In explanations of offending, cognition is implicit in the theories of a number of writers who suggest that various styles of thinking, such as ‘impulsive' or ‘concrete', are characteristic of criminal populations (Ausubel 1958; S. Glueck and E. Glueck 1950). The link between cognition and crime is perhaps at its most explicit in the controversial writings of Samuel Yochelson and Stanton Samenow.
In the first volume of The Criminal Personality, Yochelson and Samenow (1976) describe their conclusions drawn from interviews with male offenders referred to their hospital for determination of competency'. From these interviews, they claim to have discovered the ‘criminal thinking patterns’, which characterize all criminals. While cautious social scientists might exercise some restraint in accepting the conclusions of investigators who assert that ‘The criminal has revealed the working of his mind to us', and ‘We fractionate the criminal's mind and then synthesize it', it cannot be denied that Yochelson and Samenow present some fascinating and idiosyncratic views. They describe a number of styles and errors of thinking - Ross and Fabiano (1985) have counted fifty-two - which define the criminal mind. These thinking patterns include concrete thinking, fragmentation, failure to empathize with others, a lack of any perspective of time, irresponsible decision-making, and perceiving themselves as victims.
While there is much of interest in this work, there are a number of criticisms and shortcomings, some methodological, others theoretical. The sample used by Yochelson and Samenow is an unusual one in that it consisted of 240 persistent offenders, many of whom had been judged ‘Not Guilty by Reason of Insanity'. It is questionable just how far it is possible to generalize from this sample to the remainder of the offender population. Their conclusion that all criminals share the same thinking pattern (see also Samenow 1983) must therefore be treated with caution. Further, as they failed to include a non-offender control group, it is not established that the thinking patterns they describe are peculiar to criminal populations.
In summarizing the cognitive approach of Yochelson and Samenow, Ross and Fabiano (1985) express regret at the shortcomings of this work, which has led to a great deal of controversy rather than establishing a valuable contribution to clinical criminology. However, the approach of Yochelson and Samenow is not that of other cognitive models of criminal behaviour. More recent studies have looked for more explicit cognitive variations between offenders and non-offenders. Although the distinction may not be exact, Ross and Fabiano distinguish between impersonal cognition and interpersonal cognition. The former is that aspect of cognition which deals with the physical world; the latter is concerned with understanding ‘people and their actions', sometimes called social cognition. While impersonal cognition may be a factor in the development of criminal behaviour - for example the research which points to intellectual imbalances in offender samples (Andrew 1977) - Ross and Fabiano argue that interpersonal cognition may well be more important in understanding crime. In reviewing studies of social cognition with criminal populations, Ross and Fabiano described a variety of types or styles of cognition, which characterize offenders.
A failure in self-control, leading to impulsivity, figures repeatedly in explanations of crime (Ahlstrom and Havinghurst 1971; S. Glueck and E. Glueck 1950). This failure, they further suggest may be due to one of a number of factors: a failure to learn to stop and think; a failure to learn effective thinking'; a failure to generate alternative responses; or a reflection of hopelessness. Some studies suggest that offenders are more impulsive than non-offenders (Rotenberg and Nachshon 1979), others that offenders are unable to delay reward (Stumphauzer 1972). However, other studies have not found such differences (J.T. Saunders et al. 1973). The difference between studies may be due to differing definitions and measures of impulsivity, and the heterogeneity of the offender population. Uncontrolled episodes of anger may result from impulsivity or a tendency to follow impulses instinctively and without thought for the consequences. It has been suggested that this is a common characteristic of most offending behaviour, i.e. the satisfaction of immediate needs. Impulsivity is strongly associated with psychopathy and anti-social personality (Blackburn, 1993) and can be measured using the Minnesota Multi-phasic Inventory (MMPI), though there are alternative measures which tap into an inability to defer gratification, or carelessness in performing motor tasks.
The concept of locus of control refers to the degree to which individuals perceive their behaviour to be under their own internal control, as opposed to being controlled by external agents such as luck or authority figures (Rotter 1966). A number of studies have shown that offenders tend to external control, that is they explain their behaviour as being controlled by influences beyond their personal control (Beck and Ollendick 1976; Kumchy and Sayer 1980). However, other studies have failed to show any difference in locus of control between offender and non-offender samples (Drasgow et al. 1974; Groh and Goldenberg 1976); while Lefcourt and Ladwig (1965) found offenders to be more internally controlled than non-offenders. The varied findings are probably due to two unfounded assumptions: that locus of control is a unitary concept, and that offenders form a homogeneous population. With regard to the former point, a number of studies have shown that there are several dimensions to locus of control, such as belief in control over one's immediate environment as opposed to belief in control over political events (Mirels 1970). In terms of the latter point, locus of control within an offender population may be a function of race (Griffith et al. 1981); type of offence, for example, violent offenders tend to external control (Hollin and Wheeler 1982); or time spent in prison (Kiessel 1966).
Linked to locus of control explanations of crime is attribution theory, which has become a central component of social cognition. We tend to explain our own and others’ behaviour to ourselves by way of attributing motives. These are based on external factors but are often influenced by other factors too, such as prior assumptions. For instance, the way we assign blame and responsibility tends to be harsher in respect of others than in relation to ourselves. Thus hostile attributions may be made very quickly in a specific context like a pub or club and lead to aggression, while responsibility for this outcome is conveniently avoided by externally allocating intention. For example, I just pushed you because you deliberately knocked my drink (rather than it being an accident). Palmer and Hollin (2000) found that self-reported delinquency in young offenders was associated not only with lower levels of moral reasoning but also with increased tendencies to inaccurate attributions of hostility, especially in ambiguous situations where it may be difficult to accurately ascertain intentions.
If social perception is the ability to discriminate social cues and understand their significance, then the next step requires using these perceptions to decide upon suitable social behaviour. This requires a range of cognitive skills such as the ability to generate a range of solutions to a social problem, considering the various consequences of the different solutions, and planning the steps to achieve the different outcomes (Spivack et al. 1976). These cognitive skills have become more widely known as ‘cognitive problem-solving' over the past few years. Several studies have found that offenders show poorer social problem solving than non-offenders (Freedman et al. 1978; Higgins and Thies 1981; J.J. Platt et al. 1973).
In total the evidence does indicate some differences in social cognition between offenders and non-offenders. However, two points require clarification: the relationship between impersonal and interpersonal cognitive abilities, particularly with regard to offender populations; secondly, and perhaps more importantly, the process by which the cognition results in offending. It seems likely that the answer to this latter point will come from studies, not yet attempted, of social decision-making in offenders.
While social decision-making in offenders has not been researched, decision-making regarding the commission of crime has received a growing amount of attention. This line of enquiry portrays the offender as a rational decision-maker, a ‘reasoning criminal'. The beginnings of this cognitive view of crime are, somewhat ironically, to be found in studies, which focused on, environmental rather than individual factors in explaining the crime. Simply, this view suggests that for a crime to be committed, the opportunity must be afforded for the offence to occur; the criminal is one who decides to take advantage of this opportunity. L.E. Cohen and Felson (1979) elaborated on this view of ‘crime as opportunity', suggesting that for example the increase in burglaries over the past decade can be explained in terms of increased opportunity as more family members go out to work, so leaving houses empty for longer periods. Studies have similarly investigated the effects of opportunity on car theft (Mayhew et al. 1976), violence (Poyner 1981), robbery and theft (Mayhew et al. 1979), and vandalism (Mayhew et al. 1976). The cognitive component involves the decision-making process when the opportunity to offend arises. A crime is seen therefore as the result of the offender's consideration of the risks and costs of offending; indeed previous studies have addressed the relationship between risk perception and likelihood of offending (Stewart and Hemsley 1979).
Cornish and Clarke make the distinction between criminal involvement and criminal events. The former is the process by which individuals choose to become involved in crime, while the latter refers to the short-term, immediate decisions in a given set of circumstances.
As an example Cornish and Clarke work through the process of an individual's initial involvement in the crime of burglary, followed by the event decision for a specific burglary, and finally the process underlying the individual committing more and more such offences. The initial involvement includes factors such as type of upbringing and other social variables; the event decision considers the influence of factors such as the presence of a dog or a burglar alarm on the decision to enter a particular house; while the increase in frequency is related to a growing sense of professionalism' in committing the crime, and accompanying changes in self-evaluation, life-style, and peer group. As well as burglary the rational choice model has been applied to shoplifting (J.Carroll and Weaver 1986), robbery (Feeney 1986), and drug use (T. Bennett 1986). The rational choice model can also be applied to the decision to give up crime (Cusson and Pinsonneault 1986).
In total, the traditional learning theories have concentrated upon the acquisition of criminal behaviour through reinforcement and modelling. The cognitive theories have followed this trend, but have recently considered the alternative of explaining why offending does not occur even when the opportunity arises. This notion of not offending is central to another group of theories which see crime as a failure to learn control behaviour.
Recent research has concentrated on cognitive scripts (Huesmann, 1988). A script is the details of how people should behave in a certain situation and what will happen if they behave that way. These are learnt from the environment in direct experience and from watching others, and from the media. But each script is unique to an individual, yet resistant to change. They become more resistant with use and rehearsal over time. For example, if insulted, a man with an ‘aggressive script' will respond violently. He will justify this behaviour by seeing the insult as aggression, and aggression must be faced by aggression.
During high levels of physiological arousal, people resort to largely unthinking behaviour, and thus well-rehearsed scripts' take over. So to teach non-aggressive scripts' will reduce violence in situations of high arousal (Zillmann (1988))
Along similar lines, Dodge (1986) has argued that much violence comes from Hostile Attributional Bias. Ambiguous actions, like accidentally standing on a person's foot, are interpreted as threatening and must be countered with action.
The source – effective from another socially powerful offender
The message – agreeable information presented first. Immunisation against persuasion – weak arguments against crime easily countered – eg “Yes, you could be caught, but the odds in your favour are 20 to 1, and only mugs get caught”.
The channel – face to face, in a pleasant context
The receiver – recent failure – uses cognitive rehearsal – e.g. “sleep on it”
Primary food, drink, sex
Sensory boredom, seeking new experiences important at the beginning of a career and for person crimes
Monetary important for late in career, property crimes
Social increase in social contacts
Status/power built up from a series of successful crimes
Self-evaluative professional pride.
Skills and resources skilled in physical attack, cracking safes
Opportunity to obtain same objective by legal means, relevant to acquisition stage, those at performance stage combine legitimate and criminal activities
Alcohol/drugs, possession of firearms, factors that increased the likelihood of a criminal act. Override rational thinking
Drugs, need to steal to pay for drugs.
Alcohol, this inhibits behaviour. More confident but less capable. Higher crime but also higher chances of being caught. Also increases helplessness in potential victims. Cohen et al (1956), bus drivers more optimistic about driving buses through small gap but were less successful.
Arousal level, too much arousal, less successful.
Positive outcomes vs. negative outcomes. Criminal money invested in legitimate business, which becomes successful leading to increase in social status. Many crimes not reported nor cleared up so rewarded many times. Reinforcement schedule is intermittent (variable).
Cognitive consequences and distortions. Cognitive activity takes place over time. Some persons spend more time in thought than others, but with the exception of a few fictional Rambos no one is totally involved in action, with no intervals for reflection. Moreover, the fact of cognitive activity means that considerable distortions of objective reality may occur. It is central to much of social psychology that people try to maintain cognitive consistency between their attitudes and their actions, and that they experience a subjective sense of discomfort when there is inconsistency. It is easier to resolve this by changing one’s cognitions than one’s behaviour (Berkowitz 1969).
Bandura (1986) has listed the mechanisms through which the internal control of behaviour is selectively activated or disengaged. The selectivity is aimed at avoiding the subjective discomfort caused by a discrepancy between an action and a previously held belief or attitude, for example, a first or early criminal offence carried out by someone who sees himself as generally law-abiding.
1. Moral justification. This operates on the nature of the behaviour itself. “What is culpable can be made honourable through cognitive restructuring... reprehensible conduct is made personally and socially acceptable by portraying it in the service of moral ends” (Bandura 1986, p. 376). As an example, Bandura points to military training: people who have been taught to deplore killing as immoral can be transformed rapidly into skilled combatants. In the criminological context moral justification is likely to be associated with political crimes.
2. Euphemistic labelling. “Actions can take on very different appearances depending on what they are called. Euphemistic language thus provides a convenient device for masking reprehensible activities” (Bandura 1986, p. 378). Thus, mercenaries speak of “fulfilling a contract” when they mean murder, and the nuclear power industry terms a disastrous explosion an “energetic disassembly”. Euphemistic labelling will be found very widely in both political and corporate crime.
3. Advantageous comparison. A bad act can be made to seem trifling, even benevolent, by contrasting it with something manifestly worse. Stalin justified the destruction of countless Russian smallholders in the 1930’s by pointing to the eventual triumph of socialism for which it would pave the way. Many other acts of physical aggression and much property destruction, particularly in the context of political crimes, exemplify this mechanism.
4. Displacement of responsibility. This operates by obscuring the relationship between actions and the effects they cause. Nazi death camp commandants and their staffs felt little personal responsibility for their enormous crimes; they were simply carrying out orders. There are many examples of offences, often in the political context, for which those responsible made a similar claim, but shortly after World War II the Nuremberg Accords declared that people could not shelter behind this excuse: they had to take personal responsibility for their illegal actions.
5. Disregard or distortion of consequences. “When people choose to pursue activities harmful to others for personal gain they avoid facing the harm they cause or minimize it... especially... when they act alone and cannot easily escape responsibility” (Bandura 1986, p. 381). Offences against persons and property inevitably involve victims in pain and suffering or at least personal loss: examples of cognitive distortion will be given below.
6. Dehumanisation. The more victims are seen as fellow human beings the harder it is to avoid sympathy for their distress. If, instead, they can be dehumanised — by being given a derogatory label such as “nigger” or “whitey” — they are no longer viewed as people with feelings and can more easily be damaged. Both property and person crimes provide many examples of dehumanisation by offenders.
7. Attribution of blame. Offenders seek to exonerate themselves by attributing the blame for their actions to the victim. The most obvious example is that of rape — a claim that in the past was frequently accepted by the courts. It will be found also in other person crimes and to some extent in property crimes.
These mechanisms are not applied with equal intensity and frequency to everyone, but much more to victims or potential victims: friends and family remain for the most part the objects of tender concern. Nor do they immediately transform a person responsive to the distress of others into one able to damage them without apparent remorse. Rather, there is a gradual disinhibition of concern for victims over time, and as an offending career proceeds it is likely that the mechanisms listed above operate in advance of the offence, making them easier to operate and increasing the likelihood of the offence occurring (Feldman 1977).
Some persons, for example, those high on the Eysenckian personality dimension of psychopathy, may find this easier than others right from the beginning of an offending career. And other individual differences are also likely. Feldman and McCulloch (1971) suggested that variations between clients in response to various forms of behaviour therapy could be accounted for in part by learned or innate individual differences in the cognitive rehearsal, outside the treatment situation, of the events which occurred in treatment. Effective rehearsal supplements and strengthens effective learning (Bandura and Jeffrey 1973, Meichenbaum and Goodman 1971). In general, more obsessional and more introverted people may rehearse more frequently and intensely. This notion may be particularly relevant, to the task of explaining particularly violent crimes such as some sexual offences and sadistic murders.
There may also be individual differences in the content of what is rehearsed, with some persons more attentive to pleasant information, others to unpleasant information (Merbaum and Kazaoka 1967). If so, this will help to explain why some people persist in crime, despite repeated lack of success, while others desist following only occasional punishment: the former may filter out failure and rehearse only their successes; the latter may do the opposite. It is tempting to identify these two sets of persons as, respectively, highly extroverted and highly introverted.
Cognitive rehearsal, strengthens learning, obsessional and introverted people are more likely to do this. Extroverts might attend to pleasant information emphasising positive outcomes and filtering out failure suggests Feldman (1993).
Attitude change. Initially unfavourable attitudes towards property crime may be modified by a persuasive message from a high status associate. After the event the attitude would be strengthened if the act was successful (positively rewarded). A variety of attitudes would be built up depending on the particular class of criminal opportunity; based on success and likelihood of success. There would be situations in which an individual would never transgress and those where transgressions would take place frequently.
Observational learning would rapidly increase the acquisition of the necessary skills to perform property crime. The modelling effects are dependent on the status, the interpersonal attraction and the ethnic group membership of the model concerned.
Stimuli associated with successful property offences (for example, chain stores with few staff, open car doors, etc.) will become cues for future theft behaviours. So overtime positively reinforced responses are retained and negatively reinforced ones are dropped.
The offender would choose social settings that result in the acquisition of the necessary skills for property offences. So children growing up within a family that has never committed a crime and attending a school where few children are involved in crime would form strongly held beliefs that crime is wrong and therefore would be unlikely to become involved in crime. The other extreme would be the child being brought up by offending parents in an area where many children are involved in crime and therefore the child is more likely to hold positive views about offending and is therefore more likely to offend.
The performance of offences is likely to be influenced by:
Incentive. Cash, which can be turned into primary reinforcers such as food and sex. An enhanced status amongst offenders. Sensory incentives such as fun or excitement are most likely to be relevant early on in an offending career.
Risk. Property offences have a low clearance rate. Punishments can be mild for property offences compare with offences against persons. The deterrent value of punishment is reduced when the proceeds of a crime can be hidden away and enjoyed upon release.
Target. Winchester & Jackson (1982) found that there were two main factors influencing whether a building was burgled; whether the house was occupied and the siting of the building with regard to easy access: detached houses are more likely to be burgled than middle terraced houses. Apartments guarded by a doorman are unlikely to be burgled (Waller and Okihiro 1978). Walsh (1978) reported that shoplifters are more likely to target self service stores (such as department stores) rather than small shops; American chain stores report that shoplifting, as a percentage of sales, has risen to 7 ½% for fashion accessories and is 2% overall (The Economist 1991).
Skills and resources. Offenders will match themselves to a potential target in terms of the resources required for a successful crime. Weaver and Carroll (1985) recruited self-described expert and novice shoplifters by means of a newspaper ad followed up by a telephone sift. The final pool of 17 experts claimed a median of 100 shoplifting acts overall and at least 10 in the past year; the 17 novices reported zero acts. The interviewees (aged 18—62, with no difference between the two groups) then walked through retail stores with instructions to think aloud concerning criminal opportunities. Half of each group received an additional instruction to form an intention to steal during the shopping trip. The verbal protocols indicated that the experts were much more efficient and “strategic” in their shoplifting considerations, selecting appropriate items by size (larger items were less attractive because they were more difficult to conceal) and working out ways to overcome obstacles such as store personnel and security devices. In contrast, the novices were deterred by almost any difficulty. The selection level for “expertness” seems rather low in this study; 10 thefts in the last year smacks more of an experienced amateur than a professional thief. Nevertheless, the point is made; the more expert the shoplifter, the more rational he/she is in appraising and overcoming obstacles (or in aborting the operation completely — see Walsh, 1978, for another study of experienced shoplifters).
Opportunity for legitimate gain
The concurrent presence of an offending model or models especially if high status.
Reduced self-esteem e.g. failure at school
The victim. The easier the target the more likely the crime, but ‘wrongness’ also plays a part; a lost wallet would more likely be returned to a pensioner than to an affluent businessperson.
Psychopathology. Moore (1984) found that over two-thirds of shoplifters shoplifted for financial benefit rather than for psychological “compensation” (as in the Winona Ryder case, 2001/2).
1. Drugs Substances such as cocaine and heroin have an indirect effect on property crime, as indicated earlier. Those on a run may require immediate cash, which can be obtained only by crime. In desperation foolhardy criminal acts will be attempted.
2. Alcohol. The same lack of exact appraisal is likely to be associated with the ingestion of more than a small amount of alcohol. This has the important effect of impairing the judgment of even skilled performers, so that the wrong target is selected and their performance is impaired —a double hazard. This will be even truer of beginning property offenders.
3. Guns. While the possession of a gun will improve the ability of a skilled and experienced robber to overcome resistance without having to inflict physical damage, in the hands of an inexperienced user — or one affected by alcohol or a drug - it might lead to selecting a too well-defended target, and hence to either total failure, or a serious person crime, which is pursued vigorously by the police.
4. Arousal. This is of particular relevance to person crimes, but a too-high level of arousal will influence the selection and outcome of property crimes also. As indicated earlier, the more complex the decision, or the more difficult the task, the lower the optimal level of arousal. Someone who is easily excited by relatively minor events will frequently be incapable of full rationality. Major emotional upsets which impair the judgment and performance of most people will do so for offenders also. Considerable experience helps people to perform effectively despite exposure to highly arousing events, but it must often be the case that such events cannot be predicated fully and hence guarded against, and may hinder the performance of even the best planned crime.
External reinforcement. Clearance rates tend to fall as the size of property crimes increases (McClintock and Gibson 1961). This suggests that the bigger prizes are sought and won by the more competent offenders. It is clear that both the personal experience of a successful outcome and the observed successful experience of others are widespread.
Even if a crime leads to an eventual punishment the outcome may not be entirely negative, depending both on the actual return (money, property, social status, etc.) and on when the returns are received. Clearly, a $100,000 haul, which is retained, will help to mitigate the pains of a prison sentence; the latter will be experienced as much more negative if no money was retained from the offense.
The powerful influence of frequent rewards in maintaining criminal behavior is illustrated by an interview with Danny, aged 20, who claimed 700 burglaries in a 5 year period, interspersed with two short stretches in a juvenile penal institution and one in prison (Beattie 1982). Danny was “recruited” at the age of 15 by a 16-year-old friend, his first burglary being a huge success — easy access, and a significant sum of money spent quickly and enjoyably. (This is reminiscent of the beginner’s luck of many compulsive gamblers, Dickerson 1984). Danny quickly acquired the relevant skills and moved from general burglaries to a specialized interest in porcelain figures, taken from jewelers, which had a ready market with collectors. He claimed never to have been caught on the job, only through information supplied by police informers, and went on: “In ten years time I’ll either be doing a ten-year stretch or living it up. I’m not going to change my life-style... burglary is the only real skill I’ve got.”
Cognitive consequences and distortions. In the context of offenses against property this means making statements to oneself which derogate the victim (“he’s a mug anyway,” an example of dehumanization), deny the extent of his distress (“he’s covered by insurance” — this is distortion of consequences because he may not be insured and victims suffer in non-monetary ways), or assert that the responsibility for any damage is really the victim’s (“he should have been more careful with his money,” an example of attribution of blame).
For example, Danny (Beattie 1982) didn’t spare much regret for his victims: “Why should I, the people I burgle can afford it, and jewelers are all bent and bump up insurance claims. Another thing, I never burgle poor people or old people.” (It could be said that the greater ease of such targets is more than offset by the poorer returns available.)
There is no space here to include any more of this but Feldman (1993) continues to explain Social learning theory in relation to ‘Offences against the person’ in general and ‘rape’ in particular. Well worth reading.
An egocentric level of cognitive development has been linked not only with a deficit in moral reasoning but also with a lack of empathy, or an inability to sense another’s feelings. Piaget suggested that small children are literally unable to take another’s perspective, and that moving towards a position where this is possible is crucial in terms of more advanced cognitive understanding (moving from the pre-operational stage to the concrete operational stage). Taking someone else’s perspective, however, is not simply a cognitive process since it also involves understanding and relating to their feelings. It is therefore suggested that continuing egocentricism may be linked to a failure to respect other people and develop empathy. Being able to imagine the distress of others can act as a brake on the behaviour most likely to cause this, so it is not surprising to discover that a lack of empathy is significantly associated with psychopathy (Blackburn, 1993). It has proved difficult to develop reliable measures of empathy, however, though findings, which suggest that women are more empathic than men, might present strategies for developing empathy in offenders, as well as explaining why fewer women are involved in crime. For instance, anger is said to disrupt empathy (Yates et al. 1984) and if the display of aggression is an integral part of male socialisation then anger management training might be a useful first step towards developing empathy.
Most of these cognitive explanations of crime assume that dysfunctional thought patterns produce inappropriate behaviour, and therefore challenging these patterns and replacing them with more flexible thinking will reduce offending. Cognitive restructuring now forms a major part of most contemporary treatment programmes.
Kevin Brewer (2000), Psychology and Crime, Heinemann, ISBN 0-435-80653-X.
Philip Feldman (1993), The Psychology of Crime, Cambridge, ISBN 0-521-33732-1
Julie Harrower, (2001) Psychology in Practice, Crime, Pages 33-37, Hodder & Stoughton, ISBN 0-340-84497-3.
Clive R. Hollin (1989), Psychology and Crime, Routledge, ISBN 0-415-01807-2.
Fiona Lintern, Merv Stapleton & Lynne Williams (2004), Study Guide for OCR Psychology: A2 Level, Hodder & Stoughton, ISBN 0-340-81626-0