Violence in school: a few orientations for a worldwide scientific debate
by Eric Eric DEBARBIEUX, Université Victor Segalen Bordeaux2, Director of the International Observatory of Violence in Schools

Theme : International Journal on Violence and School, n°1, May 2006

The very existence of a third World Conference on Violence in Schools, held in Bordeaux in January 2006, and the probable extension to two other conferences in the next four years, is a sign of the mobilisation of the scientific community on this research topic. The creation of various observatories both within and outside the network of the International Observatory of Violence in Schools is also an important sign. Crucially, international research, which was for a long time confined to the subject of School Bullying, is now broadening its scope and is no longer conducted only in the developed countries. The number of research programmes carried out in Latin America, for example, is impressive, while the African continent is showing increasing interest in the subject, as are Japan and China, among others. This article will first present some of the issues covered by this international research, demonstrating how its legitimacy is to be established in worldwide terms against a backdrop of continuing doubt about the research object itself and its possible findings. Next, we will present results of international comparative research in a North-South space covering Europe, Africa and Latin America. The aim is to show how knowledge can be built against this doubt, but also how other questions must be raised about the definition and interpretation of a phenomenon which, in the literature, is marked by a conceptual overrepresentation of the Northern countries and which thus lacks certain realities. Lastly, the areas for new action in this emerging dialogue are highlighted.

Keywords : .

English text (download the pdf file here)

1. A common base: manipulation and "omission” of the victims

A sensational story

The sensitivity of public opinion to violence among young people in school is often exacerbated by tragic news items which dramatise "stories” (Brownstein, 2000) of a violence which is spectacular and bloody, but rare in this extreme form. We are thinking, of course, of the school massacres in the United States, which have made the front pages of newspapers the world over. The outcry was the same in Argentina after the Islas Malvinas shooting by a 15-year-old pupil which cost four lives. Another country, Japan, must also be mentioned, because this case shows how the emotional charge resulting from deaths at school can have genuine effects on the evolution of the legislation.

Japan experienced two killings in schools which horrified public opinion. The first, in 1997, was shocking: a 14-year-old boy decapitated an 11-year-old pupil and left his head at the entrance to the school. The second, in 2004, was also horrendous. A 12-year-old schoolgirl, Satomi Mitarai, was stabbed to death with a cutter by one of her classmates, an 11-year-old girl. The number of murders in Japan has constantly fallen since the Second World War and there is no evidence of a rise in killings committed by minors: they remain extremely rare. Japan is still one of the safest countries in the world in this respect, with a very low rate of violent juvenile crime. The rarity value of these killings has stirred public opinion even more, creating a real collective hysteria. This hysteria is fed by a certain type of press, of course, but has also affected the politicians, who have gone overboard. After the 1997 case, parliament lowered the age of criminal responsibility from 16 to 14. Talking about a different case, a Japanese minister, Yoshitada Konoike, publicly declared that the parents of the young killer should be "dragged through the town and beheaded”. An article on 9 August 2004 by the Japanese correspondents of the Washington Post gives an idea of the media treatment of the killing of Satomi Mitarai. The article starts like this:

"Sasebo, Japan – On a cloudless afternoon in this peaceful port town, an 11-year-old girl covered with blood and holding a cutter in her hand came into the canteen of her primary school. The teachers and pupils shuddered, believing that this pupil, known for her light-hearted nature, had seriously injured herself – but she quickly dissipated this impression when she pronounced these terrible words: ‘It's not my blood'”.

For the journalists, this murder was the latest in "an extraordinary series of juvenile crimes in Japan”, directly related to the "violent, sexually charged youth culture in Japan, now exported all over the world through cartoons, comic strips and video games”. The authors did not hesitate to speak of a "juvenile crime wave” cutting through the sense of security of the Japanese, and repeated the allegations of another newspaper, Mainichi, which described classrooms transformed into "battlefields”.

In the public mind and in many reports, the overriding impression is often one of an epidemic of brutal rage, a growing sense of danger in schools. What really attracts the attention of the public are the school massacres, the highly televised media reports of which are such an important element that the guidelines to head teachers in the USA now include "how to face the press” in the event of a major incident, given that the flood of TV crews can be just as traumatic for the community as the massacre itself (Cornell, 2003). What is represented is an eruption of savagery among children and teenagers (Debarbieux, 1997). Are school massacres so common that they should be seen as an epidemic?

What people most remember about school massacres are the ones committed by youngsters. However, it should be pointed out that these killings are far more often carried out by adults. The mass murders perpetrated in schools since the early 1960s are easy to find. There have been fifteen of them in all, if we define "mass murder” as a killing with more than three victims (if we set the number at three, we must add three more episodes including Moses Lake in 1996, and West Paducah in 1997).

Other incidents could be added, such as the latest tragic event in Russia, in 2004, the number of victims of which is still unknown – several hundred at least, including 250 children killed by terrorists during the controversial assault by the police. The table below summarises the key aspects of the massacres committed in schools.








42 years

Cologne, primary school, FRG



FLP terrorists

Maalot, Israel



Members of the Ba'ath party

Allepo Artillery School, Syria



Adult – Anti-feminist

Ecole Polytechnique, Montréal



Adult-25 years

Stockton, California, USA



20 years – former pupil

Oliverhust- California, USA



43 years

Dunblane- Scotland




Sanaa- Yemen



13 and 11 years

Jonesboro Arkansas USA



18 and 17 years

Columbine Colorado, USA



Adult - 37 years

Osaka- primary school- Japan



19 years – expelled pupil

Erfurt, Germany



Pupil 15 years

Islas Malvinas Argentina



Adult, intruder

Ruzhou High School, China



16 years

Red Lake, USA

Total killed




Total killed by adults




Table 1: School massacres – 1964-2005

There have been fifteen massacres causing 177 deaths in schools since 1964, including five in the USA. Out of those 177 victims, 128, or more than 70%, have been killed by adults, with various motivations: terrorism, sexist ideology (in Montreal), madness. The first case, in 1964, was perpetrated by a Second World War veteran who entered a primary school in Germany armed with a flame-thrower, killing eight children and two teachers and seriously burning many others. Massacres in schools are far more often committed by adults than children, and this distinction would be even clearer if we added the one in Russia. In 2004, in Beslan, the biggest killing ever recorded in a school took place. On 1st September that year, a group of thirty-eight terrorists broke into a school, taking almost 1,200 hostages in a siege that lasted forty-eight hours and ended with the death of at least 335 of them, mainly children. The hostages were killed on the morning of 3 September when the Russian troops were ordered to attack, and it remains impossible to know how and by whom these hostages were killed. At all events, if this case were added to the total, more than 90% of massacre victims in schools would have been killed by adults. Let us return briefly, for example, to the main school massacre committed in Japan: the one perpetrated in Osaka in 2001 by a 37-year-old, Mamoru Takuma, whose only regret after being condemned to the death penalty after killing eight children was, ‘I should have used petrol, that way I would have killed more'. While none of this diminishes the tragic nature of the murders perpetrated by young people, it does serve to show them for what they are: exceedingly rare.


Two ways of forgetting the victims

This focus on news stories and the more dramatic events has brought with it with a trivialisation of these acts, considered merely as representative of a "rise in violence”. The journalist Katherine Ramsland wrote an article entitled "The List”, an end-to-end record of those killed in schools in the United States: to find it, visit the website, select the serial killers section and then look for school killers. The List starts with these words: "In the late 1990s, it seemed like an epidemic had hit American schools: Children were acquiring guns and bombs, and then going to school to kill teachers and classmates”. Hasty generalisations about this rise in barbaric behaviour are legion (see, for example, in Canada, Gauvin 1995, in France, Debarbieux 1997, and an international summary in Debarbieux 2006). They are nonetheless disproved by the facts.

Let us return to the emblematic example of the United States. According to the National School Crime and Safety Survey (NCES, 2004, 1 p.2) using data from the Center for Disease Control and Prevention, since 1992 young people aged between 5 and 19 have run around 70 times less risk of being killed at school (or on the way to school) than they have in other circumstances. Between 1992 and 2000 in all American schools, there were 234 homicides and 43 suicides of youngsters aged from 5 to 19. Considering that the number of pupils in this age bracket in American schools is around 50 million, this figure amounts to a risk of less than one homicide or suicide per school year per million pupils, either at school or on the way to school. Over the same period, 24,406 children and teenagers in that age bracket were homicide victims, and 16,735 committed suicide. Lastly, although between 1992 and 1998 there was no evidence of either an increase or a decrease in these murders (between 28 and 34 homicides per year), since then there has been a sharp fall in number, from 33 homicides in 1998 to 14 in 2001-2002. The NCES report rightly concludes that there is a morbid exaggeration which consists in polarising attention on the very rare cases of extreme violence. The FBI itself (Cornell, 2003) totally refuses to conduct profiling of pupils capable of this sort of behaviour, judging that such cases are so rare that they cannot justify the stigmatisation of certain populations which might be considered at risk.

The first way of forgetting the victims is precisely this trivialisation of their experiences, the number of which remains small. Trivialising these victims by incorrectly pigeonholing them as part of a continuous series of bloody crimes at school is a way of disrespecting them. All the more so in that this trivialisation is often accompanied by a demagogic discourse on security, an example of which we have seen above in Japan, but which we have also studied in Brazil (Debarbieux, 2006 p.41 passim) with the manipulation of survey figures by a media-proclaimed "sociologist”.

The number of juvenile delinquents committing extreme violence is relatively small (Farrington, 1986) and, without underestimating it, the quantity of crimes and misdemeanours in schools remains low (Gill & Hearnshaw, 1997, Lindström, 2001, Gottfredson, 2001). After a remarkable review of the subject in the United States, Denise Gottfredson, certainly one of the world's leading specialists in the field, concludes that victimisation in the school environment has barely evolved between her first studies in 1985 (Gottfredson and Gottfredson, 1985) and her more recent ones (Gottfredson, 2001): the personal experience of victimisation, both for pupils and teachers, is related to minor incidents, while more serious victimisation is very rare. In 1985 the Gottfredsons affirmed that their surveys in the USA had shown that the real problem stemmed more from a high frequency of minor victimisations and indignities than from hardcore delinquency. We are more often in the margins of misdemeanour (what Cusson calls "peri-misdemeanours” (Cusson, 2000)), than in the realm of open delinquency.

However, placing emphasis on cases of minor victimisation, or micro-violence, as we now prefer to call it, should not lead us to underestimate the scale of the problem. This would be tantamount to placing the suffering experienced in the ignominious category of "insecurity fantasy”, which is a right-thinking negation of victims: the other way of forgetting them. Indeed, one of the aspects best identified by international research is the scale of this micro-violence, not when each episode is considered separately but when they are accumulated, the effects of which are well described. In short, what counts is not a minor victimisation but its repetition, associated with other incidents of micro-violence; this repetition can have serious consequences on the victims, or even on the social corps. If we take the case of school bullying, we will remember with Smith (Smith, 2002) that "the victims of bullying often suffer from anxiety and depression, loss of self-esteem, and physiological and psychosomatic disorders (Williams et al., 1993). In extreme cases, they may become suicidal (Kaltiala-Heino et al., 1999)”. The victimisation situation may be repeated, then, and research suggests that it concerns a small minority of pupils (cf. synthesis in Blaya 2006) who nevertheless are sometimes in great distress. This situation is therefore quite different from a single victimisation situation (terrorism, mugging, for example), both for victim and perpetrator (or perpetrators), because it requires a regular relationship between the two parties. It is an interaction which is constructed over time. Violence is not always an isolated, unpredictable, accidental incident: in part at least, it is built around the tenuous and the continuous. A serious research programme on violence in schools must therefore include not only a study of the frequency of the acts but also a measure of their repetition and their prevalence. This also implies the acceptance of a broad definition of violence, including incidents of seemingly little importance but the repetition of which should be measured (for example, Vettenburg, 1998). What is considered as violence may vary according to the country or the culture, but there could be a universal mechanism which is that of repetition, with dramatic consequences for those who suffer from homogenous or heterogeneous micro-violence.

Three prerequisites to scientific dialogue

As a result, three requisites to scientific dialogue on violence in schools could be set out. The first is an absolute ethical position involving a rejection of the sensational AND of the negation of victims. The second is a methodological trend: the development of international victimisation studies to provide an independent measurement of the phenomenon which separates exaggeration from reality but which also takes account of the repetition of violence in order to understand the victim's experience better. The third is the contextualisation of results (Gottfredson, 1986, 2001; Astor and Benbenishty, 2005; Debarbieux, 2006) in order to understand the differences – and similarities – noted in different countries, without necessarily privileging a single approach. Below we will attempt to show in concrete terms how we have implemented this international contextual approach, in a comparison between the countries of the North and those of the South.


2. A North-South comparison


In the developed countries, violence in schools mainly occurs in deprived areas, and analysis must bear in mind this sociological context. Of course we know that violence in schools does not just have one causality and that only a combination of individual and environmental risk factors can explain it. Poverty or unemployment alone cannot cause violence. However, when the factors of social exclusion are cumulated, the risk of becoming a victim or aggressor increases (for example, Farrington, 1989; Coie & Jacobs, 1993). As Farrington wrote (Farrington, 1986): "Children who live in deprived inner city areas tend to be from ethnic minorities, tend to have parents with low status, low paid jobs, or no job at all, and tend to have friends who commit deviant acts”. It therefore seems legitimate to adopt the thinking of Denise C. Gottfredson, after her own review (Gottfredson, 2001 p. 64): "Poverty and income inequality, educational attainment, mobility of the population, ethnic composition, population density, and high proportions of female-headed households are correlated with area crime rates”.

The globalisation of violence, particularly violence in schools, could therefore be a very non-egalitarian globalisation due to the "hierarchical nature of the world economy” (Adda, 2002, p.47), "globalisation being a profoundly unequalising movement” (Moreau Defarges, 1997). We therefore accept this socially unequal development of school violence as a working hypothesis. To put it to the test, we have studies that we have conducted directly or indirectly in several countries of the South, with a standardised methodology adapted locally: in Brazil, Burkina Faso, and the Republic of Djibouti.


The European Observatory of School Violence proceeded to empirical comparisons using, among other things, a questionnaire on victimisation and school climate (Debarbieux 1993), adapted by local researchers with a standard test and retest procedure.

A research tool: the questionnaire on victimisation and school climate

The purity of our measurement instrument is good, because in all its versions Cronbach's alpha is higher than 0.65 (0.70 in France, 0.76 in Djibouti, 0.69 in Brazil, 0.72 in Burkina Faso and 0.74 in England), easily reaching the international standard (more than .60). This questionnaire is composed of scaled questions aiming to test the quality of school climate, summarised by a School Climate Index, questions on the commonest victimisations, and questions on the way transgressions are dealt with (punishment).

The fieldwork was carried out by or with local researchers, who we helped train. In Brazil it was a UNESCO team that did all the fieldwork with the questionnaires, in Burkina Faso it was one of our students (Lompo, 2005), and in Djibouti it was a team of three young researchers from the Statistics Bureau and a trainer from the Teacher Training College who accompanied us and then took over in the field (Debarbieux/Unicef, 2005). These studies were also complemented by individual and group interviews and by ethnographic observations.

Samples used

In this article, we will use samples from two countries from the North, France and England, and three from the South, Brazil, Burkina Faso and Djibouti.

In France:  

    sample 1 N= 11,147 pupils from 12 to 16 years, working-class secondary schools (Priority Education Zone, in the most socially sensitive 5% of schools, urban zones of Paris and suburbs, Marseille, Roubaix)
    sample 2: 2,714 pupils from 9 to 12 years questioned in 23 primary schools – randomised sample from Aquitaine (city centre, working-class suburbs and rural)

In England:

    1,545 pupils from 16 secondary schools in three areas from among the most deprived 1% of urban and suburban zones (North London, Liverpool…)

In Brazil:

    9,800 pupils from 12 to 16 years from the big Brazilian cities, randomised sample (140 schools) – city centres, suburbs, state schools.

In Burkina Faso: 

    304 adults and 1,125 pupils from 6th through to 12th grade (12-19 years), in a total of 80 schools, or 16.13% of all secondary schools

In Djibouti:

    randomised sample of 1,699 pupils from 9 to 14 years, questioned in 19 schools in urban, outer-urban and rural areas (application of 23% of the school population concerned)





Brazil is not a poor country but it is one of the world's most unequal countries: the wealth of the richest 2% of Brazilians who own 60% of the national assets only serves to highlight the conditions of extreme exclusion in which the majority of inhabitants live. At the time of our study, Brazil was rated 73rd out of 177 countries in the Human Development Index proposed by the UN following work by the economist Amartya Sen. This medium ranking masks extreme disparities between the different urban areas and between the social strata – the relative lack of a middle class being one of the features of the socioeconomic stratification of Brazil.

Is violence in schools a problem that is comparatively greater in Brazil than in more developed countries? It would not seem to be the case, according to the perception of Brazilian pupils, compared first with French pupils.

Dependence is highly significant. chi2 = 1 386,40, ddl = 4, 1-p = >99.99%.

Table 2: Perception of violence by French and Brazilian pupils. Source: Debarbieux/Unesco 2004, and Debarbieux 2000, France N =11147, Brazil N= 9800

The perception of violence among Brazilian pupils of between 12 and 16 is clearly lower than it is for their French counterparts. 31.1% of pupils from the French reference sample against 12.4% of Brazilian pupils consider that violence is very present in their school (‘an enormous amount' and ‘a lot'). This difference is just as significant in the perception of the 1,545 English pupils questioned in working-class schools by Catherine Blaya (Blaya, 2006): 26% of English pupils think that violence is very present in their school, which is less than in France but much more than in Brazil. Reported victimisations are much lower in Brazil than in France and, in the main, in England: almost 25% of pupils both in France and in England say that they have been struck, against 4.8% in Brazil; 23% of French or English youngsters say they have received racist insults against 9% of Brazilians, with insults generally being more present in France (75%) and England (65%) than in Brazil. School peace is also expressed through better relations with teachers – more than 25% of French pupils against 11.5% of Brazilian pupils express rejection of them.

Dependence is highly significant. chi2 = 763,95, ddl = 4, 1-p = >99,99%.

Tableau 3: Comparison of pupil-teacher relations in France and Brazil.  Source: Sources: Debarbieux (2003), Unesco (2004)


In fact, every dimension of school climate is better, according to the Brazilian pupils. In addition, ethnographic observations over several months at a school in the favela of La Rocinha, in Rio de Janeiro, conducted by Benjamin Moignard, a European Observatory researcher, agree with this factor of school peace – against a backdrop of extreme violence in the community. Better school climate and fewer victims, in schools which are solely working-class and often very poor: this is a first solid result of our comparative studies and a first paradox.


The Republic of Djibouti is ranked 154th out of 177 countries on the Human Development Index (France is 12th) and the infant mortality rate is 138 per thousand (it is, for example, 6 per thousand in Canada). Life expectancy is 45 years. Gross national income per inhabitant is US$ 910 and there are considerable difficulties to obtain access to water, even in certain districts of the capital.

We questioned a representative sample of pupils in the last two years of elementary education, in both rural and urban areas, at schools which often survive in difficult material conditions in terms of both staffing and facilities (up to 1,200 pupils with 60 per class, major sanitary problems, sometimes a total lack of water and electricity, etc.). Nonetheless, we must first emphasise the quality of the climate in these schools. It is true that in our research, primary schools very often have a better climate than secondary schools. But in this country, our results are most impressive. All the indicators show a high level of satisfaction among young Djiboutians. Here are the answers to our first question, "Your school is…?” with choices ranging from "terrible” to "great”.

Dependence is highly significant. chi2 = 139,79, ddl = 4, 1-p = >99,99%.

Table 4: Comparison of the general assessment of school by Djiboutian and French schoolchildren N= 1413 (Source Debarbieux, 2004 and 2005)

Djiboutian pupils have an excellent view of their school, with almost 49% finding it great and 33.6% judging it good, for a total of 82.4% who rate their school in the top two categories on the evaluation scale. 6.8% find their school terrible or not very good. This is not solely due to the age of the pupils; it is specific to the place of schools in Djiboutian society. If we compare these results with those of our most recent survey of French primary schools, the Djiboutian responses do seem very positive. Their teachers are judged "great” for 51% (against a third of French pupils). Similarly, relations between pupils seem quite good, with more than 76% of favourable or very favourable opinions (compared with 61% in France). The feeling that violence exists in their school is affirmed by a fairly significant fraction of Djiboutian pupils, yet this fraction is smaller than in France, while the proportion of pupils who think that violence is totally absent is much higher, as the table below shows.

Dependence is highly significant. chi2 = 567,13, ddl = 4, 1-p = >99,99%.

Table 5: Violence perceived in school by Djiboutian and French schoolchildren

While violence among peers is far from absent in Djiboutian schools, in comparative terms physical and verbal attacks seem to be less common. Only 40% of pupils report insults, against 78% in France. 25.4% of Djiboutian pupils say they have been struck by someone else – in France half of the primary school pupils say this is the case. 42.6% say they have had something stolen from them against more than 51% in France. The same paradox as that seen in poor secondary schools in Brazil emerges: relative school peace, against all predictions.

Burkina Faso

This paradox is also present in the secondary schools (junior and senior high schools) of Burkina Faso, according to the survey by Lompo (Lompo, 2005), whose thesis was presented under our supervision. Burkina Faso is one of the world's most deprived countries – indeed, it is 175th out of 177 on the Human Development Index. Gross national income per inhabitant is less than $US 1 a day and life expectancy is 46 years. Another major problem is that only half the population has access to a source of drinking water. The literacy rate among adults is 13%, and barely 8% among women. Joseph Dougoudia Lompo adapted and administered our questionnaire in a total of 80 schools, in other words 16.13% of the 496 secondary schools in the country. Once again, the results are disconcerting: according to both pupils and teachers, violence would appear to be almost totally absent from schools. 61% of them clearly state that there is none at all, while 28.8% think that there is "a little”. Only 2% see violence as very present. As for victimisations, although they are present, they are less frequent than in the Northern countries we studied. For example, insults are not far off half as widespread as in France with less than 45% of occurrences. This does not mean that there are no major problems, particularly regarding pupil-teacher relations which seem to be sometimes tense with the fairly common practice of sexual blackmail by certain teachers towards young girls in order to improve their "sexually transmitted good marks”. Nonetheless, school delinquency among Burkina Faso school pupils seems to be limited.


3: Discussion: Explaining the paradox. A broader contextualisation


The example of these countries is therefore disconcerting. In Brazil – despite a high level of violence linked to the traffic of drugs and mass social exclusion – and in Djibouti and Burkina Faso – two of the poorest countries in the world – violence in schools is not as great a problem as it is in the so-called developed countries, although this does not mean it is absent. The countries ranked 73rd, 149th and 175th on the Human Development Index have lower levels of school violence than certain developed countries among the richest on the planet. There is, to say the least, something in there to challenge the accepted model of risk factors based on poverty and social exclusion.

Several explanations can be put forward, aside from any fascination for cheap exoticism. Some of these explanations are structural, aiming to show that problems other than violence are posed in these schools, while other, more sociological explanations oscillate between reification (the incorporation of the habitus in its school form by the dominated of the planet) and the use of an "immaterial social capital” by the communities in the very poor areas we worked in.

Material conditions and structural factors

First, there are the structural explanations linked to material conditions. In the case of Brazil, there is a simple solution: owing to the lack of material means and investment in education, pupils in state schools often only have lessons for half the day, either in the morning or in the afternoon. In short, the less time they spend in school, the fewer opportunities there are for delinquency. A second, complementary structural factor is present to varying degrees in all three countries: the non-schooling of many pupils, particularly teenagers. Sociologically speaking, it is probable that only a limited elite has access to secondary education in these countries. The thesis by Joseph Lompo looks in detail at the effects of the clientelism of local potentates on the possibilities of pupils continuing their studies. The real problem of exclusion in these countries is not school violence in frustrated reaction to inequality, but exclusion from school. This structural factor may also have psychological effects, as school is still seen as a desirable social commodity rather than a tedious obligation and its value has not been challenged, unlike the mass-consumption school of western countries (Dubet, 1994).

Among the material reasons, the work carried out by children is also important, especially in Africa (grazing cattle and helping out at home), and in Brazil, for symmetrical reasons, the early age at which certain children join gangs, distancing them from the world of school. It could be said that the most violent adolescents are absent from schools, having started a (sometimes short) delinquent life or taken a job to scrape a living.

"School peace” as incorporation of domination?

Another explanation for this "school peace” might be sociologically and psychologically valid: that of the incorporation of the domination experienced by the most deprived social classes on the planet. In other words, acceptance of the unfair world order is such that rebelliousness is inconceivable, because there is a belief in the natural destiny of this order, strengthened by centuries of colonisation and slavery both in Africa and in Brazil – one of the first countries to have abolished slavery. Is this not one of the lessons given by the economics Nobel Prize winner, Armatya Sen: "inequality and persistent exploitation often spread by making passive allies of those who are maltreated and exploited”, because "those who are at the bottom of the ladder can come to consider their fate as something that it is virtually inescapable, something to be borne placidly and calmly” (Sen, 1990). This can refer to symbolic violence defined as "power which manages to impose meanings and to impose them as legitimate by dissimulating the balance of power which is the basis of its strength” (Bourdieu & Passeron, 1970, 18). For Bourdieu, the "dominated products of an order dominated by forces draped in reason (like those which act through the verdicts of the school institution or through the diktats of economic experts) can but give their acquiescence to the arbitrary of rationalised force” (Bourdieu, 1997, p.99). The importance of corporal punishment is such, in this "bodily apprenticeship” that is the dominated habitus, that we must raise questions about the permanence of this practice in some of the countries studied.

Corporal punishment: a solution?

A "culturalist”, commonsense explanation comes to light in the African schools in our sample, and it is important to raise the issue because it opens up an important debate in this international dialogue. Our survey reveals the still-prevalent practice of corporal punishment. An incorrect question could be posed: is it because corporal punishment persists in these schools that there is no violence among children? This is a leap easily made by many of the people we spoke to: "In the West you have a lot of violence in schools because you no longer have any authority, you are too lax”. Is "Western” condemnation of corporal punishment[2] simply ethnocentrism preventing a good, simple solution?

The types of school punishment actually practised in Djibouti include a certain dose of corporal punishment. The commonest is "the pipe”, a plastic pipe used to smack the calves or hands of children. Our study shows that 27.6% of Djiboutian schoolchildren state they have been hit with this instrument, while 19.5% have been told to kneel, often for long periods, with their hands behind their head, and 17.1% have been slapped or spanked.

Although most people we talked to agree that these practices are diminishing, many question the real likelihood of them disappearing altogether, seeing a risk of powerlessness and the knowledge that they are a strongly-rooted tradition, popular among parents. However, it is worth wondering whether this culturalist "tradition” interpretation is no more than a naturalisation of corporal punishment. There is no "unanimous tradition”. If there were, we would find the same number of pupils beaten in similar schools. Yet the proportion of pupils who have been hit with the pipe varies from around 11% to nearly 45%, those who have been smacked or spanked from 4% to almost 38%, those who have received other corporal punishments from 4% to over 33%, and for those made to kneel from 3% to 40%. The wide variety of situations is the rule.

One of the schools with the lowest rate of corporal punishment is also one of the schools which currently achieves the best academic results, and it is a school that can be described as semi-rural, without water or electricity or telephone, with 24 two-speed classes. It nonetheless achieves the best school results in the canton, with a success rate (for the secondary school entrance exam) of over 60% over the last four years, AND it barely practises corporal punishment. In fact, whatever the area (rural, semi-rural, peripheral or centre urban), the schools with the best academic results are those where corporal punishment is the least prevalent, and it is in these schools that the pupils' sense of security is highest.

Conversely, an increase in victimisations by pupils is strongly correlated to an increase in victimisations by the teachers. The more a pupil is victimised by other pupils, the more he or she is beaten by the teacher. Just one, eloquent example: there is a school in which 80.9% of pupils say they have been beaten by their teacher. This is also the school where we have noted the highest rate of pupils struck by their classmates: 68% against the average of 25%. We can thus answer unambiguously that corporal punishment does not explain school peace in Djibouti – quite the contrary. It is in the schools where punishment is most prevalent that violence among pupils is most widespread. Corporal punishment is not just a tradition; it is also an educational choice. And it is the wrong choice.

Social capital and violence

The explanation involving the incorporation of the habitus, even by violence, is, however, too mechanistic. Without denying socio-historical or economic factors, our observations lead us to look at other, less fatalistic explanations, with the notion of "social capital”. The notion of social capital was first proposed to describe the gap, not only material but also in terms of power networks and of advantage of position in the networks for the dominant classes (Bourdieu, 1980). But it can also highlight certain particularities of the dominated communities and explain the paradoxical results of our studies. The way in which this theory is currently used by urban research in the United States seems to us to be particularly interesting.

Social capital can be material, first: the social, sports and cultural facilities of a community are correlated to the differences in victimisation in the schools (Gouvis Roman & Moore, 2003), providing that this social capital is accessible to all and that its management involves the genuine investment of the community.

However, the social capital of a community is also immaterial: it increases according to the shared values, cultural and religious activities of a community while increasing its social control (Cusson, 1983; Reitzel & Khey, 2003). To develop this point, let us take the example of Djibouti. The "material social capital” in urban areas, which are better off than the rural areas, is nonetheless very limited; we do not have enough space here to describe the extreme destitution of the districts and villages in which we conducted our study. The fight for survival is not misplaced as a description. But social capital is not only material and in Djibouti, it must be emphasised, the strength of the community remains very important. It is one of the key explanations for the good school climate given by teachers and head teachers. The sense of belonging to a community is very strong, and comes as part of a deep attachment to the district, village or community. Among other things, this sentiment seems to be linked to the resonance of the Muslim faith. And yet the community link is not just a matter of faith, and here, Islam seems to be benefiting from a long tradition that it has strengthened rather than created. Rather, it is the link of proximity which creates a strong regulation. Only the sense of belonging to a community can explain why, in the conditions of extreme poverty in which the vast majority of Djiboutian children live, they seem to appreciate their living environment more than French children who live in much better conditions: almost 43% of pupils questioned in the shanty towns of Djibouti think their district is "great”, compared to 31% of French youngsters.

It is clear that in saying this we are neither romanticising poverty nor praising destitution. But after this research in very economically deprived communities, it seems clear to us that there is a "strength of the weak” which can help close-knit survival and education, with the school benefiting all the more because the teachers themselves are in tune with a social environment that they often come from. The presence of parents at the school is common – in the playground, in the head teacher's office, at the canteen. The school is very much the community school, and is part of the social link. This feature seems to be common to many African countries (Latouche, 1998), including urban Africa; perhaps what we are seeing here is what Anne Cécile Robert calls the simple virtues of the social link (Robert, 2004). Having said that, community protection in Brazil is also related to an arrangement between the schools and the drug dealers, as the head teachers have affirmed. Thus, only a refined contextualisation can give an approach to the sense of the paradox, one which will require more research to confirm or disprove our interpretations. This is one of the aims of the observatories which are currently being set up, not so much to "find violence” in all countries as to understand, comparatively, the best resistance to it by some of them. Here, for example, it is highly likely that the "South” could teach us a lesson about the importance of the social link of proximity. In the end, is it not a fact that, as well as the costly "programmes” or "plans” against violence in school, much in vogue on the "school security market” (Devine, 2001), it is also worth attempting to understand how to make the most of those day-to-day protectors of the school that are parents and the neighbourhood community?

    Download the pdf file here


ADDA, J. (2002). La mondialisation de l'économie. Paris: La découverte. 2 tomes.

BENBENISTHY, R. & ASTOR, R.A. (2005), School Violence in Context: Culture, Neighborhood, Family, School and Gender. New York: Oxford University Press.

BLAYA, C. (2006). Violences et maltraitance en milieu scolaire. Paris: Armand Colin.

BOURDIEU P. (1980). Le capital social. Notes provisoires. Actes de la recherche en sciences sociales, n°31, janvier 1980, p.2-3. (repris in BEVORT A. et LALLEMENT M. (2006). Le capital social. Performance, équité et réciprocité. Editions MAUSS/La Découverte.

BOURDIEU, P. (1997), Méditations pascaliennes. Paris, Seuil.

BOURDIEU, P., PASSERON, J.-C. (1970). La reproduction. Éléments pour une théorie du système d'enseignement. Paris: Minuit.

BROWNSTEIN, H.H. (2000). The Social Reality of Violence and Violent Crime. USA: Needham Heights. Canada:

COIE, J. D. & JACOBS, M. R. (1993). The Role of Social Context in the Prevention of Conduct Disorders. Development and Psychopathology 5: 263-276

CORNELL, D.G.  (2003). Guidelines for responding to student threats of violence.  Journal of educational administration, vol 41 n°6.

CUSSON, M. (1983). Le contrôle social du crime. Paris: PUF.

CUSSON, M. (2000), Criminologie. Paris, Hachette.

DEBARBIEUX, E. (1996), La violence en milieu scolaire - 1- Etat des lieux. Paris, ESF

DEBARBIEUX, E. (1997), « Insécurités et clivages sociaux: l'exemple des violences scolaires ». Les annales de la recherche urbaine, n°75, pp. 43-50.

DEBARBIEUX, E. (2004). « Les enquêtes de victimation en milieu scolaire: leçons critiques et innovations méthodologiques ». Déviance et Société. Vol 28, n°3, pp. 317-333.

DEBARBIEUX, E. (2006). Violence à l'école: un défi mondial ? Paris: Armand Colin.

DEVINE (J.) (2001), "Le marché de la violence scolaire", in Debarbieux E . et BLAYA C. (2001), Violences à l'école et politiques publiques, Paris, ESF.

DUBET F. (1994). Les mutations du système scolaire et les violences à l'école , Les Cahiers de la sécurité intérieure, n° 15

FARRINGTON (D.P.) (1986). Unemployment, school leaving and crime. British Journal of Criminology, 26, 335-356.

FARRINGTON, D.P. (1989). Early predictors of adolescent agression and adult violence. Violence ans Victims, n°4, pp. 79-1400

GAUVIN (M.) (dir.), (1995). La violence en milieu scolaire au Québec 1988-1992: l'état de la question. Analyse des rapport sociaux inhérents à la production du discours sur le sujet. Etudes et Recherches Vol.2, n°1. Québec: Université Laval-CRIRES.

GILL (M.) & HEARNSHAW (M .) (1997), Personal Safety and Violence in Schools. Research report RR21. Londres, DFEE.

GOTTFREDSON (G. D.) & GOTTFREDSON (D.) (1985), Victimisation in schools, New-York: Plenum Press.

GOTTFREDSON, (D.C.) (2001). Schools and delinquency, Cambridge: University Press.

GOUVIS ROMAN(C.) & MOORE (G.E.), (2003), « Community organization as social capital: Theory, Measurement and Practice ». The American society of criminology 55th annual meeting. Denver, Colorado.

KALTIALA-HEINO, R., RIMPELA, M., MARTTUNEN, M., RIMPELA, A., RANTENAN, P. (1999).  Bullying, depression, and suicidal ideation in Finnish adolescents: School survey.  British Medical Journal, 319

LATOUCHE (S.), 1998, « Marché et marchés » in L'autre Afrique. Paris, Albin-Michel.

LINDSTRÖM, P. (2001). School-violence, a multi-level perspective. International Review of Victimology, 8, 141-158.

LOMPO, J.D. (2005). Violence et climat scolaire dans les écoles secondaires du Burkina Faso. Thèse de doctorat en Sciences de l'éducation sous la direction d'Eric Debarbieux. Bordeaux, exemplaire photocopié.

MOREAU DEFARGES, P. (1997). La mondialisation. Paris: PUF.

NCES (National Center for Education Statistics) (2004), Indicators of school Crime and Safety 2002. Washington, US department of education.

REITZEL (J.D.), KHEY (D.N.) (2003), "Applying Messner and Rosenfeld's Institutionnal Anomie  Theory: Homicide Rates at the Neighborhood Level”. The American society of criminology 55th annual meeting. Denver, Colorado.

ROBERT, A.C.  (2004) L'Afrique au secours de l'Occident. Paris, L'atelier.

SEN, A. (2003). L'économie est une science morale. Paris: La découverte.

SMITH, P.K. (ed.), (2002). Violence in Schools: the response in Europe. London: RoutledgeFalmer Tremblay & alii 1996, Kauffman, 1997

VETTENBURG, N. (1998) Violences à l'école: sensibilisation, prévention, répression. Rapport du Symposium tenu à Bruxelles (Belgique), 26-28 novembre 1998. Editions du Conseil de l'Europe.

WILLIAMS, K., CHAMBERS, M., LOGAN, S., ROBINSON, D. (1996) Association of common health symptoms with bullying in primary school children.  British Medical Journal, 313.

Read also

> Editorial
> Dynamique démocratique et violence scolaire
> Keeping Violence in Perspective
> La violence au préscolaire et au primaire : Aperçu de la situation canadienne
> Revues systématiques dans le champ criminologique et le groupe crime et justice de la collaboration Campbell
> The Effectiveness of School-Based Violence Prevention Programs for Reducing Disruptive and Aggressive Behavior: A Meta-analysis
> Violence prevention : schools and communities working in partnership

<< Back