The walk: how offenders read a stranger, and what that does not mean

The camera was set up in an area of New York City with a high rate of assaults. It recorded more than 60 people, chosen at random, walking past. Betty Grayson and Morris Stein then showed the footage to 53 inmates who had been convicted of assaulting strangers, and asked them to rate how easy each pedestrian would be to attack (Grayson & Stein, 1981).

The inmates tended to agree with each other. Age played a part: older people were rated more likely targets than younger ones. When the researchers analysed the footage, the clearest difference lay in how each person moved. Walkers whose movement looked less whole and less organised were rated as easier targets than those who moved in a coordinated, consistent way (Grayson & Stein, 1981).

The study is famous now, often retold as advice: walk with purpose and you will not be chosen. That reading gets the research backwards, and it is worth seeing why.

The reader, not the walker

Grayson and Stein measured what offenders perceive. Their data are judgements made by people who had committed assaults, looking at strangers on a screen. Nobody in the footage was attacked. The study tells us something about how a particular kind of offender reads a street.

Later work followed that thread into personality. Wheeler, Book and Costello showed short clips of people walking to university students who had completed a self-report psychopathy measure. Students with higher scores judged the walkers' vulnerability to victimisation more accurately, measured against the walkers' own reported histories of victimisation (Wheeler et al., 2009).

Book, Costello and Camilleri then ran the same task with 47 inmates of a maximum-security penitentiary in Ontario. Higher scores on the interpersonal and affective factor of the PCL-R were related to more accurate judgements. Those inmates were also more likely to say, unprompted, that they were reading the person's gait (Book et al., 2013).

Put together, the finding is narrow and unsettling. Some people are better than others at reading vulnerability in a stranger's movement, and the ones who are better at it tend to score higher on traits associated with exploiting others. It is a finding about a skill, and about who has it.

Where victimology came from

The study of victims has an uncomfortable history. Hans von Hentig's 1948 book The Criminal and His Victim treated the victim as a participant in the criminal event rather than only its object, and sorted victims into types by the characteristics that, in his view, exposed them to crime (von Hentig, 1948). The insight was real: crimes are interactions, and who is harmed is not random. The framing aged badly. Typologies of victims shade easily into the idea that victims bring harm on themselves.

Modern victimology kept the insight and dropped the blame. The question is not what the victim did wrong. It is how events come together, so they can be prevented.

Three things in one place

Lawrence Cohen and Marcus Felson gave that question its most useful form in 1979. Direct-contact predatory crime, they argued, needs three things to meet in time and space: a likely offender, a suitable target, and the absence of a capable guardian. Remove any one and the offence does not happen. They used the idea to explain a puzzle: US crime rates rose sharply after 1960, during years when incomes and education were improving. Their answer was that daily life had moved away from the household. More homes stood empty in the day, and more people spent more time among strangers, so offenders and targets met more often with nobody watching (Cohen & Felson, 1979).

Seen through routine activity theory, the walking studies are about one of the three elements, the offender's perception of a suitable target. Guardianship, the third element, is where prevention lives. A street with people on it, a shop with staff who look up, a transport system with visible staff: none of these require the potential victim to change anything about themselves.

Once bitten

The other great finding of victimology is concentration. Graham Farrell and Ken Pease, in a 1993 report for the UK Home Office, set out the evidence that being victimised is one of the best predictors of being victimised again, and that repeats often come quickly (Farrell & Pease, 1993). A burgled house is at raised risk of burglary in the weeks that follow. A person harmed once by someone they know is at risk of being harmed again by the same person.

This has a practical consequence that is easy to state. Because harm falls repeatedly on a minority of people and places, protecting the people who have just been victimised prevents a share of crime far out of proportion to the effort. Prevention can follow the victim, not in order to judge them, but because that is where the next offence is most likely to land.

The stranger problem

Stories about predators are mostly stories about strangers: a person picked out of a crowd by someone who has never met them. The walking studies feed that picture, because a stranger on a street is exactly what they test. The crimes people fear most are often not like that. Much serious harm comes from people the victim knows, and public fear tends to run toward the rare stranger instead. Criminologists have traced how fear of strangers can grow into a moral panic about strangers poisoning children without a single confirmed case behind it.

So the research on gait should be held at its true size. It shows that some offenders are skilled readers of strangers, and that the skill tracks with certain traits. It does not show that victims choose to be victims, and it does not show that changing a walk prevents crime. No study in this line followed walkers to see whether training them reduced assaults.

What follows

Victimology at its best keeps two things in view at once. Victims are not random, and patterns in who is harmed are real and worth knowing. Responsibility for an offence belongs to the person who commits it. The patterns are for prevention: for putting guardians where targets and offenders meet, and for protecting people who have already been harmed once. The person walking down the street in 1981 was being read. The research is a study of the reader.

Short answers

Do offenders choose victims by how they walk?

Research suggests offenders make fast judgements of vulnerability from movement. In a 1981 study, inmates convicted of assaulting strangers broadly agreed on which filmed pedestrians looked easy to attack (Grayson & Stein, 1981). Later studies found people higher in psychopathic traits judged vulnerability from gait more accurately (Wheeler et al., 2009; Book et al., 2013).

Does that make victims responsible?

No. The research describes how offenders perceive people. The decision to offend, and the responsibility for it, belongs to the offender. Victimology studies victims to understand events and prevent them, not to assign blame.

What is repeat victimisation?

Being victimised is one of the best predictors of being victimised again, often soon after. Because harm concentrates on a minority of people and places, protecting recent victims prevents a disproportionate share of crime (Farrell & Pease, 1993).

References

1. Book, A., Costello, K., & Camilleri, J. A. (2013). Psychopathy and victim selection: The use of gait as a cue to vulnerability. Journal of Interpersonal Violence, 28(11), 2368–2383. doi.org/10.1177/0886260512475315
2. Cohen, L. E., & Felson, M. (1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44(4), 588–608. doi.org/10.2307/2094589
3. Farrell, G., & Pease, K. (1993). Once bitten, twice bitten: Repeat victimisation and its implications for crime prevention (Crime Prevention Unit Series Paper No. 46). Home Office.
4. Grayson, B., & Stein, M. I. (1981). Attracting assault: Victims' nonverbal cues. Journal of Communication, 31(1), 68–75. doi.org/10.1111/j.1460-2466.1981.tb01206.x
5. von Hentig, H. (1948). The criminal and his victim: Studies in the sociobiology of crime. Yale University Press.
6. Wheeler, S., Book, A., & Costello, K. (2009). Psychopathic traits and perceptions of victim vulnerability. Criminal Justice and Behavior, 36(6), 635–648. doi.org/10.1177/0093854809333958