“ WHAT EXACTLY IS A PAEDOPHILE ? ” : CHILDREN TALKING ABOUT INTERNET RISK

Le discours réducteur habituellement tenu au sujet du media Internet exerce une forte pression sur l’appropriation de celuici par les enfants, comme le montrent les données tirées de l’évaluation du programme Educaunet au Royaume-Uni et présentées dans cet article. Outre l’observation et l’évaluation ante post réalisée autour du Kit européen Educaunet, les auteurs de cet article se sont attelés à comprendre, par de nombreux entretiens, quelles représentations des dangers de ce nouveau média se sont forgées les jeunes. Ils proposent ici un cadre d’analyse des discours qui ont émergé de ces entrevues, posant leur réfl exion au carrefour des mythes urbains et du « protectionnisme » qui peuvent sous-tendre les connaissances et visions d’Internet que se font les enfants. Ces interviews et leur analyse révèlent la valeur des formes exploratoires du discours, et la nécessité de laisser les jeunes “jouer” avec les mythes pour pouvoir trouver leur voie au coeur d’un raisonnement plus rationnel.


Introduction
Reports tell us that the internet is opening new dangers to children, including online grooming, exposure to pornography and financial scams (Carr 2004;Gardner 2003;UK Home Office 2001;O'Connell 2003).The result has been various initiatives which attempt to teach children safe surfing habits.The UK Home Office "ThinkUKnow" campaign featured advertisements on the radio, internet and cinemas, targeting teens and preteens with the message that the person they are chatting to "may not be who you think they are".There are indications that such campaigns have had an impact on children's awareness of "stranger danger" on the internet (Livingstone/Bober 2003).However, many organisations are still struggling with the question of how best to prevent internetrelated harm to children.
Children are exposed not only to advertising campaigns about stranger danger but also sensationalist stories about, for example, what happens to girls who enter chat rooms.When a teenage girl goes missing, police investigations routinely include looking at the girls' online activities, and tabloid media frequently make the connection between missing school girls and chat room activities.These connections are firmly embedded in the minds of the children we interviewed for the study we will be discussing.Alongside the very rational and prohibitive discourse coming from campaigns which warn children against any chat with strangers, sit the folkloric stories about girls meeting up and getting killed by paedophiles.The challenge to educators is to find an approach which will engage with both sets of discourses.
The study on which this paper is based involved piloting teaching materials which focus on various internet-related risks.Although the materials involved a school-based and therefore rational approach, they were designed with the aim of seeing risk taking as part of children's learning experience.This paper uses the data from the evaluation of the programme, particularly interviews with small groups of pupils, to discuss the discourses surrounding children's experience and knowledge of internet risk.We want to identify broad patterns in the ways children talk, and consider what these show about their perceptions of, and engagement with, varieties of internet risk.

3.
Theoretical context: discourse Kress and van Leeuwen define "discourse" as "knowledge of (some aspect of) reality" (2000).The key point about their conception of discourse is that it begins from the assumption that ordinary people have some agency in the deployment of discourses.By "agency" here we mean the power to direct social processes.Our view of discourse is, therefore, at some distance from the well known Foucaultian model (1980), in which social agents are at the mercy of deterministic discursive structures, which operate as vast historical forces in which power is inevitably concentrated in the hands of overarching state institutions.By contrast, our view sees ordinary people as, to some degree, the knowing authors of their utterances.In this way, it resembles more the vision of Volosînov (1986Volosînov ( /1929)), for whom the act of utterance was on the one hand to absorb meanings from the flow of ideology, but on the other hand for individual people to reshape them internally before returning them to the ideological flow a little altered.However, we want to imagine all possibilities along the spectrum between a deterministic model and one that emphasises social agency.Children talking about internet risk may be constructing a knowledge of reality in which they are entirely dependent on stories they have heard and imperfectly understood; and in this respect, they exhibit a weaker social agency.At the other extreme, their perceptions may be based on detailed practical knowledge and on complex syntheses of rationalistic discourse.In this case, we would see a much stronger social agency, a much greater control of the discourses they deploy, and a greater ability to reshape these for their own purposes.Many of them we would expect to fall between these extremes, expressing their understandings and their social motivations in hybrid discursive patterns.
At the risk of oversimplification, we can see the rhetoric at play in these interviews as broadly inclining on the one hand towards folkloric types of the urban myth or legend variety; and on the other hand towards rationalistic discourses.In our field of media education, this kind of polarity is common.The media cultures which are most attractive to children and young people often incorporate spectacular and thrilling genres which by their very nature appeal to the irrational impulse in human culture, such as horror films, computer games structured around combat, and comic-strip narratives of magic and contemporary myth.A problem for the media educator is that these cultures and the pleasures they provide collide sharply with the rationalistic discourses of analysis typical of media education, revolving in the UK particularly around a familiar conceptual framework of media institutions, texts and audiences (Buckingham 2003).In the case of internet education, as in other media genres, the ideal solution seems to be to help students develop skills of rational analysis without delegitimizing the pleasures of the medium, though this is much easier to say than to do, of course.
Our model of discourse, like the Kress and van Leeuwen model referred to above, is derived from social semiotics, which emphasises the social function of forms of communication, and the social interest of the communicator.The axis of social interest we will construct here assumes that at one pole, the social interest is in prohibition -don't go into chat rooms, don't give out your e-mail address or phone number, don't take risks.While the dangers may be real, the problems with prohibition and teenage risk-taking are manifold: it treats the listeners as objects of instruction, it collapses complex categories of risk into simple, exaggerated ones, and it closes down discussion.Furthermore, as Livingstone points out (2002), children become experts in subverting adult attempts to constrain their use of the internet.The Educaunet project is based on recognition, spelled out by Buckingham, that "the attempt to protect children by restricting their access to media is doomed to fail.On the contrary, we now need to pay much closer attention to how we prepare children to deal with these experiences; and in doing so, we need to stop defining them simply in terms of what they lack" (Buckingham 2000, p. 16).
The other extreme we want to posit, implied by Buckingham's argument, is exploration.We see this as an opposite of prohibition: where prohibition closes down possibilities, exploration opens them up; where prohibition leads to risk-avoidance, exploration involves risk-taking.As a different form of social action, we can imagine that this can be, at one extreme, wildly reckless, and at the other, informed and considered.Naturally it is the latter tendency that media education tries to encourage in students; but there are many variations on this theme, and finding a balance between reckless forms of learning and carefully considered procedures is not always easy.
In this article, we will develop a model of discourses related to internet risk based on the axis of discursive structure (folkloric-rationalistic) and the axis of social motivation (prohibitive-exploratory).This model, shown in Figure 1, produces four types, with the possibility of many complex inter-positions.A good deal of our analysis of the recorded and transcribed talk of the students and teachers focuses on modality, which is to say, the mechanisms by which the discourse makes particular kinds of truth claims.In systemic-functional linguistics (Halliday 1985) and in social semiotic theory (Hodge and Kress 1988;Kress and van Leeuwen 1996) this system is part of the metafunction of communication concerned with the establishment of interpersonal relations.Lemke (1998) extends the traditional social semiotic model by proposing seven types of semantic evaluation which propositions can contain -a system through which speakers express an evaluative attitude to their own proposition.In this way, we can distinguish where our interviewees code their propositions about the internet and its various risks as truthful or doubtful, desirable or reprehensible, verifiable or not, and so on.

4.
Folkloric discourses Starting with the left side of Figure 1, these are the discourses which characteristically inflate internet risk and render it spectacular.Like folkloric accounts of the world in general, they are attempts to explain reality which, in our culture, are usually perceived in opposition to scientific/positivistic accounts of the world.Also, as Imler et al (2003) point out, "one of the more widely accepted traits of any urban folklore is its reflection of what are considered to be social or individual problems" (http://www.units.muohio.edu/psybersite/cyberspace/folklore/).We will identify two sub-types, proceeding from related but quite distinct social motivations.The first is the prohibitive-folkloric type, characteristic of some forms of parental protection.The second is the exploratory-folkloric type, characteristic of peer cultures in which talk about the internet resembles in many ways other discourses of horror and risk, where real social fears are exercised and arguably partially dispelled and controlled through thrilling and pleasurable oral narratives.

Prohibitive-folkloric
This discourse is motivated by adult-generated anxiety about dangers which, while they may really exist, are here often understood vaguely, and only loosely related to experience: paedophiles, pornography, the "dark side of the net".These are irrational anxieties not because real dangers do not exist; but because these perceptions of them emerge from ignorance, and from a distorted assessment of the actual risk.Like anxieties about paedophilia in general, with which they are closely implicated, they do not assess the risk to children rationally by comparison with, for instance, the statistical risks of sexual abuse in the family, or of the physical danger of road accidents.We do not have direct examples of this discursive type, as our project did not include interviews with parents; so this part of the model is to some extent speculative.However, in a small number of cases we can hear this discourse clearly at second hand, echoed in the words of the children as they rehearse parental injunctions.
This form of prohibitive discourse will depend on the exotic figures of urban and media mythology, employing them as a warning device.Essentially, this prohibition, usually the prohibition of the home, is "Don't do that or the ‚bogeyman' will get you" (see below); or, in this case, "don't go into chat rooms or the paedophile will get you".Its discourse markers will be: -Strong demand modality -imperatives, typically linked with consequences -Brief, condensed structures representing the technology, disguising uncertainty or confusion -Dense codings of risk in nouns that carry considerable emotive and mythic freight (eg bogeyman, virus, stranger), and in representations of their attributes (appearance, clothing, instruments) and actions.The clearest example of this was in a Year 6 (age 10 -11) interview, where a group of girls echoed the injunctions of their parents, discussing their mothers' anxieties about paedophiles in the neighbourhood.When asked where they had heard about paedophiles, one of the girls mentioned receiving notices about where paedophiles had "attacked", and the following extract exemplifies parental anxieties: Daniella: my mum's been warning me about [paedophiles] saying "oh don't dress too old" ...

All: ya
Claire: don't dress up because they'll go after you

Daniella: it attracts them
Claire: ya my mum was talking to me this morning

Becky: ya like a magnet
Daniella: it's like a flea to a cat The girls' statements here, as well as being about their peer relations, also reflect anxiety from the home about childhood generally, and more particularly, about dangers to girls.This discourse is also reported in a study by Buckingham and Bragg (2004) in which girls tell about their mothers' warnings about dangers in their neighbourhood and subsequent restrictions placed on girls' mobility.These girls' discussions display echoes of the prohibitive variety of the folkloric discourse, in which warnings by parents are characterised by strong imperatives ("don't dress up"), simple consequences ("because they'll go after you"), and colourful, dramatic simile ("like a magnet"; "like a flea to a cat").

Exploratory-folkloric
Looking at Figure 1, the bottom left corner (in some ways opposite to prohibitive-folkloric) contains the second version of this discourse, which emerges from the children themselves.This peer discourse is one of horrified fascination, based on urban myth and the community folklore, often dependent on tabloid media coverage, which feeds it.For children, the figure of the paedophile assumes the shadowy shape of the bogeyman, a conventionally exotic and spectacular figure which, at its most folkloric, features as star attraction in the genre of the campfire story.The campfire story, particularly in the U.S., is the tradition of gathering round the campfire to tell scary stories, with narrators competing to produce the most pleasurable thrill or fear.Especially in American and British oral tradition, the bogeyman is the imaginary expression of children's fears, in particular their fear of the dark.In the popular culture of the mass media, one of the best-known representations of this figure is Freddie Krueger of the Nightmare on Elm Street franchise.As has often been pointed out (e.g., Buckingham 1996), the curious thing about Freddie, a multiple abuser (Jenkins 2001) and child-murderer who lives on in children's nightmares, is that his young audiences regard him with a mix of terror and affection.This horrifying icon is something which repels, warns, shocks; but which also fascinates, entertains and, ultimately, reassures.Explanations for this vary, though for the purposes of the present study, we may speculate that giving a vague fear specific (fictional) shape enables children to deal with it, and at the same time allows them to indulge their attraction to risk and danger in a relatively safe fictional form.
There is an important difference between the horror monster of popular fiction and the bogeyman of the urban myth or campfire story, however.In the first case, the child is able easily to make modality judgements which distinguish between fiction and reality (Nightingale et al 2000).In other words, the child is able to read the ways in which these stories make particular claims to be true, and assess these claims in the light of her experience, both of real life, and of these genres of fiction.In the second case, the peculiar attraction of urban myths is precisely their claim to be true -they confuse the modality of the fairytale and the news bulletin.Rather than making claims to truth status which are consistent, and which either say: this is a fantasy, or this is a documentary report, they mix together these systems so that those who listen to them and tell them are uncertain of their truth status.
However, one or two of the older pupils were aware of the process of transmission of this kind of knowledge.Pupils mentioned that they had learned about risk through their friends and relatives, as this girl (age 11) relates: This account of how knowledge or urban myths spread seems an accurate description of the way knowledge of any taboo topic is spread, especially sexual knowledge.The exploratory-folkloric discourse is characterised by: www.medienpaed.com> 12.9.2017-Narrative structures -Displaced agency, or attribution of the relevant action or event to another (it didn't happen to me, but to a friend/cousin/classmate) -Direct speech attributed to characters in the story, so that the apparent representation of the "real words" spoken at the time makes the story seem more immediate, dramatic and convincing -The truth-claim structure of the urban myth -an insistence on details of time and place, and other modality markers which raise the apparent truthfulness of the story -Exotic or exaggerated events or characters There were various versions of this type, all with their own specific characteristics.They ranged from the obviously untrue though vigorously asserted horror tale, such as that of the 9-year-old boy, who claimed to have shot a paedophile "in the nuts", to the very convincing talk of a 14-year-old boy, who showed great knowledge of the internet, with confident tales of his use of phoney ID cards to get into pubs, and of friends who had racked up hundreds of pounds on his parents' phone bill visiting porn sites.These claims, carefully elaborated with details of websites and amounts of cash, turned out to be wild exaggerations or complete fiction, on closer inspection.
An interesting feature of this discourse in relation to paedophilia was when it relayed false stories of real people, so that the truth-status became very confusing for the children constructing the stories.The clearest example was the children age 9-10 who were convinced that the Soham murder victims, Holly Wells and Jessica Chapman, were entrapped by using a chatroom: "That's what happened to Holly and Jessica.They went on the chatroom and asked to meet up with someone.""Yeah, they went on the chat rooms and asked to meet up with someone, and said they'd be wearing their Man United stuff." This talk, backed up by the other children in the group agreeing and adding details, has some of the narrative characteristics, descriptive detail and assertive modality of the urban myth.
(NB: this conversation referred to the murder of two children in England in 2003 by a school caretaker.The media reported at one point that the girls might have met their murderer through the internet; though this was later shown to be false.The images of the girls in the shirts of their favourite football team, Manchester United, were widely reproduced in the print media; though at no time was it suggested that these images had been related to chat rooms).
The popular press and television news as a source of the raw material of urban myth was a pervasive theme of the interviews.In all the interviews the pupils reported hearing news reports, advertisements or storylines related to internet danger on television or in newspapers.Kidnapping frequently arose as part of the discussions, and other heavily reported child murders which had oc-curred the previous year were referred to as well as the Soham murders (see above).The pupils also said that viruses were frequently reported in the news, and one girl mentioned an episode of a television sit-com (Friends) in which a virus transmitted via email destroyed the hard drive of a computer.When asked about the greatest risk connected with their use of the internet, almost all the pupils said either paedophiles or viruses.
Several groups in the secondary school discussed the frequency of incidents of risk on the internet, particularly in connection with paedophilia and viruses.Most groups felt that paedophilia was commonly reported in the news, and therefore a frequent risk to children.As one boy (age 13) remarked, "just from listening to the news, the most occurring thing that comes up is usually to do with chat rooms and children getting kidnapped or taken away or they've gone walkabout or something".Only one group (age 13) thought that the seriousness of paedophilia made it more likely to be represented in the news, however that was not an indication of frequency.These girls (age 11) reflect the anxiety around frequency of paedophilia: This again is the folkloric discourse, characterised by exaggeration, wild claims, and insistently high modality.In this case, it is also a discourse of anxiety, similar to discourse in the quadrant above it on Figure 1 (prohibitive-folkloric), which exaggerates dangers related to children's use of the internet and also exerts control and authority over children's lives.Children and young people, therefore, are positioned as vulnerable and innocent, and particularly passive in their reception of internet risk.In summary, the modality (truth claim) of exploratory-folkloric discourse is constructed from a complex of specific detail about place and time, compensating for the unlikely nature of the event.It is, most importantly, at odds with young people's actual experience of the internet -they either have little experience, or their actual experience is relatively confident and not anxious or worried.And finally, we might speculate that it is all the more active when given permission, partly as a result of the lack of opportunity to talk about it at school.

5.
Rationalistic Discourses Moving now to the right side of Figure 1, we will identify two branches of this discourse, again motivated by prohibition on the one hand and by exploration on the other.Rationalistic discourse is of course typical of Western educational institutions, rooted in the Enlightment ideal of Reason, and dedicated to the rooting-out of superstition, ignorance and myth.These are all value-laden terms, however; ideals which seem to have nothing to do with reason, such as fantasy, the imagination, pleasure and play have also had an important part in shaping modern education.The proponents of these ideals at their most extreme have directly opposed rationalistic values, most obviously in the tradition of Romantic thought, and especially, in the English Romantic tradition, in the work of William Blake.Blake was a poet, artist and publisher (1757-1827) for whom Reason was a tyrannical false god, and institutional education one of his most pernicious products.Many of his poems and illustrations personify Reason as the god Urizen ("Your Reason"), and he is shown blighting the energy of childish visionary innocence and of human sexuality.
Clearly, there is a balance to be struck here; and, as we suggested earlier, media education has an important role to play in offering clear conceptual structures for the analysis of culture while maintaining and fostering the pleasures of popular cultural practices and the often irrational ways in which they represent the world.In the case of internet discourses, however, we want to make a specific distinction in respect of rationalistic discourses.Where these are of value in education generally, they are expansive, explanatory discourses, tolerant of different points of view, concerned to explore difficult and ambiguous territory.However, in the case of socially taboo subjects, education has often had as much difficulty in promoting open and rational debate as any other sector of society.In particular, sexual practices such as masturbation, obviously crucially important to young people at the beginning of their sexually active lives, are notably absent from school talk about sex.In respect of the internet, issues such as paedophilia, pornography and child abuse may be similarly characterised by closed, reductive discourses in schools, even by silence; allowing the folkloric discourses on the left side of Figure 1 to thrive.This is in part an assumption (though informed by our experience as practising teachers of many years' standing); and it does not form part of our empirical enquiry.The rational discourse we have most evidence of is that of the child, partly attributable to their own resources, and the cultural capital they inherit from educational and family experience; and partly, we will argue, to the Educaunet programme.
Prohibitive-rationalistic This is the pseudo-rationalistic discourse of brisk, business-like warnings, characterised by brief, reductive sets of "do's and don'ts".While these may, in themselves, be unexceptionable, their effect, unlike the expansive, spectacular form of the "bogeyman" warning, is to close down discussion.In effect, it is a form of repressive discourse, which is unable to speak of the dangers it fears.Ironically, as in Foucault's "repressive hypothesis" about sexual discourses in the 19th century (1978), it only encourages eloquent speculation to erupt in other discourses of risk; namely, again, the folkloric urban myths of our first category (the left side of Figure 1).Important discursive markers of this discourse are: -Imperative verbs -Reductive formats (lists, bullet points) -Condensed forms which avoid detail or description This kind of closed discourse needs to be seen in the context of UK schools and the developing cultures surrounding their uses of the internet.The schools where the data was collected, like most schools in Britain, require pupils to use the internet.The National Curriculum for England and Wales requires schools to teach internet related skills such as searching, checking for accuracy and relevance and using email (at Key Stage 2 -ages 7-11), and discussing the impact of ICT on society (at Key Stages 3 and 4 -ages 11-16), and pupils in the study frequently mentioned the use of the internet for research related to school work.
The secondary school where the study took place was particularly enthusiastic about pupils working online.The school holds periodic e-learning days when pupils work independently on-line (in homes, libraries, parents' offices or at the school) to complete a day's study.In spite of the possibility for using communicative aspects of the internet for shared project work, the e-learning days are about pupils working on their own.The discursive practices here frame the internet as a tool for gathering information, and position pupils as responsible and competent internet users.However, like most schools in Britain, the schools' computer systems contain strong filters which position pupils entirely differently.The filters in the primary school are controlled by the local education authority and are completely inflexible.Here the filters make research for terms such as "cock fighting" impossible.The secondary school filters are controlled within the school and are therefore more flexible, so it is possible for a teacher to allow access to a particular site or to remove filters for a set period of time.
The pupils frequently mentioned their frustration with filters which blocked access to their email and instant messaging facilities.These two pupils age 14-15 describe the filtering software at the school: Ben: ... sometimes some sites are pretty safe to go on but because of the wording or something that may just pop up in the description of the site, it's filtered in school.www.medienpaed.com> 12.9.2017Katie: Or when you have free lessons you're allowed to go on any site you want but like the music sites are all filtered then.
Ben: Ya exactly.And you've got to find a new one every time that they haven't filtered because the next time they go onto the same one it's filtered again.
The message to pupils here is that actually, they are not responsible users, at least during the school day, and they need to be protected from their own uncontrollable desire to socialise as well as numerous unnamed risks on the internet.
There is a contradictory deployment of technologies and their accompanying social intentions and discourses here.Broadband internet access expresses a social intention to provide vastly expanded access for the pupils of the school, and is accompanied by a rhetoric of access associated with school internet culture: a serious commitment to work, typified by words like "research", "homework", "e-learning".Buckingham (2003) argues that this school internet culture is increasingly at odds with the digital culture of children and young people.Where it emphasises work, theirs emphasises play; where it emphasises education, theirs emphasises entertainment; where its technical focus is on information retrieval, theirs is on communication and gaming.The discursive pattern in schools in the UK is often marked by an eloquence about school-orientated uses, but where children's digital cultures are concerned, this eloquence gives way to the reductive brevity of the prohibitive discourse we have identified above.This may consist of very few interdictions -the usual rules about internet risk; a total ban on gaming; and then the discursive silence surrounding the use of filters.There is no real discussion here of, for instance, paedophiles or pornography -but the silence accompanying the filters implies these nameless dangers.Foucault's repressive hypothesis is an apt description, then, of the pupils' active discourse about these issues, fed by repression, folklore, media stories and home anxieties.

Exploratory-rationalistic
In the lower right corner of Figure 1, the second rationalistic discourse consists of either home or school discourse which attempts to explore and explain internet risk.It is the discursive form of media education, or of informed parental talk with children about what's risky, what isn't, levels of risk and how they might be handled.It is also, of course, the discourse of this article, which like any good discourse theory, acknowledges its own status as discourse.However, though it may seem that this is the discourse we are advocating for the purpose of developing internet risk awareness, this is not entirely true.We would not, for instance, advocate an approach which sets out to calmly and rationally discuss risk with children as its sole strategy.Rather, we would integrate this kind of discourse with ways of exploring the colourful and spectacular fascination of the internet through imaginative work such as role-play, simulation, and practical experience of the internet itself.
Discursive markers here are: -Tentative modality, which is to say a reluctance to make assertive claims about what is or is not true (as participants struggle to understand or explain) Questions -Explicatory structures and connectors -Classificatory structures -Moves from particular to general instances -Particular instances bracketed as examples, rather than as narratives central to the discourse -Evaluative comparisons of different risks An example is one of the 15-year-old pupils discussing pornography.In sharp contrast to the student for whom this topic was a trigger for lurid stories of cash fraud, this girl said: "I wouldn't say it's dangerous.I think that's just sick.... It's horrible, not hurting." In terms of Lemke's (1998) semantic categories of evaluation, this shows markers in two categories: it expresses tentative certainty rather than the assertive certainty of the urban myth ("I wouldn't say ..."; "I think ..."); and a strong aversion under Lemke's "desirability" category ("It's horrible").At the same time, it makes a precise distinction between types of undesirability -between dangerous and sick.Altogether, it is loose where the urban myth is tight (weak modality as opposed to strong); and precise where the urban myth is vague, struggling to distinguish shades of risk rather than collapsing them together for dramatic effect.
The risks of internet advertising and online shopping are mentioned by many children at different ages: "I have learned a few more games, yeah, so like how to really, who to trust, who not to trust on the internet and I've learned a bit about advertisements" (boy, age 13)."There's also sort of adverts that come up which is not the really best things ... Once you shop from one site it won't stop sending you spam" (girl, age 11) These kinds of remarks are characterised, again, by the tentative modality of the exploratory-rationalistic discourse ("I've learned a bit ..."; "not the really best things").They also display risks which are mundane, quite unlike the spectacular risks of paedophilia and pornography.
This kind of discourse also considered a wider range of social groups and interests than the folkloric discourse, which tended to fixate on their own peer group.They mentioned opportunities the internet provides (research, games, access for disabled): "It's good for people who are disabled because they can't walk and it's hard to like, if some shops have got stairs" (boy, age 11).
In the interviews, although, as we have seen, the folkloric discourse is still strongly in evidence in relation to paedophilia and pornography in particular, it begins to give way to more rationalistic forms of open question, admission of ignorance, evaluative comparisons of different kinds and scales of risk, and an awareness of a greater variety of types of risk.This shift in student discourse, from the left to the right side of Figure 1, is the focus of our final section, which looks at key moments when the pupil discourse seems to be in transition, particularly moving from the folkloric category to the rationalistic through the exploration mode.

6.
Transitional, hybrid discourses A group of 13-year-old pupils were telling us about the danger of giving out one's email address.In connection with this danger, the conversation moved to paedophilia.Pupils saw giving out email addresses as dangerous for a number of reasons including viruses and general stranger danger.A girl with relatively little experience of using the internet said, "You don't know if someone could just be emailing you and you don't know who it could be ... because they really want to know how old and everything you are and all the information, they want your details and everything."This girl obviously had heard stranger danger rules: never give out personal details.However, the girl here sounds as if she is talking about chat rooms instead of email.As stranger danger is often discussed in reference to chat rooms, a possibility is that the girl, who has little or no experience of either chat rooms or email, is confusing the two.This talk we see as hybrid and transitional -it proceeds from the same social motivation as the folkloric discourse, and it emerges from peer cultures, not from actual internet experience; but it lacks the typical narrative structure, and is presented in a rationalistic manner, constructing a general case which might apply to any user ("You") and any offender ("they").At the same time, it echoes the pseudo-rationalistic discourse of prohibition -it constructs a reductive rationale for prohibition in a way that leads nowhere.
The conversation continues with a girl mentioning that paedophilia is a danger she has heard about.The next extract follows this statement:

Reepa:
Everyone thinks of [paedophiles] because of the news and things.There's lots of things on the news about people being, if a teenager's been kidnapped or abducted or whatever and they like always check out their email accounts and things to see if they've been talking to anyone, uh they've not told their parents about or anything like that.They always say on the news ...

Rachel:
That's what it said yesterday on the news as well.
Mark: I know it sounds strange but I have been on the internet so many times I don't actually know what a paedophile is.What is it exactly?
Interviewer (male): Can anyone answer that?

Gareth:
A paedophile is someone who like is old, older and looks at younger people, in a rude way, in a sexual way.
In this extract we see several discourses coming into play, again in a hybrid fashion.For instance, Reepa's observations show something of the fascination for the figure of the paedophile rooted in the folkloric discourse, as does the general emphasis on this theme in the interviews.However, unlike the previous statement, she begins to transform the characteristically narrative patterns of this discourse into a genuinely rationalistic discourse, which recognises the part the media play in producing a distorted emphasis on the danger of paedophiles in chat rooms.Again, the modality is rooted in the construction of a general case ("Everyone"); and also in a lexical substitution which reaches for an authoritative journalistic language ("kidnapped, or abducted").
Equally interesting is the moment when Mark admits that he doesn't know what a paedophile is.Like Reepa who confuses chat with email, children in all the interviews express confusion over the dangers on the internet.Paedophilia, kidnapping and pornography all blend together with viruses, spam and hacking.For example, this girl (age 10) was describing using her mother's email account which was full of spam.She goes on to talk about spam: "they say it's pornography which is like naked women and everything but when you press onto it it's actually a virus".The difference is in Mark's outright admission -the point about confusion in the folkloric discourse is that it disguises itself, hidden beneath the apparent certainty of invented narrative detail, or actually employed in the weaving of suitably dark and obscure images.
Mark has highlighted an important finding from our evaluation -adults don't talk about the dangers the kids have heard about, many of the dangers remain in the realm of the taboo, and therefore children's understanding is patchy, based on snippets from the media, and often misinformed.Gareth, like Reepa, is working towards a more rationalistic discourse here, continuing the tenor of this part of the conversation, which, while it may be motivated partly by the folkloric fascination of the figure of the paedophile, is also motivated by a struggle to clarify and de-mythologize.
Another theme which was typified by hybrid discourse, at times colourfully folkloric, at times moving towards emergent rational evaluation, was internet pornography.Pornography produces anxiety both in schools and in homes.Livingstone (2003) cites studies in the US, Canada and the UK which find extensive content on the web which could be upsetting for children; and evidence that significant numbers of children have accessed such material.However, she also points out that research evidence on media effects in general has never satisfactorily established causal links between content, incidents and long term effects, and this is equally true of this area of content in particular.Furthermore, she observes that research in this area is clouded by a failure to define and categorise pornographic content; and by a surprising failure in many cases to ask children directly, relying instead on parental information.
Younger children in our interviews and in the teaching sessions we observed mentioned "naughty pictures" as an internet risk, and all the interviews contained some kind of reference to pornography.Interestingly, pupils had a hard time identifying the risk connected to pornography.Younger pupils said the risk was making their mum angry and older pupils mentioned financial risk or viruses connected to pornographic emails.As mentioned earlier, one girl de-scribed viewers of pornography as "sick", and only one girl (age 15) said that the risk of pornography was "disrespect to people".These discussions reflect the various discourses identified by Buckingham and Bragg (2004) in their research on children talking about pornography, particularly the themes of propriety (reflected in descriptions of viewers as "sick") and ideology (as in the "disrespect" quote above).
The discourses used by the children in these studies do not seem to be the same kind of talk that characterises talk about paedophiles, which is surprising, as adult perceptions of internet danger often considers these two kinds of risk as closely associated.From the adult perspective, both threaten the innocence of childhood, while, contradictorily, offering opportunities for the aspects of childhood we see as uncontrolled, incapable of self-regulation, to run riot.Buckingham (2000) describes how such perceptions of childhood develop as social constructs from long held and contradictory cultural practices, in particular the tradition of Romantic literature on the one hand, which represents the child as innocent, and traditions of religious belief in Western societies on the other, both Catholic and Protestant, which regard the child as a vessel of original sin.
In our study, these adult perceptions, which in respect of pornography take such content at face value and focus their anxieties, understandably enough, on sexuality, find no exact match in the discourse of the children.Though in some instances they judge the content as sexually deviant ("sick"), they also ignore it, associate it with viruses, as seeing it as a kind of health risk, perhaps, and consider it a violation of human dignity.Far from exhibiting the kind of fascinated exaggeration of the folkloric discourse which characterised talk about paedophiles, then, they show a sober, evaluative stance towards pornography which is closer to our rationalistic discourse.The interest of media educators must be in this move from the folkloric to the rationalistic, when pupils are drawing on their knowledge and enjoying media stories, but also making sense of their surrounding discourses.

7.
Conclusion Folkloric understandings of the dangers associated with the internet are deep-rooted and persistent.They overlap with similar patterns of talk about stranger danger in general.They reflect, contradictorily, both a fascinated pleasure in these kinds of narratives, and at the same time an anxiety founded on vague and ill-informed knowledge.The worry must be that this anxiety, fostered by adult warnings and tabloid stories, is disproportionate to the actual threat, and produces excessive anxiety in children, as Livingstone points out (2002).The other concern, specific to media education, is that to simply outlaw such narrative and mythic ways of dealing with social anxieties may simply not work -the problem is how to allow them space, and at the same time offer rational ways of exploring them.It seems clear that our experiment with the Educaunet programme did not offer enough opportunity of this kind; though the interviews, which for the students were a kind of extension of the programme, perhaps began to offer such an opportunity.
The interviews reveal the value of the exploratory forms of discourse, and the need for time to allow students to play with the myths and find their way through to rational explanations.In this process, discourses and practices of prohibition are deeply unhelpful, whether in the form of the unexplained imperative or in the form of crude filter software (which was much less of a problem in our partner countries).This course occupied the students for upwards of two hours a week for six weeks; and it was clear that this only began the process for many children.This suggests that quick-fix solutions are unlikely to be effective, and that advertising campaigns (such as those produced by the Home Office in the UK) will not of themselves have the desired effect.A combined strategy is more likely to work -and an essential component, we would argue, must be a media education approach which allows time and opportunity for active exploration of the nature and level of different risks.In this respect, joint policy initiatives between (for example, in the UK) the Home Office, the Department for Education and Skills, and the media regulator OFCOM, which has a remit for media literacy, would seem a promising route.
If there are dangers of content in school use of the internet, these must, as Livingstone (2003) argues of home use, be set against both opportunities offered by the internet and against other dangers, in particular the danger of forms of educational, cultural and social exclusion caused by lack of access.In the general context of these issues, and of the slowly emerging creative uses of the internet at school and home, we must not lose our nerve and allow risk to become a disproportionate preoccupation of teachers and parents, located within a debate whose terms have become inflated and distorted.Nor must we allow school internet cultures to become, as they are in danger of becoming in some schools, dreary, prohibitive, humourless environments, from which children can only wish to escape to the colourful, playful, exciting worlds of instant messaging, chat room fantasy and online gaming, for all their possible risks.

"
People in year 5 (age 9-10) know what [paedophilia] is.You know older people tell younger people and they tell younger people.Like my cousin in year 11 (age 15-16), she told me what, you know, her friends tell her what it is and she didn't keep it in and she tells me and I tell my cousin."