to achieve the social reality of
miracles and healings.
Critical Approach
Turning to the authors I have loosely grouped into a “critical approach,” it
is as if we entered a different world altogether. These writers from various disci-
plines have mainly asked if Pentecostal healing has “real” effects according to
biomedical standards. Their results are generally negative.2 A first group tries
to follow up individual cases of alleged healings (Nolen 1975; Randi 1989;
Rose 1971; Thomas 1999). Thus, the professional magician Randi (1989:287)
writes:
My standards are simple. I need a case that involves a living person, healed of an otherwise
non-self-terminating disease, who recovered from that disease as a result of a faith healer’s
actions and can produce before-and-after evidence to establish that fact. I have failed in any
and all cases I have investigated to obtain a response that satisfies these simple requirements.
Some of these authors show what “techniques” healers may use in order to
make miracles and healings appear real. How is a leg made to grow? The healer
just manipulated the patient’s shoe to make it look as if there was growth.
A blind person suddenly sees and can tell the number of fingers the healer is
holding before his eyes? The patient was not completely blind before. A person
in a wheelchair gets up and is able to walk? The person was able to walk
before; he was told to sit in the wheelchair brought by the healer. How does
the healer know the names and illnesses of individuals in the audience? Before
the show, the healer’s assistants gathered the information, and the healer mem-
orized the information using mnemonics or else receives it during the show
2
The publications assembled in this paragraph are very diverse. The case-oriented
studies often have a “debunking” character.
using electronic devices. So much, then, for the follow-up approach and its
results.
A second method used to establish the effectiveness of alternative healing
is the scientific experiment. In the past years, an increasing number of high-

quality, randomized experiments on the subject of distant and faith healing
have been conducted. In an update of a systematic review, Ernst (2003) con-
cluded that recent rigorous studies “shift the weight of the evidence against the
notion that distant healing is more than placebo.”
Most of the writers I have placed in the “critical” group see Pentecostal
healing quite negatively. James Randi is the most infuriated of them all. He
finds faith healing “disturbing,” “ridiculous,” and “silly.” Often, he is bewildered
not just by the boldness of the healers but also by the seeming irrationality of
the believers (Randi 1989:109). In fact, all of these writers see Pentecostal
healing not only as useless but also as harmful for the individuals engaging in
it.
While critical authors have brought some insights to the understanding of
Pentecostal healing, their negative value judgments have prevented them from
giving satisfactory answers to important questions. Thus, they have been
unable to explain just why many Pentecostals persist in believing in their
healing even in the face of seemingly contradictory evidence or why they
actually help the healers to perform their “social techniques.” As a result, they
also fail to understand why Pentecostal healing may be successful, as a social
enterprise, in the long term.
Explanatory Sociology Approach

In this article, I want to present yet another approach: explanatory sociol-
ogy. While this type of sociology remains still untried in the field of
Pentecostal miracles and healings, I argue that it can make a valuable contribu-
tion especially in those areas where former anthropological or critical
approaches tend to be weak.
Following the general tenets of the approach, explanatory sociology tries to
take a neutral stance regarding Pentecostal miracles and healings. These phe-
nomena are neither seen positively (as by some anthropologists), nor rejected
(as by some of the critical scholars). While neutrality is more an ideal than an
attribute of concrete research, explanatory sociology may nevertheless approach
the ideal by disregarding whether these practices are “good” or “bad” and
attempting only to explain what actually happens.
According to explanatory sociology, a phenomenon is explained if one can
show how it results from a set of initial conditions and a generative (and there-
fore causal) mechanism (Hedström 2005). More specifically, a sociological
explanation is given if we can show how an initial situation (macro) leads indi-
viduals to react to this situation (micro) and how, through aggregation, these
individual reactions combine to form a new social outcome (macro) (Coleman
1990).
In explanatory sociology of religion, it is important to assume “bounded
rationality.” We take it as a given that individuals have subjectively “good
reasons” for their beliefs and actions. However, we also assume that their ways
of creating beliefs may fail due to erroneous inferences, conclusions based on

unwarranted assumptions, or motivationally or socially biased evaluations
(Boudon 1997; Gilovich 1991; Stolz 2009).
From this perspective, the healing service is a special case of an “interac-
tion system.”3 Due to the causal mechanisms at work in the interaction system,
actors are able to produce miracles and healings as emergent “salvation goods”
(Stolz 2006). In what follows, we specify the elements and mechanisms of this
interaction system.
METHOD
This is a case study of an AIMG (Association Internationale des

Ministères de Guérisons) healing workshop with the speaker Ian Andrews in
2003 in Oron-la-ville, Switzerland. Case studies lend themselves well to an
explanatory sociology framework, since they allow for direct observation of
causal mechanisms (Miles and Huberman 1994).
Video and Transcription Analysis

My first main data source is a DVD of the AIMG workshop with accompa-
nying transcript.4 The DVD, produced by AIMG, shows 52 minutes of the
(almost) uninterrupted workshop. The use of a DVD and transcript has distinc-
tive advantages compared with “plain” observation. Repeated viewing leads to
the analysis of audible and visible details, of the distinct temporal structure of
the situation, and of the complete interaction as far as it appears on tape.
Furthermore, video data are more reliable than field notes in that they may be
used by other researchers (Knoblauch 2009:73).
On the downside, video data may lead researchers to the false belief of
having “everything,” when in fact they only have what is on tape. The point of
view of the actors (cognitions, values, preferences) as well as structures and cul-
tural frames that transcend the situation may be blocked out. In our case, the
use of additional interviews, participant observation and documentary analysis
addresses this concern.
An additional validity issue arises since the religious group itself has pro-
duced the video. Presumably, AIMG could have edited the video in such a way
3
I define “interaction system” as the interrelation of individuals perceiving each other
as jointly present (compare to Esser 2000:273ff ).
4
I chose the Ian Andrews workshop, since it is typical of AIMG practices and is fully
taped. I have attended many workshops, but not this particular one. I prefer to use a case
where I have a tape available, rather than one that I have attended but not taped.
as to make the miracles and healings as spectacular as possible to attain their

purpose of evangelization. While it is evident that AIMG in fact tries to evan-
gelize, a comparison of what we see on tape and what I have observed in
various other AIMG workshops shows that there seems to be very little

“effect-enhancing editing.”5
In-depth Interviews and Participant Observation

My second data source consists of 23 in-depth interviews. One interview
was conducted with the speaker of the workshop, Ian Andrews. A second inter-
view was conducted with Werner Lehmann, one of the two leading figures of
AIMG. The other 21 interviews were with attendees of AIMG workshops,
meetings, and conferences. The interviews lasted from 60 to 90 minutes; an
interview guide was used with main questions and possible probes. Individuals
were given as much space as possible in order to develop their understanding
of the situation. The interviewees were chosen by contacting evangelical, char-
ismatic, and Pentecostal congregations and by asking if there were people in
the congregation who had attended healing workshops or conferences with
AIMG. Half of the congregations chosen were in close contact with AIMG
through “healing rooms”; the other half were chosen randomly from a compre-
hensive list of charismatic/Pentecostal congregations. We made a point of
finding both regular and occasional attendees.
The resulting sample consists of (excluding Ian Andrews and Werner
Lehmann) 14 women and 10 men.6 The age range is from 19 to 85; the rela-
tively high mean age (52) fits the mean age in AIMG meetings quite well.
The interviews were taped and transcribed. I do not claim any statistical repre-
sentativeness for this sample but am quite confident that I have captured most
of the typical ways of seeing and experiencing the AIMG healings workshops.
Data Analysis
Concerning the DVD and the accompanying transcript, I made a simulta-
neous analysis of the video tape and the transcription. Using methods close to
“grounded theory,” I have interpreted the images and text sequentially, coded
and written memos, and watched the DVD countless times (Knoblauch 2009;
Strauss and Corbin 1998). For this task, I used ATLASti software. Likewise,
the interview transcripts and the documents were sequentially interpreted and
coded, and insights introduced into memos. In general, I have worked induc-
tively, trying to let the concepts “emerge” from the analysis. Insights from the
video analysis and from interview analysis were “triangulated.”
5
Two noteworthy editing activities are an introduction and additional testimonies at
the end of the DVD.
6
In two cases, a spouse and in one case another church member were also present and
were interviewed, giving us 21 interviews but 24 participant interviewees.
Methodological Atheism and Its Critique
Since this paper provides an explanation of how miracles are (at least in
part) socially produced, it is necessary to clarify my stance on how scientific
explanations and religious truth claims relate to each other. The mainstream

position in sociology of religion seems to be some kind of “methodological
atheism” as defended by Berger (1990 (1967)) and Hamilton (2001:5).
Methodological atheism implies that sociologists will (1) refrain a priori from
assessing the (transcendent) truth claims of religions (on the side of the
“dependent variable”) and (2) refrain from using transcendent factors as
explanatory concepts (on the side of the “independent variable”).
In this paper, I use methodological atheism in the sense that I do not use
transcendent factors on the explanatory (independent) side and that I assume
that most of religion’s transcendent statements can never be assessed empiri-
cally and are therefore not amenable to sociological testing. There is one
caveat where I diverge from mainstream methodological atheism (Hamilton
2001). I will not use “a priori bracketing” when looking at the dependent varia-
ble. That is, I will not exclude assessment of apparently “transcendent phe-
nomena” a priori, but will try to push analysis with sociological tools as far as
possible.
Not accepting “a priori bracketing” is important for my paper in several
ways: first, it leads to my premise (based on former research) that the biomedi-
cal efficiency of divine healing is rather limited—it is in any case much
smaller than suggested in healing workshops. Second, it leads to my central
research question: if there are only very limited healings during the work-
shop—how are the healings then produced with social techniques and how can
this be successful in the long term? Third, only against the backdrop of this
premise may my findings be readily interpreted. For example, it would be diffi-
cult to understand why the speaker uses the “shotgun technique” if it were not
for a social production of miracles.
THE HEALING SERVICE IN CONTEXT
Organizational and Historical Context

The Association Internationale des Ministères de Guérison (AIMG)
[International Association of Healing Ministries] belongs, in a broad sense, to
the Pentecostal and charismatic movements. The history of charismatic and
Pentecostal “faith healers” goes back to the late nineteenth century (Curtis
2007; Opp 2005). Many of the techniques used since the beginning of these
movements have remained surprisingly stable.7 In the last decades, however,
7
The reason for this stability is probably that the social mechanisms involved are on
an interactional level and may function independently of larger sociohistoric changes.
there has been a shift from “big name healers” to a movement in which every
Christian is considered able to heal. Individuals with a healing ministry
increasingly see themselves as “equippers” who teach others how to pray for
the sick. AIMG and Ian Andrews are thinking along these lines.

Pentecostalism in Switzerland is a small minority making up no more than
1 percent of the population. Pentecostals are decidedly middle class (except for
immigrant Pentecostal communities) (Stolz and Favre 2005) and are—in con-
trast to stable evangelical and declining conservative and liberal churches—
growing. One interesting contextual factor is that both AIMG officials and par-
ticipants are acutely aware of the “competition” by biomedicine. Since
Switzerland is one of the richest countries in the world and has an excellent
health care system, divine healing is not as strong, they feel, as in poorer
countries.
While AIMG is clearly successful and growing, there is no reason to think
that it is the French-speaking Swiss who would be especially prone to react
positively to divine healing. The reason for the movement starting in Oron is
purely contingent. Individuals wanting to restart a healing movement in the
United States contacted Ian Andrews in the UK who suggested the idea to
Jean-Luc Trachsel, the current leader of the movement, in Switzerland. AIMG
is now spreading to France, Italy, and Belgium as well as to Benin.
The Healing Service

The healing service with Ian Andrews during the international healing
conference took place in 2003 (28 May–1 June) in Oron (canton of Vaud,
Switzerland). The conference attracted more than 1,500 persons. It included
11 larger events and 16 healing workshops (of which we analyze one in this
paper). The speakers came from Germany, Australia, Burkina Faso, Central
Africa, France, Switzerland, England, and the United States.
The healing service took place in a gymnasium. In front of a black cover
blocking the sunlight, there was a stage where a music group was standing. On
two sides, two gigantic screens retransmitted the events happening on stage.
There were about 350 persons, almost all of whom were standing, although the
room was equipped with red plastic chairs. The audience was wearing everyday
clothes: tee shirts, jeans, colored short-sleeved shirts. The music group was
playing gentle worship songs. Many participants raised their hands, some
closed their eyes, others gesticulated, and still others hugged. Helpers in front
of the stage received the persons wanting to testify, or prayed with individuals
wishing to do so.
The speaker, Ian Andrews, was 64 years old at the time of the healing
workshop. We see, under some gray hair, a worried face with a deep vertical
wrinkle at mid-forehead and eyeglasses with invisible frames. The interpreter,
next to him, was a woman of about 35 with discreet golden jewelry and semi-
short blond hair. Andrews spoke in English, in full sentences or fragments, and
the interpreter immediately translated his speech into French, adopting his
intonations and gestures.
Andrews started out slowly, alternating statements of salvation, instructions
to the participants, and theoretical statements. After some minutes, he had a

song sung. In a certain manner, Andrews was preparing the participants for the
coming of the Holy Spirit who very soon, according to him, was going to take
“healing actions.” During the meeting, Andrews began to introduce more and
more “words of knowledge” into his speech, that is, he transmitted “inspired”
sayings of the Holy Spirit about healings simultaneously occurring in the audi-
ence. Increasingly often, he invited the persons named in his words of knowl-
edge to come quickly on stage to testify to their healings. His words had
effects: many participants came on stage and stated that they had been healed
of their illness at that very moment.
HEALINGS, CONFIRMED PREDICTIONS, AND TESTIMONIES
In order to explain how an atmosphere of miracles and healings is socially

produced, I propose the macro– micro –macro model in figure 1. The model
shows how preparation techniques and individual and collective audience attrib-
utes combine to produce the perception and feeling of being healed in a certain
number of audience members and how these members are then selected and
their testimonies edited in order to create the final atmosphere of miracles and
healings. The model postulates a number of social mechanisms ( probability,
FIGURE 1. Explanatory Model for Pentecostal Miracles and Healings.
selection, latency, editing and autoreinforcement effects). In what follows, I

flesh out these general ideas.

Preparation Techniques
Andrews uses a whole range of techniques to prepare the participants to
receive the subsequent healing.
Suggestion statements and actions. A first class of techniques consists of sug-
gestion statements and actions. These statements and actions are used to elicit
altered states of consciousness in the participants and to ensure that they inter-
pret the ensuing perceptions as effects of the Holy Spirit. We find the following
types.
Salvation promises. These statements, consisting of ultra-short and positive
sentences, affirm as indisputable truth that Jesus is there right now in the audi-
torium and that he is accomplishing miracles and healings. They are not
accompanied by any qualification, justification, attenuation, or reflection.
Moreover, they mostly target the here and now as well as all or some of the
participants in the auditorium. Here are a few examples: “He is here.” “All
things are possible.” “He is gonna heal multitudes of you this morning.” “But
Jesus is just touching people. All over this auditorium. Right now.”
Body instructions. Andrews also uses various body instructions; he invites
the participants to close their eyes, raise their arms, sing, remain quiet, and
relax. Moreover, Andrews advises them to forget their illness, to see it as some-
thing of little importance, and to concentrate entirely on the Holy Spirit and
healing. Readers will note that Andrews’s instructions are close to those used
in many relaxation techniques, such as autogenic training, guided meditation
exercises, or hypnosis. There, too, short, positive sentences as well as body
instructions are used in order to induce trance-like states (Heap 2001).
Healing advice. A third element used to induce special “feelings” consists of
“theoretical” sentences and instructions. In these sentences, Andrews explains
what illness is, how it comes about, and what means may be used to obtain
healing. These affirmations have an exclusively practical aim; they explain how
the participants can and should change their behavior to be healed.
Words of command. This is an operation by which Andrews heals partici-
pants (through the mediation of the Holy Spirit) (compare Csordas 2002:338;
1988:48). Andrews uses words of command in different ways: sometimes, he
heals by binding the forces of the occult, darkness, and illness, thus freeing the
person from these evil influences. Behind this procedure seems the idea that
the name Jesus Christ “contains power” and may thus expulse the illness. The
person is “liberated” or else “freed in Jesus.”
Andrews: Holy Spirit
Interpreter: Saint-Esprit
Andrews: We take authority over sickness and disease and pain in this
meeting right now
Interpreter: Nous prenons autorité maintenant dans le nom de Jésus sur
toute maladie
Andrews: And in the name of Jesus
Interpreter: Et dans le nom de Jésus

Andrews: bind its power
Interpreter: Je lie sa puissance
Andrews: And I loose you from it this morning
Interpreter: Et je vous libère de cela maintenant
Andrews: I loose you in the name of Jesus
Interpreter: Et je vous libère dans le nom de Jésus
At other times, Andrews produces healing by making healthy organs appear.
At other times again, Andrews calls the Holy Spirit to go to certain places to
carry out its work.
Music, rhythm, and audience size. Another way that participants are pre-
pared for healings in the workshop is by music, a very widespread technique in
Pentecostal healing services (Nickel 2002). In this workshop, Andrews starts
out with a song. The song has a sweet melody and the lyrics consist of only
three words: “Jesus is here.” The participants are thus expressing themselves
musically while, at the same time, they are ceaselessly told by Andrews that
the Holy Spirit is among them right now. In addition, the music group contin-
ues to sing these words softly while the healing testimonies are taking place.
Andrew also uses rhythm and pace to prepare his audience. Again, this is a
technique mentioned in the literature (Horwatt 1988:131). In our workshop,
Andrews starts out very slowly and then increasingly picks up pace, raises his
voice, and even shouts. In the second part of the workshop, Andrews also
encourages the participants to sing a loud song (Hallelujah). Then again, he
lets everybody be very quiet just to “feel the anointing.” The attempt to keep
up or increase the pace is especially visible during healing testimonies.
Countless times Andrews admonishes the participants to act “quickly.” Clearly,
Andrews would like to give the impression that a “fireworks of healings” is
occurring all over the auditorium. Here are two examples:
If you’ve been healed this morning come quickly, come to the front come.
If you can feel the power of God that’s been on you, Come very quickly. Come on.
Run.
A further technique, often overlooked but of central importance, is the size of

the group. In our workshop, there were approximately 250– 300 people in the
room. The room was “full,” which is an optimal condition for a faith healing
service to “work well.” There are several reasons for this. First, a large audience
guarantees a relatively high probability that at least some individuals in the
audience will have altered body perceptions and/or think that they are “meant”
by one of the “words of knowledge”. Second, a larger audience allows the indi-
viduals to let themselves fall into altered states of consciousness more easily.
With a small audience, individuals might become self-conscious and feel

observed by other members. Third, even 300 people might seem “few,” if they
are in an enormous hall. For optimal results from the speaker’s viewpoint, the
size of the room must therefore be adapted to the size of the group (or vice

versa). Andrews attributes much importance to this point:
Much depends on the size of a fellowship and their maturity to determine the size of the hall
that should be rented. Healing meetings depend on atmosphere, and it will be a big mistake to
get over enthusiastic and rent an enormous hall, only to have it one quarter full with people.
(Andrews 2003:180)
Giving out Words of Knowledge

One of the most important preparation techniques used by Andrews is
“words of knowledge,” that is, predictions as to healings that were supposedly
happening right there and then in the audience. According to Pentecostals,
words of knowledge belong to the category of “spiritual gifts” and are defined
as knowledge revealed by God about facts that the person could not have
known in a natural way. Ian Andrews receives these words in different ways.
According to him, he sometimes receives a
Picture that comes to my mind very quickly. It’s a bit like a flash picture. It flashes in front
of your eyes and then it’s gone. . . . Sometimes, I hear a voice. Sometimes I have the pain
come to my body, that I know it’s not mine. . . . And so when you get that flash, you imme-
diately seize on it and say it. And then see if somebody responds.
Here is an example referring to the healing of breast cancer in the healing

workshop:
Andrews: We’ve got a person straight back through there
Interpreter: Il y a une personne juste là-bas derrière
Andrews: That has got a lump in the breast
Interpreter: Qui a une une tumeur, une grosseur dans le sein
Andrews: I want you to go out to the Lady’s room
Interpreter: Et j’aimerais que vous alliez dans la chambre où il y a les
dames
Andrews: Feel for it
Interpreter: Et que vous alliez ressentir cela
Andrews: And you’ll find it’s completely gone
Interpreter: Allez aux toilettes des dames et vous verrez que c’est parti
Altogether, Andrews gives out 17 words of knowledge of varying precision.
Nine words of knowledge referred to physical problems, two concerned mental
or spiritual problems, and six applied to unspecified problems. In general, they
are (like the statements and promises of salvation) very positive and absolute.
The people in the auditorium mentioned by Andrews are in the process of
being completely healed by the Holy Spirit, the pain is disappearing entirely,
and the cancerous tumor is vanishing without leaving a trace.
Three attributes of the words of knowledge are important for them to
“work.” First, there is a noticeable lack of specificity as far as the description of

the persons is concerned. In general, Andrews does not know the gender, age,
name, home, or even the location of the person in the auditorium. Second,
Andrews gives out words of knowledge touching on a whole range of illnesses
and afflictions: from stomach ache to a bad neck, from cancer to angina, from
back problems to deafness and paralysis. Third, the speaker does not immediately
verify if each new word of knowledge is correct in reality, and he does not look
for the concerned person in the audience. Rather, after a whole series of words
of knowledge, participants come on stage and confirm more or less precisely
one or more of the predictions. Before turning to the effects of all of the prepa-
ration techniques mentioned, we have to analyze a second causal factor: audi-
ence attributes.
Audience Attributes
In fact, the best preparation would not have much effect if it were not for
certain characteristics of the audience. If Andrews had preached in front of
300 hard-core Calvinists, it is likely that not many miracles and healings
would have emerged. While I do not have quantitative data for the special
case in question, my experience of attending many of AIMG happenings lets
me estimate that in an average AIMG meeting, we have 90 percent of individ-
uals with a Charismatic, Pentecostal, or Evangelical background, usually a
majority of women (around 60– 70 percent), and that average age might hover
around 40 to 50. Our interview data tell us what motives/desires, beliefs, and
practices these participants have.
Motives. I found four main reasons why individuals come to these meetings:
(1) curiosity, often linked with the fact that they know people who help organ-
ize the meeting, (2) desire to get healing for oneself or for a person one brings
to the meeting, (3) desire to learn more about healing and be motivated in
one’s faith, (4) because they have been asked to help by praying for the sick in
front of the stage. Often we find a mix of reasons. For our purposes, it is impor-
tant to note that at least some audience members have the urgent desire to be
healed and that many others at least entertain the possibility of being allevi-
ated of some minor affliction. In addition, even if they do not come for
healing, the great majority of participants do not attend in a “passive” mode.
Beliefs. Closely linked to the first point are audience beliefs. I propose to
describe these beliefs by what I call a “Lakatos-model.” Lakatos (1978) argued
that research programs by scientists (as well as all kinds of other belief systems
like Marxism or Freudism) consist of unfalsifiable “hard-core” beliefs that are
surrounded by a “protective belt” of hypotheses, conjectures, ideas, arguments,
etc. According to Lakatos (1978:48) it “ . . . . is this protective belt of auxiliary
hypotheses which has to bear the brunt of tests and get adjusted and
re-adjusted, or even completely replaced, to defend the thus-hardened core.”
Almost all of the AIMG participants that we interviewed argued in a
similar way. They entertained a few “hard-core” ideas. Among these, the most

frequently cited were (1) God exists; he heals today just as he did in the time
of Jesus, (2) God is sovereign; in principle he could heal any and all illness,
(3) Jesus has already paid for our sins—and therefore also our illnesses—on the
cross, (4) it is always the Holy Spirit, never man, who heals, (5) the Bible tells
us that we should pray for the sick. Very often, this “hard core” was explicitly
presented as “beyond doubt.” Participants said “I cannot doubt that . . . ,” “ . . .
there’s not a shadow of a doubt,” “ . . . it’s a reality that I absolutely do not
doubt . . . .”
On the other hand, many of the elements in the “protective belt” of partic-
ipants were open to discussion. Three domains of additional hypotheses
seemed to be important: a first set of hypotheses states that there are no fixed
rules or laws to healing. Healing is “no exact science,” it comes in “always dif-
ferent ways,” “you cannot make a rule.” This absence of rules concerns the
place, time, person, and the attributes of healing, as well as the appropriate
methods in order to induce healings. A second set of hypotheses states that
there are nevertheless techniques that may be used in order to further one’s
healing. Thus, praying for the sick can be taught and learnt. Faith, openness,
and the removal of various “obstacles” can have beneficial effects on healing.
A third set of hypotheses states that in healings, natural and supernatural ele-
ments may be combined in various ways. Faced with a specific case of healing,
one can therefore never be sure how much should be attributed to the Holy
Spirit, to what is “human,” or to natural causes. For example, as many inter-
viewees told us, it is always possible that individuals will only claim to be
healed when in fact they are not. Furthermore, interviewees believed that
many maladies were healed by natural causes. With the help of these auxiliary
hypotheses, it becomes possible to explain any and all (non)healings while
firmly protecting the hard-core beliefs stated above.
Practices. Let us turn to the individual practices of audience members
outside the specific meeting. It is important to understand that the great major-
ity of participants are not confronting healing and miracles for the first time
when they attend a healing meeting. Rather, most of them practice healing
and experience miracles in their everyday lives: they pray for the sick and for
themselves when they are ill, in their churches as well as in their family, and
on their own. Most have experienced divine healings in the past and are able
to narrate key experiences. Many have had words of knowledge for others and/
or have experienced such words themselves.
All of this means that the public is largely familiar with the values and
views presented by the speaker Andrews. They already know how to fall into
some mild or stronger state of altered state of consciousness and are familiar
with the habit of “testifying.” This point—completely missed by the “critical”
authors—becomes important when it comes to interpreting the entire healing
service. It means that we cannot distinguish the performer and the audience.
Instead, Andrews and (at least much of ) the audience are performing together,
creating jointly an atmosphere of miracles and healings.

“Feeling Healed” and “Feeling Meant”
With the help of the social techniques and with the background of audi-
ence attributes, Andrews may induce the belief in participants that they are
healed or at least meant to be healed. This is done in three (mutually nonex-
clusive) ways. First, individuals may come to have a different perception of the
body and mind and therefore believe that they have been healed or are about
to be healed. In the literature, these states are sometimes called “altered states
of consciousness” or “trance.” In our healing service, the fact that such states
are involved is beyond doubt. They are clearly visible on tape and individuals
describe them when testifying on stage as well as in our interviews. For
example, they narrate having experienced heat, weightlessness, a vision of
lights, very strong emotions, peace, impressions of something or someone
touching them, a trembling of the whole body, uncontrollable laughter,
weeping as well as special feelings in specific limbs (heat, bubbling, tingling).
Often, these are immediately imputed to the Holy Spirit, as when individuals
say they had, “fallen under the Spirit,” or had “felt the freedom and love of
God.” One interviewee said:
I went to Malley and then they prayed for me and then something just amazing, that is, I felt
the power of God go through physically. And that I had never, never felt before in my life, it
was like driving me crazy. (Marie)
From an “emic” point of view, the Holy Spirit is at work. From an outside per-
spective, however, the social techniques described produce the bodily experien-
ces of participants. Second, individuals may have the feeling of being healed or
about to be healed since they come to believe that one of the words of knowl-
edge (that predicts a currently happening healing) is “for them,” they “feel
meant”. Many interviewees told us that they often think “this is for me” when
a word of knowledge is pronounced. Other times, the word of knowledge seems
to fit their case reasonably well, but they nevertheless remain unsure if it is “for
them.” They also often experience that other people “claim” a word of knowl-
edge that they could have imagined was destined for them or, alternatively,
suspect that if they have not been healed by a word of knowledge that seemed
to be destined for them, others certainly have. Here’s a typical example:
Because for me, it’s already happened to me to have a call for me and then you recognize
yourself right away because you feel something, it’s something physical, it, it, it goes through
you from top to bottom, it’s just, just talking about it, ha (laughter) there’s something happen-
ing then, and then you say to yourself – but that’s me! That’s – that’s me. (Annie)
Why do these words of knowledge “work” from an outside, sociological point

of view? First, there is an effect of probability. As seen above, Andrews expresses
his predictions very generally and he virtually goes through the list of possible
physical and mental problems. It follows that there is a high probability that

within a large group at least one person will feel implicated. Second, there is
an effect of self-validation. The predictions of Andrews incite the members of
the audience to interpret their current situation as phenomena described by
him. As soon as individuals agree to recognize Andrews’s description in the
experiences that they are having and testify accordingly, there is “confirmation”
of the prediction. Third, individuals may come to believe that they are healed
or about to be healed even in the absence of any “confirming evidence” (such
as a word of knowledge or an altered state of consciousness). In these cases,
they just “take it on faith.”
Note that it is not easy for individuals to know if they “really” have been
healed. After all, they have to decide on the spot, without medical assistance
and on the grounds of often somewhat shaky indicators. A good example is the
following participant who recounts how he thought he was healed and testified,
but then found that there had, in fact, been no healing:
Even me, it happened to me one time when I thought my tinnitus had disappeared after a
prayer that someone said on stage for people who had tinnitus and the Lord did something
then . . . and . . . I even went to testify after to say that it had gone away but it was it was
too much all at once I couldn’t make sure it wasn’t quiet enough to make sure. . . . But it
seemed to me. (Eduardo)
We thus see that individuals have varied possibilities to get themselves to

believe that they have been or are about to be healed. These motivations then
lead them to testify.
Testimonies
A final element of the healing service can be found in the “healing and
prediction testimonies.” We can define Christian testimonies as statements
whereby individuals reveal their Christian faith to others. In the case that
interests us, individuals are invited to announce healing testimonies. As a side
effect, they may also confirm predictions, thus attesting to further miracles.
While healing testimonies and prediction confirmation always go together, it is
useful to treat them separately. Let us first look at confirmed predictions.
Testimonies of confirmed predictions. During the healing service several indi-
viduals come on stage and confirm Andrews’s predictions. The confirmations
are sometimes very explicit; one individual starts out by saying: “I’m the one
whose neck was healed. I had a car accident. . . . .” At other times, the confir-
mation is made more indirectly, such as when one participant says: “You were
talking about occultism, um, about the occult. And last night I had a dream
about a very bad spirit.”
From the emic point of view, these happenings may be seen as miracles.
But can we account for these happenings from an outside, sociological point of
view? Let us first describe more in detail just what kinds of confirmations we
are dealing with. From the 17 words of knowledge given out by Andrews, 10

are specific enough that they might be “confirmed.” Six of these 10 are then
actually confirmed either explicitly or implicitly by testimonies. However, a
closer inspection of the confirming testimonies shows that the predictions and
the experiences by those who “claim” the word of knowledge often do not
match very well. For example, Andrews prophesies that a woman is healed
instantaneously and completely from breast cancer. The woman who claims
the healing says that her breast is still afflicted by a tumor, but “feels softer
than before.” In another prediction, Andrews prophesies that a person in the
audience “is in the occult” and “knows about power.” The person who “claims”
this word of knowledge has just had a dream about a bad spirit.
We therefore see what we would have expected in the presence of a
shotgun technique and self-selection effect: Andrews “takes a shot” at predict-
ing healings and audience members are selecting themselves by thinking “this
is me” and then testifying, even though the “fit” with the prediction might be
rather slight.
Healing testimonies. In the workshop, various healings were claimed. One
woman can suddenly hear. In the midst of cheers, she holds up her hearing aid
and says, “I believe I don’t need it anymore.” A man claims that a “growth” in
his leg is “going away.” A woman seems to have been healed after 20 years of
depression. Another woman’s neck that had been stiff because of a car accident
is healed. In the following example, an elderly woman affirms that she has
been healed of deafness.
Participant: J’entends, j’entends, j’entends, j’avais besoin de ça (shows
hearing aid), j’entends maintenant
Interpreter: I can hear now, I can hear now, halleluiah, halleluiah
Participant: Oh Jésus!
Andrews: How long have you been deaf?
Participant: Oh ça fait longtemps, que j’ai un appareil assez longtemps J’ai un
j’ai un appar . . . j’avais un appareil acoustique depuis très longtemps Euh euh
euh maintenant il me semble que j’en ai plus besoin!
Interpreter: And now I think I don’t need it anymore (shows the hearing
aid to everyone). Halleluiah
Andrews: Bless you, bless you, bless you
The testimonies may be seen as small interactions between Andrews, the inter-
preter, and the testifier; they are most often structured in a very similar
manner. Andrews begins with an invitation to speak (“What happened?”,
“What happened to you?”). One or two short sentences follow in which the
person tells his or her illness or problem and how he or she was healed. Then,
Andrews sometimes asks the person for details of her story (for example, he
asks when the depression first began), or else he incites the person to demon-
strate her healing (for example, show that she can hear with the formerly deaf
ear). The purpose of this is to dramatize the healing. Then, Andrews praises
Jesus (“Thank you Jesus,” “Isn’t Jesus wonderful?”) and places his hand on the

person’s forehead. With this gesture, the person often falls backwards.
“Catchers” then hold her up, thus preventing her from injury. In what way are
the testifiers healed?
A close look at the testimonies shows that the 24 individuals (20 women,
four men) who testify are by far not all “completely healed” as the words of
knowledge by Andrews would have had it: 37.5 percent claim to have been
healed, 20.8 percent say that they have been partly healed, 25.0 percent of
individuals do not seem to be healed, and 16.7 percent individuals do not
claim any healing (but testify to an emotional experience). What problems
and what types of healing do the participants mention? In 54.2 percent, they
pertain to physical illnesses (for example, arthritis, neck problems, cancer). In
29.2 percent, they concern emotional or spiritual problems (mental confusion,
bad dream, depression, and grief ). In 16.7 percent cases, the persons do not
mention a problem, but a joyful experience in which the Holy Spirit inter-
venes. Again, this state of affairs fits our model well, showing that many things
may trigger testifying: perceived healings, bodily and emotional experiences,
complete or only expected healings, etc.
Selection effect. Testimonies are not governed only by the desire of individ-
uals to testify. The setup of the healing meeting includes at least three impor-
tant additional mechanisms that render the testimonies more powerful:
selection, editing, and latency. The selection effect is produced by the norm
requiring that only healings (but no nonhealings) should be allowed on stage.
Very repeatedly, the participants are called to come on stage if they are healed.
Those who have been healed now you can come forward and give a testimony of the healing
of God.
If you’ve been healed this morning come quickly, come to the front come.
If you can hear come up, I want people that have been touched by the Holy Spirit.
The norm is not just repeated countless times by Andrews, it is also enforced
by helpers who do a “screening” before leading witnesses to the stage.
However, the norm is not always followed. At one interesting point, Andrews
sends a person who does not respect the norm back to her seat.
Andrews: How long have you been deaf?
Participant: Oh oh les médecins ils disent que je suis sur. . . . , non non
mais je suis . . . je n’entends pas bien encore hein. . . .
Interpreter: But she can’t hear any . . . yet.
Andrews: Right, but I don’t want that, I want people who have been healed.
It’s the Holy Spirit is doing the healing, not me.
Editing effect. Another mechanism is an “editing” of the testimonies by
Andrews. These techniques make the testimonies more impressive than they
would otherwise have been. One technique consists of not going into detail as
to what has actually happened and just assuming that the person testifying has

been totally healed. Andrews uses this often. He always receives the testimo-
nies in a very positive manner and validates them. Even if the “witnesses”
express only very unspecific things, such as a happy face or trembling, Andrews
immediately takes it as confirmation of a healing. In the following segment,
Andrews has questioned individuals who have been “healed of depression.”
Since the witnessing woman looks happy, Andrews simply assumes that she has
been healed of depression, without going into further detail.
Andrews: Look at this lady
Interpreter: Regardez à cette jeune dame
Andrews: Isn’t she beautiful?
Interpreter: Est ce qu’elle n’est pas belle?
Andrews: How long have you been depressed?
Interpreter: Pour combien de temps avez-vous été dépressive?
Participant: Twenty years
Andrews: Twenty?
Participant: Twenty
Interpreter: Pendant vingt ans
Andrews: Twenty years! Look at her
Interpreter: Pendant vingt ans. Regardez-la
Andrews: Oh Jesus loves you
Interpreter: Le Seigneur vous aime
Another way of editing is the dramatization of the healing by emphasizing
either the importance of the malady in the past or the completeness of the
healing in the present. Thus, Andrews may insist on the fact that a woman has
been depressed for 20 years, before finally being healed this day or he may
“test” the newly found hearing of another woman by whispering words into her
formerly deaf ear.
A final way of editing is the “handling” of nonhealings. When people
without a healing succeed in getting on stage (against the norms of the interac-
tion system), Andrews does not take their testimony as evidence that the Holy
Spirit has not fulfilled expectations. Rather, he prays for these people and
“heals” them on the spot. This is mostly followed by the persons falling to the
floor under the power of the Holy Spirit, where they continue lying without
further being interviewed. In any case, there is no immediate “confirmation” of
their healing, which has the advantage of excluding possible negative
messages.
Latency effect. A third mechanism enhancing the effect of testimonies may
be called an effect of latency. The fact that the different previsions are not veri-
fied one by one but are grouped together in a set of verifications (testimonies)
masks the fact that (1) many testimonies match the content of the words of
knowledge only superficially, (2) many words of knowledge are not confirmed
at all, (3) some individuals testify even if they have not been healed in this
workshop, but at some other time and place.8

Auto-amplifying processes and charisma. We have already cited Ian Andrews
as saying that “Healing meetings depend on atmosphere.” I agree but must ask
why atmosphere is indeed so very important. The reason is that healing
services depend on a close interaction between the speaker and the audience.
This leads to an important self-reinforcement effect. Thus, if nobody or only a
few obtain healing, the speaker and the audience lose confidence, the individu-
als “cool off,” nobody falls into altered states of consciousness, and this results
in even fewer healings. Conversely, once a few strong healings have been
claimed, this “builds faith” and a healing service can rocket off to a frenzy of
miracles and healings.
This also means that the “charisma” of the speaker is highly vulnerable. It
helps, of course, to be well known and to have a reputation for miraculous
deeds. However, much like a rock star or comedian, if the speaker does not
“deliver” on the spot, if he does not succeed in enthralling his audience, his
long-earned reputation is not worth a cent. As Weber (1978:1114) noted, cha-
risma depends on success and if there are no healings, the speaker does not
seem to be imbued with the Holy Spirit. In our case, Ian Andrews is well
known, but by no means a “star.” Nevertheless, his technique is good enough
to produce quite an important amount of miraculous and healing activity.
APPLYING THE MODEL: THOUGHT EXPERIMENTS

AND COMPARISONS
Explanatory sociology looks for “causal mechanisms.” It implies that chang-

ing the causes will change the effects. In a social scientist’s ideal world, we
could do randomized social experiments, setting up healing services with
various parameters and then measuring the outcomes. For the moment, I will
use other (and less costly) methods: thought experiments and comparisons. In
thought experiments, social scientists change the explanatory variables in their
models only mentally, trying to imagine what should happen to their depend-
ent variables (Maxwell 2005:58ff.). Furthermore, they may compare their case
to other cases in social reality that show these different parameters—and check
if their predictions are confirmed. While this is admittedly a rather poor substi-
tute for the “real thing,” it nevertheless may lead to interesting insights and
plausibility checks.
8
This contrasts with Andrews’s view that “95 percent plus” of words of knowledge are
verified.
Let us imagine what would happen if we took away the self-selection and
latency effects. Each healing action or word of knowledge would be uttered for
a specific person and immediately verified. We would predict a much higher pro-
portion of falsified healing actions and inapplicable words of knowledge and

therefore a less intensive atmosphere of miracles and healings. Interestingly, I
have seen a speaker at one AIMG meeting trying to do this. As predicted by
our model, a lot of nonhealings came to light and the whole atmosphere in the
room got increasingly embarrassed. Shortly after, the leaders of the congrega-
tion took over the meeting from the unlucky speaker, resorting to their usual
“shotgun technique.” The same reasoning explains why healings in “healing
rooms,” where a number of intercessors pray for one specific individual, are
reported to be much less spectacular, not necessarily instantaneous and may
easily be “lost” (compare to Csordas 2002:336).9
To take a second example: what would happen if we increased the cha-
risma of the speaker, the size of the audience, and the selection effect? Our
model would lead us to expect more spectacular healings and miracles. A clear
example of this can be seen in the Benny Hinn healing services (Nickel
2002). Hinn is highly charismatic, uses much larger audiences and a strong
screening and selection of testifyers. As our model predicts, the miracles are
spectacular: Hinn can point at a person and the person is seemingly thrown to
the floor by an invisible hand. Hinn screams “power” and about a hundred
people seem to fall over like a row of dominoes, etc.
What, finally, would happen if we were able to change audience attributes,
for example increase the desire to be healed? Again, our model would lead us
to expect more impressive healings and miracles. A good example are the
evangelistic meetings by Reinhard Bonnke in various African countries, where
the biomedical supply and general conditions according to the Human
Development Index are less favorable than in the Western world. The miracles
apparently happening on stage are spectacular, including the raising of the
dead (see the documentary by Thomas 1999).
BOUNDED RATIONALITY AND THE PERSISTENCE OF HEALING

PRACTICES
Even if we accept everything said thus far in this paper—some major ques-
tions remain. The explanatory sociology approach assumes that individuals use
(bounded) rationality. Now, it is easy to explain why individuals with incurable
maladies turn to faith healing (Gilovich 1991:126). It seems to be rational to
try a fringe cure if everything else has failed. However, the majority of
9
I have to acknowledge that our interviewees also narrate cases in which miraculous
healings happened in a small circle of individuals “under their eyes.”
individuals who witness faith healing services are not “terminal cases.” If
divine healing has only a placebo effect—why are these people often deeply
convinced even in the long term?
My position is that Pentecostals use (bounded) rationality just as well as

anybody else. Yet, since they begin with different assumptions, they end up
believing in the reality of healings even if the evidence (from the outside)
seems lacking (compare to Dericquebourg 2004). For lack of space, I can only
sketch the argument that will entail three points. First, many participants
showed a belief structure that is no less rational than many other belief systems,
ideologies, or even scientific theories. As shown above, they hold a “hard core”
of beliefs and a protective belt of auxiliary hypotheses that lets them protect
the hard core and explain any apparent “falsification.”
Second, the reasoning most participants use when evaluating healings and
miracles seems to follow a practical, everyday rationality (Esser 1993). In fact,
participants were far from thinking that everything they saw in the healing
workshops was “true” or “real.” Rather, their answers showed a great amount of
skepticism. Faced with what was presented as a divine healing, participants
thought that it was always possible that the healing would have occurred
anyway (naturally), that it was only feigned by the person, or that the person
honestly thought she was healed but in fact was not. Respondents pointed out
that some testifiers are “human” and just want to be on stage, that others
“claim” a healing since they think that they will be healed with a higher prob-
ability and that still others only think they are healed because of the “manipu-
lative” atmosphere of healing meetings.10 Likewise, however, an apparent
nonhealing can just as well have many different reasons and certainly cannot
falsify the claim that God heals still today. Respondents produced a long list of
arguments (this is the “protective belt”) in order to explain nonhealings: if a
person has not been healed, this may be the case because the person did not
have enough faith (e.g., person looked at the malady, not at Jesus), did not really
want to be healed,11 was faced with external obstacles (e.g., not lived in accord-
ance with the Bible), was faced with internal obstacles (e.g., anger, nonpardon),
or an unfavorable situation (e.g., the speaker did not have enough faith).
Alternatively, perhaps a healing has occurred, but due to special circumstances,
has not (yet) appeared as such. Thus, it may be that the person was healed,
but: she/he was not able to “keep the healing” through faith, the healing has
not yet been “manifested” but will do so in the future, the healing arrived in a
different way than expected (e.g., emotional instead of physical).12
10
Many respondents were acutely aware of the fact that some speakers try to “condi-
tion” the audience in order to “produce” more apparent healings.
11
Many of these explanations were also strongly rejected by various interviewees.
12
Alternatively, it may be that there has been a healing, but not to the person for
whom people had prayed.
Third, most interviewees were able to point to “good evidence” (at least in
their eyes) that supported the claim of divine healing (Boudon 1997). Most of
them had acquaintances, friends, and relatives who had been healed in spectac-
ular ways in the past. Furthermore, many thought that the Holy Spirit helped

them effectively to fight maladies in their everyday life. From an outside per-
spective, this latter point is easy to understand. First, the human body cures
most illnesses by itself. If a person is ill, uses an—even inefficient—medicine
or treatment, and then gets better, the cure seems to have worked (Gilovich
1991:128). Second, maladies are never stable. Periods of greater suffering alter-
nate with periods of relief. If individuals use an inefficient treatment when
they suffer greatly, chances are that their suffering will diminish because of the
normal variability of the malady. Again, the cure will seem to have worked
(Nolen 1975:69ff.).
To sum up, the persistence of Pentecostal healing practices is not due to
“irrational” beliefs and practices. Pentecostals use completely “normal”
(Lakatos-)belief structure, they use practical everyday reasoning, and rest their
case on (at least subjectively) “good evidence.”
CONCLUSION
This article has addressed the question of which social techniques produce
miracles and healings in Pentecostal healing workshops. It has also tried to
explain why such workshops may persist in the long term, even though people,
in time, might learn that these healings and miracles are often greatly exagger-
ated. My answer, concisely, is as follows.
Miracles, that is, confirmed predictions of healings, are explained by the
way the “interaction system” works. Andrews uses the “shotgun technique.”
That is, he calls out a number of supposed healings in the audience, going
through a list of possible maladies. Since the audience is large, he has a high
probability that at least one person will accept the description as “fitting” him-
or herself (effect of probability). When this person comes forward to testify,
there seems to be a miracle, although the prediction produced its own valida-
tion (effect of self-validation). Since predictions are never immediately veri-
fied, but only after a certain lapse of time, predictions that are not confirmed
are forgotten and predictions whose confirmations match badly are not noticed
(effect of latency).
Healings are also produced through the mechanisms of the interaction
system. Andrews uses certain preparation techniques (words of knowledge, sug-
gestions, music, rhythm, and audience size) in order to create altered states of
consciousness and the feeling of being “meant” in various members of the audi-
ence. These audience members have a strong impression of being (about to be)
healed. In a next stage, Andrews makes sure that only individuals who have a
healing to report are allowed on stage (using norms and helpers). Finally, he
“edits” the testimonies to make them look much more impressive. If he does
this successfully, by way of an effect of autoreinforcement, individuals in the
audience will perceive miracles and healings happening, increase their faith in
the charisma of the speaker, and have a higher probability of producing yet

more healing testimonies.
I have argued that these results are not in conflict with the basic assump-
tion that Pentecostals use (bounded) rationality. Pentecostals have a “normal”
belief structure including a “hard core” and a “protective belt” of auxiliary
hypotheses, they are often quite skeptical and use practical everyday rationality
to explain both healings and nonhealings, and they rest their case on a body of
(at least subjectively) “good evidence.”
My article seeks to make a theoretical and methodological contribution.
Theoretically, I have shown what social mechanisms are at work that create the
social reality of “miracles and healings” in a specific Pentecostal healing work-
shop. Thought experiments and comparisons with other formats show that a
wide range of other Pentecostal healing phenomena can also be explained with
the model. I have argued that such an explanatory approach has a distinctive
contribution to make in comparison with anthropological approaches that do
not focus on causal mechanisms as well as in comparison with critical
approaches that fail to understand the social context and the beliefs and practi-
ces of the actors involved. Methodologically, I have shown that qualitative
methods may be used fruitfully in order to do explanatory, mechanism-based
sociology. It is simply not true that “causal analysis” can only be done with
quantitative methods.
This being a case study, the limits of my argument are clearly visible. The
results of case studies are never “representative” for a given population of cases,
even if we are dealing with “typical” specimens. Only future research on a
wider range of phenomena and including possibly quantitative research designs
will tell if this is a promising avenue for research on Pentecostalism. If that
happens—I predict—all things are possible.
FUNDING
This research was partially supported by a research grant from the Faculty
of Theology and Study of Religion (FTSR) from the University of Lausanne.
REFERENCES
Anderson, Allan. 2002. “Pentecostal Approaches to Faith and Healing.” International

Review of Mission XCI:523 –34.
Andrews, Ian. 2003. Equipped to Heal. Columbus, GA: TEC.
Berger, Peter L. 1990 (1967). The Sacred Canopy: Elements of a Sociological Theory of
Religion. New York: Anchor.
Boudon, Raymond. 1997. The Art of Self-Persuasion. London: Polity.
Burgess, Stanley M. 2006. “Introduction.” In Encyclopedia of Pentecostal and Charismatic

Christianity, edited by S.M. Burgess, xiii –xiv. New York: Routledge.
Coleman, James S. 1990. Foundations of Social Theory. Cambridge, MA: The Belknap Press
of Harvard University Press.
Csordas, Thomas J. 1988. “Elements of Charismatic Persuasion and Healing.” Medical
Anthropology Quarterly 2:121 –42.
———. 2002. “The Rhetoric of Transformation in Ritual Healing.” In Body, Meaning,
Healing, edited by T.J. Csordas, 11 –87. New York: Palgrave Macmillan.
Curtis, Heather D. 2007. Faith in the Great Physician. Suffering and Divine Healing in
American Culture, 1860 –1900. Baltimore, MD: The Johns Hopkins University Press.
Dericquebourg, Régis. 2004. “Comment les thérapies religieuses sont-elles plausibles?”
Politico Hermetica 18:11 – 25.
Dow, James. 1986. “Universal Aspects of Symbolic Healing: A Theoretical Synthesis.”
American Anthropologist, New Series 88:56 – 69.
Ernst, Edzard. 2003. “Distant Healing - an ‘Update’ of a Systematic Review.” Wiener
Klinische Wochenschrift 115:241 –45.
Esser, Hartmut. 1993. “The Rationality of Everyday Behavior. A Rational Choice
Reconstruction of the Theory of Action by Alfred Schütz.” Rationality and Society 5:7–31.
———. 2000. Soziologie. Spezielle Grundlagen. Band 2: Die Konstruktion der Gesellschaft.
Frankfurt: Campus.
Geertz, Clifford, ed. 1993. “Religion as a Cultural System.” In The Interpretation of Cultures.
Selected Essays, 87 – 125. London: Fontana.
Gilovich, Thomas. 1991. How We Know What Isn’t So. The Fallibility of Human Reason in
Everyday Life. New York: The Free Press.
Hamilton, Malcolm. 2001. The Sociology of Religion. Theoretical and Comparative Perspectives.
2nd ed. London: Routledge.
Heap, Michael. 2001. “The Nature of Hypnosis.” A report prepared by a Working Party at
the request of The Professional Affairs Board of The British Psychological Society.
Leicester: The British Psychological Society.
Hedström, Peter. 2005. Dissecting the Social. On the Principles of Analytical Sociology.
Cambridge: Cambridge University Press.
Horwatt, Karin. 1988. “The Shamanic Complex in the Pentecostal Church.” ETHOS
16:128 – 45.
Knoblauch, Hubert. 2009. “Videography. Focused Ethnography and Video Analysis.” In
Video Analysis: Methodology and Methods. Qualitative Audiovisual Data Analysis in
Sociology. 2nd ed. edited by H. Knoblauch, B. Schnettler, J. Raab, and H.-G. Soeffner,
69– 83. Frankfurt am Main: Peter Lang.
Lakatos, Imre. 1978. The Methodology of Scientific Research Programmes. Philosophical Papers,
vol. 1. Cambridge: Cambridge University Press.
Laurent, Pierre-Joseph. 2001. “The Faith-Healers of the Assemblies of God in Burkina
Faso: Taking Responsibility for Diseases Related to ‘Living Together’.” Social Compass
48:333 – 51.
Maxwell, Joseph A. 2005. Qualitative Research Design. An Interactive Approach, vol. 41. 2nd
ed. Applied Social Research Methods Series. Thousand Oaks, CA: Sage.
McGuire, Meredith B. 1987. “Ritual, Symbolism, and Healing.” Social Compass 34:
365– 79.
——— 1991. Ritual Healing in Suburban America. New Brunswick, NJ: Rutgers University
Press.
Miles, Matthew B., and Michael A. Huberman. 1994. Qualitative Data Analysis. An
Expanded Sourcebook. Thousand Oaks, CA: Sage.
Moerman, Daniel E. 1979. “Anthropology of Symbolic Healing.” Current Anthropology
20:59 –80.

Nickel, Joe. 2002. “Benny Hinn: Healer or Hypnotist?” Skeptical Inquirer 26.3 (May/June).
Retrieved 1 March 2011 (http://www.csicop.org/si/show/benny_hinn_healer_or_hypnotist).
Nolen, William A. 1975. Healing: A Doctor in Search of a Miracle. New York: Random
House.
Opp, James. 2005. The Lord for the Body. Religion, Medicine, and Protestant Faith Healing in
Canada, 1880 – 1930. Montreal and Kingston: McGill-Queen’s University Press.
Randi, James. 1989. The Faith Healers. New, Updated Edition. Amherst, NY: Prometheus.
Rose, Louis 1971. Faith Healing. Harmondsworth, UK: Penguin.
Rossi, Ilario. 2009. “Pluralité religieuse, médecine et santé: recoupements et interactions.”
In La nouvelle Suisse religieuse. Risques et chances de sa diversité, edited by M. Bauman,
and J. Stolz, 337 – 52. Genève: Labor et Fides.
Stolz, Jörg, and Oliver Favre. 2005. “The Evangelical Milieu. Defining Criteria and
Reproduction across the Generations.” Social Compass 52:169 – 83.
Stolz, Jörg. 2006. “Salvation Goods and Religious Markets: Integrating Rational Choice
and Weberian Perspectives.” Social Compass 53:13 – 32.
———. 2009 “Gods and Social Mechanisms. New Perspectives for an Explanatory
Sociology of Religion.” In Raymond Boudon. A Life in Sociology, edited by
M. Cherkaoui, and and P. Hamilton, 171– 88. Oxford: Bardwell.
Strauss, Anselm, and Juliet Corbin. 1998. Basics of Qualitative Research. Techniques and
Procedures for Developing Grounded Theory. 2nd ed. Thousand Oaks, CA: Sage.
Thomas, Antony. 1999. A Question of Miracles. Documentary. http://www.imdb.com/title/
tt0447686/.
Warrington, Keith. 2006. “Gifts of Healing.” In Encyclopedia of Pentecostal and Charismatic
Christianity, edited by S.M. Burgess, 232– 36. New York: Routledge.
Weber, Max. 1978. Economy and Society. An Outline of Interpretive Sociology. Berkeley:
University of California Press.

"All Things Are Possible": Towards A Sociological Explanation of Pentecostal Miracles and Healings

Uploaded by

Copyright:

Available Formats

"All Things Are Possible": Towards A Sociological Explanation of Pentecostal Miracles and Healings

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

"All Things Are Possible": Towards A Sociological Explanation of Pentecostal Miracles and Healings

Uploaded by

Copyright:

Available Formats

Sociology of Religion 2011, 72:4 456-482

“All Things Are Possible”: Towards a

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Pentecostalism is one of the fastest growing religions worldwide (Burgess

*Direct correspondence to Jörg Stolz, University of Lausanne, Observatoire des religions en

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

ACCOUNTING FOR MIRACULOUS PREDICTIONS

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

While the anthropological approach has produced many important insights,

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Explanatory Sociology Approach

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

This is a case study of an AIMG (Association Internationale des

Video and Transcription Analysis

as to make the miracles and healings as spectacular as possible to attain their

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

In-depth Interviews and Participant Observation

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

THE HEALING SERVICE IN CONTEXT

Organizational and Historical Context

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

The Healing Service

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

HEALINGS, CONFIRMED PREDICTIONS, AND TESTIMONIES

In order to explain how an atmosphere of miracles and healings is socially

selection, latency, editing and autoreinforcement effects). In what follows, I

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

A further technique, often overlooked but of central importance, is the size of

With a small audience, individuals might become self-conscious and feel

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Giving out Words of Knowledge

Here is an example referring to the healing of breast cancer in the healing

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Why do these words of knowledge “work” from an outside, sociological point

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

We thus see that individuals have varied possibilities to get themselves to

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

APPLYING THE MODEL: THOUGHT EXPERIMENTS

Explanatory sociology looks for “causal mechanisms.” It implies that chang-

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

BOUNDED RATIONALITY AND THE PERSISTENCE OF HEALING

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Anderson, Allan. 2002. “Pentecostal Approaches to Faith and Healing.” International

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

Downloaded from https://academic.oup.com/socrel/article/72/4/456/1613371 by University College London user on 07 December 2021

You might also like