What suicide notes look like in the social media age

By Jack Flanagan on October 29th, 2013

It is natural, if morbid, to try to romanticise death. It’s something we’re all aware will happen, so we soften the blow: we make death into “eternal sleep”. It’s more than a little jarring to this comforting appreciation of the end that there are number of people who limit themselves to a Spartan 140-character limit when they compose their final testament: a Twitter suicide note.

“This really is the digital age” was the feeling of a friend when I told him I was researching a piece on suicide notes posted to social media. In a very grim way, that is true. Putting your suicide note online – for Facebook, YouTube and various internet forums are said to host this behaviour as well – is qualitatively different from the ‘traditional’ method of writing it. Primarily because: it’s live, and because of that, it’s very similar to writing your suicide note while several hundred people peer over your shoulder.

This year has been dotted with digitally announced suicides. In early January, Freddy E. – an American rapper – posted a series of tweets which revealed his intention to kill himself.

Freddy E short

The tweets are curt – each one is only a few words with a conclusive full-stop. Judging by the way the stream of thoughts begin, it looks unplanned – which suggests this is less of a suicide note and more of a stream of consciousness which ends in the act. A live feed of his despair.

In July, a New Mexican teenager by the name of Carlos Vigil killed himself after posting a suicide note screen-grab to his twitter. His family had no idea of his intentions. The note was a little longer – at 515 characters – and was more true to the typical format of a suicide letter, with a feeling that the world would be better without him. Mourners united on both online publications and, of course, twitter.



Among these and similar events which signpost a trend, a study was released earlier this month which looked at suicidal sentiments tweeted. The study – called ‘Tracking Suicide Risk Factors Through Twitter’ – looked at over 1.5 million tweets posted somewhere in the US, and identified 37, 000 users as ’at-risk’. To be clinically at-risk, the researchers looked for phrases like “I feel abused” “depressed” or “empty”, evidence of a troubled family or self-harm, as well as a few others. They then compared the number of risky tweets in a state to the same state’s actual numbers for suicides.

They found that the number of people posting on Twitter who would be at-risk of suicide was similar to the number of people committing suicide in the same state. For example, Alaska, which has the highest suicide rate by area in the US, had the relatively largest number of at-risk twitters, while Texas – the largest state – had the most at-risk, at 3000 users.

I spoke to one of the researchers on the project – Josh West – who surprised me by saying the results were not entirely unexpected. “We’ve seen before that Twitter is useful for supporting trends we’ve gotten through other means” – he’s speaking about pen-and-paper surveys, which traditionally have been the only way to get public health information. “We weren’t really surprised that these patterns held out to suicide rates, but we didn’t expect the patterns to be so strong.”

The difference between pen-and-paper compared to Twitter data is that, in the case of suicide, the former is mostly comes after the person is already dead e.g. in reports taken from hospitals. But in the case of Twitter, the reports are from people who may still be alive. If the connection between Twitter and real-life is as strong as the study found it means this data could be life-saving.

Josh and team think of themselves as “Health Technologists”, because they, like many, have seen the connections between online technologies and the medical field become stronger and stronger over the years. You can track arguably anything using social media data – problem drinking, drug abuse or social problems – and use that information to support the current efforts to curb them.

Putting the data to use

With this information in our hands, there is now the possibility to act upon information on Twitter. Firstly, we need to appreciate that Twitter can be used effectively to monitor feelings of despair in people that use the site, while appreciating its limits: the demographic of Twitter is biased towards young people (18-29, 26%). It’s not about replacing surveys – which are more detailed and less subjective – but to help us act quicker and more precisely.

Secondly, to make sure intervention is timely and precise. I asked Josh about “false positives”, or whether there was chance more than a few of these tweets were coming from people who were insincere about their depression. “The potential definitely exists: we have too many expressions, “Oh, I hate myself for doing that”, which is not really a reflection of their self-esteem. [We need] stronger validity testing on social media, do these patterns really exist? Before we can proceed.”

Is picking amongst the syntax of a tweet practical in the time frame of a person at-risk of suicide?

This is a particularly sensitive issue. When Twitter has apparently become a home for trolls and depressives alike – not a good mix, by any measure – how can we take any one person seriously? Because there certainly isn’t the resources to take them all at their word. For instance, in 2009 Demi Moore received the tweet “gbye… gonna kill myself now.”. She retweeted and asked whether or not it was serious, followed by a flooded response from Moore fans which put the tweeter, ‘sandieguy’, into police custody.

Demi Moore received an outpouring of good feeling. And the tweeter sandieguy followed up with “Thinking that it’s absolutely amazing that complete strangers love her. Bless you all and thanks, you don’t know what it means.” But the incident smacks more than a little of attention-seeking: she prompts Moore with an urgent “now” in her initial tweet, and had a track record of trying to contact Moore and husband Kutcher before. But if this was a lazy ploy to be retweeted by her choice celebrity, is it worth the risk to not intervene? And is picking amongst the syntax of a tweet practical in the time frame of a person at-risk of suicide?

Finally, and really the first question people ask but perhaps the best to close on, is “why?”. Why, as their last act – which is ironically often timeless when the actual person’s time is up – would someone take to a site which is conceivably the greatest topical melting point on the internet? A place where their plea or testament will go mostly unheard, or even denigrated.

We romanticise suicide: elevating it from the last minutes of a terrible situation to a picture of stately human tragedy. A sky-blue bird, character-limit and a massive lack of privacy contradict our sense of the humble nature of suicide. Suicide perhaps isn’t humble in the sense we’d like to think it is: rather than quietly relinquishing an unhappy life, suicidal people can be bitter, torn, aggressive and, indeed, desperate for positive attention. Twitter is a little function of our computers or smartphones we can go to for some attention. A Twitter reply, or Facebook like, are small bursts of self-esteem.

Self-esteem is often the critical quality missing. So low, in fact, it feels wrong to go to another person and burden them with extremely serious issues, however much they need it. Twitter is at once heavily populated and strangely anonymous: when a suicidal tweeter makes use of the platform he is not making his address to anyone in particular, but just to be heard. Perhaps in some situations this works, but at least we know in some the issues they have are not, and are addressed with timid, often confused, replies.

This is indeed the digital age: the internet is a repository of every single fascination that people will have. In acknowledging that people are now bringing to Twitter some of their deepest emotions, and occasionally their last sentiments, there is a need to make people aware that it is not all a news bulletins and trolling. And that intervention can be effective.

Of course Twitter is not the only source of suicide talk online, but it is perhaps the most populous, and that’s why it gets the airtime it has. Also, in dedicated suicide forums there is often the feeling more of a ‘culture’: interest in suicide, but not the motivation to do it. But, to be honest, right now the question of what is and what isn’t of concern is incredibly blurred.

Still: we are one step closer to understanding the digital Wild West, which we are building and adapting to in parallel. There is a need to address this issue and somehow develop a precision in how we deal with people who feel there have no other outlet than their online presence allows. Better yet, we can deal with the threat of suicide, which is increasingly pervasive around the world as we move deeper into the 21st century.

For more information about suicide prevention or to speak with someone confidentially, contact the National Suicide Prevention Lifeline (U.S.) or Samaritans (U.K.).