70 The role of the Internet in suicide prevention

Among websites relating to suicide, approximately 40 per cent represent a suicide-preventive view. However, research shows that pro-suicide sites usually rank higher on search results pages, and are more frequently searched for. These pro-suicide sites and other forms of online pro-suicide communication have tried to spread a view of suicide as an acceptable solution to life's problems. They have also increased access to potent suicide methods, through the long lists they contain. They have added to persuasion and group pressure to fulfil suicide plans, glorifying those who have committed suicide, and given rise to a new form of suicide pact—‘net suicides’.

This chapter also discusses the scope for statutory prohibition and regulation of pro-suicide websites, and ways in which Internet-based systems are used as suicide-preventive support.

The Internet is no make-believe world. It is populated by real people from diverse social and cultural backgrounds. And what happens on the Internet has repercussions on actual human beings. Parts of the Internet can, and should, be seen as virtual social environments where people meet and exchange thoughts, feelings and experiences. Just as in physical environments, there are both positive and constructive virtual social settings and those that are clearly destructive.

In one study on how a suicide event was presented in various Internet media, and of the comments from readers to which this event gave rise, inadequate reporting of the event proved to elicit comments that simplified suicide problems to a high degree (Sisask et al. 2005). As the authors state, we should not underestimate the role played by the Internet in shaping attitudes towards suicide among the general public.

A search for the keyword ‘suicide’ in 1997, using the Inktomi/Hotbot search engine, generated 130,000 ‘hits’ (Baume et al. 1997). In 1999, 50,000–100,000 results were obtained with what were then the most popular search engines (Thompson 1999). A comparison with the many millions of hits obtained from a search for the same word in 2008 may be interesting. A search on Google, for example, now yields more than 65 million hits for ‘suicide’. Does this tell us that suicide must, as many people think, be regarded as a growing problem on the Internet? All it shows, in fact, is that a vast number of web pages contain the word ‘suicide’. Unless we investigate the content of these pages we cannot know what they are about.

One quantitative study of how the subject of suicide is presented on the Web was based on hits for the words självmord (Swedish for ‘suicide’) and ‘suicide’ generated by Google (Westerlund 2008). Overall, this study shows that web pages of institutional origin (public agencies, other organizations and companies) on the subject predominate on the Internet (84 per cent) and that the content provided by these institutions concerns research and prevention, and may thus be termed ‘suicide-preventive’. But besides these institutional pages, whose manner of communication is largely reminiscent of the more traditional mass media, there are private pages (16 per cent) characterized more by multiple communication, personal confessions and narratives, and an alternative, pro-suicide stance.

Notwithstanding the predominance of the institutional websites, representing a suicide-preventive attitude, the Internet has thus provided a previously non-existent opportunity to publish material and discuss, confess and seek contact on a subject that has always been strongly taboo and therefore ‘belonged’ to only a few voices in public discourse. This opportunity has resulted in both constructive and strongly destructive contributions. Summing up, the study indicates two parallel trends in how the subject of suicide is represented on the Net:

It extends and supplements the presentation of suicide and suicide prevention in traditional mass media.

It provides virtual social environments (both constructive and destructive) where new forms of discourse and formerly unheard voices—with no possible place in public and mass-media discourse previously—put forward alternative explanatory models on the subject of suicide.

Both the supplementary suicide-related material now found on the Web and the scope for new forms of communication about suicide can presumably help to change the way in which suicide is perceived and portrayed. Accordingly, by extension, they can also affect the views and notions about suicide that prevail in our society and culture (Westerlund 2008).

In another study based on material collected using search engines, Biddle et al. (2008) analysed the results obtained when they used 12 simple keywords and phrases (such as ‘suicide’, ‘suicide methods’ and ‘how to kill yourself’) to search for information about suicide, using the four search engines Google, Yahoo!, MSN and Ask. Since studies of online behaviour have shown that users seldom look beyond the first ten hits on the first results page (iProspect 2006, Hansen et al. 2003, Eysenbach and Kohler 2002), the authors confined themselves to these ten results for each search phrase on each search engine. This yielded a total of 480 results to analyse. Summarizing these findings, nearly 30 per cent of the results may be said to consist of web pages whose content was dominated by information about suicide methods, and whose messages about suicidal acts ranged from incitement or clear encouragement to non-rejection.

In contrast to the above-mentioned web pages, 121 (25 per cent) of the hits were pages focusing on suicide prevention and pages that were clearly opposed to suicidal acts (Biddle et al. 2008). A further 70 results (15 per cent) were pages with academic or policy-oriented content. Overall, these two categories, totalling 40 per cent of the results, may be interpreted as representing a suicide-preventive view to some degree.

Although web pages with a more suicide-preventive attitude thus predominated among the results—a finding that tallies with that of Westerlund (2008)—Biddle et al. (2008) nevertheless point out that pro-suicide web pages, pages containing factual information about suicide methods and chat rooms with general discussions about suicide were the results where the highest number of hits were usually found on results pages. They also found that the three websites topping the search results were clearly pro-suicide, with the well-known (notorious) alt.suicide.holiday (ASH) ranked first. The website ranked fourth was Wikipedia, which also has a long list of suicide methods. The study showed, furthermore, that these four top-ranking websites not only describe various suicide methods but also evaluate them.

The authors found, moreover, that the four search engines differed (Biddle et al. 2008). Google and Yahoo! generated the largest number of pro-suicide results while MSN generated most suicide-preventive and academic results. This indicates that it is possible to influence the presentation of search results. It should be fruitful for Internet companies providing search engines, through optimization strategies, to attempt to maximize the probability of suicidal individuals finding websites that offer help and support instead of the risk of being ushered onto websites that are dangerous to life.

In an American study designed similarly to those discussed above, the starting point was to investigate the websites that suicidal people may encounter if they use online search engines (Recupero et al. 2008). The search terms used were ‘suicide’, ‘how to commit suicide’, ‘suicide methods’ and ‘how to kill yourself’. The first 30 search results for each phrase, using Google, Yahoo!, Ask, Lycos and Dogpile, were collected and analysed. After elimination of the websites that had already been found once (‘repeat listings’), 373 unique URLs remained. These were classified as either ‘pro-suicide’, ‘anti-suicide’, ‘suicide-neutral,’ ‘not a suicide site’ or ‘error’. As in the studies by Biddles et al. (2008) and Westerlund (2008), the findings showed that ‘anti-suicide’ websites (29.2 per cent) were more frequent in the search results than websites categorized as ‘pro-suicide’ (11 per cent). However, the authors point out that although the pro-suicide websites are fewer, in quantitative terms, they are highly accessible, as are the websites containing very potent suicide methods and describing them in detail. The study also shows that much of the pro-suicide content and ‘how-to’ suicide information in the search results were connected with the alt.suicide.holiday or Church of Euthanasias sites alone. The authors emphasize the importance of psychiatric staff asking patients about their use of the Internet. Depressed and/or suicidal patients who use the Internet are particularly in the risk zone. Staff can also help patients to find websites with preventive content and supportive functions, so that the Internet use can exert a therapeutic effect instead of constituting a threat.

In discussing the media and suicide, many researchers stress two facts: that suicide is the most frequent cause of death among teenagers in most countries (Wasserman et al. 2005) and that the suicide-attempt rate is high in adolescence (Becker and Schmidt 2004). Several studies have also shown that youthful individuals are more susceptible to, and influenced by, media presentations of suicide events than adults (Phillips and Carstensen 1986, Schmidtke and Schaller 2000). The findings of two studies that specifically explored the relationship between high-frequency Internet use (‘Internet addiction’), depression and suicidal ideation among adolescents showed that a significant correlation among these variables exists (Ryo et al. 2004, Kim et al. 2006). The above factors, combined with the emergence of the Internet as a key medium for young people, have aroused many researchers' fears.

A study by Hagihara et al. (2007) set out to investigate this very question—whether there is any association between Internet use and the incidence of actual suicides. The authors are of the view that although Internet-based communication is extremely widespread today, knowledge of how it affects suicidal behaviour is virtually non-existent. The article shows that Internet use in Japanese households rose from 0 per cent in December 1992 to 88.1 per cent in March 2005, and that in 2005 there were more than 17,000 Japanese websites containing information about suicide and suicide methods. There is also the fact that by international standards Japan has had very high suicide rates, especially since the economic recession in 1998. Parallel to the analysis of a possible association between Internet use and the incidence of suicide, associations between accounts of suicide in the daily press and actual suicides, on the one hand, and between unemployment and suicide on the other were also investigated.

Suicide figures for Japanese men and women on a monthly basis were compiled for the period 1987– 2005 and subsequently analysed in a linear model. In this model, the number of articles about suicide in the daily press, households' rate of Internet use and the number of unemployed people in a particular month were the independent variables. Suicide figures in the subsequent month were the dependent variable. The results showed a significant association between articles about suicide in the daily press and the number of suicides among men and women alike (Hagihara et al. 2007). As for the influence of Internet use and unemployment on suicide, there was a significant association for men, but not for women. The former, in the authors' opinion, can presumably be explained by the fact that men have greater access to the Internet and use it more than women. Regarding unemployment, the authors found that previous studies, too, had shown that this factor was correlated with male, but not female, suicide rates. It should be emphasized that the results show that press articles about suicide, on the one hand, and unemployment on the other are considerably stronger indicators of the incidence of suicide than Internet use, and—as we have seen—particularly among men.

There are a number of methodological weaknesses in the investigation, as the authors themselves admit. For example, the analytical model used does not take into consideration other known risk factors that affect suicide rates (such as increased misuse of alcohol and drugs, domestic violence and poor access to mental-health care). But perhaps the gravest objection that may be raised is that the study leaves out the account of how the Internet has been used; it considers only the fact of its use. Thus, we do not know if the users in the study came across any online material about suicide during the time when they used the Internet, let alone whether they were influenced by such material. Globally, moreover, during the period in which Internet use has become widespread all over the world, suicide rates in many countries have declined or, at least, not risen pari passu with a rise in Internet use (Galtung and Hoff 2007). Hagihara et al. (2007) also explored whether suicide rates in Japan had risen more rapidly after the introduction of the Internet than before, but found no differences. Summing up, more studies are needed, with more precise measuring methods, to allow us to see whether Internet-based communication affects the incidence of suicide and, if so, how.

Many researchers (such as Baume et al. 1997, Thompson 1999, and Biddle et al. 2008) have reflected on the existence of so many pro-suicide websites. They have pointed out that these sites recommend suicide as a solution to life's problems, and contain detailed descriptions of methods yielding the maximum effect, and also suicide notes, death certificates and pictures of people who committed suicide. People who try to discourage others from putting their suicide plans into effect are ‘evicted’ from websites of this type. Instead, suicide is advocated as a solution to individuals' problems in life, and they are advised not to seek psychiatric or other help since this is described as worse than useless. One of these pro-suicide websites, ‘The Church of Euthanasia’ (‘Save the Planet: Kill Yourself’), has suicide as the first of its four ‘pillars’ or principles. It expresses the view that suicide is something intrinsically beneficial that serves the ultimate goal: saving the planet by ‘a massive voluntary population reduction’ (http://churchofeuthanasia.org/coefaq.html). The website has had the intention of launching a pro-suicide hotline to support people in their suicidal process and give them advice about methods and the like (Thompson 1999).

The pro-suicide Internet source that is presumably best known, alt.suicide.holiday or ASH, started as a Usenet newsgroup online at the end of the 1980s. Although the initial idea was to create a forum for discussing why the number of suicides increased during holiday periods, discussions rapidly came to be about the right to take one's own life and how this could ‘best’ be done. Nearly 20 years after its inception, ASH is still the central source of practical information about suicide, and channel of suicide communication, on the Internet. Information about methods collected by means of the many contributions and discussions has spread to numerous websites. ASH has emerged as a discussion forum about suicide as a whole, with message boards for those who are considering suicide or wish to leave suicide notes. Thompson (1999) noted that more than 900 such messages were being left monthly in 1998–99, mainly by people who were contemplating suicide or from those who had previously attempted suicide. It should also be mentioned that active discussion participants in the ASH newsgroup in 2000 opened up the possibility of chatting in real time (Internet Relay Chat, IRC) about suicide on the website alt.suicide.bus.stop (ASBS), which is close to ASH in terms of content and values.

Today, there are a large number of pro-suicide websites in several different languages on the Internet, and they often rank high on the search engines' results pages (Biddle et al. 2008). Many of these sites show marked similarities and also often link to one another, giving the impression of a loosely connected virtual network of pro-suicide websites whose content is shared by visitors and producers alike. The similarities among these websites consist not only in the common subject; there is also a clearly shared subcultural and countercultural standpoint, opposed to the suicide-preventive attitude and, in general, to the basic values and morality of the community at large. This is clearly reflected in both the content and the nature of these sites.

Examples of how subcultural attitudes can be constituted, in general terms, are formulated by Dick Hebdige:

Style in subculture is, then, pregnant with significance. Its transformations go ‘against nature’, interrupting the process of ‘normalization’. As such, they are gestures, movements towards a speech which offends the ‘silent majority’, which challenges the principle of unity and cohesion, which contradicts the myth of consensus.

(Subculture: The Meaning of Style, 1979: 18).

The countercultural aspect is found in explicit and implicit forms in statements about suicide that are clearly diametrically opposed to the explanatory models of society as a whole. Becker and Schmidt (2004) have expressed this aspect as the existence, underlying the pro-suicide websites, of a strong ‘anti-psychiatric’ view. This concealed view manifests itself through the provision of information about how to take one's own life most effectively, and also through message to the effect that suicide is something the individual alone is responsible for.

The pro-suicide websites thus express an attitude that sees society and its institutions alone as a threat to the individual's self-evident rights. It may also be deduced that individuals have few or no obligations towards others and that we are all, according to this view, accountable solely to ourselves. The outcome is a low degree of solidarity. All forms of legislation, regulation, taking into custody or other institutional intervention in people's lives are considered to be evil (Westerlund 2008).

From a psychological perspective, producing the pro-suicide message can perform at least two functions (partially working together):

The above-mentioned considerations may thus partially explain why pro-suicide websites are produced and published on the Internet. But what are their possible repercussions on visitors to these websites? What specific risks can the pro-suicide websites pose, especially for people who are already in the risk zone for suicidal acts?

A number of case studies conducted to date show that individuals have committed suicide or carried out serious suicide attempts after obtaining information about suicide methods online (Haut and Morrison 1998, Adekola et al. 1999, Becker et al. 2004a). Becker and Schmidt (2004) refer to two cases in which teenagers went online to discuss their suicide plans and find effective methods of committing suicide. These authors' writings show that they are chiefly shocked by the number of websites describing in detail, for example, the preparations and doses necessary for attaining a lethal result. Similar experience is also described by D'Hulster and Van Heeringen (2006) in an article about two serious suicide attempts that were prepared and carried out with the assistance of information obtained from the web. In the authors' view, online access to pro-suicide information can lower the threshold and exacerbate the risk of suicidal behaviour in individuals in the risk zone.

Baume et al. (1997) followed up three cases in which people had left messages on ASH message boards, saying that they were thinking of taking their own lives. The researchers found that these people left a series of messages that appeared to be more in the nature of diary entries of some kind about their suicidal thoughts than suicide notes as such. The messages showed how the individuals were proceeding in their suicidal process, first asking others in the discussion group which methods were best and how they should go about getting hold of the requisites. Two of the three cases, as far as the researchers could judge, resulted in completed suicides after the people concerned had obtained highly detailed information about how to perform the acts (for example, exactly how to angle a pistol in one's mouth to obtain the maximum effect) and also been encouraged by other discussion participants to fulfil their suicide plans. In one of these cases, the discussion continued afterwards with participants arguing about whether the incitement to commit suicide that had been given was appropriate. In the third case, the person did not die from the suicide attempt (carbon monoxide poisoning) but was admitted to a psychiatric clinic. After this, he urged others in the discussion forum to avoid the method he had used, but did not comment on whether he regarded himself as being helped by the psychiatric treatment he had undergone.

Summing up, Baume et al. (1997) pinpointed the ambivalence that many of the participants initially show in their messages in the ASH discussion forum, but also how their decision to take their own lives is strengthened under the influence of other forum members' encouragement. Ambivalence is a highly striking and prominent factor in most individuals' suicide processes (Wasserman 2001). From a therapeutic point of view, the ambivalence factor in a suicidal person is very important since it can exert a protective effect until support and treatment can be provided. In contrast, however, ambivalent and vulnerable individuals who discussed their suicidal thoughts on ASH were subjected to such massive persuasion by other discussion participants that they were unable to back out or seek help for their problems. In one of the cases, this is extremely clear: the person in question wrote ‘I'm gonna do it any day now. Really. I promise!’, showing that he felt obliged to proceed with the plans he had begun discussing and obtained reactions to from other participants (Baume et al. 1997). These plans and thoughts might possibly have been reviewed, and revised, if the circumstances had been different. The danger of websites of this type is thus what is more or less a chorus of voices in unison, inciting and urging individuals to commit suicidal acts, while voices with other solutions to life's problems are largely excluded.

Becker and Schmidt (2004), too, hold the view that since no opinions other than pro-suicide ones are tolerated at these websites their effect is to easily tip the ambivalent stance that most individuals with suicidal tendencies have regarding whether to go on living or end it all. Teenagers and other (young) adults in the risk zone are at risk of losing their doubts about and fear of suicidal acts when they are exposed to this one-sided encouragement to proceed with their suicide plans. Clear risk factors on these websites are the group pressure to commit suicide that arises and the scope offered to agree with others to commit suicide jointly. Moreover, ‘chatters’ or discussion participants who have previously committed suicide are glorified. This may make them attractive (ideal) models for other visitors to imitate. In the view of Becker and Schmidt (2004), the Internet thereby has a more powerful ‘Werther effect’ than print media, but more research is required since to date there have only been individual case studies concerning the Internet and imitation effects. Baume et al. (1997), too, argue that the Internet has a greater potential than other mass media to influence people in the direction of suicidal acts, mainly because of its interactive nature. Winkel et al. (2003) think that social learning theory can, with advantage, be used to explain why and how individuals are influenced by communicating in various online suicide forums.

In October 2004, nine people in Japan took their own lives in what was evidently a new type of suicide pact drawn up on the Internet. Suicides of this kind have since come to be termed ‘net suicides’ (Rajagopol 2004; Naito 2007). These incidents, which attracted a great deal of attention, were followed within two months by a further 13 deaths in four separate suicide pacts. The method used in all these suicides was carbon monoxide poisoning, using small barbecue grills to burn charcoal in small, airtight spaces. As many as 60 people a year are thought to have been dying in ‘net suicides’ in Japan alone, and the number is continuing to rise (Naito 2007). This phenomenon does not, however, appear to be confined to Japan. Suicide pacts entered into online have also received attention in other countries, such as South Korea, Norway and the United Kingdom (Lee 2003; Galtung and Hoff 2007; Naito 2007). Rajagopol (2004) emphasizes the difference in kind between the phenomena of suicide pacts and suicide clusters, in that the latter is characterized by a number of suicides that is limited in time and space, and later suicides have been influenced by the earlier ones. Nor, in a suicide cluster, do the individuals involved need to have been in contact with one another. The definition of a suicide pact, on the other hand, is a joint decision by two or more individuals that they will both or all commit suicide together, in a given place and at a given time. Suicide pacts used almost always to be entered into by individuals who knew one another well and had been in a close relationship—often married or cohabiting couples. In contrast, the web-related suicide pacts that have come to light so far, such as those described above, are characterized by the fact that, for example, individuals previously unknown to one another have ‘met’ and communicated through Internet websites and then resolved to take their own lives simultaneously. If this phenomenon becomes any more common, Rajagopol (2004) foresees a change in the epidemiological basis of suicide pacts.

Lee et al. (2005) are of the view that the Internet, besides ‘facilitating’ the drawing-up of suicide pacts, has also contributed to the spread of new suicide methods among countries and continents. In Hong Kong, for example, carbon monoxide poisoning using charcoal grills was the method chosen in 20 of the 22 suicide pacts that were publicized in 2002 and 2003. Why these charcoal grills are used to such a large extent in suicide pacts, in particular, is that this method—unlike hanging and jumping, for example—can be easily shared, and because carbon monoxide poisoning has also often been described as a relatively painless way to die. This may also result in more passive participants in a suicide pact being induced to fulfil the plan. Lee et al. (2005) conclude in their article that the use of charcoal grills and ‘cyber suicide pacts’ may be seen as examples of how globalization and the emergence of new communications technology also create new challenges to global health.

Case studies have thus shown how young, vulnerable individuals, in particular, are encouraged to commit suicidal acts and how detailed information about suicide methods is provided on the pro-suicide websites. These studies and the publicity given to the phenomenon of cyber suicide pacts have prompted a moral discussion about whether this kind of material and communication should be allowed at all in the public sphere. It may appear difficult to ban this communication, since suicide as such is not a criminal offence in most countries. But some writers are of the opinion that since many countries prohibit assistance in suicide, it should be possible to take legal action against these pro-suicide websites, or at least strongly question the ethical justifiability of Internet service providers (ISPs) allowing pro-suicide websites and pro-suicide communication on their web servers (Thompson 1999; Dobson 1999).

Where legislation is concerned, as mentioned above, many countries have laws that prohibit assisted suicide. However, there is an inherent problem in this issue: such legislation criminalizes assistance in an act that is not itself outlawed. Apparently what seems to be required, for the legislation to apply, is an active, quasi-physical form of assistance in a suicidal act. Were this kind of ban to be applied in an online context, it would presumably be very difficult to draw a sharp line between what is voluntary and what can be proved to constitute unauthorized assistance from another party.

Nevertheless, a few countries have taken the issue of criminalization of assisted suicide a step further. In Portugal in 1995, a regulation concerning encouragement to commit suicide was adopted. Accordingly, a person ‘who, by any means, promotes or advertises products, objects or methods designed to cause death’, in such a way as to provoke or incite suicide, is liable to imprisonment for up to two years or a fine proportional to 240 days' income earned by the offender (Galtung and Hoff 2007: 111).

In January 2002 France, too, imposed a ban whereby all forms of propaganda (encouragement) of suicidal acts or dissemination of instructions about suicide methods are punishable (Galtung and Hoff 2007: 106). These countries' prohibitions could accordingly, be applied to information and communication on the Internet, but to date no such cases have been documented.

Unlike the more general bans in Portugal and France, a more specific statutory amendment, criminalizing web pages about suicide, was carried out in Australia in 2005 (and came into force in 2006). The Australian Ministry of Justice issued the following description of this amendment:

The Criminal Code Amendment (Suicide Related Material Offences) Act 2005 makes it an offence for a person to use the internet, or an-other carriage service, to counsel or incite suicide or to promote or provide instruction on a particular method of committing suicide. An Internet Service Provider that knowingly hosts a suicide site may commit an offence if it aids or abets the commission of an offence (www.ag.gov.au).

The Australian prohibition is very stringent, in that it outlaws not only production and distribution of websites with method descriptions, but also receipt and possession of this type of information. Galtung and Hoff (2007) point out that since the ban is so new, there are no documented legal precedents. On the other hand, some websites were discontinued after the law came into force, and this development is to be welcomed.

Discussions about regulation and legislation are taking place in many other countries as well. In the United Kingdom, for example, discussions have been held between healthcare authorities, the Samaritans emotional support service and ISPs on what measures can be taken to combat websites that encourage people to take their own lives (Bywaters et al. 2006). In Denmark, a public inquiry on the issue has concluded by advising against a ban on websites that encourage or incite suicide (Danish Standing Committee on the Criminal Code 2005, Nordentoft 2006). Galtung and Hoff (2007), too, express the view in a report on an investigation of the issue from a Norwegian point of view that the disadvantages of this kind of ban presumably outweigh the advantages. In the authors' view, if suicide information were prohibited it would, in practice, be very difficult to formulate and comply with the content of such a law. Mishara and Weisstub (2007) may be said to summarize much of what are seen as the problems of legislation and regulation in the area. The matter is complex since, for example, it touches on issues concerning freedom of expression and the role of the Internet as a global communication space that is not accommodated within a single jurisdiction. The authors also think that there is, at present, insufficient scientific evidence that Internet-based information and communication have a causal effect on actual suicidal behaviour. Nor have suicide rates risen in most countries in conjunction with the Internet becoming widely available to the public.

However, most people who have studied the subject think it is very important to increase awareness of the risks that visiting pro-suicide websites may pose for people who are close to suicide. To reduce these potential risks, two measures are called for to block the most dangerous sites: ISP self-regulation and filtering of websites in homes, schools and environments where there are vulnerable young people (Coombes 2008). Galtung and Hoff (2007) cite the scope for using existing frameworks to try, in this way, to minimize the risks of suicide incitement and clearly pro-suicide websites on the Internet. One approach would, for example, be to work for an amendment in the European Convention on Cybercrime. Article 9 of the Convention, which enjoins each party to adopt ‘legislative and other measures … to establish as criminal offences … producing, offering or making available, distributing or transmitting …’ child pornography, could be extended to apply to pro-suicide information and communication. In the authors' opinion, the Convention could serve as a natural framework and working for such an amendment would (though difficult) be easier than starting with completely new proposals. It should be possible to argue that online communication that strongly encourages suicide, and addresses minors, can be just as harmful as child pornography.

Comparisons have also been made between encouragement of suicidal acts and ‘grooming’ on the Internet, i.e. establishing a trusting rapport with a minor for the purpose of sexually abusing him or her (Galtung and Hoff 2007). Sexual grooming has attracted attention globally and the UK, for example, has already introduced legal provisions to deal with it. It is comparable to incitement of suicide since, in both cases, a type of mental influence and manipulation is involved. BBC Online reported (in 2006) on the father of a boy who had killed himself in an Internet-related suicide. The father wrote: ‘It is illegal to groom a child for sex, but not to kill themselves. That seems wrong. What we need is for the government to make it illegal …’

As mentioned above, there is a copious supply of suicide-preventive websites on the Internet today (Biddle et al. 2008; Westerlund 2008). Becker and Schmidt (2004) point out that most people who spend time and communicate in these virtual environments are teenagers and young adults, i.e. the group usually said to run the highest risk of imitative or ‘copycat’ suicidal behaviour. On these suicide-preventive websites, people with pressing suicidal thoughts or plans are urged to seek help, and leaving suicide notes or encouraging suicide is not permitted. In the authors' opinion, chat rooms of this kind can offer individuals a chance to anonymously discuss and share their thoughts about a subject that is taboo in our culture and our society, and this may afford mitigation and relief in thoughts and feelings for some people. The links to and telephone numbers of help organizations and clinics on these websites can also make it easier for individuals in the risk zone for suicidal acts to seek help.

A number of studies have also been carried out in which the results indicate that web-based suicide communication can give suicidal individuals support and tools for dealing with their situation. An online survey in which questionnaire responses were collected at German suicide forums showed that support and confirmation from other discussion participants were valued as highly as support from friends, and higher than support from the family (Winkel et al. 2005). Individuals also tended to support one another to a greater extent in the suicide forums where no discussions of suicide methods took place. The higher frequency of social support and confirmation in these forums was also correlated with the participants' valuation of reduced suicidality. A previous user study that took place over an 11-month period showed that the discussion participants on a website focusing on the subject of suicide gave one another important ‘grass-roots therapy’ in the form of shared experience, sympathy, acceptance and encouragement (Miller and Gergen 1998). Hostile and aggressive discussions and participants were unusual. Nevertheless, the authors did not find support for the view that the discussions resulted in any major changes in the participants. Rather, the online dialogue appeared to serve more as a form of maintenance than a means of transformation.

Although they may be insufficient, various Internet-based systems for distributing suicide-preventive information and support are currently emerging. In one article, Wang et al. (2005) describe an example of the kind of communication system in which communication takes place through a third party found in proximity to the person seeking help. The system also affords scope for information specially adapted for specific cases. Another, more evaluative study compared three different technologically mediated sources of psychological and emotional first aid: a telephone hotline, personal chat and an asynchronous online support group (Gilat and Shahar 2007). ‘Asynchronous’, in this context, means ‘not occurring at the same time’ or ‘more independently from space or time’. Email communication, on which such a group is based, is asynchronous and thus more distanced than, for example, telephoning or chatting on the web where feedback is immediate. This may explain the difference in the results obtained. Threats of suicide proved to be much more frequent among participants in the asynchronous support group than among those who sought help and support through the other two options provided. Further research is urgently needed to find out who joins asynchronous online support groups of this type, what kind of communication takes place there and how it develops. Another example is the Israeli SAHAR (a Hebrew acronym for ‘Support and Listening on the Net’) project, which offers people in severe emotional distress psychological support through a website (Barak 2007). This permits anonymous communication, both synchronous and asynchronous, with knowledgeable helpers. Roughly a third of those who contact SAHAR are thought to be suicidal, and on a number of occasions the service has succeeded in helping people who have threatened to take their own lives or been, de facto, moving towards the commission of suicidal acts.

In three component studies whose purpose was to explore how suicidal individuals write about themselves online, it was found that suicidal people are more rigid in their views than both other groups of people in mental distress and individuals characterized by mental well-being (Barak and Miron 2005). Members of the suicidal group were also considerably more self-centred in their writing, expressed greater psychological pain and were more cognitively limited than individuals in the other groups. These findings tally well with those of similar investigations of suicidal writing offline, and this should, according to the authors, have a bearing on further research, psychological assessments and understanding of suicidality. A previous study, in which suicidal individuals' online writings were compared with other help-seeking individuals' writing, shows similar results (Fekete 2002).

What, then, supports the view that suicide-preventive effects can be attained through the Internet is the fact that individuals can promptly obtain emotional support and a sense of fellowship with likeminded people, whereupon they feel less socially isolated. Evidence against the view is the fact that individuals can join the virtual community without actually needing to solve their own problems or develop ‘real’ relationships with others. There is a risk of communication resulting in individuals merely imitating one another's problems.

Some ambivalence is discernible among researchers as to whether online information and communication about suicide should be seen as an opportunity or a threat (e.g. D'Hulster and Van Heeringen 2007). Alao et al. (2006) formulate this issue by pointing out that the Internet, with its tremendous capacity for information and communication, can encourage suicidal behaviour by its copious supply of descriptions of potent suicide methods and pro-suicide websites where individuals with severe mental problems are advised not to seek help. There is tolerance and sometimes even advocacy of suicide as a solution to individuals' problems, and voices advising against such acts are, moreover, sometimes excluded. At the same time, in the authors' view, the Internet is a key resource for helping potentially suicidal individuals. For example, it may be used to identify people in the risk zone for suicidal acts and then communicate with them, thereby preventing such acts. If the Internet is used properly, it is a powerful communication tool that can serve to help and support suicidal individuals.

The number of Internet users is continuing to grow steadily. In Germany, for example, the number of users aged 14 and over doubled to more than 35 million in the years 2000–05. The web has increasingly developed into a communication platform for psychiatry and psychotherapy as well, and this platform can be used for the purposes of supportive information, communication and therapy (Pfeiffer-Gerschel et al. 2005). Hawton and Williams (2002) believe that more research is necessary to allow any certainty about whether, for example, pro-suicide websites genuinely influence individuals in the direction of suicidal behaviour, although a growing number of case studies indicate that such negative influences very much exist. Becker and colleagues (2004a, b) think that irrespective of whether the Internet is primarily a risk or an opportunity for young people who want to communicate with others about suicide, certain steps must be taken. First, it is imperative for doctors, psychotherapists and parents to be informed about the existence of these suicide websites and chat rooms. Second, issues of media use should be included in psychiatric evaluations. Third, specific guidelines on web-based communication about suicide should be drawn up as soon as possible.

Investigations of the Internet and the subject of suicide carried out to date are mostly in the nature of case studies. Although their results, as such, are extremely important, it is difficult to generalize from these case studies regarding online information and communication about suicide as a whole. The assertion that the Internet has a greater ‘Werther effect’ than other mass media, owing to its interactive nature, has therefore (up to now) been considered unproven. The misgivings expressed to date about the Internet and its possible influence on individuals', and especially adolescents', suicidal behaviour may also be seen in the light of the greater or lesser outbursts of media panic to which new media invariably seem to give rise. As McQuail (2005 p. 482) puts it, ‘Each new popular medium has given rise to a new wave of alarm about its possible effects’.

On the one hand, the Internet generates new problems: information about suicide methods is readily available, suicide pacts are being drawn up, suicidal acts are encouraged and incited, and the proportion of imitative suicides is rising. On the other hand, the Internet has made it possible to discuss a taboo subject and obtain information about what one can do to help oneself or others with suicide problems. Thus, the question of whether the Internet primarily generates new problems or offers new remedies with respect to suicide cannot yet be answered unequivocally.

Since matters relating to suicide and its underlying causes are still taboo to a high degree in our society and culture, websites on the subject have come to represent an important and controversial source of information. Suicide communication on the Internet may be said to produce new forms—or bolster previous forms—of uncertainty, risks and threats. But the scope for suicide prevention afforded by the Internet is also increasing. In human terms, the Internet has meant the partial blurring of the distinction between personal and intimate communication, on the one hand, and public communication on the other. This also has implications regarding how the subject of suicide develops and is dealt with in modern cultures and societies.