The Nature of Human Violence

  • Details
  • Transcript
  • Audio
  • Downloads
  • Extra Reading

In this lecture, Professor Gwen Adshead will explore current conceptualisations of violence, using criminological, penal and psychological perspectives. She will discuss why rates of violence appear to be falling, and whether all forms of violence are the same.

She will explore the relationship between mental disorder and violence, and the concept of 'normal' violence in liberal democracies. 

Dr Adshead will suggest that it may be fruitful to understand violence as a multiply determined act which has meaning for the perpetrator.

Download Transcript

04 February 2015

The Nature of Human Violence

Professor Gwen Adshead


In this lecture I explore current conceptualisations of violence, using criminological, penal and psychological perspectives. I discuss why rate of violence appear to be falling, and whether all forms of violence are the same.  I explore the relationship between mental disorder and violence, and the concept of 'normal' violence in liberal democracies.  I suggest that it may be fruitful to understand violence as a multiply determined act which has communicative meaning for the perpetrator.


Human violence appals and fascinates in equal measure. Both in fiction and in fact, we seem to enjoy reading about violence or watching on the TV or on the big screen. In most liberal democracies and technocracies, we collect data on violence that we hope will help us devise strategies to reduce it.  Violence is a vast subject, encompassing a wide range of academic disciplines: sociology, criminology, political science, many different sub-specialities in psychology, jurisprudence, anthropology, penology, gender studies, pharmacology,  psychiatry and even sports writers.

It is with some trepidation therefore that I have set out to discuss human violence with such an august audience as may be found at Gresham lectures.  I cannot claim expertise in all the academic domains described above; my claim to experience is based humbly, and with many reservations, on my experience as a forensic psychiatrist and psychotherapist who, for the last 25 years,  has been interviewing and talking to men and women who have committed acts of violence.  I have had a chance to reflect and take stock of what I have heard; and welcomed a chance to put that experience in context with the multiplicity of information and evidence about human violence that is now available. Some of what you hear will surprise you; some will seem all too familiar.

Is it normal for humans to be violent?

I have deliberately started with a normative question, instead of the usual process of finding a good dictionary and definition. I do this because I think the position that one takes on this question drives many others and influences the way we define violence, research it and interpret studies about it.

There is an active school of argument for whom the answer to this question is emphatically "yes'' ; and any suggestion otherwise is just silly. This argument is usually based firmly on our primate heritage, and the evidence that many primates who live in groups inflict serious harm on others who are either vulnerable or in competition for territory or food. There is ample evidence that chimpanzees and baboons attack other animals (either within group or strangers) apparently in pro-active ways i.e. the attacks are not reactive and defensive. Vulnerable adults, who pose no threat, may be victimised in a variety of ways; including assault, abduction and apparently purposeless killing.  

We share at least 90% of our genetic material with primates; the estimates of the exact amount differ.  If this is the case, so the argument goes, then violence to competitors and the vulnerable is biologically normal: undesirable and unappealing perhaps, and not to be ignored; but essentially normal for social primates, especially ones with increased intelligence and resources.

To bolster this argument, one might say that one has only to look at the data on violence. If we include international and civil conflicts, the historical record produces figures that are too big to make sense of. For example, here in Europe we have had an opportunity to reflect on the death toll in WWI, which some reports at as many as 10 million young men died over that 4 year period.  Some have called the 20 century the most inhuman of periods in human history, given the loss of life in human conflicts, both national and international.

The advent of better statistics, the internet and global collaboration in relation to data collection and sharing means that we have better data than ever before, even in low income and developing countries. It is possible to compare rates of violence and homicide between countries; and even come up with a figure for the daily homicide rate: over 11000 people are murdered every day world wide. The plight of vulnerable groups seems particularly poignant: WHO figures suggest that 35% of women world-wide have experienced either sexual violence or intimate-partner violence in their life time (WHO violence against women fact sheet #239; updated November 2014). Children and young people also seem to be at increased risk of fatal violence: whether intentional or unintentional (as in war). A particularly grim statistic  from the USA CDC National Centre for Injury prevention and violence reduction is that globally more than 9 people per minute die from violence or injury in the USA; a staggering 5.8 million people of all ages and economic  groups  who die each year from either unintentional injury or violence .

However, the discerning reader might pick up on the elision of unintentional injury and violence. We might ask why intentional injury and unintentional injury have been collapsed together as if they were the same phenomenon. The CDC data states that the three leading causes of injury and violence related death annually in the USA are road traffic accidents (1.3 M); suicides (844000) and homicides ( 600000). It seems odd to put together a death from a car accident and death from a gunshot or knife blade; and if one thinks about the mind of the person who was operating the instrument that brought about death, it would seem unsafe to assume that what is going on the mind of a drive who loses control of a car is the same as what is going on in the mind of the person who loses control of his temper and shoots a gun as part of that loss.

 If we look at the CDC figures, it is true that they are dreadful in their absolute number; and that is just for the USA. The UN Global study of Homicide (UNODC 2013) estimates that 437000 people were killed by homicide in 2012. But the other question that the numerate reader might ask is: what's the denominator?  It is truly dreadful that 437 000 people die each day through murder, but what proportion is this of the world's population? The population of the world is just over 7 billion; considerably less than 1% of the world's population die every day by homicide.  Compare this figure with annual deaths from malaria (1.8M) or from road traffic accidents (1.24 M) .

Common is not normal

The numbers of deaths from malaria and care accidents are far in excess of the numbers of deaths from intentional human violence; yet no one would suggest that deaths from malaria or car accidents are 'normal' for humans; and in fact we collect data and invest in reduction of these figures precisely because we think that it is not normal for people to die early in this way. There are two other good reasons why we might doubt that violence is normal for humans. First, the rates of fatal violence are different in different countries across the globe: dramatically different.  Some countries have enormously high death rates from homicide that have not changed much in 30 years; others have equally low rates. I shall come back to say more about this below.

Second, there is good evidence that rates of homicide have been falling world wide; especially in societies with organised governments based on liberal democratic structures. For example, in England and Wales, rates of homicide have almost halved over the last 15 years; from 1100 homicides in 2001 to just over 600 in 2013. There are odd years where there are 'blips' ( one caused by Dr Shipman, although the figure is not certain); but generally speaking, homicide is a very odd behaviour indeed, compared to all the other forms of criminal rule breaking that take place. Violence of any kind accounts for only 20% of recorded crime in England and Wales (Smith 2011); theft and variations on a theme of theft and fraud are the majority of offences.

It might be argued that not all violence is reported, especially violence in the home and violence that involves sexual assault. The British Crime Survey focuses on victim experience not reported crime; and found that there were just over 1 million reported assaults between young men, and a similar number reported between intimate partners ( what used to be called domestic violence). Even if these figures were 100% inaccurate (and they have been steady for about 10 years), they still represent a minority of the population acting as violence perpetrators.  Even if it is true that everyone has the potential for violence, it is clear that only a sub-group of humans are actually perpetrating the violence.

Violence and meaning

I will have more to say about perpetrators of violence in my next lecture.  My principal thesis for this talk is that violence is not normal, even if it is way too numerically frequent. My concern is that if human violence is taken to be 'normal', then it will be dismissed as inevitable. Further, claims that violence is normal in some sense may mean that strategies to reduce and prevent violence may go off on a tangent or have a mistaken focus.

The UNODC review divides homicide into three groups:  socio-political homicide, interpersonal homicide and homicide related to criminal activities. Taking the last first, 'homicide relating to other criminal activities' is that homicide which we call organised- crime- related, including homicide relating to robbery. It is commonest in the Americas (North, South and Central); for example, in Jamaica, homicide related to organised crime or robbery accounts for 84% of homicide; compared to Sweden where it accounts for only 39% of homicides.  The UNODC consider that the homicides identified here are not the primary goal of the crime; but a side effect of disputes involving strength, territory and control, in which young men are the primary victims.

Interpersonal homicide (IPH) is found across all areas of the globe; and accounts for 14% of homicides world-wide. Women are the most frequent victims of this form of homicide. Here the homicide is a means of resolving conflict with, or punishing a person to whom one has a strong emotional attachment.  This is the majority of homicides in countries like the UK or Scandinavia; which may also explain the amount of crime fiction produced by these countries, since interpersonal homicides are stories of conflicted emotions, loss and grief. They are also the types of homicide most associated with mental distress and disorder; a topic that I return to in my second lectures. 

The final category is socio-political homicide; a category that includes civil terrorism, genocide and hate crimes. The UNODC report makes an important observation here; namely that these are homicides that a means of pursuing a political agenda, and are intended to convey a message to others. We should note especially this chilling comment: 'people are killed for what they represent' i.e. the victim's personal identity or experience is not relevant here as it is crucially in inter-personal homicide. Unsurprisingly there is less data about this form of homicide than the others. It is hard not to think that the recent atrocities in Paris and Nigeria fall into this category; and that the claim that these homicides are motivated or justified with reference to religious belief is a 'cover story' for an anti-democratic political program.

Human violence: meaning and complexity

If we do not look at the meaning of homicide for the perpetrator, we will not be able to help him or her explore whether there are other means of achieving their goals and dreams.  As the UNODC report makes clear there is not a 'one-size-fits all- approach to violence, because all violence is not the same.  Rates of organised crime related homicide fell in Italy  and El Salvador with changes in government; rates of homicide have increased in those countries like Sudan, Syria and Iraq where there have been rapid changes of government and uncertainty about who is in charge.

Governments turn out to be key players in relation to human violence; and there is evidence that rates of violence fall and rise when governments of different political colour are in power.  Professors Lee and Gilligan studied homicide and suicide rates in the USA, and showed that they rise when Republicans are in power and fall when Democrats are in power (Gilligan, 2011). Furthermore, this effect is especially pronounced in the southern states of the USA. Gilligan has argued that the effect can be understood thus: that Republican governments favour polices that promote inequality and adversarial competition between people. In austere adversarial social conditions, there must be 'winners' and 'losers'; and the losers experience shame, loss and awareness of vulnerable neediness.  In the Deep South of the USA, there is a culture of masculinity in which vulnerability, shame and dependence are incompatible with being male; real men are always in successful control of their homes and lives. If governments enact legislation that supports corporate expansion that exacerbates social inequality, and makes it difficult for individuals to change or influence their economic situation, there will be increased unemployment and loss of jobs and status. Men in the southern states who lose their status and identity tend to kill themselves; often also killing their families before they do so.

There is ample evidence from other sources that homicide rates rise with inequality of income (Equality Trust, 2011) .Research with similar findings has been carried out in the UK; although the effect is not so pronounced because the inequality is not so extreme, and the base rate for homicide is so low. The point here is that taking life is a multi-factorial action by a human being; not a meaningless instinct. For too long, it has been assumed that violence is part of our animal heritage, and that violence is inevitable because men (and especially men) will behave like animals when under pressure. In my own field of forensic psychiatry, there is still an unhealthy tendency to lump all acts of violence together; to deliberately not look for the meaning behind the violence. As the UNODC report points out, violence is often a communication to some person or group of persons; a communication of rage, hate, shame and despair.

My colleague Peter Aylward, for many years a therapist at Broadmoor Hospital, first suggested to me that the enacting of violence might be best understood as resembling a bicycle lock.  In order to open the 'lock' and release the violence, a number of factors might need to be in place. The first two factors are likely to be socio-political: reflecting attitudes to masculinity for example and/or attitudes to vulnerability and dependency.  The last 2 numbers may be individual to the perpetrator, such as a substance misuse history or the experience of childhood adversity or brain damage.  The final number that opens the 'lock' may be highly idiosyncratic to the perpetrator, and may relate to past traumatic memories, unresolved distress and unstable feelings of despair and shame. The final 'number' may also involve highly idiosyncratic interpretations or distortions of the victim's last words or actions. I am thinking here of a mother who tried to smother her baby after he had been sick on her shoulder.  Given the frequency of babies being sick on their mother's shoulder, it seems likely that there was something about that event that had particular meaning for this mother; and triggered a uniquely hostile response in her. I am thinking also here of the man who said to me of the partner whom he killed:  'I wanted to kill her, I didn't want her to die'.


It is a truism that human beings are complex; and that complexity has been the basis for drama, poetry, stories, and music  for over 500 years. Violence is a human action which is also complex; and attempts to reduce it to just a natural part of human behaviour will miss the point spectacularly.  There is nothing normal about attacking another person, especially if that attack is fatal.

So what strategies might be helpful? Atul Gawande, in his book (2010) and his 2014 Reith lectures, also makes a strong case for taking human systems and complexity seriously when trying to reduce human error and fallibility. It seems to me not unreasonable to see human violence as human error and evidence of fallibility, so I wonder if some of his approaches might be helpful.  For example, he suggests that humans use checklists when carrying out a complex operation; which not only means that apparently unimportant details are not missed out, but also means working in teams that can pool information. One area where this is already being talked about is in relation to child abuse: the kind of physical abuse and neglect that results in over 50000 children being taken into care in England & Wales (National Audit Office, 2014).

This is crucial issue for violence prevention, because there is extensive evidence that a male child who experiences physical abuse and neglect  is at substantially increased risk of being arrested for a violent offence in adulthood ( Gilbert et al 2009).  Multiple studies have shown that hostile parenting i.e. your parents treating you with cruelty, disdain or indifference is a major risk factor for childhood antisocial behaviour, which in turn increases the risk of in actual crime and violence. Experiences of physical abuse and neglect combine with genetic vulnerability and brain injury to have a negative effect on brain development ( Raine et al 2001; Widom &Brzustowicz,2006); such that the child does not develop a social mind or attachments to people.

One intervention that we could do is apply a checklist of risk factors to every pregnant woman who books into an antenatal clinic. These risk factors include: being under 21, having an unplanned pregnancy, already having a child placed in care, and having had an insecure childhood oneself.  When we identify a woman with at least 2 risk factors, we could offer her an extended psychological and social intervention that continues through the vital first 1000 days of a child's life, when key parts of the brain develop (WAVE Trust 70/30 programme). The intervention would include individual and group activities that provide psychological support and promote the mother's capacity to think about her baby's developing mind.  It would need to be developed and delivered by therapists and professionals who have experience in this field and know what they are doing.

There are small pockets of excellence in England & Wales that already provide this sort of service, but the numbers they can help are pitifully few.  There may be approximately 60000 women in our society at present who need this type of expert intervention, and they need it now.  It will be expensive.  Some will say it will take money away from other services (although it doesn't have to if resources were more equitably spent and managed). It is not glamorous and it can't be done on the cheap. It will involve that anathema to British bureaucratic life, namely joined up thinking and joint working across different systems of services.

At present, we spend millions in taking children into care because their parents pose a risk to them. It costs about £23000 to take a child into care; so the accumulated costs of all those 68000 children in care was £1,564,000,000. A sub-group of those children will be violence perpetrators of tomorrow, or will be the mothers of future violence perpetrators.  As I write this paper, I cannot find anyone to offer specialist treatment for a young woman, depressed and distressed, who mistreated her child. If she goes to prison, she will come out without any understanding of what she did or why; and she has many years of reproduction ahead of her.  She lives in an area where social and health services have been cut. Her baby will be taken into care; and already has one of the risk factors for violence perpetration in later life. This case epitomises the ecological nature of violence prevention; that strategies to reduced violence will need to operate at governmental level, and at the level of the individual potential perpetrator; and all points in between.


Centers for Disease Control and Prevention:

Ofsted report (2014) In the child’s time: professional  responses to neglect. Ofsted, Manchester.

Gawande, A. (2010). The checklist manifesto: how to get things right (Vol. 200). New York: Metropolitan Books.

Gilbert, R., Widom, C. S., Browne, K., Fergusson, D., Webb, E., & Janson, S. (2009). Burden and consequences of child maltreatment in high-income countries. The lancet, 373(9657), 68-81.

Gilligan, J. (2011). Why Some Politicians are More Dangerous than Others. Polity Press. London.

Glaser, D. (2014). The effects of child maltreatment on the developing brain. Medico-Legal Journal, 82(3), 97-111.

Krug, E. G., Mercy, J. A., Dahlberg, L. L., & Zwi, A. B. (2002). The world report on violence and health. The Lancet, 360(9339), 1083-1088.

Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A., & Lozano, R. (2009). World report on violence and health. Geneva (Switzerland): World Health Organization; 2002.

NAO  & Department of Education (2014) Children in care.

Raine, A., Park, S., Lencz, T., Bihrle, S., LaCasse, L., Widom, C. S., ... & Singh, M. (2001). Reduced right hemisphere activation in severely abused violent offenders during a working memory task: An fMRI study. Aggressive behavior, 27(2), 111-129.

 Smith K (Ed.);  Osborne,S., Lau,  Ivy & Britton A (2012) Homicides, Firearm Offences

and Intimate Violence 2010/11. London. Home Office Statistical Bulletin

UNODC (2013) Global report into homicide.

WAVE  Trust 70/30 program

WHO violence against women

Widom, C. S., & Brzustowicz, L. M. (2006). MAOA and the “cycle of violence:” childhood abuse and neglect, MAOA genotype, and risk for violent and antisocial behavior. Biological psychiatry, 60(7), 684-689.


© Professor Gwen Adshead, 2015

This event was on Wed, 04 Feb 2015

Gwen Adshead

Dr Gwen Adshead

Visiting Professor of Psychiatry

Dr Gwen Adshead was Visiting Gresham Professor of Psychiatry and currently consultant forensic psychiatrist at Ravenswood House. Prior to this post, she worked at Broadmoor...

Find out more

Support Gresham

Gresham College has offered an outstanding education to the public free of charge for over 400 years. Today, Gresham plays an important role in fostering a love of learning and a greater understanding of ourselves and the world around us. Your donation will help to widen our reach and to broaden our audience, allowing more people to benefit from a high-quality education from some of the brightest minds.