GODS IN THE FLESH - Arnaud Halloy-1

This article was downloaded by: [Arnaud Halloy]
On: 18 July 2012, At: 02:04

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T
3JH, UK
Ethnos: Journal of
Anthropology
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/retn20
Gods in the Flesh: Learning

Emotions in the Xang"ô
Possession Cult (Brazil)
a
Arnaud Halloy
a
University of Nice Sophia-Antipolis, Laboratoire
d'Anthropologie et de Psychologie Cognitives et
Sociales (LAPCOS)
Version of record first published: 17 Jul 2012
To cite this article: Arnaud Halloy (2012): Gods in the Flesh: Learning Emotions in
the Xang"ô Possession Cult (Brazil), Ethnos: Journal of Anthropology, 77:2, 177-202
To link to this article: http://dx.doi.org/10.1080/00141844.2011.586465
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-

and-conditions
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any form
to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any
representation that the contents will be complete or accurate or up to
date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not
be liable for any loss, actions, claims, proceedings, demand, or costs or
damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
Downloaded by [Arnaud Halloy] at 02:04 18 July 2012
Gods in the Flesh: Learning Emotions in the
Xangô Possession Cult (Brazil)
Arnaud Halloy
University of Nice Sophia-Antipolis, Laboratoire d’Anthropologie et de Psychologie Cognitives et Sociales
(LAPCOS)
abstract Drawing on first-hand ethnographical data from the Xangô, an Afro-

Brazilian Cult in Recife (Brazil), I defend that learning possession means in the first
place to learn to identify and react to specific emotional states in accordance with cul-
tural representations and expectations. This emotional learning would take place
through two potential processes during ritual activity. On the one hand a powerful
coupling process linking “uncanny” body arousals to mythological imagination captured
by the highly evocative content of songs, invocations, objects and substances during
ritual activity. Ritual features such as archetypality, rigidity, regularity, redundancy
and spatial and temporal delimitation are propitious for eliciting and “boosting” this
coupling process. On the other hand, a social referencing process through which
emotional and behavioral reactions towards possessed persons act for novices as
reliable indicators of how to recognize, interpret and regulate their own emotional
states associated with possession. With growing familiarity with possession’s concep-
tual and experiential background, their emotional reactions might evolve from largely
undifferentiated arousal states to “orixá-specific” somatic signature and novices might
become more active towards their own possession. They also demonstrate an increas-
ing sensitivity to isolated emotional elicitors of possession as well as a better control
over their emotional reaction.
Introduction
P
ossession1 is probably the most intimate bodily experience of gods, spirits
and ancestors in many religious traditions around the world. At the same
time, its manifestations are mostly public and highly conventional, reflect-
ing the diversity of idioms and cultural expectations as to its form and content.
Despite such heterogeneity, recent research in cognitive anthropology on the
ethnos, vol. 77:2, june 2012 (pp. 177 –202)

# 2012 Routledge Journals, Taylor and Francis
issn 0014-1844 print/issn 1469-588x online. http://dx.doi.org/10.1080/00141844.2011.586465
178 arnaud halloy
mental structures underlying the conceptualisation of spirits and mind/body

interaction seem to suggest that potentially universal mental mechanisms can
account for the success of possession phenomena around the world (Cohen
2007, 2008). Studying possession thus appears a tremendous challenge to
both social and psychological sciences – it is at once private and conventional,
cognitively constrained and culturally shaped and enacted.
Scientific studies of possession have developed rapidly since the middle of
the nineteenth century2 (see the Introduction). Unfortunately, social, medical,
political and psychological theories are usually presented as mutually exclusive,
although scholars have been prompt to recognise the manifold dimensions of
possession phenomenon. More recently, however, contemporary anthropolo-
gists have started building interdisciplinary approaches to possession (Gell
1980; Smith 2001; Klass 2003; Becker 2004; Luhrmann 2004, 2007; Rouget
2006; Cohen 2007, 2008; Seligman & Kirmayer 2008; Halloy 2009; Luhrmann
et al. 2010).3 This article follows in their steps, proposing an alternative frame-
work to psycho-medical and socio-functional approaches in anthropology,
and suggesting an approach situated at the crossroads of cognitive sciences
and ethnography.
While in principle subscribing to a naturalist approach to cultural phenom-
ena (Sperber 1996), the thesis developed here offers two necessary extensions to
the mainstream cognitive approach in the anthropology of religion (Boyer 1997,
2001; Barrett 2004; McCauley & Lawson 2004; Cohen 2007, 2008). On the one
hand, it argues for an analysis of the contextual conditions constraining commu-
nicative and cognitive processes (Severi 2007; Whitehouse 2000, 2004). The
most valuable contribution of the cognitive approach consists in drawing the
attention of social scientists to ‘cognitive constraints’ and their potential influ-
ence on the conceptualisation and memorisation processes of cultural represen-
tations. However, we should also recognise its difficulty of dealing with the
complexity of real-life situations, where contextual factors may play a decisive
role in cultural learning. On the other hand, some special and recurrent features
of cultural transmission can be found not only in conceptual forms, a position
defended by cognitive anthropology, but also in how cultural knowledge is embo-
died into distinctive bodily states, and how it is enacted in performance and the
manipulation of artifacts.4 The present study embraces these two (rather neg-
lected) dimensions by developing a cognitive ethnography,5 focused on the pro-
cesses through which possession is learnt in the Xangô, an Afro-Brazilian cult in
Recife. By ‘cognitive ethnography’, I mean a theoretical and methodological fra-
mework focused on learning processes and able to take into account and articu-

Gods in the Flesh 179
late the contextual and diachronic dimensions of learning. Cognitive

ethnography can thus be seen as a way of reconciling cognitive and anthro-
pological approaches by reconnecting cognitive processes with subjective
experience by means of perception and corporeality as suggested by Csordas
(1990, 1993, 1994). However, I do not share the phenomenological view which
rejects ‘objectivist explanations of religious experience’ and sees learning as
‘inadequate to account for the phenomena discussed [charismatic healing or
possession]’ (Csordas 1990: 31– 2). In contrast to Csordas’ skepticism, I will
show that a cognitive ethnography of possession focused on learning processes
should not be reduced to ‘stimulus-response patterns’ (ibid: 33) and that it can
‘integrate domains of perception, practice, and religious experience’ (ibid: 31)
into the same explanatory framework (see also the Introduction).
The present study starts from a conclusion formulated by Rouget (1990) at

the end of his authoritative work on music and trance:
In the genesis of every possession cult, and thus of every possession trance, there must
have been, at origin, an emotional state lived by an individual and likely to be lived
again by others, either spontaneously, or by learning. (1990: 557, my translation)
Those who had the opportunity of being in close contact with possession cults
would find it difficult to ignore their emotional dimension. However, few scho-
lars have paid close attention to it (for noteworthy exceptions, see Rouget 1990;
Aubert 2004; Becker 2004). My aim here is to offer a better definition of the type
of emotion constitutive of possession, as well as a better description of the
emotional learning process at work in the Xangô cult, by considering the role of
contextual factors (such as features of ritual action and social signals) in the
shaping of emotional responses associated with possession episodes. In prin-
ciple, spirit possession cannot be reduced to emotional learning, but I will
argue here that learning possession means primarily learning emotions. In
other words, learning emotions is a prerequisite of every (elective) possession.
Spirit Possession in the Xangô Cult6

The Xangô cult
The Xangô cult, an Afro-Brazilian possession cult of Yoruba origin, is located
in Recife, the capital of Pernambuco state, in north-eastern Brazil. According to
Bastide (1960), its name derives from the popularity of the African deity Xangô
in the city. It began in Recife in the late nineteenth century. From being some-
what restrained during the first half of the twentieth century, the cult expanded

180 arnaud halloy
rapidly throughout the city from the 1940s to the 1970s, thanks to the influence
of charismatic cult leaders (de Carvalho 1987).
The social organisation of the Xangô cult is based on ‘saint families’
( familias-de-santo). These collective entities rely on initiatory links between
their members, elaborated on the model of kinship in a biological family.
The initiators are called ‘saint-father’ and ‘saint-mother’, initiates ‘saint-son’
or ‘saint-daughter’, and these co-initiates subject to the same initiator ‘saint-
brothers’ or ‘saint-sisters’. The temple (casa-de-santo or terreiro) is conducted
by a saint-father and/or a saint-mother. Every initiate is potentially a future
cult chief, which is why initiatory parenthood is at the core of social networks
linking various temples through space and time, allowing the spread of
knowledge and a constant negotiation of power and leadership (Capone
1999). Xangô members worship two categories of ‘spiritual entities’:7 the

eguns, or family ancestors, and the orixás (pronounced ‘orishas’), the Yoruba
deities associated with natural elements like rivers, the sea and thunder – or
with human activities like hunting and iron-working. Due to the influence
of popular Catholicism, the word santo (‘saint’) is frequently used as a
synonym of orixá.
Possession and ritual practice

Possession by orixás is highly valued in the Xangô cult, in contrast with that
by eguns, which is expressly proscribed, since it is conceptualised as direct
contact with death, and thus a potential source of danger and contamination.
Egun possession is rare, whereas orixá possession is a constitutive element of
ritual life.8 Xangô members often describe the possession experience as ‘gratify-
ing’ (gratificante), underlining its positive personal and social consequences.
On the personal side, possession by orixás is a sign of a relational and affective
closeness between the orixá and its ‘son’ or ‘daughter’, leading to an intimate
feeling of self-confidence and protection. On the social side, possession is a
sign of the individual’s election by the orixá, leading to an enhanced reputation
in the religious community.
At least two orixás are ascribed to every initiate in the Xangô cult. The first is
called orixá-de-cabeça (lit. ‘orixá-of-the-head’), and the second juntó or adjuntó,
(‘joined together’). Every initiate has to worship his/her orixás by offering
them an annual sacrifice, and receiving them by possession. Every initiate can
be possessed by his/her orixás, but possession is not a condition for initiation,
nor its necessary outcome.9

Possession mostly happens during ritual events. It is very much to be

expected in public festivals (toques) and in private ceremonies such as sacrifice
(obrigação) and baths of leaves (aması́) only open to the members of one terreiro
and their invited affiliates. These private ceremonies are part of the initiation
process, where first possessions – called the ‘birth’ of the orixá – are frequent.
But they are also repeated every year for each initiate during festivals in honour
of their personal orixás. If we take into account the fact that a Xangô terreiro may
cater to between a dozen and hundreds of initiates, and that initiates from
one terreiro also take part in public and private ceremonies in affiliate terreiros,
possession opportunities are relatively frequent.
Possession may also happen in the daily life of the initiate, usually revealing
the dissatisfaction of the orixá, who has decided to punish his/her ‘son’ or
‘daughter’, for example by possessing him/her at work, in a public place, etc.

But extra-ritual possession, when it happens before initiation, can also be inter-
preted as the election of the ‘child’ by his orixá and, in most cases, as a call for
initiation (elective possession).
In all cases, contexts of possession occurrences play a determinant role in the
decision to encourage or, on the contrary, prevent a possession episode. Body
treatments and collective responses may differ sharply from one situation to
another. In this analysis, I will mainly focus on the most common situation of
possession by orixás, i.e. in the context of ritual. It is during such events that
the process of learning possession actually takes place.
Finally, it is worth noting that Xangô members are also aware that possession
needs some kind of learning. They insist that learning possession should only
take place when individuals present the first signs of an imminent possession
– what they call the stage of ‘irradiation’ (irradiaçã o) or ‘approximation’ (aprox-
imação), as we will see in more detail below. They talk about ‘indoctrinating the
body’ (doutrinar o corpo) in order to explain the process by which ‘irradiated’
novices are expected to learn how to dance, and to behave according to their
orixá’s archetypal and aesthetic prerequisites. On the other hand, another
discourse claims that orixás are omniscient, and this would explain why some
individuals – among them young children – ‘just don’t need to learn anything’,
as many initiators explained to me. Such a discourse is reinforced by the idea
that some individuals, because of their ‘blood’ inheritance, are predisposed to
entertain a privileged relation to the cult knowledge and to orixás, to such a
degree that even initiation would not be necessary (Halloy 2010). But what
interests us here is that Xangô members do not consider that the ‘irradiation’
stage of possession involves any learning process. If they have a precise idea

182 arnaud halloy
of when and how possession should occur, ‘irradiation’ is simply not a matter of
learning, but the result of the willingness and ‘acting’ (atuação) of an orixá on the
body and mind of its child. In this paper, I will try to show that this first stage of
possession is also, in my view, the result of a largely implicit emotional learning
process, characterised by specific ‘gut reactions’ and their expressive attunement
to cultural expectations.
Possession and Emotions

The idiom of possession
Xangô members distinguish between three stages of possession, based on a
set of behavioural patterns and psychosomatic states: first, the ‘irradiation/
approximation’ (irradiação/aproximação) stage that precedes the ‘full’ possession

state or ‘manifestation’ (manifestação). It is usually recognisable by an ‘introspec-
tive’ pause in movement (while singing and/or dancing), and a set of bodily
changes that will be examined later in more detail. Secondly, the ‘manifestation’
stage, characterised by the ‘full presence’ of the orixá, and which expresses itself
through dance and archetypal movements and behaviour. It should ideally lead
to loss of consciousness and to amnesia. The third stage, called axerado, refers to
the transitional state between possession and a return to normal. Those in an
axerado state frequently look as if psychologically ‘absent’ and/or physically
groggy; they may cry or laugh without apparent reason.
Within this general, three-stage classification of possession, a rich vocabu-
lary tries to capture the distinct bodily states associated with the possessed/
orixá interaction according to their quality and intensity. Xangô members
say that the orixá ‘manifests himself’ (se manifesta) in his son or daughter, or
that he ‘takes’ (pega), ‘embodies’ (incorpora) or ‘comes down’ (baixa). One
can frequently hear that the ‘child’ ‘receives’ (recebe) his/her orixá or, more
generally, that the orixá ‘acts upon’ (atua) his ‘child’. This set of expressions
underlines the imperious action of the orixá and the passive attitude of the
possessed person.
While this first lexicon depicts the ‘manifestation’ stage, a second set of
words and expressions emphasises the somatic states associated with the
‘irradiation’ stage. A frequently used spatial metaphor says that the orixá ‘gets
close’ to his ‘child’ (se aproxima10). He can do this only slightly (de leve) or, on
the contrary, suddenly and/or violently, ‘nailing down’ [se cravar] his ‘child’
or ‘taking him/her in one go’ (pegar de vez). But independently of the way
the encounter between the orixá and the body of his ‘child’ takes place, the

closest the orixá gets to his ‘child’, the more intense and vivid the somatic
changes.
Based on Xangô members’s possession self-reports, I will show in the follow-
ing section how they conceptualise the orixá/materia11 interaction as a dynamic
continuum of changes in the body. If, on the one hand, manifestação is charac-
terised by the full presence of the orixá in the body of his ‘child’ – leading
(ideally) to the latter’s loss of consciousness – irradiação, on the other hand,
highlights a large variety of somatic states, varying from shivering and goose-
bumps to irrepressible trembling, heat waves, outbursts of crying, etc.
Self-reports of possession
One self-report is by Maria-Helena, a young saint-mother, ‘daughter’ of

Yansã12, describing her first aproximação:
I felt an emotion, something like that, an emotion that made me feel like wanting to
cry . . . But I was not really ‘irradiated’ . . . I did not really feel the orixá. When I did
come to feel [the orixá] it was in Paulo’s terreiro, in my grandfather’s house. It was
Oxum’s13 celebration. They started to sing for Yansã and I felt such lightness
[leveza], and that’s it. Then I fainted [lit. ‘I lost my senses’] . . . But I was not
‘manifested’. I felt as if I was fainting . . . I felt my blood as if my blood pressure
was going up or down . . . I just know that I felt my body fainting . . .
What strikes one in this description, is the emphasis on ‘feelings’, but also
Maria-Helena’s difficulty to describe them more precisely. She starts by
saying she felt an emotion ‘that made her feel like crying’, but immediately spe-
cifies that it was not really an ‘irradiation’. And a little later she describes what
she considers to be an ‘irradiation’ through body arousals such as ‘lightness’,
‘blood pressure going up and down’ or ‘losing her senses’. Maria-Helena’s
account also clearly expresses two fundamental elements of the possession
experience: the passiveness felt by the possessed – something ‘happens to’
them – and the lack of control over the changes that occur in their own body.
Kleyde, a 45-year-old initiate, daughter of Xangô, remembers her first
‘irradiation’ episode:
I became sad . . . My heart accelerated, a heat wave invaded my torso . . . I really

believed I was about to die . . .
In this short account, another constitutive element of many irradiation experi-

ences is clearly expressed: they are mythologically informed. Kleyde is a

184 arnaud halloy
‘daughter’ of Xangô, the orixá of thunder and also associated with fire, well-
known for his impetuous character. As with many initiates of this orixá, he
manifests his presence in their body through ‘heat’, sometimes expressed as
‘heat waves’ or, as another of his ‘daughters’ once told me, as a sensation of a
‘fire inside the torso’ (fogo dentro do peito). But bodily arousals associated with
the ‘irradiation’ stage are not crystallised experiences: they are context-sensitive,
and may change according to the initiate’s familiarity with possession manifes-
tations and conceptual background. Tarcı́sio, a young adult and experienced
initiate, complained about the inappropriate occurrence and the stagnation of
his bodily arousal during the ‘irradiation’ stage:
I frequently feel intense shivers [arrepios muito fortes], but unfortunately for other saints
than mine! . . . And my orixá, when he gets close [se aproxima], he doesn’t act totally
[ele não atua totalmente] . . . I just have dizzy spells [arrepios] and things like that . . .
Until today, I’ve never been totally manifested . . .
In the last testimony presented here, Yguaracy, an experienced priest, 40 years

old, comments on the first ‘irradiation’ of his juntó:
I felt it was not an ‘approximation’ of my mother Oxum [Yguaracy’s main orixà]. It

was an ‘approximation’ of my father Orixalá14 [his juntó] . . . It was something different
. . . And it took time for me to realise what was happening in my own cult house,
because I asked myself: ‘Ave Maria, do I have Parkinson’s disease?’ Because my
muscles started trembling . . . It was something different . . . And I think . . . I’m
sure that it was the first time I’d felt something like that with Oxalá15 . . .
Through these first person testimonies, we learn that orixás ‘act upon’ (atuam)
the bodies of their ‘children’, provoking qualitatively distinct bodily sensations
from chills/shivers/goose-bumps (arrepios) to changes in heart rate and respir-
ation, trembling, dizziness, prickling in the hands and feet, uncontrollable shi-
vering, loss of balance, hot flushes, ‘blood boiling’, wanting to cry, ‘tears in
the eyes’, raising or decreasing blood pressure, sweating, ‘losing one’s senses’,
visual and/or auditory disturbance.
We should also notice that in the Xangô cult possession is not understood as
a unique and monolithic state, but rather as a continuum of mental and bodily
states oscillating between consciousness and unconsciousness, and between
feeble and intense somatic changes varying from vague to specific sensations.
In such a conceptual framework, profound loss of consciousness and total
amnesia refer to an ideal state – and also a ‘theologically correct’ discourse

(Barrett 2004: 11) – associated with the ‘full presence’ of the orixá into his child’s
body, but not to the ‘irradiation’ stage, which is much more allotropic and
dynamic.
I return now to my initial question: re emotions constitutive elements of pos-
session – and more specifically of the ‘irradiation’ stage?
Possession as ‘uncanny’ gut feelings

I will start by arguing in favour of a ‘Jamesian’ theory of possession. In
essence, the position taken by neo-Jamesian theorists of emotion is that
emotions are, in one way or another, constituted by changes in body state
responding to evoked thoughts or perceptions of the direct environment
(Damasio 1999; Prinz 2004; Robinson 2007). As William James argued more
than a century ago:
If we fancy some strong emotion, and then try to abstract from our consciousness of
it all the feelings of its characteristic bodily symptoms, we find we have nothing left
behind, no ‘mind-stuff’ out of which the emotion can be constituted, and that a
cold and neutral state of intellectual perception is all that remains. (1884: 193, my
emphasis)
James’s argument about emotions is, in my view, directly relevant to posses-

sion: possession cannot be defined without considering somatic changes as
one of its constitutive elements. In other words, I consider emotional
arousal as a precondition for possession.16 The first argument to which I
will come back in the conclusion is theoretical: we cannot abstract emotions
from the conceptual framework of spirit possession if we seek to explain its
success in so many religious cultures throughout the world. My second
(and main) argument in favour of this hypothesis is ethnographic: specific
bodily arousals are core elements of possession, and seem to correspond, as
I will show in the remaining, to what perceptual theorists of emotion call
‘gut reactions’ or ‘gut feelings’.
Gut reactions correspond to what contemporary philosophers Robinson
(2007) and Prinz (2004) respectively call ‘affective/non-cognitive’ and ‘embo-
died’ appraisals, emphasising that gut reactions are spontaneous, ‘crude’ or
‘dirty’ body reactions that need a further process of cognitive evaluation in
order to be more precisely identified. Following Judith Becker’s hypothesis, I
suggest that gut reactions at the core of possession are of the same kind as
musical or aesthetic emotions:

186 arnaud halloy
Musical emotion is still rooted in basic physiological arousal felt in the body and dis-
played by tears, chills/shivers, goose bumps, palpitation of the heart, and perspiration
. . . (2004: 52).
But the similarity with aesthetic emotions has its own limitations. First, if music
plays a central role in triggering and regulating possession (Rouget 1990; Becker
2004), many other factors such as smells, flavours, touch, images, sounds, body
techniques, memories, evocations . . . are potential candidates for provoking the
initial emotional reaction associated with possession. Secondly, Becker’s
description of musical emotions does not exactly match the description of
the ‘irradiation’ stage in the Xangô cult, when she argues that musical emotions
are selectively focused on some form of happiness or sadness, or mixed
emotions such as ‘bitter sweetness’ or ‘beauty and pain’ (ibid: 52). Gut reactions
associated with the ‘irradiation’ stage are not of this kind, nor are they ordinary
emotional reactions. What Xangô members are dealing with could be described,
in Levy’s (1973: 151) terms, as ‘uncanny feelings’17, such as particularly intense
‘goose-bumps’ or persistent ‘shivers’ as well as unusual and largely undeter-
mined combinations of feelings, leading people to give confused and changing
descriptions (see Maria-Helena’s self-report above). Thirdly, and even more
importantly, what gives the affective appraisal of possession its special flavour
is the possible consonance between bodily sensations and the mythological
identity of the orixá supposed to be acting upon the body of his ‘child’. As
both Kleyde and Yguaracy state sensations associated with their orixá’s
‘approximation’ match their orixá’s mythological trait: ‘heat waves’ for Xangô,
the orixá of thunder, trembling (and often pain) for Orixalá, the old orixá.
Fourthly, the consciousness of ‘irradiated’ persons’ is, in many cases, altered
by other psychological states such as dissociation (Lapassade 1997; Budden
2003; Seligman and Kirmayer 2008), hypnotic states (de Heusch 1993; Hell
2008) and absorption,18 which demand of individuals that they ‘attend more
carefully to their inner sensations and may intensify a detachment between
inner sensations and an external world’ (Luhrmann 2007: 99).
Far from being contradictory, the difference between ‘crude’/undifferentiated
(like intense ‘goose-bumps’ or persistent ‘shivers’) and ‘orixá-specific’ arousal
states (like heat waves) reflects, in my view, two distinct moments in the
development of the emotional process of learning possession. As the possession
idiom and self-reports suggest, possession in the Xangô cult starts with some
kind of intense emotional arousal. Later on, once participants become
increasingly familiar with possession episodes and with the conceptual and

mythological background, gut feelings may become more and more ‘encultu-
rated’, leading to orixá-specific ‘somatic signatures’. In the second part of
the paper I will focus on the learning process underpinning such bodily
enculturation. My claim is that emotional responses constitutive of the process
of learning possession in the Xangô cult become more acutely discriminated,
classified and generated thanks to a process of coupling between the affective
and cognitive appraisals which take place during ritual activity. To defend this
hypothesis, we need to go a step further in our ethnographic description and
define more precisely what kind of emotional elicitors trigger possession in the
Xangô cult, and how these are articulated within the pragmatic conditions of
ritual action as well as in the imaginative process mobilised during possession
episodes.
Emotional learning
In October 2002, I attended an aması́ of an Ode initiate – the orixá-hunter –
at the terreiro of Paulo, a renowned saint-father, in João Pessoa. When Paulo
started to sing for Ode, after preparing the decoction with fresh leaves, my
hair stood on end and I felt an intense and lasting shiver. An old initiate who
was standing by my side put her hand on mine, smiled and told me: ‘It won’t
take long, it won’t take long . . .’
This episode, even if apparently anecdotal, is an interesting starting-point
for our reflection, since it is paradigmatic of the process of learning possession
in at least two ways. As I will show, it highlights the tight bond between
‘approximation’ and ritual features, and points to the crucial role of social
environment in the self-identification and regulation processes in possession.
In other words, participants at the periphery of ritual action – mere ’observers’
– are exposed, even in a softer manner, to the same stimuli inherent to ritual
features, and to the same ’social signals’ than initiates directly involved in the
ritual. A sacrifice episode will help us demonstrate these two dimensions of
learning possession.
Obrigação for the Ogun of Taı́sa (May 2003)
On her knees in front of her orixá’s altar, Taı́sa, a 14 year-old initiate, presents the
first signs of an ‘irradiation’ at the beginning of the sacrificial song: staring into
space, her chest vibrating evenly, discreetly rocking to and fro, with a grave and
severe facial expression typical of many male orixás. When the animal’s blood
was poured over her head and shoulders, her whole body shivered and started to

188 arnaud halloy
tremble intensely. Then her initiator invited her to hold the animal’s severed head
by herself. She obeyed with difficulty, as if she had – at least partly – lost
control of her own movements. Her mother, who was standing at her side,
helped her in her task. After adding salt and palm oil to the cut throat, the initiator
asked Taı́sa to drink the blood from it. At first she touched the warm flesh with
her lips half-heartedly, but within a few seconds, she behaved more enthusiastically
until she avidly plunged her face into the meat. Such a change in Taı́sa’s behaviour
was warmly welcomed by those present, who saw in it an intensification of
the orixá’s ‘acting’ (atuação) upon his ‘daughter’. The initiator then switched from
the sacrificial to the orixá’s song repertoire. The initiate responded instantaneously:
her head shook violently from side to side and the limited motion of her
torso became powerful and rhythmic, pulsating at the same cadence as the
audience’s clapping and the small percussion instruments that had just entered.
From that moment on, people began joyfully celebrating the presence of Ogun,
who would be invited – after finishing the sacrifice – to stand up and dance
facing his initiator for a few minutes, before being released until the public ceremony
next day.
The ritual unfolding of possession as described in this obrigação episode brings

to light recurrent patterns of ritual features which play a central role in the trig-
gering but also in the cognitive framing of emotional responses associated with
possession episodes.
Ritual features and ‘perceptual attractors’

Among the ‘obvious’ features of ritualised behaviour (Rappaport 1999), as
recently described by Liénard and Boyer, three are directly relevant to the
present case study. First, ritualised actions are literal and rigid: ‘What is impor-
tant is that people strive to achieve a performance that matches their represen-
tation of past performances, and that they attach great emotional weight to any
deviation from that remembered pattern’ (2006: 816). I see two distinct charac-
teristics in this first feature: there is a correspondence with what Humphrey and
Laidlaw (1994) call ‘archetypality’, meaning that ritual actions are ‘perceived as
discrete, named entities, with their own characters and histories’ (1994: 89). And
there is also a high degree of stability of ritualised actions – at each occurrence
actors tend to perform them in a very similar manner, and their organisation is
largely constrained by a quite rigid and complex ritual syntax (Lawson &
McCauley 1990) or script (Boyer 1997).
The second feature is the redundancy of ritualised actions. As already
stressed by Lévi-Strauss (1971), this redundancy is a constitutive feature of

such actions: the same actions or gestures are sometimes repeated a number of
times. I would add that if ritualised actions and gestures are usually redundant,
they might also be regular, meaning that the whole ritual is frequently repeated
in a determined period of time, as is the case with private and public ceremonies
in the Xangô cult.
The third ‘obvious’ feature of ritualised actions is that they are spatially and
temporally delimited: within a determined space and time, ordinary behaviours
are ‘suspended’, and ‘particular acts’ can be thus performed.
My claim is that recurrence in ritual behaviour is accompanied by recurrence
in the unfolding of possession episodes. More precisely, the three features of
ritual action mentioned above are propitious for eliciting possession, thanks
to ‘perceptual attractors’, i.e. perceptions that have a highly emotional and evocative
potential.19 Among perceptual attractors in the Xangô cult we find blood and
aması́20 with their singular smell, taste or tactile sensation when they are
poured over the head and shoulders, but also specific bodily techniques such
as kneeling down in front of the orixá’s altar as well as invocations and songs
for the orixás. All these elements of ritual action are propitious for triggering
powerful inferential and emotional responses in initiates by coupling sensory
stimulation (tactile, auditory, kinesthetic, visual, olfactory, vestibular 21and gus-
tative) – the ‘sensory capture’22 – with representations and images about pos-
session and orixás’ attributes and mythological traits (the orixá’s identity, his
character and psychological profile, what he did to humans and to other
orixás in myths, etc.) – the ‘imagination capture’.
In the emotional learning process at work in possession, perceptual attractors
can be even more powerful or efficient because they remain ‘invisible’ to
initiates. One can provide at least three reasons for this, starting with the
large and dynamic diversity of emotional elicitors (smells, rhythmic patterns,
songs, bodily treatments. . .etc.) present in most rituals where possession is
expected (aması́ obrigação, feitura, toque). This diversity of ritual stimuli as well
as the ‘sensory overload’ (Cox 1969: 110, cited in Gell 1980: 233) it may entail,
prevent a clear apprehension of what actually triggers the emotional response.
A second reason for the invisibility of emotional stimuli may result from the type
of emotional reaction involved. ‘Embodied appraisals’ (Prinz 2004) or ‘affective
priming’ (Zajonc 1980; Murphy and Zajonc 1993) happen very fast, automati-
cally and below the threshold of conscious awareness, triggered by partial
and incomplete perceptual information rather than clearly conscious represen-
tations (LeDoux 1996). In other words, pre-attentive or unconscious stimuli
(Leventhal and Scherer 1987) would play a central role in eliciting the gut

190 arnaud halloy
reactions associated with the initial stage of possession. The invisibility of

emotional elicitors might also be reinforced by a shifting of attention from exter-
nal to internal factors, which may happen in two ways: through ‘absorption’
which seems constitutive of possession phenomena (Luhrmann 2004; 2007;
Luhrmann et al. 2010), and makes the ‘irradiated’ person focus on his somatic
and mental states, not on environmental features; or through ‘event coding’
of possession (Frijda & Mesquita 1997) conceptualised as the atuação or aproxi-
mação of an orixá, which makes those ‘irradiated’ understand and evaluate their
own possession according to the intensity and quality of their bodily sensations.
Because of their largely invisible nature, perceptual attractors reinforce the
‘obviousness’ and ‘reality’ of possession by giving it a bodily and emotional
foundation, and at the same time they trigger a powerful ‘evocative process’
(Sperber 1974) – the ‘imagination capture’ – by coupling the sensations and

bodily states – the ‘sensory capture’ – with potential cultural meaning and
expectations.
More specifically, imagination capture might lead to what Rouget (1990)
calls ‘identification’ with the deity thanks to the highly evocative content of
an orixá’s song and invocations. As de Carvalho (1993) convincingly demon-
strates, Xangô musical repertoire is composed of two functional groups of
chants that correspond grosso modo to two distinct phases of the ritual
process: ‘functional songs’, that are above all structuring elements of ritual
syntax, and ‘songs for the orixás’, performed during bodily treatments of the
initiate in private rituals as well as during public festivals. Songs for the orixás,
according to de Carvalho, are ‘much more emotional, dynamic, and energetic,
especially during trance occasions, when the presence of gods is celebrated
with joy’ (ibid: 205) unlike functional songs that are performed with some
kind of ‘musical indifference’, characterised by its distance ‘from the world of
emotions that each ritual creates’ (ibid). Furthermore, as Segato stresses, it is
worth noting that Xangô members also discriminate each orixá’s repertoire
from the others ‘for the state of mind it generates and for the emotions and sen-
sations it triggers when it is performed’ (2000: 237). As we can see in the obriga-
ção episode above, the triggering or intensification of possession episodes
frequently happens when the initiator shifts from the functional repertoire to
the orixá repertoire. Consequently, songs for the orixás not only provide a sym-
bolic content of the ritual by suggesting identification with the orixá, but are also
able to trigger or intensify the initial emotional response.
Partly in the Yoruba language, partly in Portuguese, the invocation of an
orixá is also highly evocative and emotionally coloured in the Xangô cult,

because of its crucial role in ‘calling’ the orixá. When possession is expected, the
saint-father will start invoking his initiate’s orixá, summoning him to ‘come
down’ and ‘take’ his/her ‘child’. Like orixá’s songs invocations may provoke
or intensify the initial body arousal and, additionally, they might also inform
the possessed on the unfolding of his/her own possession. In other words,
they are part of a series of social signals communicated by the initiator and
the audience for interpreting and regulating what is happening (emotionally)
to the ‘irradiated’ person.
‘Social referencing’ and emotional regulation

‘Social referencing’, the second way by which meaning is conferred to
sensory capture, may be defined as a complex skill in which another person’s

emotional expressions and behaviour can be used as information about
mostly ambiguous external events, involving ‘an active and complex process
of appraisal and judgment rather than a merely passive contagion of emotion’
(Camras & Sachs 1991: 27 – 8, cited in Walker-Andrews 1997: 444). During
ritual action, as we can clearly see in the obrigação for Ogun, when a person
shows the first signs of ‘irradiation’, everybody starts looking at him/her,
calling the orixá, enthusiastically singing and dancing so as to intensify the
‘approximation’ of the orixá. The initiator, more than anyone else involved,
focuses his attention and conducts the initiates’ behaviour by calming them
down or, on the contrary, by intensifying the ‘irradiation’ through his own affec-
tive and bodily commitment. Social referencing thus plays two major roles in
learning possession. It is mainly thanks to the reactions of the audience
toward him that the ‘irradiated’ person learns to appraise his own somatic
signals correctly and to distinguish between common emotions and the
somatic signs of an aproximação. Social referencing also acts as an emotional reg-
ulator: the ‘irradiated’ persons learn to react appropriately to what is happening
‘inside’ them. This has been called the ‘canonical manifestation’ of the emotion-
al regulation: ‘first, one has an emotion, then one regulates it’ (Campos et al.
2004: 385). The following episode about Rafael, an initiate around 25 years
old, clearly shows this regulation process at work during ritual activity.
Iyanle´ for Xangô 23 (July 2003)
The sacrifice that day lasted more than 6 hours. Everybody was exhausted. When
Paulo, the saint-father started the Iyanlé for Xangô, he decided to go directly to the
orixá’s repertoire, showing his wish to see Xangô ‘coming down’. Musical instruments

192 arnaud halloy
were brought in and, copying the old saint-father’s enthusiasm, the assembly threw its
last energies into singing and dancing for the ‘beloved orixá’. Nina, an experienced
initiate, ‘received’ her Xangô, who started to dance vigorously. Many other initiates
presented the signs of ‘irradiation’. Among them, Rafael, son of Xangô, who sat on
a chair and started to cry ostentatiously. Within half a minute, before his saint-
father was able to calm him down, he stood up violently from his chair and started
to dance wildly, and in a quite unusual way: he alternated between wide, powerful
and uncoordinated movements, moving fast across the room, and almost complete
immobility, during which he was still crying and shouting repetitively but also reco-
vering his breath. However, something was wrong with Rafael’s Xangô: the two
young drummers were obviously failing to take this ‘manifestation’ very seriously,
exchanging ironic smiles. His saint-father finally managed to calm him down, and
took him into the peji, the room with the orixá’s altars. He returned to the main
room a few minutes later, as if nothing had happened.
How did Xangô members interpret Rafael’s possession?

According to the young musicians accompanying the ritual, there was no doubt
about it: ‘It was an èke´!’, a simulated trance. For one of them, the orixá’s unusual
behaviour (extreme agitation, alternately walking and stopping, repetitive
shouts), his rapid exhaustion and Rafael’s reappearance after only a few
minutes without any trace of the orixá’s passage inscribed in his face or body,
as is to be expected after violent possessions, confirmed his impression of a
simulation.
But the most interesting interpretation came from Júnior, a 25-year-old saint-
father. He simply told me: ‘He [Rafael] was filled with enthusiasm, so it was not
well resolved . . .’ (Se entusiasmou, então ficou mal resolvido . . .)
According to Júnior, the Rafael episode was not a ‘false trance’, but involved
bad emotional regulation. In other words, he recognised Rafael’s ‘manifestation’
as legitimate, while also emphasising that Rafael had not adjusted appropriately
what he actually felt – his ‘uncanny’ feelings – to what was (culturally) expected
of him in terms of expressive behaviour. Instead of throwing himself desperately
into the dance, he should have stayed calm, adopting a more introspective
attitude, waiting for an intensification of his orixá’s atuação. His saint-father’s
reaction identified the maladjustment of the initiate’s emotional state. But the
issue was not only about proper (normative) behaviour. As we can observe in
this particular case, possession learning is not just about learning how one
should behave – the ‘display rules’ (Ekman & Friesen 1969) – or how to feel
in certain contexts – the ‘feeling rules’ (Hochschild 1979), – but also how to

react to such feelings appropriately (Campos et al. 2004; Eisenberg & Spirad
2004; Spinrad et al. 2006).
The dynamics of perceptual attractors

The process of emotional learning in possession is not crystallised once and
for all: it changes as the possession idiom and episodes of possession become
increasingly familiar. In this last section, I describe a tendency observed in the
development of emotional expertise, as well as certain potential strategies
developed by the experienced possessed in order to provoke or intensify their
own possession.
By following the learning process of various Xangô members’ during my
fieldwork in Recife, and comparing cult leaders’ and novices’ expertise in pos-
session, I identified two tendencies in the evolution of their sensitivity to
emotional elicitors leading to possession. The emotional response to stimuli
seems to improve with time – possession on the part of experienced initiates
requires less intense sensory stimulation than that of novices. It might well be
triggered by just one element present in the contexts of first possessions, or
even, as we will see, by imagined or remembered situations. At the same
time, such individuals develop better control over their own possession
through ‘auto-induction’ techniques or, on the contrary, by resisting their
orixá’s manifestação (even if only temporarily).
The convergence of these two tendencies – increased sensitivity to isolated
emotional elicitors and better control over emotional reaction – might look
contradictory at first sight, but it is not. It reveals, in my view, a growing percep-
tual expertise, as well as an increasing mastery of regulating emotion during pos-
session. Such emotional self-regulation helps us understand, for example, why
the possession of cult leaders and experienced initiates occurs with such fluidity
and, most of the time, in expected moments. Many external factors can inter-
vene in the development of expertise in possession. The age of the first
contact with possession is very important. The youngest person possessed
that I was able to observe was 6 years old in the Gege cult.24 His orixá – Oba-
luaye´25 – was impressive and his possession convincing. The expertise of reli-
gious leaders might also play a decisive role: some develop a particular talent
for triggering and developing adequate emotional reactions in others: vocal
inflexions while calling the initiate’s orixá, authoritative staring, repetitive
and/or specific gestures or body treatments are all potential elicitors of the
initial arousal (‘irradiation’) stage. Some experienced ilu26 players are also

194 arnaud halloy
well known for being skilful inducers of possession thanks to their musical abil-
ities.
However, we can also identify at least two ‘internal’ factors for triggering
possession. Even if it is hard to evaluate precisely how active people are in
their own possession, observations and self-reports by experienced initiates
indicate two closely related ways of triggering and/or enhancing the arousal
of possession.
The first way is a cognitive factor. Luizı́nho, a young saint-father, once
told me:
When I dance for Xangô, I see my mother with her orixá on my side. I do it and these
images come to me, of my mother with Xangô, and then I ask him a lot of good things
for her.
Paulo, the patriarch of the saint-family I studied, emphasised the affective

dimension of such images:
To sing for the orixás is not only singing . . . It is singing and having emotions
. . .When I sing I think of my father, my grand-father [both renowned cult leaders]
. . . And my entire body is shivering! This shivering is so strong! It’s a strong
emotion . . . And it is not [because] the orixá [is] close to me . . . It’s [just] me thinking
about my father . . .27
Luizı́nho’s and Paulo’s reports underline the possible role played by imagin-
ation or, more precisely, by emotional memories in triggering or intensifying the
initial emotional response to possession. But they also suggest that such
emotional memories are themselves embedded in bodily techniques such as
singing and dancing. My observation of initiates during public festivals tends
to support this idea; we can frequently observe that initiates get more intensely
involved in singing and dancing when the cult leader starts to sing for his/her
personal orixá, i.e. the ritual sequence when their possession is actually
expected. We can thus infer from these recurrent situations that increased
bodily commitment in dancing and singing might also intensify the absorption
and emotional processes involved in possession, by enhancing the coupling
between kinaesthesic sensations – the sensory capture – and imagination.
Conclusion
On the basis of this ethnographic analysis, I will advance at this point several
theoretical claims about emotions as constitutive elements of possession.

First, if the emotional states at the origin of possession episodes are ‘uncanny’
gut reactions, as I argued here, cultural influence might go as deep as to ‘stamp’
largely spontaneous emotional responses. As Mauss (1936/2003) and many of
his successors (Bourdieu 1972; Parlebas 1981; Warnier 1999; Crossley 2004; Wac-
quant 2005) have argued, the ‘habits of the body’ are social-biological-psycho-
logical constructs. The present case study is an invitation to enlarge this
category from sensory-motor or ‘kinesthesic socialisation’ (Bateson 1975: 152)
to deeply rooted bodily states such as gut reactions. These ‘somatic signatures’
of the orixá, as I call them, are not prior to cultural categorisation as they directly
contribute to the ‘identity conditions of our emotions’ (Prinz 2004: 143). In the
case of the Xangô cult, learning possession means first of all learning to know
how it feels like to be ‘irradiated’ by an orixá and to react appropriately to
such emotional reactions. Thus the ‘socially informed body’ (Bourdieu 1972/
2000) is also an emotional body, undermining the evidence of a clear frontier
between biological and cultural dimensions of emotional learning in real-life
contexts (Wikan 1990; Desjarlais 1992; Leavitt 1996; Surralès 2003; Gieser 2008).
The second implication, as already suggested by Levy (1973), is that emotion-
al reactions are not necessarily felt and expressed through explicit emotional
words or categories. If gut feelings, in the present case, are triggered by the
action of a deity upon one’s body, the appropriate cultural idiom might be
largely metaphorical, translating the quality and intensity of body sensations
associated with possession into ‘physical’ or ‘proxemic’ terms (irradiação or
aproximação respectively).
A third theoretical implication exposes the limits of the cognitive approach
in anthropology, as argued in the introduction to this article. In this ethno-
graphic study I underline the pivotal role of ritual practice in triggering and
shaping emotional reactions associated with possession. I also delineate two
processes potentially lying at the core of this mechanism: a coupling process
between sensory and symbolic captures, spurred on by specific features of
ritual action such as archetypality, rigidity, redundancy, regularity, spatial and
temporal delimitations, and a social-referencing process through which
emotional and behavioural reactions towards ‘irradiated’ persons act as reliable
indicators of how to react to emotional states associated with possession. As a
consequence, one could argue that religious practice contributes in an essential
manner to cultural transmission by anchoring cultural ideas about possession in
deeply rooted emotional reactions as well as in social interactions. If one wishes
to provide a ‘synthesised explanatory account of spirit possession’ (Cohen 2007:
96), one should take into account that possession is lived through patterns of

196 arnaud halloy
affects, percepts and interactions which along with the conceptions of posses-
sion that cognitive approaches have rightly identified (Cohen 2007, 2008) are
good candidates for explaining its wide success around the world.
A fourth implication refers to the importance of a diachronic approach to pro-
cesses of learning possession. As I suggest in this article, Xangô members might
become more active by triggering or intensifying their own ‘irradiation’ thanks to
active affective recollection and/or intensification of body commitment in ritual
action. This requires long-term involvement with Xangô rituals, and increasing
familiarity with the possession experience and its mythological background. At
the perceptual level, more experienced participants demonstrate a growing
sensitivity to isolated emotional elicitors, as well as better control over their
emotional reactions. At the emotional level, their bodily reactions seem to
evolve from largely undifferentiated arousal states to ‘orixá-specific’ somatic

signatures. Finally, as illustrated by Rafael’s episode, a necessary step in the
process of learning possession consists in correctly evaluating and adjusting
emotional responses to highly conventional and expressive behaviour.
As I have showed in this paper, the process of learning possession in Recife’s
Xangô cult offers a striking example of the close and dynamic interaction
between thinking and feeling and of the psychological and pragmatic features
of learning contexts. It puts forth a novel approach, cognitive ethnography,
which brings together the naturalist and ethnographically informed approaches
to culture.
Acknowledgements
I am very grateful to Vlad Naumescu, Joël Candau, François Berthomé and three
anonymous reviewers for Ethnos for many useful comments and criticisms of
earlier drafts of this paper.
Notes
1. My use of ‘possession’ in this article will refer primarily to ‘possession trance’, as it is
defined by Bourguignon (1976). See the introduction to this special issue for an
historical and epistemological discussion on possession studies.
2. In Brazil, academic interest in possession started at the end of the nineteenth
century with Rodrigues’s work (1900), which initiated a psycho-medical approach
developed up to the end of the 1930s by Ramos (1934), Querino (1938), Cavalcanti
(1935) and Fernandes (1937), for whom trance-like phenomena were mostly ident-
ified with psycho-pathological states. This approach suffered a serious setback
when influential culturalist Herskovits (1943, 1958) emphasised the normal, cultural
character of religious possession. Most anthropologists studying Afro-Brazilian
cults follow this line, among them Bastide (1958) and Ribeiro (1950 –51, 1978), both
of whom insist on the socially adaptive character of possession. Contemporary

studies explore dimensions of possession from diverse perspectives: post-structural

(Goldman 1987), socio-economic (Motta 1993), sociological (Ortiz 1990), psycho-
logical (Augras 1983; Segato 1995; Lepine 2000), ethno-musical (de Carvalho
1993), socio-political (Dantas 1982, 1987; Maggie 1988; Boyer-Araujo 1993; Capone
1999), phenomenological (Wafer 1991), socio-historical (Johnson 2002) and
cognitive (Cohen 2007, 2008).
3. For a brief overview of original approaches to possession in recent anthropology,
see our introduction to this special issue.
4. I develop this last dimension in a forthcoming paper (Halloy 2012).
5. First coined by Hutchins (1995), the term ‘cognitive ethnography’ is currently used
to define a specific way of thinking about cognition in a situational and distributed
framework, with a particular interest for spaces inhabited by tools and material
devices (Hollan et al. 2000)
6. This research draws on first-hand ethnographical data, gathered during 14 months
of fieldwork between July 2001 and September 2003.
7. The vernacular expression ‘spiritual entities’ designates all the spiritual beings
present in Afro-Brazilian religions.
8. Many Xangô temples also worship ‘spiritual entities’ (Exus, Caboclos, Mestres etc.)
from the Jurema cult, a well-liked and quite popular Afro-Amerindian cult in
Recife. In the saint-family I studied, however, Jurema and Xangô are strictly separ-
ated in time and space, the former remaining at the margins of candomble´ which is
strongly inscribed in a prestigious kinship sustained by an ‘agnatic’ conception of
transmission (Halloy 2010). For these reasons, my analysis will focus on possession
by African deities.
9. “Contrary to other modalities of Afro-Brazilian cults (Verger 1957; Bastide 1958;
Elbein dos Santos 1975; Vogel et al. 1993; Vatin 2005), almost no explicit teaching
takes place during initiation in the Xangô cult. What novices are concretely
exposed to is an intense impregnation process of what makes the materiality of
gods (sensations, emotions, interactions with persons, substances and artefacts),
oriented towards the transformation of their sensitive experience of the spiritual
world (Halloy 2005).
10. This spatial metaphor also connotes an affective closeness between the child and his
divinity, underlining the ‘intimate’ and affective nature of the relationship between
the orixá and his ‘child’ (Opipari 2004; Halloy 2007, 2009).
11. When referring to possession episodes, Xangô members usually designate the
initiate’s body by the expression ‘material’ (materia).
12. Yansã is the goddess of storms and wind.
13. Oxum is the goddess of soft water and of fertility.
14. Oxalá is an old orixá, sometimes considered as the father of all orixás.
15. Orixalá and Oxalá are synonyms in the Xangô cult.
16. In this paper I will not discuss the cases of ‘false’ possessions, which are part of most
‘elective’-possession cults (Leiris 1958). My point here is that most good ‘pretenders’
in the Xangô cults are also ‘ possessed for real’, because one needs first to learn what
a real possession is in order to simulate it accurately.
17. ‘Uncanny feelings’ can be defined as a type of feeling specific to the context of the
‘unknown’, the ‘strange’ or at least the unusual (Levy 1973: 151).

198 arnaud halloy
18. ‘“Absorption” is best understood as the mental capacity common to trance,

hypnosis, dissociation, and much other spiritual experience in which the individual
becomes caught up in ideas or images or fascinations’ (Luhrmann et al. 2010: 75).
19. A perceptual attractor can also be defined as a perceptual salience associated with
semantic elements and emotional intensity.
20. Aması́ or agbó designates the plant decoction prepared during the aması́ ceremony.
21. In a seminal article, Gell (1980) suggests that ‘assault on the equilibrium sense’ in
religious practices of the Muria is a central ‘psycho-physiological mechanism’
leading to possession.
22. I borrow this expression from Hell (2008: 20), giving it my own interpretation.
23. Ritual sequence of food offering for Xangô.
24. The Gege cult is another Afro-Brazilian modality of cult in Recife.
25. An orixá healer, Obaluayé is also associated with skin diseases and land (terra).
26. Ilu means ‘drum’ in Yoruba. In the Xangô cult, powerful rhythmic patterns are
played on three ilus, the bass one improvising in time with the orixá’s dancing.
27. Paulo is not allowed to be possessed because he is a babalaô, an oracle expert in the
Xangô (Halloy 2010). Paulo’s case is particularly interesting because it shows how
very similar emotional response to similar events and situations can be interpreted
and treated within the same cult in distinct ways according to the social status of
the affected person.
References
Aubert, L. 2004. Les Feux de la Déesse. Rituels villageois du Kerala (Inde du Sud). Lausanne:
Payot.
Augras, M. 1983. O duplo e a metamorphose. Petrópolis: Vozes.
Barrett, J. 2004. Why Would Anyone Believe in God? Cognitive Science of Religion Series.
Walnut Creek: Altamira Press.
Bastide, R. [1958] 2000. Le Candomble´ de Bahia. Transe et possession du rite de Candomble´
(Bre´sil). Paris: Plon.
——. 1960. Les Religions africaines au Bre´sil. Paris: Presses universitaires de France.
Bateson, G. 1975. Some Components of Socialization for Trance. Ethos, 3(2):143–155.
Becker, J. 2004. Deep Listeners. Music, Emotion, and Trancing. Bloomington and
Indianapolis: Indiana University Press.
Bourdieu, P. 1972/2000. Esquisse d’une theórie de la pratique. Paris: Seuil.
Bourguignon, E. 1976. Posssession. San Francisco: Chandler & Sharp Series in Cross-
Cultural Themes.
Boyer, P. 1997. La religion comme pheńomène naturel. Paris: Bayard.
——. 2001. Et l’homme creá les dieux. Comment expliquer la religion. Paris: Robert Laffont.
Boyer-Araújo, V. 1993. Les traditions risquent-elles d’être contaminées? Paradigmes
scientifiques et orthodoxie religieuse dans les cultes de possession au Brésil.
Journal de la socie´te´ des ame´ricanistes, 79:67–90.
Budden, A. 2003. Pathologizing Possession: An Essay on Mind, Self, and Experience in
Dissociation. Anthropology of Consciousness, 14(2):27–59.
Campos, J., C. Frankel & L. Camras. 2004. On the Nature of Emotional Regulation. Child
Development, 75(2):377–394.

Capone, S. 1999. La quête de l’Afrique dans le candomble´. Pouvoir et tradition au Bre´sil. Paris:
Karthala.
Cavalcanti, P. 1988. As seitas aficanas do Recife. Estudos afro-brasileiros, trabalhos apresen-
tados ao 18 Congresso Afro-Brasileiro em 1934 em Recife, 243–257. Recife: Fundação
Joaquim Nabuco, Massangan.
Cohen, E. 2007. The Mind Possessed. The Cognition of Spirit Possession in an Afro-Brazilian
Religious Tradition. Oxford: Oxford University Press.
——. 2008. What is Spirit Possession? Defining, Comparing, and Explaining Two Posses-
sion Forms. Ethnos, 73(1):101–126.
Crossley, N. 2004. Ritual, Body Technique and (inter)Subjectivity. In Thinking Through
Rituals. Philosophical Perspectives, edited by K. Schilbrack. pp. 31–51. New York,
London: Routledge.
Csordas, T. 1990. Embodiment as a Paradigm for Anthropology. Ethos, 18(1):5–47.
——. 1993. Somatic Modes of Attention. Cultural Anthropology, 8(2):135–156.
Csordas, T. (ed). 1994. Embodiment and Experience. The Existential Ground of Culture and
self. Cambridge: Cambridge University Press.

Damasio, A. 1999. The Feeling of What Happens. Body and Emotion in the Making of
Consciousness. San Diego, New-York, London: A Harvest Book Harcourt Inc.
Dantas, B.G. 1982. Repensando a pureza nagô. Religião e Sociedade, 8:15–20.
——. 1987. Pureza e poder no mundo dos candomblés. DE MOURA, E. (org). Candomble´.
Desvendando identidades. pp. 121 – 128. São Paulo: EMW Editores.
D e Carvalho, J.J. 1987. A força da Nostalgia. A Concepção de Tempo Histórico dos Cultos
Afro-Brasileiros Tradicionais. Religião e Sociedade, 14(2):36–61.
de Carvalho, J.J. 1993. Aesthetics of Opacity and Transparence: Myth, Music and Ritual in
the Xangô Cult and in the Western Art Tradition. Latin American Music Review,
14(2):202–229.
de Heusch, L. 1995. Possédés somnambuliques, chamans hallucinés. In La transe et
l’hypnose, edited by D. Michaux. pp. 19–46. Paris: Auzas Imago.
Desjarlais, R. 1992. Body and Emotions. The Aesthetics of Iillness and Healing in the Nepal
Himalayas. Philadelphia: University of Pennsylvania Press.
Eisenberg, N. & T. Spirad. 2004. Emotion-Related Regulation: Sharpening the
Definition. Child Development, 75(2):334–339.
Ekman, P. & W. Friesen. 1969. The Repertoire of Nonverbal Behavior: Categories,
Origins, Usage, and Coding. Semiotica, 1:49–98.
Elbein dos Santos, J. 1975. Os Nago e a morte. Petropolis: Editora Vozes.
Fernandes, G. 1937, Xangôs do Nordeste. Investigções sobre os cultos negŕo-fetichistas do
Recife. Rio de Janeiro: Civilização Brasileira.
Frijda, N.H. & B. Mesquita. 1997. The Social Roles and Functions of Emotions. In Emotion
and Culture. Empirical Studies of Mutual Influence, edited by S. Kitamaya, H.S.
Markus. APA Science Volumes, pp. 51–88. Washington: United Book Press.
Gell, A. 1980. The Gods at Play: Vertigo and Possession in Muria Religion. Man,
15:218–247.
Gieser, T. 2008. Embodiment, Emotion and Empathy. A Phenomenological Approach
to Apprenticeship Learning. Anthropological Theory, 8(3):299–318.
Goldman, M. 1987. A construção ritual da pessoa: a possessão no candomblé. Candomble´.
Desvendando identidades. D e Moura, C. (org.), pp. 87– 120. São Paulo: EMW editores.

200 arnaud halloy
Halloy, A.. 2007. ‘Un anthropologue en transe. Du corps comme outil d’investigation
ethnographique’. In Corps, performance, religion. Etudes anthropologiques offertes à
Philippe Jespers, edited by P. Petit & J. Noret. pp. 87–115. Bruxelles: PubliBook.
——. 2009. ‘Incorporer’ les dieux. Les ressorts pragmatiques de la transe de possession
religieuse dans le culte Xangô de Recife (Premières pistes). In La conscience dans
tous ses états, Approches anthropologiques et psychiatriques, cultures et the´rapies, edited
by S. Baud & N. Midol. pp. 77–95. Elsevier Masson Editions.
—— 2010. ‘Chez nous, le sang reigne!’. Apprendre la tradition dans le culte Xangô de
Recife. Terrain, 55:10–27.
——. 2012. ‘Objects, bodies and gods. A cognitive ethnography of an ontological dynamics
in the Xangô cult (Recife – Brazil)’. in Making Spirits Materiality and Transcendence in
Contemporary Religions.. London: I.B. Tauris (Tauris Academic Studies).
Hell, B. 2008. Négocier avec les esprits tromba à Mayotte. Retour sur le ‘théâtre vécu’ de la
possession. Gradhiva, 7:7–23.
Herskovits, M.J. 1943. Pesquisas etnologicas na Bahia. Bahia: Museo d’Estado.
——. 1958. The Myth of the Negro Past. USA: Harper & Brothers.
Hochschild, A.R. 1979. Emotion Work, Feeling Rules, and Social Structure. Amercian
Journal of Sociology, 85(3):551–575.
Hollan, J., E. Hutchins & D. Kirsh. 2000. Distributed Cognition: Toward a New
Foundation for Human-Computer Interaction Research. ACM Transactions on
Computer-Human Interaction, 7(2):174–196.
Humphrey, C. & J. Laidlaw. 1994. The Archetypal Actions of Ritual. A Theory of Ritual
Illustrated by the Jain Rite of Worship. Oxford: Clarendon Press.
Hutchins. 2005. Cognition in the Wild. Cambridge: MIT Press.
James, W. 1884. What is an Emotion? Mind, 9:188–205.
Jonhson, C. 2002. Secrets, Gossip, and Gods. The Transformation of Brazilian Candomble´.
Oxford: Oxford University Press.
Klass, M. 2003. Mind over Mind. The Anthropology and Psychology of Spirit Possession.
Oxford: Rowman & Littlefield.
Lapassade, G. 1997. Les rites de possession. , Anthropos. Paris: Laureys Steven.
Lawson, E.T. & R.N. McCauley. 1990. Rethinking Religion: Connecting Cognition and
Culture. Cambridge: Cambridge University Press.
Leavitt, J. 1996. Meaning and Feeling in the Anthropology of Emotions. American
Ethnologist, 23(3):514–539.
LeDoux, J. 1996. The Emotional Brain. The Mysterious Underpinnings of Emotional Life.
New York: Simon & Schuster Paperbacks.
Leiris, M. 1958. La possession et ses aspects theá̂traux chez les Ethiopiens de Gondar. Paris:
Plon.
Lepine, C. 2000. Os estereótipos da personalidade no candomblé nagô. In Candomble´.
Religião do corpo e da alma, edited by C. De Moura. pp. 139–163. Rio de Janeiro:
Pallas.
Leventhal, H. & K. Scherer. 1987. The Relationship of Emotion to Cognition: A Func-
tional Approach to a Semantic Controversy. Cognition and Emotion, 1(1):3–28.
Levi-Strauss, C. 1971. Mythologiques IV: l’homme nu. Paris: Plon.
Levy, R.I. 1973. Tahitians. Mind and Experience in the Society Islands. Chicago: Chicago
University Press.

Liénard, P. & P. Boyer. 2006. Whence Collective Rituals? A Cultural Selection Model of
Ritualized Behavior. American Anthropologist, 108(4):814–827.
Luhrmann, T. 2004. Metakinesis: How God Becomes Intimate in Contemporary US
Christianity. American Anthropologist, 106(3):518–528.
——. 2007. How Do you Learn to Know that it is God Who Speaks? In Learning Religion.
Anthropological Approaches, edited by D. Berliner & R. Sarró. pp. 83–102. New York,
Oxford: Berghahn Books.
Luhrmann, T.M., H. Nusbaum & R. Thisted. 2010. The Absorption Hypothesis: Learning
to Hear God in Evangelical Christianity. American Anthropologist, 12(1):66–78.
Maggie, Y. 1988. O medo do feitiço: relações entre magia e poder no Brasil. Rio de Janeiro:
Thesis, National Museum.
Mauss, M. 1936/2003. Les techniques du corps. Sociologie et anthropologie. Paris:
Quadrige Presses Universitaires de France, pp. 363 – 386.
McCauley, R. & T. Lawson. 2004. Bringing Ritual to Mind. Psychological Foundations of
Cultural Forms. Cambridge: Cambridge University Press.
Motta, R. 1993. La Rationalité dans la Magie: Economie du Xangô de Recife. In Jeux

d’Identite´: Etudes Comparatives à partir de la Caraibe, edited by M-J. Jolivet & D.
Rey-Hulman. pp. 355–365. Paris: L’Harmattan.
Murphy, S. & R. Zajonc. 1993. Affect, Cognition, and Awareness: Affective Priming
with Suboptimal and Optimal Stimulus. Journal of Personality and Social Psychology,
64:723–739.
Opipari, C. 2004. Le Candomble´ : images en mouvement, São Paulo-Bre´sil. Paris:
L’Harmattan.
Ortiz, R. 1990. A morte branca do feiticeiro negro. Umbande e Sociedade brasileira. Rio: Editora
Brasilense.
Parlebas, P. 1981. Contribution à un lexique commente´ en science de l’action motrice. Paris:
INSEP.
Prinz, J. 2004. Gut Reactions. A Perceptual Theory of Emotions. Oxford: Oxford University
Press.
Querino, M. 1938. Costumes africanos no Brasil. , Bibliotheca de divulgação scientifica. Rio
de Janeiro: Civilização Brasileira.
Ramos, A. 1934/1988. O negro brasileiro. Recife: Fundação Joaquim Nabuco, Massangana.
Rappaport, R. 1999. Ritual and Religion in the Making of Humanity. Cambridge, U.K:
Cambridge University Press.
Ribeiro, R. 1950 –51. O Negro em Pernambuco – Restrospecto de suas praticas religiosas.
Revista do arquivo Publico, 5 –6(7 –8):571–587.
——. 1978. Cultos afro-brasileiros do Recife : Um estudo de ajustamento social. Recife: Instituto
Joaquim Nabuco de Pesquisas Sociais.
Robinson, J. 2005/2007. Deeper than Reason. Emotion and its Role in Litterature, Music, and
Art. Oxford: Clarendon Press.
Rodrigues, N.R. 1900. O animismo fetichista dos negros baianos. Salvador: Reis & Co.
Rouget, G. 1980/1990. La musique et la transe. Esquisse d’une theórie géne´rale des relations de
la musique et de la possession. Paris: Gallimard.
——. 2006. Transe : théâtre, musique et neurosciences. Cahiers de musique traditionnelle :
chamanisme et possession, 19:211–220.

202 arnaud halloy
Segato, R. 1995. Santos e daimones. O politeismo afro-brasileiro e a tradiçao arquetipal. Brasilia:

editora UNB.
——. 2000. Okarilé: Uma toada icônica de Iemanjá. Revista do Patrimônio Histórico e
Artı́stico Nacional, 28:237–253.
Seligman, R. & L.J. Kirmayer. 2008. Dissociative Experience and Cultural Neuroscience:
Narrative, Metaphor and Mechanism. Culture Me´dicine and Psychiatry, 32:31–64.
Severi, C. 2007. Le Principe de la Chimère. Une anthropologie de la me´moire. Paris: Presses de
l’Ecole normale supérieure.
Smith, F.M. 2001. The Current State of Possession Studies as a Cross-Disciplinary
Project. Religious Studies Review, 27(3):203–212.
Sperber, D. 1974. Le symbolisme en geńe´ral. Paris: Hermann.
——. 1996. La contagion des ideés. Paris: Odile Jacob.
Spinrad, T., N. Einsenberg, A. Cumberland, R. Fabes, C. Valiente, S. Shepard, M. Reiser,
S. Losoya & I. Guthrie. 2006. Relation of Emotion-Related Regulation to Children’s
Social Competence: A Longitudinal Study. Emotion, 6(3):498–510.
Surralès, A. 2003. Au coeur du sens. Perception, affectivite´, action chez les Candoshi. Paris:
Chemins de l’ethnologie.
Vatin, X. 2005. Rites et musiques de possession à Bahia. Paris: L’Harmattan.
Verger, P. [1957] 2000. Notas sobre o culto aos orixás e voduns. Editora da Universidade de
São Paulo: São Paulo.
Vogel, A., D a Silva Mello, M.A. & Pessoa de Barros, J.F. 1993. A Galinha d’Angola. Iniciação
e identidade na cultura afro-brasileira. Rio de Janeiro: EDUFF/FLASCO/Pallas.
Wacquant, L. 2005. Carnal Connections: On Embodiment, Apprenticeship, and
Membership. Qualitative Sociology, 28(4):445–474.
Wafer, J. 1991. The Taste of Blood. Spirit Possession in Brazilian Candomble´. Contemporary
Ethnographic Series. Philadelphia: University of Pennsylvania Press.
Walker-Andrews, A.S. 1997. Infants’ Perception of Expressive Behaviors: Differentiation
of Multimodal Information. Psychological Bulletin, 121(3):437–456.
Warnier, J-P. 1999. Construire la culture mate´rielle – l’homme qui pensait avec ses doigts.
Paris: PUF.
Whitehouse, H. 2000. Arguments and Icons. Divergent Modes of Religiosity. Oxford: Oxford
University Press.
——. 2004. Modes of Religiosity. A Cognitive Theory of Religious Transmission. Cambridge:
Altamira Press.
Wikan, U. 1990. Managing Turbulent Hearts. A Balinese Formula for Living. Chicago,
London: University of Chicago Press.
Zajonc, R. 1980. Feeling and Thinking: Preferences Need no Inferences. American
Psychologist, 35:151–175.

GODS IN THE FLESH - Arnaud Halloy-1

Uploaded by

Copyright:

Available Formats

GODS IN THE FLESH - Arnaud Halloy-1

Uploaded by

Document Information

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

GODS IN THE FLESH - Arnaud Halloy-1

Uploaded by

Copyright:

Available Formats

This article was downloaded by: [Arnaud Halloy]

On: 18 July 2012, At: 02:04

Gods in the Flesh: Learning

Version of record first published: 17 Jul 2012

To link to this article: http://dx.doi.org/10.1080/00141844.2011.586465

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-

abstract Drawing on first-hand ethnographical data from the Xangô, an Afro-

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

mental structures underlying the conceptualisation of spirits and mind/body

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

late the contextual and diachronic dimensions of learning. Cognitive

The present study starts from a conclusion formulated by Rouget (1990) at

Spirit Possession in the Xangô Cult6

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

1999). Xangô members worship two categories of ‘spiritual entities’:7 the

Possession and ritual practice

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

Possession mostly happens during ritual events. It is very much to be

‘daughter’, for example by possessing him/her at work, in a public place, etc.

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

Possession and Emotions

approximation’ (irradiação/aproximação) stage that precedes the ‘full’ possession

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

One self-report is by Maria-Helena, a young saint-mother, ‘daughter’ of

I became sad . . . My heart accelerated, a heat wave invaded my torso . . . I really

In this short account, another constitutive element of many irradiation experi-

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

In the last testimony presented here, Yguaracy, an experienced priest, 40 years

I felt it was not an ‘approximation’ of my mother Oxum [Yguaracy’s main orixà]. It

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

Possession as ‘uncanny’ gut feelings

James’s argument about emotions is, in my view, directly relevant to posses-

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

Obrigação for the Ogun of Taı́sa (May 2003)

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

The ritual unfolding of possession as described in this obrigação episode brings

Ritual features and ‘perceptual attractors’

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

reactions associated with the initial stage of possession. The invisibility of

(Sperber 1974) – the ‘imagination capture’ – by coupling the sensations and

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

‘Social referencing’ and emotional regulation

sensory capture, may be defined as a complex skill in which another person’s

Iyanle´ for Xangô 23 (July 2003)

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

room a few minutes later, as if nothing had happened.

How did Xangô members interpret Rafael’s possession?

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

The dynamics of perceptual attractors

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

Paulo, the patriarch of the saint-family I studied, emphasised the affective

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

evolve from largely undifferentiated arousal states to ‘orixá-specific’ somatic

ethnos, vol. 77:2, june 2012 (pp. 177 –202)

studies explore dimensions of possession from diverse perspectives: post-structural

ethnos, vol. 77:2, june 2012 (pp. 177 –202)