5 - European trends in research into violence and deviance in schools
by Carra Cécile, IUFM du Nord/Pas de Calais-Université d'Artois, RECIFES, CESDIP-CNRS

Theme : International Journal on Violence and School, n°10, December 2009

This article has set itself the task of putting into perspective the European trends in research into violence in schools, the dominant definitions, the preferred methodologies, their underlying problems, and the type of data available. It questions the implications, simultaneously shedding light on grey areas. It is based particularly on national summaries from approximately ten European countries produced as part of the CRIMPREV action financed by the European Commission and coordinated by the European Group of Research into Normativeness (GERN =Groupe Européen de Recherches sur les Normativités) while also making use of international reports. This investigation is attempting to clear the way for future research work.

Keywords : Violence in schools, deviances, standards, risks, anti-social behavior, socialization, the actor’s strategy.
PDF file here.
Click on the title to see the text.

The 1990s witnessed, in the most European countries and rich countries outside the European Union, an explosion of violence in schools as a social problem which was accompanied by a host of solutions devised to deal with the situation. Researchers took over the topic with a view to attempting to assess its extent and nature in order to produce a profile. The explanatory objective would frequently only develop during a second phase. Can we assert today that research into violence in the school environment has acquired a certain maturity? How is it structured in Europe? What are its main results? What are its grey areas? Have researchers been able to highlight cases of social problems of violence in schools, cases which tend to impose their construction and interpretation charts? Is the "violence in schools" object still attempting to define itself? How is it put together today, in terms of definitions, concepts, methodologies? What are the preferred explanations? On what theories and disciplines are they based? Do public answers integrate information produced by scientists? These are the major issues that researchers from ten countries in Europe have been led to examine: Germany, England, Belgium, Spain, France, Italy, Lithuania, Luxembourg, Portugal and the Czech Republic . Be they European countries under the Treaty of Rome, or former Eastern Block countries that have recently joined the European Community, all these countries provide a range that can be used to gain an idea of European trends in research into violence in schools and its different degrees of development.
This text is not an overview of national summaries which are, nevertheless, at the heart of this special issue of the IJVS. It is definitely more than an attempt to put into perspective European trends in research, their underlying problems, the type of data available and to be gathered, but also issues and problems some of which have already been the subject of exchanges upheld during the conference which was held in Paris from the 8th to 10th January 2009 , bringing together all rapporteurs from the countries mentioned earlier. Although it is also based on all of the national reports, this contribution nevertheless transcends them, drawing on other European works. In order to give substance to this outlook, this approach follows in the footsteps of other international reports (in particular, Carra & Sicot 1996; Vettenburg, 2000; Up & Blaya, 2001; Carra & Faggianelli, 2003; Smith, 2003). Putting issues into perspective involves sacrificing details and nuances but allows us to identify the main thrusts and to examine their implications .

Faced with the continued media and political coverage of the topic, researchers will endeavour to assess the extent of the phenomenon. For this purpose, the first reaction consists in referring to institutional data, police and judicial data in most cases and, more rarely, because they are frequently non-existent, to school-related data. When they are available, institutional statistics are, however, inadequate or irrelevant for researchers who will then build their own indicators. The 1990s witnessed the production of statistics that were generally descriptive, tending to use categories created through public debate, policies, the school institution etc. thus bringing behaviour into the classification of violence, by virtue of systems constructed outside the scientific field. Moreover, any correlation is rarely disputed because it helps to establish a cause and effect interpretation. The following investigations will be more wary of these pitfalls, moving from the production of descriptive statistics to that of explanatory statistics, and attempting to objectify the phenomenon.
If statistics currently provide more data on violence, they do not, however, all address violence in schools. In some countries, researchers use statistics for understanding violence in young people, whether or not this violence takes place in school. This approach may reveal a lack of specific statistics; more often, it implicitly demonstrates the role generally attributed to contexts and, in particular, to schools. The underlying problems are then those of juvenile delinquency addressed by international surveys into self-reported delinquency (ISRD 1 and 2) or victimisation (ICVS ) or those on health disseminated by epidemiological surveys and by international mechanisms (ESPAD , HSBC ). Rarer are those who seek to objectify the phenomenon within the school. We must, however emphasize the importance of surveys on school bullying and the Olweus «bullying/victims» questionnaire (1993) and its variants. Rarer still are those which take into account not only the socio-school context but also its role in creating violence: we can, however, quote the Debarbieux victimisation and school climate questionnaire (1996) deployed in France but also in other European countries: Germany, Belgium, England, Spain and the Czech Republic. These investigations reveal a high dark number for juvenile deviance and high pupil victimisation .
The widely preferred quantitative surveys are, however, variously systematized depending on the country. But some have managed to build temporal series in order to provide an answer to the recurring question of the changing phenomenon of violence. Results converge, demonstrating that not only has there not been any explosion of violence but, what is more, violence very often remains at fairly stable levels. Although its frequency does not appear to have altered much, we still need to query this change from the viewpoint of another criterion: social inequality and the distribution of violent behaviour: this phenomenon is far more present in areas of social relegation and appears to be on the increase in some countries.
With the media coverage of school bullying and its new forms (especially cyber bullying), the quantification of violent deeds has again become a pressing issue in recent years and has given rise to the production of new statistics. The preoccupation concerning the measure accompanied by a request for an international comparison has caused the definition of the violence and the categories used in different countries to be questioned.

The strongest trend today consists in considering violence in the broad sense, that is to say, not reduced to physical violence, or limited to legal categories. However, the definitions used form part of different problems. Three of them currently appear to predominate:
? the Olweus definition (1993) is based on the «school bullying» concept; it broadly involves the idea of repeated bullying of students or harassment between peers. It implies the deliberate intention of the student or group of students to cause harm to one of their peers, an imbalance of forces and repetitive acts. Therefore, we need to pay attention to interpersonal violence between pupils and to focus on the assessment of the psycho-social risks to be prevented.
? the World Health Organization (WHO, 2005) definition with categories selected on the basis of the injury suffered or damage inflicted, where human, social and economic costs should be reduced: self harm (attempted suicide, legal or illegal drug taking etc.) or other (homicide, aggression etc.), damage to society (vandalism, discrimination etc.). The key concept then becomes the «risk behaviours » found particularly in the epidemiological surveys.
? A definition derived from that of delinquency and using legal categories while extending these to «anti-social behaviour» (physical violence or threats against persons, other forms of delinquent behaviour such as theft or drug taking, offenses associated with the status of minor such as frequenting cafés, violation of established rules and especially school rules). The author and his intentionality can be found at the centre of this approach.
All these definitions have the effect of viewing violence from the standpoint of the individual, in this case, the student or even young person, the adolescent and the minor all grouped together within the concepts applied: interpersonal violence, behaviour at risk and anti-social behaviour. Despite these points in common, reaching a common definition of violence appears extremely difficult: should one start from the perpetrator and his intention in order to define violence, or from the victim and the harm suffered? Two major criticisms emerged from the seminar: the perpetrator may not realize that his behaviour is violent while the victim may not know that he has been a victim. Regardless of the approach preferred, that of the perpetrator or that of the victim, the difficulties persist when establishing the categories selected, the recurrent issues being those of the serious nature of the act and of category objectivity. Are playground skirmishes sufficiently serious to be included in the violence classification? Is the ostracism category too subjective? Clearly, a choice will need to be made, a choice which will be based on reference to standards and the respective or presumed weight of each within the working groups resulting in the construction of data collection tools. It is a choice which will also come from opportunities deemed to be put forward should such or such category be selected and more widely if such an approach is preferred (being selected for inclusion, for example, in a project that is funded or even being able to compare data). Occasionally, it will be a default choice when researchers are unable to identify a formulation capable of being understood by young students such as homophobic related violence.
Irrespective of the definition applied, the results converge to show that the very essence of violence in schools permeates daily life in the form of micro-violence, minor victimisation, rudeness, disrupting the school order and more broadly, the social order. Researchers talk of low intensity violence, below the criminal threshold, reiterating the rarity of serious events. The severity of an incident should not be sought in terms of the form taken by the violence but rather in terms of repeated victimisation. However, this finding does not produce feedback on social crystallization social regarding this type of behaviour, but to short and sharp assaults on views propagated by the media.

If the research into school violence has proliferated since the 1990s, not all, especially at the beginning of this period, incorporate a clarification perspective but accumulate data for the purpose of providing a description of the situation. Devoid of a theoretical framework, they are most often founded on an inventory of behaviour regarded as problematic. If this situation has changed significantly since then, research into violence in schools can still be based on the theoretical models that are little, if at all, explicit while several theoretical models can coexist in the same research work. These trends can be seen in major international surveys based on compromises. The works can also make do with empirically proven data in order to produce a data collection tool.
The variables used, in the same way as the forms of violence selected, reveal explanatory trends that are more or less detailed and more or less identifiable. Trends develop at a different rate depending on the country’s scientific tradition and on inputs from the various disciplines. Sociological research tends to investigate the contextual and social variables inherent in violence. Research into education will address the impact made by violence on the school climate and on learning, while considering prevention and intervention strategies. Criminology tends to examine individual causes and effects on the Law etc.
For all that, we can mention two major problems. The first, the most classic, is the one that is founded on the paradigm of standards and deviances. This issue includes the breakdown of the social structure and social disorganization, a lack of socialization and lack of integration in order to explain a violation of standards. On the fringes, we also have cultural conflicts and stigmatization as part of the creation of deviance. The second problems, which tends to develop quite strongly, has risks and prejudices as its paradigm. The perpetrator develops strategies according to his interests but his strategies are based on taking inappropriate risks. This risk-taking is then construed as a pathology causing harm – from self-prejudice to the harm caused to the community - , individual pathology (inadequacy of the individual, interruptions to individual development) or social pathology (dysfunctional family environment, criminogenic environment). Concerning the first problem, the social order is taken as normed; violence is a violation of the social order through the breach of standards on which this social order is based. Concerning the second, the social order is built within interrelations and violence is a product that is detrimental to this order thus negotiated.
Transversally, positivism and aetiology typify the most strongly displayed trend. The aim consists in finding out what differentiates deviant individuals from others, deviance being considered by some approaches as a characteristic of the person. This trend coupled with practical prevention and security practices undoubtedly explains the success of the risk factor approach. The recognized legitimacy of the social order appears implicit.

This explains the dominant trend observed in research into centration involving individual and family risk factors. Individual factors such as gender and age are two variables closely correlated to violent behaviour. If boys constitute by far the greatest number of perpetrators – especially when physical violence is involved –, they also represent a very large proportion of the victims. Violence in schools develops at about the age of 12 years with a peak at about 16 years after which the violence drops off to a low level. Only a small proportion of young people will continue to develop these behaviours. Deviance is then interpreted as the normal transition into adolescence. Other approaches, in contrast, seek the causes of violations in the statistical association between poor intelligence, socio-cognitive shortcomings, hyperactivity, impulsiveness and violent behaviour. Failure at school and membership of an offender peer group also appear to be strongly related to the problematic behaviour of the individual. The study of family factors is particularly developed. The family factors such as parental criminality, intra-family conflict, parental violence, lack of control and random parenting (alternating rigid rules with indifference or permissiveness) or even poverty in the home appear to be significantly correlated to violent behaviour in children.
Rare in the 1990s, the introduction of contextual factors is today more frequent; however, the time taken to include the factors related to the school itself in its models may, however, be surprising. Moreover, there is no consensus on the role attributed to them. However, if one examines the factors applied, they tend less to typify the context but rather refer again to the individual. Accordingly, the factors used for schools are frequently limited to school failure, to weak attachment to the school, to absenteeism or to school dropout rates, to frequent changes of school, to disciplinary problems or even in low commitment to school activities. Therefore, scant attention is paid to the establishment, its team, its operation, its organization, its professional practices, its management of violence, the definition of violence, its climate which constitute as many aspects which nevertheless seem important if one refers to work on "the establishment effect." They show that in sociologically comparable environments, the establishment may make a difference in contributing to mitigate or exacerbate the phenomenon of violence. However, these works still seem to have little influence on the dominant research guidelines in terms of definition and explanation. In contrast, at the outset, the school is recognized as playing a prevention role if only through the socialisation process prevalent in this environment but also through the proliferation of programmes and actions which take place there in order to prevent, detect, monitor and treat behaviours at risk. The dominant representation which emerges by implication is one of a school which is affected by violence which is alien to it and which it must manage.
The socio-political and economic context is also neglected in numerous research works and this non contextualisation constitutes an obstacle to the understanding of the emergence of the social problem of violence at school. Among the research works which address this issue, certain elements continue to recur: educational system reforms where violence can be interpreted as one of the effects of the massive expansion; waves of migration, violence appearing as the sign of a lack of integration, a lack of socialization and sometimes as a cultural characteristic; changes in the political system and the problem of adjusting values; globalisation of the economy and increasing social inequality. The links between, on the one hand, the development of the violence phenomenon and the disparities between countries and, on the other hand, the changes to political and educational systems must be questioned and examined in depth.

Solutions for responding to violence in schools have proliferated during these last two decades, from managing conflicts in the classroom to the deployment of national or federal programmes or even government plans, from establishment policy to the creation of experimental schools, from team work to school-police-justice partnerships, from citizenship to laws, not overlooking internal regulations etc.
Political, institutional and professional mobilization seems important even if it is differentiated and is occasionally reduced to announcements on action or vague local attempts.
Some of these responses are closely linked to research such as school bullying prevention programmes tailored to suit the needs of different national contexts. These prevention programmes have been extremely widely disseminated within the European countries, including in the former countries of the Eastern Block that have recently joined the European Union, and well beyond. Other responses entertain apparently weaker links to research in the same way as the security measures which have proliferated in a number of countries in Europe as evidenced by the legal output of recent years in the field of juvenile justice. They are part of a radicalization of professional practices, accompanied by an inflation of rules, a tightening of sanctions and an increasing number of cases referred to the courts. This last approach is an important aspect of the problem which has evolved in terms of risk management. The aim consists in deterring individuals from developing certain behaviours by levying heavier penalties. This punitive prevention approach tends to combine with an initiative seeking to prevent the occurrence of situations in an attempt to render risk situations rarer through the development of technological and human monitoring resources; at the same time, the aim consists in forestalling the danger of victimisation. The premise is that the individual at risk will rationally analyse the risk in terms of costs and benefits and that the threat of repression burdening the loss side, prevents him from progressing with the action.
This risk management problem has a third aspect, that of preventing treatment based on the identification of individuals at risk, individuals whose negative psychosocial development has to be corrected, acting on their direct environment and their families in most cases. The premise involved here is that the causes of behaviour at risk have their origin in individual and family dysfunctions. It may be surprising to see, in this issue of risk management, the apparently contradiction between the co-existence of a concept of the individual at risk and a rational perpetrator in some cases, and the individual whose behaviour is determined by the individual or family deficiencies in other cases. Irrespective, in each of these cases, we have to manage risks from an essential defensive position.
In parallel, and tending to blend during the last decade with the - largely dominant - approach in terms of risk management, actions forming part of a social prevention logic have been deployed. In the latter case, developed in very different ways depending on the country, the aim consists in addressing the presumed causes of the problem by combating social and urban relegation or even failure at school and dropping out. The concept underlying this view maintains that problems involving violence originate in the living conditions of the individual. These answers are accompanied by actions that are supposed to contribute to the socialization not only of students but also of their parents. Developing social skills becomes a recurrent objective in a context where the problems of violence are interpreted as relational, communicational and especially behavioural problems.
Thus, there is a transition from social prevention to the general prevention – or primary prevention – of behaviours at risk.

Focusing on the student, particularly the school pupil, and on the methodologies used, explains the availability of data concerning age, gender, school career, membership of an ethnic group etc. However, the latter poses a problem in some countries whereas, in others, it forms a central criterion. This variable, unlike others, continues to be hedged in by political disagreements which jeopardise its deployment. Where it is available, results tend to converge: pupils who belong to an ethnic minority group are significantly over-represented in victimisation, this not applying in the case of the perpetrators of violence.
These data extend to the pupil’s family circle. However, the number of parent-related data is astonishing as, unlike students, parents are rarely directly involved in the survey. It is true to say that access to families is not easy. Therefore, the role of the family and especially of the working class family is strongly queried as part of the origins of violence unlike other major socialisation structures such as schools. This is clearly also what enables us to understand the little data available on teachers. To this has to be added the strong resistance opposed by the profession when questions focus on the teacher and his professional practices. Nevertheless, existing research work demonstrates the importance of the role played by socialisation, contributing to the construction or deconstruction process applicable to violence found in teacher-pupil interactions.
The quantitative approach dominates and statistics are numerous. For all that, longitudinal studies are rare. When they do not exclusively form part of a descriptive objective, they are of value because they objectivise the phenomenon and stabilise the definition of correlations. In the trend that currently prevails, these correlations are seen as just so many risk factors – or protection factors. However, this approach cannot be used to open the «black box», to question the processes implemented, to bring out into the open situations where behaviours will be regarded as violent, to report on interactions that will lead to violence. We need to rely on qualitative research work that is more capable of accounting for these analysis levels and which can enrich illustrative views.
The most interpretative currents and the most critical views are marginalised in the current scope of research into violence in schools. The social order, the inequalities it can produce, the social effects these can generate, are not sufficiently questioned for their part in creating violence. It is undoubtedly because of this that data appears to be placed into context in different ways. What role do macro-social variables play in the violence phenomenon? Here again, research is needed in order to reply to these questions, especially as a comparative approach to national contexts. Answering this questionnaire is not just part of the additional research that has to be undertaken, particularly as more than a few have already been carried out. It is also a theoretical comment. The lack of theory, which is very noticeable in some countries, renders the causal attributions that can be made on the basis of correlations that are recorded fragile without being able to remedy these using the sophistication provided by statistical calculations.
A salient concern emerged during the conference: producing a common definition and harmonizing violence categories in order to be able to undertake comparative research. However common definitions are circulating, especially those put forward by supranational entities. They tend to home in on violence perpetrated by and between students. There is much less data on violence between students and the professionals working in educational establishments, especially when the student is the victim. If we believe the definition of violence to be – too – broad, it has, however, become somewhat depleted theoretically, marginalizing structural violence, symbolic violence and even institutional violence. Little deconstruction is applied to the categories selected and the implications are not questioned enough. What do they tell us about the way in which our societies work, about ideas on childhood, about relations with young people, about the role played by the school and our attitude to deviance? Behind the words, the categories, the interpretations, the replies, there are views of the world, standard and cultural systems that are insidiously imposed on researchers. However, the make-up of the social problem that is violence cannot be exclusively perceived as changes in student behaviour. It acquires meaning within the contexts; it is formed within social relations where it becomes a power issue, the power to define what is violence, the power to decide who is violent, the power to define the solutions to be devised. Nowadays, violence in schools which, we must not forget, is linked to normal academic and social disruption, is primarily interpreted as individual pathologies questioning what seemed to form an achievement of multi-discipline research: the normal nature of deviance during adolescence...


Allemagne : Marek FUCHS, Université de Kassel.
Angleterre : Carol HAYDEN, Université de Portsmouth.
Belgique, Flandres : Johan DEKLERCK, Université catholique de Leuven (KULeuven).
Belgique, Wallonie : Marie VERHOEVEN et Eliz SERHADLIOGLU, Université catholique de Louvain (UCL).
Espagne : Anabel RODRÍGUEZ BASANTA, Association Centro de Estudios de Seguridad.
France : Yves MONTOYA, Université Victor Segalen, Bordeaux 2.
Italie : Franco PRINA, Université de Turin.
Lithuanie : Robertas POVILAITIS, Président de l’Association Lithuanienne de Psychologie.
Luxembourg : Georges STEFFGEN, Université du Luxembourg.
Portugal : Susana CARVALHOSA, Université de Lisbonne.
Tchéquie : Magdalena KOHOUT-DIAZ, IUFM d’Aquitaine, Université Bordeaux 2 et 4.
CARRA C., (2009), Violences à l'école élémentaire : l'expérience des élèves et des enseignants, Paris, PUF.
CARRA C., FAGGIANELLI D., (2003), « Violences à l’école, tendances internationales de la recherche en sociologie », Déviance et Société, 27, 2, 205-225.
CARRA C., SICOT F., (1996), « Perturbations et violences à l’école », Déviance et société, 20, 1, 85-97.
DEBARBIEUX E., (2006), Violences à l'école : un défi mondial ?, Paris, Armand Colin.
DEBARBIEUX E., (1996), La violence en milieu scolaire, Tome I : Etat des lieux, Paris, ESF.
DEBARBIEUX E., BLAYA C., (2001), La violence en milieu scolaire. Tome 3 : Dix approches en Europe, Paris, ESF éd.
GROENEMEYER A., (2007), « La normativité à l'épreuve. Changement social, transformation institutionnelle et interrogations sur l'usage du concept de déviance », Déviance et Société, 31, 4, 421-444.
OLWEUS D., (1993), Bullying at school : what we know and what we can do, Oxford, Blackwell Publishers.
SMITH P. K., (2003), « La violence à l'école : perspective européenne », CONNECT UK-001, accessible en ligne :
VETTENBURG N., (2000), « Rapport général » in Violence à l'école : sensibilisation, prévention, répression, Strasbourg, Conseil de l'Europe.
WORLD HEALTH ORGANIZATION, (2005), Injuries and violence in Europe, Copenhagen.

Read also

> Summary
> 1 - Problem behaviour and prevention
> 2 - Schools being tested by violence
> 3 - Deviant behaviour and violence in Luxembourg schools
> 4 - School violence in Spain
> 6 - Extra - Stakes of violence in education in Africa

<< Back