Psychology Research Paper

The Development of Narrative Identity in Late Adolescence and Emergent Adulthood: The Continued Importance of Listeners Monisha PasupathiUniversity of Utah Timothy Hoyt University of New Mexico Research on narrative identity in late adolescence and early adulthood has not extensively examined how conversational storytelling affects the development of narrative identity. This is a major gap, given the importance of this age period for narrative identity development and the clear importance of parent– child conversations in the development of narrative identity. The authors present a series of 3 studies (n 220) examining how late adolescents and early adults construct narrative identity in ways that are shaped by their listeners. The findings suggest that late adolescents and early adults construct more meaning-laden, interpretive accounts of their everyday experiences when they converse with responsive friends. Further, even within this sample’s abbreviated age range, the authors found evidence for age-related increases in the factual content of personal memories. Such findings illuminate the importance of friends in the construction of narrative identity during this key developmental period.

Keywords:narrative identity development, autobiographical memory development, parent– child remi- niscing, social construction Late adolescence and early adulthood are accorded special sig- nificance for the development of self and identity (Arnett, 2000; Dusek & McIntyre, 2003; Erikson & Erikson, 1997; Harter, 1998; Kroger, 2003), narrative identity (Habermas & Bluck, 2000; McLean, 2005), and autobiographical memory (Holmes & Con- way, 1999; Rubin & Schulkind, 1997). In fact, work on the reminiscence bump suggests that adolescence and early adulthood are a privileged developmental period for the encoding of auto- biographical memories, which in narrative terms constitute the stuff of which selves are made.

But despite the widespread acknowledgment of the importance of late adolescence and early adulthood for identity development, there are some important gaps in existing research. Perhaps most notable among these is a relative lack of attention to the micro- processes by which identity in general and narrative identity in particular develop during this period. In the present article, our focus is particularly on the microprocess of conversational story- telling as a critical process by which narrative identity more broadly develops. Further, we show, across three studies, that having responsive, attentive friends as listeners for conversational storytelling helps further narrative identity development in late adolescence and early adulthood. Next, we outline the unaddressedissues in identity development work that can be better examined with the narrative and, more specifically, conversational storytell- ing frameworks. Identity Development in Adolescence and Early Adulthood Erikson and Erikson (1997) originally defined ego identity as entailing a sense of uniqueness or individuality, an emerging commitment to a place in society, and a sense of continuity over time. Much of the extant work on identity development, as well as work on self-development, has emphasized the uniqueness or individuality aspect of self and identity by focusing on explicit self-descriptive statements that participants do or do not en- dorse. Research in this arena has suggested that identity devel- opment in terms of exploring alternative identities and commit- ting to some rather than others is a nonlinear process that appears to occur primarily in late adolescence and early adult- hood (Constantinople, 1969; Meilman, 1979; Waterman, 1982).

Over adolescence and early adulthood, self-descriptions be- come increasingly multifaceted and complex, and changes in the content of self-descriptions also occur into early adulthood (Dusek & McIntyre, 2003; Harter, 1998; Sutin & Robins, 2005).

The ability to recognize contradictions in the self emerges by midadolescence, but the capacity for resolving them is a phe- nomenon of later adolescence (Harter & Monsour, 1992), con- sistent with increasing cognitive and epistemic abilities across this period (e.g., Kitchener, Lynch, Fischer, & Wood, 1993).

Contextual factors such as parent– child relationships, peer re- lationships, and exposure to college and larger sociopolitical contexts are linked to identity development over this age range (Dusek & McIntyre, 2003; Hair, Moore, Garrett, Ling, & Cleve- land, 2008; McNulty & Swann, 1994; Waterman, 1982).

This literature underscores the importance of late adolescence and early adulthood for identity development but is relatively Monisha Pasupathi, Department of Psychology, University of Utah; Timothy Hoyt, Department of Psychology, University of New Mexico.

These studies were supported by National Institute of Mental Health Grant 1R03MH64462 awarded to Monisha Pasupathi. Our gratitude is due to Benjamin Rich, for his extensive efforts in data collection, as well as to Martin Cryer and Leslie Rheinhold, who assisted with data coding. In addition, Carol Sansone, Kate McLean, and Frank Drews provided valu- able comments on earlier drafts of this article.

Correspondence concerning this article should be addressed to Monisha Pasupathi, Department of Psychology, University of Utah, 390 South 1530 East, Beh-502, Salt Lake City, UT 84112. E-mail: [email protected] Developmental Psychology© 2009 American Psychological Association 2009, Vol. 45, No. 2, 558 –5740012-1649/09/$12.00 DOI: 10.1037/a0014431 558 uninformative about the microprocesses by which identity is con- structed in that age period and does not capture the way in which people construct a sense of themselves as continuous across time.

As an alternative, narrative approaches to identity development capture both individuality and the continuity aspects of identity, and their focus on the construction of identity is one step closer to an examination of the processes of identity formation. Narrative Identity in Late Adolescence and Early Adulthood: Closer to Process Narrative approaches to identity development examine how people construct meaning in relation to their experiences and thus further their sense of self and identity (McAdams, 1996). The construction of narratives about past experiences also involves the creation of a sense of continuity over time—as the past self is represented and interpreted by the present self. Thus, narrative conceptualizations of identity hold the promise of integrating Erikson and Erikson’s (1997) aspects of identity as well as illu- minating the processes by which individuals develop identity. For example, narratives construct personal continuity over time by linking past to the present and future. They also highlight individ- ual uniqueness and, at the same time, make use of cultural scripts and schemas for organizing one’s experience in narrative form. In fact, proponents of narrative identity research (McLean, Pasupathi, & Pals, 2007; Pasupathi, 2001; Thorne, 2000) have argued that the process of constructing narratives about the personal past is the paramount process by which narrative identity in particular, and self and identity more broadly, develop. In making this case, they specifically suggest that narrative identity develops via the micro- process of constructing specific narratives in specific situations.

That is, narrative identity emerges out of countless actions of narrative construction. The study of narrative identity development consequently requires the exploration of processes that influence narrative identity creation in the moment.

It is important to note that narrative and narrative identity are not synonymous. Although all stories must provide an account of what happened—the setting and actions that occurred— each sto- ryteller may have a unique sense of how the actions were con- nected, which were important, and what the broader implications and associations of the experience may entail. These interpretive, meaning-laden features of a narrative render a set of facts uniquely reflective of the individual whose story they compose, and it is these features that both reflect and construct narrative identity.

Moreover, it is these features that across various operationaliza- tions are related to macrolevel development of self and identity in early childhood (e.g., Bird & Reese, 2006; Bohanek, Marin, & Fivush, 2006; Fivush, 1991; Fivush, Bohanek, Robertson, & Duke, 2004; Harley & Reese, 1999; Howe, Courage, & Peterson, 1994) and in adolescence and early adulthood (McLean & Pratt, 2006; see also Sutin & Robins, 2005), and to well-being and maturity throughout adulthood (King & Patterson, 2000; McAdams, Reyn- olds, Lewis, Patten, & Bowman, 2001; Pals, 2006a, 2006b).

The overall elaboration of narratives, including both factual and interpretive information, increases across childhood (e.g., P. J.

Bauer, 2006; Fivush & Nelson, 2004). Little is known about changes in general elaboration thereafter. Researchers focusing on adolescence and adulthood have emphasized broader, global types of meanings; these are conceptually connected to interpretiveelaboration. The prevalence of those larger identity-related ele- ments in narratives increases with age into middle adulthood (J. J.

Bauer & McAdams, 2004a, 2004b, 2005; Bluck & Glueck, 2004; McAdams et al., 2006; McLean, 2005; McLean & Thorne, 2003; Pasupathi & Mansour, 2006), and they become more positive and more nuanced (McAdams et al., 2006). This work suggests, but does not explicitly demonstrate, increases in the prevalence and sophistication of interpretive elaboration in narratives from ado- lescence into middle adulthood. Moreover, this work is largely mute about just how such macrolevel increases might come about; that is, it does not necessarily investigate the processes by which people come to vary in their construction of narrative identity. Conversational Storytelling: The Process of Constructing Narrative Identity Conversations about the personal past with caregivers are a major force in shaping young children’s ways of constructing narratives and developing narrative identity (see, e.g., P. J. Bauer, 2006; Fivush & Nelson, 2004). Some parents engage children in highly responsive, elaborative conversation about the past, whereas others are more focused and repetitive (Cleveland & Reese, 2005; Fivush & Nelson, 2004; Harley & Reese, 1999; Reese & Fivush, 1993; Wang, 2004). Children with more elabo- rative mothers subsequently engage in more elaborative remem- bering with other listeners and about other events. Maternal lis- tening, then, influences the story created within that conversation but over time exerts cumulative, long-term effects on children’s style of remembering. Converging findings from experimental (Peterson, Jesso, & McCabe, 1999) and prospective longitudinal designs (e.g., Haden, Haine, & Fivush, 1997) provide strong evidence for a causal role for parent– child remembering. For very young children like those studied in this work, parental respon- siveness is important for both factual and interpretive elaboration in narratives. Further, the development of basic capacities for narration and the development of narrative identity are tightly intertwined during early childhood.

Within a burgeoning literature on narrative identity in late adolescence and early adulthood (e.g., McLean et al., 2007), however, we know surprisingly little about the role played by conversational storytelling and the listeners for such stories. This is in spite of the acknowledged importance of identity develop- ment in this age range and the fact that this period coincides with a shift from parents as the primary audience for personal story- telling to friends as an additional, increasingly important audience for such stories (McLean, 2005; see also Buhrmeister, 1996; Cooper, 1999; Daddis, 2008; Hartup, 1996; Hazan & Zeifman, 1994; Updegraff, McHale, & Whiteman, 2006).

Some tiny pieces of evidence from studies of adolescents and emergent adults suggest that listeners are important in the con- struction of narrative identity in this age range. For example, conversational interactions with parents do matter. More open and free-ranging conversations with parents are associated with greater autonomy and identity development in adolescents (Cooper & Grotevant, 1985). By early adulthood, friends and parents appear to serve similar purposes as listeners for personal events, according to retrospective reports (McLean, 2005). Further, conversations with friends during adolescence are quite strikingly concerned with identity (Korobov & Bamberg, 2004; Parker & Gottman, 559 NARRATIVE IDENTITY DEVELOPMENT 1989), and work with late adolescents suggests that listeners prefer narratives that entail more sophisticated identity-related insights— indicative of a listener preference for more interpretive elaboration (Thorne, McLean, & Lawrence, 2004). Some work with young adults has linked the action of telling a personal story to changes in more traditionally assessed self-conceptions or event percep- tions and has shown that variations in the way listeners respond to that story, or variations in the goals with which the story is told, correspond to immediate self-perception changes (e.g., McGregor & Holmes, 1999; McLean, 2005; Pasupathi, Alderman, & Shaw, 2007; Pasupathi & Rich, 2005; Thoman, Sansone, & Pasupathi, 2007).

A handful of studies have explicitly looked at variations in listener behavior and storytelling among young adults (Bavelas, Coates, & Johnson, 2000; Dickinson & Givon, 1995; Pasupathi, Stallworth, & Murdoch, 1998). These studies examined story- telling by young adults to a stranger and experimentally varied the stranger’s behavior during storytelling. That variation cre- ated either distracted or disagreeable listeners; the comparison conditions always involved a responsive, agreeable listener.

Generally, responsive and agreeable listeners concurrently elicit more elaborative stories than do distracted or disagreeable listeners, whether about brief film stimuli (Dickinson & Givon, 1995; Pasupathi et al., 1998) or about personal experiences (Bavelas et al., 2000). In fact, listeners in the responsive and agreeable conditions behave more elaboratively than do listen- ers in the distracted conditions, in these studies. That is, they ask questions, make sympathetic exclamations, demonstrate that they have understood the speaker’s point, and follow the speaker’s interests rather than impose their own agenda.

These behaviors are quite similar to those observed among more elaborative mothers in work on parent– child reminiscing.

One goal of the present studies was to demonstrate that the development of narrative identity is affected by the process of storytelling, by showing that the extent to which late adoles- cents and early adults construct elaborative, richly detailed personal stories is connected to the behavior of their listening friends.

But as we noted earlier, not all elaborations are equally relevant for narrative identity. Interpretive information in par- ticular makes memory into narrative identity by tying recalled experiences to the individual’s goals, feelings, thoughts, and beliefs. In childhood, when memory conversations involve learning how to remember, responsive listening is clearly linked to elaboration of all types of information, although for self- development the elaboration of interpretive or evaluative con- tent is more important. By contrast, in adulthood, responsive listening may be of particular importance for interpretive infor- mation. In fact, Bavelas et al. (2000) showed that distracted (and therefore unresponsive) listeners were selectively impaired in their capacity to respond to the meaning of what storytellers were saying, although they did not examine whether this cor- responded to a selective suppression of storytellers’ inclusion of such meanings. Thus, our second aim was to examine whether responsive friends were especially important in facilitating the construction of richly interpretive accounts of personal experi- ences, that is, in helping late adolescents infuse memory with narrative identity. Overview of the Present Studies The major goal of the present studies was to provide laboratory- based evidence for the microprocess. Critically, these studies were not designed to address the macrodevelopmental process of nar- rative identity process but, rather, to address the critical micropro- cess building block for narrative identity—the process of conver- sational storytelling. Other work suggests that if responsive listening has immediate importance for narrative identity in the short term, it will also have long-term implications for macrolevel self-development, an issue to which we return in the discussion of our findings.

The present studies examine in vivo storytelling about personal experiences to close, same-gender friends by late adolescents and early adults (defined as the age range from 18 to 35 years). We chose to examine storytelling to same-gender close friends because narrative identity suggests that close friends and family are the primary audience for personal storytelling in adulthood and be- cause close friends and parents appear to serve similar roles as listeners among late adolescents and early adults (McLean, 2005).

Restricting the focus to same-gender friendships allowed us to reduce the possible impact of different relationships between lis- teners and speakers (i.e., romantic vs. primarily platonic relations).

We compared responsive and unresponsive (distracted) friends during storytelling across three studies, examining how variations in responsiveness influenced the elaboration of factual and inter- pretive aspects of storytelling.

We chose experimental methods because exploratory work in our laboratory suggested that under normal conditions, most friends are fairly responsive listeners in a laboratory setting in which there are few distractions and their defined task is to listen.

Experimentally creating an unresponsive listener via distraction ensured that some of our participants would not function like highly elaborative mothers by asking questions or expressing strong interest. Experimental methods also permit stronger con- clusions about causality and link the present studies to work on young adults.

In addition to examining our major goals, we also examined two other issues in the present studies. First, although age differences, particularly in this restricted age range, were not a primary focus of our study, we examined whether age was associated with increases in the elaboration of factual or interpretive information, given other findings on autobiographical memory development during this age period. Second, by late adolescence and beyond, people also have the capacity to engage in deliberate, strategic self-presentation (e.g., Baumeister, Stilman, & Wotman, 1990; Tice, 1992). Such deliberate self-presentation would suggest that people strategically change their stories, and strategic changes in narration imply a different type of identity-related implication than do changes in narrative identity construction that occur without a great deal of self-awareness on the part of the individual. Study 1 Study 1 provides a first examination of how late adolescents and young adults construct narrative identity around positive experi- ences, in collaboration with listeners who vary in their responsive- ness. We expected that late adolescents and emergent adults talk- ing with distracted friends would construct less elaborated stories 560 PASUPATHI AND HOYT in general but would be especially likely to construct less inter- pretive accounts. Method Participants Participants were 40 same-gender pairs of friends (50% men, 50% women) recruited from the introductory psychology subject pool at the University of Utah. To recruit pairs of friends, we first recruited target participants (assigned to the speaker role), and these participants were asked to bring a close, same-gender friend to the experiment. The average age of speakers was 21.6 years (SD 3.2). Participating friends had known each other an average of 3.2 years (SD 3.8). 1The majority of participants were European American (90%). Pairs were randomly assigned to an attentive or distracted listener condition; conditions did not differ in the length of time friends had known one another or in ratings of friendship quality. Equipment failure caused loss of transcripts in two cases. Sample sizes for analyses reported below vary accordingly. Procedure After arriving, participants completed several measures, includ- ing personality, relationship quality, and background measures.

Speakers identified recent undisclosed events they considered to be positive, selected one event to describe to the friend, and provided some preliminary ratings of the events. Meanwhile, friends were separated from speakers for instruction. Attentive friends were told, “Listen to your friend the way you typically do when you’re being a good listener.” Distracted friends were told, “We are interested in how conversations go when one person is distracted.” They were asked to count all words beginning with the lettersthwhile listening to the speaker and practiced this task on 10 prepared stories read aloud by the experimenter, who provided feedback about task accuracy after each story. Listeners were offered an incentive of $5 if they were accurate in counting words beginning withth, with accuracy defined as coming within a 4-point margin of error during the subsequent conversation with the speaker. This target was chosen on the basis of listener per- formance in the training session, because listeners were rarely especially accurate but achieved this standard of performance at a rate of about 50% of the time. Pilot work revealed that participants proved to have difficulty ignoring their friend’s storytelling in favor of thethtask, and the incentive improved their level of distraction. The dyad was then reunited and videotaped while the speaker described the chosen experience to the listener. Following the conversation, the speaker completed additional ratings of lis- tener responsiveness and agreeableness, and rated the extent to which the story was detailed and coherently told. Participants were fully debriefed and given information about both listener condi- tions, and were told about our interest in how listeners influence storytelling.

Measures Manipulation check.As a manipulation check, we obtained speaker ratings of the extent to which the listener agreed with his or her version of events (a single item) and of the listener’sresponsiveness (an average across five items, including, for exam- ple, “The listener was very responsive”; Cronbach’s .84).

Agreement and responsiveness often covary, but they are concep- tually distinct (e.g., Pasupathi, Carstensen, Levenson, & Gottman, 1999). Both were assessed because manipulations of listener re- sponsiveness sometimes influence both.

Speaker’s perceptions of story quality.To assess deliberate changes in storytelling, we obtained speakers’ ratings of the qual- ity of the story they told across nine items, using 7-point Likert- type scales in which 1 indicated lower quality and 7 higher quality.

Example items include “My story used few details/many details” and “My story was very incoherent/very coherent.” These nine items were averaged to create a single index of story quality (Cronbach’s .90).

Coding of conversational data.Each recorded conversation was transcribed by undergraduate research assistants. Transcripts were then divided into “idea units” that roughly corresponded to verb phrases. Idea units were then coded according to their con- tent. We examined two main types of narrative content. The first main category was event factual information, that is, information available to the perceptual capabilities of any bystander. Examples of this category include such as statements “Mom and I had a talk while you were gone,” “We talked for almost two hours,” and “She started crying and everything.” The second category wasinterpre- tiveinformation, or information that emphasizes the subjective, meaning-laden, and internal aspects of experience—in short, nar- rative identity. Examples of this category include “It was just a good talk,” “She’s always sappy like that,” “We have become closer and everything,” and “It makes it feel like things have really changed since I left for school.” Additional categories included statements aimed at regulating the conversation (establishing un- derstanding, providing minimal back channels, etc.), off-topic statements (references to the experimental setting), and uncodable statements. The Appendix presents an entire conversation, in units and coded.

To establish reliability between coders, the two authors first independently coded several conversations. These independently coded sets were then compared by means of Cohen’s kappa. Once the authors established substantial independent agreement ( .91) and resolved remaining discrepancies through discussion, the second author then trained a team of two additional undergraduate coders. After learning the coding scheme, the coding team applied the coding scheme to the 38 conversations in the study. The coding team held weekly meetings in which problematic coding units were discussed and agreement on content was sought. Fifteen percent of the conversation pairs were coded independently by two coders to quantify reliability across the full set of content catego- ries in the coding scheme. Across 373 idea units, coders agreed 77% of the time on the content category ( .68). For each pair, we computed the number of speaker-generated facts and evalua- tions. These content code variables served as the primary depen- dent measures for our analyses. 1The length of time partners had known one another was unrelated to any of the results reported here; nor did relationship quality have any impact on any of the findings. 561 NARRATIVE IDENTITY DEVELOPMENT Results Manipulation Check A general linear model examining the effects of experimental condition and gender on perceived listener attentiveness and agree- ableness revealed only an overall effect of condition,F(2, 37) 26.8,p .01, p 2 .59. Univariate tests showed that condition effects were evident for perceptions of listener attentiveness,F(1, 38) 6.3,p .02, p 2 .14, and for perceptions of listener agreeableness,F(1, 38) 54.9,p .01, p 2 .59. Speakers perceived distracted listeners to be less responsive (attentive:M 5.4,SD 1.0; distracted:M 3.1,SD 1.0) and less agreeable (attentive:M 6.1,SD 1.4; distracted:M 4.9,SD 1.7). Impact of Listener Responsiveness on Amount and Type of Information Recalled We expected that when faced with distracted listeners, our partic- ipants would construct shorter stories in general, and especially for interpretive, as opposed to factual, content. To test this hypothesis, we computed a general linear model (multivariate analysis of variance) with the number of speaker-generated facts and evaluations as depen- dent variables, listener condition and gender as between-subjects variables, and type of information (facts or evaluations) as a within- subjects variable. The results revealed significant main effects of condition,F(1, 34) 5.2,p .03, p 2 .13, and type of information, F(1, 34) 6.0,p .01, p 2 .15. We also computed simple effects tests examining the effect of listener condition within each type of information separately. These contrasts revealed that distracted listen- ers elicited significantly fewer facts,F(1, 34) 5.2,p .03, p 2 .14, and tended to elicit fewer interpretations,F(1, 34) 2.9,p .10, p 2 .08 (see Table 1). These effects are shown in Figure 1. As seen here, factual information predominated. Distracted listeners elicit less information overall, and given the absence of a significant interaction between information type and listener condition, the information decrement appears consistent across information type, contrary to our hypotheses. Are Speakers Aware of the Impact of Listeners on Their Stories?

To evaluate whether participants were aware of the impact of listeners on their stories, we computed an analysis of variance examining perceived story quality as a function of listener condi- tion and gender. There were no differences in self-reported story quality as a function of listener condition,F(1, 36) 1,ps .40 (attentive:M 3.8,SD 1.4; distracted:M 3.8,SD 1.0).

There were also no significant effects involving gender. Cross-Sectional Age Differences in Story Content To examine age differences in the content of participants’ sto- ries, we computed a partial correlation between age and factual and interpretive units, controlling for the perceived attentiveness of listeners. Older participants tended to report more facts than younger participants,r(35) .31,p .07, but they did not differ significantly in their inclusion of interpretive information,r(35) .18,p .20. These correlations are not significantly different from each other, however, and with greater power both types of infor- mation may in fact increase with age. Study 2 The results of Study 1 suggested that (a) late adolescents and emergent adults constructed less rich and elaborated stories when they spoke to distracted friends; (b) these same participants’ rat- ings of their own stories did not differ as a function of their friends’ behavior; and (c) across even this restricted age range and with limited power, there were possible age differences in the elaboration of factual information in conversational narration.

Our initial expectation, that distracted listeners would be especially problematic in relation to interpretive content, was not confirmed.

However, we had relatively little statistical power given our small sample size and, in particular, given that the stories told in this study had more factual than interpretive content. Further, we had examined only positive events in this study, whereas earlier work on narrative identity emphasized the potential for negative events to involve self- construction in childhood and, to some extent, in adulthood (Bird & Reese, 2006; Thorne et al., 2004). Thus, in Study 2, we expanded our focus to examine both positive and negative events. Participant pairs were randomly assigned to a positive or negative event condition.

Moreover, the addition of a negative event condition also increased our statistical power for examining age differences. Aside from this change, Study 2 was an exact replication of Study 1. Method Participants Participants in this study were 81 same-gender friend dyads (50% men, 50% women) recruited from the introductory psychol- ogy subject pool and through flyers posted at the University of Utah. The average age of speakers was 19.6 years (SD 1.9), again, firmly within the age range of late adolescence and early adulthood. Within dyads, participants had known each other an average of 4.7 years (SD 6.0). 2One pair was excluded because the speaker was not a native speaker of English, leaving a total sample of 80 pairs. The sample predominantly comprised Euro- pean American participants (76.5%), with Hispanic participants composing an additional 12.3% of the sample. Participants were randomly assigned to one of two event-valence conditions (posi- tive or negative) and to one of two listener conditions (attentive or distracted). As in Study 1, there were no differences by condition in the length of time participants had known one another or in their ratings of relationship quality. 2The length of time that partners had known one another and ratings of the quality of their relationship did not influence the results presented below. Table 1 Means and Standard Deviations of Dependent Measures in Study 1 MeasureAttentive Distracted MSDMSD Facts 40.1 31.6 21.0 15.0 Evaluations 29.1 23.8 17.7 15.1 562 PASUPATHI AND HOYT Procedure The procedure was identical to that in Study 1, except that half the participants assigned to the speaker role were asked to identify recent, undisclosed negative events, and the other half were asked to identify recent, undisclosed positive events.

Measures Manipulation check.Speakers rated their perceptions of the listeners’ responsiveness across the same five items used in Study 1 (Cronbach’s .83) and rated their perception of the listeners’ agreement with their views on a 7-point Likert-type scale, again, as in Study 1.

Speaker’s perceptions of story quality.Speakers rated their perceptions of story quality along the same nine items employed in Study 1, and we averaged the items to create an index of perceived story quality. The internal consistency of this measure was lower in this study than in Study 1 (Cronbach’s .68).

Coding of conversations.Coding employed the same system as in Study 1. Twenty-five percent (n 20) of the conversation pairs were coded independently to quantify reliability; coders agreed on about 82% of 1,169 idea units ( .75). As in Study 1, we computed the number of speaker-generated event facts and evaluations as primary dependent measures. Results Manipulation Check A general linear model testing for effects of listener condition (attentive vs. distracted), event valence (positive or negative), andgender on the speaker’s perceptions of the agreeableness and responsiveness of the listener revealed a main effect of listener condition,F(2, 72) 21.4,p .01, p 2 .37, and a main effect of gender,F(2, 72) 4.3,p .02, p 2 .11. Examination of the univariate effects for these two main effects suggested that listener condition influenced perceptions of listener responsiveness,F(1, 73) 42.9,p .01, p 2 .37, but not perceptions of listener agreement,F(1, 73) 1.2,p .20. As expected, listeners in the distracted condition were perceived as less responsive (M 3.6, SD 1.2) than listeners in the responsive condition (M 5.3, SD 1.2). Gender was also related to perceptions of listener responsiveness,F(1, 73) 4.8,p .05, p 2 .06, but not to perceptions of listener agreement,F(1, 73) 1.8,p .15. Across conditions, women rated their listeners (who were also women) as more attentive (M 4.7,SD 1.6) than did men (M 4.2,SD 1.3). Our manipulation of listener behavior was successful, as perceived by the speakers, but had a more narrowly focused impact on responsiveness than in Study 1. Note that even testing for effects of listener condition on agreeableness within the positive event condition did not result in perceptions that listeners differed in agreeableness. Differences in attentiveness by gender did not interact with the manipulation. Do Speakers Change Their Story Depending on Listener Condition and Event Valence?

We conducted a general linear model with information type (factual, interpretive) as a within-subjects factor and gender, lis- tener condition, and event valence as between-subjects factors.

There were main effects of type of information,F(1, 72) 4.4, p .05, p 2 .06, and gender,F(1, 72) 4.0,p .05, p 2 .05, 0 5 10 15 20 25 30 35 40 45 50 s n o i t a t e r p r e t n I s t c a F Type of Information Number of Idea Units AttentiveDistracted Figure 1.Effect of distracted versus attentive friends on information during conversational remembering in Study 1. Error bars represent standard errors of the mean. 563 NARRATIVE IDENTITY DEVELOPMENT and a trend toward a main effect of listener condition,F(1, 72) 3.8,p .06, p 2 .05. In addition, there was an interaction of listener condition and type of information,F(1, 72) 5.6,p .03, p 2 .07.

In contrast to the results of Study 1, conversations focused more on interpretive information (M 33.8,SD 48.8) than on facts (M 23.9,SD 23.4), and women produced a greater number of units on average, across all types of information (M 71.6,SD 85.1) than did men (M 43.1,SD 23.2), which is consistent with the fact that women in this study rated their listeners as more attentive than did men across all conditions. Attentive friends tended to elicit longer accounts (M 72.0,SD 85.2) than did distracted listeners (M 43.4,SD 26.4), but this effect was qualified by an interaction with information type. To follow up the interaction effect, we tested the simple effect of listener condition within each type of information. Only one of these simple effects tests was significant: that for interpretive information,F(1, 72) 5.3,p .03, p 2 .07.Participants told attentive friends nearly twice as many interpretive units (M 45.7,SD 64.9) than they told to distracted listeners (M 21.1,SD 16.9). Figure 2 displays the amount of information contained in participants’ stories as a function of information type and listener condition.

Thus, there were no significant effects of listeners on factual content and no effects of valence on memory content. See Table 2. Are Speakers Aware of the Impact of Listeners on Their Stories?

As in Study 1, we examined whether speakers’ perception of the quality of their story differed significantly as a function of listener condition, event valence, or gender. There were no significant main effects or interactions, and speakers in the two listenerconditions rated their stories quite similarly (attentive:M 4.3, SD 1.3; distracted:M 4.0,SD 1.0). Cross-Sectional Age Differences in Story Content As in Study 1, we computed a partial correlation between age and factual and interpretive information scores, controlling for perceptions of listener attentiveness and this time, based on the results above, for gender. Age was again significantly and posi- tively correlated with the inclusion of factual information,r(76) .23,p .05, and was not significantly correlated with the inclu- sion of interpretive information,r(76) .04,p .50.

To summarize, the results of Study 2 showed that (a) late adolescents and emergent adults told significantly less interpretive stories to distracted friends than to attentive friends, across both positive and negative events; (b) these same participants were not aware of the changes in the stories they told; and (c) once again, older participants, compared with younger participants, told stories that involved more factual content. We also observed main effects of gender on both perceptions of listener responsiveness and the overall length of stories. These main effects were quite congruent:

Women in this study perceived their listeners as more responsive and produced longer stories.

In contrast to the results of Study 1, participants did not reduce the factual content of their stories when talking to distracted friends. Notably, distracted friends in this study were perceived as less responsive but as equally agreeable. It is possible that when people perceive their listener to disagree, they repeat and otherwise emphasize interpretive content to persuade or convince listeners to accept their perspective. Thus, in Study 1, listeners’ unresponsive- ness and disagreeableness may have had opposing effects on interpretive information, diminishing the impact of responsiveness 0 10 20 30 40 50 60 s n o i t a t e r p r e t n I s t c a F Type of Information Number of Idea Units AttentiveDistracted Figure 2.Effect of distracted versus attentive friends on information during conversational remembering in Study 2. Error bars represent standard errors of the mean. 564 PASUPATHI AND HOYT on this type of information. Notably, earlier work on film clips, in which participants perceived a listening stranger as both distracted and disagreeable, obtained findings similar to those of Study 1 (Pasupathi et al., 1998)—significant decrements in both factual and interpretive information.

Work on children’s personal storytelling typically examines naturally occurring variability in maternal responsiveness and elaborativeness; elaborative mothers usually combine warm, sup- portive, and generally agreeable attitudes with high levels of attentiveness to the child’s story. Listening in adulthood likewise combines responsiveness with warmth and agreeableness under most circumstances but not all (see, e.g., Pasupathi et al., 1999).

Ideally, the ability to disentangle responsiveness and agreement experimentally would permit us to explore more directly the im- plications of responsiveness, as distinct from agreement, for the types of information participants include when constructing a personal story in conversations. Study 3 was designed to attempt to disentangle the implications of agreement from listeners and re- sponsiveness from listeners for the construction of narrative iden- tity and to do so in an experimentally controlled manner. This disentangling was motivated by the desire to reconcile the con- flicting findings from Study 1 and Study 2. In addition, however, responsiveness combined with disagreement may have some of its own particular developmental implications, perhaps especially in the context of the dynamics of individuation that are so important for adolescents and young adults. Thus, although the primary purpose of our project was to examine variations in listener re- sponsiveness, variations in listener agreement with the speaker may be relevant for the development of narrative identity in a distinctive way, and Study 3 also provides preliminary evidence of that. Study 3 The major task in Study 3 was to find a way to examine how unresponsive and disagreeable listening differentially influence the construction of identity in conversational narratives. One way to approach this was to borrow from experimental social psychology and to create an experience in our laboratory that permitted some conversational narration and that was at least potentially relevant to narrative identity construction but also somewhat standardized across participants. We did this by asking participants to try a computer game called Sims (Electronic Arts, Redwood City, CA) for the first time. First-time activities, though not of great impor- tance in an individual’s life, offer opportunities for thinking about new aspects of one’s identity and for construing novel situations in terms of enduring identities.The Sims game was also ideal for our purposes because it entails being a simulated individual. Players must “live” in the game by taking care of basic needs such as eating, bathing, and toileting, as well as more elaborate needs for income via work, social interac- tion, and self-enhancement. Needs, indicated via bars on the screen, are addressed by choosing to take particular actions (e.g., clean the apartment, call a friend). Unpredictable, random events such as fires, burglaries, and loss of a job occur and must be dealt with by the player. The game has relatively high appeal for women, in contrast to many computer games, and permits some degree of identity investment on the part of players.

Participants all played the game for the first time, and were then asked to talk about this experience with a same-gender friend, following procedures similar to those employed in Studies 1 and 2.

However, because pilot work showed that such conversations nearly always revolved around evaluating whether the game was fun, we were able to create a condition in which listeners could disagree with speakers’ views of the game. This was done by asking players to evaluate the game just after playing. The players’ evaluation was then employed to instruct their friends to convince them of the opposite view: If players thought the game was fun, as most participants did, then friends were asked to express the view that the game was boring.

This procedure raised three methodological issues. First, are descriptions of playing a computer game for the first time analo- gous to stories about real personal experiences? Second, in those descriptions, how does one address information about how the game is played, which is neither factual nor, in the usual sense of our coding scheme, interpretive? And third, does disagreement about the evaluation of the game change the storytelling task to one involving persuasion in ways that would undermine compar- isons with the previous two studies?

With respect to the first issue, first-time experiences with an activity, of which the present experiment can be seen as an example, are in fact among the personal experiences that we frequently disclose to others. Recently collected data in our labo- ratory showed that of 59 people asked to report on a recent (within the last 2 weeks) novel experience, 53 had already told others about that event, and across different types of events (self- discrepant, self-typical, repeated, and novel), novel events were the most likely to have been disclosed to others. So storytelling about new experiences is an ecologically valid subcategory of the more general category of personal storytelling. Moreover, it is likely that storytelling about first-time experiences is a very im- portant aspect of the construction of narrative identity, in the sense that such storytelling may revolve around the potential role that the Table 2 Means and Standard Deviations of Dependent Measures in Study 2 MeasureAttentive Distracted Positive events Negative events Positive events Negative events MSDMSDMSDMSD Facts 26.3 23.0 24.8 26.0 20.5 19.2 24.2 14.1 Evaluations 44.0 52.6 49.0 76.5 22.6 17.3 19.5 16.8 565 NARRATIVE IDENTITY DEVELOPMENT new experience or activity may or may not play in a person’s life (see Thoman et al., 2007).

To address the second issue, we expanded our coding scheme to include a category calledschematic inferences. This category included idea units that referred to information about how the game could be played, such as the rules, goals, and opportunities within the game. Examples of this category included such state- ments as “You gotta spend all this time beefing up your dude,” “You’re just watching her go through her daily routine,” and “You have to make friends.” Such information was not, strictly speaking, factual in the sense of reporting on what happened during game play. Like interpretive information, it required inferences based on the participants’ experience, but unlike interpretive information, those inferences were about how to succeed at the game, rather than about the broader category of interpretations related to the participants’ experiences and evaluations. As a consequence, we were able to examine how variations in listener behavior influ- enced three types of information: factual details about events that occurred during the participants’ game experience, schematic in- ferences about how the game works, and interpretive information about the overall game experience, including the extent to which the game was fun or interesting to play. Note that schematic information is also an aspect of personal remembering that has been extensively examined in early childhood, and when young children narrate novel experiences, they often focus extensively on this type of information (e.g., P. J. Bauer, 2006). Thus, this addition also permitted Study 3 to explore the construction of schematic knowledge in the context of responsive and unrespon- sive listening in a sample of early adults.

The third issue, which is whether disagreement about the eval- uation of the game changes a storytelling task to a persuasion task, is crucial. However, most remembering to friends or family re- volves around making some type of interpretive claim and backing it up with selected facts (Hyman, 1994). In other words, the storytelling in Studies 1 and 2 also reflected attempts to tell a story that made one or another point, but the specific points at stake were heterogeneous given the varied events participants chose. Method Participants Participants in this study were 104 same-gender friend dyads (49% men, 51% women) recruited from the introductory psychol- ogy subject pool at the University of Utah. The average age of speakers was 20.8 years (SD 3.4), again, consistent with the targeted age range. Within dyads, participants had known one another an average of 4.6 years (SD 5.5). 3This sample predom- inantly comprised European American participants (88.5%). Par- ticipants were randomly assigned to one of five conditions: atten- tive agreeable (n 20), attentive disagreeable (n 22), distracted (n 21), distracted agreeable (n 22), and distracted disagree- able (n 20). On the basis of initial analyses of our manipulation check measures, we collapsed these categories into attentive– agreeable (n 20), attentive– disagreeable (n 22), and dis- tracted (n 63) conditions. Procedure Upon arrival, participants were assigned to speaker and listener roles. As in previous studies, the pairs were separated to completebackground measures not included in the present study. In contrast to what participants did in previous studies, rather than generate a list of recent experiences, speakers played the Sims game for 15 min. The experimenter first explained the video game and pro- vided brief instructions. All participants began the game at an identical starting condition, in terms of simulated money, skills, and status; women played a female character and men a male character. After playing the game, participants completed ques- tionnaires assessing interest in the game.

During this game play time, listeners were given separate in- structions. Agreeable and attentive friends and distracted listeners were given instructions as in Studies 1 and 2. Disagreeable and attentive friends were told that the speaker liked or disliked the game, according to the speaker’s ratings of the game postplay, and were asked to convince the speaker of the opposite evaluation.

Distracted–agreeable and distracted– disagreeable listeners were asked to practice either mild confirmatory responses (e.g., “yeah”) or mild disconfirming responses (e.g., “hmm”) while engaging in the same practice of countingthwords as the distracted listeners. Measures Manipulation check.As in Studies 1 and 2, speakers rated listeners’ responsiveness along the same 7-point Likert-type items indexing indicators of listener responsiveness (Cronbach’s .82). Because agreement from listeners was being manipulated in this study, we added two items assessing listener concordance (“Your friend supported your opinion of Sims” and “Your friend agreed with your view of Sims”), again rated on 7-point Likert- type scales in which low ratings indicated no concordance and high ratings indicated higher concordance. These items were av- eraged to provide a measure of perceived listener concordance (Cronbach’s .84).

Speaker’s perceptions of story quality.The same nine items used in Studies 1 and 2 were averaged to create an assessment of overall story quality (Cronbach’s .87).

Coding of transcripts.Each transcription of a conversation was transcribed and divided into idea units by undergraduate research assistants. These idea units were coded for content through the same coding scheme used in Studies 1 and 2, with the addition of the schematic inference category. Weekly meetings were held in which problematic coding units were discussed and agreement on content was sought. Twenty percent of the conver- sation pairs (n 20) were coded independently to quantify reli- ability. Intercoder reliability for Study 3 across the detailed coding scheme was good (83% agreement; .79). Results Manipulation Check and Preliminary Analyses As noted, participants were assigned into three conditions: an attentive and agreeable condition (n 20 pairs), an attentive but disagreeable condition (n 22 pairs),and a distracted condition (n 63 pairs).Reanalyzing the manipulation checks revealed that the two attentive conditions differed from the distracted condition 3Again, the amount of time known and the quality of the relationships did not affect the results presented for this study. 566 PASUPATHI AND HOYT on attentiveness alone, that the distracted and disagreeable condi- tions together differed from the attentive–agreeable condition on agreement alone. In other words, the distracted condition in this study, as in Study 1, is better described as a distracted– disagreeable condition. Although the full disentangling of agree- ableness and attentiveness that we attempted was not achieved, the present study did succeed in providing a partial disentangling of agreement and attentiveness. Specifically, the disagreeable condi- tion combined disagreement with high responsiveness, whereas the distracted condition reflected perceptions of disagreement with low responsiveness. The attentive and agreeable condition resulted in listeners perceived as highly responsive and highly agreeable.

Do Speakers Change Their Story Depending on Listener Behavior?

We conducted a general linear model with information type (factual, game schematic, and interpretive) as a within-subjects factor and listener condition (attentive–agreeable, attentive– disagreeable, and distracted) and gender as between-subjects fac- tors. The results suggested a significant main effect of condition, F(2, 90) 8.4,p .01, p 2 .16, as well as interactions involving listener condition and type of information,F(4, 180) 5.7,p .01, p 2 .11, and listener condition, type of information, and gender,F(4, 180) 2.8,p .03, p 2 .06.

The main effect of listener condition was reflected by the fact that across all types of information, participants constructed stories with more information of all types when they had an attentive– agreeable (M 31.3,SEM 3.6) or attentive– disagreeable (M 31.0,SEM 3.6) listener, in contrast to stories constructed for a distracted– disagreeable friend (M 17.7,SEM 2.1). The two attentive conditions did not differ from each other. The Condi- tion Type of Information interaction, however, suggested that this difference was not consistent across the different types of information.

Post hoc comparisons examining the effect of listener condition separately for each type of information suggested that listener condition significantly affected all three types of information, 3.4 F(2, 90) 13.5,ps .05, p 2 .07. For facts about the game, participants constructed more elaborated stories for attentive–agreeable friends (M 45.1,SEM 6.9), compared with attentive– disagreeable friends (M 22.3,SEM 6.9) and distracted– disagreeable friends (M 26.5,SEM 4.0), who had similar effects. Differences between the attentive–agreeable con- dition and the other two conditions were marginally significant (ps .07) based on pairwise comparisons.

For schematic inferences about the game, participants con- structed stories with more schematic inferences for attentive– disagreeable listeners (M 35.7,SEM 3.8), followed by attentive–agreeable listeners (M 23.9,SEM 3.8) and distracted– disagreeable listeners (M 14.8,SEM 2.2). Pairwise comparisons showed that the difference between the attentive– disagreeable and distracted– disagreeable condition was significant (p .01), whereas the difference between the attentive–agreeable and attentive– disagreeable condition was marginal (p .09), and the difference between the attentive–agreeable and distracted– disagreeable condition was not significant (p .12).

Interpretive information followed a highly similar pattern, with attentive– disagreeable listeners eliciting the highest number ofsuch statements (M 34.5,SEM 4.0), followed by attentive– agreeable listeners (M 25.1,SEM 4.0) and distracted listeners (M 11.8,SEM 2.3).In this case, pairwise comparisons showed no significant differences between the attentive–agreeable and attentive– disagreeable conditions (p .27), but the distracted condition differed significantly from both of the other conditions (ps .02).

Finally, to explore the three-way interaction with gender, we again examined the effect of listener condition separately for each type of information, but this time we also did this separately for men and women. The resulting contrasts showed effects of con- dition on schematic information for men,F(2, 90) 3.8,p .03, p 2 .08, and women,F(2, 90) 8.6,p .01, p 2 .16, and on interpretive information for both men,F(2, 90) 5.2,p .01, p 2 .10, and women,F(2, 90) 10.3,p .01, p 2 .19. However, for factual information, whereas listener condition had a signifi- cant impact on men,F(2, 90) 5.1,p .01, p 2 .10, there was no comparable effect for women,F(2, 90) 1. Pairwise compar- isons of men and women across all other cells of the design revealed one marginal difference and one significant difference between men and women. Both differences were evident in the attentive condition. Men in the attentive condition tended to pro- duce more factual accounts than women (p .07) and produced significantly less interpretive accounts (p .02). This difference in the baseline, in which women produced relatively nonfactual accounts, is the likely reason that listener condition did not sig- nificantly influence factual content for women. These results are shown in Figure 3. Table 3 presents means and standard deviations for all types of information by gender and listener condition. Are Speakers Aware of the Impact of Listeners on Their Stories?

As in Study 1, speakers rated their stories along the same indicators. An analysis of variance of overall story quality as a function of gender and listener condition revealed no significant differences in perceived story quality as a function of listener condition,F(2, 96) 2.6,p .08, p 2 .05. There were also no effects involving gender. Notably, this effect is a trend. Examina- tion of the estimated marginal means suggests that the overall trend is due to the fact that stories told to attentive and agreeable listeners were rated more highly (M 4.4,SEM 0.2) than stories told to distracted listeners (M 3.8,SEM 0.1). Exam- ining the items that make up the composite story quality rating suggested that only a single item was rated differently by partic- ipants with distracted listeners. Specifically, stories told to distracted– disagreeable listeners were rated as significantly less emotional than stories told to attentive–agreeable listeners, al- though not differently from stories told to attentive– disagreeable friends. In sum, as in Studies 1 and 2, participants did not report awareness of telling different stories to friends whose listening behavior differed. Cross-Sectional Age Differences in Story Content As in Studies 1 and 2, we computed a partial correlation be- tween age and factual, interpretive, and schematic units, control- ling for perceptions of listener attentiveness and gender based on the analyses above. Again, age was positively and significantly 567 NARRATIVE IDENTITY DEVELOPMENT correlated with factual content,r(92) .30,p .05, but not with interpretive content,r(92) .03,p .70, or schematic content, r(92) .07,p .40. General Discussion The present studies demonstrate that the development of narra- tive identity during late adolescence and early adulthood, in the microcontext of conversational storytelling, is shaped by the re- sponsiveness, and perhaps the agreeableness, of listening friends.

Specifically, by early adulthood, responsive listening is critical for the inclusion of interpretive information in storytelling. Given that interpretive information is important for long-term effects on self and identity assessed in a variety of ways (Bird & Reese, 2006; McLean & Pratt, 2006; Pasupathi, 2007), the findings both confirm the role of relationships in identity development and point to a process by which relationships influence identity development.

In Study 3, the effects of distracted listening were similar for interpretive information and schematic inferences. The literatureon the early childhood emergence of autobiographical remember- ing has addressed not only the emergence of self (e.g., Bird & Reese, 2006) but also the emergence of scripts and schemas (e.g., P. J. Bauer & Fivush, 1992). Conversations about a novel activity involve both scripts and interpretive content, and faced with un- responsive friends, our late adolescent participants were impaired in both. Tentatively, this suggests that distracted listeners more broadly suppress meanings, both those that revolve around “what I’m like” and those that refer to “how things get done.” Because both of these kinds of meanings are important for developmental trajectories, this finding warrants further investigation across a broader range of age groups, listeners, and activities.

The role of responsiveness on the construction of narrative identity in the moment was distinguishable from agreement, which was related to whether participants constructed elaborated factual narratives. This has some important developmental implications to which we return below. Finally, these effects of listeners were not reducible to deliberate self-presentation, as our participants were 0 10 20 30 40 50 60 70 80 Facts Schematic Interpretations Facts Schematic Interpretations Males Females Number of Idea Units Attentive/AgreeableAttentive/DisagreeableDistracted Figure 3.Effect of distracted versus attentive and attentive– disagreeable listeners on information during conversational remembering in Study 3. Error bars represent standard errors of the mean. Table 3 Means and Standard Deviations of Dependent Measures in Study 3 MeasureAttentive Disagreeable Distracted Men Women Men Women Men Women MSDMSDMSDMSDMSDMSD Facts 57.7 50.0 32.4 22.8 19.4 15.8 25.2 26.5 25.8 29.8 27.2 28.4 Schematic inferences 21.7 15.8 26.0 12.6 33.0 17.2 38.4 27.1 16.3 17.6 13.3 11.4 Evaluations 15.3 15.6 34.9 26.5 32.3 23.0 36.8 32.3 11.6 9.7 12.0 11.8 Note.Gender is included because this study revealed differences between male and female conversations. 568 PASUPATHI AND HOYT not aware of constructing different kinds of stories for different listeners. Finally, consistent with the idea that this age period is a privileged one for autobiographical encoding and recall (Holmes & Conway, 1999; Rubin & Schulkind, 1997), we found consis- tently small but reliable age-related increases in the factual content of memories.

Below we consider these findings in the context of their impli- cations for narrative approaches to the development of memory and identity (Conway & Pleydell-Pearce, 2000; Fivush & Nelson, 2004; Hooker & McAdams, 2003; McLean et al., 2007; Pasupathi, Mansour, & Brubaker, 2007; Pasupathi & Rich, 2005). Through- out, we discuss specific limitations in the areas where they are relevant to our interpretations.

Developing Narrative Identity in Late Adolescence and Early Adulthood Despite the ubiquity of conversational remembering (e.g., Bohanek et al., 2006; Rime´ , Finkenauer, Luminet, Zech, & Philli- pot, 1998), previous empirical work on narrative identity in emer- gent adulthood has focused primarily on written or interview- based narratives, in which interviewers strive to avoid influencing speakers. Although narrative identity work clearly acknowledges the constructive nature of memory that has been part of cognitive work for decades (Bartlett, 1932/1995; Conway & Pleydell- Pearce, 2000; Schacter, 1996), extant research emphasizes how storytellers’ goals shape their remembering, ranging from endur- ing, characteristic motivations of rememberers (McAdams, Hoff- man, Mansfield, & Day, 1996; Moffitt & Singer, 1994; Woike, 1995; Woike, Gershkovich, Piorkowski, & Polo, 1999) to exper- imentally induced motivations (Dudukovic, Marsh, & Tversky, 2004; McGregor & Holmes, 1999; Sanitioso, Kunda, & Fong, 1990; Tversky & Marsh, 2000). Although the existing sense of self or personality that a storyteller brings to narration is important and clearly influences subsequent remembering for adults (viz.

McLean, 2005; Pasupathi, Alderman, et al., 2007; Sutin & Robins, 2005) and young children (Bird & Reese, 2006; Reese, Haden, & Fivush, 1993), the role of listeners has largely been ignored for emergent adults.

Previous work examining conversational storytelling in this age group has focused primarily on relationships between more tradi- tional self-conceptions and conversational storytelling. Those find- ings indicate that such stories are an important medium for the development of self in social contexts among late adolescents and early adults (Pasupathi, 2006; Pasupathi, Alderman, et al., 2007; Pasupathi & Rich, 2005; Thoman et al., 2007). The present results expand these findings to the interpretive aspects of conversational narrative and connect to work on early childhood storytelling, memory, and self-development. Moreover, participants’ consistent lack of awareness that their stories were changing for listeners suggests that the effects of listeners on young adults’ stories were not a result of deliberate self-presentations; this is important be- cause social psychological work suggests that deliberate self- presentation does not always exert influences on subsequent self- perceptions (e.g., see Rhodewalt, 1998, for a review).

From the childhood work, evaluative content, related to what we have termed interpretive content, is the most important content of memory narratives for self-conception development (e.g., Bird & Reese, 2006). Interpretive aspects of remembering in early child-hood serve, in part, to reflect and create a sense of agency. As children come to retain memories over longer periods, and to think of themselves as agents that endure across time, they begin to construct an agent in their stories that has more long-lasting qualities—an agent with beliefs, values, and long-term goals as well as immediate emotions and desires (see Habermas & Bluck, 2000). From this perspective, listeners can help or hinder the development of narrative identity in storytelling, in part by either supporting or constraining the agency of storytellers, regardless of whether the storyteller is a child or a young adult (see also Cleveland & Reese, 2005). The experience of agency and its extension in time, of course, change with development. But this change is likely to be reciprocally interrelated with the process of narrating one’s experiences in ways that come to construct and reflect a more enduring agency.

Agency in storytelling, however, is multilayered. It includes both the agent within a story, who has feelings, beliefs, desires, and identities expressed in interpretive content, and the storyteller him- or herself, who is giving voice to the story. The interpretive content of autobiographical memories is also the aspect of those memories that is most unique and individuated. Thus, another way to think about distracted listening is that it silences the unique voice of a storyteller. From this perspective, it is not only that unresponsive listeners do not allow us to talk about what experi- ences mean for our self-views— how we as agents think, feel, and believe over time. It is that unresponsive listeners do not allow us to be who we are and who we are becoming, by telling stories in ways that express and further that uniqueness.

Generally, these findings suggest continuity of developmental processes, in that listeners were of great importance for the re- membering of personal experiences in emergent adulthood. The process of parent– child remembering is already known to be critical in the development of narrative identity, memory, and self in early childhood (Fivush & Nelson, 2004). As children move into late adolescence and begin the transition to adulthood, the role played by listeners in the development of narrative identity may, in fact, become more specialized. That is, in early childhood, parents help children learn what and how to recall about the personal past, and this means that parents, as listeners, are shaping both narrative capacity and narrative identity. Responsive, elaborative mothers (and fathers) help children learn to tell elaborate, vivid, detailed, and interpretive stories about their past, and such children develop more advanced self-conceptions and identities in other ways as well (Reese et al., 1993). By late adolescence, narrative capacity is present—all our participants were capable of telling an adequate story about their personal experiences. However, the responsive- ness of listening friends continued to matter for the construction of stories that were richly reflective of narrative identity— of the specific, individual experience of our participants.

Methodologically, however, and conceptually, our findings also differ from prior work in childhood, and those differences bear some consideration here. One issue concerns the comparability of distracted friends and low-elaborative mothers. In some ways, these two groups can be conceived of as having similar effects, in that both groups’ behaviors may convey a lack of interest in the construction of an elaborative, highly interpretive story. Other work from our laboratory demonstrates relationships between overall memory elaboration and naturally occurring variability in friends’ responsiveness but has not examined interpretive and 569 NARRATIVE IDENTITY DEVELOPMENT factual content separately (Liu, 2005). However, in other respects, these two groups are not the same. First, low-elaborative mothers are probably more responsive, within a storytelling interaction, than distracted friends. Distracted friends are unresponsive on average, to the entire story, whereas low-elaborative mothers are selectively responsive to specific aspects of events (often factual in published examples) that they construe as important but are unre- sponsive to other aspects in which their child may have more interest. Thus, low-elaborative mothers may convey interest in a preferred version of a story, whereas distracted listeners convey no interest. Both types of listeners may reduce the likelihood that a person constructs an elaborate sense of narrative identity in the moment but in somewhat distinct ways. Second, distracted friends in these studies were unresponsive on a single occasion but, given the generally established nature and high quality of these friend- ships, are probably more responsive on other occasions. Low- elaborative mothers, by contrast, are seen as engaging in a stylistic approach to the construction of narrative identity with their chil- dren—an approach that has some consistency across various rem- iniscing occasions and over developmental time frames. This brings up the larger issue, which is that the idea that conversational storytelling is a microcontext for the development of narrative identity requires consideration of the ways in which that micro- context is linked to more macrolevel developmental outcomes.

One possibility is that at least some people have friendships that routinely and consistently discourage elaboration of interpretive aspects of remembering. Although this may seem unlikely, con- sider our gender effects in this context. In terms of storytelling content, we observed only one gender difference: Women’s stories about a computer game differed from men’s, primarily in the context of attentive friends. In that context, men’s stories were more factual and women’s were more interpretive, possibly be- cause of expertise in computer gaming generally, as expertise is linked to the nature of storytelling about an event (e.g., Pasupathi, Alderman, et al., 2007).

Other differences by gender, however, were more connected to the idea of stable and consistent differences in the responsiveness of friends. Women also perceived their listeners as more respon- sive in general (Study 2) and specifically in the distracted condi- tion (Study 3). Further, women were more elaborative in general (across factual and interpretive information) in Study 2, suggesting congruence between their perceptions and their construction of narrative identity. Such findings are congruent with findings about parent– child reminiscing practices that differ between male and female children (e.g., Fivush, 1998) and are potentially due to real differences in listener behavior by gender. Past work suggests both men and women prefer female listeners (Clark, 1994), consistent with the idea that women may be more responsive listeners in general and perhaps are better able to remain somewhat responsive even in the face of dual task demands. This capacity, in turn, suggests that the gender of friends may be connected to the extent to which those friends facilitate the construction of narrative identity, and generates some predictions about the prevalence of friendships with women and the construction of narrative identity in this age range.

Further, thinking about gender effects in this way suggests the potential importance of looking at continuity and discontinuity in the kinds of listening done by good friends and the kinds of parent– child reminiscing practices that characterized someone’searly childhood, as well as the parent– child reminiscing practices that characterize their current experience. Recent qualitative work (Weeks & Pasupathi, in press) suggests that at least in some cases, there may be interesting continuities between problematic parent– child reminiscing and problematic friend–friend reminiscing in late adolescence, continuities broadly consistent with attachment perspectives on development (Hesse, 1999).

In addition to the possibility of some friendships entailing consistently low elaboration, there is also the almost certain like- lihood that friends do engage in unresponsive listening with one another periodically and that there are systematic features of that unresponsive listening. Although in the context of high-quality relationships like those we examined here, friends are seldom unresponsive across the board, they are certainly unresponsive some of the time and to particular kinds of content over time. In the long run, repeated unresponsiveness likely serves to silence a particular aspect of identity within that specific relationship con- text, and perhaps to silence that aspect of identity more broadly, depending on the nature of that relationship for the person in- volved. This set of implications is more consistent with the be- havior of low-elaborative mothers, who are in some sense shaping a conversational story toward a particular, mother-driven form.

This kind of selective unresponsiveness can vary in its impact on young adults, from simply making it more difficult to express and explore, in reflection, one’s interest in particular experiences or activities, as in the case of unpopular majors or hobbies, to a real experience of silencing, as in the case of minority sexual identities.

The extent of the impact will also depend on the individual’s motivation and capacity to seek out alternative social relationships within which particular identities can be elaborated, reflected upon, and thus further developed—and indeed, on the very avail- ability of alternative relationships. Factual Information: Development and Disagreement The disagreeableness of listeners tended to suppress factual content in remembering, and across all three of our studies, older participants consistently included more facts in their stories. The latter finding is consistent with the idea that memory development continues even into early adulthood (P. J. Bauer, 2006; Holmes & Conway, 1999; Rubin & Schulkind, 1997). Additionally, it is broadly consistent with developments in knowledge about life (e.g., Glueck & Baltes, 2006; Pasupathi, Staudinger, & Baltes, 2001) and in memory for news events, music, and other happen- ings (e.g., Holmes & Conway, 1999) across this same age range.

This finding clearly suggests the potential for future work, espe- cially in looking at a broader age range and at more particular types of information.

Factual information was suppressed in the context of disagree- able listening for both men and women in Study 1, for men in Study 3, and in past work (Pasupathi et al., 1998). Though tenta- tive, the suppression of factual information in disagreeable dis- course is worth noting, because factual information can ground interpretive information, and interpretations or evaluations without sufficient factual basis may be tenuously held, vulnerable to vari- ability over time or even simply not truly reflective of a person’s mental representation. This is the assumption behind the Adult Attachment Interview, that broad statements about a “good mother” are not plausible without being grounded in factual details 570 PASUPATHI AND HOYT (Hesse, 1999). Although interpretive information was our major focus, the facts of our lives also have developmental import.

Further, it is worth noting that disagreement combined with re- sponsiveness can facilitate narrative identity development very well—a fact consistent with the earlier work of researchers inter- ested in autonomy and identity development (Cooper & Grotevant, 1985) and an important reminder that responsiveness is not syn- onymous with positivity or agreeableness.

Taken together, these studies demonstrate that the development of narrative identity during late adolescence and early adulthood, at a microlevel, is fundamentally intertwined with the extent to which listeners are responsive and encourage elaboration. Further work that attempts to examine the processes by which identity develops needs to take into consideration the context of conversational storytelling and, within that context, the collaboration of listeners. From childhood through adulthood, responsive listeners are critical for allowing each of us to express and further our own unique voice. References Arnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties.American Psychologist, 55,469 – 480.

Bartlett, F. C. (1995).Remembering: A study in experimental and social psychology.Cambridge, England: Cambridge University Press. (Origi- nal work published in 1932) Bauer, J. J., & McAdams, D. P. (2004a). Growth goals, maturity, and well-being.Developmental Psychology, 40,114 –127.

Bauer, J. J., & McAdams, D. P. (2004b). Personal growth in adults’ stories of life transitions.Journal of Personality, 72,573– 602.

Bauer, J. J., & McAdams, D. P. (2005). Interpreting the good life: Growth memories in the lives of mature, happy people.Journal of Personality and Social Psychology, 88,203–216.

Bauer, P. J. (2006).Remembering the times of our lives: Memory in infancy and beyond.Mahwah, NJ: Erlbaum.

Bauer, P. J., & Fivush, R. (1992). Constructing event representations:

Building on a foundation of variation and enabling relations.Cognitive Development, 7,381– 401.

Baumeister, R. F., Stilman, A., & Wotman, S. R. (1990). Victim and perpetrator accounts of interpersonal conflict: Autobiographical narra- tives about anger.Journal of Personality and Social Psychology, 59, 994 –1005.

Bavelas, J. B., Coates, L., & Johnson, T. (2000). Listeners as co-narrators.

Journal of Personality and Social Psychology, 79,941–952.

Bird, A., & Reese, E. (2006). Emotional reminiscing and the development of an autobiographical self.Developmental Psychology, 42,613– 626.

Bluck, S., & Glueck, J. (2004). Making things better and learning a lesson:

Experiencing wisdom across the lifespan.Journal of Personality, 72, 543–572.

Bohanek, J. G., Marin, K. A., & Fivush, R. (2006, March).Family narratives and adolescent identity.Paper presented at the biennial meet- ing of the Society for Research on Adolescence, San Francisco, CA.

Buhrmeister, D. (1996). Need fulfillment, interpersonal competence, and the developmental contexts of early adolescent friendship. In W. M.

Bukowski, A. F. Newcomb, & W. W. Hartup (Eds.),The company they keep: Friendship in childhood and adolescence(pp. 158 –185). Cam- bridge, England: Cambridge University Press.

Clark, R. A. (1994). Children’s and adolescents’ gender preferences for conversational partners for specific communicative objectives.Journal of Social and Personal Relationships, 11,313–319.

Cleveland, E., & Reese, E. (2005). Maternal structure and autonomy support in conversations about the past: Contributions to children’s autobiographical memory.Developmental Psychology, 41,376 –388.Constantinople, A. (1969). An Eriksonian measure of personality devel- opment in college students.Developmental Psychology, 1,357–372.

Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system.Psychological Review, 107,261–288.

Cooper, C. R. (1999). Multiple selves, multiple worlds: Cultural perspec- tives on individuality and connectedness in adolescent development. In A. Masten (Ed.),Cultural processes in child development(pp. 25–57).

Mahwah, NJ: Erlbaum.

Cooper, C. R., & Grotevant, H. D. (1985). Patterns of interaction in family relationships and the development of identity exploration in adolescence.

Child Development, 56,415– 428.

Daddis, C. (2008). Influence of close friends on the boundaries of adoles- cent personal authority.Journal of Research on Adolescence, 18,75–98.

Dickinson, C., & Givon, T. (1995). Memory and conversation: Toward an experimental paradigm. In T. Givon (Ed.),Conversation: Cognitive, communicative, and social perspectives(pp. 91–132). Amsterdam: Ben- jamins.

Dudukovic, N. M., Marsh, E. J., & Tversky, B. (2004). Telling a story or telling it straight: The effects of entertaining versus accurate retellings on memory.Applied Cognitive Psychology, 18,125–143.

Dusek, J. B., & McIntyre, J. G. (2003). Self-concept and self-esteem development. In G. R. Adams & M. D. Berzonsky (Eds.),Blackwell handbook of adolescence(pp. 290 –309). Malden, MA: Blackwell.

Erikson, E. H., & Erikson, J. M. (1997).The life cycle completed: Extended version.New York: Norton.

Fivush, R. (1991). The social construction of personal narratives.Merrill- Palmer Quarterly, 37,59 – 82.

Fivush, R. (1998). Gendered narratives: Elaboration, structure, and emo- tion in parent-child reminiscing across the preschool years. In C. P.

Thompson, D. J. Herrmann, D. Bruce, J. D. Read, D. G. Payne, & M. P.

Toglia (Eds.),Autobiographical memory: Theoretical and applied per- spectives(pp. 79 –104). Mahwah, NJ: Erlbaum.

Fivush, R., Bohanek, J., Robertson, R., & Duke, M. (2004). Family narratives and the development of children’s emotional well-being. In M. W. Pratt & B. H. Fiese (Eds.),Family stories and the life course:

Across time and generations(pp. 55–76). Mahwah, NJ: Erlbaum.

Fivush, R., & Nelson, K. (2004). Culture and language in the emergence of autobiographical memory.Psychological Science, 15,586 –590.

Glueck, J., & Baltes, P. B. (2006). Using the concept of wisdom to enhance the expression of wisdom knowledge: Not the philosopher’s dream but differential effects of developmental preparedness.Psychology and Ag- ing, 21,679 – 690.

Habermas, T., & Bluck, S. (2000). Getting a life: The development of the life story in adolescence.Psychological Bulletin, 126,748 –769.

Haden, C. A., Haine, R. A., & Fivush, R. (1997). Developing narrative structure in parent– child reminiscing across the preschool years.Devel- opmental Psychology, 33,295–307.

Hair, E. C., Moore, K. A., Garrett, S. B., Ling, T., & Cleveland, K. (2008).

The continued importance of quality parent-adolescent relationships during late adolescence.Journal of Research on Adolescence, 18,187– 199.

Harley, K., & Reese, E. (1999). Origins of autobiographical memory.

Developmental Psychology, 35,1338 –1348.

Harter, S. (1998). The development of self-representations. In N. Eisenberg & W. Damon (Eds.),Handbook of child psychology(5th ed., Vol. 3, pp.

553– 617). New York: John Wiley and Sons.

Harter, S., & Monsour, A. (1992). Developmental analysis of conflict caused by opposing attributes in the adolescent self-portrait.Develop- mental Psychology, 28,251–260.

Hartup, W. W. (1996). The company they keep: Friendships and their developmental significance.Child Development, 67,1–13.

Hazan, C., & Zeifman, D. (1994). Sex and the psychological tether.

Advances in Personal Relationships, 5,151–177. 571 NARRATIVE IDENTITY DEVELOPMENT Hesse, E. (1999). The adult attachment interview: Historical and current perspectives. In J. Cassidy & P. Shaver (Eds.),Handbook of attachment:

Theory, research, and clinical applications(pp. 395– 433). New York:

Guilford Press.

Holmes, A., & Conway, M. A. (1999). Generation identity and the remi- niscence bump: Memory for public and private events.Journal of Adult Development, 6,21–34.

Hooker, K., & McAdams, D. P. (2003). Personality reconsidered: A new agenda for aging research.Journals of Gerontology: Series B: Psycho- logical Sciences and Social Sciences, 58B, P296 –P304.

Howe, M. L., Courage, M. L., & Peterson, C. (1994). How can I remember when “I” wasn’t there: Long-term retention of traumatic experiences and emergence of the cognitive self.Consciousness and Cognition, 3,327– 355.

Hyman, I. E. (1994). Conversational remembering: Story recall with a peer versus for an experimenter.Applied Cognitive Psychology, 8,49 – 66.

King, L. A., & Patterson, C. (2000). Reconstructing life goals after the birth of a child with Down Syndrome: Finding happiness and growing.

International Journal of Rehabilitation and Health, 5(1), 17–30.

Kitchener, K. S., Lynch, C. L., Fischer, K. W., & Wood, P. K. (1993).

Developmental range of reflective judgment: The effect of contextual support and practice on developmental stage.Developmental Psychol- ogy, 29,893–906.

Korobov, N., & Bamberg, M. (2004). Positioning a “mature” self in interactive practices: How adolescent males negotiate “physical attrac- tion” in group talk.British Journal of Developmental Psychology, 22, 471– 492.

Kroger, J. (2003). Identity development during adolescence. In G. R.

Adams & M. D. Berzonsky (Eds.),Blackwell handbook of adolescence (pp. 205–226). Malden, MA: Blackwell.

Liu, T. (2005).The planning function of imagined conversation.Unpub- lished master’s thesis, University of Utah.

McAdams, D. P. (1996). Personality, modernity, and the storied self: A contemporary framework for studying persons.Psychological Inquiry, 7, 295–321.

McAdams, D. P., Bauer, J. J., Sakaeda, A. R., Anyidoho, N. A., Machado, M. A., Magrino-Failla, K., et al. (2006). Continuity and change in the life story: A longitudinal study of autobiographical memories in emerg- ing adulthood.Journal of Personality, 74,1371–1400.

McAdams, D. P., Hoffman, B. J., Mansfield, E. D., & Day, R. (1996).

Themes of agency and communion in significant autobiographical scenes.Journal of Personality, 64,339 –377.

McAdams, D. P., Reynolds, J., Lewis, M., Patten, A. H., & Bowman, P. J.

(2001). When bad things turn good and good things turn bad: Sequences of redemption and contamination in life narrative and their relation to psychosocial adaptation in midlife adults and in students.Personality and Social Psychology Bulletin, 27,474 – 485.

McGregor, I., & Holmes, J. G. (1999). How storytelling shapes memory and impressions of relationship events over time.Journal of Personality and Social Psychology, 76,403– 419.

McLean, K. C. (2005). Late adolescent identity development: Narrative meaning making and memory telling.Developmental Psychology, 41, 683– 691.

McLean, K. C., Pasupathi, M., & Pals, J. L. (2007). Selves creating stories creating selves: A process model of narrative self-development.Person- ality and Social Psychology Review, 11,262–278.

McLean, K. C., & Pratt, M. W. (2006). Life’s little (and big) lessons:

Identity statuses and meaning making in the turning point narratives of emerging adults.Developmental Psychology, 42,714 –722.

McLean, K. C., & Thorne, A. (2003). Late adolescents’ self-defining memories about relationships.Developmental Psychology, 39,635– 645.

McNulty, S. E., & Swann, W. B. J. (1994). Identity negotiation in room- mate relationships: The self as architect and consequence of social reality.Journal of Personality and Social Psychology, 67,1012–1023.Meilman, P. W. (1979). Cross-sectional age changes in ego identity status during adolescence.Developmental Psychology, 15,230 –231.

Moffitt, K. H., & Singer, J. A. (1994). Continuity in the life story:

Self-defining memories, affect, and approach/avoidance personal striv- ings.Journal of Personality, 62,21– 43.

Pals, J. L. (2006a). Constructing the “springboard effect”: Causal connec- tions, self-making, and growth within the life story. In D. McAdams, R.

Josselson, & A. Lieblich (Eds.),Identity and story: Creating self in narrative(pp. 175–199). Washington, DC: American Psychological Association.

Pals, J. L. (2006b). Narrative identity processing of difficult life experi- ences: Pathways of personality development and positive self- transformation in adulthood.Journal of Personality, 74,1079 –1110.

Parker, J. G., & Gottman, J. M. (1989). Social and emotional development in a relational context: Friendship interaction from early childhood to adolescence. In T. J. Berndt & G. W. Ladd (Eds.),Peer relationships in child development(pp. 95–131). Oxford, England: John Wiley and Sons.

Pasupathi, M. (2001). The social construction of the personal past and its implications for adult development.Psychological Bulletin, 127,651– 672.

Pasupathi, M. (2006). Silk from sows’ ears: Collaborative construction of everyday selves in everyday stories. In D. McAdams, R. Josselson, & A.

Lieblich (Eds.),Identity and story: Creating self in narrative(pp.

129 –150). Washington, DC: American Psychological Association.

Pasupathi, M. (2007). Telling and the remembered self: Linguistic differ- ences in memories for previously disclosed and undisclosed events.

Memory, 15,258 –270.

Pasupathi, M., Alderman, K., & Shaw, D. (2007). Talking the talk: Col- laborative remembering and self-perceived expertise.Discourse Pro- cesses, 43,55–77.

Pasupathi, M., Carstensen, L. L., Levenson, R. W., & Gottman, J. M.

(1999). Responsive listening in long-married couples: A psycholinguis- tic perspective.Journal of Nonverbal Behavior, 23,173–193.

Pasupathi, M., & Mansour, E. (2006). Adult age differences in autobio- graphical narratives: Integrating experiences with the self.Developmen- tal Psychology, 42,798 – 808.

Pasupathi, M., Mansour, E., & Brubaker, J. (2007). Developing a life story:

Constructing relations between self and experience in autobiographical narratives.Human Development, 50,85–110.

Pasupathi, M., & Rich, B. (2005). Inattentive listening undermines self- verification in personal storytelling.Journal of Personality, 73,1051– 1086.

Pasupathi, M., Stallworth, L. M., & Murdoch, K. (1998). How what we tell becomes what we know: Listener effects on speakers’ long-term mem- ory for events.Discourse Processes, 26,1–25.

Pasupathi, M., Staudinger, U. M., & Baltes, P. B. (2001). Seeds of wisdom:

Adolescents’ knowledge and judgment about difficult life problems.

Developmental Psychology, 37,351–361.

Peterson, C., Jesso, B., & McCabe, A. (1999). Encouraging narratives in preschoolers: An intervention study.Journal of Child Language, 26, 49 – 67.

Reese, E., & Fivush, R. (1993). Parental styles of talking about the past.

Developmental Psychology, 29,596 – 606.

Reese, E., Haden, C. A., & Fivush, R. (1993). Mother-child conversations about the past: Relationships of style and memory over time.Cognitive Development, 8,403– 430.

Rhodewalt, F. (1998). Self-presentation and the phenomenal self: The “carryover effect” revisited. In J. M. Darley & J. Cooper (Eds.),Attri- bution and social interaction(pp. 373–398). Washington, DC: American Psychological Association.

Rime´ , B., Finkenauer, C., Luminet, O., Zech, E., & Phillipot, P. (1998).

Social sharing of emotion: New evidence and new questions.European Review of Social Psychology, 9,145–189.

Rubin, D. C., & Schulkind, M. D. (1997). Distribution of important and 572 PASUPATHI AND HOYT word-cued autobiographical memories in 20-, 35-, and 70-year-old adults.Psychology and Aging, 12,524 –535.

Sanitioso, R., Kunda, Z., & Fong, G. T. (1990). Motivated recruitment of autobiographical memories.Journal of Personality and Social Psychol- ogy, 59,229 –241.

Schacter, D. L. (1996).Searching for memory: The brain, the mind, and the past.New York: Basic Books.

Sutin, A. R., & Robins, R. W. (2005). Continuity and correlates of emotions and motives in self-defining memories.Journal of Personality, 73,793– 824.

Thoman, D. B., Sansone, C., & Pasupathi, M. (2007). Talking about interest: Exploring the role of social interaction for regulating motivation and the interest experience.Journal of Happiness Studies, 8,335–370.

Thorne, A. (2000). Personal memory telling and personality development.

Personality and Social Psychology Review, 4(1), 45–56.

Thorne, A., McLean, K. C., & Lawrence, A. M. (2004). When remember- ing is not enough: Reflecting on self-defining memories in late adoles- cence.Journal of Personality, 72,513–541.

Tice, D. M. (1992). Self-concept change and self-presentation: The looking glass self is also a magnifying glass.Journal of Personality and Social Psychology, 63,435– 451.Tversky, B., & Marsh, E. J. (2000). Biased retellings of events yield biased memories.Cognitive Psychology, 40(1), 1–38.

Updegraff, K. A., McHale, S. M., & Whiteman, S. D. (2006). The nature and correlates of Mexican-American adolescents’ time with parents and peers.Child Development, 77,1470 –1486.

Wang, Q. (2004). The emergence of cultural self-constructs: Autobio- graphical memory and self-description in European American and Chi- nese Children.Developmental Psychology, 40,3–15.

Waterman, A. S. (1982). Identity development from adolescence to adult- hood: An extension of theory and a review of research.Developmental Psychology, 18,341–358.

Weeks, T. L., & Pasupathi, M. (in press). Autonomy, identity, and joint autobiographical remembering in adolescence. In K. C. McLean & M.

Pasupathi (Eds.),Narrative development in adolescence: Creating the storied self.New York: Springer.

Woike, B. A. (1995). Most memorable experiences: Evidence for a link between implicit and explicit motives and social cognitive processes in everyday life.Journal of Personality and Social Psychology, 68,1081– 1091.

Woike, B., Gershkovich, I., Piorkowski, R., & Polo, M. (1999). The role of motives in the content and structure of autobiographical memory.Jour- nal of Personality and Social Psychology, 76,600 – 612. (Appendix follows) 573 NARRATIVE IDENTITY DEVELOPMENT Appendix Sample Transcript With Coding Example: Pair 29 Content S: All right, well my experience was a couple days ago 1 we ran out of toilet paper. And2 Mary was like, okay,2 I’ll go get toilet paper, because2 it’s her turn, right?3 Yeah.1 Ten minutes later, she’s like, hey Eileen, 2 can you get some toilet paper?2 I was so mad.3 Yeah, like, yesterday all day long she’s like, 2 I don’t have anything to do all day.2 I’m like, why don’t you go get some toilet paper because 2 we’re out of toilet paper.2 She’s like, no,2 I just want to sit here and2 play on the Internet.2 L: Oh.1 S: So mad.3 Finally today I went and got it.2 (inaudible)9 I was like, napkins are good.2 L: That’s smart.3 S: Well, then there’s the part where3 I’m the younger3 (inaudible),9 supposedly, because she’s twenty-six.3 L: Oh.1 S: And she’s like, no,2 I don’t want to.2 Being lazy.3 She’s like, I’ll pay you back.2 I’m like, Smith’s is two minutes down the road. 2 Go get some.2 L: That’s funny. Funny.3 Roommate troubles.2 S: Drama, drama, drama3 We’re done.1 Note.S speaker; L listener; 1 regulatory; 2 event; 3 interpretive; 9 uncodable. Received August 9, 2007 Revision received September 29, 2008 Accepted October 14, 2008 574 PASUPATHI AND HOYT