|The Medico-Legal Society
A meeting of the Society was held at the Royal Society of Medicine, 1 Wimpole Street, London W1, on Thursday, 13th November 2003. The President, Miss Eleanor F. Platt QC, was in the Chair.
Geographical Profiling of Criminals
Professor David Canter, PhD, AcSS, FAPA, FBPsS, C.Psychol.
Professor of Psychology,
Director of the Centre for Investigative Psychology,
The University of Liverpool
© David Canter 2004
The President: Ladies and gentlemen, this evening we are honoured to have Professor David Canter with us, who has come from Liverpool. He is the Professor of Psychology at the University of Liverpool, and has been for almost the last 10 years, not quite 10 years. He has a long and varied CV. He is a Fellow of the British Psychological Society and the American Psychological Association and a member of the Forensic Science Society. He has a PhD from Liverpool and is a Chartered Forensic Psychologist and is one of the first psychologists to be elected as an Academician to the Academy of Social Sciences. He is interested in a number of things, but he provided expert evidence in the case of prisoners, for example, in the Risley Remand Centre charged with riot, that their actions were an understandable response to the environment and the regime. I am not quite sure what he is talking to us about tonight. His recreations are listed as clarinet and musical composition. We are honoured to have Professor David Canter to speak to us this evening. (Applause)
Professor Canter: What I quite unashamedly decided to do this evening was to talk through the theme of my new book Mapping Murder, which is available in all good book shops. (Laughter) When I was asked to give this talk about a year ago it was really quite fortunate that it coincided with the recent publication of the book, so I thought I would talk through the issues that are dealt with in the book, which grow out of the work I have done as a psychologist working with the police and then, moving on beyond that work, to carry out research work that is of relevance to police investigation.
I realise this is a very august gathering, though a rather difficult one for a psychologist to speak to, with doctors, who will tend to regard psychology as a footnote to their discipline, and lawyers, who will regard psychology as probably what they do anyway, and neither of them will be particularly familiar with the concept and techniques of the science of Psychology. So the presentation I have here is the one I gave to some sixth form students just a while ago, and hopefully that will cover ground that will be of interest to you.
I came into crime rather late in life, having been really an applied psychologist working in a variety of different areas of the application of psychology, including into the study of how the physical environment has relevance to people. I started off in a school of architecture as a psychologist lecturing to architects, and that was how I got caught up in looking at the design of the Risley Remand Centre and at the riot that happened there, just in case you were curious about that connection. It is quite an involved story as to how I got caught up in crime and you really have to read my first popular book Criminal Shadows to get the full story there, because what I thought I would talk about today is the development since that early work, really in the mid 80s, when I first became involved with police work, and since that time I have been involved in a number of police investigations trying to contribute in various ways to those investigations, but, as I say, really moving beyond that out into research matters.
It is worth starting with how this unfolded historically. In the mid 80s there were a series of murders around London and a series of rapes that the police linked together to a common offender and, by some curious roundabout route, the police asked me to give some comments on that. I will talk a little bit about that investigation, but it was really as a result of that that I then was given access to data that the police hold and was able to start doing some proper research on it, because the police are quite remarkable in that they won’t give you access to any information unless they think it is going to be of some use to them, and as a scientist it is terribly difficult to be of any use unless you have some information to work with, and so there was a bit of a pulling yourself up by your bootstraps. I had to help as best I could in that early investigation working from first principles without being able to test the veracity of those principles in the particular context of violent crime.
The police became very enamoured at the result, as shown by the sort of Press cuttings in Figure 1. These were a product of the police selling the idea that psychologists were now contributing to police investigations and that here they were at the cutting edge of the use of new scientific procedures. This recognition of the potential helped to open doors that allowed us to start doing proper research. This early case really opened up a particular way of thinking about criminal activity that then turned out to be very productive.
Figure 2 is a map of London. Each of the points is a location where crimes had been committed, mainly rape, over a four-year period. This is typically what the police would have, a record of where the crimes had occurred, for operational purposes to give them some basis for starting to decide where they were going to deploy officers and how they were going to carry out their investigations.
I was asked if I could help. I still remember very clearly the day when I went to where a number of senior police officers were meeting at Hendon, the training college for the police, and being asked “Can you help us catch this man before he kills again?”. The full details of how my response to this unfolded are given in my book Criminal Shadows. But, in essence, I started by exploring how the crimes unfolded over time as indicated in Figure 3.
As an environmental psychologist, interested in how people make sense of their surroundings, it seemed to me that there was likely to be some sort of process involved there, and this is what we found: in 1982 a series of rapes had happened in one limited area, in ’84 there had been a few more in a broader area, in ‘85/’86 there had been some more spreading out even wider as can be seen in Figure 3. There was no indication of anything happening in ’83 that could be tied into these cases. Then there were three murders that happened really further away. It seemed just painfully obvious that there was some development in this process, with the murders occurring round the edge and, as can be seen, the movement out from the area of the earliest crimes. When we looked at the details of the behaviour there also seemed to be a psychological development.
The early crimes seem to have been committed by two individuals, tended to be at the weekend, one of the individuals even seemed to be a bit remorseful and on occasion apologised to the victims that they had raped. The later crimes tended to be during the week, tended to be by an individual alone, were much more determined, much more planned. This indicated the possibility of a developing commitment to a criminal life, and related criminal activity. That psychological development seemed to be reflected in the growing geography of the crimes. Therefore it seemed very likely that what we should be thinking of is the early stages of the crimes being in a sense more amateur, almost sort of ‘recreational’, if you can describe rape in that way, and less committed, less determined, less part of a convinced career than in the later part. If that were the case then if we ran the crimes backwards we would actually be able to end up with the idea of the area circumscribed by the initial three 1982 crimes being where the individuals started from. Within that area, I thought, perhaps they were less likely to be thinking through what they were doing, acting as part of their day-to-day activities. I thus suggested to the police that they were looking for individuals who might be based in the Kilburn area of North West London in 1982.
I also made some other comments, like the fact that this individual particularly who had raped and murdered young women probably had a history of being violent against women and was quite a nasty character and would be known to be so. This turned out to be important in the sense that one of their suspects was in the police records and being looked at only, as far as they were concerned, because he had violently attacked his estranged wife at knifepoint and raped her. It was this that had brought him onto the list of suspect but at that stage some police officers considered this to be ‘just’ a domestic; they thought it was a tussle between a married couple and they couldn’t see this individual going off to commit a series of rapes and murders. Consequently, the comment that the individual they would be looking for would actually have some history of that type of dealing with women was something I helped to emphasise.
There were a variety of other issues I mentioned to the police but the most significant one was the suggestion that the individual lived in a limited, identifiable area of London. It turned out that of all the possible suspects they were considering - and there were really only half a dozen that they had very serious consideration of - only one of them lived in the proposed area. So, they were then able to put their surveillance resources on to that individual, John Duffy, and that gave them the information of where he was moving around, how he was watching and following women. He was seen in the areas where crimes had been committed. When they arrested him they found some very strong forensic evidence that led to a conviction. Duffy had indeed lived at the centre of the area marked out by the three rapes committed in 1982 at the time of those rapes.
What is interesting is that there are two photographs available of Duffy from around the time of these offences. As you can see in Figure 4, in terms of witness statements and descriptions of an offender, you would have really quite different descriptions potentially from different victims. So that there is a sense in which the geographical or spatial information is more indicative and more reliable in this particular case than maybe the sightings and the descriptions by witnesses, which is the normal source of what the police deal with. So it is quite an important and interesting example of the way in which trying to get to grips with the geography of the crimes is giving us some indication of what the characteristics of the individuals are. This has become an important theme, in exploring indicators of criminal behaviour, trying to make more sense of these geographical processes. My new book Mapping Murder explores those ideas in a number of different ways.
Now one of the things about this type of research is that it is potentially very frustrating. You generate some hypotheses, have some ideas about the criminal and what their characteristics are and you have to wait for the police to catch somebody to see whether you may be correct, and then they have to go through the legal process, and even then some people may get it wrong along the way, before you have got any real indication as to whether your own explorations may be appropriate.
It was therefore very exciting to learn that fourteen years after his conviction Duffy was willing to talk about the crimes and indeed to indicate who had been committing the earlier rapes with him. He claimed that David Mulcahy had been involved in the murders as well as the rapes. This came to court a couple of years ago. That then gave us the opportunity, in Duffy’s account, to see whether or not we had been correct in making these inferences from the geographical and spatial pattern about the individual’s behaviour. What Duffy actually said in court could have been taken from the speculation in my book Criminal Shadows, but I don’t think he had read it ( there is always a bit of a risk that offenders play back to you what they think you want to hear). He said “To begin with It was in areas we knew well. We would plan it quite meticulously. We would have balaclavas and knives. We used to call it hunting. We did it as a bit of a joke, a bit of a game. It added to the excitement”. And then he went on to say “You get into the pattern of offending - it is very difficult to stop”. So he is describing the developing process that we had inferred from the geographical pattern.
What is also interesting is that these crimes were referred to as the crimes of “the railway rapist”, because it was noticed that a number of the crimes occurred near railways stations and it was assumed that the offenders were travelling by train and that they had a very detailed knowledge of the railway system. In court Duffy described a situation whereby they had worked out that young women on their own tended to come out of railway stations at certain times of the evening and that therefore going to railway stations would give them that opportunity for finding these victims. But they were actually driving around to do this; they weren’t travelling by train at all. This just shows how important it is from our point of view, from a psychological point of view, to try and get a more detailed understanding of what these processes are that underlie the actions and patterns of behaviour and how they are shaped by the experience of the individual. We cannot only consider the geometry of what is going on.
The publishers of my book Mapping Murder assigned the sub-title The Secrets of Geographical Profile. I am not sure that was such a good idea. It implies that there is some sort of clever geometry or mathematics, to me anyway, seeing that on the cover. So in an early chapter I made clear that the secret of geographical profiling is to go behind the dots on the map and to try and understand the psychological processes involved. This follows a general view I have that where people go to do all sorts of things does tell us something about their lives and their relationships to others. For criminals where they choose to commit their crimes does tell us something about the personal narratives that they are developing.
What I am describing grew out of the process that is known as ‘profiling’, and so in a general introduction to these ideas I am very concerned to get across that what we are talking about are the very early days of a rather complex discipline, in which there are certainly not very many hard and fast rules. The media tends to create the image of the ‘profiler’ as some sort of super hero, some sort of sleuth figure that has great insight that goes beyond the everyday and that every crime the hero tackles is a success.
So I think it is quite useful to mention one example of a case in Alaska, where the murders of five females were linked together to the actions of one unknown serial killer. The FBI, who claimed to have developed the whole procedure of profiling, came up with the opinionthat the individual was single, around 40 years old and not in the armed services. The actual individual, a chap called Bunday, was married with two children, 34 years old and an Air Force instructor. (Laughter) What is important about this is that it slowed down the investigations, because originally this guy was excluded from the computer searches due to this profile being available.
A very similar thing happened in the Washington sniper case, where people, for totally bizarre reasons with absolutely no scientific basis, were quoted in the press as saying that the individuals involved in the sniping were likely to be Caucasian, were not black, were not from ethnic minorities and that it was likely to be one offender on his own. It turned out to be two people, one black and one from an ethnic minority group. They had been stopped a number of times but had not been looked at very closely because they didn’t fit this ‘profile’ that was offered up.
Really what I am trying to get to is that there is a whole ethos, a whole literature, a whole quasi-fictional account of profiling that suggests that there is a lot of expertise, there is a lot of knowledge out there that can be drawn on very readily. The police, in desperation, in many countries, will draw upon whoever is prepared to offer up an opinion without really evaluating the basis of that opinion is. Since the Duffy case I have become increasingly aware of how much general chatter is around, without any real foundation, and the need to develop a proper scientific discipline that can be the basis of contributing to police investigations.
One way of developing the science is to explore the internal representations that offenders have of their surrounding; trying to see how where they go and where they choose to operate is influenced by their knowledge and their understanding of their locality. One way of doing this is to explore what psychologists have called their mental maps. It is not a terribly good term mental maps, it is misleading in a number of ways, but essentially what you do - and I will show you somebody doing this in a minute - what you do is you ask an offender to draw a sketch map of where they commit their crimes and to indicate on it where the crimes occurred and where they were living at the time.
Figure 5 is such a mental map drawn by a burglar. The crosses are where he committed his burglaries - as you see, he is a busy chap - where he lived is marked a ‘home’ towards the middle of the right side of the area. You can see a number of interesting things begin to emerge about his mental representation in this sketch map. For instance, there are very clear boundaries; although there is a canal clearly indicated to the right running down the map, there is nothing indicated on the other side of this canal from his home. The canal acts as a sort of mental barrier that he doesn’t go beyond. He is quite happy to commit crimes near the police station, but going past and coming back is problematic for him. (Laughter) Another outer limit is the DHSS, where he gets his Giro, and clearly he steps out of his familiar area to the bottom once he gets to the roundabout, which seems to act as another boundary point. Besides the clear boundaries is the very interesting issue of the way in which the whole areas is shaped by where he had his base, where his home was.
In Figure 5 is also the actual map of that area. The comparison of the ‘mental map’ with the actual map shows a lot of processes that we find in all mental maps, as I discussed in my book many years ago The Psychology of Place. For instance, we tend to simplify our ideas of crossroads as simple right angles when in fact they are usually much more complicated than that. The gap to the right above his home makes sense because it is a sports ground and there aren’t any houses there. For him it does not have any presence. Further the comparison of the two maps shows how his understanding is very much structured by the dominant road that operates through the middle of the area. So, he has shaped his crimes around what he knows and around that development of the understanding that gives him of what the opportunities for crime are. That understanding is very strongly influenced by where he has his base; where his home is.
Perhaps we can look at the video, and you will see, as we call him, Bob the Burglar, not his real name, going through this process.
(The video was then played)
It is a very interesting interview, not least for hearing the term swag used without any sense of irony. (Laughter)
Of the many matters that the interview illustrates, for our purposes here there are just two points I would draw attention to within that interview. One is that this is an individual who claims he had a very disorganised lifestyle; he was on drugs and moving around. Yet the logic of where he went was still very powerful; he was still very rational. This is one of the points that comes through our research over and over again. What the person is doing may be bizarre in all sorts of ways and the reasons for them doing their crimes, whether it is rape or murder, or burglary or arson, these reasons may be actually very difficult to fathom and may be very complex, but the processes that they are going through still have a very strong logic to them. They are still trying to avoid detection and they are still trying to achieve their criminal objective as effectively as they can. Consequently there is likely to be a pattern to these activities that we can make sense of, if we can understand that logic.
The second point I would draw your attention to is, as I have already mentioned earlier, the significance of the home, the fact that the individual is moving out from where they have that base in order to carry out the activities in locations that are open to them and where they know those crimes are possible. This gives us the basis for starting to develop more involved models of location choice that can be very interesting and indeed very directly helpful to police investigation.
The idea is a very simple one. Just to get the model started, we assume the individual has some sort of base. In fact, if you are a vagrant and moving around it is actually quite difficult to continue a criminal career without either being picked up or really without making yourself vulnerable to detection in a variety of ways. So we find that many of the offenders we are dealing with do have some sort of base. If they have that base, and if the opportunities for crime are evenly distributed around that base, then they are going to move to a location for the first time that is likely to be a distance away from their base that is comfortable, is convenient for them. This location, as Bob mentioned will not be too close to their base with the risk that they may draw attention to themselves when people are looking for who committed the crime, or where they may be recognised. However, they will also not want their crime location to be so far away that they are moving into some sort of alien territory and they don’t know where they are or they may have difficulty getting back. Having been to a location for a crime, that area is now potentially one in which they are vulnerable. But the same processes in terms of distance are likely to operate in other directions, so they’ll move on in other directions to carry out their crimes. Now that new area is one where they may be recognised and where they may be more at risk, so they will then move on in other directions. These simple ideas lead to the build-up of a pattern of crimes, of individuals moving out from a base as illustrated in Figure 6. We have given a name to this pattern of ‘marauder’. It is quite intriguing what potential this pattern has.
Figure 7 is a map that were sent to me by a colleague from New Zealand. It shows the locations where a man was seen prowling. Where is that individual likely to be based; where are they likely to be living? From my previous argument I would suggest that one possibility is that the offender is living within the areas circumscribed by the crimes, possibly quite central to that area. The easiest way of being precise about the area and location is to take the two crimes furthest from each other and to draw a circle with those two crimes as the ends of the diameter. The hypothesis is that the offender is based near the centre of that diameter, at the centre of the circle. This circle is drawn on the sightings in New Zealand in Figure 7. The offender’s home is also indicated and as can be seen it is remarkably close to that central point. This is just one example of many crime series we have explored. As we have shown in published papers, although this result certainly does not occur in all cases it is remarkable how often it does. In burglaries we find it applies in around half of crime series, but for more violent crimes like serial rape and serial murder it seems to be apply to as many as three-quarters of crime series.
This simple idea gives us the starting point for developing more mathematical models about offenders’ crime location selections, admittedly based on a number of assumptions, as I have outlined, but at least giving us a starting framework that we can test out in a variety of contexts.
Having become aware of the surprising prevalence of these geo-behavioural patterns I began to look into a broader literature on the study of people’s movement around an area. I was delighted to discover that there is quite a long history, going back to the 1850s, of looking at patterns of human behaviour and seeing how the geography of those actions reveals some interesting and useable structures.
I am sure in this august setting you are all familiar with the work of John Snow, the father of modern epidemiology? But, for those of you who can’t remember his seminal work, there was a cholera outbreak in the Soho area of London, and in those days there was still a debate, I am told, about whether cholera was airborne or waterborne. Snow’s argument was that it was likely to be waterborne and therefore the pumps from the wells that were used to get water were likely to be the main source of the cholera. To test this idea he did something very elegant and precise. He marked on a map where the cholera cases had occurred and also marking on the map where the various wells were. This showed very clearly that the majority of the cases occur around only one particular well. The story is told that he went to the well and took the handle off the pump over the well so that it could not be used and the cholera epidemic then subsided.
This shows a process parallel to that I have been describing for criminals, whereby people, minimising their effort, will go to the nearest pump. The cumulative effect of this is to mark out the pump that is the source of the disease. There are some rather nice analogies between a criminal moving out and spreading his particular sort of criminal contagion, by similar analogous processes, to the way cholera operates, and on the basis of this we have been able to start developing a research tool that helps us to study patterns of crimes and to draw out interesting results that are also of value for police investigations.
The idea behind our research tool, which I have called Dragnet after the old movie series, can be seen in Figure 8. Here we have a series of crimes, five crimes in this particular instance. What we are trying to do is to see each crime as a source of some process that relates to where the offender is living. Therefore around each crime there is what might be considered a kind of force field. Indeed the mathematics of gravitational fields are very applicable here. As you get further away from any crime then the chances of the offender living there are reduced. By overlapping the fields you get a distribution of probabilities. This gives a more detailed way of calculating where the offender may be based.
To illustrate the application of a Dragnet analysis the five crimes in Figure 8 are located in relation to each other as are the crimes of Jack the Ripper in the Whitechapel of 1888. Figure 9 gives the names of the victims at each location for the five victims that most people agree are likely to have been murdered by the same man. I have superimposed on that map the Dragnet analysis showing the highest probability areas for the base of Jack the Ripper. This points to an area, not too far from Liverpool Street Railway Station, to the West of Whitechapel. Well, it is an intriguing analysis. It shows you how you can use these ideas with existing material.
Of course, we don’t know who Jack the Ripper is, but a quite fascinating document emerged that I rather like, because it is apparently written by a Liverpool cotton merchant. The document, often referred to as The Diary of Jack the Ripper, purports, in effect, to be the confession of the Whitechapel murderer, written by James Maybrick, who bought and sold cotton in Liverpool. ( I rather like the idea that the first serial criminal in modern history is a Scouser, especially as Liverpool has now been designated the European Capital of culture for 2008!).
There would need to be a whole, rather involved lecture around James Maybrick, and the diary that has been published. But in summary, what happened was that about ten years ago a document emerged that professed to be the diary written by James Maybrick. It described his activities as Jack the Ripper recounting the various crimes that he committed and what he got out of them and why he was doing it. This ‘diary’ raises many fascinating issues and I can happily answer questions about it later. But for the moment there is an interesting challenge to what I have been suggesting to you in the whole existence of that diary.
James Maybrick was a Liverpool cotton merchant who lived 200 miles away from Whitechapel. Yet my analysis of the five Whitechapel crimes suggests that this individual was likely to have a base in Whitechapel. What is intriguing about the diary - it is a curious document, it is more or less a rather rambling journal that mixes attempts at doggerel with comments on how cold his hands are with details of the murders, but what is intriguing about it is that it says briefly in it “I took some rooms in Middlesex Street”. It is the only reference in the diary to any specific location in the Whitechapel area, and, would you believe it, Middlesex Street is exactly in that area of Whitechapel that I mentioned, as shown in Figure 9. So it is intriguing that the diary quote fits our Dragnet analysis?
It is one example of a case that will go on forever. One example no more makes a science than a swallow does a Summer, but it provides a taste of what may be possible and is an interesting illustration and raises some significant questions, but really the important thing is to begin to develop the approach further testing it on many other examples, which is what we have been doing, giving rise to many publications in various journals.
However, let me show you one more example of the application of this approach to illustrate further its potential. This is an interesting illustration is in relation to a series of bombs that were left around London in banks and supermarkets about six years ago as part of an extortion campaign. These ‘bombs’ were homemade devices in video cassette cases and he put on the cases “The Mardi Gra Experience”, (annoying misspelling Gras which really confuses spell-checkers). This criminal became known as The Mardi Gra Bomber. These devices were left around London in the locations indicated in Figure 10, which has the Dragnet analysis superimposed. The locations were the devices were left are spread London. But it is clear that there is a cluster in the Western area of London and another in the Eastern area. In other words, the dragnet analysis, intriguingly, draws attention to two foci of activity. I wasn’t involved in this investigation, but the police did use something very similar to this. What they were able to do was to hypothesise that the area to the West was the dominant one for this individual and, and to identify a core of that area where he was drawing his extorted funds from ATM machines. The police were therefore able to put surveillance on credit card machines in that area of the West of London. This paid off because they were able to spot him using a machine and thereby arrest him. So it was helpful in narrowing down the area of attention for the police. It turned out that the man, Edgar Pierce, lived very close to the location indicated in the Dragnet plot. The intriguing thing, of course, is the other focus of his activity in the East of London. Well, it transpired that his ex-wife lived in that area and he used to visit her on occasions and sometimes prepare his bombs there. So there were two bases that he was operating from as indicated in the geo-behavioural analysis.
We are beginning to develop out of this is an exploration of the variety of strategies that offenders use and how these can be modelled in various ways. Some of them we know don’t move out from a base in the same way that the marauders do. They move out to an area where they know crimes are possible and then back again, and we have used the term commuters to describe these individuals. More recently in Mapping Murder I prefer to refer to them simply as travellers’ because there are many strategies involved, not just that of travelling from home to a crime area and back again as implied in the ‘commuting’ analogy. The residential location of these travellers is much more difficult to pinpoint. You can’t use the sort of Dragnet modelling that I have been talking about to identify the base of these individuals.
This therefore raises the question of what proportion of offenders operate in these different ways. I know it is getting a bit late to show you graphs and charts and I realise there are members of the legal profession here who, from my experience, break out into a cold sweat if you give them a percentage, unless it’s a fee for them - no, that’s not fair, a cheap joke. But I just want to show you some science behind this, and there are actually some interesting analyses that emerge.
What we have in Figure 11 - which is derived from work carried out by one of my graduate students, Brent Snook - is the proportion of the Dragnet area that has to be searched before you find the individual. We have three samples here, serial burglars, rapists and killers. Interestingly enough, virtually all of these individuals seem to have a base within the overall Dragnet region, as shown by the fact that nearly 100% of the individuals in the sample are found within the whole Dragnet area. It took me a long time to figure out what was going on there, but it must mean that most serial criminals commit some crimes close to home. If the opportunity is there close to home they will commit the crime there. The graph therefore shows the proportion of the area that needs to be searched, in order to find the cumulative proportion of the sample who are resident in that area. Now, if this graph just crept up steadily I wouldn’t be talking to you, because it would mean that there was nothing of note happening if criminals’ residential locations were just spread out generally across the area available. But what you can see is an elbow in the graph, which means that quite a high proportion, 70% typically, of the offenders are within about 15 to 20% of the area defined by the analysis. Now I am always very struck by this. The same general shape is found for serial burglars, serial rapists and American serial murderers - you have to go to America to get enough serial murderers to actually do the analysis. You can see they all have broadly the same shape to them. To the bottom left of the graph are those individuals who are close to the highest probabilities for the Dragnet calculations, the ‘marauders’. To the top right are those who are rather further from the added highest probabilities and therefore are more peripheral to the area surrounded by the crime locations. These are the ‘commuters’ and other travelling offenders.
Besides the practical value of such findings it strikes me that this does tell us something about criminals, that really in the end criminals are fundamentally lazy. They are seeking easy opportunities that are available to them within easy reach of the base from which they are operating, and although there are some who are travelling further and are much more spread out in terms of the area in which the operate, there is still quite a high proportion who are operating within their particular context
The individuals who are travelling offenders operating over different sort of structures and routes from the marauders are the ones that we need to explore much further, and indeed one of the big research questions at the moment is whether there is anything we can tell from the pattern of the crimes which will allow us to know whether or not the offender is a commuter or a marauder. We are still exploring various possibilities of how we may be able to tell that, and part of that we are doing by exploring the strategies that different offenders use.
Figure 12, for instance, is a mental map drawn by a woman who used to do basically shoplifting to order, collected as part of Karen Shalev’s studies. She would be given a list of things that her neighbours and associates wanted and then she would go off to a Marks & Spencer, or whatever else it was, to steal them. You can see it is a very, very different structure from what I showed you earlier. This is an individual for whom the route to where she could find the opportunities for crime is dominant. She is not really aware of any other possibilities other than where she can get to. She knows the security systems in different supermarkets and different department stores in the area. She knows where she may be recognised, and it is really getting to them and getting back which is the dominant issue, and so the route there becomes the dominant part of her way of thinking about it.
Figure 13 is an individual where also his crimes are construed in terms of the journeys and the opportunities and the motorways that were available to him. He thinks of where he goes to commit crimes in terms of major motorway journeys. He has no ‘mental map’ in the sense of a knowledge of an area related to his crimes. He just has an idea of the dominant routes. Certainly his crimes will be shaped much more by route availability than where he is living.
Also there are individuals for whom the crimes do not have an objective in the same way. Figure 14 is a youngster whom I interviewed. This is what he drew when I asked him to draw where he committed his crimes. He used to steal cars and drive them away, and when asked to draw his mental map of the crimes, what he did was to draw himself in a car, with the skid marks, being chased by the police, very aware of the helicopter view with the car number on it. He just wrote on the sketch “ many cars chasing all over most areas”. It is really the excitement of the chase that is important to him, rather than the particular localities and routes that operate. Following on from this we are doing some research now looking much more closely at the emotions associated with different types of crime and what the actual emotional benefits are that criminals seem to be getting out of their crimes.
So you can see we have moved on quite a bit from the studies in the late ‘80s, where the geography was the dominant process. We are now beginning to look at a variety of strategies that offenders use and trying to get much closer to the way in which these strategies really reflect an internal narrative and account that the offender has about what the nature of his life is, what the nature of their criminality is, and how that is shaped around the locations that they move to. We are considering the way in which the geographical journey reflects a psychological journey and reflects different modes of transacting with people in the criminal context. It is trying to read that process out of the map, as well as the patterns of behaviour that have been involved. That is opening up a whole new area of research, taking us a long, long way away from the early idea of profilers, who were really more concerned with trying to indicate some psychopathology that was producing the criminal activity, rather than trying to understand the logic that drives these crimes and makes it possible for offenders to get away with them for so long.
So I am happy to leave it there and to respond to questions. Thank you very much. (Applause)
The President: Any questions. Who wants to put the first question? Could you give your name and your specialty.
Mr Straw: I am Robert Straw, a barrister. I was intrigued sort of watching, as this is developing, the areas you are looking at. With the sort of areas where it might be a homogenised, fairly standard sort of urban gridlocky town centre and somebody who is on foot, I can see how effectively it shows up your acquisitive burglar feeding a habit. I can see how this is “Well, yes, just draw the dot in the centre”. Is there any way that you can take into account the topography of the area? Bearing in mind that if somebody is travelling by car, for instance, or a van or a truck, or is some sort of odd traditional sort of serial killing lorry driver, can you actually take that into account with the software you are doing, so effectively, because somebody goes there, one might say “Oh, that’s totally off the graph. It might be somebody different. Perhaps he turned back there. That’s a 10-minute journey and that thing was half a mile away from the 10-minute journey”. How do you take….
Professor Canter: That is precisely the development of our research activity. I think the misleading point would be to suggest that the software can do it all. For instance, one of the cases that I use in Mapping Murder that illustrates the issue very clearly is Robert Black, who travelled up and down the country abducting children and then leaving their bodies 80 to 100 miles away from where he had taken them. It turned out that there was nobody else who had ever travelled those sort of distances and disposed of the bodies so far away from where the child was first abducted. It is not hugely surprising that that individual had a job that took him round the country. You mentioned lorry drivers, but there is still the point that there is something that takes them round the country that the crimes follow, and it is very unusual to have individuals who actually shape their geography solely in relation to the opportunity for crime. Without flogging the book too much, there was a very interesting case in Belgium, Detroux, where very consciously and deliberately he would set out on a mission to abduct girls to sell on in various ways and was very deliberately going to the four corners of the country. He admitted as much. So what I am saying is that the scale of the operation and the way it is structured can tell us something about the individual, if we know how to read it. Also we know from studying, for instance, the difference between rural offenders and urban offenders that although the same sort of patterns occur, the scale of the patterns are different, and those graphs I showed with the elbows in them, what we are doing is looking at the proportion of the search area, we are not looking at the size of the area.
Mr Straw: Does that graph apply to an urban area/a rural area somewhere in the
UK/somewhere in the US? Is that a common factor, that 60 to 70 per cent of people will only commit crime…
Professor Canter: There are not that many studies, but studies done in Australia have found that 60 to 70 per cent of rapists have a similar pattern and about 50 per cent of burglars, which is what we found in Britain. It would seem that there is always a proportion of offenders who operate close to home, and in fact when you start looking into the literature there are studies done in the ‘20s in Chicago that give the average distance that burglars are travelling from home as about a mile to a mile and a half. We have got data from Liverpool in the ‘90s that is very similar. So some of these processes are consistent, but I wouldn’t want to give the impression that it is all very simple. There are lots of other things going on. You mentioned topography. Actually, when I submitted a paper exploring this I got my knuckles rapped and was told “In fact, it is land use you are talking about, not topography”. That is clearly an important issue, and in fact this very afternoon I was discussing with somebody how we can get more of a grip on the different patterns of usage of areas and how that is going to influence where criminals travel, and clearly it does. That is certainly the direction the research is moving in.
Mr Samuels: Alec Samuels. One can understand that the criminal may not want to go too far from home and may want to be able to get back to his home, as it were. How far does he have any kind of intimate knowledge of the areas which he is going to in your kind of circular location? Does he know this area, or is it just simply proximity, rather than knowledge?
Professor Canter: Well, that is part of the issue that needs to be explored. There is also the issue and complicating factor, which is how recognisable he is. In other words, for instance, if you look at people knocking on doors and saying that they have come to read the meter and then stealing things, clearly they are very recognisable, because you can’t do that with a balaclava on, so that they are going to be distributed rather differently from an individual who climbs into the back of the house, is never seen and climbs out again. So that does have an influence on that geometry, on those factors of activity. So there are those issues as well. The other thing that complicates matters is evolution over time. As I have hinted, that can be quite important, but it can operate in a number of different ways. Individuals may operate in a particular area and then move on to another area where they operate, so in both of those locations their base is important. But also one of the things that happens (and Duffy illustrates it remarkably well) is that as individuals begin to become more committed to certain sorts of crime their criminal activity takes over, and so they may start off burgling on the way to work or on the way to some sort of casual job or to a pub that they use, but once that becomes something that they feel they get some benefit from, they may then start wandering around an area and start looking out for opportunities. Certainly for serial rapists who attack strangers there is a lot of evidence that they will develop a knowledge of an area, in terms of what the opportunities are for committing that crime, so then they will very deliberately build up an understanding and quite clearly become much more sensitised to the opportunities. A somewhat frivolous example is if you are trying to post a letter you suddenly become aware of all the red objects in the area that may be post boxes. You become much more alert to where the post boxes are. Once you have posted the letter that then fades into the background, your attention is not drawn to that. Well, that is a very natural psychological process whereby our attention is focused in terms of our objectives, and quite clearly criminals become very alert to opportunities.
There is quite a lot of research interviewing burglars about how they decide what house they are going to burgle, and they do become very alert to the cues that the house gives off about, for instance, whether it is occupied or not, or, as our man Bob said, areas where there is going to be something that they can steal, or they become alert to the fact that a place is run down and the individual living there may be old and vulnerable. So they become sensitive to those opportunities, and that may show their knowledge and familiarity, but it starts off from proximity. That is the start of it.
The President: Can I ask you, Professor - the examples you have given us have been about several crimes probably by the same individual - does any of your work assist in cases where there is only one crime?
Professor Canter: Well, the basic process is relevant to any crime, and the very interesting…
The President: Well, you can’t get the sort of round picture.
Professor Canter: But you are still saying that the individual is likely to be local, for example. The very interesting example is the murder of Jill Dando, whereby, because of her public image and the fact that so many people knew about her and the way the whole police process became caught up in a huge media interest, there was a tendency for the police to explore very widely all the possibilities. Quite early on in the investigation a name was offered to them of an individual who lived round the corner. Now using the principles that I am talking about - and I have discussed this directly with the senior investigating officer, whom I have a lot of admiration for and he had a tremendously tricky job, but he admits that this basic idea had never occurred to him in that early stage in the investigation. So when this name Barry George was offered to them early on in the investigation it just wasn’t looked at seriously, and that meant, when after six months of the investigation they realised they were getting nowhere, they were so organised they could then restructure their investigation and say “Well, what have we left out? They then realised they had these individuals, including Barry George, who for a variety of reasons should have been a relevant suspect, and if they had added the extra ingredient that he was actually quite local, they could have looked at him more closely. It was then very late by the time they arrested him. They could have been interviewing him within the first couple of weeks. My point is, whether he was guilty or not, at that stage there may have been more evidence available to support the case one way or the other, but they didn’t get him until very late in the day, because this basic idea that really goes back to the notion of crimes being committed by local residents. As I’ve mentioned there is information from the 1920’s on the study of how criminals travel that fits this general idea. It is not true in all cases. I As I said, Black is an obvious case, Robert Black, who travelled up and down the country abducting children, in which the idea of locality was not the important one. In fact, in Robert Black’s case his van was his home, and his van was the place in which he had all the tools for abducting his victims and binding and controlling these girls, and that is what was moving around. So I must say proximity is a significant issue whether you have a series of cases or just one case. It is just that the more information you have about the criminal the more you have got to work with.
The President: Thank you. I think we have got time for one more question, please.
Dr Josse: Edward Josse, medical practitioner. In 1985 it so happened I examined one of the rape victims of the Duffy duo, so to speak. She was a German au pair and she was very, very badly injured. In my experience of 40 odd years of examining rape victims by and large they have not been very badly physically injured. This girl was very badly injured, and when Duffy split on his partner there were two specimens at the time, I remember, and 15 years later I looked at my original notes, because I had to redo them for the court, and the guy was eventually tagged, so to speak, from DNA. DNA had virtually not started in 1985, but it was available in 1990 when it eventually came to Old Bailey Court No. 1, and of course their series of rapes were also, as you stated, associated with murders. I just dealt with that one case and I didn’t know that any profiling was going on. What I wasn’t entirely clear was did the serious nature of the injuries - because I think the other rape victims were also badly injured - did that help with the profiling and did that give the police a clue, or was it really the shopping of Duffy, after so many years in prison, deciding “I’ve had enough”, and he gave the information to the police?
Professor Canter: Well, the behaviours are very varied across cases and of course there is always the problem, that still exists to some extent, as to which were the cases that were actually linked. I have skated over that, but that is a big issue, if there is no forensic information, as to which cases are linked. Also the point is certainly that in the mid-80s when the statements were being taken the amount of detail that was recorded was often very weak, and certainly in those early cases we were not given any of the forensic medical examiners’ reports. So although we took some account of the varieties of violence, I wouldn’t have said at that stage we would be very alert to the issue that you have raised. Since then we have been able to do a lot more studies on rape and we now have models of the varieties of behaviour that occur in rape, and therefore we can identify what are the rare behaviours and what are the common behaviours. We didn’t know that at that stage. For example, when I first started nobody had any idea about how prevalent anal intercourse was in rape. The police would say “We have got a series of rapes. We are linking them together because there was anal activity in all of these rapes”, and we said “Well, perhaps that is common. How do we know? Where is the database for that?
It is quite interesting, isn’t it, that pretty well in the last two months I have learned that Duffy actually had a history of sadistic violence against his wife. I was never given that information. So what you mention could actually have been more relevant and indeed, although there is also the prejudicial issue, could have been important in terms of certainly narrowing down the whole investigation. It really draws attention to what is really the fundamental problem, and that is really effective databases, really effective information. The information, the knowledge you have, is almost certainly not recorded in any systematic central way that researchers can get access to and people can work with, and that is what we are still up against in many cases. The police have information about crimes, they have information about offenders, but rarely are these two sets of information integrated, so drawing on trends is extremely difficult.
So, as always happens at these gatherings, I learn something more each time, and particularly about that case, because it is amazing how many people have some sort of indirect contact with that case. So that is very, very interesting what you say.
The President: Do come and sit down, Professor. Thank you very much indeed for a fascinating evening. We have all learnt something, I think, this evening we didn’t know before - just what you were saying just now - and we must go away, I think, and read your book, or books, and perhaps we ought to give you something a little extra to go and purchase one of your choice.
Professor Canter: Oh, thank you. I would also like to record my thanks to Bert Raffan for the remarkable skill he has shown in taking detailed notes of my presentation in order to facilitate this publication of my lecture.
The President: Thank you very much. (Applause)