For OCTOber 2003 issue

SCIENCE IN CRIME DETECTION-121

CAN VIOLENCE BE PREDICTED?

 ‑Dr. Anil Aggrawal

This time I am going to talk about a very interesting subject – Can violence be predicted? If we could develop a way to predict violence, we could take appropriate steps to curb it, and the crime rate, especially that of homicide in our society would come down a great deal. That would indeed be a very great contribution of criminology to modern society.

A few things have to be understood at the outset. First of all, homicide is very infrequent in our society. And the prediction of all infrequent things – be they earthquakes, or sunspots or homicide – is very difficult. According to Uniform Crime Reports, the 1982 rate of homicide in the USA was just about 9 per 100,000. That is out of all 100,000 crimes committed, only 9 were those of homicide. Even if we consider only violent crime, homicide still makes up only about 0.4% of all violent crimes. Predicting an event which has such a low frequency is definitely dangerous business.

Prediction of violence can have many practical applications. First and foremost is of course, that we could take interventional steps to prevent homicide. Furthermore, if prediction of violence was possible, quite likely, in cases where courts have to grant bail to dangerous criminals, they would first ask scientists, if it was okay to release those suspects on bail. The scientists could examine the personality of the suspect and advise the court whether violent behavior could be predicted from them or not. If it could be predicted, the courts would not grant bail, otherwise they would.

There is another very practical use of prediction of violence. In Texas, USA, when it has been proved that a person committed a murder, the courts would grant capital punishment only when it could be demonstrated that there was a continued probability of future violent conduct from that criminal.

From a prediction standpoint, one can say that there are certain high risk individuals, i.e. those persons who have a very high probability of showing violent behavior. In addition there are certain high risk situations, i.e. those situations where a person is more likely to show violent behavior. For instance certain forms of mass murder are seen more commonly when a person is fired from his job, or when he humiliated, or when domestic problems supervene.

One area that has to be looked at is, what will be the rate of accuracy of such predictions? Another area that we have to look at is, what will we do, once we are indeed able to predict violent behavior. How are we going to curb violence then? What will be the interventional strategy? What will be its cost? Can we, kind of introduce mass inoculations against violent behavior?

Alan Gruber, a prominent neurologist wrote in the New York Times of September 17, 1985, “I would submit to you we can predict human aggressive behavior 90% of the time”. But the question is, can we accurately predict 90% of the time? That is a word often left out. Can we accurately predict?

Some years back, it was reported in the US, that some income tax officials were murdered when they went out for field tax audits. They found out that there was a two or three fold increase in the number of assaults during audits. Consequently they asked investigators to look into this whole incident. And perhaps let them know what kind of situations would lead to such assaults. These would be termed high risk situations. A training program was then developed for IRS (Internal Revenue Service, equivalent to our own Income Tax department) officers to consciously avoid those situations.

A very interesting study was done by Henry J. Steadman, Chief of the Bureau of Evaluation Research, New York State Office of Mental Health, Albany, New York. He along with a colleague Felson, interviewed more than 500 individuals. One third of those individuals were ex-mental patients, one-third were offenders and the rest were randomly selected from the general population.

They interviewed them concerning disputes that had occurred in their lives with primary emphasis on their most recent one. Three types of incidents were focused upon: when they got angry and did nothing; when they got angry and screamed; and finally when they got angry and pushed, hit, kicked, shoved someone. Some interesting findings emerged from this study.

It was found that

The more serious incidents more often involved younger people and males

More serious incidents occurred when the involved party/parties were drinking

When strangers/third parties were present, there was a higher level of seriousness. Interestingly all three groups – mental patients, offenders and general population – tended to be serious in the presence of strangers/third parties.

When gun/knives were used, the disputes more often involved property.

Disputes that ended in yelling and anger, were more often domestic ones.

In disputes that ended being less serious and harmful, it was seen that people used some form of “account giving”, i.e. they used verbal mechanisms to defuse the situation rather than having them going on to become more serious. This was an area where mental patients were less skilled. They less often gave accounts to try to cut off the level of intensity of a dispute. Steadman and his colleague suggested that this was something therapists might think about in terms of training mental patients to function in the community.

S. Armstrong has developed his “seersucker theory”, which says, “No matter how much evidence exists that seers do not exist, suckers will pay for the existence of seers”. Steadman quotes this in his paper to stress that actually prediction of homicide is very difficult, yet we would probably continue to believe in it.

There are violence-prone individuals and there are violence-prone situations. One can reduce homicide, if both are reduced to some extent by some mechanism. The stress should be on preventing a violence prone-individual meeting a violence-prone situation.

Some things that can be done by the physicians/psychiatrists are:

Ask patients about their past history of violence, including arrests and incarcerations. Questions like this can be asked:

What’s the most violent thing you have ever done?

Have you ever threatened your wife?

Have you hit her?

Use of disinhibiting agents, such as alcohol, barbiturates, amphetamines, hallucinogens, inhalants.

Use of availability of weapons

One question that has been raised again and again is whether the criminality of a person depends on his genetic makeup or on his environment. Studies have been conducted on various kinds of twins – both monozygotic and dizygotic. Monozygotic twins are those that have the same genetic makeup. Dizygotic twins have different genetic make up. Environmental influences are the same in both monozygotic and dizygotic twins. It has been seen that persons who share common genes are more likely to be like each other as far as criminal conduct is concerned. Similarly criminalities in adopted children has also been studied, and conclusions are the same. The adopted children only share the environment, but not the genetic makeup of their parents. If the parents have criminal tendencies, the adopted children do not tend to show them. This means that genetic makeup is important. Some kind of screening of such children is important.

It is also seen that people of certain genetic makeup have more criminal tendencies. For instance, people who are XYY tend to show more criminality that the rest. Some scientists have studied XYY configuration in male prisoners and have come to the conclusion that this configuration was associated with greater violence.

Another important area, which doctors are looking forward to now, is a place in the brain called the limbic system. It comprises of cingulum, hippocampus, thalamic and hypothalamic nuclei as well as the basal ganglia, mid brain and the amygdala. The last three are now also included in the limbic system by modern brain physiologists.

Captions to the figures supplied

This article has (1424) words

Dr. Anil Aggrawal, M.D. Professor of Forensic Medicine Maulana Azad Medical College New Delhi‑110002

My notes:

This article is Based on the following papers:

1. Predicting Violence leading to Homicide by Henry J. Steadman, Ph.D., Chief, Bureau of Evaluation Research, New York State Office of Mental Health, Albany, New York. Bull. N.Y. Acad. Med. Vol. 62, No. 5, June 1986, pages 570-578

2. Comments on Dr. Steadman’s Paper by John R. Lion, M.D., Clinical Professor of Psychiatry, pages 579-581

3. Medical Deterrents by Stephen Nordlicht, pages 582-588