Western Trauma Discourse
1
Running head: EXPOSURE TO WESTERN TRAUMA DISCOURSE
Western Trauma Discourse Exposure and Posttraumatic Symptoms among Burundians with
Traumatic Event Histories
Peter D. Yeomans, James D. Herbert, and Evan M. Forman
Drexel University
Author’s Note
Peter D.Yeomans, Department of Psychology, Drexel University; James D. Herbert,
Department of Psychology, Drexel University; Evan M. Forman, Department of Psychology,
Drexel University.
The authors wish to acknowledge Adrien Niyongabo, Ernest Ndayishimiye, and Jean-Marie
Nibizi, for their assistance in conducting this study. This paper is based on a Masters thesis.
Correspondence concerning this article should be addressed to James Herbert, Department
of Psychology, Drexel University, MS 988, 245 N. 15th Street, Philadelphia, PA 19102-1192.
(ph: 215-762-1692; fax: 215-762-8706). Email: jh49@drexel.edu
Word count: 5876
Western Trauma Discourse
2
Abstract
Posttraumatic Stress Disorder (PTSD) has been increasingly applied in diverse cultural settings,
despite controversy over the degree to which the symptoms of PTSD are biologically based and
therefore relatively universal or are culturally constructed. We hypothesized that prior exposure to
Western trauma discourse would be associated with PTSD symptoms among indigent Burundian
trauma victims. Analyses indicated that exposure to Western ideas about trauma was predictive of
more severe PTSD symptoms, and yielded a predictive trend when controlling for quantity of event
types experienced. Despite severe trauma histories, the sample reported relatively low levels
ofPTSD symptoms. The implications of the findings in relation to the validity of the PTSD
construct in non-Western settings are discussed.
Western Trauma Discourse
3
Western Trauma Discourse Exposure and Posttraumatic Symptoms among Burundians with
Traumatic Event Histories
Western aid to impoverished countries has increasingly included mental health services
(Summerfield, 1999). Such services have also included intervention programs for the psychological
sequelae of trauma. Some scholars, however, have raised questions about the applicability of
Western models of traumatic stress response to the non-Western world (Kagee & Del Soto, 2003;
Summerfield, 2004). It is critical to examine the relevance and the effect of these models when
exported to cultures that may hold a different understanding of terror and loss.
Posttraumatic Stress Disorder
Posttraumatic Stress Disorder (PTSD) as a diagnosis was first recognized in the third edition
of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III) (APA, 1980). Since its
inception PTSD has been embroiled in debate over multiple issues. Controversies include the
influence of the political climate in which it was conceived (Herbert & Forman, in press; McNally,
2004), the broadening definition of the range of events that qualify as traumatic stressors (McNally,
2004; Mol et al., 2005; Rosen, 2004), and the notion that traumatic memories can be actively
repressed such that they are inaccessible (Lynn, Knox, Fassler, Lillienfeld, & Loftus, 2004;
McNally, 2003). Researchers and clinicians also debate the degree to which the traumatic stress
response as identified by PTSD is largely biologically determined or culturally constructed.
Is PTSD a universal disorder?
Historical and cross-cultural research represent two methods for investigating the
universality of PTSD. Evidence in British military history suggests that the severity and nature of
posttraumatic reactions are largely the result of cultural forces (Shephard, 1999). Herbert &
Sageman (2004) point to the cultural evolution of traumatic stress symptoms from paralysis to
Western Trauma Discourse
4
mutism to trembling over the last one hundred years. Summerfield (2004) argues that PTSD is born
out of an era of presumed vulnerability over resilience as a normative reaction to traumatic events.
Whereas traumatic events destabilize people everywhere, the exportation of Western models of
PTSD to the rest of the world presupposes the construct’s applicability (Kagee & Del Soto, 2003;
Pupavec, 2002). The application of PTSD to foreign populations risks inadvertently pathologizing
people who might otherwise display significant resilience. A professional community’s expectation
of protracted symptomatology may be as much the agent of the perpetuation of posttraumatic stress
symptoms than the traumatic event itself. These dissenting perspectives suggest that caution should
be exercised in applying the PTSD model in non-Western cultures.
PTSD in Africa
A comprehensive review of PTSD prevalence rate studies found that only 6% of studies (8
out of 135) used samples from developing countries (De Girolamo & McFarlane, 1996). There is
predictably a paucity of data on the assessment of traumatic stress reactions in Africans in nonWestern countries. McCall and Resick (2003) found that 35% of a sample of Ju/’hoansi (Kalahari
Bushmen) of Namibia met criteria for PTSD; 85% reported at least some avoidance/numbing
symptoms, but not to the degree that DSM-IV criteria were met. Fox and Tang (2000) assessed a
sample of Sierra Leonean refugees in The Gambia and found that 49% of the sample yielded scores
indicative of PTSD. Eighty and 85% of the sample scored above the clinical cut-off levels for
anxiety and depression, respectively. Medicins Sans Frontières assessed for PTSD symptoms
among 245 Internally Displaced Persons (IDP’s) near Freetown, Sierra Leone, and found that 99%
of respondents had scores indicative of PTSD (Raymond, 2000). This study used the Impact of
Events Scale (IES; Horowitz, Wilner, &Alvarez, 1979), a measure not yet validated in this region of
the world, and that has been abandoned by other researchers after determining accurate translation
Western Trauma Discourse
5
to be unfeasible (Terheggen, Stroebe, & Kleber, 2001). Dyregrov, Gupta, Gjestad, and Mukanoheli
(2000) conducted a study one year after the Rwandan Genocide involving 1,830 Rwandan children.
Seventy-nine percent of the children exceeded the IES cutoff for PTSD. Some of the studies cited
above used supplemental measures to assess symptoms beyond the domain of PTSD while others
did not. The omission of an assessment of possible symptoms outside of the diagnosis of PTSD
risks failing to identify the full range of symptoms experienced.
A few studies have applied less structured interview techniques in an effort to capture a
breadth of symptoms. Paardekooper, de Jong, and Hermanns (1999) conducted semi-structured
interviews with 216 Sudanese children living as refugees and with 80 Ugandan children who had
not experienced war and flight. Sudanese refugees reported more traumatic events, more memory
disturbances, more worries about their future, and more suicidal ideation. Baron (2002) reported a
consistent pattern of anxiety, somatic complaints, depressive symptoms, estrangement from friends
and family, and loss of motivation to care for family and self among Sudanese refugees in Uganda.
Limitations of the current literature
The literature on traumatic stress reactions in people living in non-Western cultures is
limited and has yielded diverse results. The research to date has found highly variable prevalence
rates of PTSD and posttraumatic symptoms (Marsella, Friedman, Gerrity, & Scurfield, 1996). Many
studies suffer from various methodological limitations, including absence of back-translation, the
use of unvalidated measures, failure to assess symptoms beyond PTSD, the potential influences of a
pre-existing power imbalance between participant and researcher, social desirability, and the
possible benefits of secondary gain.
The effect of prior Western trauma discourse exposure
Western Trauma Discourse
6
In addition to these factors, we hypothesize that prior exposure to Western trauma discourse
(WTDE) may have inadvertent suggestive effects. Research has demonstrated that both direct and
indirect morbid suggestion can impact the nature of psychological symptoms (Beckman, 2003;
Rothman & Weintraub, 1995; Skelton, Loveland, & Yeagly, 1996). However, none of the studies of
PTSD reviewed above assessed the possible influence of prior exposure to Western
conceptualizations of the psychological response to traumatic events. Radio programs, written
literature, and psychoeducational workshops are potential sources of information as to how the
Western medical establishment conceptualizes traumatic stress reactions. Kagee and Del Soto
(2003) theorize that prior exposure to Western trauma discourse and its concomitant expectations
about protracted symptoms may increase the likelihood that posttraumatic symptoms will persist.
The present study aimed to examine whether prior WTDE would predict the nature and
severity of symptoms associated with response to traumatic events in a sample of rural Burundians
who had experienced traumatic events. We hypothesized that WTDE would be positively associated
with greater severity of PTSD symptoms, and that WTDE would be more highly associated with
PTSD symptoms than with general symptoms of anxiety, depression, and somatization.
Additionally, we predicted that material complaints and general psychological symptoms of anxiety
and depression would be more frequently endorsed than specific PTSD symptoms in response to an
open-ended interview.
Methods
Participants
Participants were recruited through a trauma healing and reconciliation program run by a
small nongovernmental organization in Burundi in central Africa. Following 30 years of
intermittent violence, a civil war erupted in 1993 between the mostly Tutsi government and Hutu
Western Trauma Discourse
7
rebel forces. As the two ethnicities lived interwoven in the same communities, widespread
neighbor-upon–neighbor conflict ensued, throwing entire communities into disarray. Studies
estimate that well over 200,000 people have been killed in the conflict since 1993 (AFSC, 2001).
Individuals who had been referred to the reconciliation program were invited to participate
in the study. Among the 78 participants, 28 (36%) were female and 50 (64%) were male. The mean
age was 37.7 years (SD = 13.6). Only 14% of the sample had completed more than six years of
education. The entire sample lived in a rural area in the north central region of the country
approximately 50 miles from the paved road connecting Burundi’s two largest towns. All
participants in this study had been directly victimized by violence during or since the civil war
began in 1993. The participants were referred to the workshop through a network of local elders
who identified them as experiencing ongoing distress related to traumatic events associated with the
civil war, with hopes that their distress would be ameliorated by the workshop. According to the
elders, the referred individuals were representative of the larger population of distressed individuals
in the community, and there was no reason to believe that they were particularly responsive to the
influences of Western trauma models. Participants received reimbursement for transportation
expenses, participated voluntarily, and gave fully informed consent to participate.
Participants were administered Part I of the Harvard Trauma Questionnaire (HTQ;
Mollica et al., 1992) to asses traumatic event history. Participants indicated whether they had
experienced, witnessed, heard about, or had no exposure to each event. The mean number of
types of events experienced was 9.5 (SD = 1.9) and the mean number of types of events
experienced, witnessed, or heard about was 16 (SD = 3.0). Events endorsed as directly
experienced included combat situation (100%), narrowly escaping death (78.2%), unnatural
death of a family member (71.8%), serious physical injury from combat (17.9%), and being
Western Trauma Discourse
8
forced to hide among the dead (12.8%). The sample had minimal formal education and limited
contact with Westerners. All had experienced one or more Criterion A events even by the
strictest of definitions.
Measures
All instruments were translated into Kirundi by native Burundian speakers with fluency in
both Kirundi and English who lived either in Burundi or in the United States. Burundian natives
then backtranslated the instruments and discussed refinements in a dynamic process with the
principal investigator.
Event history. Each participant’s history was collected using the HTQ (Part I), a 19-item
event checklist that specifies whether various traumatic events were directly experienced,
witnessed, or heard about. Mollica et al. (1992) report an interrater reliability of .93, internal
consistency of .90, and test-retest reliability of .89 for Part I of the HTQ.
Semi-structured interview of symptoms of distress. Using methods derived from Kagee
(2004), we utilized a semi-structured interview to assess symptoms associated with traumatic
response. Open-ended questions were used to solicit how each participant had been affected by a
self-identified “most-distressful” traumatic event. The central question was, “What are the main
problems that affect you as the result of those events?” (Wilk & Bolton, 2002; Kagee, 2004).
Secondary questions explored how participants remember the experience, what they associate with
it, and in what ways other people perceived them as different from prior to the experience. These
interviews in Kirundi were audio taped and then translated by the interviewer into English the same
day.
Interview responses were coded over a process of two reviews. In the first review, items
were categorized as to whether or not they represented a symptom of PTSD. In a second pass all
Western Trauma Discourse
9
remaining responses were sorted into other categories. Responses that had already been coded as
PTSD were not recoded into these additional categories, thereby effectively reducing the responses
that could have been considered for other symptom categories. The non-PTSD symptom categories
were derived by the first coder through a dynamic process in which categories were developed in
response to emerging themes in the data. A second coder was trained in the resulting categories and
the coding system. The second coder was then randomly assigned 25% of the data for coding. An
inter-rater reliability of 87.8% was calculated the percent agreement (i.e., dividing the number of
items agreed upon by the total number of items).
Quantitative symptom reports. The Hopkins Symptom Checklist-25 (HSCL-25; Hesbacher,
Rickels, & Morris, 1980) was used to assess symptoms of distress. The HSCL-25 was designed as a
self-report measure and uses a 4-point Likert scale (1 = not at all to 4 = extremely) for an anxiety
subscale (10 items) and a depression subscale (15 items). Given that somatic symptoms have been
reported as reactions to trauma in prior studies of non-Westerners (Marsella et al., 1996;
Paardekooper et al., 1999), the standard HSCL-25 was modified by adding the somatic subscale of
the HSCL-90 (Derogatis, 1994). The HSCL-25 total score can be used universally as a global
measure of emotional distress (Mollica, Wyshal, deMarneffe, Khuon, & Lavelle, 1987). When
matched to diagnoses based on clinical interview, the HSCL-25 has been shown to have a
sensitivity of .88 and specificity of .73 (Mollica et al., 1987), and internal reliability of .86-.95
across multiple languages (Kleijn, Hovens, & Rodenburg, 2001).
The Harvard Trauma Questionnaire (Part IV) (HTQ; Mollica et al., 1992) uses a 4-point
Likert scale (1 = not at all to 4 = extremely) to assess severity and nature of traumatic stress
symptoms. The HTQ – Part IV is a symptom checklist of PTSD symptoms as defined by the DSMIII-R. Mollica et al. (1992) reported a sensitivity of .78 and a specificity of .65 when validated
Western Trauma Discourse 10
against results from a diagnostic semi-structured interview. Mollica et al. reported an interrater
reliability of .98, internal consistency of .96, and test-retest reliability of .92 at one week for Part IV
of the HTQ. The HTQ has been translated into multiple languages and consistently yields sufficient
reliability (internal reliability of .74-.89 in Russian, Serbo-Croatian, Farsi, and Arabic; Kleijn et al.,
2001).
Western Trauma Discourse Exposure. Participants were asked to report on the degree to
which they had experienced trauma psychoeducation workshops, radio programs, and written
materials. These data was coded according to a scale based on rankings by Burundian natives
assisting with the project. Each Burundian had received a list of the components of the WTDE
construct and was asked to rank them as indicators of exposure to Western models of
traumatization. Once the data was collected, a second coder was utilized as described previously.
An inter-rater reliability of 85.0% was calculated based on the percent agreement method. Media
contact was determined by summing the scores of both trauma-related radio contact and traumarelated reading. Psychoeducational workshops were tallied independently.
Procedures
Interviews were conducted in Kirundi by two native Burundian staff. The principal
investigator was not present during the interviews but remained nearby to consult with staff in the
event of questions or ambiguities. Measures were administered verbally due to participant illiteracy.
The metric of the Likert scale was demonstrated visually with glasses containing varying degrees of
water (see Terheggen et al., 2001).
Each participant first endorsed traumatic event items on the HTQ – Part I. They then
responded to open-ended questions about current symptoms and prior Western Trauma Discourse
Exposure (WTDE). This was followed by the HTQ - Part IV, the three HSCL subscales, and a short
Western Trauma Discourse 11
sociodemographic form. The open-ended interview preceded the standardized measures to avoid
biasing participants’ reports with suggestions from these measures.
Results
Symptom reports
Anxiety, depression, and somatization. Mean scores on the HSCL subscales (anxiety,
depression, and somatization) are reported in Table 1. For the purposes of comparison, established
norms for various groups are also included (Derogatis, 1994). Anxiety and somatization were
markedly higher than found in a Western psychiatric inpatient sample, whereas depressive
symptoms were comparable to what would be found in such a sample. Mollica et al. (1987)
established a critical cutoff of 1.75 on the HSCL-25 indicative of “substantial distress” in a nonWestern southeastern Asian sample. In the present sample, 32.1% exceeded this cutoff on the
depression subscale and 57.7% on the anxiety subscale. Though Mollica et al.’s cut-off does not
specifically apply to the somatization scale from the HSCL-90, it is noteworthy that 56.4% of the
sample exceeded the cut-off in the somatization subscale. Thus, these nonspecific symptoms of
anxiety, depression, and somatization generally exceeded inpatient psychiatric norms.
Posttraumatic stress measure. Mollica et al. (1992) determined a critical cut-off of 2.5 for
the HTQ – Part IV in an indigenous southeastern Asian sample and stated that scores above this
threshold are indicative of being symptomatic for PTSD. Our sample’s mean score on the HTQ –
Part IV was 1.83 (SD = .47). Only 11.5% of the sample exceeded the cut-off.
The HTQ - Part IV also offers a traumatic stress construct that includes an additional 14
items intended to capture more culturally variable traumatic stress reactions. Mollica et al. (1992)
showed that the addition of these 14 items significantly improved the accuracy of the scale for the
Southeastern Asian sample used in their study. The authors state that the same cut-off of 2.5 is
Western Trauma Discourse 12
indicative of being symptomatic for PTSD. Only 9.0% of our sample exceeded this cut-off (mean
1.75 (.49)).
Exposure to Western trauma discourse
Participants reported the frequency with which they had been exposed to particular
sources of Western models of traumatization. The mean score for degree of exposure to Western
models was 5.9 (3.6). Frequencies for each question are found in Table 2. The majority (85.9%)
had never attended a workshop or similar training program on the topic of trauma. The frequency
with which participants had been exposed to trauma-related media (e.g., radio programs, written
material) was more variable and normally distributed, with the modal response, endorsed by
39.7% of the sample, indicating 3 to 4 instances of such exposure.
Primary hypotheses
WTDE associated with PTSD symptoms. Our first hypothesis, that WTDE would positively
correlate with severity of PTSD symptoms, was first assessed using the symptom report from the
HTQ score. This score represents responses to the items that reflect the specific DSM criteria for
PTSD, as opposed to Mollica et al.’s (1992) larger HTQ construct that includes 14 items reflecting
additional traumatic stress symptoms reported by refugees. Western trauma discourse exposure was
significantly correlated with traumatic stress symptoms as assessed (r = .28, p = .02). The more the
participants had been exposed to Western models of trauma, the more their symptoms fit a PTSD
symptom profile.
WTDE associated with PTSD symptoms relative to nonspecific symptoms. Our second
hypothesis was that WTDE would be more positively correlated with PTSD symptoms than with
non-PTSD symptoms. The WTDE and the combined HSCL three subscales were not significantly
correlated (r = .12, p = .31); as noted above, the correlation between WTDE and traumatic stress
Western Trauma Discourse 13
symptoms (HTQ) was significant (r = .28, p = .02). Hotelling’s test of the difference between two
dependent correlation coefficients revealed a significant difference between these two correlations (t
= -2.38, p =. 02). Exposure to trauma-related media and workshops was more strongly associated
with posttraumatic stress symptoms than with non-PTSD clinical symptoms.
A hierarchical multiple regression was conducted predicting traumatic stress symptoms, first
entering general symptoms (sum of 3 HSCL subscales) and then entering WTDE. This resulted in a
significant change in R2 (r = .85, R2 = .72, adj. R2 = .71, ∆R2 = .02 (p = .02); HSCL (Β = .81, p <
.001); WTDE (Β =.15, p = .02). Thus, even while controlling for more general symptoms of
distress, WTDE was significantly predictive of traumatic stress symptoms.
To better elucidate the relationship between traumatic stress symptoms (HTQ) and WTDE,
we reexamined this relationship while controlling for total of types of events experienced. A
hierarchical multiple regression was conducted predicting traumatic stress symptoms, first entering
total number of type of events experienced and then entering WTDE. Though ∆R2 fell just slightly
short of significance, the correlation and the amount of variance explained was notable (r = .53, R2
= .28, adj. R2 = .27, ∆R2 = .03 (p = .065); total type of events experienced (Β =.46, p < .001); WTDE
(Β =.19, p = .065). Thus, even while controlling for number of event types experienced, WTDE
showed a trend for predicting traumatic stress symptoms.
Response to open-ended questions. Frequency of endorsement of different responses to the
open-ended questions were counted. No participants endorsed sufficient symptoms qualitatively to
meet criteria for a PTSD diagnosis. Complaints of a material nature (70.5% had at least one
complaint) were comparable in frequency to psychological complaints of any kind (69.2% had at
least one complaint). PTSD symptoms were less frequent than depressive symptoms and as frequent
as anxiety symptoms. Frequencies are reported in Table 3.
Western Trauma Discourse 14
Discussion
Exposure to Western trauma discourse
The present results revealed that WTDE was a significant predictor of traumatic stress
symptoms, even when controlling for the number of event types experienced. These results suggest
that well intentioned psychoeducational efforts may be associated with an increase in PTSD
symptoms. It is possible that psychoeducation that forecasts vulnerability to pathology may actually
undermine resiliency or at least alter the symptom profile in the Burundian population. One might
conclude that WTDE is associated with a general increase across all symptoms categories. A partial
test of this would be to compare the correlations between WTDE and trauma symptoms and WTDE
and general symptoms. WTDE was also more strongly associated with posttraumatic symptoms
than with more general clinical symptoms of anxiety, depression, and somatic concerns. This
pattern of results suggests that trauma psychoeducation is not associated with an increase in clinical
symptoms generally, but rather with a specific increase in symptoms of PTSD.
Given the cross-sectional nature of the study design, we cannot determine the direction of
causal effects. It may be that individuals who were experiencing PTSD symptoms sought out
media content and workshops that described the symptoms they had acquired. However, that
WTDE and event history were so weakly correlated (r = .19) makes such an explanation less
likely. Determining the chronology of the occurrence of symptoms and of exposure was beyond
the scope of this investigation. Future research is needed to identify causal relationships.
Event history and symptom levels
Our sample was drawn from a population of rural Burundians, all of whom reported
histories of multiple extremely distressful events. These events included being forced to harm or kill
others, the murder of family members, and rape. In most cases, the worst of the events took place
Western Trauma Discourse 15
more than ten years prior to the investigation. An average of nine events was endorsed. Given that
the list was predetermined and given that there was no solicitation of additional events, it is likely
that the participants experienced additional traumatic events.
Despite significant histories of multiple traumas, the participants reported relatively low
levels of PTSD symptoms. According to Mollica et al.’s (1992) cutoff for the HTQ, only 11.5%
could be considered symptomatic for PTSD. One explanation for the low level of PTSD
symptomatology relative to the substantial trauma history is a gradual abatement of symptoms over
the years. However, this explanation stands in contrast to conventional claims that PTSD is
unremitting without treatment (e.g., traumatized Vietnam veterans who experienced trauma over 30
years ago; see Rosenheck & Fontana, 1994). Such findings are strikingly similar to Bryant’s (2004)
reports that the vast majority of people either recover naturally or are resilient such that they never
develop full-scale PTSD.
A second possibility is that participants were underreporting. However, the Burundian
interviewers, whose presence should have facilitated disclosure, reported that they did not see
indications of underreporting. It is difficult to support an argument for the possibility of the specific
underreporting of PTSD symptoms, when material complaints and certain other clinical symptoms
were endorsed at substantial levels. Underreporting is also unlikely given that the mean scores on
the HSCL anxiety and somatic subscales far exceeded inpatient clinical means.
A third explanation for the low levels of PTSD in this sample is that PTSD symptoms do not
accurately capture the type of posttraumatic stress reactions of these individuals. In the quantitative
data, many more participants exceeded Mollica et al.’s (1987) cutoff for substantial distress on the
HSCL than they did on the HTQ. In the qualitative data, response rates between estimates of PTSD
symptoms and nonspecific anxiety and depression symptoms were comparable. Clearly, a trauma
Western Trauma Discourse 16
history in this sample is associated with diverse elevated clinical symptoms as well as material
complaints, rather than being composed of predominately PTSD symptoms. The absence of severe
PTSD symptoms in the presence of profound and multiple stressors support continued skepticism as
to how well the PTSD construct captures the experience of traumatized indigent non-Western
peoples.
Qualitative symptom reports
The qualitative data reveal that PTSD symptoms are just one of a number of symptom
profiles associated with a traumatic history. Material complaints were endorsed as often as
psychological ones, and symptoms of anxiety and depression were more common than specific
PTSD symptoms. These findings are consistent with Baron (2002), who used open-ended questions
as opposed to symptom checklists to determine the nature of the distress people were experiencing.
The interview questions employed in the current study were adopted from Kagee (2004) who, while
finding some presence of PTSD symptoms among South African torture survivors, reported that
these symptoms are significantly outnumbered by somatic and economic concerns.
Study strengths and limitations
Our procedures did not allow for careful assessment of the specific content that was
captured in the WTDE construct. The data collected qualitatively may have been compromised by
inconsistent efforts on the part of the interviewers to follow-up responses to questions. Additional
training and preparation of the interviewers in future studies would strengthen the validity of the
qualitative data. Our assessments were largely symptom-focused, and as in most similar studies,
failed to assess functional impairment.
Despite these methodological limitations, the study possessed notable strengths. Measures
were carefully translated and backtranslated by native Burundians, and the study procedures were
Western Trauma Discourse 17
refined in consultation with them. Interviews were conducted entirely by local Burundian staff in
the native Kirundi dialect. The use of open-ended questions informed a more culturally sensitive
approach to interviewing and decreased the demand characteristics inherent in standardized
measures. The sample was markedly provincial with minimal exposure to Western culture. The use
of open-ended interview methods as a complement to standardized quantitative instruments permits
assessment of a clearer picture of traumatic stress reactions. The chronological relationship between
WTDE and the development of symptoms would be an important issue to examine in future
research.
The application of a standardized PTSD symptom measure without careful consideration of
the possible effects of social desirability, a power differential between researcher and an indigenous
sample, and the need for a broader assessment of symptoms may lead to a premature conclusion as
to the degree to which PTSD captures a universal response to trauma. The present results suggests
that posttraumatic symptoms among an indigenous African population are diverse, are not confined
to a discrete PTSD construct, are substantially material in nature, and may be subject to
psychoeducational or cultural influences. The support for the proposed hypotheses suggests that
additional research on the effects of WTDE on symptom presentation in such settings is
recommended. The current results speak to the importance of appropriate caution when presuming
vulnerability in non-Western populations, especially vulnerability constructed in the image of
PTSD.
The debate as to the degree to which PTSD is more biologically or culturally determined
will continue. The possibility remains that Western trauma models capture a traumatic stress
response that is as of yet unarticulated by indigenous groups. On the other hand, the degree to which
PTSD is “universal” may be substantially driven by the degree to which the cultural ideas inherent
Western Trauma Discourse 18
in contemporary Western trauma discourse are exported to foreign lands. Our findings do not aim to
minimize the intense suffering that our sample of participants reported. The morality of the horrors
they experienced is independent of a determination of the nature of their distress. Further research is
critical to discern the degree to which the application of Western trauma models promotes recovery
or constitutes a risk of shaping clinical symptoms and even pathologizing normal responses to
traumatic events.
Western Trauma Discourse 19
References
American Friends Service Committee (AFSC). (2001, August). Burundi Update. Philadelphia: Author.
American Psychiatric Association (APA). 1980. Diagnostic and Statistical Manual of Mental Disorders.
(3rd ed.) Washington D.C.: Author.
Baron, N. (2002). The Mental Health of Refugees and Internally displaced people. In Green, B.,
Friedman, M., DeJong, J. Solomon, S., Fairbank, J. & Keane, T. (Eds.) Trauma in War and
Peace: Prevention, practice, and policy. New York: Kluwer Academic/ Plunum Publisher.
Beckman, Mary. (2003). False memories, true pain. Science, 299, 1306.
Bryant, R. (2004). In the aftermath of trauma: Normative reactions and early interventions. In Rosen, G.
M. (Ed.) Posttraumatic Stress Disorder: Issues and Controversy (pp.187-213). Sussex, England:
Wiley & Sons.
De Girolamo, G. & McFarlane, A. C. Epidemiology of posttraumatic stress disorder among victims of
intentional violence: A review of the literature. In Mak, F. L. & Nadelson, C. C. (Eds.). (1996).
International review of psychiatry, Vol. 2. (pp. 93-119). Washington, DC: American Psychiatric
Association.
Derogatis, L. R. (1994). SCL-90-R: Administration, scoring and procedures manual-third edition.
Minneapolis, MN: National Computer Systems, Inc.
Dyregrov, A., Gupta, L., Gjestad, R., & Mukanoheli, E. (2000). Trauma exposure and psychological
reactions to genocide among Rwandan children. Journal of Traumatic Stress, 13, 3-21.
Fox, S. & Tang, S. (2000). The Sierra Leonean refugee experience: traumatic events and psychiatric
sequelae. Journal of Nervous and Mental Disease, 188, 490-495.
Herbert, J. D. & Forman, E.M. (in press). Posttraumatic Stress Disorder. In Fisher, J. E., & O’Donohue, W.
(Eds.), Practice Guidelines for Evidence Based Psychotherapy. New York: Springer.
Western Trauma Discourse 20
Herbert, J. D. & Sageman, M. (2004). “First Do No Harm:” Emerging guidelines for the treatment of
Posttraumatic Reactions. In Rosen, G. M. (Ed.) Posttraumatic Stress Disorder: Issues and
Controversy (pp. 213-232). Sussex, England: Wiley & Sons.
Hesbacher, P. T., Rickels, K., Morris, R. J. (1980). Psychiatric illness in family practice. Journal of
Clinical Psychiatry, 1980, 41, 6-10.
Horowitz, M. J., Wilner, N., & Alvarez, W. (1979). Impact of event scale: a measure of subjective
distress. Psychosomatic Medicine, 41, 209-218.
Kagee, A., & Garcia Del Soto, Arancha (2003). Internal displacement and trauma: The need for a
broader paradigm. In C. Brun & N. M. Birkeland (Eds.) Researching Internal Displacement:
State of the Art. Conference proceedings. Acta Geographica, Series A, No. 6, NTNU,
Trondheim, 229-243.
Kagee, A. (2004). Do South African former detainees experience post-traumatic stress? Circumventing
the demand characteristics of psychological assessment. Transcultural Psychiatry, 41, 323-346.
Kleijn, W. C., Hovens, J. E., & Rodenburg, J. J. (2001). Posttraumatic stress symptoms in refugees:
Assessments with the Harvard Trauma Questionnaire and the Hopkins Symptom Checklist-25 in
different languages. Psychological Reports, 88, 527-532.
Lynn, S. J., Knox, J.A., Fassler, O., Lillienfeld, S. O., & Loftus, E.F. (2004). Memory, Trauma, and
Dissociation. In Rosen, G. M. (Ed.) Posttraumatic Stress Disorder: Issues and Controversies
(pp. 163-186). Sussex, England: Wiley & Sons.
Marsella, A. J., Friedman, M. J., Gerrity, E. T., & Scurfield, R. M. (1996). Ethnocultural aspects of
PTSD: Some closing thoughts. In Marsella, A. J., Friedman, M. J., Gerrity, E. T., & Scurfield, R.
M. (Eds). Ethnocultural aspects of posttraumatic stress disorder: Issues, research, and clinical
applications. Washington, D.C.: American Psychological Association.
Western Trauma Discourse 21
McCall, G. & Resick, P. (2003). A pilot study of PTSD symptoms among Kalahari Bushmen. Journal of
Traumatic Stress, 16, 445-450.
McNally, R. J. (2003). Remembering Trauma. Cambridge: Harvard University Press.
McNally, R. J. (2004). Conceptual problems with the DSM-IV criteria for Posttraumatic Stress Disorder.
In Rosen, G. M. (Ed.) Posttraumatic Stress Disorder: Issues and Controversy (pp. 1-14). Sussex,
England: Wiley & Sons.
Mol, S. S. L., Arntz, A., Metsemakers, J. F. M., Dinant, G., Vilters-Van Montfort, P. A. P., &
Knottnerus, J. A. (2005). Symptoms of post-traumatic stress disorder after non-traumatic events:
Evidence from an open population study. British Journal of Psychiatry, 186, 494-499.
Mollica, R. F., Wyshal, G., de Marneffe, D., Khuon, K., & Lavelle, J. (1987). Indochinese versions of
the Hopkins Symptom Checklist-25: A screening instrument for the psychiatric care of refugees.
American Journal of Psychiatry, 144, 497-500.
Mollica, R. F., Caspi-Yavin, Y., Bollini, P., Truong, T., Tor, S., & Lavelle, J. (1992). The Harvard
Trauma Questionnaire: Validating a cross-cultural instrument for measuring torture, trauma, and
posttraumatic stress disorder in Indochinese refugees. Journal of Nervous and Mental Disease,
180, 111-116.
Paardekooper, B., de Jong, J.T.V.M., Hermanns, J.M.A. (1999). The psychological impact of war and
the refugee situation on South Sudanese children in refugee camps in Northern Uganda: An
exploratory study. Journal of Child Psychology and Psychiatry, 40, 529-536.
Pupavec, V. (2002). Therapeutising refugees, pathologizing populations. UNHCR, New Issues in
Refugee Research, 59.
Raymond, N. (2000). The trauma of war in Sierra Leone. The Lancet, 355, 2067-2068.
Western Trauma Discourse 22
Rosen, G. M. (2004). Further Thoughts on Criterion Creep, and the Creation of Pretraumatic Stress
Disorder. Scientific Review of Mental Health Practice, 3, 46-47.
Rosenheck, R. & Fontana, A. (1994). Long-term sequelae of combat in World War II, Korea and
Vietnam: A comparative study. In Ursano, R. J., McCaughey, B. G., Fullerton, C. S. (Eds).
Individual and community responses to trauma and disaster: The structure of human chaos. (pp.
330-359). New York: Cambridge University Press.
Rothman, A. L. & Weintraub, M. I. (1995). The sick building syndrome and mass hysteria. Neurologic
Clinics, 13, 405-412.
Shephard, B. (1999). “Pitiless Psychology”: The role of prevention in British military psychiatry in the
Second World War. History of Psychiatry, 10, 491-510.
Skelton, J. A., Loveland, J. E., Yeagley, J., L. (1996). Recalling symptom episodes affects reports of
immediately-experienced symptoms: Inducing symptom suggestibility. Psychology & Health,
11, 183-201.
Summerfield, D. (1999). A critique of seven assumptions behind psychological trauma programmes in
war-affected areas. Social Science and Medicine, 48, 1149-1462.
Summerfield, D. (2004). Cross-cultural perspectives on the medicalization of human suffering. In
Rosen, G. M. (Ed.) Posttraumatic Stress Disorder: Issues and Controversy (pp. 233-245).
Sussex, England: Wiley & Sons.
Terheggen, M., Stroebe, M.., & Kleber R. (2001). Western conceptualization and Eastern Experience: A
Cross-cultural study of traumatic stress reactions among Tibetan refugees in India. Journal of
Traumatic Stress, 14, 391-403.
Wilk, C. M. & Bolton, P. (2002) Local perceptions of the mental health effects of the Uganda AIDS
epidemic. Journal of Nervous and Mental Disease, 190, 394-7.
Western Trauma Discourse 23
Table 1
HSCL mean scores and norms
__________________Current Sample________ Published Norms ______________
Sample mean Nonclinical Psychiatric
meana
outpatient meana
Psychiatric
inpatient meana
HSCL-25 total
1.83 (.54)
.33 (.37)
1.63 (.91)
1.61 (1.07
Depression subscale
1.68 (.52)
.36 (.37)
1.79 (.94)
1.74 (1.08)
Anxiety subscale
2.07 (.69)
.30 (.37)
1.47 (.88)
1.48 (1.05)
Somatization subscale 1.96 (.58)
.36 (.42)
.87 (.75)
.99 (.84)
3 subscales combined 1.87 (.52)
_______________________________________________________________________
Note. Standard deviations in parentheses. HSCL = Hopkins Symptom Checklist.
a
Derogatis, L. R. (1994). SCL-90-R: Administration, scoring and procedures manual third - edition.
Minneapolis, MN: National Computer Systems, Inc.
Western Trauma Discourse 24
Table 2
Responses to questions regarding exposure to Western Trauma Discourse
Have you ever attended workshops or trainings about how people are affected by extremely
frightening or traumatic events?
Never: 85.9%
< 1 day: 7.7%
< 2 days: 1.3%
2 days:
1.3%
2+ days: 3.8%
Have you ever listened to radio programs/read literature about how people are affected by
extremely frightening or violent events?
Never:
19.2%
1-2 times: 16.7%
3-4 times: 39.7%
4+ times: 15.4%
7+ times:
9.0%
Western Trauma Discourse 25
Table 3
Percentage of participants reporting one, two, or more than two psychological symptoms or other
complaints in reference to their “most distressful” event.
Number of responses within each category
Category
1
2
>2
Met diagnostic criteria
for PTSD/MDE
PTSD
30.8% 6.4%
2.6%
None
MDE
20.8% 1.3%
0%
None
Nonspecific depression 38.5% 2.6%
0%
Nonspecific anxiety
30.8%
3.8%
0%
Material
70.5% 15.4%
3.8%
Somatic/medical
23.1%
2.6%
0%
Anger
11.5%
0%
0%
Bad/evil thoughts
21.8%
1.3%
0%
Thoughts of revenge 12.8%
0%
0%
Note. PTSD = Posttraumatic Stress Disorder. MDE = Major Depressive Episode.