A Critical Discourse Analysis of Women ’ s Portrayal in News Reporting of Sexual Violence

This study explores and compares the portrayal of women in the news reporting of crimes of sexual violence against women between two newspapers from different cultures, the Jakarta Post and the Guardian. The Jakarta Post is an English quality newspaper published in Indonesia, and the Guardian is a quality broadsheet from Great Britain. To explore the representation of women, this study accounts the portrayal of men as well since the two entities are strongly inter-related. The analytical tool used in this study is naming analysis of social actors, which is a part of critical discourse analysis. This analysis is aimed at probing the representation through the choice of lexical items in representing the main news actors. The findings of the analysis indicate that the choices of the naming categories used by both newspapers are different. The Jakarta Post mostly functionalises both the victims and the perpetrators in terms of their legal status in the criminal cases. This suggests that the broadsheet tends to view them as part of the legal processes instead of as people. The Guardian typically classifies the victims in terms of their age and gender and refers to the perpetrators with their surnames instead of as parts of the criminal cases. The Guardian’s tendency to represent both perpetrators and victims as people instead of parts of legal processes indicates that the paper is attempting to focus the reports more on the crimes themselves rather than the participants involved in the cases.


INTRODUCTION
Language and gender are two topics frequently put together as the focus of research.Some of the works related to this research topic are women, men and language by Coates (2004), gender identity and discourse analysis by Litosseliti and Sunderland (2002), language and gender by Eckert and McConnell-Ginet (2003), and feminist stylistics by Mills (2004).Those studies, along with many others, have shown how researchers and experts are interested in exploring the concepts of language and gender.
'Gender' is distinguished from 'sex' because 'gender' is a socially constructed term rather than a biological classification (Coates, 2004).It is claimed as a social construction since the beliefs about gender are inherently in stilled through all aspects of our social life (Eckert & McConnell-Ginet, 2003).Gender is then attached to the discussion of language as there are claims that gender bias is established through language, e.g. the bias is made when the world is portrayed as patriarchal or in favour of males.
Language is strongly related to the contexts of where and for what it is used.This has then shifted the focus of studies on language and gender to a contextualised discourse.One of the genres in which language and gender are explored and enacted is media discourse.In this genre, the controversial gender representation issue is generally related to the portrayal of females.Females' portrayal in the media can be both visual and verbal.The visual one includes the exposing of sensual figures of women that seems to characterise women as sensual beings rather than as intellectual ones.
The verbal one, on the other hand, deals with the use of language in representing women.Poems, novels, songs, editorials, and news reports are among the verbal means in which the representation of women is established.News on sexual assaults against women like rape cases is one example; Clark (1992) has shown that women are sometimes portrayed as the blamed participants of the attacks.This negative representation of women was concluded by Clark (1992) after conducting a study of news reporting in The Sun newspaper (London).This kind of representation has been shown to be valid at least in the context of popular newspapers.What has been happening in some other broadsheet newspapers is investigated in this present study.
This study was an attempt to investigate the portrayal of women in two quality newspapers.This study aimed to extend the research that Clark (1992) did, but this time with a different type of newspapers.In this case, this study also considers the factor of comparing two quality newspapers from different cultures.The broadsheets selected were The Guardian and The Jakarta Post.In general, The Guardian might be seen as representing the quality newspapers in the UK society.The Guardian is widely recognised as a broadsheet with a balanced representation of gender, while the other one might be seen as a representative of Indonesian society which seems to be influenced by Islamic culture as the country is a majority of Muslims.The basic supposition is that there might be some disparities in the way these two quality papers portray the news actors particularly the female ones in news describing crimes of sexual violence against women owing to the diversity in cultural background.
Discussion of gender or portrayal of women cannot be separated from accounts of ideology and power.These two concepts are crucial in the language used in portraying the world.It is widely believed that the linguistic choices made by the writers and speakers of language are ideologically motivated as "language is not neutral, but a highly constructive mediator" (Fowler, 2013, p. 1).In this vein, Montgomery (2008) has argued that language shapes our view of the world, our ways of thinking, and our social interactions.As stated earlier, the field of language and gender is in the scope of Critical Discourse Analysis (CDA) (Fairclough, 2013) since this social practice is closely related to the transfer of ideology.However, the certain ideological value and power embedded in the discourses cannot be easily identified because it is not explicitly expressed.Therefore, critical linguistics suggests that linguistic analysis can make it possible to reveal the hidden ideologies.This is what most of CDA's work is concerned with, including this study.This study explores the language of representation in relation to the use of lexical items in naming the social actors involved in the news reporting of sexual violence against women.This naming analysis can unpack the ideological values packed into the news discourse.

LITERATURE REVIEW
In an attempt to provide sufficient theoretical background of this study to support the analyses, this section highlights some previous literature regarding the portrayal of women in the media, concerning CDA, and the naming of social actors.

The Portrayal of Women in Media
The issue of 'Women in Media' has been widely discussed among researchers and feminists as it comprises three key notions: gender, language, and media.The interface of these notions can be identified through how women are positively or negatively portrayed in the media.The portrayal seems to foster gender inequality where females are commonly exposed as submissive beings that are perceived or portrayed as weak and are the objects of sexuality (Wood, 1994).Females are often illustrated as objectified beings by displaying their body parts that relate to sexual arousal.If men are then drawn to the attractive look of these females, it will be viewed as the women's responsibility.The women will be the ones blamed for the effects caused by such under-rated representations.
This asymmetry can also be found in news discourses where the female actors tend to be over-represented.The fact is that the female social roles are not overrepresented in respectful ways but in opposed ones.In news reports, women are not described in relation to their sensual images but in lexical and grammatical choices used in narrating the news events in which the women are participating.In the news depicting sexual assaults against women in a tabloid like The Sun, the females who are actually the victims of a sexually violent attack tend to be represented as the blameable ones for the assault that was aimed at them.The ideal representation, however, is that the females' social roles are viewed as the affected individuals and the males are the ones who are actually accountable for the assaults (Clark, 1992).
The phenomenon above seems to be shaped by the views put forward by Spender (1985).She has argued that the world is viewed from the male's point of view because language is dominated by man.This infers that the media tend to use language in a "patriarchal order" which is known as the stereotype of Western society (Simpson, 1993).In this order, males seem to be viewed as more influential than women which results in control over the language used.Furthermore, this order indicates that females should be the ones who follow what men want and are represented in ways that are favoured by males.This cultural order might shape the ways that the media frames the information dealing with male and female actors in the news.
The discussions above show how gender bias is encoded in language and then is depicted in the media.As the media is working to serve the public's interest, implicitly these institutions have taken a role in shaping the society's perception of the world.In relation to this, the area of gender representation has become the area of interest of CDA as will be dealt with in the following section.

Critical Discourse Analysis (CDA)
Studies on gender representation have been frequently approached from the perspective of CDA which is related to studies of language and gender "because of its concerns with power" (Jule, 2008, p. 29).Power can be projected through the choices of linguistic resources in projecting meaning and social values.Sameer (2017, p. 140) notes that the focus of CDA is "the relationship between discourse, power, dominance and social inequality and how discourse reproduces and maintains these relations of dominance and inequality".Therefore, the critical linguist tries to explore how linguistic devices are incorporated to permeate power and ideology in certain social contexts because they are ideological.In relation to the primary aim of CDA which is to unmask the use of ideologically charged language, this study tries to reveal the ideology hidden within the language used in portraying the social roles by using naming analysis of social actors.Clark (1992) in her study of The Sun newspaper found that men will be called by bad names if they attack innocent women, while blameless names will be assigned to the men who attack disreputable women.How a social role is portrayed might be affected by how other roles are viewed.This indicates that the ways all the roles involved in a discourse are perceived are interconnected (Richardson, 2007).
The social roles involved in an event can be identified based on some features including occupations, names, ages, appearances and relationships to others, etc. (Leeuwen, 2003).The possible uses of these categories in the discourse of news reporting in The Sun newspaper as investigated by Clark (1992) is that men are typically foregrounded in terms of their occupations, while women are in terms of their sexuality and relationships with others (Mills, 2004).
In line with the concept of naming, Cruse (1977) has suggested a concept of lexical specificity.However, Cruse's focus is on spoken use not on written texts as it is in this study.The lexical specificity is in some ways relevant to what the naming deals with since it attempts to look at how the specificity assigned to nouns implies certain pragmatic effects.However, for this study, the framework proposed by Leeuwen ( 2003) seems more relevant in many ways.His model of naming can be the analytical tool for this study as some of the categories do match with the data used in this study.

Context and Text Selection
The data for this study were obtained from two newspapers, The Jakarta Post (hereafter, JP) and The Guardian (hereafter, TG).JP is the largest quality daily English newspaper published in Indonesia and TG is a broadsheet published daily in the United Kingdom.These two papers come from two different cultures which imply that some differences in using the language to represent women might occur.This study focuses on the portrayal of female news actors in news on crimes of sexual assaults against women.The texts are extracted from each newspaper's online website.All the articles are on the same topic, sexual violence against women.The corpora of the two broadsheets were roughly similar in word count, 5.825 for JP and 5.843 for TG.

Analytical Tools
The analytical tool used in this research is naming.Naming is a tool used to investigate how social roles are portrayed in particular contexts (Leeuwen, 2003).He has proposed a set of categories dealing the representations of social actors.This study employs these proposed categories, but only certain categories that are suitable for the data in this study.The categories incorporated in this study are functionalisation, appraisement, classification, relational identification, and physical identification, formalisation, semi-formalisation, informalisation, obscuration, and titular honorification.
Functionalisation is done when the social roles are named based on their occupations such as 'the teacher', 'the schoolgirl', and 'the prostitute'.The social actors can also be labelled "in terms which evaluate them, as good or bad, loved or hated, admired or pitied" (Leeuwen, 2003, p. 58).This sort of naming is called appraisement which can be achieved by "the set of nouns, adjectives and idioms that denote such appraisal as for instance, 'the darling', 'the bastard' ..." (p.58).These two sets are overlapping in some ways, especially when they deal with functions implying negative appraisal such as 'prostitutes'.The label can be a functionalisation as well as a positive or negative appraisement.
Classification is used when the social roles are characterised or classified in relation to their "age, gender, provenance, class, wealth, race, ethnicity, religion, sexual orientation, and so on" (Leeuwen, 2003, p. 54).The instances of this group are 'a twenty year old woman' and 'a Briton'.Furthermore, when the social actors are identified by using "their personal, kinship or work relation to each other" (p.56), this is called relational identification.This group can be characterized by "a closed set of nouns denoting such relations" as 'aunt', 'daughter ', and 'friend' (p. 56).Another naming group is physical identification where the social roles are categorized based on physical traits which "uniquely identify them in a given context" (p.57).Nouns and adjectives indicating physical characteristics such as 'blonde', 'tall', and 'bearded' are used to achieve the purpose of this category.
The remaining six naming sets deal with unique identities of the social actors.Generally, they are made up by proper nouns which can be formalised, semiformalised, informalised, and obscured with honorification as shown below: Briefly, this study aims to identify how the social actors are portrayed based on the categories outlined above.

RESULTS AND DISCUSSION
This section details the findings of the naming analysis dealing with how the social actors are referred to in the news reports.The focus is on the categories in which the two newspapers mostly based their naming of social actors.In this case, the core participants to be analysed in relation to how they are named are the perpetrator(s) and the victim(s).The naming categories that are used for the model are as set up in the earlier methodology section.Although Table 1 has a lot of data on naming analyses, it will be more useful to bring in some lexical items used in naming men and women into this discussion.Based on Table 1, the categories that are worth looking at through their use of lexical items are functionalisation, classification, negative appraisement, relational and physical identification, formalisation, semi-formalisation, informalisation, and obscuration since they seem to contribute to the greatest extent in setting up the typical representations of the key participants.
The first feature to look at is how the entities are functionalised.As stated earlier, men in JP seem to be generally referred to by using functionalisation.Approximately 69% of the 91 occurrences are of functions in legal terms.Men are frequently referred to by using words like 'suspect', 'rapist', 'suspected rapist', and 'perpetrator'.Those words are mentioned more than once particularly the word 'suspect'.From 63 occurrences, the word 'suspect' is used 44 times to name the men in the whole corpus of the JP.In terms of their jobs, men in the JP are only referred to in 28 occurrences which are only 31% of the whole occurrences.This imbalanced distribution suggests that the JP does not really functionalise men in terms of their occupations but how they are viewed in the legal world.
The typical use of functionalisation in representing men also happens in the representation of women in the JP.The JP tends to view women in relation to their positions in the legal system by using such words as 'victim', 'detainee', and 'client'.
However, the JP does not use as many types of words for women as for men.Only three types of word are used but with many tokens, especially for the word 'victim'.This is used in more than 50% of all the occurrences of legal terms.
The frequency of functionalisation in terms of legal terms is not that significant for both perpetrators and victims in TG (the occurrences were only 12 and nine times, respectively).The words used for men were 'serial rapist', 'sex attacker', 'suspect', 'rapist', and 'attacker', while there was only one type of word used for women: 'victim'.
In the use of classification, the group of victims in both corpora seems to have a considerable number of occurrences even though with various percentages.In TG, classification is the category with the highest frequency of occurrence that is assigned to the victims.The occurrence of classification was 34 times or 50% of the total occurrence of all categories.Although classification in the JP is not in the first position but in the second rank, the frequency of the category was also high.This classification was used in 48 occurrences or roughly 36% of the whole occurrences.The victims in both JP and TG seem to be typically classified in terms of their gender and age such as 'women', 'girls', 'female student', and '19-year-old woman'.The JP, however, also classified the victims based on their nationality but this only occurred five times.The lexical items used in such classification were 'Swedish tourist', 'Swedish teenager', and 'foreigner'.Briefly, both broadsheets tended to classify the victims by both gender and age.
The use of classification in naming the group of perpetrators in both newspapers was not as frequent as in describing the other groups.However, the classification category takes the second position of the total occurrences with slightly different frequency (men in the JP are classified in 27 occurrences, while in TG men are classified 26 times).In this category, the men in both papers are generally classified in terms of gender and age.'Man', 'boys', 'a 52-year-old man', and 'ten-year-old boys' are examples of words that are used in this category.This suggests that both newspapers tend to classify both perpetrators and victims in terms of gender and age.
The next class is appraisement.Only negative appraisement needs to be discussed since positive ones were not significantly used in either of the papers.TG seems to use negative appraisement more than the JP does.TG negatively appraises the perpetrators more than the victims.A number of 14 negative appraisements of men occurred, whilst for women there were only three.Some words that were used in naming the men were 'sex offender', 'homicidal maniac', and 'evil person'.All of these words carry negative values that mirror the dreadful behaviour of the men.There was a striking difference in the way TG viewed men and women in terms of negative appraisement.As shown from the examples of the words used in referring to the men, they were negatively appraised due to their terrible actions.However, regarding the women, TG does not negatively appraise them in terms of their action, but registers their desperate conditions such as 'terrified women', 'vulnerable' and 'naïve'.
There are not many things that can be discussed about the use of negative appraisement in the JP as that newspaper did not appear to appraise the main participants particularly the men for whom there were no occurrences of negative appraisement.On the other hand, for women, there were four occurrences where women were negatively appraised in terms of their miserable fates as in TG such as 'raped student' and 'pregnant rape victim'.
The focus now moves to the category of identification which covers both relational and physical identification.Since the occurrence of physical identification is not significant (only three occurrences of physical identification in the case of men in TG), this discussion is now focused on relational identification.TG seems to refer to the two main entities in terms of relational identification more than the JP does.This is as shown in Table 1 where both men and women in the TG are labelled in terms of their relationship to others 11% and 19% of the overall occurrence of all categories in each group, respectively.The occurrences in the JP are far less than in TG (only 1% for men and 6% for women).TG seems to set up a balanced representation for men and women in terms of their relations to others, while the JP does not really focus on this kind of category in the representation of both actors.The words that are generally used by TG for both men and women in this category are 'daughters', 'sisters', 'their daughters', 'father', and 'the woman's step uncle'.
Another major category is regarding how the two papers referred to the victims and perpetrators in terms of their formal, semi-formal, and informal naming.In this case, it is necessary to clarify that the naming system in the Indonesian context is rather different than the English since Bahasa Indonesia [trans.Indonesian language] does not have any very fixed rules for naming.This is also caused by the fact that many Indonesian people only have one name (without a family or clan name).Based on what is going on in the community, the system of naming in Indonesian context can be classified into two divisions.Firstly, if the people are native to Indonesia, formalisation will be realised by their full name and the naming by their first name or either part of the name will be considered as informalisation.There is no semi-formalisation for the first model.Secondly, if the people are foreigners or native to Indonesia but their names are in English style (first names plus a family or clan name), they will be named in accordance with the English system.
For the category of formalisation, the group of men in TG seems to lead in the frequency of occurrences.Apart from being in the top frequency for the group's own ranks, the frequency of this category is also the highest one amongst the other groups.This indicates that TG tends to refer to men formally.The other indication is that JP does not use formalisation significantly for both groups.
As stated earlier, formalisation is the category with the highest percentage assigned to men in TG.The men in TG tend to be referred to by using their surnames or family names only such as 'Newton', 'Grant', 'Sahodree', 'Malik', and 'Tweed'.On the other hand, for the group of victims, there is only one surname mentioned 'Bowman', but with three tokens.
In the case of semi-formalisation, the group of perpetrators in TG is still the one with the highest frequency of occurrence, while the other groups do not have significant frequency.So, the men in TG also tend to be referred to by using their full name (first name plus surname) instead of surnames only, such as 'David Newton' and 'Shahzad Malik'.
The last category in referring to the participants by their names is informalisation.Unlike the previous two categories, the men in JP seem to lead the frequency of this category.The perpetrators are referred to by using first names only such as 'Deni', 'Anand', and 'Salvador'.Although the percentage does not seem that significant, it is still the highest one among the four groups.While men led the frequency in the formalisation and semi-formalisation in TG, there was no occurrence in TG in this category.This might be due to the different cultural backgrounds.It is unusual for TG to refer to the news actors by first names in the context of crimes.However, it seems fine for JP to do so as long as the ones referred to are the ones that do not need to be respected.
Obscuration, the last category to be discussed, is still related to references to the participants but uses pseudonyms or initials in place of real names, e.g.'Ms.X' or 'Bunga' [trans. 'flower'].This kind of reference is aimed at hiding the entities' real identity for legal or other reasons.The group of participants that has the highest occurrence frequency in this category is the group of victims in JP.The other three groups do not have any significant occurrences and the group of victims in TG does not even have any occurrence in this category.In JP, the victims are referred to by using pseudonyms such as 'Kuntum' [trans.'bud'] and 'Rose', and some are by using initials only like 'S', 'D', or 'CEB'.These choices might indicate that JP tends to protect the victims' identity by not revealing their dreadful experiences to the public, whereas TG is doing what is regulated by British journalism policy which is to state the real names instead of pseudonyms.
Having thoroughly discussed the significant findings from the naming analysis, several factual points regarding the representation of social actors in JP and TG have been found.These findings indicate some salient distinctions in the categories each broadsheet chooses to use to base their references for the key news actors.
First is the distinction in dealing with the preferences of each broadsheet in using naming categories to refer to the group of victims.TG mostly classifies the victims according to their age and gender, while JP tends to both functionalise them in terms of legal terms and classify them in terms of their gender and age.Although the JP's frequency of functionalisation is rather balanced with its frequency of classification, the JP's frequency of functionalisation is far higher than that of TG.These different choices imply that TG tends to considerably personalise the victims rather than to functionalise them as JP does.This indicates that TG does not really view women as the ones involved in and affected by the crimes.TG tends to highlight the violations more than the participants involved by representing the females as part of the legal processes.Meanwhile, JP attempts to represent women by their social roles that are involved in and affected by the men's crimes.JP in contrast to TG puts more emphasis on the participants of the crimes rather than on the crimes themselves.
Another distinctive feature is related to the system of naming used to refer to the victims.The salient distinction here is that there are some occurrences where JP refers to the women with informal names.Although the frequency is not that significant (only three times), this is still worth noting since it only occurs once in TG.It was very unusual to refer to someone by their first name only, in the context of crimes, since the informality is identical with being friendly.However, the JP did this three times.Based on the contexts in which this kind of reference occurred, the three occurrences in the JP all referred to teenagers.Presumably, this was because, in Indonesia, teenagers are normally referred to by their first names only to show that they do not need to be respected as adults.In TG, the reference using a given name did not occur in the main clause but was only a part of the police story about the crime.So, it was not prominent.This confirms that referring to the participants by using the first names in such a context as this is unusual.This seems to indicate that the JP tends to use the normally accepted way of referring to teenagers in daily life and TG did not do this.
A difference between the two groups of victims also occurs in terms of obscuration.The JP had a higher frequency of masking the identity of the women than TG had.This might be due to a distinctive journalism policy of each newspaper.Hence Indonesian journalism tries to protect the identity of the victims by obscuring it using a pseudonym or initial(s).In the British context, journalists have to state the real names of the defendants.This also implies that the JP treated the victims as important participants who needed protection and support.
The next difference relates to the representation of men in both papers.As for the group of victims, the JP mostly functionalises the perpetrators in terms of their legal terms, while TG tends to balance references to the perpetrators in terms of formalisation, classification, and functionalisation.Although both newspapers functionalise the perpetrators, the frequency in the JP was far higher than in TG.On the other hand, the frequency in TG of formalisation was much higher than in the JP.This indicates that each newspaper has its own tendency in focusing the references into certain categories.Through this choice, the JP also wanted to view how the men's actions affected the other participants in the crimes.For TG, it seems that not only the women were by the paper as news actors who did not seem to be participating in the crimes but also the group of perpetrators.By referring to the men mostly in terms of their names, gender and age, instead of their functions in legal terms, TG appeared to situate them beyond the crimes.However, there was a slight difference in the way that TG perceived the victims and the perpetrators.As has been discussed, TG tried to put the women outside the criminal cases by representing them as people not as part of the crimes.Although TG attempted to exclude the men from the crimes as well by referring to them in terms of their names, TG still referred to them in terms of their legal functions but with lower frequency than in formalisation.The men's frequency of functionalisation was slightly higher than the women's.This means that TG was attempting to view the men's functions in the legal processes to a higher degree than that of the women.

CONCLUSION
Having explored the findings of the naming analyses, some conclusions can be drawn.The choices of naming categories used by the two papers were different in both of the newspapers, The Jakarta Post and The Guardian.Each newspaper had its own typical tendency in deciding about which categories the paper wanted to mainly base their references to the main social actors.The Jakarta Post mostly functionalises both the victims and perpetrators in terms of their legal status in the criminal cases.This suggested that the paper tended to view them as part of the legal processes instead of being people.Additionally, The Jakarta Post seemed to expose how the participants involved in the processes were affected.The Guardian had a distinctive tendency which was to typically classify the victims in terms of their age and gender and refer to the perpetrators with their surnames instead of as parts of the criminal cases.The Guardian's tendency to represent both perpetrators and victims as people instead of as parts of the legal processes indicated that the broadsheet attempted to focus the reports more on the crimes themselves rather than the participants involved in the cases.

Formalisation
first name only) Obscuration : Mr. X or Ms. Y (letters or numbers to replace names) Titular Honorification : Prof., Dr., Mayor (the addition of standard titles or ranks)

Table 1 .
The portrayal of social actors in The Jakarta Post and The Guardian.