Original Paper

Psychopathology 2008;41:245253 DOI: 10.1159/000125558

Received: October 26, 2006 Accepted after revision: July 3, 2007 Published online: April 11, 2008

Possession States in Northern Sri Lanka

Daya Somasundaram T. Thivakaran Dinesh Bhugra
Department of Psychiatry, University of Jaffna, Jaffna, Sri Lanka

Key Words Possession states Cross-cultural psychiatry Phenomenology South Asia

Abstract Background: Possession states are still commonly seen in developing societies as acceptable cultural phenomena in normal persons as well as in those with psychiatric illness. Possession is defined here as the experience of being taken over, controlled or occupied by another spirit or force. Sampling and Methods: This is a descriptive cross-sectional study of possession states among psychiatric patients, general population and popular adepts in Northern Sri Lanka, using semi-structured questionnaires and clinical observations. Results: Thirty psychiatric patients were identified as having possession states. They were compared with 30 controls each from the general population admitted to a general hospital outpatient department and selected popular adepts in the community. The latter are individuals who are well known as having possession states. An analysis of social factors and other variables showed that education, marital status, age, employment, strength of belief, alterations in personality, past or family psychiatric history, previous exposure to similar phenomena, help-seeking behavior and treatment outcome differed between the three groups. Religion of the subjects or recent changes in values showed no correlation with possession while monetary gain from possession states showed only a partial correlation. Western medical treatment was of value only when possession states were seen as part of psychotic illness. Discussion: Possession is a spectrum of experiential and behavioral phenomena seen in culturally acceptable form in normal people, popular ad-

epts, as well as manifestations of psychotic illness. Possession states which fit normal cultural stereotypes could, if necessary, be better managed by traditional methods. However, clinicians need to be familiar with culturally abnormal forms of possession which are manifestations of psychotic illness that benefit from western psychiatric treatment.
Copyright 2008 S. Karger AG, Basel

The [possession] merely opens the door to the secret recess of the heart where man hides those thoughts and emotions, those unfulfilled wishes and dreams or crude desires that are not permitted gratification in normal everyday life It is from this untamed area of the unconscious, semiconscious or subconscious that the storms originate, and under certain circumstances whirl him into insanity. August Wimmer (1924)

Introduction

Among the varied phenomena found in normal life and psychiatric practice, possession states are perhaps the most colorful, dramatic and exotic but least understood. It has been at once described as a subcultural variation of normal life, a form of hysterical dissociation or actingout behavior with a multitude of neurotic causes or manifestation of psychosis [15]. Indeed, Wimmer [6] equated possession states with psychogenic psychoses reactive psychoses to psychological trauma with a good prognosis. Even in orthodox western psychiatry, possession was first considered a form of passivity or delusion of control, a first-rank symptom of Schneider, or diagnostic of schizophrenia [1, 5]. However, it was soon recognized as
Prof. Daya Somasundaram Department of Psychiatry University of Jaffna Jaffna (Sri Lanka) Tel. +94 212 222 073, E-Mail manathu@gmail.com

2008 S. Karger AG, Basel 02544962/08/04140245$24.50/0 Fax +41 61 306 12 34 E-Mail karger@karger.ch www.karger.com Accessible online at: www.karger.com/psp

a subcultural phenomenon fitting a culturally sanctioned stereotyped behavioral pattern in non-western countries [1, 3]. Thus the pendulum swung to the other extreme where the modern tendency now is to discount all forms of possession as merely subcultural and benign and thus of no serious psychiatric significance. Looking at possession states from an anthropological view, Boddy [7] notes that the otherness of the phenomenon demands explanation. Lewis [8] emphasizes that it is crucial to note that scholars are embedded in cultural frameworks and hence it becomes important for research within the same culture to be looked at by people who are embedded in these frameworks. Lewis [8] talks of possession within the rubric of ecstatic religion, and as Boddy [7] emphasizes, the typologies are achieved at the expense of context and inevitably reflect the interests, values and fascination of the society which is being analyzed and that is being presented here. We also agree with Crapanzanos [9] suggestion that an adequate account of possession must grant the spirits existence in the lives of their hosts which is the key point of the study. Looking at Sri Lankan ascetics possessed by ghosts and deities, Obeyesekere [10] noted that cultural symbols inform and are informed by individual experience. He observed that those possessed are unwell and subordinate women, but the flexibility of the spirit idiom is evident. This is almost exactly what we see in the community sample where personal symbols become socially and emotionally meaningful at once. This needs to be and is clearly differentiated from those whose possession state is related to their psychiatric state. Kapferer [11] argues that in Buddhist Sri Lanka demonic exorcism is rooted firmly in the context of colonial history and social class; however, in our sample which is largely Tamil, social class remains important but the impact of colonialism would be of a different kind as this group is already a minority and their life experiences are likely to be colored by their relationship with the majority community. Possession intersects with numerous cultural domains including medicine and religion but is itself reducible to none [7]. This seems to be true of the general population [12]. However, in psychiatric settings those manifesting possession states are often found to be suffering from psychoses. Seligman [13] describes the dynamic interactions between individual and social characteristics and cultural environment using data from a study set in Brazil. The individuals expectations based on culturally held etiology of mediumship become embodied in mediums. The mediums appear to have little access to power and resources and thus these marginal positives might contrib246
Psychopathology 2008;41:245253

ute to the possession. Using the practice of Candombl in her study, Seligman [13] demonstrates that the subjects are victims of both racial and gender discrimination. Seventy-one individuals were interviewed with the aim to investigate why certain people become Candombl and what distinguishes them from those who do not become mediums. Of the 71 individuals aged 1865, some were mediums and others were seen as controls. She found that the social and economic marginality of Candombl participants was confirmed. We set out to study the clinical and social aspects of possession states. The three types of possession states seen are those in the psychiatric patients, in medical patients, and in people who are revered and recognized in the community. For the purposes of the study we defined possession as the episodic experience of being controlled by another spirit or force with the loss of ones personal identity often manifested in culture-specific, stereotyped movements and behavior with or without an associated trance state. The local, Tamil, words for these phenomena include: kalai (artistic) and uru, or more descriptively, uru edethu adethal, literally meaning form or conforming to form and to dance taking a form. Other words which are used in this context are adkollal (being taken over), sami vanthidu (the spirit has come), and pey addichitu (being hit by a demon or ghost). Most people would simply indicate that a spirit or force has come in, is staying within or has a presence.

Method
This is a descriptive cross-sectional study. The main population under study are psychiatric patients receiving inpatient treatment at the Psychiatric Unit at the District Hospital Tellippalai, in whom possession experiences have occurred during the period from August 1, 2003 to January 31, 2004. The Psychiatric Unit at the District Hospital Tellippalai functions as the main inpatient facility for all patients from Jaffna and the entire Northern Province, having four other districts namely Mullaitivu, Kilinochi, Mannar and Vavuniya. Some patients also come from other provinces. The northern region is mainly of a rural nature with small towns, recently caught up in a civil war situation. The predominantly Tamil society is very conservative and caste based from the middle and lower socioeconomic classes, belonging during this period mainly to Hindu and some Christian religious faiths. The description was both quantitative and qualitative, based on a semi-structured questionnaire with the data collected from the available records and interviews of the patient and relatives. The interviewer-administered questionnaire elicited basic demographic information: previous, present, family and socioeconomic history, phenomenological data and details of the possession and a mental state examination.

Somasundaram/Thivakaran/Bhugra

30 Community Medical Psychiatric patients Numbers Numbers 20

20

15

Community Medical Psychiatric patients

10

10

5 0 Male Sex Female

0 <10 1019 2034 3554 Age (years) 5574 >74

Fig. 1. Gender distribution.

Fig. 2. Age distribution.

The psychiatric patient group was compared to two other population groups that had experienced possession states. The psychiatric patient group included all those who had experienced a possession state which was not necessarily the presenting complaint nor the main manifestation, but part of the clinical picture. Another sample was selected from all those who visited the general medical outpatient clinic (not psychiatric clinic) at the General Hospital, Jaffna, on 3 days. After a preliminary screening of all attendees, those who acknowledged having had recent possession experiences were invited to participate in the study and the first consecutive 30 patients were included. The third population studied included people who were well known in the Jaffna community for their possession states which were claimed as supernatural within the cultural context. They often provide a social function or service through their possession such as divination, oracle, or mediumship. The sample size aimed at was 30 in each group. Thus the three groups included in the study consisted of subjects with either psychiatric possession, medical possession or community possession. Ethics approval for the study was obtained from the Ethics Review Committee, Faculty of Medicine, University of Jaffna, Sri Lanka. Informed consent was obtained from the participants after a complete description of the study.

The frequency of possession states in the psychiatric patient population (30 out of 426, or 7%) and medical patients (30 out of 1,492, or 2%) was noted. A comparative analysis of some significant variables is presented in what follows. Gender In all three samples, possession states were twice as likely in females compared to males in each of the three samples (fig. 1). Age Possession states were either not seen or seen rarely at the extremes of age none under 10 years and in no patients over 74 years of age (fig. 2). All subjects in the community possession group were over 35 years old. Nearly 50% of the possession states occurred in the age group 3040 where all three populations share roughly the same proportion. Educational Level The majority of persons having possession experiences across all three groups had had an education up to the 10th year (fig. 3). Occupation Not surprisingly, there were a large number of housewives (fig. 4). Although no statistically significant differences emerged, the highest number of housewives were found in the community group (over 50%). There were fewer possessions among white-collar workers, while unsurprisingly more of the psychiatric patient group tended to be unemployed.
Psychopathology 2008;41:245253

Results

During the period of the study, a total of 426 patients (196 males; 230 females) were admitted to the Psychiatric Unit at the District Hospital Tellippalai. A total of 1,492 persons visited the general outpatient department at the General Hospital, Jaffna, and were screened for possession in the mornings of 3 days (November 21 and 28, 2003, and January 2, 2004). However, only the first 30 acknowledging a recent possession experience were studied.
Possession States

247

20 15 Numbers 10 5 Community Medical Psychiatric patients

30 25 Numbers 20 15 10 5

Community Medical Psychiatric patients

0 Over ordinary level Under 5th grade Formal education No data

0 Divine Ghostlike Humane Category Others More than (ill defined) one type

Fig. 3. Educational level.

Fig. 6. Possessing spirit or force.

20 Community Medical Psychiatric patients

15 Numbers

Marital Status Over 81% of possessions were reported in married persons (fig. 5). Significantly, all in the community group and 90% of the medical group were married, while in the patient group only 53% were married. The Possessing Spirit or Force The great majority (66 or 73%) of those experiencing possession states across all the groups attributed it to divine forces (fig. 6). In the medical group, the possessions were by subcultural divine and ghostlike forces. Even in the community group, a person with multiple possessions had been possessed by both divine and ghostly forces. Three persons (in the category others in fig. 6) also described their possessing force as good, something akin to divine, which was perceived to have a positive effect on them and their lives. In the psychiatric patient group, over 20% claimed they were being possessed by a bad spirit or force, which was perceived to have a negative effect on them. In the psychiatric patient group, 3 persons complained of multiple possessions compared to 1 in the community group. In addition, 4 individuals in the psychiatric patient group had been possessed by humans compared to 1 in the community group. In one of these cases, the patient said she was being possessed by a scientist working for NASA indicating an impact of globalization. Details of this case are given in Appendix 1. Religion The religion of the possessed was representative of the local population across all three groups. Twenty-six (87%) of the community group, 20 (67%) of the medical and 27
Somasundaram/Thivakaran/Bhugra

10

0
-c o w llar or k Sk i la lled bo r Un sk i la lled bo r Un em pl oy ed ife Ho us ew to th sp er w ec is ifi e ed

hi

te

Occupational status

Fig. 4. Occupational status.

35 30 25 20 15 10 5 0 Married Not married Status

Numbers

Community Medical Psychiatric patients

Widow/er

Fig. 5. Marital status.

248

Psychopathology 2008;41:245253

No

(90%) of the psychiatric patient group were Hindus. Four (13%) of the community, 5 (17%) of the medical and 2 (7%) of the psychiatric patient group belonged to orthodox, conventional Christian churches, while 5 (17%) of the community and 1 of the psychiatric patient group were from the charismatic, non-orthodox Christian faiths. The religion of the possessing spirit did not necessarily correspond to the religion of the person being possessed. Hindu spirits possessed Christians and vice versa. There were a few cases where persons recently converted to a new religion (Pentecostal) seemed to have been pursued by the spirits from the old faith and as a result some had converted back to their original religion. Timing of the Possession Across all groups, possession attacks lasted less than an hour in 48 (53%) subjects. Eleven (37%) of the community, 10 (33%) of the medical and 21 (70%) of the psychiatric patient groups had long episodes continuing for several hours or even days. Those lasting for more than a day were only seen in the psychiatric patient group [14], but the phenomena had continued for less than a month. None had only a single attack. In 71 out of all the subjects (79%), the episodes had lasted more than 6 months. Diagnosis among the Patients Using ICD-10 criteria, schizophrenia was seen as the commonest diagnosis among the psychiatric patient group (n = 16, or 53%), while acute psychotic disorder accounted for 6 cases (20%). Thus psychotic illnesses were found in 73%. Other diagnoses were affective disorder (n = 1, or 3%), dissociative disorder (n = 3, or 10%) and somatoform disorders (n = 4, or 13%). Problems identified in the medical and community groups included family and social problems (n = 14 or 47% in the community group and n = 7 or 23% in the medical group), common mental disorder (n = 6 or 20% in the community group and n = 12 or 40% in the medical group), economic difficulties (n = 7 or 23% in the community group and n = 12 or 40% in the medical group) and alcohol abuse (n = 6 or 20% in the community group and n = 3 or 10% in the medical group). Five from the community and medical groups (16%) had features of psychotic illness while 1 adept from the community group had chronic psychotic illness. Change in Economic Status Economic gain through the possession state was reported in the medical (27%) and community groups (43%). This often meant some payment for services renPossession States

dered while in the possessed state such as oracle function (fortune-telling, explaining the past, providing solutions for problems) as well as healing and advising on various problems such as relational, family or social. Some charged a fee; others accepted donations whether in cash or kind. Of the psychiatric patients, nearly half (n = 14 or 47%) had lost money due to their behavior, mainly due to loss of work. Change in Personal/Social Status Twenty-four (80%) of the community group and 25 (83%) of the medical group claimed that they gained in personality or social status by being possessed, while only 6 (20%) psychiatric patients reported a positive gain. This included more respect, dignity, increased self-esteem, better relationships and understanding by others. On the other hand, 21 (70%) patients reported a decrease in their personal or social status following possession especially if possessed by negative or evil forces where fear, anger, lowering of status, feeling bad, social ostracism, and isolation were predominant. Family History of Possession States or Psychiatric Illness There was a family history of possession states in 13 (43%) of the community group, 12 (40%) of the medical group and 9 (30%) of the psychiatric patient group. A family history of psychiatric illness was obtained in only 1 person from the community group, 5 (17%) of the medical group and 13 (43%) of the psychiatric patient group. Witnessing Possession States Elsewhere Among psychiatric patients, 14 (47%) had seen a possession state manifestation, and of the community group, 24 (80%) had witnessed a possession state earlier. Treatment Sought Overall 41 (46%) people with possession states had sought any form of treatment, with 40 (44%) denying any need for it. In the medical group, 26 (86%) had not sought treatment. Around 13 (43%) of the community group felt they could be without treatment. Of patients who were seen in the psychiatric services, only 1 went away after assessment denying the need for treatment. Initially, many did not have a choice as they were usually brought by relatives who insisted on treatment. Half of them had also sought traditional treatments such as religious rituals, exorcism, trying of charmed thread, prayer, blessing, holy gaze, sprinkling or application of various holy materials, elimination of
Psychopathology 2008;41:245253

249

evil objects, among others, prior to psychiatric intervention. Treatment Outcome Western medical treatment showed a much better outcome (n = 20 or 72%) than traditional methods (n = 5 or 33%) in the psychiatric patient group. However, traditional methods showed a better outcome (n = 3 or 75%) than western medical treatment (n = 2 or 33%) in the community group. In the medical group, western methods were helpful in none, while 1 person out of 2 benefited from the traditional methods. See Appendix 2 for a case that illustrates social, economic and psychological factors.

Discussion

The prevalence rate of possession states (7%) in our study is much higher than that from the community survey done three decades earlier in the South of Sri Lanka which found 37 cases out of 7,653 or 0.5% [12]. Indeed, a study at the outpatient department of the General Hospital, Jaffna, found a high number (over half) of people with psychosocial problems [14]. The possession behavior could be a manifestation of the psychosocial distress as well as a way of coping. The twofold female preponderance in all three samples could be explained by the perceived and real inferior social status of women in this culture; thus the possession states may be a way of gaining attention and expressing suppressed feelings or somatization. The proportions are equal in the three groups irrespective of the outcome of the behavior, i.e. accepted or disapproved (in case of psychiatric illness) by the community. The Southern study had also found a threefold female preponderance [12]. The occurrence of possessions in 35-year-olds in the community group may indicate that they are well adapted to this state or have good control over themselves. In other words, their possession states are accepted by the community, perceived as supernormal by the society and hence themselves. When looking at the educational level of those with possessions, the majority had had an education up to grade 10, while those with possession experiences and more than a grade 10 education were more likely to belong to the psychiatric patient group. Thus it would appear that psychiatric patient status is more likely to be associated with possession than simply educational status. The cultural expectations of marriage and social status accorded to wives would indicate that a housewife has
250
Psychopathology 2008;41:245253

a greater need for expression through this particular mode or that the community perceives a housewifes possession as more culturally acceptable. Gender is an important factor especially as women can be marginalized in patriarchal societies and they may have been taking this control back through possession. They may also be taking this control back on behalf of their families, children, friends and colleagues. Kapferer [15] points out that because Sinhalese women are considered more firmly attached than men to relationships of the human world, their femininity culturally prefigures them as prone to demonic attack. There are differences between Sinhalese and Tamil women, but there is no reason to believe that similarities do not exist, particularly in the subcultural family dynamics. Boddy [7] highlights urban rural differences in her review and suggests that ancestry and family play a key role. An issue threading through the literature is that of selfhood or identity. Boddy [7] and Crapanzano [9] indicate that possession must be viewed as an idiom for articulating a certain range of experience. Thus, as Boddy points out, one should explore how possession as an idiom of communication is constructed and used in specific societies, which is what we set out to do. The fewer white-collar workers among those with possession states could be due to possession being a hindrance or stigma for white-collar workers. The higher number of unemployed subjects among the psychiatric patients would obviously be due to the disability caused by psychiatric illness. The surprisingly large number of married persons exhibiting possession, particularly in the medical and community groups, confirms that marriage confers social acceptance while being single could increase the tendency towards social pathology. Those with psychiatric illness would also tend to remain single. In the psychiatric patient group, the number with malignant or evil possessions increases (over 20%), which may explain the need for help, change, cure, treatment or some intervention. The increase in culturally abnormal forms of possession in the psychiatric patient group such as multiple or human possessions indicates the tendency in this group towards a form of possession deviating from the normally accepted pattern. The example of possession by an NASA scientist illustrates the pathoplastic effect of modern technology and knowledge on the substratum of morbid passivity experience which was being experienced or explained by the patient through cultural idioms or forms (possession) albeit in a deviant manner. Goff et al. [16] studied 61 chronically psychotic outpatients (25 had a history of delusional possession compared with 36 who did not). Of those who reported deluSomasundaram/Thivakaran/Bhugra

sional possession, higher levels of childhood sexual abuse, higher dissociation scores, more cannabis use and more experiences of thought control were noted. The authors suggested that possession beliefs in some psychotic patients may reflect childhood trauma and dissociation. Thus there may be cultural differences in our sample and the American sample of Goff et al. [16]. Those belonging to the nonorthodox, charismatic religious belief system were overrepresented in the medical group. Interestingly, the religious affiliation of the possessing spirit cut across the religion of the person (a Hindu could be possessed by a Christian spirit and vice versa), showing the overriding influence of the underlying common culture. Past experience or exposure can play an important role in causing, shaping and maintaining possession states. Recurrence or chronicity may correlate with social integration when community tolerates or accepts the behavior; with control and mastery of the timing and pattern even veneration is achievable. Sims [2] in considering the view of Jaspers on possession emphasizes the importance of psychiatric diagnosis in assessing psychopathological form. It could also work the other way, psychopathological form determining the psychiatric diagnosis. Culturally abnormal forms of possession should suggest a psychotic illness, while the more benign, cultural stereotypes found in the general population would suggest, if a psychiatric diagnosis is to be made, a neurosis, particularly hysterical dissociation. Here the second personality that episodically takes over the self is the possessing spirit or force. The difference may be that in possession the second personality invades and takes over, possesses the original self. Although possession states per se are said to become rare in western countries [5], except when looked for in immigrant populations from non-western cultures [4, 17], modern-day variants in western culture could include multiple personality disorder [3], and in the general population states during religious charismatic movements [12]. In addition to a reduction in stress, these mediums interviewed by Seligman [13] also experienced an increase in secondary gain as seen by more social support and general emotional benefits. The individuals reflected on their own self-narrative in spiritually significant terms. In our sample, we did not explore this, but the financial gain and positive changes in personal and social status through possession in the community and medical groups could explain the occurrence and persistence of possession in this culture.
Possession States

The finding that 24 (80%) of the community group had witnessed a possession state (and many of the general group had a family history) supports the notion that possession could be seen as a learned behavior, something akin to modeling. The high exposure could have contributed to the mastery and popularity of the members of this group. It is probable that many in the medical group and even in the psychiatric patient group would have seen or heard about possession states as they are so pervasive in this society. They may have forgotten about this past encounter or there could be an element of denial. It could also exist as a cultural archetype in the collective unconscious. Another possibility is psychological stress including war trauma as the primary pathology and the possession then occurring as a cultural coping strategy. Wijesinghe et al. [12] described a high number of comorbid disorders (46%) in their cohort with possession serving the psychological function of projection of aggression, manipulation of the environment, catharsis and, perhaps, ventilation of emotions. Thus, possession states could be considered a variation of behavioral patterns beginning with the culturally ascribed stereotypes found in the general population; then the more venerated states seen in well-known community members who would be using possession for wellknown social ends like mediumship, divination or oracle function, to the supernormal, higher trances of saints, and finally the pathological types found in psychiatric illness, particularly in psychosis. A clear religious sanction for divine possession is found in Hindu holy texts and philosophy where the highest achievement is said to be the passive experience of the divine taking full control of the self and carrying out all actions. Interestingly, the study did not detect a single case of consciously stimulated possession states, which could have been for financial gain or personal status. Overall only 41 (46%) from all three groups have sought help for their condition, showing that possession is a well-accepted cultural phenomenon. Surprisingly, only 13 (34%) of the community group felt they could be without treatment. We would have expected a much higher number to feel satisfied with the enhanced social acceptance they enjoy. Their behavior should be causing psychological distress to themselves if not to the society. This could be due to an underlying disease or disturbance of the normal psychological processes along with the behavior change. Treatment outcome shows some remarkable findings. It could be that there is only an objective measurement of
Psychopathology 2008;41:245253

251

treatment outcome in the psychiatric patient group. In the other groups it is subjective, relying on the perception of the subjects and society. To correlate it with objective measurements one would need long-term case-control studies or the interviewers should observe the subjects before and after the traditional treatment. Kirmayer and Santhanam [18] emphasize that the higher prevalence of conversion symptoms in some societies may be attributable to the fit of symptoms with local ethnophysiological notions and cultural idioms of distress as well as the stigma attached to frankly psychological or psychiatric symptoms. They also suggest that dissociative phenomena can be seen in socially sanctioned rituals under the guidance of a healer or religious leader. Thus bodily control may be seen as a form of social control or vice versa. Western medical methods are more effective in the psychiatric patient population, in those with a clear-cut psychiatric illness. The improvement would be in the illness and not in the possession phenomenon itself which may cease as a manifestation of the illness. As pointed out, those who had benefited from traditional methods would not seek western medical treatment. Significantly, none of the medical group was helped by western medical methods. This may indicate that their state fitted the normal cultural stereotypes for possession or that there is a need to change the type of treatment to more traditionbased methods. Another more progressive way forward would be to develop a better relationship and understanding between both systems. The western medical system could refer those with possession states from the general population to the traditional system and the traditional sector could recommend those with psychotic illness for western medical treatment. While being familiar with the culturally appropriate possession behavior patterns, clinicians from both sectors could be more aware of some of the abnormal manifestations and patterns of behavior of pathological possession states found in psychotic illness.

Appendix 1
A 24-year-old student sat for the ordinary level examination last year with average results. She had no family history of possession or psychiatric illness. Six years ago, she had shown transient retarded behavior for 2 weeks following a febrile illness. For the whole previous month now, she had been restless, and behaved aggressively towards her mother. She also showed irritable outbursts, fear and perplexity. Although she claimed that a scientist from NASA who lived in London had taken over her and controlled all her activities, at times she was convinced that this force was not the scientist but the Amman (female goddess); other beliefs included cheivinai (a subcultural belief of witchcraft). Her belief in the NASA scientist as a possessing force was firm, continuously present and was claimed to give a good feeling, not distress. The family members brought her for treatment as they did not believe her. They said that Amman could do such things but not a scientist. After receiving a provisional diagnosis of acute psychosis, she responded well to high-dose antipsychotics and her belief in the scientist as a possessing force disappeared. Her mood and speech returned to normal.

Appendix 2
Mrs. J., a 50-year-old assistant manager of a well-known bank, was an orthodox Hindu, described as hard working, god-fearing and responsible. Being the eldest she had been dominant and obsessive to some degree though helpful and sociable. She had frequently seen possession states (Amman Uru and soothsaying) in her village during her childhood. One of her uncles had had a possession state called Annamar. There was no family or past psychiatric history of note. She was married to an executive bank officer and had had an unremarkable happy family life from which her children had now left for their higher education. All family members were ardent devotees of the Goddess of Nainatheevu Island (a female, cobra goddess called Nagapoosany Amman) for whom they had built a shrine in their premises a few years back. On an average day, she had to get up early in the morning at 4 oclock and do the household chores, clean the shrine and go to work before 7 a.m. to reach the bank on time. She returned home at 7 p.m. along with her husband and stayed awake till 11 p.m. to midnight to complete the work at home. She felt that the maintenance of the shrine was a problem and she believed it would be better for her to stay at home to look after the deity. In fact, 2 years ago, she had dreamt that the goddess asked her to do so. But premature retirement would make her lose financially on the pension benefits and the only way to preserve it and to retire prematurely was to be discharged medically for illness. In the last 2 years, she had consulted a few doctors for nonspecific symptoms of tiredness. Six months ago at the office while on duty with several other senior staff officers, she suddenly started to dance like a snake, and speak in a strange voice with unusual impoliteness (eha vasanam). The facial expression was said to be strange with fixed eyes and a challenging or angry look (similar to a cobra). The dance was typically like a snake with movements of the upper

Acknowledgements
This study of possession states is a fulfillment of a suggestion Prof. Andrew Sims made in 1988. His continuing guidance in designing the study and preparing the questionnaire ensured its completion. Prof. Channa Wijesinghe made available to us the results of his thorough and detailed study of possession states in Southern Sri Lanka and made constructive contributions to the analysis of the data and writing of the report.

252

Psychopathology 2008;41:245253

Somasundaram/Thivakaran/Bhugra

limbs, neck and trunk. Her voice claimed to be the Goddess who instructed the people around not to torment the poor woman and to help her. Almost all the staff fled from the scene. The behavior subsided within a few minutes. Since then this behavior had reappeared now and then at work and home, especially in front of the shrine, and at temples always disappearing again after 515 min. She noted that she was not in control of the behavior but that the deity made her do all the things. She acknowledged that this behavior really made her feel better and that she was able to help others by uttering prophecies, which was confirmed by some villagers and her husband. However, she complained of fatigue,

tiredness, aches and pains all of which she attributed to the excessive office work, because of which she had neglected the house and shrine. Psychiatric assessment revealed mild depression and sick role behavior but no psychotic features. Then the frequency of the possession increased from once for many days to more than twice per day and the utterances became more authoritarian insisting on quick retirement with full pension benefits. After admission to the psychiatric unit for a few days she improved with counseling, traditional relaxation exercises and a short course of anxiolytics. Though the behavior could be precipitated by interviews, both the frequency and duration decreased.

References
1 Leff L: Psychiatry around the Globe A Transcultural View. London, Gaskell, 1988. 2 Sims ACP: Symptoms in the Mind An Introduction to Descriptive Psychopathology, ed 3. Edinburgh, Saunders, 2003. 3 Castillo RJ: Spirit possession in South Asia, dissociation or hysteria? 1. Theoretical background. Cult Med Psychiatry 1994;18:121. 4 Pereira S, Bhui R, Dein S: Making sense of possession states: psychopathology and differential diagnosis. Br J Hosp Med 1994; 53: 582586. 5 Trethowan WH: Exorcism: a psychiatric viewpoint. J Med Ethics 1976;2:127137. 6 Wimmer A: On possession states (original 1924; translated by Schioldann-Nielson J). Hist Psychiatry 1993;4:413419. 7 Boddy J: Spirit possession revisited: beyond instrumentability. Annu Rev Anthropol 1994;23:407434. 8 Lewis IM: Ecstatic Religions: A Study of Spirit Possession and Shamanism, ed 2. London, Routledge, 1989. 9 Crapanzano V: Introduction; in Crapanzano V, Garrison V (eds): Case Studies in Spirit Possession. New York, Wiley, 1977, pp 139. 10 Obeyesekere G: Medusas Hair: An Essay on Personal Symbols and Religious Experience. Chicago, University of Chicago Press, 1981. 11 Kapferer B: Performance and the structure of meaning and experience; in Turner VW, Bruner EM (eds): The Anthropology of Experience. Chicago, University of Illinois Press, 1986, pp 188206. 12 Wijesinghe CP, Dissanayake SAW, Mendis N: Possession trance in a semi-urban community in Sri Lanka. Aust NZ J Psychiatry 1976;10:135139. 13 Seligman R: Distress dissociation and emotional experience: reconsidering the pathways to mediumship and mental health. Ethos 2005;39:7199. 14 Somasundaram DJ: War trauma and psychosocial problems: patient attendees in Jaffna. Int Med J 2001;8:193197. 15 Kapferer B: A Celebration of Demons: Exorcism and Healing in Sri Lanka, ed 2. Washington, Smithsonian Institute Press, 1991. 16 Goff DC, Brotman AW Kindlon D, Waiters M, Amico E: The delusion of possession in chronically psychotic patients. J Nerv Ment Dis 1991;179:567571. 17 Hale AS, Pininnti NR: Exorcism-resistant ghost possession treated with clopenthixol. Br J Psychiatry 1994;165:386388. 18 Kirmayer LJ, Santhanam R: The anthropology of hysteria; in Halligan PW, Bass C, Marshall JC (eds): Contemporary Approaches to the Study of Hysteria: Clinical and Theoretical Perspectives. Oxford, Oxford University Press, pp 251270.

Possession States

Psychopathology 2008;41:245253

253