Narrative Theory Applied to the Autobiographies of Three Life-Course Offenders

Abstract

The written autobiographical accounts of three life-course offenders, which were published following extended life-course offending, were analyzed using a theory-led thematic analysis. The protagonists were each responsible for a broad range of acquisitive and violent crimes, although different offense types are often studied as separate entities. The utility of narrative theory was explored as a life-course theory by contrasting its framework with these disparate areas of inquiry, along the developmental trajectory of the protagonist’s account. Findings showed that onset began with trait-driven and versatile offending, which progressed toward specialization, incorporating modus operandi and rational choice making. Specialization was underpinned by themes of violence and control. The concluding themes dealt with the process of desistance, which was facilitated by a series of cognitive shifts, allowing these offenders to retain a core element of the “self.” Results further showed that narrative-identity played an influential role in the development of specialization and eventual desistance, but less so with onset. Narrative theory has the potential to aid understanding of the criminal life-course trajectory, which in turn can assist in both detection and rehabilitation processes.

Narrative theory has the potential to provide an explanation for criminality over the extended careers of what Moffitt (1993) called “life-course” offenders, where onset begins in early adolescence and continues into late adulthood. As life-course offenders are by definition repeat offenders (Halliday, 2001), many also go through periods of “versatility” before “specialising” in particular offense types (Farrington, Synder, & Finnegan, 1988; Klein, 1984). Versatility in this context means different offense types, such as dishonesty, vandalism, and violence, and can be clustered around certain age groups (Soothill, Francis, & Fligelstone, 2002). Complexity arises because these different offense types are typically studied independently and also underpinned by different models such as “trait-driven” or “dispositions” toward aggression and impulsivity, or rational choice making models. Collectively, the field represents a number of disparate areas of inquiry.

A theoretical framework that assists in linking these disparate areas may have potential for developing an increased understanding of life-course offenders and aid in the development of detection, prevention, and desistance work. Life-course offenders are viewed as being responsible for the largest portion of recorded offenses (Halliday, 2001; Tarling, 1993). Moreover, Canter (1994) and Cornish (1994b) observed that criminals’ accounts of crime commission take a narrative form; so, taken together, there is utility in exploring narrative theory for this genre of offender.

Narratives are about bringing order and coherence to a protagonist’s life so that both sense and meaning can be made of it (McAdams, 1996). Working with notions of coherence and meaning can shed light on what guides life-course offenders to make decisions about the type of offenses that they engage with and how they manage their way out of the offending cycle.

Maruna (2001) applied narrative scripts to explore both desistance and persistence in offending. Although Maruna’sfindings are yet to be “replicated” (Presser, 2009), lack important definitions to key terms such as “script” and the epistemological orientation of the approach (Hockey, 2014), as well as its limitations in terms of why desistance occurs (Healy, 2012), the principal findings were that desisters found new “redemption” scripts. These new scripts enabled desistance through a remodeling of the “self” while retaining some coherence to the past. This allowed these ex-offenders to make sense of that past and move forward with new ways of being, including core ideas and believes from the old “self.” More recently, Youngs and Canter (2012b) argued that narrative theory when considered at the point of offending may also have the potential to link identity, emotion, and cognitive processing. They devised categories that drew on narrative theory to explain narrative themes, which influence the offender at the point of the offense. This approach builds on what Presser (2009) saw as stories being situated in the antecedents of crime.

However, Ward (2012), and Ward and Marshall (2007) see stories as metaphors rather than individuals creating fictional personas which they set out to become. Narratives contain identities, a plot, actors, and a context. This also links the past with future events, together with an individual’s goals and intentions. Two models of narrative theory briefly reviewed here suggest that the framework has potential and utility for criminogenic applications.

The M. M. Gergen and Gergen (1984) model consists of the following features. A “goal state” or “valued end” must be established to succeed as a narrative. Narratives are structured in such a way as to generate a connectedness and coherence which creates a sense of movement or direction through time. Events must be selected in such a way as the goal state is rendered more or less probable. As one moves from one event to another, one approaches or moves from the desired goal state, hence a sense of directionality. There are three forms to this: (a) progress toward the goal is enhanced, which the protagonist is developing or building; (b) in the regressive narrative, things get worse over time; and (c) in the stability narrative, the protagonist has incidents and images linked in such a way that things remain the same. The theoretical framework is concerned with meanings assigned to movements, rather than the reporting of specific concrete facts or the accuracy of the relationship between facts. The arrangement of facts is only important in relation to the goal state. The meanings within autobiographies are not exclusively private but are limited at the outset to a vocabulary of action that possesses currency within the culture.

In contrast, the McAdams (1996) model, while consisting of some overlapping features (e.g., less about facts and more about meanings), differs in that it composes an heroic narrative of the self that illustrates essential truths about the self. The heroic myth exists inside the mind and is embedded in the complicated series of accounts. It is the central story from the various episodes told. The model consists of a narrative mode (concerned with human wants, needs, and goals) as opposed to a paradigmatic mode (scientific thinking). A good autobiography puts a life into story form complete with setting, characters, recurring themes, images, and the self-conscious reconstruction of human time through narrative. Initially, elementary children think in concrete terms where they learn about rules and principles. During late adolescence, they become narrators and self-conscious myth makers, before that there is no life-story, no identity. There are four types of mythic forms: Comedy and Romance (Optimistic tone); Tragedy and Irony (Pessimistic tone). Characters now consist of motivation, which older children begin to comprehend. These motives exist within personality which derives from the internal dispositions. Theories of motivation, which are held to help organize behaviors, consist of two phases: first strivings: power, autonomy, independence, status, experiences, rich with emotional excitement; second strivings: love, intimacy, interdependence, acceptance, interpersonal experiences subsumed with the experiences of the emotion joy. At this point, children begin to see human behavior as purposeful and organized. Modality evolves into agency and communion, which organize a great many human wants (e.g., needs, desires, and goals). Agency represents a striving to separate from others and breaks down into two subsections. People high in power motivation tend toward a preference for experiences of feeling strong and dominating the environment and are likely to take risks. People high in achievement motivation tend toward a preference for experiences of competence. Communion represents a desire to lose individuality and merge with others. This in turn breaks down into the two subsections of intimacy and love. Those with overlapping desires for intimacy and love are able to relate to others in warm, close, and supporting ways. High levels of intimacy motivation are revealed in the ways they speak with friends and in social groups.

Although not every aspect of the two theories are applicable to the criminogenic environment, there remains a sufficient framework that can be applied. Although in taking-up Presser’s (2009) call to further explore the potential of narrative criminology, the study here does not advance a particular approach. It does not follow Presser’s “constitutive” approach, which epistemologically privileges language as a linguistic devise for minimizing agency and sociologically derived notions of the self, from those made available culturally. However, it does draw from McAdams (2012), in that this study is in the context of exploration of psychological theory.

The utilization of autobiographies within research is typically derived from in-depth interviews, which are subject to impression management devices and issues with accuracy (Snook, Dhami, & Kavanagh, 2011), hence the unease that some criminologists have in accepting their validity (Presser, 2009). However, where the purpose is to explore the construction of the narrative, the concern for precise detail and impression management is negated because the narrative is about how a protagonist makes sense of their life and how they cast themselves (Presser, 2009). The events that a protagonist draws from to create their story are the construction of the narrative. Within that, the narrative-identity is the protagonist’s attempt to produce their image around their behavior. Furthermore, a protagonist’s account can be compared with other sources, such as official data, criminal records, and empirical literature. If the account conforms to these sources, it is more likely to derive from actual experiences. In an exploratory approach such as this, starting with accounts that are grounded in this way links a protagonist’s narrative construction to what is already known about life-course offenders, independent of the narrative framework.

One way in which to consider whether narrative theory has utility for understanding life-course offenders is to explore the written autobiographies of ex-offenders and map these disparate areas of inquiry onto the accounts, while contrasting the results with the narrative framework.

As one purpose of narrative theory is to identify how the protagonist renders the events into a story, including making sense of disruptions (Murray, 2003), the written autobiography provides a rich source of descriptive data relating to what the protagonist sees as important to their story, including its presentation. The accounts used here contain post-event rationalizations and reflections upon offending activities, some of which may not always be encapsulated by interview methods.

Method

Participants/Materials

The written autobiographies of three United Kingdom, male ex-offenders were analyzed (Boyle, 1995; Foreman, 2008; Smith, 2004). These protagonists were chosen because they were a good fit to Moffitt’s (1993) life-course offenders; Halliday’s (2001) repeat offenders, and Farrington et al.’s (1988) and Klein’s (1984) versatile offenders. Jimmy Boyle was a gangland enforcer in Glasgow during the 1960s. He was eventually given a life sentence for murder, during which time, and in his 30s, wrote his autobiography. Freddie Foreman also became a gangland enforcer in London during the Krays era and beyond. He received a number of long prison sentences, though not a life sentence. His autobiography was written during his late 70s, some years after his release. Noel “Razor” Smith, also from London, eventually received a life sentence resulting from repeated armed robbery convictions and prior sentences from the 1970s through to the late 1990s. Smith wrote his autobiography during his early 40s while serving his life sentence. All three offenders originated from underprivileged backgrounds, were frequently in trouble at school, began offending during their early teenage years, served a number of prison sentences throughout their criminal lives, and eventually desisted during the mid or later stages of life.

Procedure/Analysis

A three-stage exploration of the data was conducted. First, all offense-related material was collated. This included pre and post rationalizations, descriptions, and reflections. The data were grouped and coded according to the following broad themes: trait-driven, rational choice, versatility/specialization, Modus Operandi (MO), and desistance. These were later refined by the analysis (see results below). As the purpose of the study was to compare the trajectory of life-course offenders in the context of these competing theoretical orientations and a general narrative framework, the theory-led approach (Hayes, 2000) was utilized. This is incidentally compatible with both realist and relativist epistemologies (Braun & Clarke, 2006) as well as several methods of analysis (Willig, 2003). This approach seeks to identify existing theories that explain the data rather than identifying only data that support a priori theory or devising new theories from the data. Arguably, alternative theories may have been compatible with narrative theory or with some of the data groupings, but that would have defeated the purpose of the study, which was to explore the capacity of the narrative framework to handle the complexities of competing theories linked to versatility and specialization.

This approach had two purposes: first, it offered a measure of validity to the autobiographies by comparing the accounts with what is known about the offending patterns of some long-term offenders, while alerting us to how some offenders actually transverse a number of theoretically conflicting areas of study. Second, it facilitated organizing the groupings into temporal stages to discern the relationship between narrative theory and the temporal stages of the criminal life course. In practice, these two stages were integrated. The third stage of the analysis explored the meaning of the narrative for the protagonist(s). Here, existing theory was not used to guide the analysis. Meaning was explored through the emplotment and depending on the epistemological approach taken, meaning can be attributed through privileging language as in Presser’s (2009) “constitutive,” post-positivist position or within a realist epistemology (May, 1996) as taken here. Nested within Presser’s (2009) constitutive approach, Youngs and Canter (2012a, 2012b), and Canter and Youngs (2012) stress the importance of the “here and now” of offense decisions. This in turn results in an appeal to a culturally available narrative-identity, played out by the offender at the point of offending. A realist epistemology does not reject the notion that offenders make “here and now” decisions. Rather, it emphasizes enduring dispositions and constructs such as cognition. The purpose of a realist approach is to explain, while a post-positivist is typically to describe. The McAdams approach, which emphasizes internal constructs, falls within the realist epistemology, while culturally available resources may become the vehicle which a protagonist subsequently adopts or aspires toward.

Within realism, language is not privileged in its own right, rather extracts are treated as items of data (Hockey, 2014).

Results

These first five themes deriving from the first two stages of the analysis relate to different chronological and spatial periods in the trajectories of these life-course offenders. The third stage yielded the narrative-identity, emplotment, and the resulting story.

Theme 1: Trait-Driven Behaviors

This theme is about cyclical behaviors that dominate the lives of those individuals who neither understand these drives nor are able to control them, despite the deeply destructive consequences that arise from them and was particularly relevant during the early adolescent period.

Research has shown that many young offenders have traits that hinder normal development and socialization (Crick & Dodge, 1994, 1996) which are stable across the life course (Moffitt, Caspi, Harrington, & Milne, 2002). These traits typically manifest as “aggression” (Huesmann, 1994; Huesmann & Eron, 1989; Huesmann, Lefkowitz, Eron, & Walder, 1984) or impulsivity (Ross & Fabiano, 1985).

Huesmann et al. (1984) administered a variety of procedures (e.g., Minnesota Multiphasic Personality Inventory Scales [MMPIS] and peer nominated aggression score) to 870 participants from a semi-rural New York state county, whose modal age was 8 years at the time these initial measures were taken. They also interviewed 75% of the parents. At age 19 years, 211 boys from the original cohort were re-interviewed, of which, 198 males were again re-interviewed at aged 30 years, along with 165 spouses. Furthermore, data on 542 of the original cohort (some of whom had been re-interviewed) was obtained from the New York state criminal justice system. Results revealed that where high levels of aggression were shown in 8 year olds, aggression remained stable at 19 and 30 years of age, including acts of criminality (e.g., r = .30, p < .001, n = 190, MMPIS; and r = .29, p < .001, n = 193, self-report). Impulsivity is viewed as an absence of information processing between impulse and action (Ross & Fabiano, 1985), exhibiting a lack of self-control (Gottfredson & Hirschi, 1990) resulting from social skills deficits (Freedman, Rosenthal, Donahoe, Schlundt, & McFall, 1978).

Boyle aged 13 years experienced his first period of custody, “The thought of returning to the remand home was worrying me very much as the few hours I’d spent there had been enough and I only wanted to go home and I’d never get in trouble again” (p. 39).

Although he may have meant it at the time, it did not stop him from reoffending and soon found himself experiencing the same anxiety:

I was overjoyed when I was out and sincerely meant it when I told my Ma that I would never get in trouble again. I told her this again and again, and with my whole heart and soul this was what I meant. Why when feeling like this did I return time and time again? I never wanted to get into trouble and this feeling was passionately strong. (p. 48)

Smith wrote, “At the age of fourteen I was a criminal in the making. I had very little fear of anything, I was willing to take on dangerous gambles on little more than a whim, and I hated the Police” (p. 44).

Foreman reflecting on his late teenage years of violence wrote about his lack of forethought and loss of self-control:

Although young men at the time we never considered the implications, if someone died as a result of a fight you could very easily find yourself facing a murder charge. In spite of that, you continued to risk life and liberty by doing battle with others.

And “I went over the top on a number of occasions and I would put that down to the madness of youth,” and “As young men we had no fear. Nobody wanted to lose their freedom or their life, but there was literally nothing that could frighten us” (p. 12).

The responses to prevailing events were spontaneous, and although they may have regretted the consequences later, it did not prevent repeat behaviors. In other words, they were repeatedly impulsive and aggressive, they could not control their urges (Gottfredson & Hirschi, 1990). Traits can dominate the adolescence of an individual, particularly if there are insufficient interventions during childhood. For Huesmann and Eron (1989), it is the innate factors that produce the cognitive structures that predispose the child to habitual aggression. These internal dispositions (McAdams, 1996) are the innate aspects of an individual’s character and are observable in 2- to 3-year-olds. For example, studies on twins have shown that antisocial behavior maybe more deeply rooted in the genes than prosocial traits (Krueger, Hicks, & McGue, 2001). Here, 170 pairs of monozygotic and 106 pairs of dizygotic twins completed a 198-item multidimensional personality questionnaire and an antisocial self-report questionnaire. Individual differences for negative emotionality and a lack of constraint were highly correlated with antisocial behavior for those in non-shared environments. Of the different models tested (i.e., genetic variance with shared environment [AC], and genetic variance with non-shared environment [AE]), it was found that antisocial behavior was linked to AE, χ² = 1.79, p < .05. What appears to influence onset in life-course offenders (Moffitt et al., 2002) are a number of dysfunctional traits that have been observed to endure across the life course (Huesmann & Guerra, 1997), particularly in the absence of intervention or self-rehabilitation. The presence of these traits may explain the difference between children within the same families or social backgrounds, experiencing similar social risk factors who offend while others do not (see Farrington, 1996, for a review of multilevel risk factors).

Smith reflecting on his adolescence wrote, “though I was slowly learning to suppress my own natural instincts, I was not yet fully mature” (p. 193), while Foreman wrote, “Looking back I have learned to understand some of the forces that dictated my actions” (p. 12). They were often confused by the cyclical process, suggesting that they lacked the capacity to be able to appreciate or control these drives. Furthermore, during their adult-offending years, they each made references to committing random offenses impulsively and out of desperation, despite long periods of offending as self-proclaimed professionals.

An alternative view is Presser’s (2009) “Constitutive” approach in that cultural identities influence a protagonists narrative-identity and role. However, given that onset was in early adolescence and prior to identity formation (McAdams, 1996), the question for the Constitutive approach here would be “how can they choose a cultural influence.” That is not to deny an interaction with cultural influences, but with early teenage offending being pre narrative-identity and the weight of evidence relating to internal dispositions, it appears more logical that dispositions, not cultural influences, predominantly drive early offending. As later stages in the analysis below suggest, cultural influences may play an increasing role.

Theme 2: Opportunity and Versatility

This theme is concerned with how individuals seek-out or take up new opportunities as they are presented. At this stage, there is no commitment to a specific set of criminal offenses but there is an underlying consistency to the process of criminality.

Two strands of the research field, “opportunity,” underpinned by Cornish and Clarke’s (1986) Rational Choice theory, in conjunction with Cohen and Felson’s (1979) Routine Activities theory (RAT), and “versatility” (Farrington et al., 1988; Klein, 1984), appear to converge during this period of offending. Rational Choice theory is concerned with the principle of “least effort” and the calculation of minimizing the risk of failure against possible gains. This operates in conjunction with RAT, in that offenders seek-out the necessary role of “opportunities” (Clarke, 2010) within the environment typically habituated.

At 12 years old, Boyle would steal from shops on the way home from school, “We found it exciting and soon we were deliberately stealing, going into shops on the way home to take whatever we could” (p. 27). Reflecting on a new opportunity, Boyle aged 13 years recalls the time when chewing-gum machines were being installed to shop fronts all over the district, “These machines made a fortune as everyone was using them, but their accessibility was very tempting to us who roamed the street” (p. 35). Boyle was eventually caught and given another period of custody. By the time he was released, the chewing-gum machine opportunity had gone. However, Boyle was soon stealing coal to sell, “Eventually we gave-up trying to get the coal legitimately and decided to steal it” (p. 44).

As a teenager, Smith was still finding his way in the criminal world and had not settled on a particular offense type in which to earn money, “we would drive around looking for things to nick and bring back to sell” (p. 243). Although he continued to do random car-related offenses, when an opportunity came along to cash stolen giro checks, he took advantage of it, “The Giro firm were always looking for a new face to take the checks into the post offices, and I did a bit of that for a while” (p. 255).

At aged 18 years, Foreman joined a gang of girl shoplifters, “I teamed up with a load of girls known as the Forty Thieves” (p. 41). After that was over, Foreman continued to take up any viable opportunity that came along, “George and I were always open to fresh opportunities, like the one that presented itself at George’s own workplace” (p. 67).

These examples reveal that throughout their adolescence, each of them moved from one money making scheme to another as new opportunities became available, while previous schemes ceased to exist or remain viable (Crawford & Evans, 2012).

Adolescent versatility appears to be influenced by the opportunities that are available at a given moment. An instrumental element of versatility then is opportunity. Opportunity, in part, derives from an offender’s environmental constraints (RAT). Suggestively, an impoverished calculation of risks versus benefits, albeit impulsively assessed, informs the decision making of offending.

This period of versatility, while not exclusively related to the adolescent years is certainly dominant during this stage and appears to precede specialization. A pre-requisite to offense specialization is the evolvement of “offence scripts” (Cornish, 1994a).

Theme 3: Habit, Routine, and Offense Scripts

This theme is characterized by the procedural knowledge of repeated sequences and refining the procedures through experience. Offenders develop a system or set of procedures which facilitate the smooth running of a criminal activity. Through direct experience or vicarious means (Bandura, 1989), they learn how to build techniques into the routines, which increase their prospects of committing offenses (Cornish & Clarke, 1987).

Script theory (Schank & Abelson, 1977) is considered to be a mode of processing information at a basic level. The main characteristics of the theory are that there is a routine which develops from rehearsal or habit. The routine is designed to achieve a concrete goal such as eating in a restaurant. As the script holder becomes more familiar with the routine, less cognitive processing is required and so it becomes more automated. The notion of scripts is further elaborated by Cornish (1994b) who viewed them as a particular form of schemata. Schemata’s are mental constructs that organize information about something (e.g., a restaurant). Scripts are the “dynamic” form of the schemata because they move the script holder (e.g., a customer) through the scenario. A number of studies regarding the cognitive processing of criminal behavior across a range of different offense types have compared the basic principles of the theory with the processing styles of habitual offenders (Cornish, 1994a, 1994b; Cornish & Clarke, 1987; Hammond & Brown, 2005; Hockey & Honey, 2013; Huesmann, 1994; Huesmann & Eron, 1989; Michael, Hull, & Zahm, 2001; Nee & Taylor, 2000; Tunnell, 1992; Ward & Hudson, 2000; Wright & Decker, 1994). Furthermore, script-style processing directs attention toward behaviors consistent with it (Ceci, Fitneva, Aydin, & Chernyak, 2011), hence the cyclical process. In the context of offenders’ routine lifestyles (Brantingham & Brantingham, 1984; Cohen & Felson, 1979; Wiles & Costello, 2000), habit appears to develop in two ways: first, the utilization of the routine “into” a typical activity and, second, the mode of processing as a minimalist function, perhaps as a heuristic device (e.g., Cornish & Clarke’s, 1986, principal of least effort) at a more general level.

At age 13 years, Boyle was learning about methods for committing burglary, “One of these methods was known as ‘grafting’ which was finding a shop with an empty house next to it or above it and making a hole in the roof or through the floor and going in that way” (p. 52). During these burglaries, Boyle had acquired a series of techniques designed to enable the burglary and reduce the risk of immediate apprehension.

Smith had developed into a bank robber during his late 20s and had acquired a smooth, efficient routine for committing the offense. His gang targeted banks at the time when security was at its most vulnerable and the potential gains were at their highest, “For us the hard work was in finding out which day the reserve was expected and this was mainly done by picking a good-sized bank and putting it under observation for a while” (p. 11).

At age 24 years, Forman had developed into a prolific burglar of commercial premises and had a series of techniques built into his routine designed to reduce the risk of getting caught, “We would return to the store and walk around inside the shop, trying to set off hidden alarms” (p. 74). They would then hide for a while to see if the police turned up before going back inside the building and carry on with the burglary.

Nee and Meenaghan (2006) found that experienced burglars displayed the characteristics of expertise as they talked their way through the decision making of a hypothetical burglary scenario. Topalli (2005) also found that the decision making of violent offenders equated to specialization when compared with non-offenders. This “here and now” crime scene efficiency is often confused with intelligence, which has an unclear role in its contribution to crime (Blackburn, 2001) and is an over inflated perception of the rational choice making process. For example, Cornish and Clarke’s (1987) cost benefits analysis results from the rehearsal of simplified routines, which focus on immediate or short-term goals. An overlooked but central component is that these offense scripts are “here and now” (Ross & Ross, 1995) goal-tied problem solving mechanisms, similar to Piaget’s (1962) concrete operational thinking stage. Just as the child develops more elaborate detail in the games being played through daily experience, the habitual offender develops more elaborate detail to the crime scene behavior through similarly limited processing. It is this observed temporal development that creates the illusion between the here and now thinking of the trait-driven behaviors and the perceived efficacy or agency underpinning the misunderstood rational choice model, hence Cornish and Clarke emphasis a “bounded rationality.” The realization of what one has done to one’s own life as well as to others does not occur until later (see the desistance and emplotment themes below). Moreover, Hockey and Honey (2013), and Hockey (2016) found that a group of apprehended offender primarily focused on how to commit a burglary, while the non-apprehended offender group primarily focused on how to avoid detection. The processing capacity of the apprehended offenders at the crime scene remained within the parameters of concrete processing (Ross & Ross, 1995; also see Farrington, 1996, who noted that habitual offenders tended to do less well on abstract tasks than non-offenders, while performing well on concrete tasks). Indeed, Canter and Youngs (2010) described offenders as “limited” people, while Morgan (2002) provided a breakdown of the dysfunctional characteristics of the prison population. This suggests that agency is bounded and is more about how to commit an offense or “how to construct a narrative-identity” (Ward & Marshall, 2007) than the offender having the abstract capacity to see beyond the “here and now” of offending.

Theme 4: Specialization, MO, and Narration

This theme is characterized by choice making in terms of offense type and how it is committed. In this context, offense type refers to the “specialization” of an offender to focus on particular crimes deriving from early versatility. However, research has indicated that specialization is more about criminogenic themes (e.g., Donald & Wilson, 2000; Soothill, Francis & Fligelstone, 2002; Yokota & Canter, 2004) rather than Klein’s (1984) specific offense types (Youngs, 2004). Another strand to this line of investigation has been pursued through studying the MO, which is effectively the offender’s idiosyncratic style embedded within the general sequence for committing a crime.

There has been long-standing support for the notion that offenders have an MO, which provides the basis of how the given offender commits crimes (e.g., Alison & Eyre, 2009; Homant & Kennedy, 2006). Canter and Youngs (2010) described the MO as a pattern of behavior that is rather more specific than the general style of an offense. An MO is thought to derive from a schema that an offender holds about the way in which a given crime type unfolds (Jackson & Bekerian, 1997). Therefore, a particular offender’s MO will hold idiosyncratic detail which is embedded within the general offense script.

However, within the context of a narrative framework, what is of interest is whether there is an interaction between offense type or specialization, MO, and narrative-identity. This includes the manifestation of a narrative-role, which is not limited to cognition as in the offense-script framework. As Youngs and Canter (2012b), and Canter and Youngs (2012) argued, narration begins to combine emotion, identity, and cognition in the decision making of the criminal act. In this respect, if an offender holds a narrative identify, decisions may derive from the offender’s perception of which behaviors are appropriate or desirable to adopt in a criminogenic situation, hence a narrative-role. Presser’s (2009) notion of “narrative influencing lines of action” may become salient at a given moment resulting in a behavior selection (e.g., Ward’s, 2012, “narrative product”). Narrative-identity broadly equates to an offender’s self-concept or self-identity (Canter & Youngs, 2010) but with a dynamic output within a social context (K. J. Gergen & Gergen, 1988).

Youngs and Canter (2012b) developed a four-category typology of the central themes that they found to be important in offenders’ immediate criminal actions. Building on themes (i.e., Professional, Hero, Victim, and Revenger) derived from the Canter, Kaouri, and Ioannou (2003) study, Youngs and Canter (2012a) administered their “Narrative Roles Questionnaire” to 71 offenders with an age range from 21 to 61 years old and a mean of 34.5, incarcerated for a mixture of offense types. The results showed separate self-awareness on the identity-focused items. For example, while some offenders saw themselves as being weak (e.g., I was helpless; Victim theme), others saw themselves as being strong (e.g., I was like a professional; Professional theme). Within these categories, the “professional,” as with Merton’s (1957) innovator, appears to be the closest fit to the protagonists here. The “Professional” category denotes the offender as someone who is doing a job. They are qualified and specialized. They see themselves as highly skilled, intelligent, and competent.

Furthermore, their choice of offense type was related to their perceived capabilities, which included comfort with “confrontation” over what Walsh (1986) described as “people avoiders” (i.e., typically burglars are people avoiders, while shoplifters, robbers, and pick-pockets are people acceptors). Similar groupings have been further developed (Youngs, 2004), and accordingly, the protagonists here would additionally fall into the “expressive” category where the offender uses violence as the style of interaction to exert control.

Boyle’s principal criminal activities involved an element of providing security to protect his moneylending and alcohol selling black-market business, as well being a gangland fighter. “We were now the top mob so with a little luck we would expand and develop things and make plenty of money,” and “We were going from strength to strength and the financial side was blooming. Moneylending was a lucrative business” (p. 140).

As a bank robber, “Smith” would have needed to be prepared to take people on in a direct confrontation. The central component to his offending relied upon the threat of or if necessary the actual engagement with violence to obtain money, “Both me and Andy were doers, up-front men who enjoyed being in the thick of it” (p. 440).

Foreman’s post-adolescent criminal days also relied on a physical presence,

Whenever we did a bank or vault job, I never did the burning open that was left to Ron and big Georg and Mick. Me and Alf used to cop for anyone who came back to check us out. We’d wrap them up and make sure it was safe to carry on. (p. 89)

Within the narrative framework, the narrative-identities and corresponding narrative-roles of Boyle, Smith, and Foreman were that of “hard men,” “tough guys” who would use violence as a means to achieve their goals. They each had far more confidence than doubt in their abilities to take on opposition during times of physical conflict, and this was reflected in the type of crimes that they committed (i.e., acquisitive crimes but underpinned with violence).

The propensity toward a genre of offense types or “specialisation” derives from certain characteristics (Donald & Wilson, 2000) including an offender’s perceived skill set or personality traits. Within the narrative framework, this appears to influence narrative-identity formulation and offense types, which together with the offender’s related event-schema, influences the subsequent “narrative-role” (Youngs & Canter, 2012a). The narrative-role in repeat offenders is presumably expressed through the offense script (Cornish, 1994a, 1994b), which is the general sequence to committing the offense, embedded within that is necessarily the offender’s MO.

Theme 5: Desistance

This theme is characterized by the somber tone of reflection on a past life that failed to live up to early expectations and hopes. It is a recognition that something has gone terribly wrong and cannot be put right.

According to Maruna (2001), desisters hold a redemption script which casts them as victims of society, where they were involved in crime to achieve power in an otherwise bleak situation. Within his model, deviance becomes a trap, but when the chance to desist occurs, desisters can accomplish what they were always meant to do and they can give something back to society. This suggests that there comes a point when, if an offender is given the right opportunity at the right time, they will rework their scripts to facilitate desistance. Within the redemption script, a series of internal shifts take place to enable this, but these shifts occur as a result of favorable changes in the holders’ external environment. The past is made sense of and rendered as not being completely pointless by retaining perceived culturally valued elements of the old self, which support the growth of a new man. This is particularly so if they can find roles or occupations that provide the same sense of empowerment and potency that they were seeking through criminality (Maruna, 2001).

Boyle used hatred of the guards as motivation to him to keep going:

The only way to pass time was to do exercise and reading, and thinking thoughts of hatred. My former life was far in the distance and I would try not to think of my family as it hurt too much. (p. 179)

He began to develop a mission in which to expose the brutality of the guards where he languished, “For the next week we concentrated on getting the story out to the press. We had this obsessive belief that getting it into the papers would be the panacea although earlier experience had shown us differently” (p. 216). This was a mission about survival more than about reform. He hated the system and the people who worked in it. For him, they were bullies and cowards who took advantage of their superior positions.

Eventually an opportunity for desistance was presented. He was initially highly suspicious and reluctant to accept it. He was sent to the newly established Barlinie special unit, designed for a small number of prisoners. Rehabilitation was far more humane than he had previously experienced. It took time for adjustment to this, but eventually he recognized how a shift in emphasis could keep his mission alive while also improving his own quality of life, “I could see that bit more clearly now, realising that the unit could have the potential to show that change was needed and could be made within the system” (p. 243).

Smith also used hatred as a means of survival:

The more they prodded, poked and ordered me around, the more I grew to hate the bastards. I could feel my hate roiling and rumbling within me, like thick molten lava seeking the smallest crack to erupt out of, but I clamped down on it, set my face on neutral and used hate as the fuel that drove me. (p. 57)

He found he was able to control it enough to call upon when needed to either fight the system or motivate himself to keep going during periods of despair. As time went on he discovered a new way of channeling this hatred, “By now I was more into fighting the system with the written word than throwing digs at screws” (p. 409). This eventually changed again as his mission began to develop into a more constructive tool.

The prison governor visited Smith soon after his reception and offered him a deal, “I could stay in Highdown and not go back to Dartmoor on condition that I would cause no trouble for his staff and I would agree to start up an in-house magazine” (p. 422).

This series of cognitive shifts in their missions did not occur without these opportunities. This was not simply a re-storying of their lives, but change occurred through a deeper and more fulfilling purpose (Ward, 2012). A simplified illustrative model might look something like this: (a) hating the system and using violence to fight it, (b) motivated by hating the system but fight it through legitimate means, and (c) not hating the system, but still fighting against its perceived injustices with legitimate means. Boyle and Smith had both experienced many moments in custody where they realized that their lives up to that point had been for nothing and were simply squandered with little future ahead of them. Although similar to Maruna (2001), where “new scripts” could be constructed with the narrators’ retaining some elements of the “old self,” new ways to move forward were quite different here. The protagonists’ attributed their offending to their own character inadequacies and their retention of old character elements was to enable a continuation of their fight against the system, which was in contrast to Maurna’s (2001) desisters who made sense of the past and used it to emerge as better people. This point becomes important when considering the “emplotment.”

Theme 6: Emplotment

This third step of the analysis is more about meanings (McAdams, 1996) than patterns of events, while also drawing from different theories to help make sense of those narratives (Murray, 2003).

Using dominant discourses (Willott & Griffin, 1999) that effectively sustain and reproduce the social status quo, the protagonists describe themselves as “good” criminals, “chaps,” and “professionals,” while today’s criminals are seen as people without honor or code.

Boyle wrote,

The “Criminal Code” isn’t a thing that has been written up by top gangsters. It is an unwritten code of ethics. There are done things and things that are not done. It isn’t the done thing to “grass” or inform on anyone. It isn’t the done thing to “bump” or cheat someone from a robbery that you have all taken part in. There are lots of these unwritten rules that could fill another book but these are just two examples that exist between guys in crime and on the whole they abide by them. (p. 99)

Smith wrote,

I had always had this strange notion that armed robbery was an honourable profession. I didn’t go around coshing pensioners for their bingo money, or burgling people’s homes or raping women and children; I was the cream of the criminal classes, a “blagger,” one of the heavy mob, a modern-day highwayman. I didn’t steal from poor people’s pockets, I robbed the rich, the banking and insurance institutions. People who could afford it. The only “victims” of my crimes were the faceless money-shufflers who had to tally up what was nicked. I was almost a Robin Hood figure, for fuck’s sake! Or was I? (p. 7)

Foreman wrote, “I never robbed poor people or broke into homes. We robbed only business and factories and did jump-ups with lorries. We considered that respectable and honest thieving” (p. 82).

They describe themselves as superior criminals to others. Offenders often talk about other offenders in a downward social comparison (Maruna, 2001). These statements are examples of how they see themselves while reflecting back. These references to “code of ethics,” “honourable,” and “respectable” denote highly valued virtues in society, Presser’s (2009) “cultural context” and the appeal to these underpin the reconstruction of their past lives while neutralizing or minimalizing the effects of their role in offending. This suggests a naïve attempt to make sense of their past so as to position themselves as having been in control of their lives when that is not certain (i.e., particularly the trait-driven phase), prior to the development of McAdams’s (1996) narrative-identity phase. In many ways, this is at the core of the emplotment because it is not only about justifying the lifestyle but also trying to make sense of it after desistance and while retaining the self-concept through the self-narrative and positioning it within socially valued conventions. Without these rationalizations about their pasts, it is very difficult to not feel that their entire lives were wasted up to that point.

The emphasis of some stories over others facilitates this coherence (McAdams, 1996), and cultivates an identity of being one person over time and across circumstances. These self-defining memories are essential to ones sense of self (McLean, 2005). It is the concordance from discordance (Ricoeur, 1994), or the winning out over a lifetime of disruptions (Murray, 2003) in the form of cyclical prison sentences and the preceding behavioral traits that they could not make sense of at the time. Furthermore, techniques of neutralization are seen by Presser (2009) as a forerunner to narrative criminology; this rationalization part of the emplotment corresponds to the basic principles that support Sykes and Matza’s (1957) assertions. However, how these neutralization techniques are implemented in the emplotment when considering Hamlin’s (1988) point—that it is unclear which came first, the delinquent act or the belief justifying it—suggests an iterative process in which narrative construction evolves with criminal development. Ward and Marshall (2007) described narrative-identity as a dynamic and somewhat fluid process. Hence, the protagonists later render an account that makes sense in a way that did not during its unfolding. In doing so, they can only draw on those culturally available discourses (Presser, 2009) and so submitting to the socially dominant ideologies, retrospectively see themselves as having been fully informed agents who were in complete control of their lives.

The need to reconcile the past results from a shift in aspects of the belief system, while retaining continuity and conveying a message about who the narrator really is, but is constructed without appreciating the earlier trait-driven behaviors and lack of a developed agency. As Maruna (2001) acknowledged, there may not be a language or discourse available to describe these changes.

For example, Foreman wrote, “Throughout my life, I have always felt revolted by liberty-takers and bullies. In the eyes of the law and society I am seen as a villain, but I have always strived to be quiet, polite and well mannered” (p. 160). Here Foreman makes sense of his past without appreciating the traits that underpinned the early offending by seeing himself as having at least some justification for his lifestyle when referring to people who act in an inappropriate way. He then seeks retrospective clarity and continuity by describing himself with the socially valued attributes by always being “quiet” and “polite,” when the evidence suggests that he unwittingly held a distorted and mixed belief system.

Another important feature both of the narrative and the resulting “emplotment” is “reputation,” which appears to have played an important role in the identity formation of the protagonists here. This is a subplot of the emplotment in that behaviors and rationales relating to the reputation are consistent with the rationales of the emplotment (e.g., formidable but fair). Reputation also appears to be related to the offense type(s) in which Boyle, Smith, and Foreman specialized (Farrington et al., 1988; Klein, 1984), as well as its influence on the MO or the way in which the offenses were carried out.

After slashing another guy in a gang fight at age 14 years, Boyle recalls,

Within days I was a force to be reckoned with and some kids were saying that I was as “mad as a brush.” There was a sort of hero worship about all of this and I was placed on a higher pedestal by all my own gang, but like reputations in other fields you’ve got to deliver the goods otherwise you’re in trouble. (pp. 57-58)

As time went on, violence also began to enter his thieving. By the time he had reached the age of 18 years, he had already spent a number of periods in custody and on one such occasion recalls, “I only knew that I wanted to be out and big in crime” (p. 107).

Similarly, Smith reflects, “I craved the recognition of being a ‘face’ in, and out of, prison, and I loved the respect I got from others in my world” (p. 7). He did things to maintain his reputation, “So that other criminals could say, There goes Razor Smith, he’s a diamond geezer. One of our own” (p. 478).

Foreman reflects on his time during his mid teenage years, “I was quite well known and already had a bit of a reputation as a street fighter” (p. 38) and later in adulthood, “There was no limit to what you would do. People knew that, so you held that bit of fear and respect and it was true of the situation at the time” (p. 84).

For Stevens (2012), an individual’s identity derives from the capacity to keep the narrative going and reflects a reconceptualization of identity formulation that is not trait driven with a fixed core throughout life, but a reiterative activity. However, this view is too narrow as it ignores the developmental phase of the narrative-identity and how that may contribute to its construction and maintenance, as well as the body of evidence that does support the notion of early trait-driven behaviors. Whatever the conceptualization of identity, reputation becomes important and this is borne out by the comments of the protagonists’ here as well as other offenders. For example, Courtney (2008), who was another violent criminal during the Krays era and an associate of theirs as well as Foreman’s, alludes to this notion of honor among some criminals. He claimed to be inspired by how he saw the Krays and the presence that they exuded. Similarly, James (2003) wrote that to survive in the prison system, it was essential to at least appear like you could look after yourself in a fight. Maguire and Bennett (1982) noted that the offenders in their study were keen to be associated with those offenders with big reputations.

Discussion

Central to narrative theory is narrative-identity and according to McAdams (1996) narrative-identity begins to develop in later-adolescence. It is predisposed to create identity along certain thematic lines, including biological, cognitive, and socially related processes. This identity formulation appears to emerge around the time that specialization in offending begins to take shape (Farrington et al., 1988; Klein, 1984), which is sometime after the onset of offending for Moffitt’s (1993) life-course offenders. Indeed, Presser (2009) regarded this as unchartered territory.

In the absence of narrative-identity formulation (McAdams, 1996), early onset appears to derive from a different source and this surely needs to be addressed before narrative theory can be fully realized in the offending context. For example, it may be that traits drive early offending behavior leading to repetition, resulting in routines and habit formation, of which constitutes the offense script (Cornish, 1994b). Direct rewards such as material gains and further opportunities, including reputation, presumably also feed and reinforce the cycle. Within the McAdams approach, as late adolescence draws near, agency begins to play an increasing role but may initially be limited to refining offense decisions (i.e., offense type and MO). The offender is drawn toward maximizing material benefits by offending in ways that are not only familiar and comfortable, but which maintain and enhance reputation. Theoretically, this would account for the simultaneous emergence of offense specialization and narrative-identity, which in turn is built around the self-concept (i.e., who one believes oneself to be, not necessarily who one is; Ward, 2012; also see Fiske & Taylor, 1991, for a detailed discussion). In this respect, choice making begins to play a greater role than before, but its rationality is bounded (Cornish & Clarke, 1987) within the parameters of these life enduring traits and social settings, while agency capable of controlling the self (Ward, 2012) remains under-developed in these chaotic and “high-volume” offenders (Halliday, 2001). At the height of their bounded choice-based offending, the protagonists here could well be described within Youngs (2004) “Expressive” category, Youngs and Canter’s (2012a) “Professional” category, and Merton’s (1957) “Innovator” category. During this period, there appears to be a convergence with McAdams’s (1996) second modality of agency, which characterises high power motivation, strength, domination and risk taking.

The eventual cognitive shifts noted in the desistance theme above and to some extent Maruna’s (2001) general findings facilitate a different channeling of the offender’s traits from before and are which expressed through a change in the social context. It appears more likely that a fully developed agency, or “higher-order” cognitions (Brehm, Kassin, & Fein, 1999), needed to be able to take control of the “self,” does not emerge until much later in those offenders who are habitual due to their antisocial traits (e.g., Boyle, Smith, and Foreman desisted much later in adult life). Furthermore, although traits may fail to predict behavior in a single situation it is behavioral aggregation that enables trait claims (Matthews & Deary, 1998). In addition, in contrast to Moffitt’s (1993) life-course offenders, Moffitt’s adolescent-limited offenders presumably do not develop narrative-identities based on criminality, but for other reasons, desist from offending around this time. The timing of this narrative-identity formulation is entirely consistent with the age crime-curve for adolescent desisters and those that continue offending. Maruna’s (2001) claim, that desistance is a problem to trait theorists, is difficult to support not only in respect of these wider considerations, but in light of the protagonists here, who found ways to reform while managing their traits.

Philosophically, Ricoeur (1994) conceived of two distinct and polarized but overlapping values, which are identity as “selfhood” and identity as “sameness” or as oneness and together form the narrative-identity. Selfhood is the cognitive part, the choice making of the identity which may be based on a set of beliefs about the self (e.g., self-identity). Sameness is the component that retains a continuous element across time (Ricoeur, 1994). In the offender context, this can be understood as trait characteristics which form part of the “actual-self” or it can be understood as the part of the “perceived-self” that is always seeking to justify one’s actions within cultural values. For Ricoeur, identity takes up position in the space between these two values of selfhood and sameness, and oscillates between them. For Ricoeur, it is analogues to aging, where one’s appearance changes over time, but an identifiable element of the previous remains. In relation to narrative criminology, it is the slow development and emergence of agency over time, which ultimately wins out over the earlier trait-driven behaviors, hence the eventual desistence. The emplotment is the protagonist’s reflective sense making of the what, why, and how, which has come after the cognitive shifts have occurred. It is the necessary drawing on sameness to create a coherence to make sense of the past and prevent a complete collapse into worthlessness by creating or maintaining a value to it. A narrative-identity reframing (Yardley, Wilson, Kemp, & Brookes, 2013) derives from this process. The story that comes from this is the tragedy (Murray, 2003) of the missed opportunities and the realization of “the person one was always meant to be” (Stevens, 2012).

Limitations

The main potential limitations of this research are the use of the written autobiography and the sample size in terms of whether these accounts are representative of life-course offenders generally. In addition, they maybe self-written or produced by a ghost writer working in collaboration with the protagonist who communicates what is important to them, although this is ultimately the essence of a narrative. Written autobiographies provide an offender with an opportunity to reflect over a longer period of time and consolidate understanding with memories in a way that interviews may not. For example, where Maruna (2001) found that his desisters attributed their inability to control their offending to outside forces (e.g., selfishness as being the product of the environment), Boyle, Smith, and Foreman saw their inability as deriving from internal strivings, which took many years to understand and manage. Perhaps these discrepancies result from the differences between protagonists being able to produce their own account in their own time, compared with the dynamics of an interview setting which produces a particular rendering at the point of narration, “deriving from the immediate social contextual demands” (Prebble, Addis, & Tippett, 2013). Although there is no claim here that these protagonists are representative of the offending population, it is interesting to note that the accounts were similar in many ways beyond simply being life-course offenders as alluded to by their developmental pattern. Clearly, more autobiographies should be analyzed to explore these issues further.

Future Implications

One issue with using narrative theory as a general theory of life-course offending is that according to McAdams (1996), the principle model utilized by Maruna, 2001, “narrative-identity,” does not appear to emerge until after the early onset associated with life-course offenders. However, children do develop the cognitive skills for telling stories long before mid-adolescence (Habermas & Bluck, 2000), although with limited purpose (McAdams, 1996), and so, more work needs to be done on this unresolved issue (i.e., does early adolescent offending begin in the absence of a narrative-identity or does narrative-identity emerge first and play a key role in influencing early offending). If trait-driven behaviors are responsible for early onset, two things are plausible: that Presser’s (2009) “Constitutive” culturally available narrative-identities, as postulated by Maruna (2001), Canter and Youngs (2010), and Youngs and Canter (2012a, 2012b), iteratively form the sense-making component to the life-course offender’s “offending” (e.g., offense type and identity) retrospective to early onset, or when giving their accounts their participants draw from “culturally” dominant discourses as the only available means for them to make sense of it.

Furthermore, while aggression could alternately be explained through theories of masculinity, impulsivity still remains less well explained by alternatives to dispositions. Moreover, the characteristics of impulsivity and aggression appear to continue to play an intermittent role throughout adult offending, particularly during desperate times as the analysis above alludes to (also see Maruna, 2001). The assertions of Youngs and Canter’s (2012a, 2012b), and Canter and Youngs (2012), “that agency plays a central role in a given offence-narrative or theme,” is not without some merit as offense decisions are made. Nevertheless, further work is needed to determine the characteristics of this “agency” and how it operates (i.e., prior to or during narrative-identity formulation). For example, do offenders freely choose a life of crime or choose a criminal narrative-identity from a particularly limited pool of perceived narrative options following onset.

In Healy’s (2013, 2014) desistance work, all human beings have the potential to act agentically, which is capable of being activated based on the capacity to imagine a new self that is meaningful and credible. Persistence is not the result of psychological deficits but the absence of this imaged identity and so lack incentive to exercise agency. Healy’s research was derived from adult offenders and this refashioned hark back to traditional free-will choice making tends to become more salient in the later stages of the criminal life course and appears to have minimum impact on early onset for life-course offenders, based on the results here.

B. Smith and Sparkes (2009) found that pre-rehabilitation athletes with life-changing spinal cord injuries (SCI) perceived of a few, but common (among SCI’s) ways of future identities and narratives. For B. Smith (2013), the narrative habitus is an inner library of narratives that the individual holds. In their rehabilitation work, they seek to introduce the SCI athlete with alternative narratives that can successfully bid for a location within the organization of the narrative habitus. These ideas have implications for work around agency in terms of free-will or bounded rationality in the context of offending.

Finally, theoretical constructs such as “self-concept,” “traits,” and particular versions of “narrative-identity” are entirely compatible within a micro-level “realist” framework as they derive from notions of internal mechanisms that can be studied through standardized approaches (see May, 1996). It is in the post-positivist approach as outlined by Presser’s (2009) macro-level “constitutive” approach that such concepts are not epistemologically compatible, as there is no commitment to the notion of an accessible reality. As narrative theory develops within the forensic setting, it may be helpful for researchers to make explicit their epistemological orientation. As observed by Ward (2012), this will avoid confusion across the different ideologies underpinning the various approaches to narrative theory that Presser (2009) described.

Conclusion

Based on the limited results of this exploration, narrative theory when applied to life-course offenders does have utility, but more work is needed as an explanation for early onset. Thus far, it may be better understood as a theory that maintains offending and develops it in specific ways. In this study, its earliest emergence (in terms of narrative-identity) began with specialization, which is more about themes underpinning offense types than very specific acts. However, earlier trait-driven behaviors which are more versatile due to the randomness of opportunities and in the absence of an offender identity to guide decisions appear to provide the context and background in which specialization emerges. This specialization appears to be linked to narrative-identity (i.e., the protagonist’s perceived role in a chosen offense type) as well as the emplotment (i.e., the meanings the protagonist attaches to their role). Trait-driven behaviors and rational choice models are typically treated as disparate areas of inquiry, but it is possible to conceive how one gives way to the other in the same offender.