Concealment and evasion in crime reporting through grammar
In Chapter 2, we looked at how participants are represented through lexical choices and at how the words chosen to do this can signify discourses that influence the way we perceive (deviant) participants, events and circumstances. In this chapter, we analyse how we perceive criminals and those dealing with them can be shaped not only by word choices but also by grammar, in terms of their representation through ‘transitivity’. Again, this can play a part in promoting certain crime discourses and ideologies that are not overtly stated. We also explore how our perception of criminals can be influenced not only in terms of what they do but also by who or what they are or how they are named. As we will see, the nature and gravity of what they have actually done can be completely transformed by this process. Later in the chapter, we look at the strategies in language for representing social actors and how to analyse these. We end by showing how our linguistic social actor analysis can also be applied visually.
Transitivity is the study of what people are depicted as doing and refers, broadly, to who does what to whom, and how. This allows us to reveal who plays an important role in a particular clause and who receives the consequences of their action. A transitivity analysis of clause structure shows us who is mainly given a subject (agent/participant) or object (affected/patient) position. Deriving from the work of Halliday (1994), transitivity in this sense goes beyond traditional grammatical notions, which distinguish between verbs that take objects (‘The burglar hit the resident’) and verbs that do not (‘The burglar fled’). For Halliday, the grammar of a language is a system of ‘options’ from which speakers and writers choose according to social circumstances, with transitivity playing a key role in ‘meaning making’ in language. This means that the choice of certain linguistic forms always has significance, and is arguably often ideological. Language is always part of an intervention in the world, and we have stressed the importance of print media, in particular, in constructing our everyday world and our expectations of it through the patterns of its representation. For example, in media accounts of important political events, such as demonstrations and other acts of civil unrest, the responsibility of authorities and police may be systematically backgrounded or omitted; agency and responsibility for actions may be left implicit. As we have seen so far, news does not simply reflect reality, but is shaped by political, economic and cultural forces. This makes transitivity analysis not only a useful tool for finding out what is in texts, but also for what is absent from them. We can therefore ask ourselves why some information is omitted from a text – information which logically should have been there.
In an analysis of a text, we can therefore ask which kinds of participants are foregrounded and which ones are backgrounded, such as through the use of passive verb sentences. For example, van Dijk (2000) has demonstrated that ethnic minorities are mostly shown as active agents when they do something ‘bad’. When things are done for or against them, they are represented in a passive role:
Muslims prisoners sue for millions after they were offered ham sandwiches for Ramadan (London Evening Standard, 26 October 2007)
Muslim prisoners get their own cells after sharing row (Daily Mail, 21 June 2009)
In the first headline, Muslims are the active participants in ‘suing’ the prison, whereas in the second, they are the passive recipients of a privilege. Both headlines construct this as something negative, because prisoners should neither sue nor be given privileges. Both headlines also ‘other’ the participants in terms of their ethnicity (‘Muslims’) and in terms of their status as prisoners. Hoey (2000) points out that evaluative words that are placed at the beginning of sentences are more difficult to challenge, as the reader is not positioned to decide whether to agree or disagree with these evaluations. That does not mean, of course, that readers do not resist information they are presented with.
When analysing agency (who does what to whom) and action (what gets done), we are interested in describing three aspects of meaning:
● participants (this includes both the ‘doers’ of the process as well as the ‘done-tos’ who are at the receiving end of action; participants may be people, things or abstract concepts)
● processes (represented by verbs and verbal groups)
● circumstances (adverbial groups or prepositional phrases, detailing where, when and how something has happened). (Simpson and Mayr, 2010: 66)
For example, in the sentence ‘Three jobless hoodies attacked a pensioner yesterday’, the Actor element is the ‘three jobless hoodies’ who carry out the process of attacking. The Goal is the ‘pensioner’ who has been attacked, and the Circumstance is ‘yesterday’, which locates the process in a temporal context. So in a transitivity analysis, we first have to identify the participants in a clause and then the process types used. Halliday distinguishes six process types: material, mental, behavioural, verbal, relational and existential. Let us now take a closer look at each of these in turn.
Material Processes. Material processes describe processes of doing. Usually, these are concrete actions that have a material result or consequence, such as ‘The police arrested the burglar’, although they may also represent abstract processes, such as ‘The crime rate has fallen’, or metaphorical processes, such as ‘She demolished his argument about crime reduction’. The two key participants in material processes are the Actor and the Goal. The Actor is the part which performs the action, and the Goal is the participant at whom the process is directed (the direct object in traditional grammar). Some material processes have one participant only, the Actor, as in ‘The burglar ran away’. This is called a non-goal-directed material process or a ‘non-transactional’ process (van Leeuwen, 2008).
Mental Processes. Mental processes are processes of sensing and can be divided into three classes: ‘cognition’ (verbs of thinking, knowing or understanding), ‘affection’ (verbs of liking, disliking or fearing) and ‘perception’ (verbs of seeing, hearing or perceiving). Examples of the three classes of cognition, affection and perception are, respectively, ‘I understood the story’, ‘He feared he might be attacked again’ and ‘The police woman saw them’. Mental processes allow us to gain an insight into the feelings or states of mind of certain participants (‘People worry about the recent spate of burglaries in their area’).
Behavioural Processes. Behavioural processes denote psychological or physical behaviour, like watch, taste, stare, dream, breathe, cough, smile, and laugh. Semantically, they are in between material and mental processes. For example, ‘look at’ and ‘listen to’ are behavioural processes, but ‘see’ and ‘hear’ are mental processes. Behavioural processes are also in part about action, but, unlike material processes, the action has to be experienced by a single conscious being, that is, a person (‘The pensioner heard suspicious noises’).
Verbal Processes. Verbal processes are expressed through the verb ‘to say’ and its many synonyms. A verbal process typically consists of three participants: Sayer, Receiver and Verbiage. The Sayer can be a human or human-like speaker, as in ‘The police spokeswoman explained the situation’, but it can also be an inanimate item, as in ‘The paper says he is guilty’. The Receiver is the one at whom the verbal process is directed (‘They told me to leave at once’), while the Verbiage is a nominalized statement of the verbal process (‘The paper gave a detailed account’ or ‘The defendant said that this wasn’t true’).
Relational Processes. These are processes that encode meanings about states of being, in which things are stated to exist in relation to other things. They are often expressed through the verb ‘to be’, which is the most frequent, but synonyms such as ‘become’, ‘mean’, ‘define’, ‘symbolize’, ‘represent’, ‘stand for’, ‘refer to’, ‘mark’ and ‘exemplify’ are also classed as relational processes. To ‘have’ in the sense of possessing something is another relational process (‘She has a criminal record’). Relational processes allow us to present as ‘facts’ what could be classed as opinion (‘Many offenders have cognitive deficits’).
Existential Processes. Existential processes represent something that exists or happens, as in ‘There has been an increase in property crime’. Existential processes typically use the verb ‘to be’ or synonyms, such as ‘exist’, ‘arise’ or ‘occur’, and they only have one participant, as in ‘There was an assault’. This participant, which is usually preceded by there is/there are, may be any kind of phenomenon and often denotes a nominalized action. In the above example, the verb ‘to assault’ has been turned into a nominalization. This can have the effect of obscuring agency and responsibility.
When we look at these processes and participants out of context, as in the examples presented above, they have no particular ideological function as such. However, things are very different when transitivity is embodied as discourse. For example, the relationship between Actor and Goal can be ideologically significant if agency is supressed through the use of the passive voice. In passives, the positions of Actor and Goal are reversed (‘One demonstrator was killed by police in riot gear’), and the Actor may even be omitted completely (‘One demonstrator was killed’). Even more backgrounding is achieved through the use of a one-participant process such as ‘One demonstrator died’, in which the action appears not to be caused by the police at all. As van Dijk has shown, ‘negative acts of in-group members, such as the authorities or the police, may be reduced in effect by placing them later in the sentence or by keeping the agency implicit, for instance in passive sentences’ (1991: 215–16).
As we will demonstrate, transitivity patterns, especially in the manipulation of agency at the grammatical level, can be significant in media representations of people in terms of power. According to van Leeuwen (2008: 60–1), material processes can be ‘transactive’ and ‘non-transactive’. Transactive processes involve a Goal (as in ‘Police finally apprehended the criminal’), whereas non-transactive processes have only one participant, the Actor (as in ‘The criminal fled’). Van Leeuwen points out that the ability to transact ‘requires a certain power, and the greater that power, the greater the range of “goals” that may be affected’. This may be important in discourses in which the police need to be seen as affecting the actions of certain Goals, that is criminals.
Another important transitivity feature is the use of nominalization, where verb processes are converted into noun constructions, thereby removing any explicit expression of agency. Thus, in ‘The introduction of ASBOs marked a new era in Youth justice’, the Actor responsible (the government) is only implied. Elsewhere, relational processes (‘be’, ‘have’, ‘represent’, ‘mark’ and so on) are often used in discourse to present information as ‘facts’, as they suggest an unqualified sense of certainty. They are also very clear in apportioning blame and expressing explicit opinion and, as Conboy observes, can be ‘a strong indicator of the position of a news institution’ (2007: 52). We can see this in a headline from the British tabloid News of the World (13 April 2008), taken from a story in which a mother was found to have organized her own daughter’s kidnapping:
Shannon Matthew’s mum is a violent, foul-mouthed, chain-smoking, boozy slob
Here the relational process ‘is’ describes Matthews in terms of her supposed identity, as a woman who has violated the traditional role of the ‘caring’ mother. This is also achieved by casting her with the noun ‘slob’, which makes it possible to put any number of adjectives in front, in this case ‘violent, foul-mouthed, chain-smoking’ and ‘boozy’, which further add to her negative evaluation. Headlines and stories like this one are symptomatic of a broader political agenda in the popular press, which obsessively focuses on criminal women who do not conform to traditional notions of femininity or ‘motherhood’. In Chapter 5, we will focus on precisely this issue by looking at the often stereotypical coverage of criminal women in some sections of the media and illustrate how transitivity can play an important part in this.
Transitivity and the representation of
social action in the media
A number of feminist criminologists and media scholars have found that the press significantly contributes to the construction of gendered discourses about male and female criminals, particularly through features such as naming or labelling, thereby perpetuating powerful gender stereotypes (see Wykes, 1998; Jewkes, 2004). As we saw in Chapter 2, naming is a significant ideological tool, because different names for objects or people represent different ways of perceiving them. But alongside the powerful connotations these lexical items may have, transitivity also plays an important part in gendered representations. Kate Clark’s (1992) analysis of both naming practices and transitivity patterns in news coverage in the Sun of violent crimes and sex crimes perpetrated against women found that the paper often used linguistic patterns that suggested that women, even when they were victims of violent assaults, were partly to blame for what happened to them. Blame, lack of responsibility, foregrounding or backgrounding of a participant can all be expressed through transitivity choices. Here is an example from the Sun quoted by Clark (1992: 213), which demonstrates this:
Girl, 7, murdered while Mum drank at the pub
Little Nicola Spencer was strangled in her bedsit home – while her Mum was out drinking and playing pool in local pubs (Sun, 20 December 1986)
Both the headline and the lead have two clauses, a main clause describing the murder and a subordinate one stating what the mother was doing at the time of the daughter’s killing, implying that she is not blameless in the girl’s death. The only Actor here is the mother, for whom the paper uses active material processes (‘drank’ and ‘was out drinking’ and ‘playing pool’). The Sun therefore passes judgement on the mother’s ‘irresponsible’ behaviour rather than on the murderer, who has been completely omitted. This may not have been an intentional strategy, as it is common to use the passive (‘murdered’) when the killer is not yet known. But the effect is still that the murderer is backgrounded and the mother foregrounded.
A more recent example from the Sun (8 October 2010) actually makes agency very explicit:
Monster rapes same woman twice
A SCARRED sex monster who raped a woman twice in separate late-night attacks was being hunted last night.
This headline and lead show how language can be used to make it perfectly clear who is to blame for an act. The ‘monster’ is shown as an Actor acting upon the ‘woman’, who is the Goal. Both the active material process (‘rape’) and the way the participants are named (the demonization or ‘othering’ through the use of ‘monster’ versus the neutral expression ‘woman’ to denote that she is one of ‘us’) also leaves no doubt as to who is to blame for the crime.
The above headline is also a good example of Clark’s (1992: 224) concept of ‘fiend naming’, in which the criminal is referred to as a ‘monster’ and a ‘SCARRED sex monster’. Such strong condemnation, while apparently blaming attackers, also suggests that they are so evil and alien that they are outside humankind and society. By positioning ‘monsters’ in this way, a discussion of why society produces men who commit violent acts against women and children is sidelined. A consequence of this dehumanizing strategy is that it glosses over those very aspects of the patriarchal system that need to be challenged. No ‘normal’ man, it is implied, would be capable of such an act. The Sun online article came with an ‘e-fit’ of the suspected rapist, depicting a black man with a scar above his lip, biologically categorized by means of stereotypical features, such as the big lips and the tightly curled hair. Two readers provided the following online comments:
pls get this animal in human skin off the streets pls . . . .
This is why DNA should be taken at birth or on entry to the UK, eventually rape would be almost unheard of
The first one, referring to the rapist as an ‘animal in human skin’, shows that this reader sees rapists as basically subhuman creatures, whereas the second one contains a racial element, suggesting that many rapes would stop if only foreigners had their DNA taken ‘on entry to the UK’. A number of feminist and criminological studies, however, have demonstrated that many rapes are committed by people known to the victims, such as relatives, partners and parents (see Jewkes, 2004).
Clark also suggests that the Sun distinguishes between ‘genuine’ victims and ‘non-genuine’ ones. For example, when a violent crime is committed against a ‘respectable’ or sexually unavailable victim, labelling will be neutral, as in ‘a woman’ or ‘his daughter’. Naming can, however, be different when the crime is committed against a victim that is deemed ‘unrespectable’, as we shall see now.
The third example is from an article about the ‘Suffolk Strangler’, who killed five women within a short period of time in 2006. The murdered women had worked as prostitutes, and the following extract describes the situation of one of the five victims:
Tragic Mother called Crackhead Annie
Anneli Alderton met her violent death after being sucked into a life of prostitution. The Ipswich-born 24-year-old’s cravings were so bad that other hookers described her as ‘crackhead Annie’. Tragic Annie, whose naked body was found dumped near Nacton, Suffolk, only began walking the streets of Ipswich in recent weeks. (Sun, 13 December 2006)
In terms of transitivity, through the use of passive constructions, the woman is cast as a victim (‘. . . after being sucked into a life of prostitution), but the text also underscores her active role in it by using two material processes (‘. . . met her violent death’,‘. . . began walking the streets’) as if she was not completely blameless in her death. She is also identified and labelled through a number of personal details, such as her name, age and appearance (the article shows a black-and-white photograph of her face, and she is also unnecessarily described as a ‘blonde’, whose body was found ‘naked’. We also learn that she is a mother. According to Clark, personal details, instead of individualizing and personalizing victims, often do the opposite, and in doing so often create ‘a voyeuristic rather than sympathetic reading of events’ (1992: 222). Some of the labels used here appear to be sympathetic to the victim (‘tragic mother’ and ‘tragic Annie’), but others focus on the fact that she was a drug addict (‘crackhead Annie’) and a prostitute, although she is not labelled in this way by the paper, which only quotes ‘other hookers’ as calling her a ‘crackhead’. Quoting is an important ideological strategy, as it allows papers to write down statements but at the same time to distance themselves from them, as they were made by other people.
Clark’s (1992) conclusion to her study was that ‘under its veneer of moral indignation against fiends’ the Sun actually shifts blame subtly onto the victims of violent sex crimes, depending on how ‘respectable’ they are. We can detect the same process in the example quoted above.
This discussion of transitivity and naming patterns has highlighted the fact that much media discourse is constructed around ‘essentialist’ notions of male and female offenders (Jewkes, 2004); the notion, in other words, is that men are biologically prone to (sexual) violence, whereas women are biologically predisposed to caring and nurturing and not to violence. Women who fail to conform to the cultural stereotypes of the maternal, caring and monogamous woman often receive particularly vicious coverage in the popular press, as we shall see in one of our case studies below. We will come back to this very issue in Chapter 5, which focuses on women and crime.
Moving on to other verb processes that can also play an important role in shaping how we perceive participants, we have to discuss mental and verbal processes. This is important, as texts not only tell us what people do or what has happened but also how people feel about things. Van Leeuwen (2008: 56) points out that social roles, as reinforced in texts, prescribe not only actions and identities but also feelings.
Starting with mental processes, it is often the case that participants who are made the subjects of mental processes are constructed as the ‘focalizers’ or ‘reflectors’ of action. These actors are allowed an internal view of themselves. This can be one device through which listeners and readers can be encouraged to have empathy with that person. For example:
James Bulger’s mother still suffers 15 years later. (Daily Mail, 3 March 2008)
Here the reader is encouraged to sympathize with the mother by being informed of her suffering. We are reminded of the murder of a two-year-old boy by two youngsters who were only ten at the time of the murder. But newspapers covering the case by and large did not offer corresponding details of the feelings of these two boys and their parents. So we learn nothing of the mental processes of the two boys nor the concerns and sufferings of their parents. This serves to alienate us from these participants and, in a wider sense, to justify the way the criminal justice and penal system dealt with them.
Since mental processes are mainly about sensing and reacting, they can also convey passivity. Consider the following newspaper headline:
Prison officers fear that Muslim inmates are turning to extremism (Times, 10 October 2008)
On the one hand, in the sentence above, we are allowed an internal view of the prison officers. But the impression is that they are passive and helpless against Muslim inmates, which in the case of running a prison may worry readers.
Very often in official mediated discourses on crime, officials, experts, citizens and victims of crime are represented not only through their actions but also through their ‘reactions’ (van Leeuwen, 2008). Citizens, for example, may ‘fear’, ‘have legitimate fears’, ‘be afraid’ or ‘feel besieged’, so the question of who is reacting to whom or what can be relevant for a critical analysis. Many reactions are not expressed ‘dynamically’, as in the mental process ‘they fear’, but more statically, as in ‘they are afraid’. Van Leeuwen’s category of ‘reactions’ is an important complement to Halliday’s, as the category ‘mental process’ does not fully cover the category ‘reaction’. In our case study below of a text on Muslim prisoners, we will see that the actions of the prison staff are mere reactions to the actions of the Muslim prisoners.
Abstractions
Abstrations describe actions which are generalized and non-specific and abstract away from the more specific micro-processes that make up actions (van Leeuwen, 2008: 69). For example, in the sentence
It is important for staff to engage with Muslim prisoners
the details of what is done are obscured. In such cases, what is done may be less important than that staff appear to do something with the prisoners.
As another example of abstraction, the following extract is taken from the website of the Labour party in Rochester and Stroode, entitled
Stamp out anti-social behaviour
Medway Labour is committed to combating anti-social behaviour, kerb crawling, drug usage and alcoholism that blights communities in Medway.
We can see here that relational processes such as ‘is committed’, a phrase that has become a staple in the rebranding of political discourse, is used to abstract what Labour actually do in this respect. Being ‘committed’ is not the same as actually combatting something. The website then explains in bullet points what Labour intends to do in this respect. Here we quote three of these:
● Promote neighbourhood watch schemes and the Special Constable programme Medway-wide
● Support basic community policing and voluntarism. Empowering local people to campaign against crime
● Opposing kerb-crawling with tough action on offenders and tighten up rules on street prostitution.
All these are examples of abstractions, which do not make it clear what exactly is being done against ‘anti-social’ behaviour. When we find such abstractions at the level of social action, we have to ask why they are being used and what is being concealed.
Grammatical positioning of actions
Prepositional phrases and subordinate clauses
Another important strategy of representing social action is through a circumstance, such as a prepositional phrase or subordinate clause. Both can be useful strategies in backgrounding certain acts. Prepositional phrases begin with a preposition, such as ‘for’, ‘at’ or ‘after’. A newspaper headline might use a main clause and a prepositional clause, as in ‘Boy stabbed at school’; the prepositional phrase tells us where the stabbing happened. A subordinate clause will begin with a conjunction, such as ‘while, ‘after’ or ‘as’, or a relative pronoun, such as ‘which’ or ‘whose’. Grammatical structures like these can have an ideological effect. For example, in the sentence
Violent protests as Britain pushes through student fee hike
the reason behind the violent protest – the actions of the British government, that is, their decision to raise study fees – are de-emphasized by being embedded in a subordinate clause through a nominalization (‘student fee hikes’). Van Dijk (1991: 216) points out that ‘[e]vents may be strategically played down by the syntactic structure of the sentences, for example, by referring to the event in a lower (later, less prominent) embedded clause, or conversely by putting it in the first position when the events need extra prominence’ (1991: 216). In our example, the ‘violent protests’ are made to stand out while the actions of the government are totally backgrounded.
When analysing newspaper headlines, Richardson (2007: 207) found that prepositional phrases can be used to provide context for dominant clauses, but they thereby also reduce responsibility for an action, as it is de-emphasized. This can be seen in the following headline and lead below:
12 dead in attack on Hamas
Seven children killed as Israelis assassinate military chief (Guardian, 23 July 2002)
The headline does not reveal who is behind the killing. Although this can probably be inferred, the prepositional phrase ‘in attack on Hamas’ requires some extra effort on part of the reader to work out that the ’12 dead’ is the result of an attack by the Israelis. The subordinate clause in the lead, (‘. . . as Israelis assassinate military chief’) almost reads as if it the assassination was a separate incident, although, again, the reader would be able to make the connection. However, the lead could have stated that ‘Israelis kill seven children in assassination of military chief’, which would have made it clearer that the Israelis were responsible.
As we have seen so far, attributing agency through Halliday’s list of verb process types can be complex. We must be careful to see who is activated and also look for ways how this can be concealed or downplayed through a number of grammatical strategies.
In our case study, we conduct a transitivity analysis of a newspaper text from the Times (10 October 2008) which reports on the danger posed by radical Muslims at a ‘top security jail’. There are a number of important language choices found in this text which help to shape the story ideologically. In a later section, we look at the way in which the participants are represented plays one important role in this ideological construction. Here we are interested specifically in transitivity. The important instances of transitivity used for the prison officers have therefore been put in italics.
Prison officers fear that Muslim inmates are turning to extremism
Staff at a top security jail fear Muslim inmates are pressurising other Muslim prisoners to adopt extremist views and encouraging other offenders to convert to Islam, according to an inspection report published today.
Prison officers at Whitemoor jail also feared growing radicalisation and conversions among the 395 inmates, of whom almost one third were Muslims.
On one wing at the jail staff admitted that Muslim inmates ‘policed themselves’, the report said.
One inmate claimed to inspectors that inmates are converting to Islam because they want protection and that Muslim gangs in the jail provide it.
Anne Owers, the chief inspector of prisons, urges the prison service to do more to provide staff throughout the jail system but particularly in the top security prisons with help to deal with increasing Muslim numbers.
The number of Muslim prisoners in jails doubled in the ten years to 2006 to reach 8,243 – 11 per cent of the total prison population.
Ms Owers said the top security prisons are facing increased risks involving more gang activity, more young men serving long sentences and a small number of men convicted of terrorist offences.
She said Whitemoor jail at March in Cambridgeshire had faced a sudden expansion in black and ethnic minority prisoners to reach 150, of whom 120 were Muslims.
A committee had been created to advise on action to deal with possible gang, terrorist or extremist activity at the prison even though only 8 of the Muslim prisoners had been convicted of terrorist offences.
Despite these strong central systems, residential staff were mostly unaware of these initiatives. They expressed a fear of what they saw as a rising problem of prisoner radicalisation and an increase in Muslim conversion’, the chief inspector’s report said.
Staff were unsure about how to treat Muslim inmates other than as potential security risks. The report said officers were reluctant to engage with inmates and deliberately kept their distance from them because they feared they would be ‘conditioned’ by prisoners.
‘Some staff appeared reluctant to engage with Muslim prisoners, and the little information and training they had received about Muslim prisoners was related to monitoring them as potential threats to national security, which inevitably impinged on the way they interacted with them’, the report said.
It added: ‘Prisoners said that staff attitudes towards them changed markedly for the worse if they chose to, or happened to, associate with those prisoners.
‘There was a perception among officers that some Muslim prisoners operated as a gang and put pressures on non-Muslim prisoners to convert, and on other Muslim prisoners to conform to a strict and extreme interpretation of Islamic practice.
‘However, there was a reluctance to engage with Muslim prisoners and challenge inappropriate behaviour. An officer on one wing said that Muslim prisoners ‘policed themselves’ and that others in the staff group agreed’.
Describing the gang culture at the jail, one inmate said: ‘On the main wing it is Muslim vs whites. Staff are worried as what will they do when it all goes mainstream. They are beginning to outnumber everyone and don’t care – all this radicalisation and they’re extremely violent slashing people’.
Another prisoner said: ‘The new gang are the Muslims. The Muslim group is the big group and others are looking for protection. Those who are isolated are looking for protection and so are the ones converting as they won’t get help from screws (prison officers).
Ms Owers report said resettlement work had improved along with activities available to prisoners but she said the jail still faced challenges.
Phil Wheatley, director general of the National Offender Management Service, said: ‘The Chief Inspector is right to highlight the challenges and risks Whitemoor is facing. It is also important to recognise the action being taken to manage challenging prisoner profiles.
‘A more sophisticated approach to addressing bullying and the management of bullies and their victims is now in place and is bringing improvements.
‘Work to improve the relationships between staff and prisoners is a priority and measures have been implemented to tackle this, including training to develop staff understanding of the growing Muslim population’.
Starting with the transitivity patterns for the prison officers, we can see that they are represented through a high incidence of the mental process ‘to fear’:
Prison officers fear; staff at a top security jail fear; prison officers [. . .] also feared growing radicalization; feared they would be ‘conditioned’ by prisoners.
Then there are several relational processes:
staff were unsure; appeared reluctant to engage; are worried; were mostly unaware; are worried.
We also find three verbal processes in which officers express or even ‘admit to’ their fears.
staff admitted; An officer on one wing said . . . others in the staff group agreed; expressed a fear.
And finally two existential processes:
There was a perception among officers
There was a reluctance to engage with . . .
These noun constructions are nominalized processes, one of which is mental (‘to perceive’) and one relational (‘to be reluctant’), and which further add to the general impression of inactivity on the part of the officers.
All these processes are basically what van Leeuwen (2008) calls ‘reactive’ and add to an overall tone of passivity on part of the prison officers. All they do is ‘deliberately keep their distance’ from Muslim prisoners.
The transitivity patterns for the Muslim prisoners are entirely different. They are mostly engaged in ‘transactive’ actions:
are pressurizing other Muslim prisoners; encouraging other offenders to convert to Islam; operated as a gang; put pressure on non-Muslim prisoners to convert; are beginning to outnumber everyone; they are extremely violent slashing people.
All except one of these processes have Goals (‘other Muslim prisoners’, ‘other offenders’, ‘non-Muslim prisoners’, ‘everyone’, ‘people’) who are affected by the actions of the Muslims. Only one mental process describes their attitude (they don’t care), and one relational process presents as a fact that they are extremely violent.
Looking at how the prison authorities are represented, we encounter many verbal processes, particularly for the chief inspector of prisons, whose report is also quoted. These processes can also be seen as reactions:
urges the prison service to do more; said; the chief inspector’s report said; the report said; it added; Ms Owers report said.
Here we find that the prison authorities, the ‘experts’ make authoritative statements by using the neutral verbal process ‘said’, addressing the problem in a professional way. Only once does the chief inspector ‘urge’.
We also learn that
A committee had been created to advise on action to deal with possible gang, terrorist or extremist activity at the prison . . .
Phil Wheatley, director general of the National Offender Management Service is quoted as saying:
It is also important to recognize the action being taken to manage challenging prisoner profiles.
The italicized bits in these two sentences are examples of abstraction. We do not really know what exactly is done to address the situation in the prison, but the reader may feel assured as long as the authorities appear to do something. Edelman (1977) reminds us that one of the functions of crime-control discourse is to convey rational decisions, change and progress, even if this is not the case. The fact that there is an ‘Offender Management Service’ through which prisoners and their ‘profiles’ can be ‘managed’ demonstrates this strategy on behalf of the prison service to do something about the problem. It is also an example of the influence of managerialism and its language on the prison system, the overall concern of which is the ‘efficient enhancement of social control’ (Garland, 2001: 176).
Apart from the Muslim prisoners, other anonymized prisoners are also present in the text who are represented in terms of verbal processes:
One inmate claimed to inspectors that Muslims are converting to Islam; prisoners said; Describing the gang culture at the jail, one inmate said; another prisoner said.
We assume these prisoners are non-Muslim, although this is not stated explicitly all the time. Apart from ‘claim’, which suggests that a statement should perhaps not be believed, the neutral ‘say’ is used. This would probably not have been the case if the article had been about prisoners protesting against their conditions. In the quote below by one (non-Muslim) prisoner, in which he describes the Muslims as the dangerous ‘other’ in ‘us’ versus ‘them’ terms, it may be significant that the neutral ‘say’ is used and not the doubtful ‘claim’ or ‘allege’:
Describing the gang culture at the jail, one inmate said: ‘On the main wing it is Muslim vs whites.
To sum up, in any text, we can ask how many process terms describe actions and how many reactions. Just as with many newspaper texts on immigrants (who usually provoke reactions) in which there may be a predominance of reactions attributed to those portrayed as ‘us’ (see van Leeuwen, 2008), the Muslim prisoners in this text are also depicted as producing reactions from prison officers, who ‘fear’ them, and from the prison authorities, who react with reasonable measures. Although the main focus in this text is on the prison officers and the authorities, and the actions of the Muslim prisoners are backgrounded in subordinate clauses, the text still manages to convey an overall sense of threat emanating from Muslims who impose their religion on British culture, even behind prison walls.
The representation of participants in crime reporting
One other important language feature employed in the text analysed above is its use of representational strategies. It is probably fair to say that there is no neutral way to represent a person. All choices will serve to draw attention to certain aspects of identity that will be associated with certain kinds of discourses. For example, take the following sentences:
Three jobless hoodies attacked a pensioner yesterday.
Three unemployed men attacked a man yesterday.
Three Muslim men attacked a war veteran yesterday.
Three young men attacked an ex-soldier yesterday.
All of these sentences are used to describe the same event and the same participants. But in each case, the participants have been represented by different naming choices, each of which serve to signify a different set of associations. In each case, certain aspects of identity have been emphasized and others de-emphasized. Some of these serve to heighten the severity or immorality of the event, and others to diminish it. The expression ‘jobless hoodies’ places the attackers in the discourse of ‘feral youth’ undermining decent society. In a tabloid news article, it would be no surprise to the reader to find them attacking a pensioner, and he or she may immediately dismiss them, not as part of a symptom of a broader issue of social exclusion but of immoral behaviour and disrespect and society on the verge of collapse.
In the next line, ‘unemployed men’ makes them appear less of a ‘burden’ and appears more ‘objective’. But this is helped by the victim now no longer being a vulnerable ‘pensioner’, but simply a ‘man’.
The third sentence foregrounds the religious beliefs of the attackers and the ‘decency’ of the victim by foregrounding his established membership in society because he is a ‘war veteran’. Here the discourse signified could be the disrespect to cherished British cultural values by those who are ‘other’.
Finally, the meaning of the last sentence becomes ambiguous, as we are told of the actions of ‘three young men’, which has fairly positive connotations, and the ‘ex-soldier’ is less emotive than the ‘war veteran’.
In these sentences, all of the naming strategies could in some ways describe the participants accurately. The attackers were Muslim, wearing hoodies, unemployed and young men. They could also be many other things, relating to different aspects of their lives. They might be members of a football team, residents of a particular street, former school poetry prize winners, someone’s child. But choosing to foreground one aspect of their identity has an ideological effect. It can serve to legitimize or delegitimize. We can see this in the following sentences:
Muslim man arrested for fraudulently claiming benefits.
Father of two daughters arrested for fraudulently claiming benefits.
In fact, there could have been many other possibilities that could have been used to characterize this man. It could have been foregrounded that he was a British man, an amateur dancer, a local office worker, and so forth. But in the first place, it is his official religion that has been chosen. Again, we find the discourse of those from ‘outside’, who are a burden on the British taxpayer. In the second sentence, by foregrounding the fact that he is a father, the crime appears more forgivable, as this humanizes him. In this sense, the same act is presented in a very different way for the reader. Crime reporting usually involves creating moral others, so that the perpetrator is not like us (e.g. Wykes, 2001). In the second case, he is ‘one of us’. The effect here is that possibly the fraud was justified, as he was struggling to look after his children. In the case of the attack on the pensioner above, some of the representational strategies are more useful for creating moral others.
Classification of social actors
Van Leeuwen (1996) offers a comprehensive inventory of the ways that we can classify people and the ideological effects that these classifications may have. This inventory helps us to take a more systematic approach to analysing representational strategies in texts. We first provide an overview of van Leewen’s categories before returning to the article from the Times above on the Muslim prisoners (pp. 62–63).
Inclusion and exclusion
In the first place, we can simply consider who is excluded from a text. What kinds of participants are missing or suppressed?
Demonstrators were taken into custody throughout the day.
Here, the action has been included, but the agents of that action, in this case the police, have been excluded. It may be the case that in this text it was the actions of the demonstrators that were to be foregrounded for ideological reasons. In the following example, we can see that the police are absent from the text, as they wish to background their inability to develop a case, as is made clear in the second sentence:
Certain facts in the case were difficult to establish.
The police were unable to establish any facts.
In the following, we can see suppression rather than exclusion, as it is harder to tell who the actual agent is:
Concerns are being widely expressed in Britain that tougher measures on crime are needed.
In this sentence, who is actually expressing the concerns is suppressed through passive agent deletion. Nevertheless, someone is expressing concerns.
In texts, we may find that some social actors are completely absent. A text or video game on war may include no civilians. One of the authors conducted a short study with a group of students on WWI literature for British schools. Apart from the Home Front, there was nothing included on the lives of civilians in the areas in and around the combat zones. Anyone who has read novels by front-line soldiers will be confronted by the appalling effects of war, in the first place, on civilians, who will not be provided with supplies in the same way that the soldiers of each side are. In such cases, we can ask why such suppression has occurred.
Personalization and impersonalization
When participants are included, we can ask to what extent they are personalized or impersonalized. This simply means the difference between terms that describe participants through proper names or nouns, or even humanizing adjectives, and those that dehumanize and objectify, such as calling people ‘the rich’ or ‘the overweight’ or simply ‘problems’. This can be seen in the following two sentences:
Detective superintendant Paul Cobley thanked the public for their support during this demanding investigation.
A police representative thanked the public for their support during this demanding investigation.
In the first case, the police appear much more human due to the use of names. In the second sentence, impersonalization gives much less of the sense that the police are actually engaging personally with the public. Later in the book (Chapter 6) we will be looking at strategies used by the police in personalizing the image of the force and in fomenting the sense that they are acting in and alongside the community.
On the other hand, impersonalization can be used to make a more official statement or to obscure the actual source of information:
The Kent police force will not be accepting strike action by officers in the face of the threatened cuts.
In this case, we could argue that it is the officers who are the police force. So we can ask here who this viewpoint actually represents. Of course here, too, we find suppression of the actual agent.
Impersonalization can be carried out by objectification and abstraction. In the following sentences we can see that young men have been objectivated by being represented through a single feature or thing with which they are associated:
Hoodies are to be banned from the shopping centres.
This means that participants can be reduced to this feature. A tabloid newspaper might refer to a group of youths as ‘hoodies’ throughout an article. This can distract attention from exactly who these people are, for example, young people who have experienced poor education, have had few opportunities and have little to gain from mainstream society in terms of self-esteem.
Individualization and collectivization
It is also useful to consider how participants are described as individuals or as part of a collectivity, as is shown in the following sentences:
It was announced that police officers had been injured during the demonstration.
It was announced that police officer Crispin Thurlow, ten years with the force, had been injured during the demonstration.
In the individualization found in the second sentence, we are more likely to align ourselves with the participant. We can imagine a text in which it is the demonstrators who are all individualized, giving details of their names, ages, job, family status and so on, and in which the police are all collectivized. In this case, we would be most likely be more aligned with the demonstrators.
Specification and genericization
We can also look at whether participants are represented as specific individuals or as a generic type. In our earlier example, we saw that the person accused of benefits fraud could either be named or identified as a type. Consider the following two sentences:
A man, Mazar Hussein, a local shop owner, challenged police today.
A Muslim man challenged police today.
In the second case, the man who challenged the police is represented as a type (‘a Muslim man’). This is used here to place the story in a particular news frame. In this case, the generic category of ‘Muslim’ can place this story into a news frame in which Muslims are a contemporary problem in Britain, either due to their extremism or their complaining about their situation. In fact, the man may not have been a practising Muslim. It could be like saying, ‘Christian John Smith challenged police today’. It is the use of such generic terms that can be used to give a racist angle to a story even when the newspaper takes a stance that it is not racist.
In the following example, we see how a specific naming strategy can make us much more likely to align ourselves with an offender:
One demonstrator was sentenced by Judge Smidgely-Smithely.
The demonstrator Joan Judd, 43, mother of three and local librarian, was sentenced by Judge Smidgely-Smithely.
In the second sentence, naming strategies are used to foreground the way this woman may have been ordinary, decent or gentle. Such people would not demonstrate unless there was good reason. In the first sentence, this is suppressed in favour of a generic term in order to background the possibility that there could easily be widespread support for the demonstration.
Functionalization and identification
Participants can be identified or nominated in terms of who they are or functionalized by being depicted in terms of what they do. Functionalized terms often end in -or, -er, -ant, -ian, -ist or -eer. So we can have ‘police interviewer’, ‘rapist’, defendant’, etc. For example,
The demonstrator was injured outside the embassy.
The defendant was warned by Judge Peter Smidgely-Smidthely.
In these cases, the demonstrator and defendant are partially dehumanized by referring to them with functionalizations that highlight only their roles. Had both of these been named and identified by mention of them being mothers, for example, we would have evaluated them differently.
Functionalization can also connote legitimacy:
The local shop owner said that he needed greater police protection.
Here the functionalization positively evaluates the speaker as a legitimate, decent member of a local community. He could have been nominated as ‘a local’ which may not have had the same effect.
Representation of people through what they do can also be achieved with the use of ‘functional honorifics’. These are terms that suggest a degree of seniority or a role that requires a degree of respect. These will normally involve official roles, such as ‘President’, ‘Lord’ or ‘Judge’. These can be used to foreground the fact that they are important and official. We might find that different ideological accounts of the same set of events will see honorifics ascribed or withheld. In a press release, the police may want to emphasize that the spokesperson is very important, a chief superintendent’, but also nominate them to provide personalization.
Identification means that people are described in terms of who they are. This could include how they are classified by sex, age, race and so forth. This can also be accomplished with physical identification, such as ‘small’, ‘tall’ or ‘pretty’. How people are classified can tell us a lot about the ideology of the definer and the social climate in which this is considered relevant. Identification can also be accomplished by relationality, such as ‘an aunt’, ‘a mother’ or ‘a colleague’. Relational identification was formerly more important in Western cultures, in which individuals were asked about their family connections, as compared to more contemporary times, in which we tend to be asked what it is that we do.
It is of ideological importance whether a person is functionalized or identified in a text and, if both, in which order this is done. When the innocence of a victim is required, relational identification and classification may be more important than functionalization, as in, for example,
The petite, pretty, mother of two was attacked on her way home.
The solicitor was attacked on her way home.
In terms of order of functionalization and identification, we find that what is placed first can provide clues as to the ideological defining of identity and its history. For example, we can have a ‘woman doctor’ but not a ‘doctor woman’. In this case ‘woman’ is clearly of paramount importance. In other cases, we can have ‘policewoman’ as the term ‘policeman’ already existed.
Anonymization
Participants can also be anonymized:
A police source said today that there was no further information on the case.
Many people believe that criminals should receive harsher sentences.
The term ‘a police source’ is very common in newspapers. On the one hand, we rely on journalists to have legitimate sources, but this conceals which social groups and organizations actually provide these sources. In the second case, we can see how a politician could use such a representational strategy (‘Many people’) to avoid specification and the development of detailed arguments. It allows him or her to conveniently summon up arguments that are then easy to dismiss.
Aggregation
Aggregation means that participants are quantified and treated as ‘statistics’:
Many thousands of criminals are going unpunished through soft sentencing.
Van Dijk (1992) shows that this use of statistics can be used to give the impression of research, of scientific credibility, when in fact we are not given specific figures. Is ‘many thousands’ 3,000 or 10,000, for example. Below, we see this in a news agency feed:
One of the few suspects to express remorse over his alleged involvement in last year’s bombings on Indonesia’s Bali island arrived at court on Thursday.
In this case, how many is a few? Exactly how many have shown remorse and how many have not? And if we are not told, then why not? So in cases of aggregation, in which actual numbers are replaced by such abstractions, we can always ask what work this does for the author.
Drawing on this tool kit for considering the meanings of representational strategies, we can now think about the way these strategies work in the text from the Times (10 October 2008) above (pp. 62–4) about the Muslim prisoners
In this text, there are four categories of participant: Muslim prisoners, non-Muslim prisoners, prison staff and prison service officials (the chief inspector of prisons, and the director general of the National Offender Management Service). The members of the first category are always represented through the word ‘Muslim’. We find extensive overlexicalization, with the term ‘Muslim’ being repeated 20 times. Every time a prisoner is mentioned, even in collectivized form, such as ‘Muslim prisoners’, ’Muslim inmates’, ‘the Muslim population’ or ‘Muslim gangs’, the generic category ‘Muslim’ is used. Whenever we find such overlexicalization of characteristics to represent people, we can assume that some kind of over-persuasion is taking place, which is normally evidence that something is of ideological contestation. These prisoners are never individualized, which has the effect of saying that they are all the same. In only one case do we get a longer nominal group, describing some of these prisoners in more detail as ‘more young men serving long sentences and a small number of men convicted of terrorist offences’. Here the word ‘Muslim’ is left implicit through the use of the term ‘terrorist offences’. We do not hear any of these Muslim prisoners’ views about the supposed radicalization that is taking place in some English prisons.
The other, ‘non-Muslim prisoners’ are collectivized through terms such as ‘other offenders’ and individualized twice through the use of ‘one inmate’ and ‘another prisoner’, which is when they tell the reader that ‘inmates are converting to Islam because they want protection’, that on ‘the main wing it is Muslim vs whites’ and that ‘the new gang are the Muslims’. Although they remain anonymized, they are allowed to speak about their sense of alarm regarding what is going on.
The people who work for the prison are collectivized and functionalized as ‘prison officers’, ‘staff’, ‘some staff’, ‘officers’, ‘residential staff’ and ‘screws’, but are never individualized or allowed to speak. This further enhances the sense that they are non-active, as the transitivity analysis above (pp. 64–66) has shown, and that they are ‘overwhelmed’ by the Muslims. It is odd, though, that the reader is not informed about their views. Instead, we learn that they ‘fear’ that Muslim inmates are turning to extremism.
The other two people working for the prison service, the officials, are both individualized through nomination and functionalization, and they provide the official assessment of the situation. The chief inspector of prison, Ann Owers, is quoted at length.
A further key part of the representational strategies in this text is the extensive use of aggregation, which is provided mostly through the quotes given by Ann Owers. We learn that
The number of Muslim prisoners in jails doubled in the ten years to 2006 to reach 8,243 – 11 per cent of the total prison population.
. . .Whitemoor jail at March in Cambridgeshire had faced a sudden expansion in black and ethnic minority prisoners to reach 150, of whom 120 were Muslims.
Apart from these figures, we also find that there are ‘increasing Muslim numbers’, ‘the growing Muslim population’, ‘more gang activity’, ‘growing radicalization’ and a ‘rising problem of prisoner radicalization and an increase in Muslim conversion’, all terms that add to a more diffuse sense of an overwhelming threat emanating from these prisoners.
The story analysed here uses representational strategies to overlexicalize the word ‘Muslim’ and also makes extensive use of aggregation, such as ‘increasing Muslim numbers’. Those representing the prison authorities are individualized and functionalized to foreground their professional roles, whereas the prison officers remain the anonymized and functionalized group who ‘fear growing radicalization’ and ‘appear reluctant to engage with Muslim prisoners and challenge inappropriate behaviour’. The Muslims are represented as the dangerous ‘other’, operating in ‘gangs’ and seeking to impose their religious views on other (Muslim) prisoners. Whereas the authorities are represented as neutral and official, and some prisoners are allowed to speak, the prison officers, through their anonymization, appear largely overwhelmed by the Muslim prisoners.
In summary, the transitivity and representational strategies in this text add up to the sense that a significant number of Muslims are posing a threat to prisoners, prison staff, the prison system and, in a wider sense, to British society’s values. The story could have been written very differently, emphasizing the constant overcrowding with which some UK prisons are struggling to cope. But the ‘increasing numbers’ of prisoners, Muslim or non-Muslim, remain unproblematized. Instead, we find a very different focus on ‘Muslim’ prisoners threatening and bullying both staff and inmates to adopt their religion.
The visual representation of
transitivity and social actors
The categories for the representation of social actors and social action presented in this chapter can also be applied visually. This is something we will be doing in later chapters as part of our analysis of sample texts. This is an important step, as, of course, many texts about crime carry images. In terms of visual transitivity, we can consider the extent to which social actors are represented as acting through material processes or shown in moments of reflection. Visually, in photographs accompanying a story on a demonstration, we may find the police represented as still and thoughtful, through behavioural and mental processes, simply watching the demonstration, rather than in material acts of ‘intimidation’ and ‘enforcement’. In contrast, demonstrators, who have taken to the streets out of concern for their jobs, future, or nature of the society they live in, are represented in the material acts of destruction and violence. Linguistically, texts might tell us of the police, ‘accompanying’ the demonstration rather that restricting it. Together, visually and linguistically, the police are represented as more thoughtful, measured and certainly not threatening.
In terms of social actors analysis, we could ask to what extent the police are named and individualized. We might find the police named through honorifics, such as ‘officers’, but at the same time personalized through naming. These individuals will then be represented through their ‘concerns’. In contrast, the demonstrators might be collectivized and anonymized as ‘rioters’.
Also important in this process are basic issues of how far the social actors are positioned from the viewer and how they engage with the viewer. In the first case, is the social actor positioned in close shot or in the distance? This is an old and well-documented technique in film editing to bring viewers closer to the mental world of some participants than to that of others. Kress and van Leeuwen (1996) relate this to social space. Proximity in photography and film carries the same associations as proximity in everyday social settings. Closer proximities are associated with greater intimacy. So if we see individual, named police officers positioned in medium close shot, this increases their individualization. If a crowd of demonstrators is shown in long shot, then this further anonymizes them. Of course, extreme proximity in some cases can be inappropriate. In Chapter 4, we find a (posed) photograph of a ‘hoodie’ with a knife in extreme close up to emphasize the threat and intimidation. Journalists like to film themselves in close proximity in crowd scenes to connote ‘on the spot’ reporting, even when this leaves the broader social context completely unexplored.
We can also consider whether the social actors are depicted as looking at the viewer or not. Kress and van Leeuwen (1996) argue that when participants look out at the viewer, this is a ‘demand’ image, as they ‘demand’ a response from us, if only on an imaginary level whereas if they do not meet our gaze, we are invited to simply contemplate and observe them. This is referred to as an ‘offer’ image. Both personalize and individualize the social actors depicted, as they encourage us to engage with them, particularly ‘demand’ images.
We will be using these observations on visual transitivity and visual social actor analysis throughout the book. As with the linguistic analysis, we can observe the patterns that are found for different sets of social actors.
Conclusion
In Chapter 2, we looked at lexical analysis; we discussed how we can look for patterns in word choices and how these are laid out in lexical fields. In this chapter, we have shown how we can refine our analysis by looking at particular kinds of lexical and grammatical choices related to identity and action. On the one hand, criminals may not necessarily be evaluated in terms of what they do but rather in terms of who they are. In later chapters, we will show that it may even be difficult to establish from a text what a criminal has actually done. What is more important is who they are and how this makes the immorality of any criminal act greater. We also show that some kinds of criminals, such as those involved in corporate crime, tend to resist certain types of naming strategies, so that their actions more easily escape definition as ‘crime’.
We have also shown the value in looking at what participants are represented as doing in a text. This allows us to more carefully think about agency, blame and responsibility. In later chapters, we will use this form of analysis to consider, for example, instances in which the details of what is done are obscured, in which certain participants are represented as very active and in which others are represented more through mental processes, all of which can have considerable ideological motives and consequences.