Abstract
We examined narrative processes of identity development as they related to desistance from delinquent behavior in a sample of vulnerable adolescents. Building on a robust theoretical and empirical foundation in the field of narrative identity, we examined processes of meaning-making and agency in relation to desistance. Thirty-one adolescents were interviewed to elicit commonly studied autobiographical memories (e.g., turning points, self-defining memories). Interviews were coded for meaning-making and agency. Results showed that desistance was related to agency, but not meaning-making, which was related to higher numbers of past delinquent behaviors. Results are interpreted in terms of the possibilities for desistance and growth from a narrative perspective and are situated within a structural and cultural perspective on the development of vulnerable youth.
Engagement in delinquent behavior tends to peak in adolescence and emerging adulthood and, for most, declines in early adulthood (Hirschi & Gottfredson, 1995; Moffitt, 1993; Rhodes, 1989). For others, problematic behavior in adolescence develops into adult criminality. Given the potential hazards associated with engaging in delinquent behavior, the question of how and why adolescents desist has captured the attention of researchers (e.g., Breen, 2010; Laub & Sampson, 2003). There are a variety of explanations for desistance, including normative processes of maturation, cognitive development, and social identity transitions associated with new roles such as parenthood (for a review see Laub & Sampson, 2001). However, a less studied question is the role that socioemotional developmental tasks play in desistance from delinquent behavior, in particular, the key developmental task of this stage: personal identity development.
In undertaking this investigation we took a narrative approach to identity, focused on individual variation in narrative processes employed in constructing one’s past experiences (McAdams, 1993). From a narrative perspective, meaning-making processes—those in which individuals tell personal stories demonstrating what they have learned about themselves from past experiences—are viewed as a critical component to narrative self-development (e.g., Habermas & Bluck, 2000; McLean & Pratt, 2006; McLean & Thorne, 2003). However, little research on narrative development has examined diverse populations, particularly vulnerable youth (cf., Breen, 2010; Greenhoot, Johnson, & Glisky, 2005; Sales, Merrill, & Fivush, 2013). We wondered whether meaning-making processes may be different for adolescents who have experienced numerous negative life events, such as parental death, sexual and physical abuse, and who also live with structural violence, especially poverty. That is, when much of the content of one’s life story is overwhelmingly negative, is meaning-making a vehicle for positive development?
While there is a somewhat normative peak in problematic behavior in adolescence, adolescents who are least likely to desist come from distressed backgrounds in terms of economic resources, family functioning, and peer deviance (e.g., Sanderson & McKeough, 2005). We investigated whether there were any developmental processes we could point to that might explain the difficulty of desistance for these particularly vulnerable populations. Given that the primary developmental task of this stage is identity development (Erikson, 1968), we examined processes of narrative identity construction as potential keys to the difficulties with desistance. While many particularly vulnerable adolescents are not able to desist, a narrative perspective suggests that those who do may be able to construct a self-story that supports and guides behavioral change (Breen, 2010; Breen & McLean, 2010; Hauser, Allen, & Golden, 2006; Maruna, 2001). A narrative perspective is particularly suited to studying questions of risk and vulnerability in adolescence because narrative can shed light on the intersections of the individual and his or her familial and cultural contexts (Bruner, 1990; Hammack, 2011).
We note at the outset that overcoming delinquency, which in our sample may be inextricable from familial and structural forms of neglect and violence, is a tremendous challenge, and authoring a self-story that promotes resilience is asking a great deal of individuals who develop within contexts of structural violence. The power of stories to transform may be strained when adolescents face the menacing obstacles of poverty, addiction, and abuse. Further, individuals construct their identities by looking to their social environments for possibilities and expectations (Erikson, 1968). Thus, cycles of poverty and violence may not only be structural problems, but also narrative problems stemming from larger story structures that limit possibilities for resilience (Hammack, 2011).
Narrative Identity
McAdams (e.g., 1993) views narrative identity as a psychosocial construction, reflecting the ways in which one makes sense of the personal past, one’s current self, and future expectations. Influenced by cultural context and available story structures (McAdams & Pals, 2006), people construct stories about how they came to be, which provides them with a sense of unity and purpose. This process of story construction begins in early childhood and extends across the lifespan, but it is in adolescence when cognitive skills and social demands for developing a self-story become prominent.
Meaning-Making: Stories of Growth
It has been argued by many theorists that in constructing a coherent self-story, one must engage in some degree of reflection, even if it is simply in recounting how one felt at the time of the event (e.g., Fivush & Baker-Ward, 2005; McAdams, 1993; McLean, Pasupathi, & Pals, 2007; Thorne, 2000). One process of more effortful reflection (McLean & Fournier, 2008), which we examined in the present study, is meaning-making, defined as the active reflection on what one has learned about the self from past events (e.g., McLean & Thorne, 2003). Studies have found steady age-related increases in the frequency of meaning-making processes in cross-sectional studies of adolescents (McLean & Breen, 2009; McLean & Mansfield, 2011; McLean, Breen, & Fournier, 2010; see also Habermas & de Silveira, 2008; Pasupathi & Wainryb, 2010).
While adults who learn from, as well as resolve, difficult life events tend to have higher well-being than those who do not (e.g., Pals, 2006), the findings in adolescence have not been as clear. For example, McLean et al. (2010) found that for younger adolescent boys, greater displays of meaning-making was actually associated with lower well-being, though this trend began to reverse by late adolescence (see also Chen, McAnnally, Wang, & Reese, 2012). McLean et al. (2010) put forth several interpretations, including that boys who reported more meaning may have been experiencing especially disruptive events that require meaning-making, but that may also cause lower well-being. Indeed, in a sample of at-risk African American adolescent girls Sales, Merrill, & Fivush (2013) found that meaning-making was negatively associated with positive adjustment. Further, some studies have shown that personal reflection in the service of developing self-identity may pose risks for vulnerable adolescents as such processes have been linked to depression and suicide (Borst, Noam, & Bartok, 1991).
These preliminary data suggest that it would be prudent for narrative researchers to step back from our current models to examine whether such models can encompass the identity processes of individuals from various subpopulations within a given culture (see also Greenhoot & McLean, 2013). Our sample consisted of adolescents who were particularly vulnerable in terms of the number of negative life events they had experienced, and who also experienced structural constraints on identity possibilities, such as poverty. Given the limited research described above that suggests that there may be some hazards to meaning-making in such a sample, we also consider alternative narrative processes that may relate to desistance, in particular, agency.
Narratives of Agency: The Role of Personal Responsibility
While meaning-making captures the degree to which one learns about the self in constructing a narrative identity, we also examined the extent to which one’s stories reflected a sense of agency, defined as taking responsibility for one’s own decisions and actions. Success in the transition to adulthood requires that individuals assume responsibility for themselves (Arnett, 1997), which is reflected in several related constructs, including locus of control (Rotter, 1990) and broader definitions of agency (de Silveira & Habermas, 2011; Ryan, 1993; Sales et al., 2013). However, we took a more narrow approach to agency, given the sample with which we were working. For example, agency typically emphasizes an orientation to future action and outcomes, is concerned with goal formulation, planning, and mastery, and is connected to the individuals’ sense of meaning and purpose in life (Adler, 2011; de Silveira & Habermas, 2011; Sales et al., 2013). We recognized that the difficult life events experienced by many of the youth in this study may have constrained the participants’ possibilities for achieving their most personally meaningful longer term goals. We also recognized that certain kinds of experiences that were common in our sample, such as violent victimization, are not within the realm of personal responsibility. Thus, we limited our focus to everyday agency, including both relatively “small” decisions (e.g., completing homework) and larger decisions (e.g., ending contact with a negative peer group).
There have been only a few studies to look at similar concepts. First, in Sales et al.’s (2013) study of at-risk African American adolescent girls, internal locus of control was associated with higher well-being. Second, in a longitudinal study of adolescents who were hospitalized in a psychiatric institution, Hauser et al., 2006 found that resilient adolescents saw themselves as having some control in their lives. Thus, agency appears to be an important aspect of resilience for vulnerable youth.
In the present study we examined the degree to which the narrative processes of meaning and agency were associated with desistance from delinquent behavior. Our study included both a deductive quantitative component and an exploratory qualitative analysis, which provided insight into the participants’ identity processes within the unique contexts of their sociocultural and familial environments. The qualitative and quantitative components of our study are united by an overall emphasis on narrative. While our research focuses predominantly on examining stories as a process of identity construction, our qualitative work is guided by narrative methodology, which emphasizes in-depth analysis of the whole of individual’s stories (Riessman, 2008).
Method
Participants
The population from which we drew our sample was impoverished and primarily Caucasian, living in a small town in Northwest Washington State, with a population of about 10,000. While there were once thriving logging and rail economies, the current average income and percentage of citizens with a bachelor’s degree or higher is below that of the state average, and the percentage of citizens living in poverty is higher than the state average. Our data were collected at an alternative high school, designed for students who are not excelling in a traditional school setting for a variety of reasons (e.g., expulsion, learning difficulties, pregnancy). The sample (n = 31) ranged in age from 15-19 (m = 16.71, SD = 1.05; n = 19 female). The majority of participants self-reported their ethnicity as Caucasian (n = 27), which is representative of the area (92% Caucasian). One third were raised by both mother and father, and the rest were raised by a single mother or other combinations (e.g., stepparents). One half reported that their parent(s) received welfare. Three participants had one child, all of whom received welfare. Time spent at the alternative school ranged from 1 to 48 months (m = 12.31 months, SD = 10.24).
In terms of the contexts in which these adolescents were living we gleaned the following experiences from their interviews: parent in jail (n = 3), death of a parent, close family member, or friend (n = 11), abused by a relative (n = 8), drug-addicted or alcoholic parent (n = 10), in foster care (n = 1), ran away, kicked out, or not living with parent (n = 7), homeless (n = 2), suicide attempts, suicidal thoughts, or self-mutilation (n = 5), self or parent received psychiatric diagnosis (n = 7), sexually assaulted (n = 2), and experienced domestic violence (n = 2).
Procedure
Students were recruited for the study via class announcements and signs posted around the school. Students were interviewed by a female interviewer (the second author, who is a native of the town) during lunch or class-time in a private room, and were given US$10 gift cards to a major department store for participation. After the interview, participants completed a survey packet; relevant surveys are described below.
Measures
Interview
Interviews were audio-recorded and began with a warm-up by asking students to think of their lives as a movie, and what kind of movie it would be (e.g., romance), and which actor/actress would play them (Breen, 2010). We then asked participants to describe their childhood, if they currently felt like a child or an adult, and why, and we also asked for specific memories to illustrate their answers to these two questions. Next we asked them to narrate a life high-point, a self-defining memory, and a transgression. High points were described as experiences when the participant might have felt, joy, excitement, happiness, or some other positive emotional experience (McAdams, 2006). Self-defining memories are vivid, important, highly emotional (positive or negative), and help one to understand who one is as a person (Singer & Salovey, 1993). The transgression was described as a time the participant went against his or her sense of who one is/what one believes (Mansfield, McLean, & Lilgendahl, 2010). If details were not forthcoming, prompts included questions about what specifically happened, when it happened, how participants currently felt about the experience, whether or not they learned anything about themselves, and whether or not they feel that they have remained the same or changed since the event. Finally, they were then asked what brought them to the school, and thoughts about the future. Interviews took about 30 minutes.
Survey
After the interview, students completed several surveys, only one of which we used: an inventory of risky behavior. This inventory assesses how many times they had engaged in each risk behavior, and time since last engagement for each behavior. This scale was adapted from Tanner and Wortley’s (2002) Toronto Youth Crime and Victimization Survey (see also Breen, 2010), using only questions that focused on crime. From this inventory we created two relevant variables—the number of different kinds of risks participants had engaged in and the average time since last engagement in risky behavior (desistence). We computed the number of different kinds of risk behaviors, as opposed to the frequency of engaging in any risk behavior, to deal with what appeared to be inflated numbers of experiences with each risk behavior (see ranges below), and because engagement in multiple risk behaviors (e.g., stealing, drug use, and aggression) may be indicative of more substantial conduct problems than frequent engagement in just one category (Murray & Farrington, 2010; Sullivan, Childs, & O’Connell, 2010).
Narrative Coding
Interviews were transcribed and checked by a team of research assistants. The second author and two research assistants coded meaning-making, and the third author and one research assistant coded agency. Research assistants were blind to hypotheses.
Meaning-Making
Participants’ memory narratives were coded for meaning, based on McLean and Pratt’s (2006) rubric (intraclass r = .96). Narratives were coded according to a 4-point scale. A score of zero was assigned to narratives that contained no explanation of the meaning of the event to the self. Narratives were scored as one if there was mention of a specific lesson that the reporter learned from the event. A score of two was assigned to narratives that contained vague meaning; narratives of this sort describe some growth or change in the self, but the specifics of the change are not clear. Narratives were scored as three if there was evidence that the reporter gleaned specific emotional, psychological, or relational insight from the event that applies to broader areas of the reporter’s life. For analyses, we averaged the meaning scores across memories.
Agency
Interviews were holistically coded for the extent to which the narrator evidenced a sense of responsibility for his or her own actions on a 4-point scale (intraclass r = .80). A 0 indicated no evident sense of personal responsibility, such that the reporter seems to view the cause of and responsibility for his or her actions as residing outside of the self (e.g., with another individual, fate). A 1 was used if the reporter had an awareness that he or she should be responsible for his or/her own actions, but the sense of responsibility was inconsistent, rare, marked by ambivalence, or viewed as residing predominantly within others. A 2 was given when there was a moderate sense of personal/internal responsibility for one’s own actions, indicating that there was a tendency towards a view of responsibility as residing within the self. A 3 was given when there was clearly a high level of personal responsibility, suggesting a strong sense that one’s actions are one’s own responsibility.
Results and Discussion
Descriptive Statistics
In terms of their self-reported risk behaviors: 12 had been arrested (range = 1-250; removing the participant who reported 250 arrests: m = 6.27; SD = 14.53). Twenty-three participants reported stealing (range = 0-100; m = 6.05, SD = 18.30). All but one participant reported using alcohol (range = 1-100; m = 27.13, SD =38.64). Thirteen participants reported using over the counter drugs for nonindicated use (range = 1-100; m = 10.65, SD = 29.79). Twenty-six reported using illegal drugs at least once (range = 1-600; removing the participant who reported 600 times, m = 10.65, SD = 29.80). Nine participants reported selling illegal drugs at least once (range = 1-50; m = 3.34, SD = 10.31). Twelve participants reported carrying a concealed weapon (range = 1-100; m = 10.71, SD = 27.51). Twenty-six participants reported acting violently (range = 1-100; m = 5.47, SD = 17.77). Two reported belonging to a gang. The mean number of risky behaviors participants reported was 4.65 (SD = 1.82, range = 1-8, mode = 4).
Table 1 shows the means, standard deviations, and ranges of our main variables of interest. In terms of demographics, girls reported more agency than boys did, t(29) = −2.51, p <. 05 (d = .93). There were no gender differences in meaning-making, and there were no associations between our narrative variables and family structure (intact or not), income (received welfare or not), or age.
Intercorrelations and Descriptives of Main Variables.
Note. *p < .05. **p < .01.
Central Analyses
We will take each primary analysis in turn, using our qualitative data to unpack the quantitative findings. As can be seen in Table 1, meaning and agency were positively correlated, which we take into account in subsequent analyses.
Meaning-Making
In terms of our first question, meaning was not associated with desistance. However, meaning was associated with engaging in more varieties of delinquent behavior, and this relationship remained when controlling for agency (r = .43, p < .05). In examining the kinds of reflection in which participants were engaging, individual meanings were: (a) clearly positive in nature (“I learned that I was loved.”; n = 22) or focused on lessons (“I learned not to vandalize again.”; n = 21); (b) more ambiguous about valence (“I learned not to trust others”) or mixed in valence (“I want to be a better person than my mom”) (n = 13); or (c) negative (“I am an asshole”) (n = 13). It is the latter two groupings—ambiguous and negative, which suggest that meaning-making might not always reflect adaptive processes.
Josie’s interview represents the process of negative meaning-making. Josie was 17 at the time of her interview, and was raised by a single mother, who was on welfare. Josie has a child of her own, and she reported that her pregnancy brought her to the school, where she has been for 6 months. Josie’s past risk behavior included multiple arrests, selling and using drugs, and violence, the latter of which all occurred over 1 year before the interview. She also reported alcohol use within 6 to 9 months before the interview. Her life history includes exposure to family drug use, time in drug/alcohol rehabilitation, violence, rape, a suicide attempt, and a diagnosis of bipolar disorder.
In discussing her suicide attempt (her transgression memory), she reported, “I learned that if I get pushed enough that uh a way of feeling better is just think about things I shouldn’t think of, for example, like . . . death.” In discussing her self-defining memory, which concerns her mother’s broken promise not to have more children, Josie reports: “ . . . I’m the only person that I can rely on in my life because I’ve tried to rely on other people and I either get stabbed in the back or hurt, so I really know that I can only trust myself and rely on myself.” Josie’s meanings suggest several processes. First, she has learned that thinking about death is a coping mechanism for her. She is reflecting and learning, but such a meaning may not put her on the path to resilience, and may instead validate or intensify her potentially negative trajectory, unless she is able to find other coping mechanisms. Her second meaning concerning her lack of trust in others is slightly different. From a relationships and resilience perspective, learning to trust others and having people that one trusts in life is a critical developmental accomplishment (e.g., Bowlby, 1969), which would classify Josie’s meaning as “negative.” However, this realization might be adaptive given her background. Without environmental supports (e.g., trustworthy people), she might have more “success” on her own. This may be part of the agency that she displays at other points in her interview, which we discuss later.
To illustrate the lack of association between meaning and desistance, we present another case, Tracey, who was 15 at the time of the interview. She had been at the school for 2 months and her attendance there resulted from a violent altercation with another girl at her previous school. Her mean score on meaning was above average, compared to the rest of the sample, but her desistance score was low. Her past risky behavior included being arrested, stealing, using drugs and alcohol, and violence. Her last reported risky behavior was one month previous to the interview. Despite this checkered past she is quite thoughtful about her life. In describing her childhood, she reported:
Um, I grew up with a sister, a mom, and a dad. My dad broke his back the year I was . . . like three months . . . so I kind of wasn’t picked up that much as a child by my dad . . . didn’t get to interact and do fun stuff with my dad as much as I wanted to, so it was kind of . . . makes me different, kind of like iffy like about people like hugging me and stuff like that . . . it’s ‘cause it’s I feel weird without . . . ’cause I can’t remember . . . my dad couldn’t do things with me and I was . . . a quiet kid when I was younger; wasn’t a really active but then I got in elementary school and got in a lot of trouble and was a troublemaker and kind of did it for the attention from my parents ‘cause it was all like focused on my sister. Yeah.
Tracey connects early parenting behavior with current relationship issues in a relatively nuanced way. She also provides an autobiographical reason for her earlier behavior.
Aside from this somewhat difficult relationship with her parents, she reported having a strong and positive attachment to grandfather who died when she was in 6th grade, right around the time she started to get into more serious trouble. She recounted a story about crying in front of her grandfather, who accepted her, and her mother, who did not tolerate crying. In answer to a question about whether or not she had learned anything from her relationship with her grandfather later in the interview, she reported:
That I’m str . . . crying does make you stronger. It doesn’t always, it helped to hold things in, but it helps to get things out and talk to people.
Here she reports learning about herself, and also learning about herself with some nuance—sometimes crying helps and sometimes it does not.
Tracey points out 7th grade as the time when she started to get into more trouble; she started to hang around with a bad crowd, sneaking out of her house, and doing drugs:
. . . [I] Think back and think I’m an idiot. (laughs) I could have done a lot of things different and I would probably have my dad’s trust, ‘cause I don’t have that right now because of all the stuff I’ve done in the past.
Did you learn anything about yourself?
I uh really suck under peer pressure. I snap and I’m like, fine whatever, let’s do it.
Her connection between having her dad’s trust and her behavior is important, but she also sees her weaknesses—succumbing to peer pressure and having low self-control. That weakness seems to still be hampering her desistance, as we see in her transgression narrative about a recent fight spurred by a classmate speaking poorly about her family:
I had told myself not . . . that I wasn’t going to be mean to people, that I wasn’t going to be a jerk or anything, and this year, around Halloween, I . . . there was this girl at school that I talked to about my family and my niece, ‘cause my sister had a baby, and so I finally snapped and went to school one day, and was fine and then she decided to call me rich and I like have really bad anger issues and I decided that I wasn’t going to take it anymore so I snapped, beat her up, got kicked out of school, got assault charges and anger management classes and all that, so I was like, went against what I was planning on not doing, but I’m not one of those people that you talk shit about my family and I just sit there and be like, oh yeah go ahead. (big inhale) yeah.
Um, how do you feel about the experience now?
I . . . feel like I could have handled it better, but then I’m . . . this sounds really bad to say, but I’m happy about it because she hasn’t talked to me about my family since, and so . . . I feel bad but then I feel kind of happy about it because she doesn’t say anything about my family, she hasn’t said anything about me since.
In this excerpt we again see the nuance with which Tracey thinks about these issues. She can see the situation and the different variables involved, but she does not seem to have the skills to put those thoughts into practice. Notably, Tracey is young. Middle adolescence is a time when individuals can see contradiction, but cannot yet integrate it (Harter & Monsour, 1992). Other recent work has shown that in early and middle adolescence meaning-making is not associated with well-being, and researchers have suggested that this may be due to the inabilities of adolescents at this age to fully employ their cognitive resources to understand and act coherently in challenging situations (Chen et al., 2012; McLean et al., 2010). Thus, while Tracey’s thoughtfulness does not appear to be helping her to desist currently, it is possible that it will in time. Of final note, Tracey did report a positive attachment relationship with her grandfather, her parents are still together, and in other parts of the interview she suggests that her parents are relatively involved in her life, and she does not indicate that they are neglectful, abusive, or substance users. Thus, Tracey’s thoughtfulness may also be related to a relatively stable family situation, which we did not see in the majority of our other interviews.
Agency
In terms of agency, mean time since engaging in delinquent behavior was associated with agency, but this association became nonsignificant when controlling for meaning. Our raw data revealed two outlier participants who received zero scores on agency, but who had relatively high levels of desistance. The first was an individual who had almost no detail in his interview (often saying he could not, or did not want to, remember his past), and suggested that he was engaging in suicidal thoughts. Our impression was that he was not able to engage with the interview process because of his current distress. The second case did not take the interview seriously, offering sarcastic comments throughout, with relatively little detail on his life history (also mentioning that he had trouble remembering his childhood). He also seemed unable or unwilling to engage with the interview. When removing these two cases the correlation between agency and desistance is obviously stronger, r(29) = .45, p < .05, but the relation between agency and desistance holds when controlling for meaning, r(29) = .37, p < .05. Given our small sample size these two cases may belong to a different group of adolescents (perhaps one that is more compromised in terms of vulnerabilities for the first, and a nonreliable informant for the second), though we offer these results with caution.
Consistent with the limited past research, it seems plausible that perceptions of personal responsibility allow individuals to take charge of their actions and desist from delinquent behavior. Josie was on the higher end of desistance in our sample, and also received the highest score for agency. This score was mostly driven from her discussion of becoming a parent, in which she discussed her cessation of self-medicating, partying, and her new abilities to love another person—her son.
When I had my son was the only time that I have ever been happy in my life honestly. That’s about it . . . when he was born my life really did change like before I thought I’m going to have a child and I’m going to feel the same way. I’m going to feel like there’s nobody for me in the world like to love or whatever…and that can really change you . . . you go in there still being pregnant and then coming out with a baby so it’s like hard to believe until it happens and like . . . I loved him from the moment I had him so . . .yeah.
Others have written about the potentially transforming effect of having a child for vulnerable youth (e.g., Breen, 2010), and this may be a place where such individuals can become agents. The impression we have of Josie is that she has a current self-story of a responsible mother with goals centered on her child. When she delves into her past, where meaning-making might aid some individuals, she seems to lose the coherence of her current agentic self. For example, in her self-defining memory she reported,
Ummm well uh, it’s taken me years to know even who I am at all. Like ummm underneath everything I’ve done like being high all the time or drunk . . . I still don’t know who I am today. I mean like I know my physical name and the fact of where I live and stuff but like not deep down inside and it gets pretty bad for me.
Perhaps Josie can coherently tell her story if she doesn’t get into the details of her past—otherwise her perception of herself as being a responsible adult and mother falls apart. That is, perhaps she will be able to soldier on without reflection (e.g., McLean & Mansfield, 2011; Sales et al., 2013). However, this is very much an empirical and longitudinal question. Alternatively, she may need to engage in reflection in order to bring coherence to her understanding of self, perhaps especially concerning her role as a parent. If she is not able to find a coherent story about her past, she may compromise her abilities to provide a healthy foundation for her own child (e.g., Bowlby, 1969).
Implications
One of the clearest lessons that we draw from these data is that we know little about this important group—adolescents whose life histories and behavior make them highly vulnerable—and they do not appear to follow the same patterns as normative adolescents or adults. These data suggest that there may be other strategies used to navigate psychosocial development. Indeed, while a large body of research has shown that the construction of the meaning of the personal past can facilitate and encourage growth and resilience (see Pals & McAdams, 2004), we suggest that these stories can also define the individual in rigid ways that may actually compromise resilience. That is, story construction can also work to constrain possibilities for growth by confirming negative attributions, beliefs, and expectations (Banks & Salmon, 2013). Further, the limits of these stories are not only personal, but also societal and structural. The realities of poverty, loss, a lack of healthy social opportunities, neglect, abuse, and other forms of victimization all limit the ability to construct stories with the power to transform. We will address these ideas in turn, though we first discuss some limitations.
First, we acknowledge our small sample size, particularly in determining which correlations might be significant with greater power. Further, agency and meaning were correlated, suggesting that in some cases meaning-making may be an agentic act. However, our control analyses suggest that there is something special about agency in this particular sample (see also Sales et al., 2013). We also note the importance of the lack of a longitudinal design to examine the role of the narrative processes in predicting behavior and psychological health over time. Relatedly, measuring desistance is difficult. Those in our sample who were desisting may still reoffend, suggesting that desistance can take many failures before success, and even then, there is no clear consensus among researchers for defining success in overcoming delinquency (Laub & Sampson, 2001; Maruna, 2001). As suggested by life course models of criminal behavior, it may not be possible to determine desistance in a youth sample, as criminal behavior may reoccur some months or years in the future. Thus, while we can suggest that those participants who report ceasing to engage in delinquent behaviors are “on the right track” towards desistance, we cannot determine that they have desisted. Further, in our sample, and in the broader literature, issues of maltreatment, delinquency, and poverty are often entangled. For example, while delinquency and abuse occur in families from all social-economic backgrounds, stressors associated with poverty heighten risk for both negative family functioning and conduct problems across the life course (e.g., Cicchetti & Howes, 1991). Given these complexities, we took a holistic view of these individuals, and did not attempt to disentangle the relative influences of each of these issues.
While we found that meaning was not associated with desistance, it is possible that over time meaning could be associated with more resilient functioning in the future. Positive change is a process (Breen, 2010), and story construction and behavioral shifts take time to develop, such that the ability to reflect may be what is important in the long run, rather than the conclusions one draws at the time. While this is an optimistic hypothesis, we are not sure it will pan out for the individuals in our sample.
First, the number of adversities that one faces, and the nature of these challenges, may preclude more positive meaning-making (Lilgendahl & McAdams, 2011). For example, Sanderson and McKeough (2005) found that adolescents who were behaviorally troubled reported more negative life events, compared to a low-risk sample, and their negative life events were also qualitatively more traumatic than the low-risk sample. In particular, they found that in describing those past traumas they were likely to construct their narratives around a victim role, employing a more passive stance, echoing our findings and others about the importance of agency. Further, in impoverished communities, some of these delinquent behaviors may be seen as viable routes to autonomy and independence, lessening the appeal of desistance.
Second, there are certainly an abundance of developmentally critical interpersonal processes that were likely compromised in this sample and may contribute to the processes of narrative construction employed by these youth. For example, the description of family life suggests that some of the youth may have experienced a disorganized attachment relationship, which may result in some of the fragmentation and incoherence in the narratives (see Solomon & George, 1999). Many of the families may not have been able to appropriately cope with and discuss the traumatic pasts to which these youth were exposed, limiting the opportunities for appropriate scaffolding of self in relation to the past (Fivush, Haden, & Reese, 2006). Appropriate contexts within which to practice the dual needs of autonomy and relatedness (e.g., Allen, Hauser, Bell, & O’Connor, 1994) may also have been missing.
However, these findings speak to broader social and political rhetoric about poverty. It is becoming increasingly difficult, perhaps impossible, to overcome the experience of poverty and other forms of structural violence (e.g., Appleyard, Egeland, van Dulmen, & Sroufe, 2005). If stories reflect and produce experience, the real lack of opportunity might be reflected in the absence of alternative narratives. Family and community stories—the “master narratives” that shape and inform the stories that individuals construct—may themselves be constrained by structural violence. Cycles of poverty are not just social and structural, but also narrative (Fine & Weis, 2008).
However, the silver lining in the present study was that the more one was able to form a story that centered on taking some degree of personal responsibility in one’s life, the more likely one was to evidence desistance from delinquent behavior. In thinking about the role of agency, McAdams has argued for a developmental story of self and identity that focuses on the roles of actor, agent, and author (see McAdams & Cox, 2010). Before a self even develops there is an actor, visible by temperaments, which later become traits. What comes next is the role of agent, in which the self has needs, desires, and goals. At school age, children become better able to plan and are motivated to organize their environments to meet their needs. In adolescence and emerging adulthood the self becomes an author.
While it is clear that in low-risk samples of adolescents the author self is emerging, as assessed by indicators of meaning-making (e.g., McLean et al., 2010; McLean & Breen, 2009; see also de Silveira & Habermas, 2008), it is possible that in the current sample our participants were still working on their agent selves. Difficult and traumatic pasts may create additional barriers to finding a sense of temporal continuity and coherence (see also Chandler, Lalonde, Sokol, & Hallett, 2003). These barriers may be related to the neglectful and compromised parenting many of them experienced, as well as the victimization and powerlessness that their parents likely experienced in the face of structural violence. As well, these adolescents may have other, less obvious barriers to processes of meaning-making, such as language impairments and psychiatric disturbances, both of which occur at elevated rates among populations with conduct problems (e.g., Cohen, Barwick, Horodezky, Vallance, & Im, 1998; Cohen et al., 1998). An optimistic view would be that as they develop an understanding of their own agency, these adolescents may be able to move on to an understanding of their role as authors. A more pessimistic view is that they might continue to experience social, family, and learning environments that limit their opportunity for agency and meaning-making. Their pasts may thus limit the ability to develop their own “voice” as a self within narratives of past experience, as well as “voice” in terms of being heard by others and larger power structures (see Fivush, 2010).
Researchers have viewed the dynamic between master narratives and one’s individual story as a dance of positioning the individual towards or against one’s larger culture (Hammack, 2011; Thorne & McLean, 2003; Weststrate & McLean, 2010), such that one can take the larger story structures available and use them to story one’s own life, or resist those master narratives and conform to a counter-narrative (Weststrate & McLean, 2010). McAdams (2006) has argued that the story of redemption—when bad turns to good—is a prototypical narrative to which Americans are particularly responsive, and is viewed as a legitimate, and encouraged, way to story hardship. McAdams (2006) argues that this narrative has a long history in America, including in oppressed and marginalized groups. Indeed, we see the master narrative in Josie’s comments about the possibilities for change: “ . . . if you look at the pattern of life everything eventually does change. I mean we made history recently about having a president who wasn’t Caucasian . . . I think that if that can happen then anything can happen.”
However, we wonder how much redemption these adolescents see in their lives, aside from that in the media. Are these adolescents living a life in a world where negative life trajectories are more common models than those who pull themselves out of risky contexts? Most of the adolescents referenced abuse, abandonment, neglect, and poverty in talking about their families and we wonder about the role of generations of marginalized families living with a master narrative of pulling oneself up by the bootstraps, when that is not a proximally visible option. In discussing classism Jones (2003) discusses the bootstrap narrative as “a belief that people can escape poor and working-class origins if they apply themselves. This belief supports individual efforts toward improving one’s economic situation. However, it also blinds one to oppressive structural forces that give some people an advantage over others.” (p. 816)
Thus, is there a competing master narrative of negativity and contamination in this subgroup? If there is, it might be doubly challenging given that the larger narrative is one of redemption. For example, in Hammack’s (2011) recent work with Palestinian and Israeli youth, he suggests that to change society, personal narratives and community narratives must coalesce. For the youth in his studies, being personally defined by conflict and war makes peace difficult because of the personal investment in the identity of conflict. In our study there are some ways to change the narrative, such as the discussion of the hopeful narratives that are produced by many young mothers (see also Breen, 2010; Breen & McLean, 2010), but options in these populations are more limited. It may be that the community narrative, and real opportunities for change, needs to shift so that these adolescents are able to move towards other alternatives. For example, Ungar (2008) talks about resilience as partially dependent on the environment’s ability to provide the resources necessary for the individual to overcome adversity. In this way, an individual cannot be resilient unless he or she has access to the resources necessary to become resilient (e.g., vocational opportunities, safety, positive, and stable relationships).
In closing we turn to a story to exemplify the idea of the multiauthored life story—with both the individual and the micro- and macro-culture playing roles in story construction. While spending time at this school, the second author learned of a recent episode that provides an example of changing a personal story by changing a larger story. Most of the residents of the town that houses this school have a negative view of the school and the youth who attend it. A park in town had been vandalized and students from the school were accused of doing it (by police and in a letter to the editor in the local paper), though there was no evidence that they were the culprits. Several of the youth from the school had taken on the task of trying to change the negative perceptions of the school. One student had even responded to the negative letter to the editor with her own letter, titled “Don’t Judge the School by Its Reputation.” In her letter, this student reported on the good works that the students were engaging in, including adopting the park that was vandalized and working to clean it up. In closing she stated, “It’s not fair to zero in on this one school just because it doesn’t have the best reputation, especially when we have been trying so hard to make our image better and to make this park enjoyable for everyone.” In this small effort, this student and her peers were trying to rewrite their story on a more global level, not just a personal level, and that may be exactly the kind of agency these youth need—both to see themselves as agentic and to allow the possibility for others to see them as having a valid voice in the community.
Footnotes
Acknowledgements
We would like to first acknowledge our participants who were so generous in sharing their stories with us, as well as the school for working with us to complete this study. We also thank Western Washington University for helping to fund this study. We thank the Narrative Lab at WWU, especially Meghan Bornheimer, and the Breen Lab, especially Kristen Cairney, for assistance in developing and applying our coding systems. We also thank Rebecca Goodvin and Lew Jones for providing critiques of an earlier version of this article. Finally, we thank the Society for Personology, who discussed this article at the 2012 meetings.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Western Washington University.