Introduction

It is always possible to bond a large number of people through love, as long as there are others left to externalize aggressiveness (^{Freud, 2011}, p. 60).

In that action, both sadistic aspects linked to narcissistic traits (in which the aggressive impulses turn to the outside without discrimination or delimitation of the object on which the violence is projected) and to prejudice (in which such aggressiveness can find, as a rule, from psychic mechanisms such as projection and/or reactive formation, a defined and delimited object on which the violence is projected) (^{Crochík et al., 2018}) would be present.

The 'narcissism of small differences' (^{Freud, 2011}) and the above-mentioned tendency to aggression can be understood as elements present in this type of conduct. Violence against a classmate can mean both a deepening of the process of identification among peers and blind adherence (facilitating cohesion among members of a group), as a way of satisfying the aggressiveness by choosing an appropriate target. That is because, in antipathy to classmates, one also finds self-affirmation, self-love that can encompass classmates who condone the aggression against another person while there is a mutual identification with the aggressor.

Verbal aggression, even if violent, is not the same as physical aggression. However, each situation of violence, be it verbal or physical, imprints its mark on everyone involved - even those present as observers - since it can mutilate both the consciousness and the body, which adopts a form equally conducive to sadism and brutality (^{Adorno, 1995a}).

But it is not a matter of seeking the roots for such events only in intrapsychic dynamics, as if they were separated from the external world. Violence against the weakest must also be examined in the light of the school itself, its educational processes of valuing competition and productivity in relation to the society that supports it (^{Crochík et al., 2018}). We stress here the situational character of school violence. It demands a broader look at the causing factors of such violence within society. It brings to the forefront issues related to the libidinal and aggression economy of individuals. Namely, repressions and social permissions for discharging such affections into a target relate to the configurations that life in society has assumed. One should highlight in it the exclusion of individuals, the tendency of an education for adaptation, and the naturalization of injustices (^{Crochík et al., 2018}).

If contemporary society organizes itself emphasizing technique to solve tasks in an immediate and fragmented manner, it also bears responsibility for tension and irrationality. It establishes the rules for individuals to insert themselves socially (^{Marcuse, 1964/1973}). In that case, such an organization has some relation with violence present in the relationships among them. Also, if social institutions have as one of their tenets the hierarchical division of individuals, classifying them in better or worse positions according to their social status and the so-called 'competencies.' Such a hierarchization may legitimize submission, allowing the violence of one individual over another.

In this scenario, it is necessary to remind that school plays an important role in education and can contribute to reflection and the fight against these forms of violence, and set up other directions for aggressive impulses other than domination of other human beings. One of the objectives of school education would be cultivating (self-)knowledge about violence working for a just society, therefore, consisting of individuals capable of reflecting upon themselves and the world (^{Adorno, 1995b}).

Some adaptation is necessary for coexistence, and the school contributes to this education [Bildung]. However, nowadays, such a dimension seems to have become an end in itself and not part of a process toward the education of individuals capable of producing a true consciousness. ^{Adorno considers that such a situation enables identification with the aggressor (1995b}). That would require the school to make an even greater effort to resist expressions of peer violence. Being able to express violence is, in fact, an important measure to be taken, and the school can constitute a space for that. The education provided by the school can point to a path other than the domination of one over another (^{Crochík et al., 2018}).

As mentioned before, contemporary society organizes itself by its hierarchical division. Such an organization is also revealed at school. There, occurs both the appreciation of students who adapt to the rules and respond adequately to what is expected academically - an official hierarchy - as well as the admiration of those who have a high popularity and physical strength - what ^{Adorno names unofficial hierarchy (1995a}). Such hierarchies play an important role in establishing group relations, both in the traits of identification and union among peers as in those of exclusion and stigmatization present in maltreatment situations. Maltreatment, returning to Freud's statement opening this article, can be understood as a type of conduct that expresses the ambivalence of group relationships in which love and hate are mobilized. And the school, by establishing a ranking of students based on idealized models, imposes a kind of modeling on the student in a heteronomous way (Adorno 1995c), in which it favors discrimination against those who do not fit the proposed model. This leads to an institution's consent to peer violence. Indeed, it is important to acknowledge what has been learned by some and not by others, but this does not justify appreciating some at the expense of others.

Thus, in the field of school education, it is important to consider that the rationality promoted there is not a facilitator of the symbolic psychodynamic processes necessary for the establishment of fraternal relationships. On the contrary: the proposed rationality is that of the inclusion/exclusion logic. In other words, the rationale is one of inclusion or exclusion: adding dogs to dogs and cats to cats; every dog is an animal, but not every animal is a dog. Formal thinking has many of these characteristics, inherent, according to ^{Horkheimer (1944/2015}), to subjective reason, which has an incidence in our perception of reality. These logical matrices are reflected in our interpretations to a greater or lesser extent. The hierarchical aspects are part of a cognitive scheme that allows us to access new logical forms in educational frameworks. But if they are conducive to scientific knowledge, especially in the exact and biological sciences, and to technique, as far as they reduce reality to them, as far as they do not consider themselves a product of human development but as their genesis, they eschew the very understanding of the world.

The broad spectrum of peer abuse would range from bullying - a systematic form of violence (physical or otherwise) in which one individual or group imposes himself/herself by force, placing the other in a role of submission in which he/she is prevented from reacting (^{Olweus, 1993}) - to forms of unsystematic violence in which retaliation and target rotation are present (^{Prieto Quezada, Carrillo Navarro, & Lucio López, 2015}), as well as racism and prejudice.

In this article, we define maltreatment, as did the ^{Center for Social Entrepreneurship and Administration in the Third Sector [CEAT] and the Foundation Institute of Administration [FIA] (2010}), to designate violent acts among students, whether physical, verbal, psychological or sexual; it differs from bullying since it does not return to the same target for a prolonged period. It is important to differentiate the types of violence to describe them better, to research their origins and consequences, although they may be related and bear common characteristics. In the case of bullying and maltreatment, the roles played are the same: some practice them and some who undergo them, there may be some who support them and some who are observers, and the same individual may play different roles.

In a study seeking to relate the hierarchical position occupied to the roles in bullying and school violence in general, ^{Crochík (2015}) identifies that students at the top of the unofficial hierarchy, and at the bottom of the official hierarchy, such as delimited by ^{Adorno (1995a}), there tend to be the aggressors. Those at the bottom of both hierarchies tend to be victims of school violence; those who score well tend to be neither aggressors nor victims. The confirmation of the existence of these hierarchies at school and their relationship with school violence places the school's structure as pivotal in that matter. It is up to the institution to reflect and develop measures on its role in the reproduction and even creation of social determinants that favor domination among individuals.

The intense presence of the institution of school in the lives of children and young people; the establishment of bonds between individuals' lives over there; the formation of groups among peers, and the transformation in the forms and roles played in the groups acquire fundamental importance in the education of the individual, which should be contrary to violence and focused on sensitivity; therefore, what hampers an education toward non-violence - the expressions of maltreatment - demands constant research. Thus, as noted, in this paper, we will not limit ourselves to examining only the present violence of bullying; we will start from a broader perception of peer abuse that may or may not be related to systematic bullying intimidation.

Among the researchers, we find different designations and even variation in quantity in describing the roles in bullying. We also understand that to be present in maltreatment. Most studies have in common description of the roles of aggressor and victim - sometimes called 'target.' The description of the aggressor/victim figure is also frequent. (^{Fante, 2005}; ^{Lopes Neto, 2005}; ^{Olweus, 1993}). Mixed roles, such as 'provocative victim' (^{Fante, 2005}), 'open supporter', or 'covert supporter' (^{Olweus, 1993}), are described differently according to each of the authors. It is important to note the differentiation between the types of observer. Depending on the position adopted, it is evident that such a role may also have an indirect form of participation in violence.

Deepening the investigation regarding the students who play the role of observer is also important to broaden the understanding of the phenomenon of maltreatment among peers beyond the aggressor/victim dyad. After all, it is a group phenomenon occurring with roles assigned in and by the school institution (^{Melo & Pereira, 2017}). One should note that the observer category is the largest group involved in bullying situations in several countries (^{Fante, 2005}; ^{Olweus, 1993}; ^{Santos & Kienen, 2014}; ^{Salmivalli, 2014}; ^{Mendoza-González & Maldonado Ramírez, 2017}).

The observer's figure - also called 'spectator' (^{Fante, 2005}) or 'witness' (^{Olweus, 1993}) - has different denominations, which denotes a variation in understanding their participation in the aggression. ^{Vieira, Torales, Vargas, & Oliveira (2016}), in an investigation of the attitudes of bullying bystanders among students in late elementary school and high school students in Brazil, identify observers who take no action before the aggression (in most cases), and also observers who interfere in defense of the victim or resort to an adult (on fewer occasions). Among the reasons the observers did not take any action, they identified the importance given to the social relationship between the observers and the aggressor(s), since the silence denotes veiled support of the one with whom the observer has, or wishes to have, some kind of proximity: the aggressor. The authors also found a relationship between the aggressors' motives and the observers' attitudes. These vary according to motives for aggression. Revenge and violent reactions arising from mockery and teasing were situations in which observers tended to remain passive and silent because they considered such aggression justified.

Research with Swedish students (^{Thornberg, Pozzoli, Gini, & Hong, 2017}) has shown that while most students consider bullying to be a more severe and wrong than other forms of transgression of school rules, those in the roles of aggressor and even victim (although at a different frequency than the aggressor) are less likely to understand bullying as a morally wrong action than those in the role of observer. The former also tend to show less empathy with classmates and less commitment to the school's conduct rules.

The study by ^{Melo and Pereira (2017}) is based on the distinction of types of bullying observers as established by ^{Salmivalli, Voeten and Poskiparta. (2011}). It ranges from assistants to aggressors to victim advocates, to reinforcers of aggression and passive observers. Portuguese researchers sought to understand the motivations for observers to help victims of bullying or not through the development of a moral sensitivity assessment scale. The results confirm previous research (^{Salmivalli et al., 2011}) that there would be at least five categories to be considered for allowing the observer's defense or not of the victim: the degree of perception that the aggression will be harmful to the victim, the emotions before the aggression (fear, empathy, excitement), relationships and social positions (whether or not the aggressor has prestige/popularity among peers or is a friend of the observer), moral evaluation (whether or not the aggressor considers bullying to be wrong and whether or not the victim deserves such aggression), and self-efficacy (the belief that his or her action will be effective in stopping the aggression).

^{Trindade and Menezes (2013}) research describes different dimensions of aggression among peers at school. Among them, the so-called "on the razor's edge" games and violence in a group of friends." The observer's role is strongly linked to friendly relationships, forming a kind of bond in which roles can vary rapidly within the group: sometimes victim, sometimes observer, sometimes aggressor. What seems to keep the absence of censorship before such kind of maltreatment among classmates is the fear of losing a friendship. Thus, maltreatment is seen as a 'joke' about which no one is allowed to complain, in order not to lose the friendship with members of this group. The researchers also mention other configurations in which the observer, although not bound with the aggressor by friendship, remains silent, and the aggressor interprets that as a sign of approval. On the other hand, the 'non-complicit observer', if initially, he does not keep quiet before aggression, also tends to silence when he realizes there is no support of more classmates and that he can become the next victim. The authors conclude that in situations of aggression, those involved have compromised their individual critical sense in favor of the relationship between peers and group cohesion, which attempts to endure itself by naturalizing the idea that such practices are 'jokes'.

These investigations broaden the observer's role. It is not only a figure able to reinforce systematic aggression, but also able to prevent it, depending on the position adopted before the aggression observed. The researches mentioned show that this type of aggression is not limited to personal issues between the involved. As mentioned at the beginning, group configurations and the school itself also determine the occurrence or not of the aggressions.

The ^{Mendoza González and Maldonado Ramírez's (2017}) research with Mexican primary school students on social skills and the role played in peer bullying identifies four types of participation in peer bullying: victim/aggressor in school violence, victim/aggressor in bullying, and those who are not involved in any type of maltreatment. The research identified that the emotional state and social skills vary according to the role played in violent situations. Those in the dual role of victim and aggressor, whether in bullying or other school violence, tend to be less willing to cooperate with peers and to be less empathetic.

Another study that seeks to correlate emotional and behavioral problems with roles played in violence situations among students was that of ^{Resett (2014}). It investigated students from Argentine secondary schools. The author distinguishes bullying from other forms of harassment among students and identifies three roles: victims, aggressors, and aggressors/victims. It concludes that students involved in bullying have from emotional problems (the victims and aggressors/victims) and conduct problems (the aggressors).

The last two studies cited, although they contribute to the description of the different roles present in peer violence and its assessment, strengthen the understanding of bullying and other forms of school violence as something arising from individual characteristics. That may ignore society and school's role in its origins, and consequently, in the forms of participation before the phenomenon, as understood by ^{Adorno (1995b}). It is also not possible, in this type of approach centered only on aggressor/victim figures, to investigate alternation in roles played (aggressor, victim, or both), since personality characteristics are associated with specific forms of school violence. This prevents the perception of changes in relationships between students over the years.

Regarding the type of aggression practiced, the study by ^{Santos and Kienen (2014}) on the perception of bullying by Brazilian students and teachers identified that students perceive bullying more easily when physical aggression occurs, such as kicking and pushing. The same is true of verbal aggression. Indirect forms of violence such as exclusion and gossip, although they also occur, they are identified as an aggression to a lesser extent by students. The authors conclude that the most subtle violence practices may be the ones causing the most serious harm precisely because they are not identified and tackled by the school community.

As for the higher or lower incidence of violence as years of education vary, the highest risk of occurrence has been identified within primary education in both Brazilian and European research (^{Lopes Neto, 2005}; ^{Olweus, 1993}). The study by ^{Mendoza González and Maldonado Ramírez (2017}) concludes that elementary school students would be at greater risk of victimization, what corroborates that view. ^{Vieira et al. (2016}) identified a higher incidence of bullying among the youngest (62% between 13 and 14 years versus 39% in the 17-18 age group) but without significant difference between genders. However, perhaps what happens is much more a change in maltreatment forms to more indirect and subtle expressions than an effective decrease in violence. For instance, the study by ^{Prieto Quezada et al. (2015}) with Mexican university students identified significant rates of maltreatment at that education level: exclusion and isolation (more than 30%), verbal violence - nicknames and cursing - (more than 20%), and the use of their belongings without permission (more than 10%). Such results, in addition to pointing to the continuity of bullying in later years, also point out a paradox for the authors, since it would be expected that a longer period of education would result in "[...] a better strategy to relate to peers" (^{Prieto Quezada et al., 2015}, p. 46).

Considering what has been presented so far, the objectives of this article are: to verify if there has been a variation in the type of aggression between the primary and secondary education levels, as defined in Argentina²; to examine the roles played by students concerning maltreatment, to verify whether or not these roles are maintained, as well as to verify whether there is a relationship between the types of roles, i.e., whether there would be a tendency to play a role in one of the educational levels and to assume another at the other level, such as being the perpetrator of aggression in elementary school and the supporter in secondary education, being a victim at one level and an observer at the other.

Method

Analysis and discussion of results

The results and their respective analyses will be presented in the same order as the objectives. The first set refers to the verification of the maintenance of the frequencies of the type of violence in the two levels of education for the different types of roles involved.

In the following tables, the frequencies and proportions of each type of aggression are presented, by the level of education, with reference to the roles played in these peer aggressions.

According to the data in Table 1, memories of aggression in secondary school are more frequent than in elementary school, which contradicts several studies' research (^{Lopes Neto, 2005}; ^{Olweus, 1993}). But it must be emphasized that this study deals with memories, not occurrences. Older memories resist being remembered more, as psychoanalysis shows us; thus, the data obtained here do not challenge these authors but only illustrate how conscious we are or not of them. For example, ^{Prieto Quezada et al. 2015}) show the persistence of the aggressions even in secondary education.

One can see in Table 1 that insulting is the most frequent aggression remembered in both secondary and primary education; threats of physical aggression, gossip, and exclusion are the most prevalent. It should be noted that ^{Santos and Kienen (2014}) indicate that the most visible aggressions are the most well-noticed, and in the case of this study, the most remembred. The correlation obtained between the frequencies of aggression remembered in the two education levels was significant (r = 0.92; 8g.lib; p < 0.01). That means that the practices of aggression most remembered in one level of education were also the most remembered in the other, and vice versa.

It is important to highlight from the data that the type of maltreatment tends not to change from primary to secondary education. That would be expected because, as school education progresses, students might have other forms to express violence: from the most direct to the most symbolic, even if the latter is still violence, it can be challenged verbally, while physical violence also needs more immediate action to be stopped. In other words, it is not a matter of saying that violence would be less severe if it changed its expression form, but that the form to confront it could be different.

Table 2 presents these data for the role of the supporter.

Memories of support for various types of aggression are far less frequent than practices of aggression. But just so, there are more memories from secondary school than from primary school. In secondary school, the most frequent ones coincide with those of the practices. Still, it draws attention that in elementary school, the memory of having supported the exclusion of classmates is the most frequent. The correlation between the frequencies of support memories at the two levels of education was not significant (r = 0.62; 8g. of lib; p > 0.05). That means that the type of support remembered was sometimes more frequent at one level of education, sometimes at the other education level. The fact that the memories of supporting are less frequent than those of maltreatment authorship may indicate that it had less meaning for these students than the former; in this case, it is not a matter of forgetfulness but of leaving fewer marks on the individual.

Table 3 shows the frequency data of these types of aggression concerning the observations remembered.

According to the data in Table 3, the memories of the aggressions observed are more frequent than those of practice or support, following several studies (^{Fante, 2005}; ^{Olweus, 1993}; ^{Santos & Kienen, 2014}; ^{Salmivalli, 2014}; ^{Mendoza-González & Maldonado Ramírez, 2017}). In secondary education, the first seven were noted by more than half of the sample; they are also frequently remembered by several primary education students. The frequencies of memories of observations at both levels were significantly correlated (r = 0.98; 8g.lib; p < 0.01), i.e., the more one type of aggression was remembered at one level, the more another was also remembered at the other level.

Table 4 contains these data for the aggression underwent.

As for the remembrance of undergoing the various types of aggression, in both types of teaching, it is noted that insults are more frequent, but threats of aggression and physical aggression, gossip, and exclusion are also remembered. The frequencies of the memories of the maltreatment undergone are significantly correlated (r = 0.91; 8 g.lib; p < 0.01), that is, the more one remembers having experienced a type of aggression in secondary school, the more one remembers the same form of violence in elementary school.

In short, regarding the frequency of aggressions, it ought to be considered that:

- There are more memories in secondary school than in primary school;

- There are more memories of observing than practicing, supporting, and undergoing aggression.

- In all cases, insult is the most frequently remembered aggression, but aggression, the threat of aggression, exclusion, gossip, and nicknames are also very frequent.

- There are significant correlations between the frequencies of memories in the two education levels concerning practice, observation, and being abused.

So far, one can conclude that peer aggression is quite frequent, which is something to be considered by educators, and that the most frequent types of violence at the primary level tend to remain at the secondary level. Thus, it can be said that there was no variation of the type of violence in both levels of education in three of the four roles under consideration.

This study's second objective concerns keeping or not of the roles in maltreatment in the two levels of education researched. The third objective is to assess the relationship between the roles in the levels of education investigated. The following tables bring data to verify these objectives.

Table 5 contains the means and standard deviations and the minimum and maximum values obtained from participants regarding the four roles assumed in the aggressions between pairs, for secondary and primary education.

Regarding the data in Table 5, it can be considered that in all the roles examined - aggressor, supporter, observer, and target - some did not participate in any of them. Some reached the maximum in practices and observations. The variability among participants was appreciable. It can also be noted that the highest means are those of observation in both secondary and primary education, which corroborates the result found concerning the last objective. However, one has to bear in mind that they are not high in general. That does not mean they are less severe, though. These data reveal that many students do not participate in the maltreatment in any of the roles. Others are little involved. Thus, at least partially, there are not solely violent students, neither only those who undergo school violence.

Table 6 shows the correlations between the various types of roles assumed in maltreatment among primary and secondary education peers.

Considering only the significant correlations to 0.001, pointed out in the table above, it can be verified that there was a significant relationship between the memories of aggression in the two levels of education: i.e., one who tends to be an aggressor during primary education tends to continue to be an aggressor during secondary education. The same can be stated regarding supporting of aggression, the observation of aggression, and the aggression underwent; this strengthens the previous considerations.

One may also note from the data in Table 6 that the practice of aggression and support of aggression have a significant relationship at the two levels of education. However, these are of lesser magnitude than the correlations obtained between the same roles at the two levels. It implies that, if there is a relationship between being an author of aggression at the primary level and a supporter at the other level (r = 0.44), such a relationship is less significant than in being an author of aggression at the two levels of education (r = 0.69). The memory of aggression undergone at both levels tends to relate to both forms of observation, as per data in Table 6. The memory of observations in secondary education, though, relates to being targeted during secondary education; and observation memories in primary education relate to undergoing them in primary education, and to a lesser extent, to undergoing them in secondary education; i.e., having more observation memories in primary correlates with being a victim at that one level of education, but not at the other level. The case is the same regarding observer memories in secondary education, which were only correlated with being targeted at that one level of education.

To confirm these data, a factor analysis was calculated using the Principal Component method, Varimax rotation (KMO): 0,593; X2 = 214,254; 28 g. lib; p < 0,001), involving such variables. The factor loadings obtained are in Table 7.

Table 7 shows the confirmation of most of the observations extracted from the correlations in Table 6: 1- there was the maintenance of roles from primary to secondary education, except for observation; 2- whoever is a supporter or aggressor in primary education tends to be an aggressor in secondary education; 3- whoever is a supporter tends to be an aggressor and vice-versa; 4- whoever suffers aggression in primary education tends to suffer it in secondary education; 5- whoever observed aggression in primary education tends to suffer more aggression in both levels; and observations of aggression in secondary education did not associate with any of the other variables

Regarding our objectives, there was a tendency to keep the same roles at both levels of education. Being a supporter and author of aggression are associated. Still, the same cannot be said of the relationship between being an observer and a victim of violence at both education levels since this relationship tends to occur only at the same level of education.

Final considerations

Regarding the found results, it should be pointed out that school education has not been acting to its fullest in reducing peer violence, neither in its frequency nor in its form. That reinforces, quite a while later, ^{Adorno's (1995}b) statement that education has failed, at least partially, in its objective of de-barbarizing individuals. Even worse, if in the mid-1960s, he could propose an education against violence, it either did not succeed or was not sufficient. It seems that there has only been the establishment of an environment contrary to barbarism, but in coexistence with the environment conducive to violence.

From what the research data indicate, the formation of groups that are targets of aggression perhaps occurs to enable the union of the group itself. That depicts the narcissism of minimal differences (^{Freud, 2011}) as continuing to be necessary, as well as the intention to end all forms of tension through violence - the death drive. The maintenance of aggression from six to 18 years of age, even more frequently of physical violence, points out that the possibilities of civilized relations, conducive to respect for others and their rights, especially that of not being harassed, remain. The objective conditions that led to Auschwitz being supported by a multitude of pseudo-educated individuals seem to persist. It could be argued that maltreatment and reaction to it are part of 'healthy living', something natural to men, but that would be to reverse the terms of a pacified humanity, which Kant was able to propose in his Perpetual Peace, and which ^{Freud (2011}) can glimpse in his critique of the religious precepts of loving everyone, including and especially enemies: one cannot love everyone, but if others respect us, we can repay them the same respect, and this we could learn during our school years.

As a limitation of this research, it is important to highlight that we were able to collect the memories only of those who entered higher education; a study with young people who attended primary and secondary levels but who are not in higher education anymore could bring complementary data.