RITUALS, SOCIAL SHARING, SILENCE, EMOTIONS AND COLLECTIVE MEMORY CLAIMS IN THE CASE OF THE GUATEMALAN GENOCIDE

Carlos Marín Beristain, Darío Paez^* and José Luis González^**

University of Deusto, * University of the Basque Country and ** University of Burgos

In this article we will review both theory and data pertaining to the emotional effects of collective remembering by means of social sharing and funeral rituals. Data is based on the experience of Guatemalan Mayas, who were victims of a genocide during the nineteen eighties. Results show that that Mayan subjects did not report lower levels of subjective and emotional reactions compared to the Latino community. A finding which contradicts the idea that this is a less expressive and emotional culture. Rituals had a more important buffer effect for Mayas in comparison to Ladino communities. Material losses did not imply, and affect, both groups in the same way. Commemoration activities were most important for those subjects who had been affected by collective massacres. Silence was an adaptive form of coping in the past, although that may not be the case at the present moment.

Rituales, comunión social, silencio, emociones y demandas de Memoria Colectiva en el caso del genocidio Guatemalteco. Este artículo examina la teoría y los datos sobre los efectos emocionales de los procesos de memoria colectiva como los rituales funerarios y la comunión social. Los análisis empíricos se basan en la experiencia de los Mayas Guatemaltecos, quienes fueron víctimas de un genocidio durante la década 1980-1990. Los resultados muestran que los Mayas no declaran menos reacciones emocionales y subjetivas en comparación con los guatemaltecos ladinos. Esto contradice la idea que los Mayas comparten una cultura menos expresiva y emocional. Los rituales tienen un efecto amortiguador más importante para los Mayas que para los Ladinos. Las perdidas materiales han golpeado y afectado a ambas comunidades, ladinos y mayas, de forma diferente. Las actividades de conmemoración fueron más importantes en el caso de personas que vivieron masacres colectivas. El silencio fue una forma de afrontamiento adaptativa, aunque puede no serlo en el momento actual.

Collective trauma in Guatemala

70% of the Guatemalan population is of Mayan origin. Four million Mayas live in the highlands. The 1981-1986 army counter-guerrilla campaign produced an intense social impact: a) 50.000 adults (a large majority of them civilians) have been killed since 1980 in order to wipe out an estimated 3.500 guerrilla soldiers and 10.000 local irregular forces. The US Embassy (not a likely support for the guerrilla forces) estimated that the guerrilla had between 30 and 60 thousand actively organized supporters and that 80% of them were mayan. Approximately 200.000 people were killed or disappeared (ODHAG, 1998). b) One million people were displaced (25% of the highland population), c) USAID estimated than 400.000 people had been hit by repression (10% of the highland population suffered the destruction of harvest, homes, disruption of normal activities and displacement), d) war and collective repression provoked 400.000 refugees in Mexico, Belize and the USA, e) 20.000 subjects organized in Popular Resistance Communities (illegal and allegedly pro-guerrilla communities) lived in the jungle under army fire and 50.000 refugees around the mexican border were in similar conditions, f) 700.000-900.000 Mayas had been recruited for civil patrols (local counter-guerrilla forces). Nearly 10% (mostly those who worked the land and were more than 18 years old) of the guatemalan population has been militarized (ODHAG,1998).

Historical and cultural background

In order to understand the emotional experiences, rituals and collective memory claims of the indigenous guatemalan people, we shall summarize some aspects of their culture. Mayan culture has been under spanish and latinamerican domination for 500 years. This culture assimilates external pressure: today’s indigenous dress was imposed by the spanish and is now a symbol of ethnic pride. Cofradias (syncretic catholic brotherhoods) and individual land ownership, both originally imposed, are now pillars of indigenous society (Wearne, 1994). Before armed insurrection and state repression, mayan culture had been eroded by the lack of land and by migration from the highland to the plantations on the Pacific coast of Guatemala. Guatemala has the most unequal land distribution in Latinamerica. In 1979, 88% of all guatemalan farms were too small to meet household food requirements. Traditional mayan earth cult was not dominant even before state repression. Catholic chatequists during the nineteen seventies criticized elders’ domination and catholic brotherhoods, and in doing so questioned alcohol consumption, dancing and the use of the saint’s images, all central to cofradia celebrations (Wilson, 1995). Mayan communities were not ideal islands of harmonious life: class differences were important and ethnic clashes took place between different groups.

However, the war stopped the communities’ traditional rites. War accelerated social changes, reinforcing the incorporation of indigenous population into national guatemalan society. This brought about, in these groups, a massive process of acculturation. The economic crisis and repression displaced a large number of people to the cities, to other lands and into exile, and reinforced the commoditization of rural labor relationships. On a cultural level, war fostered universal indigenist beliefs and provoked an ethnic revival (ODHAG,1998; Wearne, 1994; Wilson, 1995).

a) The maya - and guatemalan - culture is a collectivistic or sociocentric culture, marked with less materialism and to some extent less individualism in comparison to Latinamerican mestizo cultures. For the indigenous culture the value or worth of a person and all his identity marks (language, clothes, religion) are linked to the community. Social identity refers usually to the village or community first, linguistic group second, and only recently, after the turnmoil suffered during the nineteen eighties, has a mayan identity emerged (Wearne, 1994; Wilson, 1995).

A collectivistic culture probably reinforces the impact of repression in community members other than the nuclear family. This could lead us to state that in a collectivistic culture such as this there could be a larger amount of resilence among these subjects. Some data with Latinamerican (mexican) subjects confirms that events happening to members of the social network (other than the self, spouse and children) were important predictors of emotional distress but not for more individualistic subjects such as mexican americans or caucasians in the USA (Hough, Canino, Abug & Gusman, 1998).

b) The maya - and guatemalan - culture is also a high power distance society in which respect towards elders and local officials, responsability, honesty and hard work are valued (Wearne,1994; Wilson, 1995). The balance and harmonious relationships with nature and the respect due to people are a part of mayan cosmovision. In mayan culture, any indication that someone is hostile implies that the person is not good and cannot be trusted. Moreover, in Latinamerican culture health and well-being are related to social relationships and tranquility. Anger is a threat which disrupts social relationships (Woodrick, 1995). Indigenous subjects are less prone to anger and their emotional balance is reflected in a lower level of accidents and violence in comparison to mestizos (Wearne, 1994). Self-modification and indirect control is the preferred coping style in Latinamerican societies similar to the guatemalan culture such as Mexico (Diaz-Loving, 1998). Among the mayas we find a more subtle expression of anger, not its absence. As Zur (1996) noted, mayas tend to express distress by means of somatization or somatic metaphores and not by catharsis, a word which in fact they do not have in their language. Direct expression of emotions is viewed as a factor which may jeopardize social relationships and one’s social status.

Cultural rejection of anger and self-modification coping may help normalize stress by means of resignation and stoicism when faced with suffering (Hough, Canino, Abueg & Gusman, 1998).

c) Farming is the traditional maya activity and land ownership is central to social identity. Milpa (plots of mountain lands) and maize are major facets of mayan culture. A milpa is an essential symbol of mayanity or mayanness. Land has a religious significance: it is the home of the most important maya god. Land also produces the sacred maize which is the basis of every meal. Traditional mayan subjects see themselves analogically as ‘Men of Maize’. Representations of the landscape are central to community and individual mayan identities (Wilson, 1995).

Material and animal loss are usually also a symbolic loss, a very important factor in a culture in which land is essential for social identity. Displacement disrupts the relationship with the mountain gods, and impedes the undergoing of transition and mourning rituals.

d) Mayan subjects develop syncretic religious practices, integrating «magical» rituals in catholic forms. Mayan culture has an all-pervading sense of the religious, the magical and supernatural. Animals and nature are related to humankind, they command love and respect and are personifications of good and of ancestors’ souls. Some, but not all, of the mayan ethnic groups believe in animal alter egos, usually called nagual. An animal spirit companion of the person lives in the mountain, and spirits reside in each mountain (Wilson, 1995). Tzotzils, an important minority of the indigenous inhabitants of Mexico, and present also in Guatemala, have a spiritual concept of the self, expressed in the idea that an animal soul defines and influences personal destiny. Animal souls and person co-essence are a repeating figure in mayan folk theories of well-being and fate (Gossen, 1994).

Violence directed towards land and animals also implies inflicting symbolic wounds to the earth and god and potentially to souls and co-essences. Although the importance given to animals is a feature of mayan culture it is nevertheless less prevalent than spiritualism and the importance attached to the land.

e) A holistic and analogical mode of thought, opposed to the western individualistic analytic and categorical mode of thought, characterizes maya culture (León-Portilla, 1986). Things, people and earth are conceived of as interrelated. Indigenous thought is more holistic and analogical, and so many western analytic categories cannot be found. Death does not mean absence of life, and there are different forms of life. For Mayas, in a way similar to that of buddhists, the deceased community and family member becomes an ancestor, with whom the bereaved may remain in contact. From a cultural point of view for the mayas death is not the opposite of life. The dead are «alive» in the community. Ancestors are present in everyday life. The living and the dead are both on Earth. Cleansing corpses and offering drinks or clothes to the dead are important aspects of mayan funerals. People offer cigarettes and alcohol to the dead in commemorations. Exhumations, secundary burials of the bones of the deceased and annual celebrations are rituals which reinforce a positive relationship between the deceased and the bereaved in mayan culture (Woodrick, 1995).

Absence of these rituals disrupt the normative positive relationship between the living and the dead, and cultural beliefs and practices probably reinforce a sense of presence of the deceased (Averill, 1979).

Massive loss, mourning rituals and collective memory

Grief is a basic reaction to loss. Mourning rituals, funerals and commemorations were intended to allow open expression of grief through ritual. Some features, typical of collective violence, like sudden death, massive loss, and absence of mourning rituals, are risk factors for chronic grief symptoms (Worden, 1991). Because of the importance of corpse related rituals and the belief that ancestors are present in daily life, massive loss, sudden death and absence of rituals could be even more negative in the case of the mayas. Many survivors of the guatemalan genocide did not bury their beloved - they did not know what had happened to their relatives, or where the corpses were. Repression also did not allow funeral rites and ceremonies. Displacement outside of the country, absence of corpses or repression hindered funerals and rituals. Mutilated corpses also provoked pain («they were killed like animals») and inhibited normal rituals (ODHAG,1998). Authors such as Parkes and Weiss (1983) posit that collective commemorations and mourning rituals have a functional role in order to assimilate loss. On the other hand, an absence of mourning rituals are a risk factor for psychological reactions and problematic grief (see Worden, 1991). Freud (1917) conceived the work of mourning as the process by which a person reorganizes the internal world after bereavement. In the same vein for Malinowski a funeral rite and a ceremony function to reduce or aminorate the tensions created in individuals by death (Howard, 1989). Some studies confirm the positive effects which commemorations have on subjective well-being. For instance, Levav et al. (1988) found that fathers and divorced/widowed mothers of a son killed in an accident showed higher mortality than fathers and divorced/widowed mothers of a son killed in the 1973 Yom Kipour war. This suggests the positive role which collective commemorations and rituals may have on health. Notwithstanding, data concerning the positive effects on individual emotional states of commemorations and rituals is scarce. For instance, in a sample of 30 bereaved partners of men with Aids, participation in, and acceptance of, ceremonies of leave-taking as satisfactory were unrelated to psychological symptoms 12 months later (Weiss & Richard, 1997). One explanation to this contradictory picture is that factors which correlate with objective health and behavior may be uncorrelated with distress and emotional experience (Pennebaker, Mayne & Francis, 1997) or can be negatively correlated with emotional recovery.

Mourning rituals and activities of collective remembering may benefit the individuals and also serve important social functions. Mourning rituals, memorials and commemorations, like other rituals of transition, insert individuals into the whole society, reaffirming its continuity (Averill, 1979). Durkheim argued that rituals intensify shared emotion and bind together those who have had their emotions augmented in a feeling of solidarity. The loss of a group member entails funerary rites. There is a coming together and a closeness. The results of mourning rituals are similar to more positive rites and commemorations: an emotional effervescence through the sharing of a common feeling, even if it is sadness. A sense of unity with others arises and even when facing death, provokes a renewed interest in life. In sum, rituals reinforce emotions and strengthen social cohesion and mobilization (Durkheim 1912/1982; Kemper, 1993). According to Radcliffe-Brown’s structural-functionalist approach, strongly influenced by Durkheim, funeral rites are one of the elements which help maintain social order and group cohesion - the behavior of the bereaved serves to reaffirm the values of society and promote group solidarity (Howard, 1989). Available data on collective disasters or catastrophes confirm that rituals (e.g. memorial services) help people strengthen common links between members and come to terms with events by providing a structure (Hodkingson & Stewart, 1991).

Halbwachs, a Durkheimian sociologist who wrote on the topic of collective memory, proposes that commemoration and rituals are a form of collective remembering which supports a moral self-definition and helps consolidate memories of important events. Remembering is a normative process which allows people to have a personal and social identity, to transmit and learn a moral lesson (Páez et al., 1997; Halbwachs, 1950/1968). In a similar vein, Frijda (1997) posits that commemorations and rituals actualize feelings towards deceased people, both positive like affection, but also negative such as sadness and anger. Emotions stemming from past traumatic collective events are usually alive and acute, and commemorations re-actualize these emotions. By this token reviving bonds with the deceased confirms a person’s social identity and is a step towards the reappropriation of the past - an effort to extract a moral lesson from the past which supports a moral self-definition (Halbwachs, 1968; Frijda, 1997; Pennebaker, Paez & Rimé, 1997).

Participation in rituals and social sharing not only helps to enhance social integration and restore self-concept and self-esteem, but also helps to construct a collective memory (Rimé et al, 1998). It may be possible that participating in rituals reinforces emotions such as sense of injustice and also reinforces claims for collective remembering and sociopolitical demands.

Culture, silence and social sharing of emotions

Different authors propose that inhibition is unhealthy and that social sharing and disinhibition helps to overcome emotional and traumatic events (Pennebaker, 1990). Evidence also shows that the social sharing of emotional events intensifies affective reactions or that social sharing does not help to bring about emotional relief, even if data supports the idea that social sharing is associated with improvements in health and subjective well-being in the medium and long term (Rimé et al., 1997). Rimé et al suggest that social sharing helps restore the self-concept, affected by an emotional event which challenges the systems of beliefs about oneself, others and the world. Social sharing also allows the social milieu to provide the affected person with culturally accepted ways of defining, expressing and coping with the experience.

Silence may be imposed by a situation of generalized insecurity and fear. Some authors posit that inhibition of social sharing accentuates the distinctions between the self and others, thereby destroying solidarity among groups. Self-silencing exacerbates people’s feeling of isolation and reduces opportunities for collective response (Lykes, 1994). At the same time, silence in response to social disaster and repression is often an adaptive strategy for survival (Comas-Diaz, Lykes & Alarcon, 1998). Correlational and experimental evidence suggests that in the midst of a threatening situation, such as a social disaster, people hold out their emotions and prefer not to talk about the current disaster. Emotional inhibition and silence are quite adaptive if feelings may interfere with performance and if such strategy allows people to avoid punition. Once the social disaster is over, subjects can reflect on the trauma and their emotions (Pennebaker, 1990, p.164).

It seems reasonable to believe that these are cultural differences concerning emotional vividness, frequency and effects of emotional inhibition and silence. In our case these differences would be between the ladino or mestizo culture, more individualistic and relatively westernized, and the more collectivistic indigenous mayan culture. This latter culture values harmony and so the verbal non-expression of negative emotions is quite frequent. On the other hand, non verbal communication is very important due to the fact that mayan culture has a highly contextual communicational style. For instance refusals are not stated directly, they must be infered from non verbal and paralinguistic communication, people must learn to understand when a «yes» means a «no» (ODHAG, 1998).

Collective trauma and emotions as antecedents of collective memory

Some type of traumatic events, like wars and genocides, reinforce the need to perform rituals (Frijda, 1997). Moreover, collective threats and repression provokes not only more social disruption, but more emotional reactions and also more social mobilization (Martin-Baró, 1990). Different emotions, due to their specific social functions, can reinforce claims for collective remembering. Sadness, the dominant emotion in grief, by strengthening social bonds, increases social cohesion and group survival. Sadness is also supposed to enable to take a careful look at reality, accept loss, and facilitate plans for a better performance in the future. Fear, another dominant emotion in the social climate shared by victims of violence, motivates flee and flight, and by focusing attention on the threat can guide self-protective behavior. Nevertheless, fear may not be a socially sanctioned emotion, and as such, be much more inhibited than anger. Finally, anger and sense of injustice mobilize energy and sustain goal-directed activity - not necessarily aggression. Authors such as Oberschall posit that anger and a sense of injustice are the psychological basis of social revolt. However, it is important to notice that the collectivistic mayan culture has a negative attitude towards anger and reinforces inhibitions (i.e. an angry man is not allowed to be close to an infant).

Method

Sample

6.494 interviews were collected in 18 regional departments concerning the army repression and violence which had taken place during the period 1981-85. 400 interviewers participated in the field work. 61% of interviews were conducted in 15 native languages - 39% in Spanish, 19% in Qéqchi, 15% in Ixil and 11.8% in Kíche. It was a volunteer sample and data was collected during the period 1995-97. Data finally used in the statistical analysis comprised 3424 subjects, this was due to the fact that comparing results using 2.000 interviews and 3.400 results were similar.

Instrument

A structured interview with open answer questions was used. Seven questions were asked: a) What happened; b) Where and when?; c) Who? d) Effects of violence in persons, families and collectivities, e) How do you cope with the events; f) Why? attribution of causality; g) What to do to prevent violence? The Catholic Church sponsored the field work by means of the Reconstruction of Historical Memory Project (REHMI) of the Guatemalan Catholic Church Human Rights Department (ODHAG). Interviewers were trained in 220 workshops. Interviewers belonged to mayan communities and of course were fluent in their language and usually bilingual (they also spoke spanish). Interviews were usually tape recorded. All were translated to spanish by bilingual staff and were coded by trained coders. Intercoding agreement was above 70% after training. Code categories were the following : a) type of repression: individual, collective; murder, missing people, torture, b) antecedents and nature of violence; who: army, police, paramilitary forces, civil patrols; c) psychological reactions: fear; sadness; anger; helplessness; rumination; nightmares; sense of injustice; uncertainty; prolonged and intense grief; d) losses within one’s family: spouse, father/mother, son, or other members; e) family crisis: family disintegration, economic problems, change in roles, overload; f) material losses: houses, harvests and animals; g) community effects: disorganization, mistrust, community harassment; h) individual and collective coping: individual flight; collective flight; exile; displacement; reconstruction of social groups and social support; vigilance; community organization; silence and emotional inhibition; talking and searching for emotional social support; direct informational coping; self control and contention; religious coping; enduring extreme conditions; political engagement; positive reappraisal; helping others or altruistic behavior, contention and self-control; i) explanations for violence (God, sociopolitical, ethnic, etc.) and what to do to prevent violence and with respect to the traumatic past: good will, reclaiming Human Rights, to know the truth, no impunity, demanding punishment for those responsible, the Church should act, collective remembering of the deceased; reparation to victims; the army and guerrilla should change, sociopolitical change is needed.

The code categories were constructed by three members of the REHMI professional team, and were based on previous work (20 focus groups with refugees on emotion prototypes and coping), inductive construction of categories using the first 50 interviews, and a review of literature on PTSD, coping and collective behaviour.

Code categories were not psychopathological: fear, sadness and anger were conceived of as normal emotions or normal reactions to extreme situations. For instance, prolonged and intense grief does not imply a psychopathological criteria for chronic or complicated grief (Prigerson et al., 1995). However the criteria was similar to the main feature of complicated grief: the person being interviewed could not speak of the deceased without experiencing fresh and intense grief (Worden, 1991, p.75).

Results

Frequency of types of repression, rites, emotional reaction, coping responses and collective memory claims

Frequencies in the 3424 interviews were the following : a) type of repression: individual murder 31%, missing people, torture, bomb threat 38%; collective massacre, 26%, individual murder and collective massacre, 5%. b) Only 49,5% of the subjects know were the corpses are and only 34% performed funerary rites - and usually under the threat of repression. c) psychological reactions: past fear, 31%, and present fear 26%; sadness past, 29%, and present 15%; sense of injustice past, 15,4%, and present 9,1%; prolonged and intense grief past, 6.9%, and present 8.1%, current intense grief, 13.7%. d) individual and collective coping: direct and cognitive coping, 52%; displacement and collective flight, 35%; helping others or altruistic behavior,14%; vigilance, 9,3%; social sharing and searching for emotional support 6,5%; reconstruction of social groups and social support from one’s family, 4,5%; community organization; contention and self-control, 4,6%, silence and emotional inhibition, 4,1%; enduring and resisting extreme conditions, 4,7%; political engagement, 3,1%; positive reappraisal. e) what to do to prevent violence and with respect to the traumatic past: reclaiming Human Rights, 17,8%; to know the truth about the traumatic past, 9,2%; God’s will, 9,2%; no impunity, punishment for those responsible, 8,1%; the Church should act, 3,2%; reparation to victims, 4,2%; collective remembering of the deceased, 1,7%; the army and guerrilla should change, 3,9% and 1,5% respectively, sociopolitical change is needed, 5,3%.

Dimensions of emotional reaction, coping responses and collective memory claims

Psychological responses were submitted to exploratory factor analysis and to categorical factor analysis (Prinscal). Two factors were found. Sense of injustice, anger, despair and uncertainty were related and load in the first factor. Sadness, fear and psychosomatic problems (starvation) were related and load in the second factor. Intense grief loads in both factors and was analyzed as a specific reaction. Items related to family losses, effects of repression on the community and material losses converged in one factor per area.

Coping responses were submitted to exploratory multivariate analysis. Four clusters or dimensions appeared. In a first dimension of confrontative and instrumental coping we find altruistic behavior, cognitive coping, reconstruction of social support, searching for information and reappraisal. In a second dimension of adaptive emotional coping there converges social sharing, self-control and contention, enduring an extreme situation, silence or inhibition of social sharing and direct cognitive coping. A third dimension was composed by collective flight or displacement and coming back. A fourth dimension comprised transformation of community organization and taking precautions when facing danger.

Factor analysis of items concerning losses of a family member (spouse, parents, sons and other members), family crisis items (family disintegration, economic problems, change in roles, overload), material loss items (houses, harvests and animals) and community effects items (disorganization, mistrust, community harassment) produces only one factor by domain. Higher factor scores mean strong family losses and crisis, material losses and community problems.

Responses to the question what to do to prevent violence were also submitted to the same multivariate exploratory analysis. Three dimensions were found. In a first factor we encounter: God’s will; active protection of Human Rights; to know the truth about the traumatic past, to put an end to impunity, reclaiming justice and punishment for those responsible of the massacres, and that the Church should do something. This dimension represents the reappropriation of the past and prevention of violence. Responses stating that the army and guerrilla should change, and that sociopolitical change is needed load in the second factor or dimension. This dimension represents sociopolitical change demands. Finally, reclaiming mourning rituals, respect and restauration of the dignity of the deceased and reparation of victims load in a third factor, which represents a commemoration and moral and material reparation dimension.

Comparison between mother tongue (indigenous or spanish) interviews

In order to analyse the results, two types of analyses were conducted. Pearson correlations and multiple regressions were performed using total standarized or factor scores between psychological reactions, being a victim of collective massacre, the social sharing and silence dimension of coping, participating in funeral rituals and different dimensions of collective memory claims. Crosstab analyses were performed between participation in funeral rites, burials and ceremonies, and the pyschological and coping responses. Results are presented in the form of Phi coefficients, a dichotomic form of an association coefficient such as the Pearson correlation.

Psychological responses were differenciated in past emotions (when the account confirms that subjects talked about emotions lived in 81-85) or current emotions (when the account confirms that people talk about current emotions).

The interviewed subjects’ mother tongue was unrelated to a global score of mental and emotional reactions, disconfirming the idea that mayas are more self-constrained and less emotionally expressive. However, those whose mother tongue was spanish show a higher report of the anger and sense of injustice dimension (Spanish=1; Indigenous language=2; r (3127)=-.07, p<.001), supporting the hypothesis of a higher cultural rejection of anger and hostility in the maya culture. Having spanish as one’s mother tongue was related to a higher report of both social sharing (r=-.07, p<.001) and silence (r=-.10, p<.001), suggesting that the mayas show lower levels of coping. Having maya as mother tongue was slightly related to a higher report of fear and sadness (r=.03, p<.07).

The indigenous population also talk more about material losses (r=.05,p<.03), especially those concerning plots of land or harvests, and not so much those regarding animals. This may be due to various reasons, one would be related with the fact that these animal and material losses are also considered to be symbolic losses in which the earth, one’s ancestors and even God are inflicted a series of «wounds». Another answer to this situation would be to consider events just as material (and not symbolic) losses. As we have seen the mayas scored higher on the fear/sadness and starvation factor, and so we could posit that these losses were in fact very important and real economic losses and so affected the community’s and family’s economic development. Mayan subjects reported less community crisis (r=-.05, p<.003).

The mayan population talks on average less than the spanish speaking population (r=-.10, p<.001). This last group of subjects seem to prefer a more direct coping strategy (r=-.049, p<.03), self-control and contention (r=-.04,p<.03). They also are more in favour of getting to know the truth of what happened in Guatemala and putting an end to the impunity of government officials, police forces, the army, etc. (r=-.07, p<.004) and sociopolitical changes (r=-.032,p<.003). To conclude, spanish speaking subjects report higher anger related reaction, higher coping reactions, community crisis and collective memory claims than mayan subjects. Mayan subjects report more fear reaction and material losses than spanish speaking persons.

Associations between collective massacre and social sharing, silence, funeral rites and emotions

In order to contrast the hypothesis concerning the higher impact of massive and sudden loss on emotional reactions, correlations were performed between being victim of a collective massacre (Yes=2,No=1), participation in rituals, social sharing, silence and factor scores of past and present emotional reactions. Being a victim of collective massacre was associated to past sadness/ fear (r=.15, p<.001), to past anger/sense of injustice (r=.13, p<.001), to participation in funerary rites (Phi=.12,p<.001), to social sharing (Phi=.06,p<.001) and to silence (Phi=.035,p<.03). Being a victim of collective massacre was also associated to present sadness/ fear (r=.12, p<.001) and to present anger/sense of injustice (r=.09, p<.001). Being a victim of collective massacre provokes higher emotional reactions, more interpersonal and ritual coping.

Effects of collective remembering by means of funerals, burials and ceremonies on emotions

In order to contrast Freud and Malinowski’s assumptions concerning the positive emotional effects of conducting rituals, a series of correlations between the presence/absence of funeral rituals and psychological reactions (total z scores and specific reactions) were performed.

Participation in funeral rites was associated with past sadness/fear (r=.17, p<.001), with past and present anger/sense of injustice (r=.13, p<.001 and r=.064, p<.003), to social sharing (Phi=.08, p<.001) and silence (Phi=.09, p<.001). However, participation in funerary rites was unrelated to present fear/sadness.

Participating or conducting funeral rites was associated in the past to enhanced sadness (Phi=.18, p<.00), fear (Phi=.12,p<.00), sense of injustice (Phi=.11,p<.00), helplessness (Phi=.088,p<.00) loneliness (.059,p<.00), anger (Phi=.049,p<.02) and intense grief (Phi=.033,p<.08).

Participation in, or conducting, funeral rites was unrelated in the present to fear, sense of injustice, anger and intense grief. Taking part in, or performing, funeral rites was associated in the present to enhanced sadness (Phi=.077,p<.00) and loneliness (Phi=.054,p<.01).

Participation in rituals reinforces negative emotional reactions and was unrelated to emotional recovery from trauma.

Effects of collective remembering by means of funerals, burials and ceremonies on coping

In order to contrast the validity of Durkheim and Radcliffe-Brown’s assumptions regarding the positive social effects of conducting rituals, a series of correlations between presence/absence of funeral rituals and coping reactions (total z scores and specific reactions) were performed. Participating in rituals was associated with the total z score of direct and social cohesion type coping (r=.22, p<.001).

Participation in rites was also correlated with specific coping responses in the moment the event took place. Participation in funeral rites was associated to more coping: direct cognitive coping (Phi=.18,p<.00), altruistic behavior (Phi=.17,p<.00) and reconstruction of social support (Phi=.099, p<.00). It was also related to religious coping (phi=.098,p<.00), social sharing inhibition (Phi=.09,p<.00), social sharing (Phi=.076,p<.00), positive dreams (Phi=.071, p<.001), and to reappraisal (Phi=.062, p<.004)

Performing funeral rituals was associated in the past with enhanced sadness (Phi=.18, p<.00), fear (Phi=.12, p<.00), sense of injustice (Phi=.11, p<.00), helplessness (Phi=.088, p<.00), loneliness (Phi=.059, p<.00), anger (Phi=.049, p<.02), and intense grief (Phi=.033, p<.08).

Participation in, or conducting, funeral rites was unrelated in the present to fear, sense of injustice, anger and intense grief. Taking part, or performing, funeral rituals was associated in the present to enhanced sadness (Phi=.077, p<.00), loneliness (Phi=.054, p<.01), lower levels of helplessness (Phi=-.059, p<.00), and lower levels of political disengagement (Phi=-.04, p<.05).

Globally, participation in rituals reinforces adaptive coping and decreases responses related to lower collective self-efficiency, confirming the positive social function of rituals.

Effects of collective remembering by means of funerals, burials and ceremonies on collective memory claims

In order to contrast Halbwachs’ suggestions on the effects which rituals have on collective memory, various correlations were conducted between the presence/absence of funeral rituals and collective memory claims (total z scores). Participating in funeral rituals was associated with a stronger demand for collective memory, more reparation (moral, material) (r=.07, p<.01), and with direct and social cohesion type coping (r=.22, p<.001). It also produced more demands for knowing the truth of what had happened, and reappropriation of the past (r=.09, p<.001).

Cultural differences: interaction between mother tongue (indigenous or spanish) interviews and funerals

In order to contrast cultural differences with regard to funerary rituals, a series of Anovas were performed using mother tongue and participation in burials as independent variables and factor scores of emotional reactions in the present, and coping and collective memory claims as dependent variables. Marginally significant interaction effects were found for anger related reactions and commemoration and reparation factor scores [F (1,1636=3.11, p<.08 and F (1,1636=3.09, p<.08 respectively]. There were no differences in current fear/sadness or intense grief.

Inspection of means (see Figures 1 and 2) shows that participation in funeral rituals decreases anger/sense of injustice in mayan subjects, quite the opposite of what happens for spanish speaking respondents. Participation in burials implies reinforcing more strongly the demand for commemoration in spanish speaking persons and to lesser degree in mayan subjects. Having performed or taken part in funeral rituals or burials clearly decreases anger/sense of injustice in mayan speaking subjects.

Antecedents and emotional effects of social sharing and inhibition of social sharing

In order to contrast the effects of social sharing and inhibition of social sharing as a form of coping with traumatic events, correlations were conducted between the use of silence and talking (No=1, Yes=2) and emotional reactions (total factor scores and specific reactions in the past and in the present). Social sharing (talking and seeking emotional support) and inhibition of social sharing (silence and hiding emotions) were associated to being a witness of collective massacres (Phi=.07, p<.001, and Phi=.14, p<.001 respectively). Social sharing and silence were similarly correlated with factor scores of past sadness/fear (both r’s=.15, p<.001) and with past anger/sense of injustice (r=.21, p<.001 and r=.19, p<.001 respectively). Talking was related with present fear/sadness (r=.22, p<.001) and anger/sense of injustice (r=.23, p<.001). Silence predicted current fear/sadness (r=.15, p<.01) and anger/sense of injustice (r=.14,p<.01).

Talking and silence were related to a higher report of specific emotions in the past, similarly to participating in funeral rites. They correlated with fear (Phi=.14, p<.001, and Phi=.15, p<.001 for social sharing and silence respectively), sadness (Phi=.11, p<.001, and Phi=.10, p<.001), anger (Phi=.17, p<.001, and Phi= .12, p<.001), sense of injustice (Phi=.12, p<.001, and Phi=.10, p<.001) and with intense grief (Phi=.054, p<.001, and Phi= .035, p<.002).

Talking and silence were also related to higher report of emotions in the present: sadness (Phi=.16, p<.001 and Phi= .10,<p<.001), anger (Phi=.14, p<.001 and .08, p<.001), sense of injustice (Phi=.18, p<.001 and .12, p<.001), intense grief (Phi=.14, p<.001 and Phi=.084, p<.002) and to fear (Phi=.12, p<.001 and Phi=.15, p<.001).

Silence and talking reinforce emotional reactions in the past and in the present, suggesting that the forms of coping with the traumatic events had a more important impact in the current mood states of persons than rituals, which were associated mostly to past emotional reactions and only to current sadness and loneliness.

Associations of social sharing (talking) and inhibition of social sharing (silence) with other forms of coping with traumatic events

Talking was associated to the factor score of direct and social coping (r=.25, p<.001), and the need of knowing the truth of what had happened and reappropriation of the past (r=.19, p<.001). Talking also reinforces the need for commemorations and general reparations for what had happened (r=.08, p<.001). Silence reinforces direct/social cohesion coping (r=.17, p<.001), knowing the truth and punishing those responsible (r=.18, p<.001), and commemoration and reparation (r=.08, p<.001).

Talking and silence were related to direct and cognitive coping (Phi= .22, p<.00 and Phi=.14, p<.001), altruistic behavior (Phi= .18, p<.001 and Phi=.13, p<.001), and reconstruction of social support (Phi=.10, p<.001 and Phi=.084,p<.001). Social sharing and silence were also related to contention (Phi= .11, p<.00 and Phi= .12, p<.001) and collective flight or displacement (Phi=.09, p<.00 and Phi=.06, p<.001).

Talking and silence show a similar profile of associations with coping and collective memory claims, suggesting that both forms of coping were adaptive in the past.

Specific influences of collective massacre, funerals, emotions, talking and silence on social cohesion coping

A multiple regression attempted to separate the effects of type of repression (collective massacre yes/no), participation in rituals (burial, ceremonies), social sharing and silence, and emotional dimensions (anger/sense of injustice and sadness/fear/grief) in predicting confrontative and instrumental coping. This dimension included altruistic behavior, cognitive coping, reconstruction of social support, searching for information and reappraisal and is a good index of social cohesion. One multiple regression was performed. The fear/sadness emotional reaction was strongly related to direct and social cohesive coping (beta=.27, p<.001). Anger/sense of injustice also shows a significant association with social cohesion (beta=.26, p<.001). Participation in funerary rituals also shows a specific influence (beta=.10, p<.001). Finally, being victim of collective massacre, social sharing and silence predicts social cohesion (beta=.08, p<.001; beta=.13, p<.001 and beta=.07, p<.001, respectively).

Cultural differences: interaction between mother tongue (indigenous or spanish) interviews and silence and talking

Anovas conducted using mother tongue and social sharing as independent variables and emotional reactions and collective memory claims as dependent variables show some cultural differences (see figure 3 and 4). Talking strongly reinforces in mayan subjects the emotional reaction of anger/sense of injustice [F (1,3121)=5.94, p<.02) reinforcing in mayan speaking populations the non dominant response - it is important to notice that mayas report lower levels of anger/sense of injustice. Talking also reinforces in the mayan group the need for commemorations and general reparations for what had happened [F (1,3121)= 3.29, p<.07). Spanish speaking subjects demand commemoration and reparation more than the mayan population.

Silence increases intense grief more strongly for mayan native speakers [F (1,1911)=3.0 p<.08). Results suggest that silence has a more negative effect for mayas and that talking reinforces the set of anger/sense of injustice and commemoration/reparation claims - this emotional reaction and collective memory demands appear as associated (see below). Effects of interpersonal communication appear as more important in subjects who share strongly the more collectivistic and high power distance mayan culture.

Collective massacre, social sharing, silence, emotions and funeral rites as antecedents of collective remembering claims

Being a victim of collective massacre was related to higher claims for collective remembering: it was related to reappropriation of the past (r=.14, p<.001), to commemoration and reparation (r=.10, p<.001) and to claims for sociopolitical change (r=.10, p<.001). Participation in funeral rites was related, but with less intensity, to collective remembering claims such as reappropriation of the past (r=.09, p<.001) and commemoration and reparation (r=.08, p<.001), but was unrelated to sociopolitical change.

Coping with traumatic events by means of social sharing and silence was similarly related to reappropriation of the past and the responsabilities’ claims (r=.19, p<.001 and r=.18, p<.001 respectively), to commemoration and reparation (r=.084, p<.001 and r= .08, p<.001) and to claims for sociopolitical change (r=.09, p<.001 and r= .07, p<.001).

Sadness/Fear shows the strongest association with reappropriation of the past and the responsabilities’ claim dimension (r=.42, p<.004). Anger and sense of injustice also show a significant but lower association with reappropriation of the past (r=.34, p<.001). Anger/sense of injustice and sadness/fear were similarly related to sociopolitical change (r= .19, p<.001 and r=.20, p<.001 respectively). Anger/sense of injustice shows a stronger association with commemoration and reparation (r=.23, p<.00). Sadness/fear shows a significant but weaker association with commemoration and reparation (r= .15, p<.00).

A multiple regression attempted to separate the effects of type of repression (collective massacre yes/no), participation in rituals (burial, ceremonies), social sharing and silence, and emotional dimensions (anger/sense of injustice and sadness/fear/grief) in predicting collective remembering claims. Three multiple regressions were performed. The fear/sadness emotional reaction was strongly related to reappropriation of the past (beta= .33, p<.001). Anger/sense of injustice also shows a significant association with reappropriation of the past (beta= .16, p<.001). Finally, being victim of collective massacre, social sharing and silence also predicts reappropriation of the past (beta= .08, p<.001, beta= .13, p<.001 and beta= .07, p<.001, respectively). Regression coefficients confirm that being a victim of collective massacre and the emotional reaction of anger and sense of injustice were specifically related to commemoration and reparation (beta= .06, p<.005 and beta=.24, p<.001 respectively).

The fear/sadness emotional reaction and anger/sense of injustice were related to claims for sociopolitical changes (beta= .11, p<.001 and beta=.17, p<.001 respectively). Being victim of collective massacre and social sharing also predicts demands for sociopolitical change (beta=.05, p<.03; beta=.05, p<.03 respectively).

To conclude, being a victim of a collective massacre, social sharing and silence, the fear, sadness and the grief dimension of emotional reaction was specifically related to claims for past reappropriation. Anger/sense of injustice and being a victim of collective massacre were related specifically to commemoration, restoration of the deceased people’s dignity and the victim’s reparation claims.

General discussion

Cultural differences in emotional reaction and coping

In different studies, collectivistic subjects have mentioned coping responses less than members of individualistic cultures. This has been interpreted as showing that the intensity of emotions is lower and that its social character is higher amongst collectivistic subjects, which is why they regulate their expression less. Following the assumptions of some anthropologists’, in the maya’s collectivistic culture self-disclosure is less important than in more individualistic cultures and open expression of emotions are not reinforced.

Anger related reactions (sense of injustice) were lower in mayan interviews, confirming that collectivistic maya culture regulates anger more. As we have seen, anger related reactions and coping reactions were lower in mayan speaking subjects, showing that collectivistic cultures regulate more anger, and that emotional intensity is also lower and emotions are more «socialized». However, mayan subjects did not report lower general levels of subjective and emotional reactions, disconfirming the idea that it is a less expressive and emotional culture.

Material and symbolic losses were more important among the mayan community and were a predictor of intense grief. This may be because these losses were symbolically important for the mayas due to their respect for the land, animals and ancestors’ burial grounds and tradition, or due to the fact that indeed these losses were economically important for the individuals, the families and the communities, causing famine, starvation and making survival even more difficult.

Results show that there was more declared coping in those subjects whose mother tongue was spanish (e.g. silence, talk, direct coping). These subjects also reported more collective flight, transformation of community organization and self-control and contention.

It is also quite noteworthy to state that the mayan population present less community crisis than the ladino or mestizo population. Collectivistic cultures seem to be more resilient, although in the ladino community crises tend to imply more coping.

It is interesting to note that mestizo and not mayan subjects stress more collective flight, while at the same time reassuring having had more support. This result is in line with the fact that the Popular Resistance Communities which mainly included ladino subjects (who participate more in cooperatives) had to flee from their own land and migrate, but they tended to do so together, and so suffered more community crisis, although at the same time they tended towards more coping, specially of a self-contention type.

Rituals show stronger effects in the case of mayan subjects. Funeral rituals decrease anger/sense of injustice and do not reinforce so strongly the quest and demand for commemorations among the mayas. Both results suggest that rituals have a more important buffer effect for the mayas. A tentative explanation could be due to the very nature of social organization in the maya community which is more cohesive and considers that burials or funeral rituals put an end to the situation once the ancestor has been honoured by giving them a proper and adequate burial. Cultural beliefs and the fact of performing rituals could interact and support the presence of the deceased in the community. Loss is buffered by the idea that funeral practices help the ancestor to be in the «correct» place, and alive in the community. In other terms, among mayas funeral rituals are more effective due to their beliefs in the presence of ancestors, coessences, etc. which guard them and their community against evil, and because the mere meaning of rituals may be more important among mayas than among ladinos. In a similar vein, Matsumoto (1994) recalls how family losses are less frequently cause of sadness among asian subjects who believe in the presence of ancestors in comparison to western subjects who do not believe in this «real» presence of the deceased.

The maya´s culture is also related to a highly contextual communicational style: understatements, indirect statements and silence are used as a way to communicate a negative and to avoid directly disagreeing with others. Disconfirming some anthropologists’ claims, silence and emotional inhibition were not more typical of collectivistic mayas - at least using mother tongue as a cultural marker. Some evidence was found for cultural differences in the effects of talking and silence. However globally associations between talking, silence, and other forms of coping and emotional reactions were similar in the ladino and the maya ethnic group. Interpersonal communication reinforces anger responses and demands for commemorations, factors which are mutually linked. Communication reinforces the less frequent responses among mayas, suggesting that the function of reinforcing emotional intensity and collective memory claims is more important in cultures where verbalization is less frequent. However, silence reinforces stronger intense grief among the mayas. In other words, inhibition is a stronger risk factor among more collectivistic cultures. This means that a less frequent form of coping as silence produces more negative effects in a collectivistic culture, independently of if it is related to verbalization or inhibition of communication. In general we could state that communication is more important in collectivistic cultures than in relatively more individualistic cultures: social sharing reinforces active emotional reactions (anger and commemoration claims) and inhibition of social sharing reinforces emotional reactions related to loss (intense grief).

Antecedents and effects of collective remembering: Funerary rituals and emotional reactions

Being a victim of collective massacre was related to higher emotional reactions, to more rituals, to higher claims for collective remembering (reappropriation of the past, commemoration and reparation) and to claims for sociopolitical change, confirming that wars and genocides reinforce social mobilization and the need for rituals.

Results show that the absence of mourning rituals was not related to intense and prolonged grief in 81-85 nor in 95-96. Mourning rituals are related to higher frequency of sadness, fear, sense of injustice, uncertainty, anger and intense grief in the past - and only to current sadness. Results suggest that funerary rites enhance current sadness, and do not protect against negative emotions and grief. These results disconfirm Freud’s assumption that rituals support the work of grief and help achieve emotional recovery.

It is important to be aware that mourning rituals were limited and usually performed in negative conditions and there is an absence of collective commemorations before 93-95. The fact that funeral rites sometimes were performed in extreme conditions can explain why they do not have a positive emotional effect.

However, other data collected with a western sample who perform normal ceremonies shows a similar profile. Bowlby’s review states that rites and social support are a protection against social isolation, but not against emotional isolation related to the loss of an attachment object (Bowlby, 1980). Weiss & Richards’ (1997) results are similar to ours, in the sense that satisfactory participation in funerals were related to better social functioning and not to a decrease in emotional distress. A bereavement scale was related to a more social oriented scale and unrelated to psychological scales measuring depression and rumination. Specifically, the category «Acceptance of ceremonies of leave-taking as satisfactory» predicts a Positive State of Mind 12 months after the bereavement. The aim of the PSOM scale was to capture quality of social functioning (capacity to meet expectations in ordinary roles, maintaining social life and so on). Their description of a satisfactory leave taking ceremony shows how rituals are not only a social step in the closure of the relationship with the deceased but are also related to intense emotions:« We had a memorial service on Sunday down there. I did a lot of crying down there which was good. It was just a very emotional, a very nice ceremony that kind of helped put another conclusion - close one more door» (Weiss and Richards, 1997, p. 890).

A well-known anthropological study on rituals also concludes that rituals not necessarily help in grief: «the psychic process of grieving only partially intersects with the performance of death rites... rituals sometimes aid the process, but it could as easily be no help at all, or even an extra burden to bear» (Metcalf & Huntington, 1999, pp.4-5). Finally, in a review of funeral rites, Pradelles (1996) suggests that mourning rituals in Africa are not relevant for the individual process of bereavement but for social cohesion and reorganization. This conclusion is similar to previous available data and to Hertz’s hypothesis about mourning rituals: the rite of passage symbolizes changes in the corpse, the deceased person and changes in the social order. The fate of the corpse in mourning rituals reflects the reordering of social relations (Hertz, 1981).

Effects of collective remembering: Funerary rituals, emotional reactions and social cohesion

Results confirm Durkheim and Radcliffe-Brown’s position: rites and social sharing about past traumatic events reinforce emotional arousal, and elicit social cohesion and solidarity. Rites were associated in the past with higher coping frequency as a reconstruction of social support, altruistic behavior, and to a lower sense of helplessness and disengagement, confirming Durkheim’s contention. Direct cognitive coping can be seen as an antecedent of funeral rites, but this is not the case of the other types of coping which can be interpreted as related to social cohesion.

Collective commemorations and mourning rituals have a social function: activities of collective remembering (mourning rituals, memorials and commemorations) insert individuals in the whole society, reaffirming the continuity of society (Averill,1979).

According to Radcliffe-Brown, funeral rites are one of the elements which help maintain social order and group cohesion - behavior of the bereaved serves to reaffirm the values of society and promote group solidarity. Reactions of grief, sadness and anger were irrelevant for Radcliffe-Brown and the purpose of the ritual was not to amplify or reinforce negative emotions related to loss, but to strengthen the social bond, togetherness and produce positive feelings (Howard, 1989; Metcalf & Huntington, 1999). For Durkheim, collective commemorations and mourning rituals reinforce emotional reactions of grief, sadness and anger. Rituals put pressure on people to put their emotional behavior and feelings in conjunction with bereaved relatives. By means of ritual practices intense emotional reaction is induced in an structured manner. To share the same emotional climate is associated to a positive statement of their commitment, not only to the suffering group members, but also to social values. For Durkheim, the amplification of negative emotions by means of ritual is an important step in order to strengthen social cohesion and mobilization (Durkheim 1912/1982; Kemper, 1993). Our results confirm the Durkheimian position against Radcliffe-Brown’s: funerary rituals reinforce social cohesion, but, negative emotional reactions also has the same effect, even when controlling for participation in rituals - by means of multivariate analysis. Moreover, with respect to collective memory claims, rituals do not show a direct influence and they reinforce collective memory indirectly by means of intensified emotional reactions of anger and sadness.

Effects of collective remembering: Funerary rites, social sharing and collective memory claims

Results also support Halbwachs’ ideas: rites and social sharing about past traumatic events reinforce the construction of collective memories, because they are related to higher levels of collective memory claims. Commemoration and rituals are a form of collective remembering which supports a moral self-definition and helps to consolidate memories of important events. Remembering is a normative process that allows people to have a personal and social identity (Halbwachs, 1968; Pennebaker, Paez & Rimé, 1997). By means of memorials, commemorations and rituals, reviving bonds with the deceased confirms a person’s social identity and is a step towards the reappropriation of the past - an effort to extract a lesson from the past which supports a moral self-definition (Halbwachs, 1950; Frijda, 1997; Pennebaker, Paez & Rimé, 1997).

As we can see, participating in funeral rites and social sharing not only helps to enhance social integration and restore one’s self-concept and self-esteem, but also to foster collective memory (Rimé et al., 1998). This could be because participating in rituals reinforces emotions such as sense of injustice while also reinforcing sociopolitical demands and collective remembering.

Effects of communicative forms of coping: Social sharing or talking, and inhibition or silence

Social sharing was also related to past and current emotional reactions - people who mention using silence or social sharing report a higher frequency of fear and sadness, grief and anger. Rimé et al. (1998) also have concluded that social sharing does not help to emotional recovery.

Silence shows a similar profile and both social sharing and silence were related to intense grief. Moreover silence and talk clustered together and were related to other adaptive forms of coping. A factor analysis showed that social sharing and silence load together in the same factor, suggesting that they are not contradictory forms of coping. Moreover, both social sharing and silence were related to adaptive forms of coping, like self-control and direct cognitive coping. Social sharing and inhibition show a similar association with other forms of coping: both were associated with instrumental coping. Other available data suggests that subjects directly affected by traumatic events at the same time show more social sharing and more avoidance (Páez, Basabe & González, 1997). That is, subjects whose relatives have experienced traumatic events talk and inhibit more, suggesting that traumatic events in general mobilize more coping. These results suggest that silence in the midst of repression was not inadaptive and that silence is a coping response associated with social sharing and other forms of coping. This of course does not mean that permanent silence is adaptive or does not have a psychological cost: previous research shows that the less subjects talk about past traumatic events that actually affect members of their groups, the more they have a positivistic view of current society (Páez, Basabe & González, 1997). Inhibition and silence may not only have negative effects on well-being and health, but also can play an important ideological role of justification. Moreover, associations of sharing and inhibition with current emotions were stronger than in the case of rituals. They were stronger in the present than in the past and in the case of talking than in the case of silence. This suggests that coping by means of emotional support causes more long lasting emotional reactions - funerary rites were associated only with current sadness - and that social sharing provokes a stronger impact than inhibition.

Commemorations and emotions as antecedents of collective memory claims

Reactions of fear, sadness and grief were strongly associated to reappropriation of the past and the responsabilities’ claim dimension of collective memory. Anger and sense of injustice were strongly associated with commemoration, moral reparation of the deceased and material compensation for victims, supporting some specific social functions of emotions. Reappropriation of the past and the responsability dimension of collective memory appear associated to social functions of loss acceptance, plans for a better performance in the future and with focusing attention on potential threats demanding self-protective measures - functions associated to sadness and fear. The commemoration dimension of collective memory appears as associated to a social function of mobilizing energy which sustains a more proactive goal-directed activity: restoring the deceased’s dignity and victim’s reparation claims. Sense of injustice and anger mobilize energy to demand for moral and economic reparation.

Finally, funeral rites do not show a direct influence in claims for reappropriation of the past, commemorations, moral and material reparation of deceased and non deceased victims of violence. Funeral rites reinforce shared emotions, and by this mechanism, indirectly propulse collective memory claims. However, both social sharing and silence, as complementary forms of coping with traumatic events, reinforce claims for collective memory, supporting the idea that interpersonal coping helps to construct collective memories.

Some limitations of our results

Nevertheless, our data has some important limitations. We worked with retrospective open accounts submitted to reconstructive memory processes. However, comparisons of data obtained in the REMHI project and socio-demographical data show an important convergence (ODHAG,1998).

Effect size were lower and usually they were positive in dichotomic cases. Lower effect size was expected because of the data characteristics: dichotomic variables, with «high noise». However, analyses using only good quality interviews produced results similar as those obtained with the whole sample. Empirically the effect size (with non signification) decreased with the increase of the sample size: both error and the effect of unmeasured variables increases continuously with a larger sample size (Scherer, 1997). At the same time, and casting doubts on the idea that we are in presence of tendency to give more complex answers in some categories, people who participate in funerary rites report lower compromise.

Our data is based on open responses to a structured interview. The interview was focused on the reconstruction of repression and it tapped accessible knowledge and reconstruction of events which had taken place during the last 3 to 14 years (e.g. only 4% mention rumination or nightmares in 81-85, usually above 70% of traumatized subjects report rumination). Reconstructive effects are probably important.

Our code categories were not psychopathological: fear, sadness and anger were conceived of as normal emotions or normal reactions to extreme situations. For instance, prolonged and intense grief does not imply a psychopathological criteria for chronic or complicated grief (Prigerson et al., 1995). However the criteria was similar to the main feature of complicated grief: the person being interviewed could not speak of the deceased without experiencing fresh and intense grief (Worden, 1991, p.75).

However, the code categories were culturally sensitive and adequate to the content of interviews. Categories were constructed by three members of the REHMI professional team, and were based on previous work (20 focus groups with refugees on emotion prototypes and coping), inductive construction of categories using the first 50 interviews and a review of literature on PTSD, coping and collective behaviour.

Respondents were volunteers and the work was sponsored by the Catholic Church. This could influence the type of responses (more motivated people and «politically correct» answers). Nevertheless, thousands of people were interviewed and some of them were reluctant to give information during the first contact.

Some implications of our results

1. Contrary to a stereotypical view, those with a stronger implication in mayan culture did not show less general emotional activation. They did show a lower reaction of anger and sense of injustice and less coping reactions.

2. Material losses had a stronger impact in those people with a relatively more maya culture and they also stated more fear reactions. This means that possible steps directed towards material reparation are even more important for this group

3. Rituals seemed to have clearer, and in some sense positive, effects for those with a stronger mayan culture. This suggests that commemorations are more important and may have more positive effects in this population.

4. Moreover, open communication and accounts concerning what had happened, and avoiding silencing attitudes may also produce a more positive increase in emotional climate in the case of the mayas.

5. Our results also suggest that undergoing commemorations has a paradoxical effect: to remember is to relive and reinforces those emotions associated to the loss, although at the same time it also reinforces social integration, probably due to the positive affect linked to social support, and produces a positive moral image of oneself. Other data suggests that well performed commemorations, which value the disappeared, and allow their relatives to play a positive role have positive effects on mental and physical well-being.

It is important for commemorations not to be repetitive and focused only on suffering. They must be performed in an «aesthetic» way, regaining a positive image of the deceased, and helping to obtain a positive moral lesson for the future.

6. Silence and the search for emotional support during repression were adaptive coping forms in the past. Nevertheless, prolonged inhibition of a trauma may very well have negative effects. These effects will not only be psychological, but in fact in the case of silence may be associated with a «non realistic» and positivistic view of the present. Verbal repetition about what has happened not only helps to understand it better or to extract lessons for the future, but may also be a form of rumination and of keeping alive wounds inflicted in the past.

7. Two types of different tasks seem to emerge from our results. First of all the task of knowing the past, identifying and punishing those responsible, and also erecting barriers so that repression may not take place again. This is of utmost importance in order to make the climate of fear, anxiety and sadness disappear. In second place, recognizing the moral value of those killed, and morally and materially reparing the survivors is an essential social task in order to dissipate the emotional climate of indignation caused by repression. In other words the anger and sense of injustice experienced due to what had happened.

Averill, J.R. (1979). The functions of grief. In C.Izard (Ed.). Emotions in personality and psychopathology. New York: Plenum.

Bowlby, J.(1980). Attachment and loss. Vol.3. Loss, Sadness and Depression. London: The Hogarth Press.

Comas-Diaz, L., Lykes,M.B. & Alarcon, R.D. (1998). Ethnic conflict and the psychology of liberation in Guatemala, Peru and Puerto Rico. American Psychologist, 53, 778-792.

Díaz-Loving, R. (1998). Contributions of mexican ethnopsychology to the resolution of the etic-emic dilemma in personality. Journal of Cross-Cultural Psychology, 29, 104-118.

Durkheim, E.(1912/1982). The Elementary Forms of Religious Life. Madrid: Akal.

Freud, S. (1917). Mourning and Melancholia. Complete Psychological Works of Sigmund Freud,14. London: Hogarth Press.

Frijda, N. (1997). Commemorating. In J.W.Pennebaker, D. Paez & B. Rimé (Eds.). Collective Memory of Political Events. Mahwah, NJ: Lawrence Erlbaum.

Gossen, H. (1994). From Olmecs to zapatistas. American Anthropologist, 96, 553-570.

Halbwachs,M. (1950/1968). La mémoire collective [Collective Memory]. Paris: Presses Universitaires de France.

Hertz, R. (1981). Death and the Right Hand. Madrid: Alianza.

Hodgkinson, P.E. & Stewart, M. (1991). Coping with catastrophe. A handbook of disaster management. London: Routledge.

Hough, R.L., Canino, G.J., Abueg, F.R. & Gusman, F.D. (1998). PTSD and related stress disorders among hispanic. In A.J. Marsella, M.J. Friedman, E.T.Gerrity & R.M.Scurfield (Eds.). Ethnocultural aspects of Post-Traumatic Stress Disorder. Washington: APA.

Howard,M.C. (1989). Contemporary cultural anthropology. Glenview, Illinois: Scott, Foresman and Co.

Kemper, D. (1993). Sociological Models in the explanation of emotion. In M. Lewis & J.M. Havilland (Eds.). Handbook of Emotions. New York: The Guilford Press.

León-Portilla, M. (1986). Tiempo y realidad en el pensamiento maya. Mexico: Unam. [Time and Reality in Mayan Thought].

Levav, I.; Friedlander, Y. et al (1988). An epidemiologic study of mortality among bereaved parents. New England Journal of Medicine, 319, 457-461.

Lykes, M.B. (1994). Terror, silencing and children: international, multidisciplinary collaboration with guatemalan maya communities. Social Science and Medecine, 38, 543-542.

Martin-Baró, I. (1990) Psicología Social de la Guerra. El Salvador: UCA Editores.

Matsumoto, D. (1994). People. Psychology from a cultural perspective. Pacific Grove, CA: Brooks/Cole.

Metcalf,P. & Huntington, R. (1999). Celebrations of Deaths. The anthropology of mortuary rituals. Cambridge: Cambridge Univesrity Press.

ODHAG, Oficina de Derechos Humanos del Arzobispado de Guatemala: Informe Proyecto InterDiocesano de Recuperación de la Memoria Histórica (1998). Guatemala: Nunca Más. Vol. I, II y III. Impactos de la Violencia. Tibás, Costa Rica: LIL/Arzobispado de Guatemala. [Guatemalan Catholic Church Human Right’s Department. Historical Memory Recovery Project Report. Guatemala: Never Again].

Páez,D., Basabe, N. & González, J.L. (1997). Social processes and collective memory. In J.W.Pennebaker, D. Paez & B. Rimé (Eds.). Collective Memory of Political Events. Mahwah, NJ: Lawrence Erlbaum.

Pennebaker, J.W. (1990). Opening up. New York: W.Morrow & Co.

Pennebaker, J.W., Mayne,T.J. & Francis, M.E. (1997). Linguistic predictors of adaptive bereavement. Journal of Personality and Social Psychology, 72, 863-873.

Pennebaker, J.W., Paez, D. & Rimé, B. (1997). Collective Memory of Political Events. Mahwah,NJ: Lawrence Erlbaum.

Pradelles, C. (1996). Les morts et leurs rites en Afrique. L’Homme, 138, 137-142. [Death and their Rituals in Africa].

Prigerson, H.G. et al. (1995). Complicated Grief and Bereavement-related Depression as distinct disorder. American Journal of Psychiatry, 152, 22-30.

Rimé, B. et al. (1998). Social Sharing of Emotion: New evidence and new questions. In W. Stroebe & M. Hewstone (Eds.). European Review of Social Psychology, 8.

Scherer, K. (1997). The Role of Culture in Emotion-Antecedent Appraisal. Journal of Personality and Social Psychology, 73, 902-922.

Wearne, P. (1994). The Maya of Guatemala. London: MRG Editions.

Weiss, R. S. y Richards, T. A. (1997). A scale for predicting quality of recovery following the death of a partner. Journal of Personality and Social Psychology, 72, 885-891.

Wilson, R. (1995). Maya resurgence in Guatemala. Norman: University of Oklahoma Press.

Woodrick, A.C. (1995). A lifetime of mourning: grief work among Yucatec Maya Women. Ethos, 4, 401-423.

Worden, J.W.(1991). Grief Counselling and Grief Therapy. London: Tavistock/Routledge.

Zur J. (1996). From PTSD to voices in context: from an «experience -far» to an «experience-near» understanding of responses to war and atrocity across cultures. International Journal of Social Psychiatry, 42, 305-317.

Aceptado el 20 de diciembre de 1999