Women, Crime and Gender in the Private Sphere: Femminicidio

Abstract

Exploring the historical gendered fixed roles for women and men in the private, i.e. heterosexual relationships, demonstrates that these prove equally detrimental to women. The femminicidio, the crime in which men kill their on-going or former partners, is a widespread phenomenon worldwide as well as in Italy. Qualitative and quantitative language investigations show that politics is dealing with this issue as an abstract phenomenon which affects women. Reporting on the topic in Italian newspapers reproduces dangerous narratives which recount how male jealousy and hurt feelings are at the core of this gendered crime. Furthermore, specific language is used to bridge the gap between a void in the legislation and this systematic atrocity perpetrated by men against women.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

In memory of Sara and all victims of femminicidio

In the introduction, I explained that this volume is divided into two parts, connected by a shared focus, the investigation of how, to various extents, gender operates in Italy, explored with same methods (e.g. corpus linguistics) and theoretical underpinnings (ideology , discourse). To introduce this section, I move the focus from the public sphere —which has been central to my investigation of language used to refer to, insult, and construct female politicians , as explored in Chapter 4—to gender in the private sphere . One cannot say that there is a clear-cut division between the two domains, because women have been historically positioned in the private sphere , and they are often misconceived when they attempt to move into the institutional, public arena. As argued in the previous chapter, attempts to reposition women and to return them to the private space, have been undertaken with the implication that this is where they naturally belong.

I here conceive of the ‘private ’ according to two different perspectives: specifically, on the one side, the space to which women have been assigned, and, on the other, the dimension of private feelings in (heterosexual) relationships. One cannot disregard that both—the space and the emotional dimension—have created gendered roles, roles that are constantly reinforced in our society (through daily practices as well as products of the media , Larcombe 2005). Romantic love is habitually seen as a universal experience and a “human staple” (Monckton-Smith 2012, p. 43). However, love has been used in society to also express an acceptance of violence and to “provoke[e] behaviours and actions which are extreme” (Monckton-Smith 2012, p. 43).

Heterosexual love and relationships are part of institutionalised discursive practices and, in them, men and women are given roles which could be seen as strict and hierarchised: to some extent, there seems to be an agreement that men are in charge and women are subservient (Monckton-Smith 2012; Ingraham 2006). Starting from this, I here explore a gendered crime , labelled as femminicidio . The term relates to women killed by their ongoing or former partners, husbands or boyfriends. I will show how the parliament and the media have reported femminicidio so as to link it to, and to reinforce and reproduce, the accepted position of women (and in binary opposition, the accepted position of men) within heterosexual relations.

I first review the existing interdisciplinary literature on Italy’s female-oriented private sphere and then move to present data and information about femminicidio in Italy. With this in mind, I here give a general overview of the methodology of the analyses that follows. These concern parliamentary acts (mozioni) from the period 2013 to 2017, newspaper reports from the period 2013 to 2016, and an XML mark-up investigation of a specific femminicidio, that of Sara Di Pietrantonio .

A Gendered Private Sphere

In Italy, women have been historically portrayed and constructed as stay-at-home mothers and wives, where the only public concessions given were to those in the roles of teachers and carers and, as presented before, a career in the entertainment industry (with the aim to please the gaze of men). This also forms part of the fascist era representation, when women were celebrated as sposa e madre esemplare (bride and model mother, Dittrich-Johansen 1995, p. 812) through Mussolini’s speeches ¹ and reproduced in cultural artefacts such as advertising. In an interview given in 1932 (Emil 1970, p. 32), Prime Minister of the Kingdom of Italy, Benito Mussolini (also known as Dux, in power from 1923 to 1943) states:

La donna deve obbedire. […] Essa è analitica, non sintetica. Ha forse mai fatto dell’architettura in tutti questi secoli? Le dica di costruirmi una capanna, non dico un tempio! Non lo può! Essa è estranea all’architettura, che è la sintesi di tutte le arti, e ciò è un simbolo del suo destino. La mia opinione della sua parte nello Stato è in opposizione ad ogni femminismo. Naturalmente essa non dev’essere una schiava, ma se io le concedessi il diritto elettorale, mi si deriderebbe. Nel nostro Stato essa non deve contare.

A woman must obey. […] She is analytic, not synthetic. Has she ever done architectural work in all these centuries? Ask her to build a shed for me, I am not saying a temple! She cannot! She is extraneous to architecture, that is the synthesis of all arts, and therefore this is a sign of her destiny. My opinion about her role in the State is in opposition to [the one promoted by] feminism. Naturally, she does not have to be a slave, but if I had to give her the right to vote, people would ridicule me. In our state, she must not count.

More than 80 years have passed since this interview and, in Italy, women were only given the right to vote in 1945. However, as shown in the previous chapters, some remnants remain as for what professional and non-professional environments women are deemed more suitable. This introduction advocates that historical facts should never be disregarded in relation to how women were framed and, to various degrees, continue to be so in present society. ²

I commence with a study of a law that appeared in the ventennio fascista (the two fascist decades) and de-construct it through a study of the extensive literature on violence against women. It was only on 5 September 1981 that the Republic of Italy changed the so-called delitto d’onore (honour killing [law]), this being part of the Codice Rocco. The Codice Rocco, named after the Minister of Justice in the Mussolini government, Arturo Rocco, is formed of the Codice penale italiano (penal code where laws on the subjects are collected) and the Codice di procedura penale italiano (code of penal proceeding, namely the code that includes the laws which regulate penal trials in Italy), both approved in the 1930s. In one of the sections, the code deals with honour killing, marking the difference between murder caused by offence to honour and murder for other reasons (as regulated under existing laws on murder). In relation to the former, the law reads:

Chiunque cagiona la morte del coniuge, della figlia o della sorella, nell’atto in cui ne scopre la illegittima relazione carnale e nello stato d’ira determinato dall’offesa recata all’onor suo o della famiglia, è punito con la reclusione da tre a sette anni. Alla stessa pena soggiace chi, nelle dette circostanze, cagiona la morte della persona che sia in illegittima relazione carnale col coniuge, con la figlia o con la sorella.

Whoever causes the death of the spouse, the daughter or the sister, is punished with a 3/7-year sentence in prison if the accused had learnt that the victims were being in an illegitimate sexual relationship and if he has acted in the state of anger determined by the offence bore to his honour or to that of his family. The same sentence is given to whom, in these circumstances, causes the death of the person who is in the illegitimate sexual relationship with the spouse, the daughter or the sister.

Articolo codice penale 587

This article of the penal code is a self-explanatory example of the political structure in respect of its institutionalised attitudes towards women, and one can assume that, while the law was abrogated, the understanding of honour to be attributed to men remained in the imaginary for longer, as the systematic analysis below shows. In relation to the gendered language used, it is interesting to note that while daughter and sister (figlia and sorella) are in their feminine forms—discursively constructing that only women can bring shame to a man or his family—coniuge (spouse, an epicene noun) is preceded by the masculine article (attached to the preposition di, di + il  = del, of, of+the = of the). While a certain ambiguity can be seen in this—where both men and women can be victim of a spouse—no ambiguity is offered in the case of a sister or a daughter. This seems to be one of those cases where, as Wykes put it, the law, through language, “may, through the symbolic legitimation of masculine violence, culturally support male violence” (1995, p. 55).

I mentioned that this law was changed in 1981 and while it could be considered as a victory (as honour was no longer justified as a defence in murder cases), male violence towards family members did not stop. Of course, this is not to say that violence is not committed by women. However, as has been widely researched, there is no evidence of a similar female version of the phenomenon as that which sees violent men perpetrating systematic atrocities against women. The United Nations ³ firmly asserts that violence against women is a violation of human rights, stressing the structural powerless role that women seem to have in the world. In their facts and figure list, the UN suggests that “35 per cent of women worldwide have experienced either physical and/or sexual intimate partner violence or sexual violence by a non-partner at some point in their lives. However, some national studies show that up to 70 per cent of women have experienced physical and/or sexual violence from an intimate partner in their lifetime” (United Nations 2017 ⁴ ).

It is safe to conclude that gendered violence exists and is perpetrated in relationships. Among the scholarly work on this subject, Monckton-Smith describes how gender is intertwined with what she refers to as ‘intimate homicide’, suggesting that this “is a gendered problem, not just in the sense that it is mainly perpetrated against women, or that the motivations to kill intimates of men and women are usually vastly different, but that gender belief systems are crucial in understanding and deconstructing what happened and why” (2012, p. 2). Therefore, the debate around the social and cultural, rather than the biological, position of men is a paramount aspect. From the same perspective, Dobash and Dobash (1998) found that women are punished for failing to meet men’s expectations in relation to their physical, sexual or emotional needs (1998, p. 359). Wood (2001) argues that men who do not feel that they are conforming to the traditional idea of masculinity can be prone to re-establish the connection with manhood by controlling others (more specifically those who are physically less strong than them). There seems to be a wide agreement on this topic, for instance, Schrock and Padavic (2007, p. 628) point out that “men who harm women often do so when their sense of traditional manhood – such as breadwinner or having women meet their often-unspoken needs – is threatened”. The notion of seeing their manhood threatened is also discussed by Monckton-Smith who contends that “there is a logic which says that where a man has been provoked by infidelity, desertion or even failure to observe gender roles, then there is the potential danger that he will respond with fatal violence. The same is not believed of women” (2012, p. 3). In line with this, Boonzaier (2008) supports the view that male violence is strictly connected with men accomplishing or doing gender (or rather masculinity); in this, men’s sense of insecurity and powerlessness leads to a gender crisis and “women become the targets of the man’s attempt to re-assert his masculinity/identity” (2008, p. 201).

In Chapter 1, I provided a definition of patriarchy based on the seminal work of Walby (1990); in relation to violence, she suggests that this is related to how men, particularly, have operated and continue to do so in society, e.g. prone to settle disputes with violence, brought up to be macho (Walby 1990, p. 134). She also discusses how, in relation to the criminal justice system, the state has tended to disregard women from an economic and welfare point of view. On the same topic, Hearn (2004) agrees that, structurally, men are dominant in several spheres of life and are largely unchallenged on this.

Ideas around the social positioning of men as a gendered group are also expressed by Wood who investigates women’s accounts of the violence they were subjected to. She convincingly argues that “widespread violence defies individualistic explanations” (2001, p. 241) and that culture is to be taken into consideration as it brings with it a set of values and meanings. However, in a paper exploring ‘citizen journalism’ (here intended as people who contribute unsolicited comments to news on the internet), Bou-Franch (2013) found that those commentators who initially supported male violence through, for instance, minimising the seriousness of domestic abuse, and/or by putting blame on the victims, eventually came to consider each case on its own merits.

There is a connection to be made here between internalised, publicly defined gendered prototypes and the violence these behaviours might produce in the private (the private sphere seen as a setting where happiness and safety are at the core of the relationship among those who dwell in it, Pahl 2016)—and the phenomena of violence against women as a public terrain, where the state as well as society have to intervene. The prevailing perspective on violence against women at the intersection between public and private , as presented in this section, is in line with that argued in previous chapters: an imbalanced gender(ed) world where men are in control. When men lose this control, they often seek to regain it by placing the ‘blame’ on women: professional women such as those in politics , and female victims (or survivors) in relation to violence.

The shifting of gender and violence from the private space where it occurs, to the public space where it is discussed, should also be seen in relation to representations of it in the media. Such depictions act as a means by which to reproduce an institutional power imbalance between men and women within both the private and public spheres (see also Wykes 1995). Tabbert suggests that the analysis of how crime is reported in the media “provide[s] a valuable insight into how society views crime” (2016, p. 3); the media construct the news and with it they rebuild realities (and truths as intended in the relation between language and discourse ). More importantly, what appears in the newspapers is, presumably, what is editorially seen to be of interest to the reader (Tabbert 2016; Busá 2014).

Having laid the foundations, I now move on to discuss the specific gendered crime of femminicidio and its deep-rooted connections with history, society and gender, in the private and public sphere .

Femminicidio

In the cultural and social scenario discussed above, several forms of violence emerge; among these femminicidio is known and debated, yet not officially recognised by the criminal penal code of Italy. Karadole (2012) attempts to explain that there is a difference between the term femicide and feminicide. The difference lies in the former (femicide) being the killing of a woman, while the latter (feminicide), is meant to include several crimes such as rape, and psychological violence. There seems to be an agreement in Italy that femminicidio is the term used to refer to the criminal phenomenon in which women are killed by someone from a former or current relationship. There are several terms in academic literature that describe this crime. Monckton-Smith (2012) uses ‘intimate partner violence’ or ‘intimate partner femicide ’; Wykes (1995) terms it ‘intimate murders’ (for both female and male perpetrators); and Johnson (1995) defines it as ‘patriarchal terrorism’ (later renamed ‘intimate terrorism’). Each term originates from the idea that male violence forms part of the male’s desire to gain or retain control. Uxoricidio, in English ‘uxoricide’, is the legal term used to refer to the murder of one’s wife, but a term that does, as explained by Dobash and Dobash (1998), is likely to be based on a biological rather than embodying the symbolic definition of gender. I use the term femminicidio here, as it has now entered the vocabulary of the Italian media, particularly in relation to the ‘politics ’ of gender violence, and the work of activists in this area.

Returning to the lack of state intervention as discussed by Walby (her work taking place prior to the 1990s) since this time there has been neither a consistent or robust political response to femminicidio in Italy. There is no mention of the term (and therefore of the specific crime) in the Italian penal code and, institutionally, only half-hearted attempts have been made to deal with this matter. The law of 2013, Legge 15 ottobre 2013, n. 119, was meant to address this specific social issue, however, this law fails to use the term femminicidio , preferring to replace it with other terms such as violenza di genere (gendered violence) or violenza domestica (domestic violence). This creates an interesting scenario: while the media seems to acknowledge this term, in line with public opinion, politics and the interconnected legal domain ⁵ lags behind. The document Femminicidio. Stalking, malamore, maltrattamenti e altre violenze di genere: i primi dati della Commissione parlamentare d’inchiesta (Femminicidio. Stalking, sick love, abuse and other gendered violence: the initial data from the Parliamentary Committee ⁶ ) published by the Senato della Repubblica in April 2017 confirms that

La legislazione italiana non contempla una definizione di femminicidio inteso come uccisione di una donna per questioni di genere, cioè come un omicidio in cui l’appartenenza al genere femminile della vittima è causa essenziale e movente dell’omicidio stesso. Non è stato previsto nemmeno, fino al Piano d’azione straordinario contro la violenza sessuale e di genere adottato nel luglio 2015, un sistema integrato di raccolta e di elaborazione dei dati sul fenomeno. Per questa ragione il numero di femminicidi accertati differisce a seconda del soggetto rilevatore e dei criteri di classificazione seguiti. In particolare, i dati forniti dalle forze dell’ordine si riferiscono a tutti gli omicidi con vittime di sesso femminile e non solo a quelli nei quali il movente del reato è costituito dal genere (ovvero i femminicidi in senso proprio).

The Italian legislation does not contemplate a definition of femminicidio intended as the killing of a woman because of gender, that is such a homicide for which the belonging of the woman to the gendered group is an essential cause and motive of the homicide. An integrated system for collecting and elaborating data has not been put in place before the extraordinary action plan against sexual and gendered violence adopted in July 2015. For these reasons, the number of femminicidi ascertained differs based on those who have collected the data. More specifically, the data provided by the police forces refer to all the homicides with female victims and not only those for which the motive is a gendered one (that is what is meant by femminicidio).

Senato della Repubblica, 2017

How widespread is this criminal phenomenon? Different sets of data available are reported in Table 5.1 with the aim of offering an overview of a period of 12 years. It is impossible to state that this data is conclusive and officially recognised, as also confirmed by the extract above: as there is no definition of this crime in law, no one has been formally sentenced for committing femminicidio .

Table 5.1 Number of femminicidi from 2005 to 2016 provided by Eures and Senato della Repubblica

The data are drawn from instances provided by Eures, ⁷ an institution for social and economic research, and those published by Senato della Repubblica which has disclosed lists of the number of femminicidi, collected from different institutions in the period 2013 to 2016—i.e. the police, the carabinieri (part of Italian police force under the authority of the Ministry of Defence), and Casa delle donne per non subire violenza (an association that works to help survivors of domestic violence). ⁸

While I could comment on increasing and decreasing numbers of femminicidi (shown in Table 5.1), I prefer to withdraw from such considerations as counting victims would somewhat emphasise the ‘mathematical’ aspect of the issue, and possibly give rise to a concomitant tendency to attribute positive values to the lowest quotas. My personal view is that every single one of these murders should count as an individual history of a woman who was killed because she was trapped within fixed and prototyped gendered roles; because of this, femminicidi should not be hidden in sums. These statistics are meant to introduce this phenomenon, show that it exists and that it needs broader attention, from both institutions and communities.

While the aforementioned Senato document seems to acknowledge the failings of politics and the law with regards to femminicidio, one can hope that the work to establish a relation between (some) homicides and gender continues in future parliaments . A glimpse of this hope can be seen in the work undertaken by the parliament (of the legislatura XVII now terminated) which approve a law titled Modifiche al codice civile, al codice penale, al codice di procedura penale e altre disposizioni in favore degli orfani per crimini domestici (Amendments to the civil code, the penal code, the code of penal procedure and other measures in favour of the orphans for domestic crimes). ⁹ However, the document insists on using the plural crimini domestici (domestic crimes).

On 25 November 2017, the Camera dei Deputati organised an unprecedented event on violence against women. Speaker Laura Boldrini gathered women who had experienced violence (directly or indirectly) to sit in the Chamber to discuss their ordeal in the context of a society that is sexist and, to various extents, misogynistic. The event—called #inquantodonna (#justforbeingawoman)—saw 1300 women attend. In her introductory speech, ¹⁰ Boldrini states: “Ogni due giorni e mezzo una nostra concittadina viene uccisa per mano di chi dovrebbe amarla. Ma sbaglia chi pensa che la violenza sia una questione che riguarda esclusivamente le donne. No, no, no, riguarda il Paese e sfregia tutta la nostra comunità. Questo fa la violenza: non è una questione di donne, è una questione che riguarda tutto il Paese” (one Italian woman is killed every two and a half days at the hand of [one] who should love her. Those who think that violence is only a matter that concerns women exclusively are mistaken. No, no, no, it concerns the country and hurts all our community. This is what violence does: it is not a woman’s issue, it is an issue that concerns the whole country).

With this background as my starting point, I now outline how I investigated the three datasets, first providing an overview of overarching methods and methodologies, then providing an exploration of parliamentary acts, and finally, the media .

Overarching Methods and Methodologies

To investigate femminicidio I decided to explore datasets and comment upon them from the perspective of Corpus-Assisted Discourse Studies (henceforth, CADS , Partington et al. 2013) for several reasons. Firstly, it switches the focus from corpus linguistics as a method for solely obtaining quantitative data; secondly, I draw on sources of information outside the corpora to both justify the research and to explain how these corpora are meaningful in relation to a discourse-type (this being data on the topic of femminicidio ); thirdly, I use specialised corpora to investigate a socio-cultural phenomenon.

To offer a wide overview of how this topic is dealt with in Italy, I explore three datasets: (1) Parliamentary acts, in the form of mozioni ; (2) Newspaper articles on femminicidi reported in the news in the period 2013–2016; and (3) Newspaper reports of a specific femminicidio, a case which has captured the attention of the media and public opinion because of its cruelty.

The aim of this investigation is to uncover “construction[s] of ideological meanings” (Jeffries 2015, p. 381) with regard to the gendered crime itself and those who are part of it. For this reason, I focus on different datasets and different aspects (politics , the crime , and one specific femminicidio). I am particularly interested in how the news outlets, as Fairclough (1995, p. 4) puts it, “decide what to include what to exclude, and what to ‘foreground’ and what to ‘background’”. Specific to the news, I emphasise patterns of inclusion and exclusion, of ways in which participants are represented and the crime narrated. I also discuss the extent to which these narrations become, in effect, ‘sensus communis ’ (Schaeffer 1990, p. 2), that is a created common understanding and consensus amongst people. I consider this here in two different ways: on the one side, reproducing the ‘sensus communis ’ outside the texts, and, on the other, strengthening it through repetition (Meyers 1997). The ‘sensus communis ’ concerns expectations about how gender operates in the public sphere and helps to create the public view on this matter. On this topic, Fagoaga (1994) argues that there has been a thematisation and routinisation of gender-based violence.

The research questions of this chapter are as follows (details on definitions used will be provided in the relevant sections):

• 1. How do parliamentary acts deal with the topic of femminicidio?

  • a. Are there differences and similarities between the left leaning parties and the right leaning ones?

• 2. How is femminicidio described in a corpus of Italian newspapers in the period 2013–2016?

  • a. What forensic narratives are used to construct femminicidio in the Italian news reports in the period 2013–2016?

  • b. Who do headlines blame for the femminicidio in the newspaper articles in the period 2013–2016?

  • c. Are there diachronic changes in the use of the term femminicidio in the years 2013–2016?

• 3. How is Sara Di Pietrantonio described in the news of her femminicidio?

These research questions are dealt with in three sections—parliamentary acts; femminicidio in the news; and, the femminicidio of Sara di Pietrantonio—and are linked to specific datasets and literature backgrounds.

Parliamentary Acts: Rationale, Methods and Methodology

In October 2013, the Gazzetta Ufficiale, an institutional publication that announces new legislation, publicised a new law, Legge 15 ottobre 2013, n. 119, that stemmed from a decree-law (that is, an intervention of the government on an urgent matter) which was signed in August of the same year. The law did not directly deal with femminicidio—as already mentioned, a term that is never used—but instead modified previous articles of the penal code that related to crimes committed by a parent of a minor toward the other parent. Amendments were also made in relation to harsher punishment for those who perpetrate crimes against people who they are or have been in affective relationships with. The law also included a piano d’azione straordinaro contro la violenza sessuale e di genere (extraordinary action plan against sexual and gendered violence); this was intended to educate men and boys in interpersonal conflicts, to warn the media to use a fair representation of gender, and to promote school projects on the topic. This forms part of my rationale for investigating the political debate on femminicidio and violenza di genere (gendered violence) in the Camera dei Deputati from 2013.

I originally thought of investigating parliamentary debates (as in my Ph.D., Formato 2014). However, because the topic is not debated in the Chambers, I decided to explore other parliamentary acts, namely what is referred to as mozione parlamentare. The mozione parlamentare ¹¹ is defined by the Camera dei Deputati as:

[un intervento con il] quale è possibile proporre un dibattito e una deliberazione in Assemblea e che contiene una determinata direttiva al Governo.

[an intervention that] proposes a debate and a vote in the assembly which contains a defined direction to the government.

The reason why I chose this type of parliamentary act lies in the act’s structure and its function. It is a deliberate document proposed by one MP who has been supported by, either other MPs of the same party—as is mainly the case in the mozioni investigated here—or by a range of MPs from different parties (signalled as ‘several parties’ in the list in Table 5.2). As such acts occur at the beginning of the parliamentary procedure, its structure and its function is that of highlighting a theme to compel the attention of the government. Therefore, these acts contain ideas about the theme rather than strict information about laws and bills. The dataset of mozioni has similarities, e.g. the topic, and differences, e.g. the audience, with the news corpora investigated below.

Table 5.2 List of parliamentary mozioni on the topic of femminicidio in the 2013–2017 parliament divided into gender, political party and political orientation

To collect the dataset, I consulted the available database of the website of the Camera dei Deputati (http://aic.camera.it/aic/query.html): I searched for the term femminicidio, selected the parliament —XVII from 2013 to 2018—and opted for ‘mozioni’ in the menu of the options for parliamentary acts. I then read the 25 documents obtained to decide whether the main topic was femminicidio or if the term was used to support another topic, as in the case of some mozioni that, for instance, dealt with female genital mutilation. In Table 5.2, I list the mozioni that made the selection, and details of the date, the MP who presented it in the parliament , their political party and their political orientation. These details were collected with the aim of offering an overview on possible differences in the treatment of the topic based on political ideologies. The political orientations of the Italian parties range from left and right, to centre-left and centre-right, ‘other’, as in the case of M5S, ¹² that is an emerging movement of citizens turned politicians, and ‘mixed’, which refers to the gruppo misto, a parliamentary group formed by MPs who do not belong to other groups and those who, being dissatisfied with their affiliation, changed it to join the gruppo misto.

While I discuss the political orientation later, it is interesting to note that 11 out of the 16 mozioni are led by female MPs. In a study on the UK parliament , Catalano (2009) discusses the relation between female MPs and topics which could be thought of as women related, e.g. health, and suggests that there is a correlation between participation of women and these topics. This was also seen when investigating violence against women during my Ph.D. studies, where 48.5% of female MPs intervened on the topic and only 28.4% of the men were involved (as explained in the investigation of noi forms in Chapter 4).

The individual files were converted from PDF into text files; the final number of the individual files, as well as the total divided into political affiliation, is shown in Table 5.3.

Table 5.3 Number of words of mozioni divided into political orientation

The corpus is overall a small one, one that could be defined as ‘opportunistic’, that is “represent nothing more nor less than the data that it was possible to gather for a specific task” (McEnery and Hardie 2011, p. 11). It was carefully chosen and collected to be investigated according to the topic of femminicidio . Similarities in the number of words occur between the two main political alignments—left and right—with smaller sub-corpora oriented towards other ones. One way to look at these similarities is that the topic is transversally relevant. In the analysis, I investigate whether left and right treat femminicidio differently, prioritising some discourses over others. In relation to gender quotas, female MPs lead the number of mozioni presented in the parliament , having initiated this topic 11 times (male MPs 6 times). More specifically, Binetti Paola has asked the Camera dei Deputati to debate on this topic three times; what is telling is that she presents mozioni having moved from one party to another (in 2013 she belonged to the centre party Forza Civica, in 2016 she moved to Area Popolare NCD-UDC, aligned towards the centre-right and, finally in 2017, to the mixed group). While I believe that this is unsurprising in Italian politics where changes in political affiliation are common, this might complicate the quest for a stable and solid picture in relation to left- and right-led mozioni on this topic.

Once the mozioni were collected, I created 6 sub-corpora, which could be investigated together as one corpus or as split sub-corpora based, and the files were named as left, right, centre, all parties, mixed, other. The analysis is corpus-driven, starting with the creation of a wordlist, and investigating keyness and multi-words on the online corpus tool Sketch Engine (Kilgarriff 2012). I provide more details about each of these methods in the discussion of results below.

Results and Discussion

I started the investigation into the mozioni by checking which are the most used terms, producing a word list, namely a list of the most frequent words in a corpus. In Table 5.4, I list those words which are relevant to this chapter—therefore removing grammatical words—according to their absolute frequency (AF) and frequency per one million words (PMW).

Table 5.4 Absolute frequencies and frequencies per million words of salient words in the mozioni corpus

From the Table 5.4, we see that violenza (violence, 15,49 PMW) is used much more often than femminicidio (1,48 ptw). This might suggest that the MPs want to deal with this topic in a broader way rather than just referring to one of its forms, which, it is worth remembering, does not exist in its own right in the penal code and, as explained below, is not accepted by all speakers. It could be that violenza carries greater authority, one that could convince the parliament to tackle the phenomenon. Another interesting result is the presence of those who are victims, i.e. donne (women) but no mention of those who are alleged to have committed the crime, i.e. men. Men are invisible in the wordlist (and in the concurrent conceptualisation of violence).

While I present the systematic investigation of the news in the following sections, I here use the news corpus collected (331 articles on femminicidio in the 2013–2016 period) as a reference corpus, that is a corpus used for comparison, with the aim of investigating keywords , defined here as “[a] word which appears in a text or corpus statistically significantly more frequently than would be expected by chance when compared to a corpus which is larger or of equal size” (Hardie and McEnery 2006, p. 97). One important point to be made about keywords is that they are words (or expressions for multi-words) “which occur with unusual frequency in a given text. This does not mean high frequency but unusual frequency” (Scott 1997, p. 236). These words (or expressions) being unusual, become typical of the corpus under investigation.

I have compared the news corpus with the mozioni corpus via Sketch Engine (Kilgarriff 2012), focusing on relevant multi-words, namely those key ‘compound’ words that could, arguably, replace femminicidio . These are listed in Table 5.5; those selected had a keyness score above 100.

Table 5.5 Multi-words with keyness score above 100 in the mozioni corpus

These terms seem to encompass more than homicide (or the crime of femminicidio as such), indicating that women are subject to several forms of violence: (1) some specific, e.g. violenza fisica (physical violence), violenza sessuale/violenze sessuali (sexual violence); (2) some more generic, e.g. those with forma (type), forma di violenza, forma di discriminazione, forme di violenza, forma di abuso; and (3) Those referring specifically to gender, e.g. violenza di genere (gendered violence), discriminazione di genere (gendered discrimination), violenza maschile (male violence). Based on the notion of keywords and the difference in genre between the two corpora investigated (the mozioni vs the news), we can see that violence is presented as an abstract phenomenon rather than actions committed by humans against others, confirming the absence of the terms uomo and uomini (man/men) in the word list discussed above. This is, however, not unexpected; in previous work, I investigated violence metaphors used in debates on violence against women in the 2008–2011 period and found similar results (Formato 2014). This can also be because these texts tend not to include specific events but instead talk about the topic more broadly. In the multi-words list there is only one term that refers specifically to those who (are likely to) commit violence against women and that is ex partners (with 6 occurrences and a key score below 100).

With the aim of investigating differences between the left-wing and right-wing parties, I compared the two sub-corpora based on political affiliation, those that have more mozioni, e.g. left leaning (5 mozioni, 9640 words) and right leaning (5 mozioni, 9597 words). I examined the first 25 multi-words as generated by Sketch Engine (Kilgarriff 2012). I then proceeded to categorise the multi-words in groups according to similar aspects with the aim of exploring differences and similarities between the two political orientations; the quantitative data are shown in Table 5.6, divided into the category, the number of multi-words (NM), the absolute frequency of the multi-words (AF) and the percentages (based on the categories).

Table 5.6 Number and absolute frequencies of multi-words and percentages of content categories in the mozioni corpus

Before commenting on the differences and similarities between the two political orientations, I here note that the highest number of multi-words can be found in the group law/parliament procedures, as one would expect for the arena where these texts are presented. Linee guida (guidelines), decreto legislativo (decree law), concreta attuazione (real actuation), azioni per i centri (measures for the centres) appear in the centre-left corpus; while delitto di atti persecutori (crime within persecutory acts), atto di indirizzo (measure addressed to), presente atto (current measure), custodia cautelare (cautionary custody), appear in the centre-right corpus. Terms from several fields are used by the centre-left, e.g. ‘culture’ (parità di genere/gender equality, cultura patriarcale/patriarchal culture), ‘institutions’ (commissione europea, commissione parlamentare) and, more importantly, ‘victims’ (a term which does not appear in the centre-right, i.e. minori vittime/victim minors, vittime di violenza assistita/assisted violence victim). In the right-centre, the group ‘violence’ covers the 28% (12% in the centre-left) of terms that convey a wide range of types of violence (episodi di violenza/violent cases, violenza sessuale di gruppo/gang rape, aborto forzato/forced abortion, atti sessuali/sexual acts).

I now draw some conclusion and answer the research question: How do parliamentary acts deal with the topic of femminicidio in the parliament XVII? Generally speaking, the criminal phenomenon of femminicidio is seen within a broader understanding of violence as an abstract phenomenon and its main focus is on women (as victims) rather than men (as perpetrators). The answer to the research question— Are there differences and similarities between the left-leaning parties and the right-leaning ones?—do right-leaning parties seem to focus more on what procedures can be taken in the parliament and the types of violence, while left-leaning ones seem to have a broader view of the phenomenon in relation to other institutions, victims, and different aspects of the culture surrounding femminicidio . This word does not appear in the 25 highest keywords in any of the two larger sub-corpora.

Femminicidio in the News

Literature already exists on language used in the media to describe crime: with some references to gender (Tabbert 2016), gendered violence (in Spain, Santaemilia and Maruenda 2014), and rape (O’Hara 2012; Ehrlich 2003). Abis and Orrú (2016) published a qualitative analysis of 143 new reports (in the period 2010–2015) on the topic of femminicidio , investigating descriptions of: (1) Victims (accounts of their role within the relationship and their physical traits); (2) Extenuating circumstances portrayal of their professional status (alleged) mental health issues and how these are used to describe what led to the crime ); (3) Motives (jealousy); and (4) Violence. They admit that “dagli articoli che abbiamo raccolto non è in sostanza possibile interpretare la violenza sulle donne da un punto di vista sistemico” (from the articles that we have collected it is not possible to interpret violence against women from a systemic point of view, 2016, p. 31). Starting from this claim, I aim to offer a systemic view of how femminicidio is constructed in the news, by using methodological tools—e.g. keyword analysis and collocations —which could function as a guarantee for systematicity. Systematicity is also found in Santaemilia and Marueda (2014) who, through corpus methods, investigated the Spanish daily broadsheets—El País and El Mundo—on the topic of violencia de género (gendered violence), violencia doméstica (domestic violence) and violencia machista (sexist violence), in the period 2005–2010. They conclude that the newspapers focus on the objectification of the victims who are in need of protection (from the institutions), while male perpetrators seem to be hidden, almost always excluded, and certainly backgrounded.

In the section on parliamentary acts, I review the law promulgated in 2013, which, among other measures, proposed to educate the media on how to report the violence phenomenon fairly, and, above all, remain mindful of the women involved. This is why my dataset starts from 2013. In addition, I also review public voices and opinions raised on how to, and how not to, depict violence. For instance, the activists of Non una di Meno ¹³ write that, in some media :

C’è una straordinaria coerenza di costruzione discorsiva, in cui la violenza sulle donne è raccontata dal punto di vista di chi la esercita e sublimata come parte del mito fusionale, dell’ideologia dell’amore romantico/passionale: l’uomo che agisce violenza viene rappresentato come ‘innamorato’ della vittima, il movente è la gelosia , ritenuta una “passione sana” (al contrario della violenza che è morbosa od eccessiva) oppure l’incapacità di accettare la separazione, raccontata con modalità che producono empatia ed assoluzione, deresponsabilizzando e legittimando l’autore della violenza. Un altro frame spesso impiegato è quello della relazione conflittuale, che giustifica una violenza letale come reazione a una discussione e sposta la responsabilità dall’aggressore all’intera dinamica di coppia, di fatto alludendo alla corresponsabilità della vittima. In ogni caso, nonostante l’impiego del termine femminicidio sia aumentato da 4 articoli nel 2006 a 5000 articoli nel 2013, il modo di affrontare e descrivere il fenomeno rimane ancorato alla percezione della violenza come questione individuale

There is a strong coherence of the discursive constructions in which violence against women is told from the point of view of those who commit it and seen through the lens of the fusional myth, ¹⁴ [perceived] as part of the ideology of romantic and passionate love: the men who use violence are represented as ‘in love with’ the victim, the motive is jealousy seen as a healthy element (contrarily, violence is seen as morbid or extreme) or [it is described in relation to] the incapacity of accepting the end of the relationship, constructing empathy and acquittal [towards men], denying responsibility and legitimating the author of the violence. Another frame used is that of the conflict within the relationship, as a justification for a lethal violence caused by an argument moving the responsibility from the male perpetrator to the couple dynamic, hinting at the shared responsibility of the female victim. All in all, while the use of the term femminicidio increased from 4 articles in 2006 to 5000 articles in 2013, the way of facing and describing the phenomenon is still tied to the perception of violence as an individual matter.

Non una di meno, 2016

This extract is one of great interest as it summarises how femminicidio is seemingly treated in the media, and it justifies a systematic CADS investigation into the themes implied, e.g. the shared responsibility of the victim, a cultural rather than an individual phenomenon, and, motives behind the killing. However, it was impossible to have a comprehensive picture of their study (e.g. number of articles) and how they conduct what seems a linguistic analysis. More recently, a group of female and male journalists—under the name GiULiA—composed a set of guidelines on how to report femminicidio and, more broadly, gender-related news, in the press. Their work is collected in a booklet titled Stop violenza: le parole per dirlo (Stop violence: the words to say it), published in 2017, in which they outline 10 rules to be followed. Those that are relevant to this chapter are listed below, with the original numbering:

2. adottare un comportamento professionale e consapevole per evitare stereotipi di genere e assicurare massima attenzione alla terminologia, ai contenuti e alle immagini divulgate.

5. utilizzare il termine specifico ‘femminicidio’ per i delitti compiuti sulle donne in quanto donne e superare la vecchia cultura della sottovalutazione della violenza: fisica, psicologica, economica, giuridica, culturale.

10. nel più generale obbligo di un uso corretto e consapevole del linguaggio, evitare:

b) termini fuorvianti come ‘amore‘, ‘raptus ’, ‘follia’, ‘gelosia ’, ‘passione’ accostati a crimini dettati dalla volontà di possesso e annientamento.

d) di suggerire attenuanti e giustificazioni all’omicida., anche involontariamente, motivando la violenza con ‘perdita del lavoro’, ‘difficoltà economiche’, ‘depressione’, ‘tradimento’ e così via.

e) di raccontare il femminicidio sempre dal punto di vista del colpevole, partendo invece da chi subisce la violenza nel rispetto della sua persona.

2. to adopt a professional and informed manner [when writing] to avoid gender stereotypes and to pay meticulous attention to terminology, contents and images used.

5. to use the specialized term femminicidio for crimes perpetrated on women as women and to overcome the old culture of underestimating physical, psychological, economic, juridical and cultural violence.

10. In the wider duty of using a correct and informed use of language, one should avoid:

b) misleading terms such as ‘love’, ‘burst’, ‘insanity’, ‘jealousy’, ‘passion’ to describe crimes which are guided by possession and annihilation [of those murdered].

d) to suggest extenuating circumstances and justification for the perpetrator, even if done unintentionally, motivating violence with ‘loss of job’, ‘economic problems’, ‘depression’, ‘cheating’ and so on and so forth.

e) to always narrate the femminicidio from the perpetrator’s point of view, focusing instead on those subject of violence, respectful of them.

(GiULiA 2017, p. 82–83)

Both the work undertaken by Non una di meno and GiULiA focus on an aspect that has been dealt with in the literature on gendered violence in the UK, that is, the centrality of love as a justification and explanation for acts of violence (Monckton-Smith 2012, p. 3).

Starting from observed patterns, it seems paramount to dedicate part of this study to what Monckton-Smith (2012) refers to as forensic narratives developed in the media , more specifically newspapers, in order to consider discourses about and around gender (in both private and institutionalised/public arenas) arising from this dataset.

Monckton-Smith (2012) explains that this is an umbrella term to define “those crime narratives used in formal legal processes to establish what happened and why and are constructed in police, and prosecution and defence case files” (2012, p. 73). From her background in the police, Monckton-Smith explains that facts collected become a narrative which is useful in court, more specifically for the jury, and therefore they contain moral stances and ‘sensus communis ’ beliefs about crime and those who commit them. Therefore, Forensic narratives concern how the characters are positioned within data (newspaper articles in her case)—for instance, whether they are positioned as a couple or as individuals—as well as what recurring themes and subtle clues can be found (Haaken 2010). These forensic narratives are never conceptualised outside a historical, political and cultural background, and this is why I believe that this term and its notion can be useful to investigate the 2013–2016 news. Borrowing this term from Monckton-Smith , I adapt it to reflect how I explore the people involved and other insights into how femminicidio is described and provide the following definition of the term:

Forensic narratives arise from the investigation of systematic linguistic patterns which are used to describe 1. those involved in the femminicidio and 2. Apparent reasons, motives, explanations about why the crime has occurred. These narratives shed light on ideological stances about cultural understandings of gender within heterosexual relationships in a gender-imbalanced social and cultural context, which may be, nonetheless, perceived as ‘sensus communis’.

There is one crucial point to make on this definition and on the operationalisation of it. Forensic narratives are a term which is analysis-led. This means that the systematicity of the results produce forensic narratives in what can be described as a bottom-up approach. In other words, it is the investigation of these newspaper articles that gives rise to forensic narratives based on several stories. Forensic narratives are not stories of individual femminicidi but a construction of how gender is seen through femminicidio. I use different corpus linguistics methods: single word, multi-word analysis and collocations (defined in the analysis below) to explore these narratives.

The News Corpus

I collected the data via the Nexis database with the intent to investigate femminicidio from 2013 to 2016. The news corpus and sub-corpora of this section consist of articles which Nexis found to contain the expressions femminicidio OR delitto passionale (crime of passion) in Italian language news. I chose these two terms based on the intuition that they could be used interchangeably or, to some extent, had a similar meaning. The period I investigate commences from the 1 January to 31 December of each year considered, that is 2013, 2014, 2015 and 2016. While I do not necessarily see this study as a diachronic one—namely, an examination of how the narratives changes (or does not) in the years—I take into consideration that a debate on good practices when writing about femminicidio could have been taken on board; for this reason, I explore whether there have been changes throughout the years, specifically in relation to use of the term femminicidio.

Once the data were collected through the database, I proceeded to clean the data, focusing on some criteria which would match the aims and research questions of this chapter. The corpus collected contains newspaper articles which recount: (1) episodes of women killed by somebody whom they knew, most often a(n ex-)partner, a(n ex-)husband, or someone whom these women had expressed the intention/desire not to be with and (2) details of the trials in which the male perpetrators are accused of murder or manslaughter. This corpus does not contain: (1) Editorials on the theme, that is opinion pieces where journalists or experts explain what this phenomenon concerns; (2) Events which are related to this phenomenon, e.g. exhibitions, launches of books on the subject; (3) Accounts of episodes where women survived the violence, as this would be connected to other parliamentary and legal procedures, namely, atti persecutori (harassment) and stalking; (4) Episodes of femminicidio regarding women killed outside the national border of Italy; and (5) Political views of the episodes, e.g. politicians releasing statements on specific women killed or on the progress of the law in the parliament . In cleaning the data, I realised that the perfect dataset does not exist. I was required to make clear choices in deciding what newspaper articles to consider in respect of the list above. The articles have been published in several newspapers or press agencies, more specifically Il Giorno, il Giornale, il Resto del Carlino , il Corriere della Sera , Ansa. Therefore, the corpus includes a variety of press outlets; given the difficulty in sometimes tracing this information, political alignment is not taken into consideration.

The following have been cleaned to contain only the newspaper articles, while noise such as metadata (date of the publication, newspaper, etc.) has been removed, see Figs. 5.1 and 5.2.

Fig. 5.1

Sample of newspaper article before removing noise

Fig. 5.2

Sample of newspaper article included in the news corpus after noise has been removed

In fig. 5.2, I show the elements that have been removed in order to systematically investigate the newspaper articles. More specifically, I removed the information relevant to section, length, dateline, newspaper title, date, load-date, language, publication-type and journal-code (which appear in Fig. 5.2). Some other newspaper articles contained other information, which, similarly, has been removed, e.g. section, highlight. A copy of the full set of uncleaned data was kept in order to be used for insights where needed. I left in the article headlines and conducted a study on these too.

Having explained how the corpus has been built, I present the number of articles and the number of tokens for each year, in Table 5.7.

Table 5.7 Number of articles and words in the news corpus (2013–2016) and sub-corpora divided by years

There is a discrepancy in the number of newspaper articles across the years. It is difficult to predict the reasons for this as incidences of femminicidi are, on the whole, similar across these years (as shown above). It is possible that these newspapers dedicate less space to a phenomenon which has somehow become normalised in society and, therefore, less episodes are reported. Santaemilia and Maruenda (2014, p. 257) show that, likewise, their datasets alternate the three terms used (violencia de género, violencia doméstica violencia machista) and that increasing and decreasing trends can be seen across the 5-year period. In their dataset, violencia machista (sexist violence) seems to have increased over the years (especially in 2009).

The articles I collected are investigated according to different quantitative and qualitative methods. I offer specific details on each method used within the relevant analysis sections, with the aim of giving a step-by-step overview of how the results were gained.

However, I anticipate that the methods to investigate forensic narratives is through corpus-driven approach (Tognini-Bonelli 2001), that is to investigate the corpus without prior assumptions and expectations. McEnery and Hardie define a corpus-driven approach as “a corpus method that is entirely bottom-up” (2011, p. 242), this serving the investigation of forensic narratives. ¹⁵ A qualitative analysis of telling single and multi-words is provided; these are explored according to corpus linguistic functions, e.g. collocations and concordance lines.

In the second part of the analysis, I isolate the headlines of the newspaper articles; in the third, I investigate whether the term femminicidio has changed across the 4-year period, before drawing conclusions on this dataset. The overarching research question for this study is: How is femminicidio described in a corpus of Italian newspapers in the period 2013–2016?

Single Words

In this section, I present and discuss the results of the analysis of single words in the news corpus. I conducted keyword analysis through the corpus tool Sketch Engine (Kilgarriff 2012) to extrapolate statistically significant lexical items—on their own, i.e. the single word and in a lexical group, i.e. multi-words (discussed in the following section)—comparing the news corpus with a reference one—in this case itTenTen16 (4,972,350,335 words). It proved difficult to collect data and to build a representative corpus that could be compared with the news corpus, e.g. news reporting other crimes (e.g. homicides which did not have a gendered dimension). For this reason, I used itTenTen 16 as a reference corpus as it contained texts about gender and gendered violence (the focus of the news corpus); this also helping to overcome the differences in genres between the two compared corpora (itTenTen16 contains texts collected from the internet); furthermore, the availability of the corpus on SketchEngine and the year in which it was collected (2016) meant that it was suitable for this research. The results have been ordered according to their keyness scores (Kilgarriff 2012) ¹⁶ ; a list of the scores and frequencies is available in Appendix 1. In selecting the first 50 keywords , I have decided to remove names and surnames of people involved (for both victims and perpetrators) as they are related to single stories, while I am here interested in forensic narratives (as explained in the methodology section above). However, I have left nomination strategies used to group some people (for instance, nationality) as such groups have a potential to provide insights into cultural and social typification e.g. romena (Romanian). I have grouped the keywords to form a scenario of femminicidio as recounted in the relevant newspaper articles (Table 5.8).

Table 5.8 Single words, ordered in descending order, grouped according to aspects of the femminicidio in the news corpus

It is unsurprising that terms that relate to the murder of the women are prominent (feminine endings), e.g. uccisa (murdered), ammazzata (killed), as well as the means used to cause the death, e.g. coltellate (stabbing), pistola (gun). In relation to the people involved, we see women addressed (according to the keywords ) as badante (care giver), romena (Romanian) and convivente (cohabitant, an epicene noun), this last also used for the male perpetrator to indicate the relationship between the two (signalled with (!)). The male perpetrators are similarly referred to by the role they play within the couple, i.e. fidanzato and fidanzatino (sweetheart—ino is a suffix used to add emotive connotation to a word). While he is described as omicida (killer) and assassino (murderer), terms are also used related to the action of killing—uccide (kills) and ucciderla ([to] kill her)—with regards to his role in the femminicidio. Language used to describe the relationship prior to the murder (‘pre-killing relationship events’) suggests conflict, clashes that, it is implied, lead to the killing: lite/litigio/litigavano (arguments, quarrels, they used to fight). Monckton-Smith (2012), as reviewed at the beginning of this chapter, discusses the problematisation of placing blame on the failed relationship (and therefore the couple) rather than on those who commit the crime.

Moreover, as argued by activists of Non una di Meno and GiULiA, the reasons for the murder stand out in this corpus. We find the perpetrator described as in love with the victim, suggesting ‘love’ as a justification for the killing, a consequence and demonstration of it, e.g. gelosia (jealousy), geloso (jealous), passionale (passionate). Likewise, we find the killer being guided by a temporary condition— raptus (burst)—one that seems to lessen his responsibility for the act. In other words, perpetrators are not seen as consistently violent people. In the following sub-section, I investigate the two terms gelosia and raptus, as well as the relationship between the two.

Gelosia and Raptus

In order to investigate the two terms in more detail, I checked the concordance lines of each term with Sketch Engine (Kilgarriff 2012) and then performed collocational analysis with the function graphcoll on Lancsbox (Brezina et al. 2015). Collocation is the “co-occurrence relationship between two words. Words are said to collocate with one another if one is more likely to occur in the presence of the other than elsewhere” (McEnery and Hardie 2011, p. 240, see also Glossary 2).

As for gelosia (jealousy), I investigated collocates in a span of minus 2 on the left, as intuitively the term was going to be preceded by an article or articulated preposition (and this preceded by a noun) and zero on the right, as relevant elements were thought to appear before. The results allow for further discussion on this term, as presented in Fig. 5.3.

Fig. 5.3

Collocations of the node gelosia in the news corpus

I now comment on the collocates with the aim to shed light on how gelosia (jealousy) is used as a term by the media, and how, specifically, it occurs in this corpus. I decided to examine collocates based on their frequency, starting from the list produced according to the Mutual Information score which measures the strength of the collocations . ¹⁷ The definition of the term gelosia is as follows: “Stato emotivo di dubbio e di tormentosa ansia di chi, con o senza giustificato motivo, teme (o constata) che la persona amata gli sia insidiata da un rivale” (Treccani, ¹⁸ —emotive state of doubt, tormented anxiety of those who, with or without justified reason, fear or certify that the loved person is harrassed by a rival—my emphasis). There are, in this definition, two implied gendered elements, one is the indirect object pronoun gli (masculine object pronoun) that seems to emphasise that the harassment is hurting a man, meaning, by assumption on heteronormativity, that the loved person is a woman. What is interesting here is that the sentence would make sense even without the pronoun, as in che la persona amata sia insidiata. The other indirect implication comes via the masculine indeterminate article un which precedes the epicene rivale. Both—gli and un—could be, as described in Chapter 2, used as generics. ¹⁹ If we move on to the denotation of the term, this seems to conceptualise jealousy as a sort of rivalry between two people, one inside and the other outside of the couple relationship, where the other half of the couple is loved by one and harrassed by someone else. Starting from the understanding that these news articles deal with heteronormative couples, it seems that masculinity (and the threat of one man over another one), is part of the femminicidio; therefore, the woman seems to be passive (and) caught in the middle of the rivalry. The closest and most frequent collocate is the preposition per (for, 28 occurrences) which indicates a corresponding ‘cause’, suggesting that the killings here are caused by the jealousy (of the man towards the woman), as shown in the sample concordance lines in Table 5.9.

Table 5.9 Concordances lines of the node per gelosia in the news corpus

In these concordances, we read that the cause that led the man to kill the woman (uccidere per/to kill for, fare per/to do something for) was indeed jealousy. This is confirmed by the verbs which appear in the graph above, i.e. acceccato (blinded by), scatenato (triggered by), ossessionato (obsessed by), together with the nouns tarlo (seeds of) and lite (argument), and the adjective ossessiva (obsessive). However, there is a recognisable difference between ossessionato/ossessiva (obsessive), tarlo (seeds of) and acceccare (blind by) which are killer-oriented, and scatenato (triggered by) and lite (an argument for) which seem to be event-oriented. Another interesting preposition is dalla (from the) as it positions jealousy as separate from the person, indicating something that comes from the outside: verbs that precede the construction dalla gelosia , are ossessionato (obsessed) and acceccato (blinded by). Among the collocations of gelosia, raptus (burst) also appears. Here, I first investigate the relation between the two, discussing raptus and its other collocations later in this section. The definition of raptus (burst) in the Italian dictionary is “[i]n psichiatria, impulso improvviso e incontrollato che, in conseguenza di un grave stato di tensione, spinge a comportamenti parossistici, per lo più violenti” (Treccani, ²⁰ in psychiatry, a sudden and uncontrollable burst which, consequently, to a grave state of tension, pushes [the person to have] paroxysmal behaviours, mainly violent ones). The nature of this term as sudden and uncontrollable, originated in a state of tension, does not seem to link with what has been discussed above; for instance, terms which indicate that the person had a prolonged issue with jealousy (obsessed, blinded by, etc.) (Table 5.10).

Table 5.10 Concordance lines of the node gelosia in the news corpus

The concordances above use the term raptus paired with gelosia , seemingly as pertaining to the legal discourse, sentencing perpetrators based on the motive: the women were killed because of the sudden and incontrollable ‘burst’ of men.

To gather more information on how the term raptus is used. I searched for its collocates, in the span zero (on the left) and 2 (on the right); the collocations are shown in Fig. 5.4.

Fig. 5.4

Collocations of the node raptus in the news corpus

It is not surprising that gelosia appears in the graph. The two other nouns which collocate with raptus, following the preposition di (of) which introduces a complement of specification, are violenza (violence) and follia (insanity). These two terms are quite different in nature, with the first based on action and the second based on mental dysfunction. Therefore, violenza (3 occurrences) seems to explain the behaviour that follows the sudden and uncontrollable change of state in one person, while follia (5 occurrences) seems to aggravate that sudden and uncontrollable change, into something that seems a long-standing mental illness. The two—follia and raptus —appear to contradict each other, one being a sudden realisation and the other an established condition of an alteration of reality. Arguably, one is more likely to be forgiven, based on the understanding that the behaviour is a one-off, but the pair raptus and follia indicates a different picture. The overall analysis of this term shows how problematic the use of the term raptus can be. This has been seen in prior studies, as Wykes (1995) found, men were discussed in the media as “hav[ing] acted in the heat of the moment” (1995, 54) when discussing intimate murders in the UK media. Similarly, Monckton-Smith (2012) reports that ‘just losing it’ is a common (and accepted) narrative in cases of intimate partner femicide .

Multi-words

With the aim of investigating significant pairs in news corpus, I examine the function multi-words offered in Sketch Engine (Kilgarriff 2012). The multi-words analysis provides further insights into the forensic narratives that appeared with the single word examination. These are the first 50 multi-words processed by Sketch Engine (Kilgarriff 2012, function: keyword terms: the table with the scores is presented in Appendix 2). As for the analysis above, the reference corpus is itTenTen16. Starting from the results, I reconstruct the forensic narratives before comparing them to the previous findings. Phases—pre, during and after violence, characteristics of the crime, nominalisation of the protagonists, and police and court-related terms—are taken into consideration for the scenarios presented in the 4-year news corpus. I also comment on dispersion, that is, I check that these terms are not used in one case only, but are used sparsely across the corpus (Table 5.11).

Table 5.11 Multi-words , ordered in descending order, grouped according to aspects of femminicidio in the news corpus

Starting from this table, one notes some similarities (e.g. the focus on passionale flagging the link between love and femminicidio and on the role in the relationships of the two parts involved) and differences with the single word analysis. Precisely, I investigate the categories that, I argue, are more relevant in constructing a coherent forensic narrative, more specifically, the criminal phenomenon together with murder related terms and people markers. I then move to investigate in detail a term that occurs in different pairs, namely ennesimo (umpteenth).

Criminal Phenomenon and Murder-Related Terms

In relation to the criminal phenomenon, we find patterns indicating a repetition; more specifically, both nuovo (new) and ennesimo (umpteenth) contextualise the crime as a pattern: while the former seems more neutral, the latter refers to something that has an indeterminate number, and yet a very high one (as I discuss later); similarly, ultima vittima (latest victim) is referring to the crime as such (rather than exclusively to who has been murdered) and signals a relation with time. However, some multi-words do not contain a numerical recurrence-related element, yet some contain modifiers which similarly emphasise the severity of the crime, i.e. duplice omicidio (double homicide). Incidente domestico (domestic incident) seems to be on the opposite scale of severity: in checking the concordance lines, however, it emerges that the killer (as the concordances are referred to one episode only) had pretended that the wife had died following an incident occurred in the house. On the other side, delitto passionale (passionate crime), an expression which has gathered disapproval over the years, reproduces an understanding that men “killed because they loved, which is more forgivable than killing without love” (Monckton-Smith 2012, p. 3). This links together with the reasons given for the murder, that is movente passionale (passionate motives) and futili motivi (trivial motives). If we consider the futili motivi (which appears in different texts), there seems to be discrepancies between how the violence was used against the victims and the reasons for this violence. On the one side, violence is described as a furia omicida (homicidal rampage), aggressione mortale (fatal aggression), while futili motivi refers to the triviality of the act. There seems to be an unbridgeable gap between asserting that the violence was cruel and that the reasons for it were either motivated by love, widely perceived as the opposite of cruel violence, or by solvable problems. While these two seem to attribute the fault to the perpetrator, ennesimo litigio, ennesima lite (umpteenth argument), which precedes the murder event, appears to remove the responsibility from one person only and attribute it to the couple’s dynamic, “where love in relationship has broken down” (Monckton-Smith 2012, p. 63). The terms litigio and lite also appear in the single word list, however, with the multi-words function, we are offered greater insights.

People Markers

In this category, I grouped the multi-words which referred to the two parties involved: the perpetrator and the victim. It is unsurprising that almost all terms refer to the two people in their past relationships, signalled by ex (ex), for instance ex marito (ex-husband), ex moglie (ex-wife), ex convivente (ex-cohabitant), ex compagno (ex-partner [masc]), ex fidanzato (ex-boyfriend), ex compagna (ex-partner [fem]), and, amante (lover), some of which also appear in the single word investigation. While some terms attributed to perpetrators seem to be relevant to specific incidents, it is interesting to find that these offenders are often described according to their (privileged) profession, i.e. avvocato (lawyer), guardia giurata (night guard), primario oculista (Head ophthalmologist) and agente immobiliare (estate agent). This creates an asymmetry between the two gendered groups, where women are only seen as belonging to the private (in relation to their personal space and their feelings) while men are seen within a public sphere (although, of course, this linked together with their intimate side, the relationship). Speculating on these results, it seems that men, as rounded humans, are foregrounded by having their role in society established, alongside the authority that their position affords. The roles are indeed very telling: doctor (somebody who should be looking after people), lawyer (somebody who should know what is right and wrong), and a police officer (somebody who should fight crime rather than commit it). If we return to the single word study, where women were referred to via their job as care givers, we can speculate on a clear division between the spheres in which men and women are legitimised to operate (and display in the media as a public arena where meanings are negotiated). The re-occurrence of these terms, and seeing them in groups, provide an interesting insight into Italian society.

Ennesimo (Umpteenth)

The term ennesimo (umpteenth) appears in the list of the single and multi-words of the news corpus. It refers to something serial that is high, yet indefinite, in frequency ²¹ ; in itTenTen16, it collocates with caso (case), episodio (episode) and occasione (occasion), shedding light on the relation between ennesimo and events. In a search to make sense of the function of this term in the news corpus, I utilised the framework proposed by Bednarek and Caple on news values, defined as “what’s news” (2017, p. 1). I treated the investigation of ennesimo (and nuovo) as both inherent to the femminicidio as a topic—what Bednarek and Caple (2017, p. 31) define as ‘event’—and apposite to ‘how’ the news on this topic is reported—defined as ‘story’. I believe that ennesimo builds the relation between the event and the story, and explain so by examining them through the news values of ‘superlativeness’, ‘timeliness’ and ‘consonance’. The three news values are applied to how the femminicidio is told (in other words, the “story”) but they also lead to an interesting discussion about the “event”. According to the news value framework (Bednarek and Caple 2017), ‘superlativeness’ is expressed through linguistic instances of intensification and quantification; if we focus on the latter, we can see that ennesimo expresses a vague quantity. Bednarek and Caple argue that the vague quantities “may indicate lack of knowledge but also seem to construct the ensuing number as high” (2017, p. 95). This fits particularly well with the lack of officially recorded numbers of femminicidi in Italy, as discussed above. However, the terms also have elements of ‘timeliness’. More specifically, ennesimo (and nuovo) suggests a link to the past, as well as being the latest in a series of similar events. Opposite to ‘unexpectedness’ which emphasises the surprise generated by the event, ‘consonance’ is the news value that constructs the event as typical, more specifically, in relation to a comparison with the past and the expectance of “yet another”.

The link between the ‘story’ and the ‘event’ can be found in the uncertainty of the legal status of the crime. As already discussed, the 2013 bill, which many newspapers then labelled as the legge sul femminicidio (law on the femminicidio), does not include the term femminicidio but focuses on violenza di genere (gendered violence), and violenza domestica (domestic violence), as outlined in the later document published by the Senato in 2017, and in the relevant parliamentary acts. Similarly, the penal code does not contain a legal definition of femminicidio and/or how this should be treated in criminal cases. The void left within the law, as well as in the penal code, renders the term ennesimo (and nuovo) inherent and proper to the “event” rather than merely the “story”. To conclude, I believe that this is further detrimental to the women who have been killed by a toxic, yet accepted, masculinity.

The Newspaper Headlines

One of the most commented upon practices among activists concerns the headlines of newspaper articles. For this reason, I decided to explore newspaper headlines and provide an analytical framework which investigates the positions of the participants in them. While some scholars have worked on the styles and functions of the headlines (Ifantidou 2009) criticising the misrepresentation and distortion of events in the story, I adopt a critical stance on language in these headlines as contributing to reproduce specific ideologies (as in the tradition of Critical Discourse Studies, see Van Dijk 1998; Caldas-Coulthard 1993). Starting from this, I isolated the headlines of the articles of the news corpus, exclusively considering those that mentioned the events, whether in reported speech—as in «Ha ucciso la compagna e la figlia. Date l’ergastolo al primario oculista» (He has killed his partner and the daughter. Give a life sentence to the Head Optometrist)—or in the words chosen by the newspapers—Ossessionato dalla gelosia accoltella la ex e confessa (Obsessed by jealousy, he stabs his ex and confesses). Some of the newspaper articles did not have a headline, while some other headlines were removed as they: (1) had a general headline , for instance «Vicenda tragicamente attuale» (A tragically timely episode), Marini: Incomprensibile violenza (Marini: incomprehensible violence); (2) referred to the trials and did not contain reference to the story, as in “Ci ha rubato la felicità. Ora date l’ergastolo all’assassino di Lucia” (“He has stolen our happiness. Now give Lucia’s killer a sentence for life”), L’assassino tenta la carta dell’infermità (The killer tries to be judged as having mental issues), Uccise 17enne e ferì la sorella. La perizia: “Incapace di volere” (He killed (somebody who was) 17 and injured her sister. The appraisal: “Unfit to plead”); (3) Reported more than one single femminicidio , e.g. Tre ragazze uccise in due giorni. Allarme femminicidio (Three young women killed in two days. Alert femminicidio); and (4) did not seem to highlight the ‘love relationship’ between the victim and the killer, as in Uccisa a Cesena da un uomo che poi si tolse la vita in duomo a Cervia (Killed in Cesena by a man who committed suicide in the cathedral of Cervia). In Table 5.12, I summarise the number of headlines for each year under investigation.

Table 5.12 Number of newspaper headlines divided by years

If we look at the number of the headlines that made the headline corpus (investigated qualitatively) with respect to the news corpus which contains 331 articles, we note that only 58% of the headlines strictly give information about the femminicidio. Starting from these headlines, I proceeded to group them. However, categorisation was not an easy task here as some of the elements of interest were found to be overlapping in the headlines. To explore patterns within them, I decided to build an analytical framework which would investigate who the blame for the femminicidio was attributed to.

Blame is “the most proximal attributional determinant in behavioural response, actions, judgments about appropriate punishments” (Rye et al. 2006, p. 639), stressing how it is thought to be related to what follows the event. In other words, newspapers are, arguably, indicating who is to be punished (or has been punished), for what and the extents to. While scholars tend to define and discuss blame in relation to it being unfairly placed upon women in cases of sexual rape or misconduct at the end of men, I here use the term to explore blame laid on either party in the relationship; namely, I do not wish to exclude that blame is used by newspapers to sanction the male perpetrators. There is an evident asymmetry, however, of blame being fairly put on the perpetrators and unfairly put on women who are constructed as “transgress[ing] acceptable (…) boundaries of dress, behaviour or femininity [this seen as] complicit in what has happened to them, [and as] individually responsible for their fate” (Thapar-Björkert and Morgan 2010, p. 40). The research question addressed in this section is: Who do newspapers headlines blame for the femminicidio in the period 2013–2016?

In Table 5.13, I outline the analytical framework and provide examples for each category.

Table 5.13 Analytical framework used to investigate blame in the headlines with examples

The analytical framework consists of 3 main categories—‘blame on the killer’, ‘a male perspective’ and ‘blame on the woman’. In operationalising this framework for analysis, I made some choices based on the linguistic elements that were present. For instance, the first category is split into two sub-categories: ‘killer’—where the blame is only positioned on the man who committed the crime—and ‘killer plus the couple’—where there were suggestions that the relationship was unstable, or the couple close to legal separation and divorce, therefore hinting at the idea that the couple dynamics contributed to the crime. The rationale for this is that the blame shifts from the man only to the dynamic of the couple, a dynamic which is seen as being a possible cause for the actions of the man (as also discussed above). The category ‘a male perspective’ aims to explore how men are viewed when they are reacting to events in their emotional life, with a focus on how their feelings are hurt upon being rejected, this leading to the crime according to the newspapers. In the sub-categories, I included ‘men’s emotions’ and ‘jealousy’ (which, as argued above, are used to build a specific masculinity in the rivalry between two men). The final category is ‘ blame on the woman’, similarly divided into two sub-categories: the ‘women’s behaviour’ seems to reproduce the imbalanced gendered fixed roles of heterosexual couples, where traditionally the woman is seen as passive and a mere receiver of men’s love and decisions; the second ‘women’s action versus killer’ is based upon the previous actions the woman took to report the (ex) partner’s violence. The sub-categories ‘men’s emotions’ and ‘women’s behaviour’ differ from one another based on who of the two is the linguistic subject, whether in active sentences (s/he + action/feeling) or in passive sentences (s/he + passive verb + male/female receiver).

The analytical framework attempts to systematically investigate the linguistic patterns; however, overlaps within the categorisations, while brought to the minimum, are possible. To overcome the challenges of a quantitative study, I discuss some qualitative highlights after presenting the results, displayed in absolute frequencies (AF) and in percentages (%). The percentages are based on the year (%Y) and on the 2013–2016 period as a whole (%P); the aim is to see whether there are changes across the years.

Table 5.14 indicates the trends for each year (summarized in Fig. 5.5). These show that the newspaper headlines tend mostly to put the blame upon the male offenders, increasing from 50% in 2013 to 70.58% in 2016 (this being a positive sign in how the crime is narrated). In relation to the male perspective, the motive of jealousy seems to be higher than the men’s emotions except in the case of 2016, where the latter is given slightly more focus than the former. The blame on the women across the years is the category with the lowest percentages and, while they are constant, in 2016 the percentage is similar to that which indicates the male perspective (14.70%). Interestingly, as far as blame on the woman is concerned, in 2013 and 2016, more focus is put on the actions of the women to report the violent men, while in 2014 and 2015, the headlines describe the behaviour of the woman as a possible cause for the femminicidio .

Table 5.14 Raw frequencies and percentages of headlines divided into ‘blame’ categories divided by each year

In Fig. 5.5, I present the total results of the 4 years in order to discuss the general trend, before presenting some telling headlines.

Fig. 5.5

Percentages of the 6 ‘blame’ sub-categories used to cluster the headlines in the period 2013–2016

This figure shows that, overall, newspaper headlines blame the male offender for the femminicidio . However, there is still some focus placed on men’s jealousy (16%), as, for instance, seen in the following headlines:

OSSESSIONATO DALLA GELOSIA ACCOLTELLA LA EX E CONFESSA

Obsessed by jealousy, he stabs his ex and then confesses [the crime]

UCCIDE PER GELOSIA LA EX A FOLIGNO E POI SI SUICIDA; ROMENO TAGLIA GOLA A CONNAZIONALE, POI SI ACCOLTELLA DOPO FUGA

Driven by jealousy, he kills his ex in Doligno and then commits suicide. Romanian slits a co-national’s throat and then stabs himself after fleeing

These headlines present jealousy as a possible cause for the femminicidio, somewhat justifying the perpetrator’s actions, as also argued in the keyword analysis above.

The category that follows concerns the emotions of the male offender (11.39%); one cannot exclude that this partially situates blame on the women too, for causing the negative emotions of the men which led to the femminicidio. However, I decided to consider what the language was suggesting, rather than impose extra layers of reading. Headlines with a focus on the emotion of the men—mostly their reactions to rejection—are listed and commented on below:

NON ACCETTA SEPARAZIONE, UCCIDE L’EX MOGLIE E SI SUICIDA; ‘SE TE NE VAI TI AMMAZZO’, AVEVA DETTO. CADAVERI TROVATI DAI FIGLI

He does not accept the separation, he kills his ex-wife and commits suicide “if you are leaving [me], I’ll kill you” he said. Corpses found by their kids

BIDELLO UCCIDE PROF. A SCUOLA, ‘FERITO DA INDIFFERENZA’; LA AMAVA IN SILENZIO E LEI NON SAPEVA, OMICIDIO PREMEDITATO

A janitor kills a teacher in a school, “hurt by [her] indifference’; he loved her and she was not aware, wilful murder

«NON LASCIARMI». PAZZO DI RABBIA, DUE SPARI E LA UCCIDE

“Do not leave me”. Crazily angry, two shots and he kills her

The three headlines above attempt to present the perspective of the man who committed the femminicidio, and, specifically, the emotions the men felt and how these were seen as leading to the crime. As also observed by Monckton-Smith (2012) when analysing murder and gender, the fear of being left or rejected highlights the perpetrator’s implied belief of the right to own a partner, ‘ownership’ seen even in instances of love for someone who was not aware of the killer’s feelings.

9.38% of the headlines describe women as adopting behaviours which place them outside of the traditional view of the woman’s role within a heterosexual-couple relationship. Similar to the headlines that describe the emotion of the men, in the following headlines we can see patterns of blaming male offenders for not managing rejection. However, I focus on the action of the woman, as described by the author of the texts, as in:

IL CUSTODE DELLA SCUOLA UCCIDE LA PROFESSORESSA CHE L’AVEVA RESPINTO; RAGUSA LA DONNA AVEVA 54 ANNI, MARITO E DUE FIGLI

The janitor of the school kills the teacher who had rejected him; Ragusa, the woman was 54, she had a husband and two children

VOLEVA LASCIARLO LUI LA ACCOLTELLA PER UN MESSAGGINO

She wanted to leave him. He stabs her for a text

We see that in the first headline, of the story also reported in the headline above, the journalist describes the action of the woman—the woman’s rejection of the man. This raises another point observed from reading the headlines on the topic of femminicidio (and many of those which I have investigated), namely the certainty of how events between the offender and the victim unfolded. This is the reason why exploring headlines, prior to understanding how things could be, has to be seen as a quest for unravelling underlying ideologies; the underpinning beliefs, strengthened by a society that still tends to divide men and women and assign roles and behaviours to them.

Likewise, the second headline seems to construct a larger story than the one that sees the man killing his partner, by suggesting that the woman had the intention (voleva, wanted) of terminating the relationship. This is explicitly linked to the emotion of the man, how he feared rejection and felt entitled to manage the continuity or the end of a relationship, removing the agency of, and power from, the woman.

Slightly lower in percentage (8.29%), one finds the category ‘killer plus the couple’ for those headlines that, while blaming the male offender, describe the dynamic of the couple as troublesome, in the context of their relationship and/or the event(s) that led to the femminicidio, and during their separation, therefore hinting at a shared responsibility, as seen below:

DONNA UCCISA A SARZANA: L’EX MARITO SI COSTITUISCE A MASSA; IEMMA SI PRESENTA DIRETTAMENTE IN CARCERE.TRE COLPI DOPO LITIGIO

Woman killed in Sarzana: the ex-husband turns himself in in Massa; Iemma turns up to the prison. Three shots after a fight

UCCIDE MOGLIE DAVANTI AI FIGLI, SI STAVANO SEPARANDO

He kills his wife in front of their children, they were separating

The focus on the couple is here used to divert the attention from the male offenders and enrich the story of the relationship between them and the victims. In the first headline, it is not clear who has initiated the fight, therefore suggesting that both parties were arguing, and in the second we learn that the couple was separating, conveying a shared willingness to end the relationship.

The last category is ‘woman’s actions vs killer’ and groups together the headlines which recall the actions taken by the victims against those who became their murderers. An example is given below:

UCCIDE LA EX MOGLIE CHE L’AVEVA DENUNCIATO POI SI TOGLIE LA VITA; LECCE, I CADAVERI SCOPERTI DAL FIGLIO 18ENNE

He kills his ex-wife who had reported him to the police. Then he commits suicide. Lecce [the location], the corpses found by their 18-year-old son

This group also sheds light on the inability of the legal and social system to assist those who are subject to intimate violence, not-withstanding that this topic is at the centre of political debate (as shown in the analysis of parliamentary acts) and of activists’ concern.

To conclude, the answer to the research questions is that blame is placed mostly on men but headlines seem to carry within them ideological meanings (and alleged stories) which position the parties involved in the femminicidio within fixed understandings of how women and men should behave and ‘feel’ when part of a couple, attributing, to different extents, the blame for transgressing these roles.

Diachronic Change: Femminicidio Across the Years

To conclude this section on newspaper data, I investigate whether there have been changes across the years under investigation, namely from 2013 to 2016. This can be seen within a sub-methodology of CADS , defined as Modern Diachronic Corpus-Assisted Discourse Studies (MD-CADS), as discussed by Partington (2010). While the corpora under investigation are not large in size (as pointed out by Partington, for this kind of studies), this analysis stems from the aim of “track[ing] changes in […] social, cultural and political changes as reflected in language” (2010, p. 83). The goal is to check whether the debate surrounding the topic of femminicidio, initially initiated by activists, has been assimilated in the media, or whether similarities can be found throughout. The research question addressed is: Are there diachronic changes in the use of the term femminicidio in the years 2013–2016?

To conduct this segment of research, I compare the keywords of each year (2013, 2014, 2015 and 2016) against itTenTen16. I start by looking at how the word femminicidio is used and whether changes to the term’s use have occurred.

As discussed above, the term femminicidio is a recent one and has been resisted by some while welcomed by others. An Italian politician and novelist, Michela Murgia, explains this term, reproducing a commonly heard dialogue:

“A cosa serve chiamarlo femminicidio?” – continuano a chiedere alcuni e talvolta alcune – “La parola omicidio comprende già i morti di tutti i sessi”.

Sarebbe un’obiezione vera solo se la parola ‘femminicidio’ indicasse il sesso delle morte, laddove invece indica il motivo per cui sono state uccise.

Una donna che perde la vita durante una rapina non è femminicidio.

Sono femminicidi solo le donne uccise perché si rifiutavano di comportarsi secondo le aspettative di ruolo che gli uomini e la società patriarcale hanno delle donne.

Dire omicidio dice che qualcuno è morto, dire femminicidio dice anche il perché.

“Why would we call it femminicidio?” – some men and some women as well continue asking. “The term homicide includes dead people of all sexes”

This would only be a real objection if the term ‘femminicidio’ would be used to indicate the sex of the person who has died, while it indicates the reason why they have been killed.

A woman who loses her life during a robbery is not a victim of femminicidio.

Femminicidi concern women who are killed because they rejected the expectations of how women should behave set by both men and a patriarchal society.

With ‘homicide’ we say that somebody is dead, to say ‘ femminicidio ’ is to also explain the reasons why.

Michaela Murgia (2016) ²²

In the tradition of CADS , I use this external source of information together with the number of femminicidi (NF) to identify areas for analysis and interpret the data. Table 5.15, I present the use of this term across the 4 years, in relation to its absolute frequency (AF) and the keyness score (KS).

Table 5.15 Number of femminicidi, absolute frequencies of the term femminicidio and keyness score divided by years

From the table, we can see that the term remains stable in 2013–2014 but decreases in number by the year 2016, most sharply in 2015 (together with the number of newspaper articles published on this subject in this year). While femminicidi continues to occur, the attention of the media and, proportionally, the use of the term seems to decrease irrespective of the work undertaken on this by organisations and the media itself. If we link these results to the lack of the term in Italian law and the penal code (as already outlined in this chapter), it is telling that the highest frequency is found in 2013, when the debate on gendered violence found room in a law (as discussed above), and yet decreases as the years go by.

The Femminicidio of Sara Di Pietrantonio

In this section of the chapter, I investigate a specific instance of femminicidio . This was widely reported in Italy because of the cruelty with which a young woman died. On 29 May 2016, Sara Di Pietrantonio was returning home after spending time with her boyfriend when her ex-partner chased her by car, strangled her and then burnt her body on the outskirts of Rome. The killer was initially given a life sentence having been found guilty of murder, but on recent appeal, this was reduced to murder in the second degree and a 30-year term of imprisonment. Perhaps because of the young age of the victim (in her early 20s) and for the cruelty that led to her death, Italy mourned her to an unprecedented level, with the Camera dei Deputati also dedicating a video to her. ²³ The interest of the media was also considerable. For this reason, I investigate a corpus of newspaper articles on this case. The research question I aim to address is: How is Sara Di Pietrantonio described in the news of her femminicidio?

The Sara Di Pietrantonio corpus (Sara corpus, henceforth) has been compiled on the news database Nexis , which produced all newspaper articles in the week following the crime and the week following the sentence of the first trial. After removing opinion pieces, I was left with 31 newspaper articles which described the act of femminicidio as well as the trial. The number of words for the Sara corpus reaches 14,612. I investigate the corpus through an XML mark-up that was used prior to this study to investigate female victims of male violence in legal sentences in England and Wales (Potts and Formato forthcoming). In those sentencing remarks, male perpetrators pleaded guilty to the murder or manslaughter of female victims. The female victims were either known to the killers—sisters, wives, partners and other relatives—or unknown to them—strangers. The XML mark-up allowed us to search for patterns in relation to nomination strategies and agency of the victims. In practice, this XML mark-up is a way to annotate a text with information which is useful to retrace occurrences of grammatical and discourse phenomena. The text annotated in the Sara corpus is the one referring exclusively to the victim of this femminicidio. I am interested in exploring ways in which Sara is referred to and whether, and to what extent, she has been given a voice.

I explain in detail what the linguistic tags, utilised for the analysis of the Sara corpus, are composed of and how they have been conceived. I expand on the work undertaken on the sentencing remarks, before explaining its adaptation to Italian data. Potts and Formato use the nomenclature of grammatical cases, regardless of English (and Italian in my case) not being case languages, and adapt cases to complements as they appear in the language(s). According to these complements (investigated as cases), I annotated all elements in the newspaper articles that would refer to Sara, i.e. name of the victims, pronouns, ways in which she is described, etc.

The coding was not always straightforward; English and Italian have two very different systems, and adaptations needed to be made, e.g. the coding of the pronoun when it is attached to the verb.

Examples of naming conventions and cases are offered in the tables below, where naming conventions (marked up elements grouped according to their grammatical category, Table 5.16) and cases (the complements that the elements would express, Table 5.17) are adapted from the list of mark-ups used in the English sentencing remarks:

Table 5.16 Naming conventions in the XML mark-up annotation, classified by grammatical elements and examples

Table 5.17 Case tags in the XML mark-up annotation with functions and explanations

Following on from what is presented in these tables, I now show how the text appears without the mark-up (Fig. 5.6) and with it (Fig. 5.7).

Fig. 5.6

Sample of text taken from the Sara corpus before adding the XML mark-up

Fig. 5.7

Sample of text from the Sara corpus after adding the XML mark-up

For the sake of clarity, I exemplify one marked-up sentence, presenting what information it contains and discussing the advantages of using this method:

Ma < actor gender = “f” case = “nom” name = “pro” > lei </actor > stavolta non è tornata

However < actor gender = “f” case = “nom” name = “pro” > she </actor > has not gone back this time

The tags in this sentence carry the information about the gender of the person in the text (gender = f); while I only investigate the female victim, the tag could contain a ‘gender = m’ for men and ‘gender = mf’ to annotate other participants in the data. The case follows (case =), e.g. lei which here could be translated as ‘she’, is marked-up with the nominative, being the subject of the sentence. The final information in the tag is the grammatical element (name =), in the case of lei, the abbreviation ‘pro’ which stands for pronoun. The richness of information of the tags and the study of its frequencies (based on the grammatical element or on the case) provides a systematic way to shed light on insights of the corpus under investigation. To obtain frequencies, I used Sketch Engine (Kilgarriff 2012), selecting CQL (corpus query language) with default attribute tag. This function allowed me to search the tags and have a list of concordance lines.

Naming Conventions

In this section, I investigate the quantitative results and explain them in relation to what was found in the 2013–2016 news corpus. In Table 5.18, I present the number of occurrences and percentages of the naming conventions, divided into absolute frequency (AF) and percentages (%).

Table 5.18 Absolute frequencies and percentages of naming conventions used in the Sara Corpus

As can be seen from Table 5.18, Sara is mostly referred to through categorisation. However, the results do not indicate a substantial preference: the pronominal and nominative naming conventions are very close in percentages to the categorising group. Zero only occurs 5.50% in the corpus. In relation to this, it is to be considered that, while the other categories are present in the corpus, zero has been imposed upon it, namely I have inserted the tag <name = “zero”> in sentences where normally one among pronominal, nominative or categorising would occur. In the next few sections, I investigate the most preferred conventions, starting with ‘nominative’ and then moving on to ‘categorising’. I do not discuss the pronominal choice as the forms in which this occurs in the language cannot provide discursive insights into how the victim has been described.

Nominating Sara

In relation to her given name, and responding to the tag <actor gender = “f”&name = “given”/>, Sara is referred to in three different ways in these newspaper articles, as shown in Table 5.19 (presented in absolute frequencies and percentages).

Table 5.19 Absolute frequencies and percentages of the nominative case in the Sara corpus

In the corpus investigated, the victim is mostly referred to with the first name, followed by the full name, and only once solely by her surname. Some scholars have studied how proximity and distance (Hook 1984; Leech 1999) are expressed through the use of titles, names and other strategies. However, their focus was based on spoken interaction and mainly referred to the interlocutors in conversations. In this case, we have to consider that the person these newspaper articles is referring to is external to the interaction between the newspapers and its readers. I present some concordances lines of the search for the tag <actor gender = “f”&name = “given”/> (Table 5.20).

Table 5.20 Concordance lines of the nominative case in the Sara corpus

Using the first name Sara (80.55%) seems to signal proximity with the referent with an aim to project that proximity to the readers; to make the victim a somebody who could be part of their social circle. The name plus surname strategy (18.75%) is arguably in line with newspapers telling the facts of the story, among which are the people involved. Differing from the use of first name only, this strategy can be seen as distancing both journalists and audience from the victim, giving Sara a legal status: a party involved in and affected by a (non-emotive) fact.

Describing Sara

The aim of this sub-section is to investigate which categories are used to describe Sara. To make sense of the results, I proceed to a further grouping, in order to offer an overview of the ways in which the victim has been described (presented in absolute frequencies and percentages).

Table 5.21 shows that the victim is mainly described according to her age (50%), possibly with the intent of emphasising her youth, and thus creating a specific empathy in the readers who could feel moved by the loss of such a young woman. The term ragazza (girl) is the one that is most used (28.28%), followed by giovane (youngster, 13.15%), terms which do not specify her actual age (conversely done by ventenne and ventiduenne—20/22 years of age) yet give an idea of how young she was, as in Poi Paduano incendia l’auto della ragazza (Then Paduano sets the girl’s car on fire—Paduano being the ex-boyfriend/murderer). Tabbert (2016, p. 76) discusses that mentioning age in news reports could “emphasise [the person’s] vulnerability and innocence”.

Table 5.21 Absolute frequencies and percentages of terms used to describe the female victim in the Sara corpus divided into categories

Unsurprisingly, Sara becomes a specific side of the story in the category ‘murder-related’ through the term vittima (victim) positioning her in the event rather than who she was before this, and, triggering feeling of empathy in the readers. The relationship between the two main parties of the story—the male perpetrator and the female victim—is specified through the terms ex and ex fidanzata (ex-girlfriend), signalling that a decision to terminate the relationship had been made.

In two instances Sara is reported as fidanzata (girlfriend) but only one of these suggests that there was an ongoing relationship between the two, the other instance describing her in a new relationship. This instance and the occurrences of figlia (daughter) enlarge the circle of people who indirectly suffer because of the event (referred to as indirect victimisation, Tabbert 2016, p. 76). The work-related terms, i.e. (studentessa/student), seem to de-personify the victim. She is not merely Sara but a set of social identities and, arguably, social expectations, e.g. somebody who is building her future. This might reflect the pressure to achieve that is put on young women, not just Sara, in many instances.

Investigation of the Cases

Having discussed the naming convention, I now move to explore the marked-up cases. The aim of this section is to further investigate how Sara is positioned within the text in the newspaper articles. In Table 5.22, I present the quantitative results of the investigation (presented in absolute frequencies and percentages).

Table 5.22 Absolute frequencies and percentages of XML mark-up cases in the Sara corpus

While the genitive seems to be the highest in frequency, one has to note that the accusative and the dative are similar in that they position the referent as an object (direct and indirect, respectively), therefore their frequency together (181 in total, 38.34%) is slightly higher than the genitive (164, 34.74%). In the following sections, I explore the nominative and the genitive case to investigate the agency of Sara within her relationship, with elements in the newspaper articles. These are the cases that, when studied in detail, produced the most interesting results.

Nominative Case

I begin with the case in which Sara is the doer of the action, that is the nominative case (<actor gender = “f”&case = “nom”/>). Starting from the 125 occurrences, I performed collocation analysis to see what actions were attributed to her and to what extent they can suggest her agency. The collocations were retrieved through Sketch Engine (Kilgarriff 2012; Kilgarriff et al. 2014) with the function collocation, in the span plus 1 on the right. The cut-off point I considered for the two terms collocating on the basis of Mutual Information score, was set at 3. Results below this cut-off point have been excluded. In Table 5.23, I include the collocations which will form part of the discussion.

Table 5.23 Collocations of the nominative case in the Sara corpus, presented in absolute frequencies and MI score

In exploring the concordances lines of these collocations , I found that sarebbe (would be/would have) expresses the uncertainty of how the facts have occurred, aveva is followed by a past participle that indicates that the victim had left her boyfriend (lasciato/left, 2) or had decided to distance herself from (deciso di troncare/decided to dump, 1) him. The other occurrences show that she had spoken with her friends about the behaviour of her ex-partner (confidato, 2) and reiterated to the perpetrator her decision to terminate the relationship (ribadito/re-affirmed, 1). In another occurrence, aveva (had, in its imperative form) is followed by an adjective, e,g. aveva paura (scared, 1). There is also indication of the new relationship she had started, through aveva (has + past participle) (cominciato/begun, intrecciato/build ties with, 2). These terms are interesting as they seem to portray her agency mainly in relation to the perpetrator, directly but also indirectly (explaining her fear of him or for having a new relationship). Era (was) is used to describe parts of the event (where she was, 4), who she was (figlia unica/only child, 2) and the history of when she was with the person who then became her killer (stata insieme/been together).

Lo (him) describes the relationship between Sara and Paduano and the concordance lines show that she was the one to leave him (lasciato/left, 2, similar to the verb aveva) together with the positive opinion she had of him (lo considerava/Sara considered him). For the most part, we can conclude that her agency is mainly seen in her breaking from, and talking about, her relationship with the murderer. One could speculate that the reoccurrence of this agency in relation to the ex-boyfriend could flag his unacceptance of the decision made by Sara to terminate the relationship. This is not to say that blame for the femminicidio is put on the victim, but certainly, there seems to be a tendency to describe the events from the side of his hurt feelings.

Genitive Case

The annotated occurrences of the genitive present the possessions of the victim (<actor gender = “f”&case = “gen”/>). Its examination is fruitful in describing personal and other links, as shown in Table 5.24. Similar to the nominative case, an MI score of 3 is the cut-off point for the collocations (in the span minus 3 on the left, to take into account that the node would be followed by an articulated preposition).

Table 5.24 Collocations of the genitive case in the Sara corpus, presented in absolute frequencies and MI score

As we can see from the table below, we have objects and people collocating with the victim, mainly in relation to how the criminal episode unfolded, e.g. zio (uncle) and madre and mamma (mother and mum), fidanzato (boyfriend, referring to the perpetrator preceded by ex, and the one Sara currently had a relationship with, preceded by nuovo) as well as the people who participated in the events that followed (funeral, memorials), for instance amiche (friends/fem). Similarly, casa (apartment, house) is a very telling result, as explored in the concordances. All the occurrences refer to a meeting that Sara had with the murderer in her house. This somehow rebuilds the relation between the two, shedding light on what is usually talked about, that is l’ultimo incontro (the last encounter). For violence against women associations and politicians, this refers to when the murderer asks to meet the victim to clarify the situation, this frequently turning into the last quarrel (as shown the analysis of the news corpus) and the murder (which is either contemplated or executed on that day). Women who fear their (ex)partners are asked not to agree to meet for the ultimo incontro (last encounter), somewhat imposing ways in which women are as weaker and the passive party. The word vita (life) here is not concerned with what Sara had achieved in her lifetime, but rather is used to recount her last moments that terminated in her death. As can be seen from the above list, omicidio (homicide/murder) is preferred to femminicidio possibly because this first term is used to refer to a less abstract concept than that to which the actual murder refers.

To conclude this section on the femminicidio of Sara Di Pietrantonio , some interesting and telling results were found through the innovative XML annotation (a step by step guide to will be accessible in Potts and Formato forthcoming), shedding light on how the victim is described in relation to her social identity, in relation to the murderer, and in relation to what she possesses. The corpus is a small yet an interesting one, and further study on bigger corpora or on the investigation of both parts in the crime can offer yet a more detailed investigation of how this and other gendered crimes reproduce gender stereotypes.

Conclusions

This chapter has demonstrated that women are asymmetrically treated in the private , this referring to the space which women have traditionally and historically been assigned to, and, to the dimension of their feelings in heterosexual relationships. I have focused on how this private is brought to the fore in the gendered crime of femminicidio, that is, women killed at the hands of someone who was or had been in a relationship with them. While there are some resisting political and activist voices to how victims of femminicidio are (to be) dealt within the news, there still seems to be an imbalanced perception of the parties involved.

As shown above, to explore this I investigated three different datasets—parliamentary acts, news reports, and news of a specific femminicidio—through corpus linguistic functions and methods. In parliamentary acts (mozioni), MPs deal with violence as a broad and yet abstract phenomenon that impacts women. Left-oriented political parties seem to be more inclined to work together with other institutions, while right-oriented ones tend to focus on what can be done in parliament . As for news reports, the forensic narratives that emerge seem to portray female victims mainly in their roles within the relationship (and only rarely in their working roles) while men are seen in their authoritative roles (yet this professional status in contrast to the crime they have committed). Men are also seen as having lost control of their emotions, their actions (and re-actions) based on based on (alleged) jealousy. News headlines , while mostly blaming men for their actions, also have a tendency to blame the women, by suggesting that the hurt feelings of men was a (mitigating) trigger for the crime. The analysis of news reporting of the femminicidio of Sara Di Pietrantonio , demonstrates that she was described as young (with the inference of, and possibly naïve) and mainly in relation to her agency in breaking up with the person who then killed her.

To conclude, it seems that the recommendations made on how to deal with this gendered crime, have not always been assimilated by the media and a vagueness which, almost paradoxically, reproduces fixed understandings of romantic love and those who practice it, remains. This has to be seen in the context of a heteronormative naturalised discourse and patriarchy which are reproduced, and in some cases, reinforced, in both micro- and macro-levels of beliefs, stereotypes , and understanding about gender. A key example of how this is reinforced and institutionalised is found in the fact that the Italian penal code does not have a definition for the crime of femminicidio, and the implications of this void are clearly seen in how language relating to violence against women is used in the media.

Notes

• 1. La storia ci dice che la guerra è il fenomeno che accompagna lo sviluppo dell’umanità. Forse è il destino tragico che pesa su l’uomo. La guerra sta all’uomo, come la maternità alla donna. (History is suggesting that the war is the phenomenon which accompanies the development of humanity. Maybe it’s the tragic destiny that is loaded onto men. The war equals men, as the maternity equals women.) In this maths-related sentence, Mussolini clearly divides the destiny of women and those of men in two fixed and binary categories (from the Parliamentary Speech, 26 May 1934; da Scritti e discorsi, vol. IX, p. 98).

• 2. In Dobash and Dobash (1998) there is reference to the work undertaken by Rederlechner and Ratz (1993) who explain that women were thought of as a means to create the nationhood by procreating and raising children.

• 3. http://www.unwomen.org/en/what-we-do/ending-violence-against-women. Accessed 24 October 2017.

• 4. http://www.unwomen.org/en/what-we-do/ending-violence-against-women/facts-and-figures, last updated August 2017. Accessed 24 May 2018.

• 5. Italy is a parliamentary republic where laws are discussed and approved by the parliament .

• 6. http://www.senato.it/application/xmanager/projects/leg17/file/repository/notizie/2017/femminicidio.pdf. Accessed 15 March 2018.

• 7. http://www.eures.it/il-femminicidio-in-italia-nellultimo-decennio/.

• 8. There are data for femminicidi in 2017 (114 from January to October), as reported in the news http://www.rainews.it/dl/rainews/articoli/Violenza-donne-nei-primi-10-mesi-anno-114-le-donne-uccise-80381535-efbd-46ed-94d0-8e3223ffbe58.html. In Quanto Donna, an association for women, reports that 33 women have been victims of femmininicio, from January to March, http://www.inquantodonna.it/femminicidi/femminicidi-2018/. Accessed 15 May 2018.

• 9. http://www.senato.it/leg/17/BGT/Schede/Ddliter/47784.htm, https://www.studiocataldi.it/allegati/news/allegato_28553_1.pdf, http://www.altalex.com/documents/leggi/2017/12/27/orfani-per-crimini-domestici. Accessed 15 March 2018.

• 10. http://webtv.camera.it/evento/12258. Accessed 18 May 2018.

• 11. http://www.camera.it/leg17/14?conoscerelacamera=231. Accessed 15 February 2018.

• 12. Diamanti (2014, pp. 12–13) describes M5S as “to some degree, a ‘catch-all’ party that lacks a precise ideological and social imprint, and thus presents itself to a wide and varied electoral audience with diverse expectations and objectives”.

• 13. This forms part of what was discussed in an assembly Non una di meno (Not one [woman] less) organized in Rome on 27 November 2016. https://nonunadimeno.wordpress.com/2016/12/08/report-tavolo-narrazioni-della-violenza-attraverso-i-media/. Accessed 25 October 2017.

• 14. While there is no reference to studies or more specific indication to what this referred to, I believe it is conceived to mean “the (perfect) couple”.

• 15. The other option would be to use a a corpus based approach where the corpus is investigated starting from analysts’ hypotheses on specific terms (as in my work on ministra, ministro and sindaca, sindaco in Chapter 4).

• 16. https://www.SketchEngine (Kilgarriff 2012). co.uk/wp-content/uploads/2015/07/Getting_to_know_2012.pdf.

• 17. McEnery and Hardie (2011, p. 246) define Mutual information (MI) as “a statistic that indicates how strong the link between the two things is. Mutual information can be used to calculate collocations by indicating the strength of the co-occurrence relationship between a node and collocate”.

• 18. http://www.treccani.it/vocabolario/gelosia/. Accessed 3 May 2018.

• 19. For this reason, I consulted another dictionary, Hoelpi, which describes jealousy such as “Sentimento di ansia che nasce dal sospetto più o meno ragionevole di avere un rivale in amore (feeling of anxiety that grows from the suspect, more or less reasonable, that someone has a rival in love)”. Similarly, rivale is preceded by the masculine article un. http://dizionari.repubblica.it/Italiano/G/gelosia.php. Accessed 16 May 2018. The Sabatini Olivetti dictionary has a more generic definition: “Sentimento d chi ha paura di vedersi sottratto l’oggetto del proprio amore” (feelings of who is afraid to see the object of one’s own love stolen). In this definition, the problematisation moves to conceptualising the loved one as an object. While this is not the subject of this chapter, I believe that the point raised on gli and un has to been seen in relation to dictionary constructions. Nossem, who works on queer and lexicography, suggests that: “[l]e decisioni lessicografiche, siano esse frutto di scelte personali e/o prese consapevolmente o inconsapevolmente sotto l’influsso di norme sociali, danno un peso particolare ai valori e agli ideali del_la lessicograf@. Queste decisioni lessicografiche formano, alla fine del processo redazionale, il dizionario, un’opera di autorità e di riferimento. Il dizionario finale promuove poi i valori e le norme in vigore durante la sua redazione, rafforzandoli e fissandoli grazie al suo potere e alla sua autorità” (2015: 122 lexicographical decisions, whether they are originating from personal choices and/or taken consciously or unconsciouly under the influence of social norms, give a specific weight to valued and ideals of the lexicographer. These lexicographical decisions form part, at the end of the editing process, of the dictionary, a work that is deemed authoritative and used for reference. The final dictionary promotes the values and the norms valid at the time of its editing, strengthening and established them because if its power and its authority). I agree with Nossem (2015) when she claims that this can have an impact on how the users perceive meanings.

• 20. http://www.treccani.it/vocabolario/raptus/. Accessed 3 May 2018.

• 21. The highest collocates of ennesimo ranked by their absolute frequencies are: volta (time), dimostrazione (demonstration), potenza (power) conferma (confirmation), prova (proof), tentativo (attempt), episodio (episode), occasione (occasion), and, caso (case). In this case potenza (power) refers to physics. However, the expression ennesima potenza is metaphorically used to mean something that is at its best (capacity). Interestingly, the first set of collocates (volta, dimostrazione, conferma, prova, tentativo) seem to contextualise ennesimo as a negative evaluation for something that has not been achieved before, hinting at the hope that this ennesimo is the last (time, demonstration, confirmation, proof, attempt). While with case, occasion and episode, yet maintaining a negative connotation, it seems to be also related to the future (rather than exclusively the present or the past).

• 22. Available here, https://unaltrogeneredirispettoblog.wordpress.com/2016/11/23/dillo-che-sei-mia-la-trappola-fatale-dellimmaginario-di-michela-murgia/. Accessed 29 November 2017.

• 23. http://webtv.camera.it/evento/12259. Accessed 18 May 2018.

Appendices

Appendix 1: Single Words

In this appendix (and in the following one), I present the list of the first 50 single words (and multi-words in Appendix 2), arranged in descending order of their keyness score.

In the translation column I provide information that could be useful, e.g. whether the past participle is masculine or feminine and therefore referring to a woman or a man; similarly, I have indicated the grammatical gender of some specific nouns (Table 5.25).

Table 5.25 List of single words in the news corpus, presented in keyness score (KS) and absolute frequency (AF)

Appendix 2: Multi-words

See Table 5.26.

Table 5.26 List of multi-words in the news corpus, presented in keyness score (KS) and absolute frequency (AF)