Farewell to the World: A History of Suicide 
by Marzio Barbagli, translated by Lucinda Byatt.
Polity, 407 pp., £19.99, September 2015, 978 0 7456 6245 9
Show More
Show More

Around​ forty years ago, a friend of mine took his own life in the middle of a party he was throwing in his apartment. A neighbour who happened to look outside saw him climb onto the window ledge, hesitate briefly and then jump to his death from the fifth floor. His guests were stunned when the police rang at the door.

Why did he do it? We grasp at reasons, motives, causes, triggers. Sometimes people oblige us by leaving a note behind to explain their action. It was, the note says, because of a severe illness, or unbearable pain, bankruptcy, despondency, loss of a loved one, betrayal, a slight, boredom. (My friend didn’t leave a note.) Sometimes people kill themselves just in order to leave such a note. Others were crying for help, but nobody heard them. There is a laundry list of possible reasons, but do any of them answer our question? Suicide is an anomaly: what causes some people to kill themselves when most of us live on?

There is no shortage of theories. For more than a thousand years in the West, suicide was ascribed to Satan. In the first systematic treatise devoted to the topic, published in 1637, John Sym, a Puritan, was still able to attribute suicide to ‘the strong impulse, powerfull motions, and command of the Devill’. But natural causes were increasingly being blamed: melancholy, a dysfunction of the hypochondriac organs, folly, a foul climate, the ‘English malady’. Predictably, psychiatrists staked their claim to suicide as soon as their discipline got off the ground. ‘I believe that I have demonstrated that a man does not attempt to end his days except in delirium,’ Jean-Etienne Esquirol wrote in the early 19th century, ‘and that suicides are insane.’ Others, like the phrenologically inclined physician François-Joseph-Victor Broussais, believed that suicide was caused by a defect in the still to be located organ presiding over our ‘propensity for staying alive’.

Then came Durkheim. By his own account, his 1897 study, Suicide, is a deterministic, ‘aetiological’ theory. It is a fascinating attempt to find the sole cause of suicide, the one that would render unnecessary all the other causes adduced in popular wisdom and by rival sociologists. ‘The motives that are … attributed to suicides are not their true causes,’ Durkheim insisted. If people killed themselves, it was not because of neurasthenia, race, heredity, the climate, or Gabriel Tarde’s silly notion of ‘imitation’, that suicide was the result of a psychological contagion. Durkheim argued instead that society was to blame. Poring over suicide statistics, he noted that annual suicide rates tended to remain stable in a given society over a given period. This, he argued, ‘can only be due to the permanent action of some impersonal cause that hovers over all the particular cases’. And this cause was society, a collective being ‘as real as cosmic forces’, which exacts an unvarying suicide toll from the group. In a draft rebuttal, Tarde was quick to call this society a ‘mythological entity’: Durkheim, he wrote, had hypostatised a statistical constant, making it into a ‘divine person, infinitely superior to individuals and commanding them’.

Durkheim went further. From the fact that suicide is caused by society, he deduced that it is a social pathology – or, more exactly, a pathology of the social. Why, for example, are married people better protected against suicide than single people, or Catholics and Jews than Protestants? Because they are better integrated into their social group. Conversely, the suicide rate increases as soon as the social bond slackens and the individual asserts itself against the collective. Young, politically integrated societies, such as the Roman Republic or the early Empire, had fewer suicides than decadent, disintegrating ones, such as the late Empire. Durkheim calls this type of suicide ‘egoistic’, meaning asocial. He blames ‘anomie’, the absence of the social regulations that normally curb our unquenchable passions and desires, for a second subset of suicides: people kill themselves, he claims, when they realise they will never be satisfied, but live in a state of constant disappointment. Anomic suicide is typical of modern, deregulated societies where ‘the state of crisis … is constant and so to speak normal.’ The qualifier is in order, for Durkheim’s idea of the ‘normal’ is a highly integrated and regulated society. Behind his theory of egoistic and anomic suicides is a deeply conservative diagnosis of the shortcomings of modern industrial and urbanised societies.

But​ what of the ‘inferior societies’ of ‘primitive peoples’, where the group matters more than the individual? Are they immune to suicide? Durkheim knew this not to be the case: the anthropologist S.B. Steinmetz had debunked that idea only three years earlier. He therefore made room for a third, ‘altruistic’ or social kind of suicide. In such societies, he claimed, people kill themselves out of duty, because society requires it: servants or warriors follow their king to the grave, Indian widows mount the pyre to accompany their dead husbands, old Vikings take their lives before becoming a burden to their kin. Thus people can commit suicide because of too much social integration as well as because of a lack of it. In any case, it’s still society that caused it.

Durkheim’s study has the seductive trappings of a mathematical theorem, which is undoubtedly the reason it became one of the founding texts of the new science of sociology. But as it turns out, it doesn’t square with the facts. Marzio Barbagli, also a sociologist, shows beyond a shadow of a doubt that the father of the discipline got it all wrong. Farewell to the World is a ruthless revisionist study which asks whether Durkheim’s categories can still account for the variety of reasons people kill themselves, and finds that they cannot.

Take the relation established by Durkheim between altruistic suicide and strong social integration. On the basis of his theory, one would expect to find no egoistic or anomic suicides in ‘inferior societies’, where the group supposedly trumps the individual, and no altruistic suicides in modern societies, where the individual trumps the group. But Barbagli has no trouble finding a host of counterexamples. In the main, he shows, people in hunter-gatherer, nomadic and agricultural societies take their lives for the same egoistic reasons we do – jealousy, pain, humiliation and the like. In addition, self-murder was (and still is in some cases) used in many traditional societies to protest against injustice or to shame someone. In India, China, Japan, but also in pre-Christian Europe, people would kill themselves in order to curse someone, to challenge an enemy to do the same, or to start a vendetta. This form of suicide, which Barbagli calls ‘aggressive’, was overlooked by Durkheim and clearly can’t be classed as either obedient or altruistic.

Even the most striking example of altruistic suicide, the sati of the Indian widow, makes a poor fit with Durkheim’s theory. As Barbagli points out, sati did not exist in India before the third century bce and was barely practised in the thousand years that followed, though women were just as subordinate during that long period as they were in the 19th century, when sati reached its peak (there were almost a million suicides that century before the British made the practice illegal). As for the Chinese ‘faithful widows’ who also refused to survive their husbands, they often committed suicide to resist pressure from their families to remarry. It was an act of disobedience, not of submission.

At the other end of Durkheim’s historical arc, there is today a stunning epidemic of altruistic suicides in our egoistic societies. Durkheim’s theory certainly didn’t predict this: all over the world often highly educated and ‘Westernised’ people launch suicide attacks or blow themselves up in order to inflict casualties on their enemies. We call them terrorists, but they see themselves as martyrs who are sacrificing themselves for a larger cause, just like the early Christians.

Barbagli traces this new phenomenon back to earlier forms of aggressive suicide, such as the Indian and medieval Irish practice of fasting in order to gain ‘redress’, which was appropriated for political purposes by Gandhi and Irish republicans in the form of hunger strikes. Then came the first public self-immolations, such as that of the Buddhist monk Quang Duc during the Vietnam War or the Czech student Jan Palach after the 1968 invasion; the practice was weaponised to devastating effect by Hizbullah when it launched the first suicide missions against American and French soldiers in Lebanon. This form of ‘suicide-as-a-weapon’, as Barbagli calls it, has spread like wildfire (by imitation, Tarde would say). It is clearly altruistic in Durkheim’s sense, but it can hardly be conceived of as being required by a cohesive and integrated society. Rather, it is a desperate attempt at repairing a damaged social bond.

Durkheim’s theory would lead one to expect the rate of egoistic and anomic suicides to go up in our increasingly individualistic societies. The opposite has happened. While there was indeed a steady rise in suicide rates in the West from the end of the 17th century until well into the 20th, the last thirty years have seen a striking decline. The explanation certainly isn’t that social bonds have grown stronger. Barbagli thinks the drop is a result of improvements in pain management and the medicalisation of depression: instead of committing suicide, people seek therapy and/or take antidepressants. This explanation seems counterintuitive at first, since Barbagli himself shows convincingly that it was the gradual spread of secular, often medical explanations for suicide from the 18th century on that first led to the rise in the number of self-murders. St Augustine had declared suicide a sin, the gravest crime against the Creator, and this doctrine had been enforced with utmost ferocity throughout the Middle Ages. Suicides were denied burial; their corpses were tried post-mortem, mutilated, hanged upside down from gallows, dragged through the streets; their houses were ritually torn down (a custom known as ‘ravage’) and their properties confiscated. But once suicide was no longer seen as the work of the devil, killing oneself became acceptable. ‘Life is slavery if freedom to die is wanting,’ Montaigne asserted.

But wasn’t that Durkheim’s point? The more people free themselves from society and its binding rules, he claimed, the more likely they are to commit egoistic or anomic suicide. They think they are choosing death freely, but in fact their actions are determined by society. Barbagli, interestingly, doesn’t deny that the integration/regulation pair plays a role in some of the variations in the suicide rate, notably in its increase during the 20th century. But he rejects Durkheim’s rigid determinism and his mythological concept of society: not only do people kill themselves for all kinds of reason that do not correspond to Durkheim’s scheme but, fundamentally, reasons are not causes.

Upper-class​ Englishmen began to take their own lives in droves towards the middle of the 18th century – in 1759 Lady Montagu called it a ‘fashion’ – but this striking development certainly wasn’t caused by a lack of social integration, or by what George Cheyne, in a bestselling book of 1733, diagnosed as The English Malady, a melancholy he attributed to the effects of the English weather and food on the spleen. Rather, Cheyne’s medical theory of suicide provided a template for action: if English and melancholic, then take your life. As one observer, Jacob Zimmerman, astutely put it: ‘An Englishman suffering from melancholy would shoot himself in the head. A Frenchman suffering from the same disease would take monastic vows, which, as a matter of fact, is the same thing. The Englishman would not kill himself if there were monasteries he could go to.’

Barbagli seems to argue that psychotherapy and/or antidepressants are effective in combating the desire to end one’s life, but it’s unclear that he is right: SSRI antidepressants and other psychotropic drugs have been shown to cause compulsive suicidal ideation (‘suicidality’) in some people. A better way to describe the situation would be to say that current medical and psychological theories have popularised the idea that depression (melancholy) is treatable. If depressed, talk to your doctor. Hence the drop in suicides: being in treatment has now become a socially sanctioned way of expressing distress.

Isn’t talk of the ‘social’ a concession to Durkheim? Barbagli prefers to speak of cultural factors: ‘My thesis here is that the factors that have most influenced the frequency of different types of suicide are cultural, or in other words they comprise the wealth of cognitive schemas and classification systems, beliefs and norms, meanings and symbols available to men and women.’ So if English melancholics killed themselves while present-day depressives tend not to, it is because of differing expectations, ideas and theories about suicide. In one cultural context, say radical Islam, blowing yourself up with a suicide belt will get you to heaven. In another, say medieval Christianity, ending your life will get you to hell. Different cultures present different options when it comes to acting on distress, anger or despondency, which is the reason rates and forms of suicide vary so widely.

Barbagli’s point here is that these cultural scripts or scenarios are not ‘social causes’. They inform, shape and mould people’s actions and intentions, but they don’t cause them. This is what Tarde (who is barely mentioned by Barbagli) called ‘imitation’, a mimetic process he explicitly contrasted with Durkheim’s determinism:

Durkheim has not understood what I mean here by the action of imitation. I never said nor thought that a healthy, happy, content man need only see examples of suicides around him … to be forced to imitate them. But if, being unhappy or ill, in other words in the condition required to feel the influence of this example, he comes to learn that people have killed themselves in such and such a manner, he will decide to do the same and most often in the same way.

But what are these ‘cultural factors’ and Tardian ‘examples’, Durkheim would have objected, if not the beliefs, norms and symbols of a given society? If religion, for example, protects against suicide, it is because it represents the social: ‘The details of dogmas and rituals don’t matter. What matters is that they be of such a nature as to instil a collective life of a sufficient intensity.’ Barbagli, disappointingly, never addresses this objection head-on. ‘Cultural’ is, obviously, another name for ‘social’. So the point to be made here (and Barbagli doesn’t make it, at least explicitly) is that these beliefs, norms and symbols do not represent ‘the social’, as Durkheim would have it. They are sociality itself, the ever changing network of ideas, practices and customs through which people communicate and relate to one another. There is no whole different from the parts, no society hovering over individuals and causing them to do this or that. There are only interactions between participants in a game whose rules evolve according to the players.

It is in that sense that suicide is social (unless it is the result of some psychotic outbreak or chemically induced tendency). ‘We may ask ourselves to what extent it is a social fact,’ Tarde wrote, ‘since it is a sudden and deliberate exit from society. In my view it is social nonetheless, for whoever exits in this way is driven for the most part … by pressures or ideas originating in social relationships … with his fellow beings.’ Even the most solitary and silent suicide is still a social act, because it is a way of relating to others, an idiom of distress or anger.

People kill themselves in different ways and at different rates in different cultures, but not because each society is autonomous and distinctive in the way it determines these things. Rather, the idioms of distress change as they are spoken by various locutors, some of them introducing new variations that are taken up by others. Suicide bombing was unheard of before Hizbullah, now it’s common. Jihadist sociality is the result of this new idiom, not its cause. Only history can tell us why suicide becomes meaningful and attractive to some people. Sociology cannot.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.


Vol. 38 No. 24 · 15 December 2016

Mikkel Borch-Jacobsen writes that the ‘first systematic treatise devoted to the topic’ of suicide appeared in 1637, John Sym’s Lifes Preservative against Self-Killing (LRB, 17 November). It’s true that Sym’s was the first to be printed, but the first to be written was John Donne’s Biathanatos (1608), which, against Donne’s explicit instructions, was published by his son after his death.

Jonathan Holmes

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN


Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences