## **Representations of Child Sexual Abuse in Jamaica**

 A Corpus-Assisted Discourse Study of Popular News Media

## **Representations of Child Sexual Abuse in Jamaica**

i

ii

# **Representations of Child Sexual Abuse in Jamaica**

## **A Corpus-Assisted Discourse Study of Popular News Media**

Authors

**Tatyana Karpenko-Seccombe Kenisha Nelson Christine Fray Roxanne Harvey Karyl Powell-Booth Adele Jones Nadia Wager Xiaomin Sheng**

Basel • Beijing • Wuhan • Barcelona • Belgrade • Novi Sad • Cluj • Manchester

iii


*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

For citation purposes, cite as indicated below:

Karpenko-Seccombe et al. 2022. *Representations of Child Sexual Abuse in Jamaica: A Corpus-Assisted Discourse Analysis of Popular News Media*. Basel: MDPI, Page Range.

**ISBN 978-3-0365-2467-2 (Hbk) ISBN 978-3-0365-2466-5 (PDF) doi:10.3390/books978-3-0365-2466-5**

Cover image courtesy of the None in Three Centre for the Global Prevention of Gender-based Violence (http://www.noneinthree.org).

© 2022 by the authors. Chapters in this volume are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book taken as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

iv

## **Contents**



## **References 73**

## **About the Authors**

#### **Tatyana Karpenko-Seccombe, Ph.D**

Tatyana is a Senior Lecturer at the University of Huddersfield teaching academic skills to doctorate students. She holds an MA in Language Studies and a Ph.D. in Linguistics from Lancaster University. She collaborates as a Linguistics Expert at the None in Three Centre for the Global Prevention of Gender-based Violence, based at the University of Huddersfield, UK. She uses corpus linguistic methods and discourse analysis extensively in her research and has published several articles and book chapters. Her other interests lie in the area of teaching academic English with corpora and concordancers.

#### **Kenisha Nelson, Ph.D**

Kenisha is a psychologist and full-time lecturer at the University of Technology, Jamaica. She is also a Research Assistant at the None in Three Research Centre, Jamaica. Kenisha has a Ph.D. in Psychology with an emphasis in Occupational Health Psychology from Cardiff University, Wales. Kenisha's area of research interests include stress and well-being, gender-based violence, and help-seeking behaviours for mental health related problems.

#### **Christine Fray, Ph.D**

Christine is an Associate Professor in the Dietetics and Nutrition Department of the College of Health Sciences, at the University of Technology, Jamaica (UTech) and Country Director of the None in Three (Ni3) Research Centre, Jamaica. She has been a Registered Nutritionist with the Ministry of Health since 1999 and a Licensed Master Sports Nutritionist (National Association of Sports Nutrition, San Diego) since 2008. She has a BSc. in Biochemistry and Zoology (Parasitology & Physiology); an M.Sc. in Nutrition and a Ph.D. in Nutrition from the University of the West Indies (Mona). Her areas of research are nutrition economics and gender-based violence.

#### **Roxanne Harvey, B.Sc.**

Roxanne is a Research Assistant with the None in Three Research Centre, Jamaica. She holds a Bachelor's degree in Child and Adolescent Development with a minor in Disabilities Studies from the University of Technology, Jamaica. Roxanne has worked in research aimed at reducing bullying among high school students in Jamaica, with focus on administration of surveys and data analysis.

### **Karyl Powell -Booth, Ph.D**

Karyl Powell-Booth is a Lecturer and former Programme Director at the University of Technology, Jamaica, in the Child & Adolescent Development course of study. She has a Ph.D. in Psychological Medicine which she pursued at the University of Glasgow, in Scotland, UK. She is also the Research Fellow at the None in Three Research Centre, Jamaica, here at the UTech. Additionally, Karyl is an Associate Clinical Psychologist with over 15 years of experience working in the field of mental health. The focus of her Ph.D. thesis was on factors associated with suicide risk and self-harm in Jamaica. A recipient of one of the top prizes for First Year Post Graduate Researcher Students' Presentation in 2016 at the University of Glasgow, she also has presented at several local, regional, and international conferences on her work on mental health issues including self-harm, gender-based violence, and child sexual abuse.

### **Adele Diana Jones, Ph.D**

Adele Jones OBE is Emeritus Professor of Social Work at the University of Huddersfield, UK, and Founder of the None in Three Centre for the Global Prevention of Gender-based Violence where she was formerly Director. A Fellow of the British Academy of Social Sciences and a qualified Social Worker, Prof Jones has conducted research in over 23 countries and is the author of numerous publications on topics such as child abuse, adoption, residential care, migration, parental imprisonment, child refugees, gender-based violence, and HIV-AIDS. Notable among these is a series of three books on child sexual abuse in the Caribbean. Published by Palgrave Macmillan and credited as the first scholastic texts on the topic in the region, the series addresses research and theory, policy and programming, and clinical practice.

#### **Nadia Wager, Ph.D**

Dr Nadia Wager is a Professor in Psychology at the Teesside University. Nadia specialises in understanding the causes and consequences of gender-based violence, particularly in the context of preventing and intervening to ameliorate the harm caused by sexual victimisation, including child sexual abuse. She has conducted research and service evaluations for organisations such as the Independent Inquiry into Child Sexual Abuse, the Centre of Expertise into Child Sexual Abuse, and Victim Support.

### **Xiaomin Sheng, Ph.D**

Xiaomin Sheng is a postdoctoral researcher at Tsinghua University in China. She obtained her Ph.D. from the School of Education and Professional Development, University of Huddersfield, and worked as a researcher at the None in Three Research Centre for the Global Prevention of Gender-based Violence. Her research focuses on sex and relationship education, educational strategies on violence prevention, and curriculum development.

## **Acknowledgments**

We would like to thank Mark Davies for creating a wonderful resource – the NOW corpus – and for his help with making it work for us. We are also very grateful to Hilary Nicholson of Women's Media Watch for supplying some data for this study.

## **Chapter 1. Introduction**

### *1.1. Social Construction of Childhood and Child Abuse*

Social constructionism is a theoretical perspective that refers to the process by which people describe, understand, or in other ways account for social phenomena or reality (Gergen 1985). According to Muehlenhard and Kimes, 'social constructionists are concerned with examining the words that people use and the ways in which people understand the world, the social and political processes that influence how people define words and explain events, and the implications of these definitions and explanations—who benefits and who loses because of how we describe and understand the world' (Muehlenhard and Kimes 1999, p. 234). Hegemonic conceptualizations of childhood, drawn primarily from cognitive developmental traditions, tend to define childhood as 'the lack of adult power and capacity' (Pasura et al. 2013, p. 201), reflecting the interests of those who proffer the definition (Muehlenhard and Kimes 1999), rather than of children. Yet, the ways in which childhood and, by extension, child abuse are constructed have important consequences for policy and practice (Holloway and Valentine 2000). Narrowly defined terms influence social acceptance of what is deemed normal or unacceptable behaviours (Muehlenhard and Kimes 1999). Pasura et al. (2013) refer to two contrasting ways of conceptualizing childhood in the literature as the plurality versus singularity debate. The plurality of childhood approach considers 'the social, economic, cultural and historical contexts within which children are embedded' (Pasura et al. 2013, p. 202), whereas the singularity of childhood emphasizes a single, universal, or unchanging construction of childhood as a social category. The universality of childhood has been challenged in more contemporary constructions of childhood. It is contested that no true or single definition of any concept, including childhood, exists (Muehlenhard and Kimes 1999; Pasura et al. 2013). This position may be different from the one taken by international agencies whose focus is on child agendas that emphasise fixed, stage-based biological characteristics in defining childhood (Prout and James 2015). However, the lack of a globally agreed definition of child strengthens the plurality perspective and highlights the relevance of local and cultural considerations in describing and accounting for childhood experiences.

Changes and inconsistencies in definitions and conceptualizations of childhood or child, legally or otherwise, also give rise to specific social constructions, rather than universal meanings. For instance, while in some countries child labour laws are strictly upheld, in other nations children might be expected to take on economic responsibilities at an early stage in order to facilitate their development. These varying cultural underpinnings and ideas of childhood make the implementation of universal principles and policies problematic, such as some of those contained within the United Nations Convention on the Rights of the Child (UNICEF 1989), despite this

being the most widely ratified human rights treaty in the world (Norozi and Moen 2016; Biddle 2017). This plurality versus singularity approach to the construction of childhood and abuse is evident in most fields of inquiry concerning children. Reisig and Miller (2009) for example, identify challenges that can emerge for migrant populations who move to other societies with parenting practices that differ from their own. Reisig and Miller explain that migrant populations may have difficulty integrating into countries where their child raising practices are scrutinized within the context of mainstream societal or cultural beliefs and practices, especially where these are based on different values and traditions. Importantly, migrant families may face legal implications given differences between what constitutes appropriate and inappropriate behaviours in raising children.

What is distinguishable is that, in most societies, children are viewed as being separate from adults, regardless of context and time. Inherent in this 'separateness' is the dominant notion that children are helpless and vulnerable and need to be kept safe from the dangers in the adult world. Jones (2001) argues that how the 'otherness' of children is managed is an important factor in the relationship between childhood and adulthood within societies. Jones further suggests that, rather than being viewed as *less* than adults, children should be viewed as *different* from adults, thus enabling childhood to be considered in more meaningful ways. Similarly, more recent debate has argued for the notion of the child as a 'social actor'. That is, the view of children as being able to exercise active agency over their own lives, rather than simply being the object of actions by others (Prout and James 2015; Kjørholt 2005). This aligns with the constructionist approach adopted by the authors of this monograph and who are of the view that 'children are not formed by natural and social forces but rather that they inhabit a world of meaning created by themselves and through their interaction with adults' (James et al. 1998, p. 28). Hence, as Pasura et al. (2013, p. 201) argue, 'the "socially constructed child" is a local, rather than a global, phenomenon and tends to be extremely particularistic' and although CSA is a universal problem, its manifestations and effects are mediated, therefore, through the interplay of specific social, environmental, historical, and cultural factors. Though children are its victims, how CSA is portrayed is dependent on the standpoints of adults within these contexts and the meanings of childhood they promulgate. Regardless, then, of interventions that promote children's empowerment and agency, the disparate power relations between adults and children mean that it is adult voices, rather than children's experiences, that influence how abuse is framed within popular discourse. In turn, it is the adult voice that determines who is held accountable for CSA, what determinants are important (and what are overlooked) and what actions are needed to prevent it.

#### *1.2. Global Prevalence of Child Sexual Abuse*

The World Health Organisation (WHO 2017, p. vii) defines child sexual abuse (CSA) as 'The involvement of a child or an adolescent in sexual activity that he or she does not fully comprehend and is unable to give informed consent to, or for which the child or adolescent is not developmentally prepared and cannot give consent, or that violates the laws or social taboos of society'. Child sexual abuse is considered as a widespread, global problem (Mathews and Collin-Vézina 2019; Singh et al. 2014). The World Health Organization (2020) estimates that up to 1 billion children aged 2–17 years, have experienced some form of maltreatment within the past year, including sexual abuse. In a review on child maltreatment by Moody et al. (2018), findings reveal that sexual abuse was the most common form of maltreatment. Recent data from Ligiero et al. (2019), show that of 24 mostly high- and middle-income countries, the prevalence of sexual violence against children ranged from 8% to 31% in girls and 3% to 17% in boys aged under 18 years. For low-income countries, the percentage of girls who experienced sexual abuse before age 18 ranged from 3% to 35%, while the percentage for boys ranged from 3% to 21% (World Health Organization 2020).

Among Caribbean countries, CSA is a growing concern. Prevalence rates are difficult to estimate due to under-reporting (Jones and Jemmott 2009), but similar to international figures, females/girls are the primary victims of the incidents reported. For the period 2007 to 2012, incidents of child sexual abuse among ten Caribbean countries showed a gradual increase of 762 in 2007 to 2558 in 2009. A slight reduction was observed in 2010, when the number of reported incidents totalled 2401. In 2011 however, the number of reported incidents increased to 3910.

These statistics reflect what is common among most studies, that reports of CSA are higher for girls compared to boys. Indeed, Hinds and Giardino (2020) note that the overwhelming number of victims of sexual assaults are females. However, prevalence rates among boys may be affected due to the reluctance of boys and men to report abuse. In fact, some studies show that when boys have the opportunity to self-report confidentially, they reveal comparable prevalence rates to girls (Boduszek et al. 2017). The findings from a survey of violence victimisation among over 7000 children (9–17 years) in Jamaica support this (Fray et al. 2021). Fray and her colleagues showed that, while the prevalence of sexual abuse among girls was higher than among boys, this was the case only for children who reported CSA they were subjected to *outside* the home and a greater parity in prevalence was found between boys and girls concerning CSA perpetrated by adults *in* the home. A report by UNICEF (2017) highlights the ways in which sexual violence against boys can be misrepresented and, in many instances, contribute to misperceptions about its prevalence. In Jamaica, the reporting of CSA against boys has long been hampered by archaic laws which defined carnal abuse as only pertaining to female victims and by patriarchal values which promote early sexual debut among boys as sexual conquest or 'rite of passage' (Jones and Jemmott 2009; UNICEF 2017; Harris and Dunn 2018). Of note is that the

definition of carnal abuse has now been revised to sexual intercourse with someone (male or female) under 16 years of age (Harriott and Jones 2016). Obtaining an accurate picture of CSA experienced by boys is paramount, as the research suggests that boys are also significantly affected. As with other studies on the subject, the main perpetrators of sexual victimization of children in Jamaica are males and are typically individuals known to the child, such as family members or acquaintances (Hinds and Giardino 2020).

#### *1.3. The Prevalence of Child Sexual Abuse in Jamaica*

Child sexual abuse has been a public health concern in Jamaica for several decades (Kempe 1978). Bourne et al. (2015) documented an average of 550 cases of carnal abuse during the 1970s. This figure increased almost three-fold in the 1980s and 1990s to approximately 1300 (Bourne et al. 2015). Statistics from The Children's Registry, with direct responsibility to receive, record, and store data related to child maltreatment, indicate that for the period from 2007 (from the time of its establishment) to 2017, there was an annual increase in reports of CSA incidents, with figures rising from 121 to 3103 (The National Children's Registry n.d.). One possible explanation for this steady increase in the number of reported cases of CSA is that with the establishment the Children's Registry, the awareness of the importance of reporting was rising with each successive year, and, respectively, the number of reported cases increased too. Data from the Centre for the Investigation of Sexual Offences and Child Abuse in 2016 revealed that the majority of reported cases were among children under the age of 16 years, the age of consent in Jamaica, with 97% of reports involving female victims (Caribbean Policy Research Institute 2018).

A large-scale, island-wide survey conducted between 2018 and 2019 in Jamaica among 7182 children between 9 and 17 years of age explored various forms of child maltreatment, both inside and outside the home (Powell Booth et al. 2021, in press). The findings showed that 15% of the respondents reported experiencing diverse types of sexual violence perpetrated by an adult living in their home, and 35% by an adult from outside the home. These figures may underestimate the true extent of CSA for several reasons. First, some children are likely to be reluctant to divulge CSA, even anonymously in a survey, due to such factors as guilt, shame, or fear of the repercussions for them and their families. Qualitative research (Powell Booth et al. 2021) conducted by the Jamaica team of the Ni3 project into CSA state that according to the Child Protection and Family Services Agency (CPFSA) 'just under 2500 cases of CSA were reported between January and October 2018, in a population of 2.7 million' (p. 15).

#### *1.4. News Media Discourses of Child Sexual Abuse*

Newspapers play a significant role in forming worldviews, in the construction of social identities, and influencing social relations (Fairclough 1995, p. 12). In reporting child abuse, news media can shape public opinion and construct social attitudes to this problem by foregrounding certain events and viewpoints, while marginalising others. 'Real events are subject to conventional processes of selection: they are not intrinsically newsworthy, but only become 'news' when selected for inclusion in news reports' (Fowler 1993, p. 11). Considering the scope of CSA globally and for the purposes of the present study, in Jamaica in particular, it is important to examine how the topic is presented to the wider public and to critically assess the role news media plays in framing child abuse.

The power of media to influence public perception and attitudes towards child abuse has been recognised in previous research (see, for example, Bullock 2007; Saint-Jacques et al. 2012; Davies et al. 2015), most importantly the ability of the news media to draw attention and to raise awareness of a particular issue (Mejia et al. 2012) and, as a consequence, to trigger public, government, and legal responses to the problem and influencing policy making (Weatherred 2015, 2017; Saint-Jacques et al. 2012). Previous studies also noted the power of the press to set an agenda by prioritising particular issues (Saint-Jacques et al. 2012; Lipschultz and Hilt 2014) and focusing public attention on specific events. This point is significant because the media in general, and newspapers especially, remain the main source of information about child abuse for large sections of population (Weatherred 2015; Goldman and Grimbeek 2015; Popovi´c 2018).

Considering this power of the press, it is essential to scrutinize hegemonic discourses around child abuse and to critically analyse 'taken-for-granted' and established viewpoints, because newspapers, alongside other media, not only form and control perspectives on child abuse, but also influence institutional responses and child protection policies (e.g., Thakker and Durrant 2006).

Studies of media reports on child abuse have provided an overview of the way this issue is presented in the UK (Bingham 2019), England and Wales (Davies et al. 2015), the United States (Cheit 2003; Bullock 2007; Payne et al. 2008; Hove et al. 2013; Mejia et al. 2012), Malaysia (Niner et al. 2013, 2014), Canada (Saint-Jacques et al. 2012), Bangladesh (Haque et al. 2020), Hong Kong (Ho and Chan), South Africa (Corbella and Collings 2007), and Australia (Lonne and Gillespie 2014). There are several trends in reporting that emerge from these studies. Previous research has highlighted the prevalence of high profile or extreme cases of CSA in media coverage; among cases of CSA, newspapers give precedence to the most distressing incidents (LaLiberte et al. 2011; Haque et al. 2020; Miller et al. 2014; Lonne and Gillespie 2014; Mejia et al. 2012). Another problem with newspaper coverage is that papers tend to under-report cases of child abuse within the family compared to cases of abuse by perpetrators unknown to the child, the so-called 'stranger danger' (Mendes 2000;

Niner et al. 2014; Cheit 2003; Davies et al. 2015). This contributes to the popular and erroneous myth that children are most at risk of abuse by strangers, rather than by people known to them. One of the reasons for such under-reporting is considered to be the desire to uphold traditional family values (Mendes 2000; Bullock 2007; Niner et al. 2014).

Existing studies identified two main ways in which CSA is covered in the press: as personal stories focusing narrowly on the event of abuse or as a broader societal issue. The former journalistic representation of material has come to be known as 'episodic', as opposed to the latter, 'thematic' representation (Iyengar 1990, 1996). Episodic representation is known as event-based reporting or reports focused on presenting details of separate incidents of abuse, describing victims and perpetrators, or reporting specific court cases or police investigations. Episodic coverage in newspaper reporting of CSA, according to existing research, is prevalent across many countries, for example in Hong Kong news media it constitutes 89% (Ho and Chan 2018), in the USA—80% (Mejia et al. 2012), in Malasia—68% (Niner et al. 2013), in the UK—71% (Weatherred 2015), and in England and Wales—84.6% of tabloid and 63.1% of the broadsheet newspapers (Davies et al. 2015). These studies clearly point to the disproportionate use of episodic framing within the media; however, such representations can lead to constructions and perceptions of abuse that emphasise individual pathology at the expense of socio-cultural factors. For example, they are likely to suggest to readers that cases of abuse are related to individual behaviour alone and can therefore only be remedied by individuals; such reporting often includes implicit and sometimes explicit victim-blaming messages. Furthermore, when abuse is presented as detached from its social causes and its societal roots ignored, the possibilities for community and policy solutions and preventative measures are diminished (Hove et al. 2013; Mejia et al. 2012).

The thematic framing of CSA, on the other hand, contributes to the discussion of the social aspects of child abuse and its underlying causes and encourages a deeper analysis of wider societal responsibilities and potential solutions. Thematic representation in the media is, however, relatively rare (Haque et al. 2020; Mejia et al. 2012; Lonne and Gillespie 2014; Niner et al. 2013; Weatherred 2015; Ho and Chan 2018). To the best of our knowledge, only one study has shown thematic reporting as being more prevalent than episodic reporting (Hove et al. 2013). Hove et al. (2013) looked at the presentation of child abuse in major US newspapers between 2000 and 2008 and found that 75% of all the child abuse stories had thematic frames. The study by Hove et al. reflects a recent shift in media approaches to reporting child abuse from 'an individual-level problem with individual-level solutions to a societal-level problem with institutional culpability' (Weatherred 2017, p. 3); however, thematic representation is still considerably less common than episodic. The reason for the prevalence of episodic coverage could relate to the concept of newsworthiness, one key factor of which is attracting the readership by reporting negative news

(Galtung and Ruge 1965, p. 69). The episodic representation of negative content, according to Galtung and Ruge, describes 'an event with a clear interpretation, free from ambiguities in its meaning'; they argue that this approach is preferred by the media to thematic discussions which may contain inconsistent, complex, or more ambiguous implications (Galtung and Ruge 1965, p. 66).

The Jamaican mass media plays a major role in forming public opinion about child maltreatment, in influencing specific constructions of childhood, abuse, and victimhood, and in highlighting or obfuscating culpability (Brown 1990). However, despite the importance of this topic and the scale of child abuse in Jamaica, the issue of coverage of child abuse in the Jamaican press has received scant attention. Yet, studies of newspaper discourse on child abuse can establish how balanced the coverage is, how supportive it is of victims, which aspects of child abuse are foregrounded, and which issues are hidden or pass unnoticed. While much has been written about CSA in Jamaica, in taking a novel approach to the analysis of media-constructed representations of child abuse, and the implications these representations have for children affected, the authors provide new insights into a longstanding problem. By scrutinising the popular press, we assess the ways in which specific representations of CSA influence readers' collective unconscious buy-in to hegemonic views on childhood and particular social constructions of abuse.

## **Chapter 2. Methods**

## *2.1. Corpus Analysis of Discourse*

Previous research into newspaper reporting on CSA has commonly been conducted using thematic analysis, a qualitative method based on close examination of the textual content of complete articles and their context to identify underlying themes.

The present study takes a different approach, employing a combination of discourse analysis and corpus methods to study the language of CSA reports in the press to provide new insights into constructions of child abuse. Within the general focus of this study on the representation of CSA in Jamaican papers, the study attempts to answer the following research questions:


Discourse analysis is an established method of social research. It joins the study of language to the social context of its production enabling an examination of the ways ideologies are manifested in discourses and, the role discourses play in creating, legitimising, challenging or reproducing particular worldviews through use of language (Fairclough 1995, 1996, 2003; van Leeuwen 1996, 2008; van Dijk 2001, 2013).

A corpus approach enables researchers to collect and process large amounts of language data and analyse the frequencies and salience of particular words and the context in which they occur. This can give unique insights into the way ideologies and attitudes are entrenched within dominant discourse. 'Corpus linguistics is a powerful methodology—a way of using computers to assist the analysis of language so that regularities among many millions of words can be quickly and accurately identified' (McEnery and Baker 2015, p. 2). Corpus data highlight the recurrence of particular lexical items and point to emerging patterns of usage which are then analysed linguistically by looking at concordance lines, surrounding text (co-text), extended context, and, where necessary, at whole texts so as to reveal important features of attitudes and ideologies embedded in language.

In other words, corpus analysis provides the empirical basis for understanding how language is used to construct a particular view of the world, thus providing 'a higher degree of objectivity—that is, it enables the researcher to approach the texts (relatively) free from any preconceived notions' (Gabrielatos and Baker 2008, p. 7). Corpus analysis is essential in avoiding over-interpretation of data (O'Halloran and Coffin 2004); the application of corpus methods diminishes bias in the analysis of discourses. However, corpus methods cannot be considered completely objective because 'human analysts [make] choices at every stage' (McEnery and Baker 2015,

p. 246). Corpus methods should not be considered as purely quantitative because researchers need to make sense of corpus results, and this involves 'qualitative, functional interpretations of quantitative patterns' (Biber et al. 1998, p. 5). The combination of computer-generated results with manual intervention and analysis is the essential feature of a corpus approach: 'the researcher is normally required to analyse hundreds of lines of concordance data by hand, to identify wider themes or patterns in the corpus which are not so easily spotted via collocation, key word or frequency analysis' (Baker et al. 2008, p. 277). Thus, qualitative analysis of corpus data has become an indispensable stage subsequent to data analysis. It is at this point of interpretation that corpus linguistics comes close to discourse analysis (Mautner 2009, p. 43). Corpus methodology has been increasingly used in social sciences not only to quantitively support qualitative discourse analysis, but 'to do justice to both' (Baker et al. 2008). A seminal study of the representation of refugees in the UK papers has outlined the principles of using corpus methods in critical discourse analysis research, deeming it 'a useful methodological synergy' (Baker et al. 2008, p. 273). Other studies combining a corpus approach with social sciences are the studies of discourses around Muslims (Baker et al. 2013b; Hart 2011), elderly people (Mautner 2007), mental illness (Balfour 2019), and gender stereotyping (Baker and Brookes 2021).

There is a substantial body of research using corpus-linguistic approach to study newspaper discourses. One notable strand in corpus studies of newspaper discourse is the representation of refugees, immigrants, and asylum seekers in the British press (Baker et al. 2008; Gabrielatos and Baker 2008; Hart 2011). Research has also focused on newspaper representation of Islam in the British press (Baker et al. 2013a), the reaction of press and social media to ideologically motivated murder (McEnery and Baker 2015), British press coverage of the Snowden case (Branum and Charteris-Black 2015), and representations of the notion of the Arab world in Arab and UK newspapers (Partington 2015a).

Corpus methods have also been used to explore the language of sexual crime, e.g., the language of online child grooming (Lorenzo-Dus and Kinzel 2019; Lorenzo-Dus et al. 2020) and the representation of rape (Tranchese 2019, 2020) and gender-based violence in the British media (Tranchese and Zollo 2013). However, corpus-linguistic research into representations of child abuse in newspapers is rare and limited, to the best of our knowledge, to one study by O'Keeffe and Breen (2007) who investigated newspaper articles from *The Irish Times*, between 1998 and 2000 on clerical and domestic abuse and uncovered differences in the author stance in reporting abuse in these different contexts.

The approach used in this study can be described as corpus-assisted. Partington defines Corpus-Assisted Discourse Studies, or CADS, as 'the investigation and comparison of features of particular discourse types, integrating into the analysis, where appropriate, techniques and tools developed within corpus linguistics'

(Partington 2010, p. 88). This study uses the corpus-assisted approach because it places a strong emphasis on a balanced combination of computer-generated results and manual intervention.

The research generally followed the steps suggested in Baker et al. (2008, p. 295) and was geared towards answering research questions about the way children and CSA are portrayed in Jamaican newspapers. It starts with contextual analysis of the topic, as informed by the recent comprehensive study of child abuse in Jamaica conducted by the team of researchers within the framework of the None in Three Research Centre (www.noneinthree.org). The results of this research, presented in a report '*It Affects You For a Lifetime!' Perspectives on Child Sexual Abuse in Jamaica* (Powell Booth et al. 2021) have been instrumental in investigating the broad context surrounding issues of sexual child abuse in Jamaica and helped to identify a set of words for corpus research in line with the research questions. This stage was followed by constructing a corpus of Jamaican papers for 2018, 2019, and 2020 ('*JN18-20'*) which was used to build a specialised subcorpus based on the set of keywords connected with *child*. Next, the corpus was investigated with the help of the corpus tool Sketch Engine (Kilgarriff et al. 2004).

In conducting this research, the following steps were followed:


## *2.2. Data*

The dataset for this research was collected from the NOW (News on the Web) corpus containing data from web-based newspapers, magazines, and other sources from twenty different countries which was used as a basis for building a corpus of Jamaican papers (*JN18-20*).

In building a corpus, several considerations should be taken into account: the type and the size of the corpus should be determined by the research question. In

order to look into the most recent discourses around childhood and child abuse in Jamaican papers and to capture existing media attitudes to these issues, a specific corpus was compiled by extracting from the NOW corpus all files that contained news items from Jamaican news outlets for a three-year period from January 2018 to December 2020.

Another requirement of a corpus is its representativeness. The NOW corpus is one of the most comprehensive existing corpora with an estimated size of approximately 12.1 billion words (March 2021) covering texts from twenty English-speaking countries, including Jamaica. It is a 'monitor' corpus which is updated every day by 6–10 million words, or 2–3 billion words each year (Davies 2021, p. 584). It uses automated scripts run every day to add texts to the corpus by collecting relative URLs from Google News (before July 2019) and Bing News afterwards (Davies 2017). The creator of the corpus, Mark Davies, explains the process thus:

Every hour of every day, Google News was queried (using a simple search, like all texts with the words *the* or *to*, which would find essentially all articles) to find online newspaper and magazine articles that had been released in the previous 60 minutes. This search would be repeated for each of the twenty different English-speaking countries. In mid-2019, this procedure was modified. . . . I moved to Microsoft Azure Cognitive Services to collect the URLs. Every day, I retrieve a list of new magazine and newspaper articles (from any source) from the previous 24 h, for each of the twenty English-speaking countries. In addition, each day I query Bing to find new articles (from the previous 24 h) for 1000 specific websites. (Davies 2021, p. 584)

Thus, the rigorous process of the data collection used in compiling of the NOW corpus makes it currently 'the only large (10+ billion words) "monitor corpus" of English' (Davies 2021, p. 585) containing all available newspaper articles from the searched sources. The working corpus of Jamaican papers (*JN18-20*) was compiled for this study from the NOW corpus by collecting all the newspaper articles tagged as 'jm' (Jamaican sources) available for that period in the NOW corpus. These sources comprise the *Jamaica Gleaner, Jamaica Observer, Jamaica Star Online, Loop Jamaica*, and two government sources (Government of Jamaica and Jamaica Information Service). Sources are represented in different numbers for different years depending on the availability of these papers at the time of data collection.

One challenge in creating a representative corpus of news items was presented by the fact that the government sources were included in the NOW corpus alongside the online newspapers. The Government of Jamaica and Jamaica Information Service (a government news agency) are designed to 'disseminate information that will enhance public awareness and increase knowledge of the policies and programmes of the Government of Jamaica' (https://jis.gov.jm/government/agencies/jamaica-

information-service-jis/, accessed on 1 March 2019) making the discourses of government news outlets intrinsically different from those of popular papers. Thus, the decision was taken to eliminate all government sources from the corpus, leaving the newspapers only.

The resulting corpus (*JN18-20*) comprises all the available articles from Jamaican newspapers contained in the NOW corpus covering the period from 2018 to 2020. Thus, the corpus is representative of the language used in Jamaican newspapers over these three years. It contains 9,377,187 tokens and 7,723,794 words and comprises 13,346 texts. The corpus contains articles from three main Jamaican newspapers (e-paper versions which are the same editions as the printed copy): *The Gleaner*, *The Star*, and *The Observer* (http://bit.ly/epaper-login, accessed on 1 March 2019) (see Table 1). According to the latest 'All Media Survey' by Market Research Services Ltd. (2017), Kingston, Jamaica these papers have the highest share of potential readership in the country. *The Gleaner* and *The Star* were mentioned in Surlin and Soderlund's book *Mass Media and the Caribbean* in 1990 as the most prominent print media newspapers.


**Table 1.** Corpus sources.

Source: Table by authors.

*The Gleaner* (Sunday–Saturday) dominates the market with the total number of readers reached by the paper estimated by the survey as 477,000 out of 825,000 (58% of newspaper readers). *The Star* (Monday–Saturday) reaches 210,000 (25.4%) and *The Observer* (Sunday–Saturday) 212,000 (25.6%). These numbers take into account overlaps in readership which means that the same readers can read several newspapers across different groups.

The readership of these three papers is spread almost equally between the male and female population. The number of readers of all three papers similarly tends to be higher for the 20–34 group, closely followed by the 35–49 and 50–64 groups. All three papers also show similar distribution of readership by area, with the overwhelming majority of readers being from the Kingston Metropolitan Area,

constituting about two thirds of readership for all papers. Next for all three titles come rural readers, followed closely by urban readers outside Kingston. The only tangible difference between the reader demographics of the papers is that the readership of *The Gleaner* and, to a lesser extent, *The Observer* have a higher proportion of readers from upper and upper-middle socio-economic groups, whereas the readers of *The Star* are predominantly from the low-middle and low socio-economic group (see Table 2).


**Table 2.** *The Gleaner*, *The Observer*, and *The Star* readership.

Source: Adapted from 'All Media Survey' (Market Research Services Ltd. 2017, p. 173).

*The Gleaner* was described by Aggrey Brown, writing in 1990, as the only Jamaican 'paper of record' (Brown 1990). *The Star* is affiliated to *The Gleaner* and can be seen as a 'tabloid'; its socio-economic readership profile differs from that of *The Gleaner* (see Table 2). *The Observer*, on the other hand, is a privately owned company. The corpus also contains news items from *Loop Jamaica*, a fast-growing digital media news platform launched in 2014 which has increasingly become an important source of online news for the public. Thus, the corpus of Jamaica papers built from these four main online news sources reflects the existing exposure of the public to the news. The balance of the news items as they appear in the NOW corpus has been preserved in the Jamaica corpus because the main focus of the current study was on dominant discourses in the presentation of child abuse in the news, rather than on a comparison of coverage from different sources. The study focuses on the cumulative effect of news media in creating and disseminating particular discourses, the effect that Gerbner et al. (1986), in research related to television, described as 'the cultivation of shared conceptions of reality among otherwise diverse publics' due to 'accumulated total exposure' (p. 9).

For a more targeted investigation of the presentation of CSA in Jamaican newspapers, a specialised sub-corpus was built following Baker's suggestion on building corpora for special purposes:

'One consideration when building a specialized corpus in order to investigate the discursive construction of a particular subject is perhaps not so much the size of the corpus, but how often we would expect to find that subject mentioned within it'. (Baker 2006, p. 27)

The sub-corpus was constructed from the general Jamaica newspaper corpus 2018–2020. Using Sketch Engine tools, the texts containing the words *child/children, boy/s, girl/s, youth/s,* and *pickney/s* (a Jamaican term for *child*, https://languages.oup. com/google-dictionary-en/) were identified and collected in a smaller sub-corpus (*JN18-20 Child+)* containing 619,121 tokens, or 509, 957 words, and constituting 6.6% of the original corpus. The list of words for the sub-corpus was based on the research into CSA in Jamaica (Powell Booth et al. 2021).

## **Chapter 3. Results**

## *3.1. Discourses of Children and Childhood*

## 3.1.1. Word Sketch of Child

Ambiguities and inconsistencies in the understanding of child and childhood in Caribbean nations have implications for addressing sexual violence against children (Pasura et al. 2013). As the newspapers play a significant role in forming concepts of child and childhood in the minds of their readers, it is essential to establish how children in general are framed in newspapers. To answer the first research question related to representations of children in Jamaican newspapers, the lexical context in which the word *child* appears in the articles was investigated first, because vocabulary reflects ideological representations of the world in texts (Fairclough 1996, p. 94).

A collocational profile of the noun *child* was established using Sketch Engine tools. Collocation is a statistically significant association between words which co-occur in a corpus more frequently than is possible by chance. Strength of association between co-occurring words is an important consideration. There are several ways of measuring this; Sketch Engine uses the typicality score *Log Dice* to measure the strength of the collocation between the two words: the higher the score, the stronger the collocation. Strong collocation indicates that the node and the collocate co-occur habitually and not by chance. A low score means that the words in the collocation also frequently combine with many other words. Recurrent collocates indicate the way in which ideas are typically expressed in the language of a particular corpus.

The first step in exploring collocations of the word *child* was to obtain a 'sketch' and analyse it. The 'sketch' function in Sketch Engine (www.sketchengine.co.uk, accessed on 02 October 2020) groups words according to the grammatical relationship between a collocate and a node (or a search term) which allows Sketch Engine to combine 'an analysis of both grammatical and collocational behaviour' (Walter 2010, p. 435). Collocations were derived with a cut-off point of 10, meaning that only words with a frequency above 10 in the corpus were taken into account. Figure 1 shows a visualisation of the ten most significant collocates of *child* in the corpus *JN18-20 Child+* in four categories: verbs collocating with *child* in the position of subject and object; nouns modified by *child*; and adjective modifiers of *child*. The size of each bubble is representative of frequency, the closeness to the node word *child* shows the strength of connection of the two words in the corpus.

meaning that only words with a frequency above 10 in the corpus were

taken into account. Figure 1 shows a visualisation of the ten most

significant collocates of *child* in the corpus *JN18-20 Child+* in four

categories: verbs collocating with *child* in the position of subject and

object; nouns modified by *child*; and adjective modifiers of *child*. The

size of each bubble is representative of frequency, the closeness to the

node word *child* shows the strength of connection of the two words in

the corpus.

**Figure 1.** Visualisation of *child* collocates in the corpus. Source: Figure created by authors using the corpus tool Sketch Engine (Kilgarriff et al. **Figure 1.** Visualisation of *child* collocates in the corpus. Source: Figure created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

2004). This sketch provides an initial view of the salient connections of the word *child*, indicating the contexts in which children are most commonly described. *Abuse*, for example, is the most common single-This sketch provides an initial view of the salient connections of the word *child*, indicating the contexts in which children are most commonly described. *Abuse*, for example, is the most common single-word collocate of *child* in the whole corpus with the highest typicality score of 11.4, indicating that the connection of these words in the corpus is stronger than between any other words.

**22**  word collocate of *child* in the whole corpus with the highest typicality The sketch of the nouns modified by the word *child* shows that children are commonly described in the context of danger or abuse (see Figure 2). Out of the six most frequent and statistically significant nouns, three are connected with mistreatment and harm (*abuse, labour*, and *pornography*), while two are related to death (*mortality, McCool,* the surname of five murdered members of one family). A sketch of adjective modifiers showed that out of the 17 strongest modifiers of *child* with minimum frequency of occurrence of 10 (see Table 3), 14 adjectives provide a functional description of children from the point of view of age (*young, little*), nationality (*Jamaican*), social position (*needy, school*), and status in the family (*first*,

*second, only, own*). Three adjectives describe children as being in a potentially dangerous or abusive situation or susceptible to suffering from harm (*missing, vulnerable,* and *innocent*). adjectives describe children as being in a potentially dangerous or abusive situation or susceptible to suffering from harm (*missing, vulnerable,* and *innocent*).

score of 11.4, indicating that the connection of these words in the

children are commonly described in the context of danger or abuse

(see Figure 2). Out of the six most frequent and statistically significant

nouns, three are connected with mistreatment and harm (*abuse, labour*,

and *pornography*), while two are related to death (*mortality, McCool,* the

surname of five murdered members of one family). A sketch of

adjective modifiers showed that out of the 17 strongest modifiers of

*child* with minimum frequency of occurrence of 10 (see Table 3), 14

adjectives provide a functional description of children from the point

of view of age (*young, little*), nationality (*Jamaican*), social position

(*needy, school*), and status in the family (*first*, *second, only, own*). Three

The sketch of the nouns modified by the word *child* shows that

corpus is stronger than between any other words.


**Figure 2.** Nouns modified by child (Sketch Engine). Source: Figure created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004). **Figure 2.** Nouns modified by child (Sketch Engine). Source: Figure created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

**23**  These collocates of the word *child* give a first indication of how children are framed in the newspapers. A word sketch, however, is 'a somewhat "broad-brush" approach' (Baker et al. 2013b, p. 260). In order to look in more detail at the way children are commonly described in *JN18-20 Child+,* the adjective and noun collocates were explored using a combined qualitative and quantitative approach. Adjectives were chosen because they are 'the most typical vehicle for characterizing in English' (Jeffries 2007, p. 64); they provide a description as well as an evaluation of an entity and thus are the most obvious indicator of representations of *child*. Patterns in their usage can yield useful results in revealing how a *child* is constructed in newspaper discourse. There are several ways of categorising adjectives, such as whether they express emotions, opinions, evaluations, or none of these; subjective and objective (Di Carlo 2015); prototypical and less prototypical (Givón 1993, pp.

62–64); or evaluative and non-evaluative (e.g., Camiciottoli 2013). Givón, for example, identifies prototypical adjectives as denoting the 'inherent, concrete, relatively stable qualities of entities', whereas 'less prototypical' designates more temporary, less concrete states, signalling evaluation and subjective preference. In this study, a similar distinction is made between descriptive adjectives, which give the factual account of the objective characteristics of a child, and evaluative adjectives, which express feelings, emotions, attitudes, and opinions, thus adopting Thompson and Hunston's broad definition of evaluation as 'the expression of the speaker or writer's attitude or stance towards, viewpoint on, or feelings about the entities or propositions that he or she is talking about' (Thompson and Hunston 2000, p. 5).


**Table 3.** Word sketch of *child*: modifiers.

Source: Table by authors based on the results derived by using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

#### 3.1.2. Adjectival Patterns in Representation of Children and Childhood

The next section examines the modifiers used throughout the corpus to identify the way children are presented in the papers. For more comprehensive coverage, together with *child/children,* other alternative words are considered: *boy/s, girl/s, youth/s,* and *pickney/s*.

There are several patterns in which adjectives are used to modify a noun. This section considers adjectives used both attributively and predicatively, including predicative adjectives in an attributive *who*-clause:


• *child/ren, boy/s, girl/s, youth/s, pickney/s* + who + [be] + adj.

The starting point is an analysis of the frequencies of different categories of modifiers, following Baker's observation that '[o]ne of the most basic ways in which corpus-based analysis can reveal something about discourse or attitudes is by considering the frequencies of particular words or related sets of words. Frequency can be an indicator of markedness' (Baker 2010, p. 125). The frequency data of particular attributes were used in this study to explore the context in which *child* (*boy/s, girl/s, youth/s,* and *pickney/s)* is used in Jamaican newspapers because 'frequency data can help to give the user a sociological profile of a given word or phrase enabling greater understanding of its use in particular contexts' (Baker 2006, p. 46). This section further moves towards qualitative analysis by looking into the ways corpus data can reveal information about dominant discourses. In this section, some elements of van Leeuwen's framework of representing social actors are used, which is useful for understanding how meanings are constructed in media (van Leeuwen 1996, 2008). Van Leeuwen's approach is based on Bernstein's recontextualising principles, which describe the way the addressers can 'selectively appropriate, relocate, refocus and relate to other discourses' (Bernstein 1990, p. 184). By using different ways of presenting people and their actions, an event can be 'refocused' in a particular way in a news item. Van Leeuwen suggests a comprehensive taxonomy of the ways in which social actors (i.e., categories or groups of people) can be represented in media, for example, in a generic or specific, personalised or impersonalised way, as a group (collectivised), or as individuals. This taxonomy is used in analysing the presentation of children and child abuse in the corpus.

To examine adjectival patterns, a concordance search in Sketch Engine was conducted on the node words *child/children, boy/s, girl/s, youth/s,* and *pickney/s* using the part-of-speech (adjective) context of one token to the left of the node word, as adjectives are commonly placed immediately to the left of the nouns. The search showed 2078 instances of use of modifiers, all of which were manually examined. The irrelevant cases (456) were discarded after closer context investigation. These were the cases where node words functioned as modifiers: *elite boys school, community youth club.* Titles of films, books, shows, or songs, as well as metaphoric and idiomatic expressions such as *education is a forgotten child of the Jamaican family* or *makes Jack a dull boy* were also disregarded. The remaining 1622 modifiers were then examined and classified into key themes (see Table 4).


**Table 4.** Collocational profile of *child/children*, *boy/s*, *girl/s*, *youth/s*, and *pickney/s*.


**Table 4.** *Cont.*

Source: Table by authors based on the information derived from authors' corpus by using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

Some explanation is needed in relation to the classification of adjectives *little* and *poor*, which both appear twice in the 'Examples' column. In considering the co-text in which *little* is used in the corpus, it became clear that not all the 94 instances of *little* are related to age. The investigation of the meaning of *little* in the context of Jamaican newspapers showed that it refers to age in 28% cases (26 occurrences), but in 72% of cases (68 occurrences) it is used in an diminutive meaning (affectionate or patronizing), thus supporting Stubbs' comment that *little* can have 'different uses and collocates' (Stubbs 1995) and can 'connote cuteness' (Stubbs 2001, p. 312). In this meaning it collocates twice as frequently with *girl/s* (53) than *boy/s (*23).


Thus, 68 cases when *little* was used in this meaning were classified as 'emotive', the rest 26 cases were classified under the category of 'age'.

*Poor* was also used both in the descriptive meaning of 'deprived' or 'impoverished' (7 instances) and the evaluative meaning of 'deserving sympathy' (4).

(3) *how much further damage has been and will be done to the poor child*

*Other* occurs in the meaning *'*remaining children in a group mentioned or described before' expressing some indefinite quantity and was, therefore, classified under the 'Quantity'. *Another,* meaning 'one extra child added to indefinite quantity' was also added to this category.

The collocational profile of the words *child/children, boy/s, girl/s, youth/s,* and *pickney/s* (Table 4) shows that age appears to be the most common characteristic, with 471 instances constituting 29% of all adjectives and noun-phrase modifiers.

Most common among them are *young* (150), *little* (26), and *teenage/teen* (43), as well as noun-groups specifying the age of a child (*10-year-old*). It is, of course, important, to state the age of a child, particularly when offences against children are reported. However, surprisingly, in the overwhelming majority of cases, it is the only characteristic of an abused child that is reported:

(4) *The gang is linked to at least eight murders and several rapes including that of a 12-year-old girl*

Quantity is the next most frequent attribute with 246 instances or 15.2% of all modifiers represented by numerals, adjectives (*countless*), determiners (*many* and *most*), or phrases (*as many as 100 children).* Quantification, noted by Baker with reference to the representation of refugees in newspapers (Baker 2006, pp. 78–79, see also Potter et al. 1991; van Leeuwen 2008, p. 50), is one of the main ways children are characterised in Jamaican newspapers. Quantification occurs both in terms of exact numbers and indefinite determiners of quantity such as *most* or *many*, causing different discoursal effects. When exact numbers are used (39, 2.4%) children are quantified and treated as 'statistics'.

(5) *Statistics from the Office of the Children's Registry indicated that during the period of January to June 2018, 813 reports of missing children were made. Of that number, 75 per cent of the children reported missing were females.*

However, the prevalent form (207, 12.8%) is quantification without referring to concrete cases (*countless, many,* or *most*), which makes the discourse about children vague and generic:


Here, even though the problem raised is serious, the presentation is too vague and overgeneralised to have an impact. Use of indeterminate numbers has another effect. It presents children as a group, a mass deprived of personality (collectivization, van Leeuwen 2008, pp. 37–38). Consistent use of such descriptors erases individual features, as illustrated in the following sentence:

(8) . . . *we learned that too many children are experiencing physical, emotional and sexual violence at the hands of family members and people they trust*

This effect is not limited to the use of *many* or *more*. It is also manifested in the use of plural forms: the newspapers more often refer to *children* as a group than to an individual *child*. Out of 6,932 instances of *child/children* in the corpus, the plural form occurs twice as often (4625 times, 67%) as the singular (2306, 33%). However, in 410 (6%) cases, the singular form of *child* is used generically, indicated by the indefinite article:

#### (9) *A child who runs away from home is one who is in crisis*

In this case, *a child* is presented as a general type, rather than individual, or is genericised (van Leeuwen 2008, pp. 47–48).

Thus far, several tendencies have been identified in the way children are portrayed. They are represented as statistics or quantified, as a group without distinguishing personal features (collectivised) or as a general class (genericised). Such representation is counter in the recent document 'Our Children, Our Media: A Guide for Caribbean Practitioners' (Association of Caribbean Mediaworkers 2016, p. 38) which encourages reporters to 'highlight the human stories behind the statistics'. This research shows that presentation of children as statistics still prevails.

Representation of children as a group is also demonstrated in the use of the third person possessive plural pronoun, *our children*, which is common in the corpus. A concordance search on *child/children, boy/s, girl/s, youth/s,* and *pickney/s +* possessive third person pronoun *our* shows that it is the most frequent pronoun (420 occurrences), followed by *their* (403). (Note that the numbers in Table 5 exclude pronouns which do not have possessive meaning (*them youths,* or *him pickney*) and pronouns occurring in the titles of songs, plays, films, etc.)


**Table 5.** Pronouns in the *JN18-20 Child+* corpus.

Source: Table by authors based on the information derived from authors' corpus by using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

#### 3.1.3. The Use of Pronouns

A sketch of the word *child* shows that *our* is the most frequent pronominal possessor of *child* with 317 occurrences and the highest Log Dice typicality score showing the strength of this particular collocation (9.4). *Our* appears consistently in the corpus in quotations from various public figures who make generalised, rhetorically marked, inclusive statements as in example (10):

(10) *The cases of domestic violence, the drive-by shootings in several communities, the killing of our children, and this blood thirst needs more than just a sprinkling of Zones of Special Operation or the scraping up of thousands of young men without charge unconstitutionally*

(Opposition Leader, Dr. Peter Phillips)

Van Leeuwen (1996, p. 39) noted that use of *our* in such cases 'can make an association explicit without naming the resulting social grouping'; in other words, the group of people implied by *our* remains rather vague. Similarly, Wales (1996, p. 58) describes such use as 'usefully ambivalent'. Another point is that frequent use of the possessive *our* creates, on the one hand, a sense of ownership (children seen as possessions) and, on the other, of familial closeness which can be an expression of Jamaica's collectivist culture, but can also be used for political purposes as a rhetorical strategy. The possessive *our* in *our children* also carries the meaning of inclusivity, creating the impression that the speaker is speaking on behalf of the addressees.

(11) *it is our children who are the most vulnerable victims of this widespread contagion* [pornography]

This representation, signalled by the use of *our*, is commonly embedded into the '*we*-group' rhetorical strategy indicating commonality and solidarity (Chilton 2004, p. 201). This style of presentation, however, also impersonalises children:

(12) *the abuse and murder of our children* . . .

The next extract shows how rhetorical use of generic *our* can cause confusion as to who the speaker represents and who is referred to as *we, you*, and *our:*

(13) *As we continue to promote everyday values, we ask you to keep in mind, the role the family plays in the positive reinforcement of values of our children. Remember everyone, not all forms of abuse leave bruises* (Renee Rose, Brand Manager at LASCO, Jamaican holding company with

operations in milk and food products, as well as distributional financial services).

A similar inclusive function is performed by the possessive case of *nation* and *Jamaica. Nation's* collocates with *child/children, boy/s, girl/s,* and *youth/s* in 40% of cases (73 out of 181).


While some level of generalisation is, of course, to be expected in newspaper coverage, the *JN18-20 Child+* corpus data show a distinct prevalence of generalised presentation of children. In a style of language where children are seen as quantities or large groups, the individuality of each child is obscured. Even though factual information about the number of children may be important in a particular context, the fact that quantity and age are the most common attributes of children in

newspaper articles, constituting 44.2% of all modifiers, is revealing of the way children are presented: they are seen in numerical terms as part of the statistics.

The next most frequent group, family status (202 occurrences, 12.5% of all the modifiers, see Table 4), contains kinship identifications of children within a family. It also contains a significant number of quantified representations, as family status is often construed as the order in which children were born (e.g., *first, first-born, second,* or *third child*). This group also demonstrates bias existing in newspapers against children who grow in one-parent families: the use of adjectives *bastard, fatherless,* and *illegitimate* stigmatise these children, reinforcing patriarchal familial stereotypes. A similar trend was noticed by Nicholson (2002) in relation to gender stereotyping in Caribbean newspapers.

Other groups—Nationality, ethnicity, and race; Social status and education; Suffering abuse or being in a potentially dangerous situation; Health and disability; Place of living; and Religion—also contain factual descriptions of children. The language used portrays children in a narrowly functional way, not as individuals but in terms of the roles they play in the family or in society (*affluent, poor,* or *economically challenged)*, their nationality or ethnicity (*Jamaican, Afghan, American, black,* or *white*), place where they live (*inner-city youths, urban, local,* or *deep-rural),* their religion (*Catholic, Christian,* or *Rastafarian*), their state of health (*able bodied, deaf, disabled,* or *healthy*), or educational level (*grade-seven girl* or *college girl)*. Such representation is known as 'categorisation' when social actors are represented as classes of people (van Leeuwen 2008, p. 42) rather than individuals. 'Functionalisation', which is a part of such representation, 'occurs when social actors are referred to in terms of an activity, in terms of something they do, for instance an occupation or role' (ibid, p. 54). As the major groups of modifiers show, such presentation of children prevails in Jamaican papers.

The next group (100 adjectives, 6.2%, see Table 4) contains individual physical qualities or other descriptive characteristics, and defines children in terms of gender (*male* or *female*), body shape (*fat,* or *taller*), or an outward distinctive feature (*bespectacled*), thus also classifying them on the basis of external physical characteristics.

Thus, the frequencies of adjectives used in the corpus to modify *child/ren, girl/s, boy/s, youth/s,* and *pickney/s* show that the overwhelming majority of adjectives and determiners in the corpus (85.3%, see Table 4) are descriptive and factual. Even though, in some cases, it may be important to mention these characteristics, the predominance of such descriptions of children shows that personalities of children are not in the focus of news media; they are rather presented in a way that reduces them to the roles they perform. This kind of representation has implications for how the public perceive children and may also have implications as to how the readers react to child abuse. In the following example, much more information is provided about the offender than the victim of the abuse ('*another little boy'*):

(16) *Based on information provided by the police, the 16-year-old alleged attacker is not a first time criminal accused, as he has a previous case before the court for allegedly molesting another little boy.*

As mentioned earlier, adjectives can be divided into two groups from the perspective of meaning: evaluative and non-evaluative/ neutrally descriptive (Thompson and Hunston 2000, p. 5). Evaluative and emotional presentation of material is characteristic of newspaper style. It is a rhetorical strategy commonly used in newspaper discourse (White 2004; Bednarek 2006, 2010; Reah 2002). The use of evaluative language is considered to be linked to newsworthiness (Bednarek 2006, 2010; Bednarek and Caple 2017; Harper and Hogue 2015). Against a backdrop of the generally evaluative and emotional journalistic style, the predominance of an impersonal descriptive stance towards children is unusual. In the corpus, the scarcity of personality adjectives points to the fact that children are not described in an individualised way. According to psycholexical research (Ashton et al. 2004a; Ashton et al. 2004b; Roivainen 2013; Saucier and Goldberg 1996), personality-descriptive adjectives are the principal way of describing different dimensions of an individual. In the corpus of Jamaican newspapers *JN18-20 Child+*, only 14.7% (239) of all the adjectives modifying *child/children, boy/s, girl/s, youth/s,* and *pickney/s* are related to children's character and disposition and can be categorised as personality-descriptive. These are evaluative and attitudinal adjectives which are often emotive, for example: a*dorable, best, brilliant, damaged, impressionable, joyful, lovely, malevolent, rambunctious,* or *wayward.* The fact that the number of 'personality' adjectives collocating with *child/children, boy/s, girl/s, youth/s,* and *pickney/s* is significantly lower than other attributes shows a tendency to impersonalise children. Van Leeuwen noted that such presentation 'can background the identity' of social actors (van Leeuwen 2008, p. 47). The importance of compassionate media coverage of abused children was emphasised in the checklist for reporters, editors and programme-makers in the 'Our Children, Our Media: A Guide for Caribbean Practitioners' (Association of Caribbean Mediaworkers 2016). The guide questions the reporters: 'Have you portrayed abused or exploited children as victims, criminals or as human beings with rights and dignity?' (p. 27). The results of corpus analysis show that newspapers still have some way to go to meet this obligation.

Among the group of evaluative adjectives of *child/children, boy/s, girl/s, youth/s,* and *pickney/s,* the collocates *vulnerable* and *innocent* have the highest frequency (within the span of five words to both sides, 44 and 19 occurrences, respectively). In the Word Sketch of *child* described earlier (see Figure 1 and Table 3) these adjectives rank 5 and 6 by the Log Dice typicality score (8.3 and 8.1, respectively) among modifiers of the noun *child*. It is, therefore, important to look into these adjectives further. Previous research has shown that their use is associated with particular discourses of victimhood common in news media representations of those suffering violence and abuse (see, for example, Greer 2007; van Dijk 2009; Hall 2017; Niner et al. 2013;

Ndangam 2003). In his seminal work, Christie (1986, 2018) introduced the notion of an 'ideal victim' who is innocent, weak, vulnerable, and helpless. These features of the victim are typified, building a stereotype. The implication of this stereotype is that the compassion of the media and public can only be extended to victims who comply with the expectations of this narrowly framed 'ideal' construct (Greer 2007; van Dijk 2009; Lewis et al. 2019); these victims are placed by media at the top of a 'hierarchy of victims' (Fohring 2018; Hall 2017; McAlinden 2014). Needless to say, victims in real life can be very different from the constructed 'ideal' victim identity (Christie 2018, p. 20; McAlinden 2014, p. 180; McGarry and Walklate 2015). In the corpus, the relatively high frequencies of *vulnerable* and *innocent* among evaluative adjectives indicate that a one-dimensional view of victims is reproduced by Jamaican papers (see example 17). This tendency is also supported by the occurrences of adjectives from a similar semantic field (*defenceless, helpless, fragile, powerless, voiceless,* and *damaged*) and the adjectives *little* and *poor* discussed earlier. These findings are in line with the conclusion of the 'Our Children, Our Media: A Guide for Caribbean Practitioners' (Association of Caribbean Mediaworkers 2016) which drew reporters' attention to the fact that 'the media often depict children as silent victims or charming innocents' (p. 6).

(17) *In making her commitment to be involved in the work of the charity and social support organisation, Mahbubani* [Montego Bay businesswoman] *said she thought of the "beautiful, innocent and perfect" children of Jamaica, who have no one to advocate for them.*

Extensive existing research on preferences in media coverage agrees that newsworthiness is one of the main reasons for such perspective on the victim (Hall 2017; Greer 2007; van Dijk 2009). However, 'not all victims or offenders of sexual crime hold the same level of newsworthiness and interest for the media or politicians' (McAlinden 2014, p. 182) and the media prefer to report on crimes involving an 'apparently pure and innocent victim' (Lotz 1991, p. 63). This could lead to stigmatizing a non-ideal victim: 'the romanticization of childhood innocence excludes those who do not conform to the ideal' (Kitzinger 2015, p. 148).

Predicative adjectives with 54 instances are not as numerous as attributive (1622). However, they are an important indicator of child framing in the newspapers because, when used in a position of a predicative, the quality attributed to a child becomes foregrounded. The most common group here is also a set of descriptive attributes (35%) such as ranking (*third*), physical quality (*obese, overweight,* or *slim*), kinship (*motherless, fatherless,* or *illegitimate),* and health *(sick)*. Within the category of descriptive attributes, the most numerous is that of children in a potentially dangerous situation (*missing*), at almost 11% of all cases. Evaluative adjectives constitute 22.5%, with positive evaluation (*joyful, happy, bright, well-spoken,* or *strong*)

at 17% and negative (*sassy* or *lippy*) at 5.5%. Once again, descriptive attributes outnumber evaluative.

Next to be considered were *who*-clauses, which function as modifiers to the nouns *child/children, boy/s, girl/s, youth/s,* and *pickney/s* and perform the function of 'referential identification' (Givón 1993, p. 107). A Sketch Engine search for adjectives used in this structure was conducted by choosing the lemma context *who be* and limiting the search to two tokens to the right. As a result, 164 concordance lines were generated with adjectives which were manually investigated in concordance lines and classified into semantic groups (see Table 6).


**Table 6.** *child/children, boy/s, girl/s, youth/s, pickney/s + who* + [be] + adjective.

Source: Table by authors.

Figure 3 illustrates concordance lines containing the words *child/children, boy/s, girl/s, youth/s,* and *pickney/s* collocating with *who + [be] + adjective.* As can be seen from Figure 3, they are predominantly presented in the negative contexts of hardship, abuse, crime, illness, or death (109 instances or 66% altogether), for example *boy who was reported missing, girl who was earlier reported dead, children who are potentially exposed to second-hand tobacco smoke,* and *girl who was found on a bus*. Attributive *who*-clauses which define children positively (e.g., *who are the future*) constitute only 3.7% (6 instances). This imbalance is connected to the well-researched phenomenon

of reporting predominantly negative events. Preferential coverage of negative events is, in Partington's words, 'what newspapers do for a living' (2015a, p. 242). First systematically described by Galtung and Ruge, the factors of newsworthiness include 'reference to something negative' (Galtung and Ruge 1965, pp. 62–72). This tendency for negative representation of children is traced in the corpus of Jamaican newspapers. events is, in Partington's words, 'what newspapers do for a living' (2015a, p. 242). First systematically described by Galtung and Ruge, the factors of newsworthiness include 'reference to something negative' (Galtung and Ruge 1965, pp. 62–72). This tendency for negative representation of children is traced in the corpus of Jamaican newspapers.

Figure 3 illustrates concordance lines containing the words

*child/children, boy/s, girl/s, youth/s,* and *pickney/s* collocating with *who + [be] + adjective.* As can be seen from Figure 3, they are predominantly presented in the negative contexts of hardship, abuse, crime, illness, or death (109 instances or 66% altogether), for example *boy who was reported missing, girl who was earlier reported dead, children who are potentially exposed to second-hand tobacco smoke,* and *girl who was found on a bus*. Attributive *who*-clauses which define children positively (e.g., *who are the future*) constitute only 3.7% (6 instances). This imbalance is connected to the well-researched phenomenon of reporting predominantly negative events. Preferential coverage of negative


**39**  In summary, the analysis of attributes which modify *child/children, boy/s, girl/s, youth/s,* and *pickney/s* and the context in which they occur in the newspaper corpus allows us to conclude that characteristics assigned to *child/children, boy/s, girl/s, youth/s,* and *pickney/s* in papers display consistent (85.3%) descriptive numerical and functional In summary, the analysis of attributes which modify *child/children, boy/s, girl/s, youth/s,* and *pickney/s* and the context in which they occur in the newspaper corpus allows us to conclude that characteristics assigned to *child/children, boy/s, girl/s, youth/s,* and *pickney/s* in papers display consistent (85.3%) descriptive numerical and functional collocations with quantities, terms of kinship, descriptions of race and ethnicity, etc. In contrast, emotive and evaluative 'personality' adjectival collocations constitute only 14.7%.

Representation of *child, children, boy/s, girl/s, youth/s,* and *pickney/s* is characterised by functionalisation, collectivisation, and generalisation, lacking individual features and attributes of identity. Predicative adjectives display similar dynamics with descriptive attributes prevailing over evaluative. *Who*-clause modifiers place *child/children, boy/s, girl/s, youth/s,* and *pickney/s* in the negative contexts of hardship, abuse, crime, illness, or death.

The effect of such representations on the readers is cumulative: 'writers constantly both communicate their own evaluative attitudes but can also seek to impose overtly or covertly, particular values and stances' (Partington 2015b, p. 280). It is argued in this study that the notable scarcity of evaluative identity attributes modifying *children*, together with the preference for selecting numerical and descriptive lexical items, as well as placing children in newspaper reports in the contexts of abuse and danger, disclose the newspapers' attitudes to children which

are then imparted to the readers: such editorial choices present to the readers 'a partial view of the world' (Fowler 1993, p. 11).

Frequencies of lexical items are particularly important for understanding these attitudes because the 'central factor influencing what readers understand and remember (i.e., their interpretations) is the frequency of specific collocations and the semantic/discourse prosodies they communicate' (Gabrielatos and Baker 2008, p. 20). The results support the findings of the 'Our Children, Our Media: A Guide for Caribbean Practitioners' (Association of Caribbean Mediaworkers 2016) which states that '[t]he way in which the media represents, or even ignores, children can influence decisions taken on their behalf and how the rest of society regards them'.

These findings are important because they emphasise the fact that media can play a significant positive role in changing this one-dimensional representation of children; 'by presenting positive, empathetic images of children and young people, the media may have a powerful influence in preventing, rather than perhaps indirectly promoting, child maltreatment' (Saunders and Goddard 2002, p. 4).

#### *3.2. Discourses of Child Sexual Abuse: Who Is in the Focus*

The Word Sketch of *child* discussed in Section (Figures 1 and 2) shows that *abuse* is the most frequent collocate premodifier of *child* (80 occurrences) with the highest typicality score (LogDice 11.36). This associational strength of *child* and *abuse* means that these words do not occur together in the corpus accidentally which, in its turn, points to the fact that papers dedicate a substantial amount of coverage to the topic of child abuse. This section looks at the ways the papers approach this topic—who are the main agents and whose voice is represented in the coverage.

Our main interest was in examining instances of child sex abuse (CSA), thus in order to make the search more targeted, the node words *child/children, girl/s, boy/s, youth/s,* and *pickney/s* were searched for collocates of *abuse, sex, sexual, sexually*, and *rape*. An important consideration at this point was choosing a span for the search. Even though '[t]here is no 'standard' within corpus linguistics circles with regard to what such a span should be', the commonly used span is four or five words either side of the node (McEnery and Gabrielatos 2006, p. 42). A larger span is usually not recommended to avoid showing words which do not directly collocate with the search term (Baker 2006, p. 103; Brezina 2018, p. 273). However, a narrow span has its weaknesses: 'it can lead to an omission of important meaning associations' (Brezina 2018, p. 273) and could 'disregard a good deal of potentially useful context' (Baker 2006, p. 103). Narrow span is 'less sensitive to co-occurrences over larger stretches of text' (Thornbury 2010, p. 273). Some studies used a large span of up to 25 words on both sides of the node to 'capture the maximum possible co-occurrences' (O'Halloran 2010, p. 572); this allows the researcher to investigate fuller context, and is particularly useful for discourse studies because it can 'cast new light on the workings of discourse' (Virtanen 2009, p. 1058). This study is interested in the

discourses around the broad issue of child sexual abuse in the corpus of Jamaican newspapers; thus, in order to capture and analyse all the co-occurrences of the words connected with CSA, a broad setting of 15 tokens (the largest available in Sketch Engine) on both sides of keywords was chosen. This approach allowed us to identify the sentences in which the node words were associated with *child*, even if they were quite distanced from each other, as in example (19):

(19) *The overall findings of the investigative unit indicated poor nutrition, segregation and abuse of some of the country's poorest and most vulnerable children*

The collocations of the node words *child/children, girl/s, boy/s, youth/s,* and *pickney/s* with *abuse, sex, sexual, sexually,* and *rape* allowed us to capture a wide variety of word combinations connected with CSA, e.g., *sexual assault, sex offence, sexually abused, sex predator, sexually molest*. The search returned 479 instances of such collocations, all of which were manually analysed. Duplicates and instances not related to child abuse (e.g., *substance abuse* or *sex of the child*) were discarded. The main focus was maintained on CSA, which is why instances not directly related to CSA were excluded (e.g., *verbal abuse or emotional abuse).* The cases when the use of the word *abuse* was likely to include sex abuse were included, for example, *children now face greater risk of all forms of abuse* or *home for girls who are abandoned, orphaned, abused and neglected.* The final number of 369 instances was analysed to answer three questions of interest which were identified within the second research question: 'What discourses surrounding child abuse are prevalent in Jamaican newspapers?':


Examination of concordance lines alone appeared to be insufficient for answering these questions; therefore, a larger context (several paragraphs on both sides of the keyword, about 200–300 words) was scrutinised in most cases. For some questions, particularly the question of 'voice', it was necessary to extend context investigation to the whole article. Each cooccurrence was examined separately, and, therefore, the numbers discussed below are the cooccurrences of *child/children, girl/s, boy/s, youth/s,* and *pickney/s* with *abuse, sex, sexual, sexually,* and *rape*, which can cooccur several times in the same sentence or appear in different sentences, paragraphs, or news items. Thus, there are four cooccurrences in the sentence (20) *abuse and rape of our girls and boys.*

The emergent results were grouped according to three perspectives: generic or specific presentation, thematic centre of the sentence, and the voice.

## 3.2.1. Generic vs. Specific Trends in Representation of Child Abuse

In presenting events and their participants, one of the most important choices made by the journalist is the choice between generic and specific representation.

Participants (such as children in the corpus) 'can be represented as classes, or as specific, identifiable individuals' (van Leeuwen 2008, p. 35), and actions, too, can be described 'in very general or vague terms or in very specific and precise terms' with a varying amount of detail (van Dijk 2008, p. 826). These differences have implications for the way participants and events are constructed in a discourse. The distinction between covering a particular case of child abuse (*episodic presentation*) and the description of general issues (*thematic presentation*), as well as the implications of this division have been discussed in a growing body of literature on newspaper reporting, beginning with the influential work by Iyengar in 1990 and 1996. As discussed earlier, most existing studies, which used content analysis of news articles, have consistently noted that newspapers favoured episodic or event-focused representations based on case reports (Davies et al. 2015; Niner et al. 2013; Mejia et al. 2012; Weatherred 2015, 2017; Ho and Chan 2018; Cheit et al. 2010).

This research is different in that it employs corpus methods and focuses on individual occurrences of *child/children, girl/s, boy/s, youth/s,* and *pickney/s* collocating with the words associated with CSA and on the context in which they appear in the corpus. Even though the results of this study are not directly comparable with the research mentioned above, the division of context into episodic and thematic is useful. The study of the distribution of generic and specific representations of child abuse in the corpus of Jamaican newspapers shows that *child/children, girl/s, boy/s, youth/s,* and *pickney/s* collocates with *abuse, sex, sexual, sexually,* and *rape* in the corpus *JP18-20 Child+* CSA are presented in a generic way in 72% of cases (264 co-occurrences) and in a specific (episodic) way in 28%, with 105 co-occurrences.

Examination of collocations in context showed that generic presentation of child abuse prevailed, being used two and a half times more frequently in comparison with episodic or discussion of concrete cases. Corpus results show a complex picture of generalised representations which have different implications for the way the message is framed; these implications are discussed in the sections to follow. Generic (thematic) representation is often limited, as illustrated by the examples, to broad statements (20) and statistical accounts (21) of CSA.


Thematic representations at times frame child abuse as a societal problem which can be an indication of emerging recognition of violence against children as a public issue. Concrete ways of dealing with the problem are sometimes stated:

(22) *The Centre for Investigation of Sexual Offences & Child Abuse is a branch of the Jamaica Constabulary Force which, among other things, has been working closely with schools.*

However, such instances are rare and generic representation is often limited to broad rhetorical declarations rather than careful analysis of the underlying causes of child abuse issues.

In all generic statements (which constitute 72% of all representations of child abuse in the corpus) children are collectivised (van Leeuwen 2008), or seen as a class with no identifying features, a perspective also apparent in the findings of the previous section on the modifiers of *child/ren*. Episodic presentation is based on relating the instances of concrete abuse.

The same situation or event can be viewed from different perspectives, with some parts of information highlighted or backgrounded through the management of topic–comment information structures (Tomlin et al. 1997, pp. 83–92; van Dijk 2008, p. 826). Topic or theme is the information centre of the sentence; theme, in Halliday's words, is the 'point of departure' (Halliday 1994, p. 37). It is similarly defined by Tomlin et al.:

. . . each clause in a discourse contains one key element—the *theme*—which is somehow the central referent at the moment, a *point of departure* for the clause, the referent about which the remainder of the clause predicates something. (Tomlin et al. 1997, p. 83, highlight in the original)

If topic/theme is broadly defined as the subject matter of the sentence, comment is something that is said about the topic.

To examine which information is placed by the papers in the position of information centre, this section looks at the topics/themes of sentences containing the collocations of *child/children, girl/s, boy/s, youth/s,* and *pickney/s* with *abuse, sex, sexual, sexually,* and *rape*.

According to Halliday (1967, p. 205), '[t]he sequence of elements in the clause tends to represent thematic ordering', i.e., movement from topic to comment. Thus, the theme commonly coincides with the syntactic subject of a sentence or clause (Tomlin et al. 1997, p. 84; van Dijk 1992a, p. 115, Hockett 1958, p. 201, cit. in Brown and Yule 1984, p. 70); 'it is that with which the clause is concerned' (Halliday 1994, p. 37). Themes in a text are indicative of what the author considers as the key point of the utterance; therefore, it is important to identify the theme in order to understand what/who is placed at the centre of the sentences/clauses containing references to child abuse. Table 7 presents the thematic centres of the sentences containing *child/children, girl/s, boy/s, youth/s, pickney/s + abuse, sex, sexual, rape* showing that the emphasis is placed on various aspects of CSA.


**Table 7.** Thematic centres of the utterances containing *child/children, girl/s, boy/s, youth/s, pickney/s + abuse, sex, sexual, rape.*

Source: Table by authors based on the information derived from authors' corpus by using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

These groups of collocates will subsequently be discussed in more detail, starting with generic references.

## 3.2.2. Generic Reference: Emphasis on Institutions

In sentences where child abuse is referred to generically, the emphasis is most commonly placed on institutions: government organisations, charities, and support groups, as well as projects, programmes, events, and campaigns. A quarter of all occurrences (93, or 25.5%) of collocates appear in clauses where the main emphasis is dealing with or talking about child abuse in general terms. A significant number of collocation of *child/children, girl/s, boy/s, youth/s,* and *pickney/s* with words related to CSA occurs in sentences where the emphasis is on child protection organisations and police, for example, the Centre for the Investigation of Sexual Offences and Child Abuse (CISOCA, 23 instances), Child Protection Agency (3), National Child Registry (2), International Society for the Prevention of Child Abuse and Neglect (ISPCAN, 3), and National Plan of Action for an Integrated Response to Children and Violence (NPACV, 6). These are, of course, essential organisations and programmes providing child protection, but the way they are presented in the news switches the emphasis from the child to the institution. These two sentences (23) illustrate such use (themes are underlined).

(23) *Dr Elizabeth Ward, chair of the VPA said the campaign hopes to raise awareness of the National Plan of Action for an Integrated Response to Children and Violence (NPACV), which was tabled in the House of Parliament in June this year. The mandate of the NPACV is to create and maintain a protective environment, supportive of and responsive to the issues of violence, child abuse and maltreatment of children in Jamaica.*

Here, *child abuse* appears in the context of a national campaign with emphases on the chair of the organisation, the title of the campaign, the government plan, and the mandate of the plan.

In cases when issues of child abuse are incorporated in different social contexts, they become 'recontextualised' (Bernstein 1990; Fairclough 2003, p. 139) with some elements being foregrounded and others replaced in a new context. CSA is also mentioned in sentences where the main emphasis is on businesses, as in example (24). JHTA (Jamaica Hotel and Tourist Association) and its president, the donation, and their further commitment to donate all constitute themes in the sentences of this paragraph.

(24) *JHTA President Omar Robinson said last year the group demonstrated its commitment to protecting the nation's children from abuse with a donation of \$1 million to the Centre for the Investigation of Sexual Offences and Child Abuse (CISOCA). In furtherance of this commitment, he said a similar donation will be made this year, as the problem of exploitation in any form will erode and undermine Jamaica's image and that of the country's tourism product.*

While the Board is a benefactor contributing considerable sums of money to the cause of preventing child abuse, the statement is clearly promotional: the piece places the issue of child abuse prevention in the context of *Jamaica Hotel and Tourist Association (JHTA)* thus providing useful publicity, particularly because child abuse could damage the tourist industry. Thus, the message of child abuse is being reframed to further another agenda.

## 3.2.3. Generic Reference: Emphasis on Crime

The crime of child abuse presented in general terms appears in the position of thematic centre in 17% or 64 cases. In sentence (25), for example, abuse is referred to in an abstract way as a crime without a human agent, without bringing into discussion victims or offenders:

(25) *A lot of the* [child] *sexual abuse cases stem from the lack of emotional and financial support.*

In such cases the theme of the sentence/clause is often presented through statistical data:

(26) *Between January and July of last year 598 cases of child abuse were reported to CISOCA, with more than half the amount involving sexual intercourse with a person under 16 years of age.*

Linguistically, generalised meaning is created through the use of particular structures. Most prominent is the preferential use of the plural form, *children*, used 139 times in association with *abuse, sex, sexual, sexually,* and *rape*. The noun *child*

(when not used as a noun premodifier, as in *child abuse*) occurs in the singular form 82 times, 18 of which are used generically with the indefinite article:

## (27) . . . *trained to detect the signs that a child is being abused.*

Thus, reference to a particular *child*, using the singular form, occurs only in about 30% of all cases, whereas generic reference accounts for approximately 70%.

Another characteristic feature of generic presentation of child abuse is the recurrent use of nominal rather than verbal forms of *abuse*. The use of the noun *abuse* typically conceals the agency of the abuser. However, van Dijk (2006, p. 129) has warned against 'ideological (over-) interpretation' of nominalisation which 'should never be described in isolation, but in relation to the text (co-text) as a whole and in relation to the context'. Corpus analysis provides the opportunity for such analysis and allows us to make conclusions about the preferences of nominalised forms based on frequencies of their usage in context. In the corpus, the noun *abuse* occurs 168 times in association with *child/ren, girl/s, boy/s, pickney/s,* and *youth/s*; 163 of these cases (or 97%) have a generalised meaning without reference to perpetrator. In 5 cases, the offender is clear from the context. The verbal form *to abuse*, occurs 45 times, or 3.7 times less frequently. For example, in the sentences (28) and (29) the agent or the perpetrator of the *abuse* and *violence* is concealed.


Sentences with emphasis on the perpetrator occur in 5% of all cases (17 instances, see example (30)) and are often part of police statistics (31):


It is interesting to note that specific references to perpetrators (which are discussed in Section 3.2.7) are 3.5 times more frequent than generic (17 as opposed to 59 instances, see Table 7). One of the obvious reasons for such coverage lies in the newsworthiness of the offender and the details of the crime, which are likely to engage the interest of the reading audiences.

## 3.2.5. Generic Reference: Emphasis on Victims/Survivors

Generic references to victims/survivors constitute 11% of all occurrences. Generic mention of child abuse is often found in official reports, statements, and speeches by government representatives. Collectivised representation of abuse

victims impersonalises the victims/survivors; in such presentations 'no attempt is made to delineate clearly who are victims of sexual crime; what are their individual as well as collective interests, rights and needs' (McAlinden 2014, p. 182).


To sum up, as Table 7 shows, a generic approach to presenting child abuse is prevalent in the corpus, constituting 72% (264 occurrences in the corpus). In the focus of such representations are government institutions, NGOs, and charities, as well as various programmes and projects connected of preventing abuse (25.5%, or 93 occurrences). Preferred mention of crime and victims is also generic: 17%, or 64 occurrences, and 11%, or 40 occurrences, respectively. Such presentations take the focus away from the victims, who are presented in a collectivised impersonalised manner. Victims/survivors of abuse are constructed as a random group, their individuality erased. In a quarter of the occurrences, newspapers report on institutional 'claims of the need for protection of an undifferentiated and abstract class of generic or potential victims as "vulnerable citizens"'(McAlinden 2014, p. 181). For the reader, this kind of presentation carries the message that the victims' personalities and experiences are irrelevant. When presented generically, the crime, the victim, and the perpetrator can be conceived of as 'removed from the readers' world of immediate experience, treated as distant 'others" (van Leeuwen 2008, p. 36).

#### 3.2.6. Specific Cases: Emphasis on Victims/Survivors

In about a third of the occurrences (105 or 28%) the words *child/children, girl/s, boy/s, youth/s,* and *pickney/s* in collocation with *abuse, sex, sexual, sexually,* and *rape* refer to specific instances of abuse. The reports on specific cases still reflect a tendency to exclude or background the victim/survivor of child abuse.

As can be seen from Table 7, an individual victim of abuse is the focus of only 9% of occurrences (34). While, clearly, identification of a child victim is not possible for legal and ethical reasons, anonymized victim/survivor experiences can still be reported; however, newspapers do not commonly engage with victims of abuse, do not personalise them, nor present as individuals. Eleven out of 34 specific mentions of victims are celebrities, which is in line with previous research in presentation of child abuse in newspapers which commented on the newsworthiness of elites (see, for example, Popovi´c 2018; Wilczynski and Sinclair 1999; Ndangam 2003). Discussing the issue of 'discursive domination . . . by specific elite authors' van Dijk noted that it 'may be used to express and convey a distorted view of social events' (van Leeuwen 2008, p. 822, see also van Dijk 1988, p. 87). In the corpus, the experiences of

historical physical and mental abuse were revealed by the singer Dalton Harris (34) and footballer Kyle Butler (35) which saw several mentions in more than one article.


Another characteristic feature of CSA reporting revealed by corpus analysis is that it is mainly the experiences of adult survivors that were reported, with only one case focusing on a young girl.

(36) *A nine-year-old girl who was being sexually abused by her father was saved from his clutches after she made a phone call to Crime Stop. Prudence Gentles, manager of the National Crime Prevention Fund, said that the little girl called Crime Stop after being turned away by others who labelled her as being troublesome.*

Even though in example (36) the emphasis in three clauses (underlined) is on the abused girl, she is presented impersonally; there are no individual traits of the girl in the description—just a factual mention of her age. There is no empathetic account of the girl's experience, how she coped with the trauma, or what support she needed and received. This example is indicative of the way victims appear in news reports. In referring to victims of abuse, factual mention of age and other minimal descriptors often echo the style of police or court reports in which the child victim is portrayed in an impersonal and impassive manner and which reproduces the court setting where '[t]he victim is often invisible and silent' (Kitzinger 2009, p. 83).

The emotive adverbs *brutally* (2)*, viciously,* and *violently* are used only four times (7.7% of all rape reports) to describe an assault. There is little attempt to engage the reader's empathy by describing personal context, trauma, distress, psychological effects, or the ways of seeking assistance. Similar tendencies in reporting were observed in newspapers of other countries. For example, Lonne and Gillespie (2014), in analysing news coverage of child abuse in Australia, note:

. . . reporting of maltreatment incidents tended to be conflated into generalized stories that lacked important details about the children who were harmed and the event context. Hence, in most instances, without vital information, readers were left to infer what had happened apart from the actual incident of maltreatment. (Lonne and Gillespie 2014, p. 843)

Only five reports in the corpus (1.3%) contained a description of the ordeal using the direct words of the survivors, and only one (0.3%) contained a compassionate description of the child victim reporting the crime (38):

(37) *Two months ago, Sergeant Janet Williams-Richards was sitting across from a teenage girl in distress. The teenager was shaking and in shock. Williams-Richards* . . . *said,*

*"I don't know what you have been through, but whatever it is, you will be OK." After being comforted for some time, the girl reported that she had just been raped.*

Typically, in reports on child abuse, when a word referring to victim/survivor is placed in the thematic position in a sentence, the perpetrator often remains unmentioned. For example, out of 45 cases of the use of the verb *abuse*, a third, or 14 cases, are the agentless passive forms *[be] abused*, as in example (38):

(38) *They also visited the Women's Resource and Outreach Centre and Eve for Life, which serves adolescent mothers living with HIV and girls who have been sexually abused.*

This process of passivisation deletes the agent of the abuse because the messages expressed by active and passive voice, 'are slanted in crucially different directions' (Simpson 2003, p. 99); thus shifting attention from the offender onto the victim.

## 3.2.7. Specific Cases: Emphasis on Offenders

In reporting on specific cases of abuse, offenders were identified in 59 cases (16%), which shows that the newspapers in the corpus put the offender in a thematic position almost twice as often as the victim/survivor (34, 9%, see Table 7). The reports often provide personalised information about the offender and the crime, while the victim is either backgrounded or not mentioned at all, as was shown in the previous section. The following examples illustrate such reporting style showing that more information is presented about the perpetrator and crime than about the victim.


Detailed descriptions of the crime and the offender add to the newsworthiness of the article in the same way as the preferential reporting about the abusers in positions of authority (e.g., sports coaches, teachers, or pastors). This is consistent with previous research suggesting that coverage of 'respectable' abusers is much more prominent in newspapers in comparison with the real statistics of such cases (Wilczynski and Sinclair 1999, p. 276) due to newsworthiness of such cases.

(41) *the teacher was arrested on January 12, following investigations that were carried out by detectives from the Centre for the Investigation of Sexual Offences and Child Abuse (CISOCA). The teacher reportedly taught the complainant at a St Catherine-based primary school, and coerced her to engage in sexual activities with him on more than one occasion*

Iyengar (1990, 1996) and the studies that followed highlighted another important aspect of newspaper reporting. Episodic, or event-focused representation is known to ignore societal factors of offending and assigning the responsibility for the crime exclusively to the offender, an 'individual whose actions are divorced from any social causes' (Wilczynski and Sinclair 1999, p. 276), overlooking the broader social context (Lonne and Parton 2014; Davies et al. 2015). Even though generic representation has the potential to address social factors of child abuse (and, in some cases, papers do look into the causes of abuse and the measures needed to protect children), it is not always the case; generalized statements about child abuse prevail; they are often used for rhetorical purposes, reframing the issues of child abuse within the contexts of business and political interests. A generalised way of reporting child abuse in newspapers filters out the real circumstances of the abuse and the victim/survivor's point of view and voice.

### *3.3. Discourses of Child Sexual Abuse: Whose Voice Is Presented*

Thus far, the portrayal of child abuse in broad terms was considered, with emphasis at the cumulative effect of different ways in which child abuse is presented in newspapers. Corpus methods are particularly suitable for the study of such 'accumulated' media coverage (Gerbner et al. 1986). The corpus contains all the available articles which appeared in Jamaican newspapers between 2018 and 2020. However, the accounts of child abuse in these articles were given by different categories of people—journalists, officials, concerned citizens, and, more rarely, by victims and survivors of abuse. This section sets out to investigate who is speaking about abuse. A distinction is made between the journalists writing on their own behalf and those reporting the views of others. Unsurprisingly, in the majority of cases (181 out of the total of 369, or 49%) the voice is that of the journalist; a significant part of the content is mediated through the voice of a reporter, and ultimately, of the news outlet. Other voices reported in news items about CSA constitute 51% (188 cases), see Table 8.



Source: Table by authors based on the information derived from authors' corpus by using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

3.3.1. Institutional Voice: Representatives of Police, Government, Child Protection Organisations, Support and Advocacy Groups, and Charities

The dominant voices in the corpus speaking about child abuse issues are institutional: government representatives, police, support and advocacy groups, charities, and academics. These sources represent 51% (97) of all non-journalistic speakers. This frequency provides a strong indication of 'institutionalised' discourse around child abuse.

(42) *Chief Executive Officer of the Child Protection and Family Services Agency (CPFSA), Rosalee Gage-Grey, is reminding parents, guardians and caregivers of their duty to protect their children, in keeping with the Child Care and Protection Act (CCPA) 2004*

There is no doubt that these are legitimate voices and reliable sources of information for the newspapers. However, as Fairclough noted, 'one striking feature of news production is the overwhelming reliance of journalists on a tightly delimited set of official and otherwise legitimized sources which are systematically drawn upon' (Fairclough 1995, p. 49). Among such sources, Fairclough names government and the police, which is consistent with our data. This apparently positive use of reliable sources has, according to Fairclough, a downside: it can restrict or exclude voices coming from non-institutionalised groups of activists and campaigners or the victims/survivors themselves, resulting in 'a predominantly establishment view of the world' (Fairclough 1995, p. 49).

This tendency for giving preferential attention to institutions and their representatives was noted and criticised in the Code of Ethics of Association of Caribbean Media Workers which stated: 'Resist favoured treatment of organisations or individuals including politicians, business interests or advocacy groups to influence coverage' (Association of Caribbean Mediaworkers 2016, p. 15). Representatives of both these categories appear in the news coverage as confident voices that are entitled to express their opinions and concerns or make a call for action. They are identified and presented as individuals, and their societal role and position is painstakingly reported in the news item. Such presentation is in tune with van Leeuwen's conclusion that '[t]he government is rarely backgrounded or referred to generically and often individualised and nominated' (van Leeuwen 2008, pp. 68–69). The officials are commonly constructed in the newspapers as professional, strong, and confident, which is reflected in the language used to report: *loud and clear, assured, strengthening, ensure, decisively, blasts,* and *made strong comments.* Their feelings are also made clear to the readers: *deep concern and anguish, dismay,* see (43)–(45).

(43) *Listening to the founder of Hear the Children's Cry and child care advocate Betty Ann Blaine, speaking on The Morning Agenda, on Power 106 FM radio, on the morning of Wednesday, August 7, 2019, she again made it loud and clear her deep concern and anguish pertaining to the Ananda Dean's case, and the many other children who have also been murdered.*


In contrast to victims/survivors, representatives of this group are commonly given the opportunity to express themselves directly, through verbatim quotations, which are more common for this group than any other at 17.6% (33 instances of direct speech) for representatives of various institutions. Direct speech makes the message personal and immediate, 'more vivid and perceptually engaging' (Eerland et al. 2018, p. 1). Even more importantly, it puts the speaker at the centre of the discussion, giving him/her an individual voice, which brings the quoted speech closer to the reader. Van Dijk noted that direct quotations foreground the issue and give it credibility: 'statements by prominent news actors may be newsworthy in their own right, simply because they express the interpretation or opinions of important news actors' (van Dijk 2015, p. 152). Lack of direct speech, on the other hand, moves the issue further from the reader, which is the case with victims' voices.

## 3.3.2. Business People, Sports Personalities, Church Ministers, and Celebrities

The voices of business people, sports personalities, church ministers, and celebrities are also quite prominent in child abuse discourse, accounting for 17.6% (33 instances, see Table 8). In the excerpts (47) and (48), for example, the newspaper quotes two prominent business women on the issues of CSA:


Such coverage is consistent with the general tendency in reporting highlighted by van Dijk where in a 'hierarchy of sources' newspapers tend to prefer elites as reliable 'observers and opinion formulators' (van Dijk 1988, p. 87).

In an overwhelming number of cases, representatives of these two categories (institutional and personalities) make generalised comments about child abuse (130

occurrences, 35.5% for both groups, Table 8). These comments are often rhetorical in nature and are characterised by 'collectivized' language (van Leeuwen 2008, p. 38), such as plural forms (*children, girls,* and *boys*), collective and generic nouns (*nation, citizens, Jamaica, our island,* and *Jamaicans*), and an inclusive *we* and *our* (Leech and Svartvik 2002, p. 59) which creates alignment with the audience by merging the speaker and the addressees into one group. A collectivised method of address is also used by the speakers, importantly, 'to imbue them with impersonal authority' (van Leeuwen 2008, p. 69), as in the following example:

(49) *"Jamaica has a deep, underlying issue of sexual abuse," she asserted* . . . *"In order to do something in an effective way, we need to create awareness and this can only be accomplished by sharing the solutions of prevention and education with more people* . . . *"* Montego Bay businesswoman, Kareena Mahbubani

In the speech of celebrities, business and sports personalities, and church ministers reported in newspapers, the phrase *child abuse* often appears with reference to agendas very distant from the actual issues of child abuse. The following expanded context shows that in this article the main emphasis is on investment and business opportunities in sport (in bold, emphasis authors'). Mention of child abuse comes as an afterthought, and the problem is mentioned in passing and, again, in very general terms.

(50) *Broadening stakeholder involvement to deal with issues involving athletes and enabling sporting organisations to develop and operate within a business framework, are key drivers of the Jamaica Olympic Association (JOA) initiative, Sports for Breakfast Forum*. These factors were outlined by lead principals of the JOA, the island's governing body for sport, at its initial forum, themed "Investing Now: Safeguarding our Future". It was held at the Terra Nova Hotel in Kingston.

*The main presenters were Diahann Gordon Harrison, chairperson of the Office of Children's Advocate (OCA) and Mayberry Investment Limited's CEO, Gary Peart. "The Sports for Breakfast Forum is one of our stakeholder engagement initiatives whereby we invite member associations, members of the Government, members from corporate Jamaica to share in national and international topics that are relevant not only for sport, but topics that are relevant for the country on a whole," said Ryan Foster, the JOA's general secretary.*

*"It comes under the pathway for success for vision, where we broaden our stakeholder involvement and broaden how we improve discussions on what are current topics in the Jamaican landscape," he added.*

*Gordon-Harrison, the guest speaker, dealt with child abuse and advocacy, plus the importance of children in the growth of sport, but also in the protection of children as it relates to the future of sport in this country.*

In the cases when *child abuse* is mentioned while arguing an irrelevant point, it becomes gratuitous. One example of such use comes from an article by O'Brien Chang, a businessman writing for *Loop Jamaica*. In this example (51), child abuse is totally unrelated to the topic of the article where it is used in the same paragraph as a Road Traffic Act and placement of roadside stalls (in bold, emphasis authors'):

(51) *"We will never get to the stage as a country that is truly somewhere that people want to live if we don't address the situation of law and order. I refer specifically to things like the indiscipline on the roads in terms of how the taxi men drive. The fines in the new Road Traffic Act are a joke, they need to be more than that. We won't get to a really developed society if we don't address the situation of the rampant child abuse that takes place. We won't get to that point also if we don't enforce the situation relating to the zoning laws—people being able to put up stalls any and everywhere they feel".*

Thus, in some quotations generalised mention of 'child abuse' becomes a recurrent formulaic speech pattern, or an 'ideograph', a 'virtue word' (McGee 1980) which is defined as a word or phrase that is highly abstract and vague, but whose positive or negative value is stable and 'taken for granted'. They are rarely contested and 'appeal to the common sense of the audiences' (Condor et al. 2013, p. 12). Among ideographs scholars often mention 'community', 'freedom', 'fairness', 'rights', 'tyranny', and 'poverty'. Boyd, interestingly, also mentions the phrase 'no child left behind' (Boyd 2018, p. 145), adding that '[f]ew people are willing to enter public dialogue opposing such widely accepted values' (Boyd 2018, p. 143). The use of ideographs serves the rhetorical aim of enlisting alignment with the speaker (Boyd 2018; McGee 1980). Another important rhetorical function of ideograph is that 'while appearing decisive, they do not in fact commit the speaker to any particular course of action' (Condor et al. 2013, p. 21). It seems that, in many speeches by political figures, a reference to child abuse is fulfilling the rhetorical function of an ideograph, an abstract, incontestable notion everybody agrees with, the mere mention of which is the means for 'governments to be seen as doing something in relation to the problem of child abuse and neglect' (Saunders and Goddard 2002, p. 6).

This assessment of the rhetorical use of 'child abuse' is supported by the fact that it is used with over-generalised meaning in passages where child abuse is not a primary focus. 'Child abuse' appears in strings of similar ideographs in the political speeches reported in the newspapers (in bold, emphasis authors').


Repeated generic usage of such phrases is vacuous and devoid of real-life concrete meaning. In such cases the victims are mentioned for effect, whereas 'the real rather than rhetorical interests of victims need to be put at the forefront' (McAlinden 2014, p. 191).

As has been noted in previous studies, an episodic pattern seems to dominate the representation of child abuse in newspapers across several countries (Davies et al. 2015; Niner et al. 2013; Mejia et al. 2012; Weatherred 2015, 2017; Ho and Chan 2018; Cheit et al. 2010). The prevalence of a thematic pattern focusing on social issues related to child abuse was only reported by Hove et al. (2013), who studied the coverage of child abuse in US newspapers. In Jamaican newspapers, the generic treatment of the topic also prevails; it is two and a half times more frequent than episodic; however, the thematic analysis of the causes and social context of abuse is rare. The following paragraphs are positive examples of a thematic report covering solutions at the societal level.


#### 3.3.3. Members of the Public

The second most frequent category of people who speak of CSA on behalf of the victims is that of readers; the voice of the public can be heard from rubrics such as 'A view from the outside' (*Loop Jamaica*), 'Letters to the Editor' (*Jamaica Observer*), 'Letters' and 'Kelly's world' (*The Gleaner*), and 'Tell me, pastor' (*The Star*). They constitute over a quarter of all voices speaking about child abuse (26%, 49 occurrences, see Table 8). These contributions sometimes present unorthodox views on child abuse, which the newspapers validate by giving them a platform:

(56) *It has bothered me for some time that people treat children today like they are untouchable. So all when the pickney dem wrong, it's like they have immunity because 'dem a just children'.* . . . *I'm not saying children should be abused. Heaven*

*forbid. However, if people even look at children too sternly now, everybody a call Child Development Agency and Office of the Children's Registry. (*Letter in *The Gleaner)*

Even though the 'Letters' sections are open for the public to contribute, the newspapers do not seem to reach out to abuse victims/survivors to encourage them to speak out and use 'Letters' space to share their experiences; none of the papers have the voice of CSA victims in these regular opinion pieces.

## 3.3.4. Victims/Survivors

In contrast with the voice of representatives of different institutions, businesses, personalities, celebrities, and the general public, there are only nine cases (4.8%) in which abuse is talked about by victims/survivors themselves. In four cases, the speech is indirect, i.e., mediated by those who report it:

In indirect speech, the reporter is free to introduce information about the reported speech event from his point of view and on the basis of his knowledge about the world, as he does not purport to give the actual words that were uttered by the original speaker(s) or that his report is restricted to what was actually said. Indirect speech is the speech of the reporter, its pivot is in the speech situation of the report. (Coulmas 2011, p. 3)

Consider the example of victim's feelings communicated by the regular contributor to *The Gleaner*, Michael Abrahams who is described by the paper variously on different occasions—as 'a gynaecologist and obstetrician, comedian and poet' or as 'an obstetrician and gynaecologist, social commentator, and human-rights advocate':

(57) *The child was devastated. Not only was she violated, but was also blamed for the assault. She describes her subsequent years as being very turbulent, with much time spent displaying aggressive behaviour while trying to deal with her pain.*

The report of the emotions and the long-term effect the assault had on this person's wellbeing is presented through the eye of the writer of this piece.

In three other cases, the victims'/survivors' words are also mediated through the words of the officials or police:

(58) *female child reporting that a truck driver attempted to abduct and abuse her while she was commuting from school in St Catherine*

Apart from four cases of indirect speech accounts of abuse (2.1%), there are also five cases (2.7%) of direct speech by victims/survivors where they are given an opportunity to relate their experience in first-person reports.

(59) *"I reported the matter to the police but nothing was done about it. The night it happened, I called 119 and some police came and rescued me and hand me over to*

*CISOCA (Centre for the Investigation of Sexual Offences and Child Abuse)," he said.* . . . *While the case struggles to take shape, Doe said that the politician and comedian* [the perpetrators] *have been calling, trying to pay him to drop the matter. "Dem beg me because they say they have dem reputation," he said. "The police tell me I am lying so it come in like it will never go anywhere. Nothing nah come out of it, the case never mention or nothing; it just dead like that. " Doe said that he found out he was HIV-positive after the incident and has seriously considered suicide.*

There is also a first-hand account of the ordeal by the woman who was abducted, raped, and held captive for a month when she was 14 years of age. Speaking of the effect it had on her she told the newspaper:

(60) *Me hate man with every breath me take* . . . *. I fail at everything, and life is just really wicked to me* . . . *Me nuh trust nobody* . . . *. As it relates to man, me nuh waa inna no relationship with any*

In another rare case, a victim of abuse talks after many years of silence about her feelings to do with being raped by her grandfather. The report is a mixture of direct and indirect speech (61):

(61) *However, she said that a report was made after she found out that the man also raped two of her younger cousins. " When I found out about my younger cousins, I thought that it was because I was silent for so many years that this happened to them. So I started blaming myself for it, and that was what prompted me to make a report because I thought it was just enough," she said.*

There are only five instances in the corpus of victims using their own words to describe their experiences and feelings. These results agree with previous research into newspaper presentation of CSA which also noted that '[o]verwhelmingly, journalistic representations of child sexual abuse exclude the victims. Their voices had not been heard in the stories that we analysed' (Birnie and McKee 2009, p. 107). These findings are consistent with research of newspaper coverage in Australia (Lonne and Parton 2014; Lonne and Gillespie 2014), England and Wales (Davies et al. 2015) and Britain (Ndangam 2003) as illustrated by Kitzinger (2015). 'She [the abused child] is described as a "silent sufferer of victimization" but rarely allowed to speak about her own actions as opposed to the acts committed against her. Child survivors remain faceless and inaudible; their struggles to resist and endure abuse remain largely uncharted and unheard' (Kitzinger 2015, p. 149). The corpus results give a strong indication that the reporting style of Jamaican papers also denies survivors the possibility of constructing their own narrative; it is constructed for them. The scarcity of the voices of survivors is what Fairclough calls a 'significant absence' (Fairclough 1995, p. 106); it is indicative of the fact that the victims' discourse is constructed by someone else. Significantly, a guide for Caribbean practitioners on presenting child abuse in media emphasises the importance of taking children's

views into account (Association of Caribbean Mediaworkers 2016, p. 15) and giving them 'space to speak for themselves' (ibid., p. 27). A section of the guide dealing with children's right of expression specifically urges news outlets to find the way to 'assist children to express their opinions'. The present research shows that these points seem not to have been taken on board by news media. On the contrary, as has been shown above, children's voices very often go unheard, with many other agencies speaking for them. This corpus research has, thus, enabled us to note significant tendencies which otherwise might have gone unnoticed: 'A single text on its own is quite insignificant: the effects of media power are cumulative, working through the repetition of particular ways of handling causality and agency, particular ways of positioning the reader, and so forth' (Fairclough 1996, p. 54).

Another important aspect of reporting CSA is presented by the choice of word to refer to the abused person. There are two recognised ways of referring to the abused person—as a victim and as a survivor. To investigate which form was preferred by the newspapers, searches were conducted on the words *child/children, boy/s, girl/s, youth/s,* and *pickney/s* in association with two nouns—*survivor* and *victim*. *Survivor* is found seven times in the corpus, with only one instance related to CSA where child abuse was given as a generic example of a depression trigger in a survivor. The concordance search for instances of *child/children, boy/s, girl/s, youth/s,* and *pickney/s* collocating with the lemma *victim* showed 51 cases in which children were referred to as victims of violence and crime, including sexual abuse.

There is, of course, a reason for this: *victim* is a legal term used in Jamaican legislation, specifically in the Sexual Offences Act to define a person against whom 'the offence of grievous sexual assault' was committed (Sexual Offences Act 2009, p. 7). *Victim* is also an accepted term used in academic research on sexual crime. However, it has been noted by social scientists that, outside the legal and academic framework, the use of this term has another dimension. The term *survivor* has been used in the past several decades as an alternative to *victim* (van Dijk 2009, pp. 2–3; Leisenring 2006). Hunter (2010), researching narratives of women who suffered sexual abuse in childhood, recorded significant differences between the narratives of suffering and victimhood as opposed to narratives of transformation and 'survivorhood': it is survivors who 'told the *narrative of transformation'* (Hunter 2010, p. 183; see also De Sacco 2019). Similarly, according to the research of Newsom and Myers-Bowman (2017) evocatively entitled *"I Am Not a Victim. I Am a Survivor": Resilience as a Journey for Female Survivors of Child Sexual Abuse*, a survivor mentality is essential for overcoming traumatic experiences of CSA and developing coping mechanisms and resilience in order 'to capture those parts of women's [adult survivors of CSA] experiences that are attributable to proactive behavior, rather than solely to a passive response to trauma' (Hyman and Williams 2001, p. 216).

Existing research into this issue has noted that a survivor narrative does not diminish the suffering of the victim; '[s]urviving is the other side of being a victim'

(Barry 1979, p. 39). A survivor identity incorporates both women's victimization and agency (Profitt 1996, p. 29). The recurrent use of *victim* in the newspapers with reference to child abuse shows that the emphasis is placed on childhood victimisation rather than on the experiences of survival; children are constructed as sufferers, casualties of crime. A *survivor,* which can be understood as a more positive and empowering synonym to *victim,* is used only once in the corpus with a generic reference to a survivor of CSA. From the point of view of discourse analysis, a name-interpretation rule proposed by Reisigl and Wodak (2005, p. 76) states that when a person is named in a particular way, it implies that this person has particular set of qualities associated with this type of naming. 'The labeling expressions . . . shape . . . people's identities' (Mautner 2007, p. 53). The recurrent lexical choices in naming point to the way a particular entity is framed in a discourse. Naming is a conscious decision of the writer to choose a particular noun to designate the referent, a noun which, according to the writer, best reflects the nature of the referent. Naming practices can be indicative of a particular type of discourse. 'Different styles of *naming* are associated with different social values . . . this is a highly charged sociolinguistic indicator' (Fowler 1993, p. 99). The almost complete absence of the word *survivor* in relation to CSA in the Jamaican newspaper corpus indicates that the identities of those who have suffered abuse have been constructed by the news story writers as that of victims, not survivors. The absence of some representations from the text can be as telling as high frequencies and, as Fairclough (1995, p. 106) highlighted, it is important to be sensitive 'to absences from the text, to things which might have been 'there', but are not'; *survivor* is one of such meaningful absences.

#### *3.4. Other Aspects of Newspaper Discourses of CSA*

The next sections consider other notable features of reporting CSA that emerged from the study.

#### 3.4.1. Sensationalism

The previous sections showed the prevailing lack of empathetic treatment of victims of abuse in newspaper reporting; their voices are not being heard, and their side of the story remains largely untold. However, at the other end of the spectrum of newspaper coverage is sensationalist representation of child abuse (Lonne and Parton 2014) where news outlets report abhorrent stories of child abuse due to their potential to attract readership (Cheit et al. 2010; Hove et al. 2013; Popovi´c 2018). Research into newsworthiness, drawing on Galtung and Ruge's research, was further developed by Cheit (2003). Among the catalogue of newsworthiness criteria listed in Cheit (2003), three were found in the corpus: 'extra violence', 'bizarre facts', and the 'celebrity status' of either the abuser, abused, or both.

The corpus provides strong evidence of sensationalist representation of abuse, particularly connected with the reporting of rape. *Rape* occurs in association with

*child/children, girl/s, boy/s, youth/s,* and *pickney/s* 52 times. Investigation of the context shows that one fifth of these (10 occurrences) appear in a context that provides gratuitous graphic details of the abuse, some of them quite lengthy. Such coverage, far from giving victim a voice, is connected with gaining readers' attention by unnecessarily supplying disturbing details of the abuse. The actual victim's voice is heard in only one of these shocking reports, however the report itself raises questions about the wellbeing of the survivor who was asked by the reporter to relive her ordeal in a particularly graphic account.

Another side of sensationalism, also connected with newsworthiness, is the strong presence of elite—both as 'news actors' (victims and perpetrators alike) and as sources of opinion. As van Dijk noted: 'Elite sources are not only considered more newsworthy (as news actors) but also as more reliable as observers and opinion formulators' (van Dijk 1988, p. 87). Corpus analysis also shows a disproportionate number of reports of abuse featuring celebrities. The co-text and context surrounding the use of *child/children, girl/s, boy/s, youth/s,* and *pickney/s + abuse* show that, out of 34 sentences in which the emphasis was on the victim, about one third (11) referred to the cases where victims were celebrities; out of 59 instances of CSA focusing on perpetrator, one quarter (14) were mentioned with reference to allegations of crimes committed by celebrities. Both examples (62) and (63) show reports of denials by the celebrity or celebrity's family.


Newspapers also give disproportionate attention to bizarre facts connected with abuse.

(64) *A Tennessee woman accused of placing her 8-month-old baby into a freezer has been charged with child abuse.*

It is revealed only later in the article that the freezer was disused, and the child was unharmed. In another example of sensationalist reporting, the following piece of news was discussed on several occasions with various people—public figures, police, and readers—expressing their shock and disgust.

(65) *Children's Advocate blasts social media video. Children's Advocate, Diahann Gordon-Harrison, has expressed dismay at a video that is circulating on social media*

*with a female minor describing in excruciatingly graphic details, sexual acts she claimed to have seen being done by her parents.*

The search terms *child + abuse* appear five times in the context of this case. A later news item, however, revealed that:

(66) *It was also established that the person speaking about sex on the voice note was definitely not the child* . . .

This incident, which appeared not to have happened, received much wider and much more detailed coverage based on its distressing and bizarre nature ('unexpectedness' and 'unpredictability' factor of newsworthiness, as specified by Galtung and Ruge (1965, pp. 62–72; see also Wilczynski and Sinclair 1999; Davies et al. 2015). News media use shocking material to increase readership, and some 'child abuse stories have emerged as a sales tool' (Ndangam 2003, p. 98). Newsworthiness is, of course, co-constructed by the media outlet and readers' preferences (projected by the newspapers). News outlets make their assessment of how much prominence a particular event should be given based on their ideas of the newsworthiness of the event, thus making a 'transition from events to news' (Galtung and Ruge 1965, pp. 62–72). In contrast with sensational cases, cases of real CSA are often reduced to statistics. Another side of sensationalism is that stories focus on the disturbing details of abuse rather than on the causes of the abuse: 'the press tends to concentrate on specific criminal cases of a very severe nature without providing details of the social context within which abuse occurs' (Wilczynski and Sinclair 1999, p. 275).

## 3.4.2. Representation of Images of Child Sexual Abuse

The news stories around images of child sexual abuse were also specifically investigated. Even though child protection professionals have moved away from referring to children in terms of pornography, and are using such terms as 'indecent images of children' or 'images of child sexual abuse' (the latter being the preferred phrase), this terminology is not found in the papers. There are no instances of *indecent image* in the whole corpus. Preferred terms in the papers are *pornography* and *porn,* which can be connected with the fact that the term 'pornography' is still used in legal context in Jamaica, e.g., Child Pornography Prevention Act (CPPA). A search on the words *pornography* and *porn* was conducted because these words appear in Word Sketch in the top 10 strongest collocates of the lemma *child* (see Figures 1 and 2). *Pornography* collocates with *child* 12 times in the corpus, with a Log Dice score of 9.5, and *porn* 5 times (Log Dice 8.3).

To further investigate the context in which *pornography* and *porn* co-occur with *child/children, girl/s, boy/s, youth/s,* and *pickney/s*, a concordance search was conducted (15 words to the right and left of the search terms). The results showed 34 instances of such association. The phrase '*child pornography'* appears 14 times in the corpus, 8 of which in the legislative context of the Child Pornography Prevention Act (CPPA).

Six occurrences appear in the context of two specific cases related to celebrities: two well-known singers prosecuted for child pornography, one of whom is a US national. The phrase *child porn* occurs 10 times (328.76 per million), 5 in a general context and 5 in a specific context, out of which 3 occurrences relate to the case of a celebrity in the USA. An inspection of concordance lines showed that use of these collocations followed the general pattern outlined earlier: they are overwhelmingly generic, focusing on institutions or on the crime itself. Specific use is disproportionately skewed towards celebrity cases (67):

(67) *Prosecutors said the 26-year-old singer admitted to reaching out to 30 other girls online before his arrest on child pornography charges in 2017.*

This example shows that image-related CSA offences often include online grooming and the production of the images, achieved through online grooming and coercing young people to engage in online acts that can either be live-streamed or captured as images to be later distributed. The papers, it seems, do not place such cases in the focus of their attention. A search for *groom/grooming* showed that sexual grooming is reported twice in the corpus, and criminal grooming once. A search for *online* also revealed that papers do not pay significant attention to online child abuse—out of 93 occurrences of *online* in the corpus, the word was used twice to describe a case of online grooming and other two instances were used in the context of warning and recommendations for parents as to children's online use.

As with the presentation of rape, the context and the style of sentences containing *child pornography/porn* resembled court press releases:

(68) *The defense wanted the minimum available five years. Jones of Bloomingdale, Illinois pleaded guilty in February to child pornography.*

The victim of this form of CSA was mentioned only once in the whole corpus and only in a cursory way:

(69) *Authorities in Alabama have identified the suspect and the victim in the child porn video being shared around the world.*

This news item described a case when an indecent image of a child went viral on the internet and was discussed in several newspapers by reporters and public figures alike who expressed their revulsion at the video. However, as the next example demonstrates, the attention is still not on the harm done to the victim but on the effect of the abuse on the adults (70):

(70) *The disturbing image of a little girl, probably no more than five or six years old, being forced to perform oral sex on an adult male, was just too much to watch, and I was left with a feeling of complete horror and helplessness.*

(Betty Ann Blaine, child advocate and founder of Hear The Children's Cry, writing in *The Gleaner*).

#### 3.4.3. Reframing Abuse: Normalising the Crime and Shifting Responsibility

The heightened attention to the case mentioned above and the way it was reported highlight important issues in newspaper coverage of images of child sexual abuse and of child abuse in general. First, it demonstrates a sensationalist approach to reporting on child abuse. The most extreme and shocking cases generate disproportionately extensive coverage: out of 24 occurrences of *child* in association with *pornography/porn*, 14 are connected with one disturbing case. Another concern highlighted by this case is the framing of the abuse. While the tone of the coverage in several papers was generally of condemnation, one paper, *The Star*, published an article under the headline *Child gives man blow job*. This caused understandable outrage among readers, because the use of the verb *to give* presupposes agency and initiative on the part of the abused child and places responsibility for the abuse not on the sex offender, but on the victim of the abuse, thus normalising the abuse.

Under the pressure of public indignation and complaints, *The Star* had to publish a public apology for the headline. The Press Association of Jamaica (PAJ), however, refused to interfere or condemn the publication on the grounds that passing judgement on publications is not part of the role of PAJ. The President of the PAJ, Dionne Jackson Miller, in her reply published by several papers (e.g., Loop Jamaica http://www.loopjamaica.com/content/not-our-role-paj-wont-judgestar-headline, accessed on 20 July 2018), stated: 'On occasions such as these, there are calls for the Press Association of Jamaica to condemn the offending publication. While I understand the public's desire to have an industry-specific organisation speak out in such cases, this is not, and should not be, the role of the Press Association of Jamaica'. To date, there is no Jamaican body that could adjudicate on similar cases. However, the guide to ethical reporting 'Our Children, Our Media: A Guide for Caribbean Practitioners' (Association of Caribbean Mediaworkers 2016, pp. 40, 67) clearly sets the parameters for reporting on sexual abuse in general and pornography in particular, stating that the emphasis should be on prevention of the abuse and protection of the child. This case highlights two problems with the reporting of child abuse: there are occasions when the papers choose not to abide by the guide, and the PAJ does not regulate cases of unethical representation.

Further examples of normalising abuse in newspaper coverage can be seen in the following collocations appearing in the corpus: *have sex/sexual intercourse with a minor/an underage girl/ a 12-year-old girl'* (see concordance in Figure 4).


such cases, this is not, and should not be, the role of the Press Association of Jamaica'. To date, there is no Jamaican body that could adjudicate on similar cases. However, the guide to ethical reporting 'Our Children, Our Media: A Guide for Caribbean Practitioners' (Association of Caribbean Mediaworkers 2016, pp. 40, 67) clearly sets the parameters for reporting on sexual abuse in general and pornography in particular, stating that the emphasis should be on prevention of the abuse and protection of the child. This case highlights two problems with the reporting of child abuse: there are occasions when the papers choose not to abide by the guide, and the

Further examples of normalising abuse in newspaper coverage

can be seen in the following collocations appearing in the corpus: *have sex/sexual intercourse with a minor/an underage girl/ a 12-year-old girl'* (see

PAJ does not regulate cases of unethical representation.

concordance in Figure 4).

**Figure 4.** Concordance lines *have sex with* and *sexual intercourse with.* Source: Figure created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

**75**  There are 28 cases of such collocations in which the euphemistic and misleading phrases *have sex* (20 occurrences, 32.3 per million) and *sexual intercourse* (8 occurrences, 12.92 per million) are used. This use of neutral expressions misrepresents the fact that sex with an underage or minor person is illegal and abusive because, by definition, children under the age of consent (16 years in Jamaica) cannot consent to have sex, and any case of sexual activity with a minor constitutes rape. 'Because a legal age of majority is required for consent, all sexual acts between an adult and underage child (even with child assent) are, by definition, CSA' (Murray et al. 2014, p. 323). The World Health Organisation's (WHO 2017, p. vii) definition of CSA draws attention to the fact that a minor involved in sexual activity 'is unable to give informed consent' to it.

Newspapers, however, continue to use this expression on a regular basis, conflating rape with having sex.


(73) *Grief for girls ! - More than 32% who had sex before 15 years old were forced. Almost one-third of Jamaica 's girls who had sex before the age of 15 were forced into the act.*

When speaking of similar discoursal strategy of normalisation in relation to newspaper coverage of racism, van Dijk argued that use of euphemisms is one of the ways 'to mitigate the seriousness, extent or consequences of one's negative actions' (van Dijk 1992b, p. 180). Examples (71) and (72) illustrate such strategy, whereas example (73) demonstrates even more inappropriate word use: although the newspaper reports that the underaged girls '*were forced*', it does not use the word '*rape'*, but still resorts to an innocuous expression '*have sex'*. The use of the phrase *against her will* in (74) is also inappropriate for the same reason: a child cannot give consent to sexual activity, and, therefore, any such activity is against their will and should be called rape.

(74) *Reports from the police stated that on July 31, the child was on her way home when she was attacked by the farmer, who had sexual intercourse with her against her will.*

Even more misdirected is a phrase *to rape her against her will* used in one of the reports (75):

(75) *Gurdon allegedly sexually assaulted and raped her against her will*

Among other euphemistic phrases normalizing CSA used by the newspapers are *to be sexually active, sexual exploration,* and *sexual initiation*:


Normalizing discourse strategies shift the blame away from the perpetrator to the abused child, as can be seen both in examples (77) and (78):

(78) . . . *he was drugged by men, five of whom had sex with him. According to Robert, his father found out about the incident, blamed him for it, and threw him out on the street.*

The wording of the report (78) again uses euphemisms to normalise the gang rape by referring to it as '*having sex'* and calling the crime '*an incident*'; it also presents without comment, as a fait accompli, the fact that the father blamed the youth for the abuse.

Shifting the blame from the perpetrator to the victim and claiming consent as a means of denying responsibility is a common strategy employed by abusers (Smart

2000; Lonsway and Fitzgerald 1994). In the media, the blame is also often assigned to a victim (Wilczynski and Sinclair 1999; Saunders and Goddard 2002; Ndangam 2003), to the family, particularly mothers (Niner et al. 2014; Ndangam 2003; Saunders and Goddard 2002), or child protection services (Lonne and Parton 2014). Such representations not only ignore the perpetrator, but also obscure the social factors of the abuse (Saint-Jacques et al. 2012; Weatherred 2015). Blame for child sexual abuse in several news reports in the corpus was placed on adolescent mothers (79), mothers of teenage boys who abuse younger girls (80, 83), immature (81) and abusive parents (82), as well as culture (83) and technology (84):


Superintendent of Police Charmaine Shand, who heads the Centre for Investigation of Sexual Offences and Child Abuse (CISOCA)


In the case of the viral video with an indecent image of a child, some commentators blamed digital technology and the internet, yet again concealing the deliberate acts of perpetrators who commit child abuse crimes:

(84) *The video had already gone viral, and the nagging questions in my mind were, "How do we stop this? How do we protect children from the dangers of digital technology? Computers and the Internet have made the predator's job easier. Historically, child predators found their victims in public places where children tend to gather—schoolyards, playgrounds, shopping malls. Today, with so many children online, the Internet is the predator's dream come true. The Internet provides a source for repeated long-term victimisation that can last for years* . . .

The blame here is transferred from the perpetrator onto the internet, an attitude that has previously been noted in the UK papers (Ndangam 2003, p. 179). The language used in this passage is also worth noting; instead of referring to child sexual exploitation as *crime* or *abuse,* the euphemisms *predator's job* and *predator's dream* are used. Portraying crime in neutral terms normalises it and desensitises society to inappropriate behaviour.

Furthermore, shifting blame from the perpetrators to victims and their mothers, or to the technology they use, is likely to distract from the broader causes of abuse on social, political, and institutional levels.

## 3.4.4. Newspaper Presentation of Abuse by Family Members

The Report of the Consultation on Child Abuse Prevention produced by the World Health Organisation established as early as 1999 that the perpetrator can be known to the child and can be a caregiver, or biologically related to the child (WHO 1999, p. 13).

A recent qualitative study into child sexual abuse in Jamaica highlighted the fact that CSA is commonly 'perpetrated by individuals known to the victim such as family members or family friends' (Powell Booth et al. 2021, p.10) and a survey conducted among 7182 Jamaican children between 9 and 17 years of age (2018–2019) showed that 15% of CSA (5% contact and 10% non-contact) was perpetrated by adults living in the victim's home.

To see whether and how papers report on cases of abuse by family members, a concordance search was conducted of the words *sexual abuse, sexual assault, incest, rape,* and *molest* associated with nouns of kinship: *family, father, mother, uncle, cousin, relative,* and *step.* To have a broad focus on the context, a wider span of -15 to +15 was applied. There were 86 instances of co-occurrence, 42 of which were relevant to CSA in a family. Investigation of the concordance lines in this category showed that two thirds of instances (28, or 66.7%) were related to specific cases of abuse, as in the excerpt (85):

(85) *Kashina McLean was sexually abused as a child for eight straight years by family members who should have been her protectors.*

Closer examination of the co-text revealed patterns of use similar to the ones discussed earlier. Out of 28 collocations, about one third (10) were cases connected with celebrities: seven instances of co-occurrence referred to the alleged abuse of Whitney Houston and three of Dalton Harris. Four other co-occurrences were related to the case of a victim/survivor being abused by her step-grandfather, an extreme and shocking case which met the criteria of newsworthiness and sensationalism and thus received more substantial coverage. Fourteen instances reported abuse in general terms; six of them are related to the issue of abortion in cases of incest. On

the positive note, several pieces outlined the problem of raising general awareness about child abuse within a family, for example:

(86) "*Since the start of the year, we have seen that most of our crimes are committed by persons who are known to the victims. So we now no longer tell persons (just) to be careful of strangers. So what do we tell them? These are our fathers, our stepfathers, our stepmothers, our brothers, our uncles, our cousins, who are abusing our children.* (Senior Superintendent Charmaine Shand, head of CISOCA, speaking at a meeting of the Rotary Club of Kingston)

In general, this category follows the trends in reporting child abuse which emerged from previous concordance searches. However, there is one feature specific to reporting cases of child abuse in the family: there is a strong emphasis on women as offenders and accomplices in abuse; one third, or 18 out of 59 cases, relate to women. For example, three instances refer to the case of a woman jailed for failing to report her daughter's abuse. Another illustration of the emphasis on women offenders appears in the CISOCA message (example (86) discussed above), which, while giving a general warning about child abuse in the family, provides three examples of offences committed by female family members (87), (88):


Newspapers are also giving a platform in their 'Letters' sections to unsupported claims that women are not prosecuted for sexual offenses (89):

(89) *Those of us in the courts are aware of cases in which the younger boy is preyed on by an older female, yet he is taken to court for rape or carnal abuse and no action is taken against the girl.*

In the cases of celebrity abuse, female offenders are also singled out (90):

(90) *Houston's long-time assistant Mary Jones tells filmmaker Kevin Macdonald that the late singer told her that Warwick, the sister of Dionne Warwick, molested her at a young age. Houston's brother Gary Garland-Houston also says he was molested between the ages of seven and nine by a female family member, and says his sister was abused too.*

Example (91) is the excerpt of a news item about a mother prosecuted for not reporting CSA of her underaged daughter. In the paragraph, there is only one mention of the abusive stepbrother: the central focus in this piece is the mother's

inaction in the face of the violence and abuse. This excerpt starts and finishes with a description of the mother's misconduct, whereas the actual perpetrator has less coverage, and his offence is described by using, once again, the deceptive and misleading euphemistic expression '*had sexual intercourse'*.

(91) *Mother charged for not reporting sexual abuse of 11-y-o daughter. A Manchester mother who allegedly failed to report the repeated sexual abuse of her 11-year-old daughter has been charged by the police. The woman, who now faces a fine of up to \$500,000 and prison time of up to six months, was charged for the offence of failing to report a child in need for care and protection following investigations by the Mandeville police. Reports are that the reported abuser, who is said to be the complainant's stepbrother, had sexual intercourse with the child on several occasions, along with several other sexual activities. The complainant is said to have reported the abuse to her mother, who allegedly failed to make a report to the police. The incident has led to the head of the Centre for the Investigation of Sexual Offences (CISOCA) to appeal for persons who may have information on the matter to come forward. A date for the mother's appearance in court is being finalised by the police*

Even in cases where women were neither offenders nor accomplices, they were sometimes constructed as reprehensible associates (92):

(92) . . . *she was asked not to speak about being raped because some persons thought that she would bring shame on her grandmother* . . .

Azzopardi et al., looking at the context of professional response to abuse, noted the pervasive nature of mother blaming, which reflects, according to the authors, 'sociocultural ideologies' (Azzopardi et al. 2018, p. 270). One such ideology responsible for the skewed representation of domestic violence and abuse in newspapers may be grounded in protecting patriarchal power structures.

To ascertain whether reports on child abuse by family members contained voices of victims/survivors, the larger context of collocations of the words *sexual abuse, sexual assault, incest, rape,* and *molest* associated with nouns of kinship *(family, father, mother, uncle, cousin, relative,* and *step)* was examined. The investigated context was as large as possible because reports of abuse can be as long as several paragraphs, or even up to the whole article, with a mix of direct and indirect speech. The results showed that six victims' accounts of abuse were presented using direct speech and first-person narration (10 co-occurrences, 23.8% of the general number of instances of use), which is much higher than other concordance searches described in Sections 3.2 and 3.3 (*child/children, girl/s, boy/s, youth/s, pickney/s + abuse, sex, sexual, rape*) where the voices of the victims/survivors were barely heard. The reason might lie in the newsworthiness of such reports because accounts of abuse within a family are particularly disturbing. Consider, for example, one attention-grabbing headline: *'I was raped by my grandfather*' which capitalizes on the newsworthiness of the case.

### 3.4.5. Disclosing and Reporting Abuse

In 2004, mandatory reporting came into effect in Jamaica as a result of the Child Care and Protection Act, requiring practitioners who suspect child abuse or neglect to make a report. The study by Jamaican academics (Powell Booth et al. 2021) specifically highlighted the importance of children reporting abuse and disclosing abuse to adults. To explore the newspapers' position on disclosing abuse, a concordance search was conducted on the key words *child abuse* in conjunction with verbs of speaking *tell, talk, disclose, discuss,* and *report* within 15 tokens to the left and right of the keyword to allow the maximum available context. Twenty-four instances of collocations were found in the corpus, *tell + child abuse* was not used in the corpus. *Tell + abuse* is used in the sense of reporting once in the context of theatrical performance. *Talk* appeared three times (0.21 pm), all with reference to adults representing different organisations talking about abuse in general terms:

(93) *The talk on children and how we look at the whole problem of sexual abuse, child abuse as it relates to sports and coaching is something we've been talking about*

Twenty-one co-occurrences of *child abuse* + *report* were found in the corpus (see concordance in Figure 5), nine (0.96 pm) of which related to reporting cases of child abuse (the other 12 referred to media and police reports about adult abuse.) related to reporting cases of child abuse (the other 12 referred to media and police reports about adult abuse.)


**Figure 5.** *Child abuse* + *report.* Source: Concordance created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004). **Figure 5.** *Child abuse* + *report.* Source: Concordance created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

Out of these nine cases, the issue of reporting child abuse was treated generically in seven, all of which were used in constructions that do not show who was actually reporting: the infinitive (*more*  Out of these nine cases, the issue of reporting child abuse was treated generically in seven, all of which were used in constructions that do not show who was actually reporting: the infinitive (*more persons* . . . *to report*) or passive voice (*were reported, reports were made,* or *is receiving reports*).


70

*empower them to speak up about abuses or attempted abuses.*  Sonita Morin Abrahams, executive director of RISE Life

Management Services

*44,782 reports of child abuse were made to the Office of the Children 's Registry, of which more than 380 were for sexual abuse. This national programme was therefore developed to educate children on what is appropriate and accepted behaviour* versus *those that are not, to provide treatment for at-risk boys and girls who have suffered abuse, and to* 

Montego Bay businesswoman, Kareena Mahbubani

**84** 

was reported by a girl who was abused. However, this news article only described the case and did not emphasise the importance of

Only in one instance in the whole corpus is it clear that the abuse

*more than 380 were for sexual abuse. This national programme was therefore developed to educate children on what is appropriate and accepted behaviour versus those that are not, to provide treatment for at-risk boys and girls who have suffered abuse, and to empower them to speak up about abuses or attempted abuses.*

Sonita Morin Abrahams, executive director of RISE Life Management Services

Only in one instance in the whole corpus is it clear that the abuse was reported by a girl who was abused. However, this news article only described the case and did not emphasise the importance of reporting abuse, nor did it give any information about reporting. Only one newspaper article contained the so-called 'mobilising information' allowing victims to access assistance in cases of abuse. The importance of mobilising information in cases of gender violence was highlighted in a number of studies (see, for example, Nicodemus 2004; Seely and Riffe 2021). Newspaper articles containing contact information, the telephone numbers and addresses of relevant services and useful websites help people to seek assistance and report abuse. According to Seely and Riffe (2021), 'news media have the opportunity—even a responsibility—to help people take action'. To establish whether the Jamaican newspapers in the corpus provide mobilizing information for abuse victims, the search was conducted on lemmas *call, phone,* and *dial* collocating with *abuse*; only two instances of mobilising information following a news item on child abuse were found in the corpus.


The corpus was also searched for instances of *Crime Stop,* a partnership between the community, the police, and the media designed to encourage the public to give information about crime. *Crime Stop* was mentioned in the corpus 11 times, but only once was the occurrence of *Crime Stop* accompanied by mobilising information. Even a piece reporting on the success story of a girl who was rescued from an abusive father after calling Crime Stop did not contain the telephone number to call. Insufficient information to help the public report child abuse is another significant absence from the corpus. A specific, useful action that can be taken by newspapers is to 'focus on informing people what they can do to help solve the problem and how they can link up with existing organizations and programs that promote child welfare' (Hove et al. 2013, p. 107). By contrast, many news items end with the telephone number for accessing news on mobile phones.

#### 3.4.6. Counselling for Children Who Suffered Abuse To consider further the question of assistance and support

3.4.6. Counselling for Children Who Suffered Abuse

mobile phones.

To consider further the question of assistance and support available for abused children, the word *counselling* was searched in the corpus (see Figure 6). available for abused children, the word *counselling* was searched in the corpus (see Figure 6).

specific, useful action that can be taken by newspapers is to 'focus on informing people what they can do to help solve the problem and how they can link up with existing organizations and programs that promote child welfare' (Hove et al. 2013, p. 107). By contrast, many news items end with the telephone number for accessing news on


**Figure 6.** *child/ren, boy/s, girl/s, youth/s, pickney/s, victim/s + counselling.* Source: Concordance created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004). **Figure 6.** *child/ren, boy/s, girl/s, youth/s, pickney/s, victim/s + counselling.* Source: Concordance created by authors using the corpus tool Sketch Engine (Kilgarriff et al. 2004).

The word *counselling* co-occurs with *child/ren, boy/s, girl/s, youth/s, pickney/s,* and *victim/s* 17 times (15R-15L), out of which 14 refer to children and 3 to families of missing children. The concordance lines and the surrounding co-text were examined manually to understand the context in which the word was used. The references to child The word *counselling* co-occurs with *child/ren, boy/s, girl/s, youth/s, pickney/s,* and *victim/s* 17 times (15R-15L), out of which 14 refer to children and 3 to families of missing children. The concordance lines and the surrounding co-text were examined manually to understand the context in which the word was used. The references to child counselling are generic in 43% of cases (6) appearing in the speeches of politicians, representatives of the government, and police.

counselling are generic in 43% of cases (6) appearing in the speeches of politicians, representatives of the government, and police. (98) *we need to create a safe, supportive school environment, where the*  (98) *we need to create a safe, supportive school environment, where the schools will assess children's behavioural patterns, provide caring programmes and provide psychological and counselling services to children*

*schools will assess children's behavioural patterns, provide caring*  Counselling for children who have been abused was mentioned once in general terms.

**86**  (99) *Grieving or abused school children, many without lunch money, need along with lunch the therapeutic counselling of such workers* . . .

Of 17 specific instances of collocations of *child/ren, boy/s, girl/s, youth/s, pickney/s,* and *victim/s* with *counselling*, 57% involved counselling as a result of school fires (4 occurrences), violence (2), and instances of killings on school premises (2). There is one case that mentions future plans for a facility for the victims of domestic abuse where counselling will be offered, and only one case that calls for support for a victim/survivor of CSA.

(100) *The news of a nine-year-old girl being raped while in state care, allegedly by another ward, is tragic and disturbing. We expect not only a full investigation by the authorities to determine how this tragedy occurred, but also the provision of counselling and care for the victim, and appropriate corrective measures.*

The writer of this piece which appeared in *The Gleaner* June 5, 2018, Philippa Davies, is an attorney-at-law and co-founder of Marriage Matters Jamaica. Two sentences (100) are the only mention of child abuse and victim counselling in this article, which also uses the generic inclusive *we expect* with no clear reference, all of which points to 'ideographic' use of this issue.

Newspapers seem to be missing the opportunity to make a difference by highlighting the necessity of support for abused children; just as with mobilising information, the articles mentioning counselling do not contain information about the accessing existing support for children.

## **Chapter 4. Discussion and Conclusions**

### *4.1. The Role of the News Media in the Social Construction of Childhood and Child Abuse*

According to Kitzinger (2015, p. 161), "[d]ebates about the sexual abuse of children are deeply embedded in discourses about childhood." The news media, and in this instance newspapers, play a significant role in social construction of child abuse as a major problem in society. Media constructions of childhood and of child abuse help to shape society's attitudes towards raising children and understanding what constitutes normal or unacceptable behaviours. This study used a combination of discourse analysis and corpus methods to explore the language used to frame and represent children and child abuse in the Jamaican news media. Given the prevalence of CSA and the impact of violence against children in Jamaica, there is a need for examination of the representation of children and their experience of violence in the news media, which constitute the main public access to reports of abuse of children. On the one hand, media reports that fail to accurately conceptualise and represent children who have experienced abuse run the risk of promoting discourse that further marginalises children, inhibits disclosure of abuse, contributes to stereotyping and victim-blaming, and undermines perpetrator and observer accountability. On the other hand, media can play a positive role in changing perceptions of children, heightening awareness around child abuse and its consequences, and changing public attitudes about abuse. Findings from qualitative research conducted by Powell Powell Booth et al. (2021) suggest there is a need for change in cultural norms and practices that normalise abuse in Jamaican society. Consistent with the Bronfenbrenner's (1989) ecological systems theory, the media reflects a sub-system of actors crucial to achieving the kind of social transformation envisaged by Powell Booth and her colleagues. However, this requires the use of their platforms to facilitate different discourses around issues related to child abuse which centre the child, locate accountability where it needs to be and catalyse social action.

#### 4.1.1. Generalisation and Collectivisation

The results from the analysis of the language and discourse represented in the Jamaican news media are clear. Contrary to the guidelines outlined in the 'Our Children, Our Media: A Guide for Caribbean Practitioners' published in 2016 (Association of Caribbean Mediaworkers 2016), the Jamaican newspapers have failed to humanise children in their reporting of abuse and violence. Instead, representations of children are generic, their experiences often relegated to statistical summaries which are denuded of any personal features. While it is necessary to report factual information, there is a balance that is needed in the framing of the discourse around children as survivors of abuse in the news media such that there is sensitivity in reporting the individual experience of the survivor. Experiences of child

abuse must be conceptualised and represented in the media in a way that encourages a public climate of empathy, that increases the propensity to disclose abuse, and promotes the pursuit of justice for survivors. Our findings showed, however, that representations of child sexual abuse in Jamaican newspapers either take a vague approach, which diminishes the voice of the child and minimizes the effects of harm, or are sensationalized for maximum newsworthiness, which also renders children invisible. Abuse was reflected broadly as a societal problem with little in-depth analysis of the factors that enable perpetration of abuse and the repercussions for the children. Statistics were often used in the reports of abuse as a generalised and collectivistic way of framing the incidents of abuse within society. In doing this, the intimate stories of child abuse are sanitised. When incidents of abuse are reduced to enumerative prevalence alone, this can have the effect of desensitising the public and normalising abuse. Reports from which the victim's voice and perspective are missing privilege other voices (van Leeuwen 2008), for example those of institutions, politicians, and public figures who often rely on ideographs and platitudes that offer neither direct support to children nor solutions to the harms they have experienced.

### 4.1.2. Children as Voiceless '*Victims*'

The results show that the media has the propensity to present a one-sided, stereotypical view of children who have been victimized as vulnerable, powerless, fragile, and voiceless, which may distort the public's perceptions of children and promote misconceptions of those who do not conform to these representations (Kitzinger 2015). The recurrent use of *victim* in the newspapers contributes to this construction as it emphasizes victimhood rather than survival or resilience; children are constructed as passive, wounded, and lacking agentic capacity. It is important to acknowledge that within many legal systems, including in Jamaica, the term 'victim' carries a legal definition which denotes someone who has been subjected to a crime. In contrast, 'survivor' most commonly refers to someone who has either gone through or is in the process of going through the recovery phase after the abuse has occurred and signifies empowerment for the persons who have been abused. Despite this, the term *survivor* appeared only once in the corpus and was made as a generic reference to child sexual abuse. The discursive representation of CSA is complex; children *are* disempowered when they are abused; they have neither power nor responsibility in preventing abuse, and in order for them to have justice they must first be recognized as victims of crime. Nevertheless, in omitting their voices and perspectives, understandings about the myriad factors that contribute to child victimization and the myriad ways in which children attempt to contest and resist abusive behaviours are lost. It is not always inappropriate for others to speak on behalf of children, but it is always appropriate for children to speak on behalf of themselves; one is not synonymous with the other.

Another problem with the representation of children in newspapers is limited acknowledgement of the effects of abuse on them, their families, and the wider society. Kitzinger and Skidmore (1995), in what is considered their seminal work on media coverage of CSA, reported that, at the time, 71% of newspaper and 83% of television coverage was case-based or episodic, with very little attention to the possible causes and prevention of CSA. In the current study, there was a notable lack of reporting the implications to society from a psychological, sociological, or an economic perspective. A recent study of adult survivors of child sexual abuse has shown that Jamaican children who are abused experience many consequences and carry a heavy burden that continues into adulthood (Powell Booth et al. 2021). Therefore, more accurate representations of their abuse, for example, by giving a platform to survivors, give authenticity to their experiences and would help children understand what has happened to them. A broader, recent application of this was the UNICEF initiative facilitated by the Jamaican Government in November 2019 that saw children address parliament in a special session on violence against children as part of activities to commemorate the 30th anniversary of the UN Convention on the Rights of the Child (UNCRC). The media should also take their lead from this initiative in giving a voice to survivors.

### 4.1.3. Who Is Represented?

Generally, the perspective of the survivor was least represented in the news media, which is consistent with previous research (Birnie and McKee 2009). Instead, the voice of the journalist was predominantly represented, with the second most prevalent representations made by institutions, celebrities, and members of the public, thereby backgrounding the survivors. These representatives are seen as confident and knowledgeable, with professional attachments that give their voice authority in commenting on matters relating to child abuse. This may strengthen the perception of the public that these professional perspectives are valid, which may inadvertently diminish the voice and perspective of the survivors concerning their experiences and what decisions or interventions may be in their best interests. In addition, reliance and reporting on the comments of these individuals may also skew the discourse of child abuse towards individual or institutional agendas. This style of reporting by the Jamaican newspapers gives limited access to the narrative of the survivor; instead, the discourse surrounding their experiences is constructed for them.

In a few cases, the survivor received indirect representation in which the experience is related from the writer's perspective, or a direct representation of the victim's voice is briefly given. This type of distancing and generalization in the way children were portrayed impersonalized them and reduced their narratives to blurbs, broad statements, and statistics. From the reader's perspective, there was no emotional connection to the child or the child's experience. The limited inclusion of

children's voices in the matters that affect them is perhaps a cultural issue in Jamaica. While the UNCRC (Article 12) of which Jamaica is a signatory recognises the right of the child to be heard in matters concerning them and their well-being, Jamaica still has a culture of *'children should be seen and not heard.'* Representations of children in the Jamaican media also seem to support this viewpoint. Indeed, the Guide for Caribbean media practitioners underscores the need for media to give children the opportunity to speak for themselves and to tell their own stories (Association of Caribbean Mediaworkers 2016, p. 27).

Another important finding in the corpus is that the emphasis on the specific offenders was twice as frequent as on the victims. This was particularly true for offenders who were viewed as 'newsworthy', including persons in positions of authority or those given celebrity status. Public figures and celebrities feed into the sensationalism of a story to attract an audience. However, how newspapers choose to direct readers to certain aspects of the news or individuals also frames the story for the public. That is, who and what is portrayed and whose story is emphasised have implications for public perceptions, judgments, and attribution of responsibility (Kim et al. 2002).

#### 4.1.4. Normalising Crime and Sensational Reporting

While the use of terms which normalizes sexual abuse in print media is not regulated by a public body or authority, recommendations on ethical reporting are provided in the Guide for Caribbean media practitioners (Association of Caribbean Mediaworkers 2016). The use of neutralizing terms such as "*have sex/sexual intercourse with a minor"* were found in the corpus. Research studies which seek to explore constructions of abuse sometimes use neutral terms to extract more accurate information. For example, a study of perceptions, attitudes, and opinions on CSA in six Caribbean countries (Jones and Jemmott 2009) reported the value of taking a neutral approach in order to fully uncover the extent of CSA. They explained, for example, that framed as CSA, most people would agree that this was not acceptable, no matter what the circumstances, whereas, when framed as 'sexual activity with a child', respondents were more likely to be open about their own abusive behaviours. Such exceptions apart, the use of euphemisms undermines the illegality and seriousness of sexual abuse and shifts responsibility and blame from the perpetrator to the victim (van Dijk 1993, p. 180). Reframing abuse normalises the crime and makes it difficult for the public to empathise with survivors; it also creates confusion about what constitutes abuse and further prevents reporting of abuse.

As evidenced by the findings of this study, sensationalism is an issue of concern which also limits the voice of the victim. Based on the reports of rape in Jamaican newspapers, the motivation behind some of the published reports is more intended towards capturing the attention of the reader/audience as opposed to giving the victim a voice. This is evidenced by the disproportionate attention given to cases

involving celebrities or noteworthy people in society and the disproportionately extensive coverage of the most extreme and salacious cases.

## *4.2. Conclusions and Recommendations*

This study investigated the media portrayal of children and sexual violence against them within the popular Jamaican press. The first study of its kind in the Caribbean, its findings have implications for policy and ethical guidelines that govern media practices. It is instructive at this point to relate some of the study findings to the qualitative research carried out by the Ni3 team in Jamaica, which examined the experiences of CSA from the perspectives of adult survivors and perpetrators and which, at the point of writing, is the most up-to-date research on the topic. This research by Powell Powell Booth et al. (2021) has been synthesised into a Policy Brief; this highlights some key issues for the media to take into account:


*and agency. There is also a role for the media in uncovering institutional failings in the criminal justice systems and in the responses of professionals to CSA.*

5 Boys who experience CSA face additional hurdles in disclosure and reporting, including values and beliefs which promote sexual prowess as signifiers of masculinity and the fear of negative reactions to disclosure. *The media should take a progressive stance towards enabling boys to report CSA and to catalyse debate among the public which promotes healthy and equitable gender relations and disrupts the myths which sustain and normalise abuse.*

(Policy points extracted from the None in Three Jamaica Policy Brief derived from qualitative research: http://www.noneinthree.org/wp-content/uploads/2021/ 09/Jamaica-Policy-Brief-Print-View.pdf, accessed on 11 May 2022)

Consideration should be given to infusing specific training on these issues in reporting on children and child maltreatment at the different levels of media/journalist training programmes. Continuous training through workshops, seminars, and professional groups for journalist and media practitioners may also help to reinforce the need for caution and sensitivity in media reports regarding children. Whilst there are guidelines, for example, through the Guide for Caribbean media practitioners, the findings of this study show deviations from the recommended practices. It is also instructive from the findings of this study that the newspapers need to find a delicate balance between giving children the opportunity to represent themselves and tell their own stories as a part of the public discourse, while ensuring that they are protected and not exploited through sensationalism and lack of adherence to ethical practices. While it is important to report on statistics to highlight the challenge of abuse among Jamaican children, further journalistic applications can be made to also spotlight the child as the main individual of concern. Given the prevalence of CSA and violence against children in general in Jamaica, it is important that all agencies work together to eliminate the negative norms and perceptions of children, and to cultivate a climate in society where abuse is not tolerated.

#### 4.2.1. Strengths and Weaknesses

This corpus assisted method and discourse analysis into the representations of child sexual abuse in Jamaican newspapers is the first of its kind and a noteworthy strength of this paper. Though significant research has been conducted using the corpus-linguistic approach, which is frequently cited as the fastest growing method of contemporary language research, this is a novel methodology within the Jamaican, and indeed Caribbean, context. Additionally, the current study provided an in-depth surgical analysis of child representation within the Jamaican context using the social constructivism theory to argue that Jamaica's media representation of children as vulnerable, passive, and universal serves only to harm survivors of abuse, rather than focus on their humanitarian cause. The findings of the research centres attention

on discussions that acknowledge the power of the media, and the need to hold them more accountable as agents that shape the perception of children and their realities within the Jamaican landscape.

Though the research is novel and has its methodological and theoretical strengths, the period of analysis was limited to three years, 2018 to 2020. While the corpus within this period was sizable and provided a wealth of information for analysis, it only represented more recent content and reporting by the newspapers. Future research may need to consider covering larger periods of time to understand changes in the media landscape over time with regard to their representation of children and reporting child sexual abuse. It may also be instructive that future research be focused on an interventionist approach to examine how policy and practical changes may influence changes in reporting on children and their experiences.

**Author Contributions:** This work was undertaken by Dr. Tatyana Karpenko-Seccombe, Linguistics Expert at the None in Three Centre for the Global Prevention of Gender-based Violence, based at the University of Huddersfield, UK (www.noneinthree.org). The study was conducted in collaboration with members of the Jamaica Ni3 project: Dr. Kenisha Nelson, Dr. Christine Fray, Roxanne Harvey, Dr. Karyl Powell-Booth and Directors of the None in Three Centre, Professor Adele Jones and Professor Nadia Wager. Technical assistance in preparing corpora was provided by Xiaomin Sheng. The monograph is based on original research.

**Funding:** Funded by the UK Research and Innovation Global Challenges Research Fund (GCRF), the Ni3 project in Jamaica was implemented through a partnership with the University of Technology (UTech), Jamaica under the leadership of Dr. Christine Fray, Associate Professor at UTech and Ni3 Country Director.

**Conflicts of Interest:** The authors declare no conflict of interest. The funding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

## **References**



Baker, Paul. 2006. *Using Corpora in Discourse Analysis*. London: A&C Black.


Bernstein, Basil. 1990. *The Structuring of Pedagogic Discourse*. London and New York: Routledge.












© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

MDPI Books Editorial Office E-mail: books@mdpi.com www.mdpi.com/books

www.mdpi.com ISBN 978-3-0365-2466-5 (PDF)