Introduction

Understanding the social responsibility the academy has to contribute to the prevention of violence against women, the Universidad Central’s (Colombia) Program of Social Communication conducted a research based on a comprehensive monitoring of written press in 20 national and local Colombian newspapers on cases of femicide and attempted femicide published between March, 2015 and December, 2017. The sample was collected to analyze the role played by the media in the construction of the story about femicide in Colombia, and to establish whether there has been an appropriation of Law 1761 of 2015, which criminalizes femicide in the country.

To this end, a conceptualization of femicide was made, not only from the legal perspective of Colombian law, but also from feminist theories. Regarding the concept of femicide, the researchers ^{Radford and Russell (1992)}, as pioneers in the discussion, understood femicide as “the murder of women by men motivated by hatred, contempt, pleasure, or a sense of ownership of women” (p. 23).

In Latin America, the Mexican anthropologist ^{Lagarde (2005)} enriches the discussion by arguing that this is a State crime, since it is the State that is responsible for preventing and attacking violence against women. This concept was relevant in the discussion on the responsibility of the State in crime prevention.

Likewise, the Argentine theorist ^{Segato (2012)} speaks of the:

Importance of typifying the various types of violence against women, making the difference between crimes that can be personalized, i.e., interpreted on the basis of interpersonal relationships or personal motives on the part of the perpetrator, and those that cannot be (p. 1).

According to the author, there is a special interest in keeping crimes against women in the private sphere and therefore in making them invisible in terms of social and State responsibility.

This present research was based on the classification made by the Mexican sociologist ^{Monárrez Fragoso (2005)} on the types of femicide: intimate femicide, which is subdivided into child and family femicide; systemic sexual femicide, which covers the murders of girls and young women who have been kidnapped, raped, tortured, mutilated and whose corpses are deposited in deserted lots or vacant sites (p. 167). Likewise, the typification made by the feminist lawyer Isabela Agatón, who promoted in Colombia the anti-femicide Law 1761, was also taken into account. ^{Agatón (2017)}, like Monárrez Fragoso, distinguishes systemic sexual femicide, and enriches the concept with the typification of “sexualized femicide: that crime committed because of prejudice related to sexual orientation; and aggravated femicide; when the crime is committed after the sexual aggression” (p. 47).

Other important theoretical references for the development of this research were taken from ^{Segato (2018)}, mentioned above, who develops the concept of the masculinity mandate linked to the pedagogy of cruelty as a form of male enlightenment towards women through the use of violence; a fundamental concept to think about the prevention of violence from the media and the need for the stories to be built without disclosing the logics of violence that are carried out from such pedagogy.

The sample was subjected to a content analysis that showed that the predominant journalistic genre was news pieces with 79% of the total sample, followed by news briefs with 14%, leaving genres such as chronicle and news stories in very low percentages. Although the sample was made up of both local and national newspapers, it was the local newspapers that contributed significantly to the sample since they are the ones that publish the most cases of femicide.

Regarding the characterization of the victims, it was decided to analyze the majority category: women of legal age (84.2%), victims of femicide or attempted intimate femicide. That is, femicide which is committed by a relative of the victim, including the partner or ex-partner (^{Monárrez Fragoso, 2005}).

Based on the results of the content analysis, we proceeded to identify the main theme or predominant macrostructure of the sample, understanding the macrostructure as the structure of meaning of a discourse, which is responsible for defining the core that gives meaning, or not, to the discourse (^{van Dijk, 2003, p. 4}). Considering the delimitation of the sample, it became evident that the macrostructure of the analyzed pieces is not the understanding of femicide as a social phenomenon, nor the dissemination of the Rosa Elvira Cely (anti-femicide) Law. It is the Myth of Romantic Love, understood from the concept of ^{Bosch et al. (2007)}:

To make the other person the only and fundamental aspect of existence; to live very intense experiences of happiness or suffering. To depend on the other person and to adapt oneself to him or her, postponing one’s own. Forgiving and justifying everything in the name of love (p. 135).

This myth is exposed in a non-obvious way as an excuse for femicides, and trivializes, through what is believed to be love, violence against women and its fatal outcome in femicide.

Methods

The research was conducted under a mixed methodological approach, based on quantitative and qualitative analyses. The creation and subsequent development of the methodological design was divided into four phases.

The first phase was based on a documentary research on the concept of femicide; this research considered the guidelines of the Rosa Elvira Cely law (anti-femicide), its most important characteristics and its limitations according to activists of the cause. This phase sought that the entire research team, composed of undergraduate students, would be able to identify when a case of femicide or attempted femicide occurs.

Continuing with the documentary research tool, in the second phase of the project, during three months, students² and the teacher leader of the research monitored 20 newspapers in the newspaper library of the National Library. Among the newspapers were: El Colombiano, El Heraldo, El Universal, Diario del Huila, El País de Cali, La Patria, El Nuevo Día, Vanguardia Liberal, El Liberal de Popayán, El Diario del Llano, Boyacá 7 Días, Llano 7 días, Chocó 7 Días and the newspapers Qhubo de Cali, Bogotá, Pereira and Medellín. The newspapers El Tiempo and El Espectador were monitored in their online version, as well as the national news portal HSB, which publishes information from several popular newspapers including El Extra, and other online media were monitored by volunteers from the campaign No es hora de callar. The sample was taken from May, 2015 to December, 2017, identifying 1 011 journalistic products on cases of femicide and attempted femicide, plus 117 products that responded to the case of the girl victim of kidnapping, torture, rape and aggravated femicide, Yuliana Samboní, which were not taken into account in the content analysis as they multiplied the data.

In the third phase of the methodological design, an information analysis matrix was constructed which, at the same time, served to systematize the corpus of information. The matrix took into account the following variables: name of the victim and perpetrator (or aliases), age of the victim, social stratum, ethnicity, mother or not, city, femicide or attempt, whether the perpetrator was a family member or partner or others, evidence of re-victimization by the media, place where the events occurred, weapon used, type of violence used pre-femicide and/or in the femicide, suicide or attempted suicide of the victimizer, section of the newspaper, the words most used by the media, use or not of the word femicide, what type of motive was used or implied by the media, headline, among other emerging variables. The information systematized in the matrix was subjected to a content analysis, which was carried out using the free software R version 3.6.1, and the package quanteda (^{Benoit et al., 2018}), obtaining quantitative results.

The sample showed that the predominant journalistic genre was the news piece with 79% of the total sample, followed by news briefs with 14%, and that most of the products were contributed by local, small, and popular media. Regarding the characterization of the victims, it was decided to analyze most cases: women of legal age (84.2%). Minor victims have a percentage of 10.7% of the total sample, a percentage that for the purposes of this article was not considered because their cases have different discursive and narrative structures than those of adult victims.

Once the sample was delimited, we moved on to the fourth phase of the research, which sought to identify the macrostructures, i.e., the most common themes of the products analyzed. The research made use of the analytical perspective of the Critical Discourse Analysis (CDA) by van Dijk. Although CDA does not have a designed methodology, its creator suggests the analysis of textual linguistic elements that help to identify in a discourse the relevant information proposed by the text.

For the analysis of the sample, the project was based on one of the most important textual macrorules proposed by ^{van Dijk (1995)}: selection, which consists of paying attention only to the parts of the text that are relevant to understand its general meaning. It is necessary to clarify that a strict linguistic analysis was not carried out, as is the case with discourse analysis, which differs from CDA in that, in the latter, multidisciplinarity is fundamental to understand the contexts (^{van Dijk, 2003, p. 4}).

Results

The research corpus included 1 011 journalistic products on cases of attempted femicide and femicide. Some of the results of the content analysis are presented:

Cases of femicide and attempted femicide: 18.8% corresponded to cases of attempted femicide, while 81.2% of the sample corresponded to cases of femicide.

Victimizer: In the majority of femicides the perpetrator was the victim’s partner or ex-partner, representing 61.8% of the total sample; the family perpetrator, including father, stepfather, uncle, cousin, among others, represented 20.4% of the sample; and the percentage of products for which no information was provided was 17.8%.

Suicide or attempted suicide of the victimizer: no was an 88.9% of the total sample, and yes was 11.1%.

Variables that account for the role-played by the media in reporting cases of femicide:

Use of the term femicide: 60.7% of the publications did not use the term “femicide” or “attempted femicide” and 39.3% of the sample did use the terms.

Type of source: 34.8% of the total sample responded to a judicial source, 28.9% to police, 14.8% to family, 8.9% to neighbors, 8.2% to others, and 4.4% to experts in the field.

Newspaper section: 58% of the total sample published the product in the judicial section, others were a 25%, and 17% in the daily news section.

Motivation for femicide according to the publication: 32.7% of the total sample reported no motivation for femicide, 28.9% reported abandonment of the victim to the perpetrator as motivation, 26.2% reported jealousy, and 12.20% of the sample reported other types of motivation.

Re-victimizing elements in the publication: in 46.4% of the total sample various forms of re-victimization were found, yellow journalism in 21.6%, no re-victimization was found in 10.9%, biased information was identified in 6.6%, re-victimizing interviews in 5.1%, violent photographs in 3.8%, use of wrong terms to refer to femicide such as “crime of passion” in 3.6%, and others in 2%.

Most used words in the publications: 34.8% of the sample presented the combination of words “abandonment”, “jealousy”, “love”, and “infidelity”; 32% could not be identified; “crime of passion” appeared in 16.4%; “jealousy” in 9.7%: “abandonment” in 3%; “love” in 2.4%; and “suicide” in 1.7%.

Use of the term femicide in publications by year: in 2015, 23.3% used the term and 74.7% did not; in 2016 33.6% used it and 68.4% did not; in 2017 51.9% used it and not in 18.1%.

Journalistic genre: 79% news pieces, 14% news briefs, 3% chronicle, 2% story, and 2% others.

Critical Discourse Analysis of Journalistic Products

Every text has a macrostructure that represents its meaning; thus, for a text to be understood as a unit of communication, it must have a fundamental informative core; that is, a primary theme (^{van Dijk, 2003, p. 8}).

The textual macrostructure that one would expect to find in news products on cases of femicide would be that of femicide treated as a social problem, its causes, its consequences, and the dissemination of the anti-femicide law in Colombia. However, the content analysis conducted on 1 011 notes on cases of femicide in different Colombian newspapers evidenced that 60.7% of the analyzed news items did not use the term, and that, even though from 2015 to 2017 the use of the concept in the products increased by 23.3%, the use of the term was made without explaining the anti-femicide law in Colombia or the social problematic that frames the phenomenon. Although in the totality of the sample a murder is mentioned, there is no special emphasis that the murder responds to the typology of femicide, and when it is done, there is no mention of the specific characteristics that make it a gender-based crime; demonstrating that the issue of femicide in its full dimension is not part of the macrostructures of the sample.

The fact that in most of the sample analyzed the concept of femicide or attempted femicide was not used, that when it was used it was done without explaining its meaning, and that, on the contrary, evident importance was given to issues related to the personal lives of the victims, their affective relationships linked to transgressions of female gender roles, such as abandonment or the role of the victim in her love relationships, suggests that one of the most relevant macrostructures -which does not mean that there are not others- is linked to the Myth of Romantic Love.

To contrast the above hypothesis, it was decided to use the analytical perspective provided by the CDA. It should be made clear that an analysis of the entire sample is an impossible task to accomplish, so methodologically it was decided to analyze by means of examples that are part of the research corpus and the results of the content analysis, one of the most important textual macrorules that, according to ^{van Dijk (2003)}, are indispensable for the identification of the macrostructure of any text: selection.

Romantic Love and the Naturalization of Violence

The concept of gender is understood as a social construct and it differentiates from sex, as explained by ^{Lagarde (1994)}:

We have physiological and sexual characteristics with which we are born, which are natural and do not change (at least not naturally). We call this “sex”. And, on the other hand, we differ, because each society, each culture has given a different value and meaning to these sex differences and has developed ideas, conceptions, and practices about being a man and being a woman. This set of social, economic, political, cultural, psychological, legal characteristics and norms, assigned to each sex differentially, is what is called gender (p. 24).

This helps us to understand how the processes of socialization differentiated between boys and girls have nurtured female stereotypes linked to maternal care, family care, care of the elderly and the sick, among others, which for generations prevented women from entering the public sphere of social and economic recognition. In addition, as analyzed by ^{Burin (1996)}, it is evident that while men constructed their identity as economic providers, women constructed their identity as providers of affection “... for the men, rational power; for the women, the power of affection” (p. 85).

The power of affection has made women privilege affective relationships before any other type of relationship, which has generated a feminine identity subordinated to the idea of feelings. In this construction arises the Myth of Romantic Love.

For centuries, the Myth of Romantic Love has been disseminated from literature to current film productions, ceasing to be a private reference of a desirable way of life, for a globalized desire of what is expected to be affective relationships. Journalistic products, no matter how objective they try to be, do not escape from this logic.

Considering that, in the results of the content analysis presented above, it was evident that 62% of the perpetrators out of 100% of the sample were the victim’s partner or ex-partner, the myth of Romantic Love became a constant reference in the textual structures examined.

From the macroregulation of selection to the macrostructure of the Romantic Love Myth

Next, we present elements identified in the analyzed sample that respond to the textual macrorules of selection proposed by the ACD, making an argumentative path of why the Myth of Romantic Love has been placed as one of the most important macrostructures of the analyzed corpus.

Selection

Conclusions

Although the results of the content analysis show that the use of the term “femicide” has increased from 25.3% in 2015 to 51.9% in 2017, it is not possible to say that in the analyzed sample there is an appropriation of the anti-femicide law in Colombia. This, because in the total of the research corpus, from 2015 to 2017, the non-use of the term is still higher, and because, although the concept is used, it is not accompanied by the explanation of the law or the understanding of the serious dimension of the problem. On the contrary, the term is being used in an empty way, and even, it could be said predictably, it will end up being a yellow journalism element to describe this type of murders.

Although in the national media there is a growing interest in communicating about violence against women with a responsible gender approach -without this meaning that it is always achieved-, in the vast majority of small or popular media, which publish much more frequently cases of femicide than the national media, and therefore were more representative in the corpus of the research, femicides are assumed as daily news without showing an interest in seeking the causes of this social phenomenon. On the contrary, the cases are published repeatedly, as if they were all the same, with the same re-victimizing elements, with lurid details about the facts and the private life of the victim, with elements that implicitly justify the crime through the stereotyping of love, jealousy, abandonment, and infidelity as possible motives for the murder.

Popular newspapers have a great power of influence on the imaginaries of their readers and therefore on what is understood by the crime of femicide, and it is these newspapers that should be influenced with greater certainty, seeking to modify the stories under the understanding that a femicide can never be assumed from a love story nor can it be justified from the rupture of the social role assigned to women.

As evidenced, State pronouncements in relation to femicide are almost nonexistent, and therefore their possible culpability for omission in prevention (^{Lagarde, 2005}), the administration of justice and the monitoring of the situation of the orphans of femicide (a subject that is not addressed in any academic or institutional text), should be the subject of critical analysis. This analysis must consider the value that the loss of a life represents for society in general, taking into account that the victims of femicides are not only the women who die, but also their families and especially their children who have to grow up without a mother and often with a murderer father deprived of liberty or on the run. The State must take on this problem effectively, not only by enacting laws that protect women from being murdered because they are women, but also by generating and supporting education and communication processes with a gender perspective to prevent this violence.

Likewise, it is necessary to work on the role that the masculinity mandate plays in the concurrence of the crime; this means that journalism must understand that gender roles have a very strong influence on femicides and that these logics should not be replicated or fed in journalistic accounts.

The naturalization of violence against women of legal age, in the products analyzed, is evidenced in the harmful link made between violence and love and, therefore, its consequences. This burden can only be disarticulated when, socially -with the help of responsible media when addressing these issues-, femicide is not treated as a private matter, but as a social problem. The systematic and non-reflective dissemination of the facts of violence against women, its use as a commercial element, its naturalization, and therefore its tacit acceptance in societies such as the Colombian one, is not a minor problem, it is a priority issue when preventing any type of violence against women and, of course, femicide.