Forging one’s identity as a twin: Balancing sibling cohesion and deidentification

Abstract

Little is known about how twins influence one another’s development during emerging adulthood. Although the exact mechanisms that underlie sibling influence have not been firmly established, they likely encompass processes that highlight either sibling identification (e.g., seeking similarities and imitation in the service of building cohesion) or deidentification (e.g., seeking differences that distinguish themselves to reduce competition, rivalry, and jealousy). Because past research inferred these processes rather than directly assess them, we conducted a mixed methods study in which 20 sets of monozygotic (MZ, n = 11 dyads) and dizygotic (DZ, n = 9 dyads) twins, aged 19.70 years (SD = 1.11), were interviewed individually regarding their personal identity development and sibling relationship. A grounded theory approach guided the qualitative analysis, which included coordinating twins’ interview responses at the dyadic level. Each twin also completed the Adult Sibling Relationship Questionnaire which provided a complementary quantitative assessment. 75% of the dyads prioritized differentiation and felt that college was the optimal time to pursue unique identities and goals. Those who prioritized seeking similarity were MZ twins who sought emotional closeness and support and attended the same college. Rather than pitting social learning and deidentification as opposing processes, the results suggest that deidentification is a developmental process in which twins increasingly take steps towards independence while seeking to retain the closeness, trust, and reliance they have long enjoyed in their relationship. Practical implications for helping twins balance their needs for autonomy and distinctiveness, while maintaining emotional closeness, are discussed.

Introduction

Evidence is growing that siblings influence one another’s development and well-being across the life course (Her et al., 2021; Jensen et al., 2023). This is particularly true during the transition to emerging adulthood when siblings serve as key referents for one another as they form their identities and make life decisions (Cassinat & Jensen, 2020). Although the exact mechanisms that underlie sibling influence have not been firmly established, according to Her, they likely encompass processes that highlight either sibling identification (e.g., seeking similarities through social learning in the service of building cohesion) or deidentification (e.g., seeking differences that distinguish themselves to reduce competition, rivalry, and jealousy). However, it is not clear whether one of these processes is more critical for identity development than the other. The current study uses a mixed methods approach to learn directly from emerging adult twins about how they believe they influence one another’s development and identities.

As described by Whiteman et al. (2009), previous research has been limited by the practice of inferring social learning (Bandura, 1969) and differentiation processes (Adler, as cited by Ansbacher & Ansbacher, 1956) post hoc, with social learning or imitation presumed when siblings make similar choices, and deidentification or differentiation, when siblings make different choices. However, identities are shaped by numerous influences (Bandura, 1969) and what appears to be imitation could actually been intended as differentiation, such as when the pursuit of distinct goals leads siblings to similar paths. Thus, direct assessments are needed to clarify the specific processes that underlie sibling influences (Whiteman et al., 2007), especially assessments that give voice to siblings’ multi-faceted lived experiences and perspectives.

Erikson (1968) considered identity formation to be a critical developmental process, beginning in adolescence and continuing through what we now refer to as emerging adulthood, in which an individual progressively develops a clear sense of who they are as a person and as a contributor to society. In contrast to role confusion, identity formation refers to an evolving sense of personal coherence that is relatively consistent across time, life changes, and new role requirements, and that contributes to a sense of well-being. Similarly, Bowen’s (1978) concept of differentiation of the self proposed that as individuals develop a clearer sense of who they are as they mature, they are more likely to differentiate from close family members such as siblings. Bowen’s family systems theory explicitly states that individuals’ differentiation of self is essential for forming meaningful romantic relationships that will avoid emotional reactivity, fusion, or emotional distance. Following Erikson and Bowen, in this study, we focused on how emerging adult twins conceive of themselves and one another.

The transition to college offers a unique opportunity to examine twins’ identity formation and sibling influence processes as it is a critical life event typically occurring in the early stages of emerging adulthood (Arnett, 2007). During this transition, individuals often physically separate from their families as they define and pursue professional goals, romantic relationships, and personal values and belief systems. In comparison to other sibling types, twins may feel particularly challenged to explore unique interests and goals given their extensive history—since the womb—of shared experiences and relationships (Bekkhus et al., 2016). The transition to college may be particularly salient for monozygotic (MZ) twins who may perceive stronger pressures than dizygotic (DZ) twins to behave in similar ways, maintain a similar appearance, and follow similar paths as they chart their professional and personal lives (Segal, 1999). For many twins, the transition to college may offer the first opportunity they have to choose divergent paths and forge unique identities as this is when the sibling relationship becomes more voluntary and parental influence has lessened (Hamway et al., 2019; Jensen et al., 2018; Lindell et al., 2014).

Sibling influence processes

Whereas siblings may influence and socialize one another in a variety of ways (Kramer & Conger, 2009), we focus on two frequently cited theoretical approaches: social learning theory and sibling deidentification.

Social learning theory

According to the social or observational model of learning advanced by Bandura (1969), “children have repeated opportunities to observe and to learn the behavior and values not only of parents, but also of siblings, peers, and other significant persons” (p. 251). Siblings have been found to emulate one another in various domains, including risky sexual behaviors (East, 1998), substance use (Slomkowski et al., 2005), and aggression and delinquency (Bank et al., 1996). Individuals are more likely to adopt the characteristics of siblings when they are of similar ages and gender (Wong et al., 2010), have a close relationship (Slomkowski et al., 2005), and when the family reward structure encourages such imitation (Bandura, 1969). With respect to forming life plans, Cassinat and Jensen (2020) found that young adults’ beliefs about how central marriage is to their identity and their plans to marry in the future approximated those of their closest-in-age sibling when the siblings engaged more in modeling than differentiation. In the current study, we examined individuals’ reports of the importance they placed on seeking similarity with their twin, and the ways they did so, as indices of social learning.

Sibling deidentification theory

As cited by Ansbacher and Ansbacher (1956), Adler advanced the concept of sibling deidentification to describe how siblings consciously or unconsciously define themselves as different from their sibling as a means of reducing sibling rivalry and competition for parental attention and affection. Frances Fuchs Schacter and colleagues (Schacter et al., 1976, 1978) further described sibling deidentification as a process by which siblings protect themselves from rivalry, jealousy, and conflict by seeking unique identities, adopting distinct attributes, engaging in different activities, and forming distinct social relationships. An interesting corollary of Schacter’s premise is that, through deidentification, individuals form their identity, in part, in reaction to how they perceive their siblings. In this way, siblings serve as key referents for one another as they explore and form their respective identities.

Deidentification processes play important developmental functions, such as when adolescents differentiate themselves from a sibling who is using alcohol or other substances (Whiteman et al., 2013) or engaged in other risky behaviors. Deidentification is posited to occur more intensely when siblings are objectively similar (e.g., in age and gender) and wish to highlight their unique characteristics (Schacter et al., 1976, 1978). As same-aged siblings who may share the same gender, deidentification may be especially salient to twins launching from their families of origin.

Social learning versus deidentification

Social learning and deidentification have been largely cast as opposing processes in previous research (Whiteman et al., 2007). However, as Whiteman et al. pointed out, both types of influence may occur in the same dyad, with siblings seeking to be alike in some areas, yet different in others. Furthermore, as few studies have directly and simultaneously assessed both sets of processes, it is not yet clear whether social learning and deidentification are independent processes or work in concert (Whiteman et al., 2014). In one exception to this trend, Whiteman et al. (2014) examined the sibling influence processes of social learning, differentiation, and having shared friends in a study of adolescents’ attitudes and behaviors related to alcohol. Although siblings endorsed some items reflecting both convergent (i.e., modeling, shared friends, and younger sibling admiration of their elder sibling) and divergent (i.e., differentiation) influence processes, one process tended to dominate. More similar alcohol-related attitudes and behavior were found among dyads reporting convergent modes of sibling influence; however, the effects for differentiation were moderated by sibling gender constellation. In line with Whiteman et al.’s call for research to use direct measurements of both convergent and divergent sibling influence processes to disentangle their effects, the current study investigated the occurrence of both seeking similarity and differentiation processes in twins’ descriptions of their relationship. Acknowledging the likelihood of multiple and reciprocal pathways of sibling influence, the current study also examines whether social learning and differentiation may co-occur.

In addition to understanding whether and why siblings seek to be similar and/or deidentify (their rationale), we must also understand the domains in which they do so, such as academics, arts, athletics, or social relationships (Lindell et al., 2014; Whiteman et al., 2007). Additionally, understanding how siblings behave to be more similar or different within these domains is also important. In the current study, we examined the steps twins report taking to be more similar or different in the domains that were most salient to them.

Siblings’ efforts to be alike or different may also be shaped, in part, by their perceptions of their siblings’ personal qualities and the quality of their relationship as well as through sibling comparison (Jensen et al., 2015). Whiteman et al. (2007) found that youth who emulated their elder siblings reported a warmer relationship than those who deidentified. Interestingly, no evidence was found linking deidentification with less rivalry or conflict. In the current study, we also consider twins’ interests in seeking similarity and difference in light of their perceptions of warmth and conflict in their sibling relationship quality.

Most studies of sibling influence have assumed that elder siblings act as socialization agents for their younger siblings as their greater expertise and experience make them effective models (Whiteman et al., 2014) as well as caregivers and teachers (Kramer & Hamilton, 2019). Although younger siblings also regularly influence their elder siblings (Kramer & Conger, 2009), this hierarchical approach has framed most studies of sibling socialization. Interestingly, twin relationships, with their lack of an inherent hierarchy based on age, birth order, maturity, or expertise, represent a unique developmental context for studying sibling influences—one that has been largely overlooked to date.

The current study: A dyadic and qualitative approach

We followed a constructivist grounded theory approach (Charmaz, 2014) as we conducted individual interviews with both members of twin pairs and then examined their respective perspectives collectively, as a dyad. This dyadic approach to analyzing interview data, in which contrasts and overlaps between the twins’ perspectives in their textual and subtextual responses are examined, can reveal themes related to the nature of twin relationships that may not otherwise emerge (Eisikovits & Koren, 2010). This approach is a significant departure from previous studies in which only one sibling’s reports are considered. According to family systems theory (Minuchin, 1974), mutual influences may be overlooked when the perspectives of only one sibling are examined. Thus, triangulating the reports of both twins, examining similarities and differences at the descriptive and interpretive levels, as we do in the current study, should advance our understanding of sibling influence processes.

In summary, in the current study we directly assessed processes of sibling influence during emerging adulthood using a mixed methods approach that included: (1) in-depth individual interviews with MZ and DZ twins about their personal identity development and sibling relationship during the transition to college; and (2) a standard quantitative measure of adult sibling relationship quality. We aimed to describe how twins influence one another’s identities, and decisions and plans for the future. Although we did not set a priori hypotheses about whether social learning versus differentiation would be most evident in twins’ reports of sibling influence, we were sensitive to the discussion of these processes in our qualitative coding. We also explored contributions of zygosity and sibling relationship quality.

Method

Paradigmatic approach

Following Charmaz (2014), we adopted a constructivist grounded theory approach to guide the collection and analysis of our qualitative interview data. We prioritized the voices of our participants and sought to describe and interpret how and why our participants constructed meaning and made decisions and took particular actions during the transition to college. Our approach further prioritized the contextual and structural conditions under which our participants found themselves as well as how we, as researchers, understand these conditions given our “perspectives, privileges, positions, interactions, and geographical locations” (Charmaz, 2014, p. 240). The sample size of 40 individuals (20 twin dyads) enabled us to reach saturation and we ended data collection when we found redundancies in interview reports.

Researchers’ positionality and self-reflection

Our approach was influenced by our prior knowledge of sibling relationships during emerging adulthood. The first author, a firstborn daughter, has studied the development of sibling relationships for more than 30 years. The next two authors are fraternal twin college students who (without a great deal of discussion) found themselves making virtually identical choices for their education and career goals. The identities of these two authors, which are similar to those of our participants, increased the likelihood that our team would deeply understand our participants’ narratives. The final author is a firstborn daughter currently studying sibling and other personal relationships in her doctoral clinical psychology program.

Participants

Participants were recruited using flyers physically posted, and distributed electronically, on one university campus in the northeastern US. Interested individuals were asked to complete a brief Qualtrics form requesting basic demographic information to ensure that they met the inclusion requirements of being an undergraduate with a twin who also attended (any) college in the US and would likely be interested in also participating in this study. The twins of the applicants were then contacted and invited to participate. Eleven sets of MZ twins and 9 sets of DZ twins, aged 19.70 years (SD = 1.11, Mdn = 20, Range = 18–22) met these requirements. Twelve dyads identified as both cisgender women, 4 both cisgender men, 3 cisgender men/women, and 1 cisgender woman/nonbinary. The dyads had, on average, 1 additional sibling; 2 dyads were triplets, yet their sibling chose not to participate. Participants indicated all ethnic/racial groupings with which they identified and 15% identified as Hispanic. Additionally, participants identified as White (67.5%), Asian (5%), Black (5%), biracial (20%), and other (2.5%). Most (78%) were in their first or second year of college and 10% were international students. Eleven dyads attended different universities (4 MZ, 7 DZ) and 9 attended the same university (7 MZ, 2 DZ).

Procedure

Each twin was interviewed individually online via Zoom by an interviewer who had no contact with the participant’s twin. Interviewers were undergraduate or graduate students trained to conduct semi-structured interviews. The interviews took approximately 1.0–1.5 hours and were video recorded. Following the interview, participants were emailed a link to a Qualtrics survey that inquired about the quality of the relationship with their twin. In appreciation of their effort, each participant received a $20 gift card to a major retailer.

Measures

Interview

The interview crafted for this study was based on the results of relevant research and theory regarding sibling influence processes. Open ended questions tapped: (1) their decision about attending college with or without their twin, (2) perceived similarities and differences in their choice of major and career plans; (3) the importance they placed on seeking similarity and/or being seen as independent of their twin; (4) steps they have taken to be similar and different; (5) experiences of sibling comparison and competition; (6) the meaning that being a twin has had for them; (7) patterns of similarity and differentiation while growing up; (8) others’ expectations that they follow similar/different paths; (9) history of parental differential treatment; and (10) future expectations for their relationship.

Sibling relationship quality

Lanthier and Stocker’s (2014) Adult Sibling Relationship Questionnaire-Very Short Form (ASRQ-VSF) was used to assess participants’ perceptions of warmth (6 items) and conflict (6 items). Items were rated on a 5-point Likert scale (1 = hardly at all; 5 = extremely much). Sample items include, “How much do you and this sibling talk to each other about important things?” and “How much do you and this sibling irritate each other?” Lanthier and Stocker reported satisfactory reliability and validity indices for this instrument. With the current sample, internal consistency (alpha) was .864 for warmth and .871 for conflict.

Qualitative analyses

Qualitative analyses were conducted in several phases. First, each video-recorded interview was transcribed by the interviewer; a second researcher then checked the verbatim transcripts to ensure accuracy. Second, the entire research team (6 research assistants and principal investigator) watched individual video recordings, along with their transcripts, and met over the course of 36 weeks to discuss the emergent main themes. This process resulted in an initial delineation of themes, codes, and subcodes that described participants’ perspectives and manifest and underlying meanings.

Third, working in three sets of pairs, the 6 researchers independently watched a set of recorded interviews along with their verbatim transcripts and applied the initial set of codes. Researchers were free to add new codes and offer edits or suggest deletion of existing codes to best represent the participants’ viewpoints. Researchers wrote memos to describe significant observations, questions, and hypotheses following the coding of each interview. After working independently, researchers met to discuss their codes and make adjustments through consensus. Researchers then brought their coding to the larger group’s weekly discussion, which typically led to coding system refinement. As the coding system evolved, researchers recoded videos that were coded earlier. Each video underwent coding by at least 2 individuals, with all researchers ultimately watching all videos.

Fourth, after finalizing their coding for an individual, the coding pair then reviewed the video recording for that individual’s twin. A dyadic analysis was then conducted of how the twins’ reports, when considered as a pair, converged and reflected components of social learning and deidentification. Four coders completed a “Coders’ Questionnaire” for each dyad in which they described the degree to which the dyad agreed about the importance of seeking similarity or differences as well as their rationale, salient domains, and the steps they took to be more alike or different. Coders also provided their own impressions, observations, and hypotheses. The Coders’ Questionnaires were discussed at weekly meetings with the goal of understanding shared and divergent perceptions of the dyads and resolving differences through consensus.

Fifth, the software package, Dedoose, was used to code the transcripts using the coding system described above. Dedoose also allowed for the integration of demographic variables and descriptor variables, such as zygosity, gender, attending the same or different college, ASRQ-VSF scores, and coders’ ratings.

Trustworthiness

Inter-rater agreement was estimated by having two naïve researchers independently code 8 videos and transcripts (20%) and examining coding convergence. Overall, inter-rater agreement was 84%. Additional steps to ensure the trustworthiness of these data, which addressed credibility, dependability, confirmability, and transferability (Lincoln & Guba, 1986) are outlined in the Supplemental Materials.

Results

The results presented below are organized in terms of addressing a series of questions we asked of the qualitative and quantitative data. First, how do twins influence one another’s identities, plans, and aspirations during the transition to college? What evidence is there to support social learning or deidentification as primary modes by which twins influence one another? Further, what role does zygosity play in shaping these processes? Second, does deidentification appear to be a developmental process as predicted by key developmental and identity theories (e.g., Bowen, 1978; Erikson, 1968)? Third, how does sibling relationship quality relate to patterns of seeking similarity and differentiation? Do twins who report greater emotional closeness, warmth, and low conflict in their relationship seek greater similarity and less differentiation?

RQ #1: Processes of sibling influence during the transition to college

To gauge whether twins considered social learning (operationalized to participants as “seeking similarity”) or deidentification (operationalized as “seeking differentiation”) as their primary mode of sibling influence, we examined their interview responses for descriptors of these constructs, using both the Dedoose coding system and coders’ appraisals. Participants were considered to have social learning or seeking similarity as their primary mode of sibling influence when their interviews were coded as placing importance on being alike, for example, by engaging in similar activities, emulating their twin’s characteristics or successes, enjoying mutual friendships and relationships, and not feeling ready to be separated. Twins were considered to have deidentification or differentiation as their primary mode of influence when their interviews were coded as placing importance on establishing their unique identity and sense of self, wanting to pursue different interests, goals, and relationships, and oriented towards “independence.”

Whereas all participants discussed elements of both deidentification and social learning, differentiation was considered the primary mode of sibling influence among 15 of the 20 dyads (75%), with seeking similarity characterizing the remaining 5 dyads (25%). In only one dyad did individual siblings hold disparate perspectives, with one sibling valuing differentiation and one similarity. All 5 dyads (100%) who prioritized similarity were MZ and attended the same college; in contrast, only 4 dyads (27%) who emphasized differentiation attended the same college.

Below, we describe the domains in which the twins sought similarity and differentiation, the rationales they gave for prioritizing these modes, and the steps they took to be more similar or different from one another. Participants’ names have been changed in the quotes below; however, we retained their true zygosity (MZ, DZ) and male/female/non-binary gender constellation (MM, FF, MF, F/NB).

Social learning and identification: Seeking similarity

Those with a strong desire to be similar (n = 5 dyads, all MZ) highly valued the emotional closeness and support twinship brings. This was expressed, for example, through statements such as, “We were in the womb together! There’s just a connection I can’t really explain” [Riley, MZ, FF], and echoed by her sister:

Obviously, we love each other more than any other person because we're literally like the same person, and we're, like, so close. [Penelope, MZ, FF]

Two of the 5 dyads stressing similarity viewed their twinship as an integral part of their identity and felt little need to differentiate:

I didn't need to be somewhere else to, like, prove I was different… I never needed to prove to anybody that I didn't want to be close to him, or that I didn't want to be like him, because, like he is my world… Like, all of my memories are with him. [Max, MZ, MM]

The remaining 3 dyads prioritizing similarity emphasized a strong desire to sustain their attachment, with some specifically wanting to avoid the anxiety that could come with separation:

I can't even begin to like fathom how alone and like overwhelmed I would feel if I was alone. Like having Riley [at the same college] made it like so much easier, because I knew I had someone who would like, understand. [Penelope, MZ, FF]

My favorite part of being a twin, having someone… that you know will be there because they can't leave me. [Jenna, MZ, FF]

Domains of seeking similarity

When considering the full sample (not only those prioritizing similarity), the most common domains in which twins described seeking similarity were in their personality, behavior, and conduct (expressed by 88% of participants); social relationships (75%); academic and intellectual interests and abilities (75%); athletics and sports (65%); and interests, hobbies, and extracurricular activities (58%). Less frequently mentioned domains in which twins sought similarity included morals, values, and viewpoints (33%); artistic areas (28%); and physical appearance (23%).

Few patterns were observed with respect to zygosity in the domains of seeking similarity; however, MZ participants were more likely to mention seeking similarity in athletics and sports (82% MZ, 44% DZ) and in interests, hobbies, and extracurricular activities (73% MZ, 39% DZ).

Rationale for seeking similarity

The most common rationales twins (across the full sample) gave for seeking similarities reflected the fun they have being alike and doing things together (45%) as well as for practical reasons (45%):

We just have fun all the time. (laughs)… If you… go to two different schools, you don't get that pleasure… So [at first] we decided that it was probably best to kind of get some separation, you know, for like independence and all this stuff. Um. But then when we decided to go to the same school [we realized], like, “Yeah, that was so stupid to [plan to] go to different schools because it makes so much more logistical sense to be together.” Like it's so much easier. [Cody, MZ, MM]

Similarly, life was thought to be “better together” as they can do more by joining forces, especially when one’s strengths can compensate for the other’s relative weakness (45%).

…it was better when we were together. Better socially. I’d say better mentally, ah, just holistically, everything is better… We're the perfect team. Because I don't, like, take emotion really that much into account when I'm making decisions. “Does this make logical sense? If so, we should probably do this.” And he's more of like, “Oh, we're going to be happier if we do this thing”. Um. So that's the reason when we're together, we're so much better than when we were apart…Like I'd make a lot of mistakes that I knew if Max was there with me, he would tell me that this is a mistake. [Cody, MZ, MM]

Some participants (25%) discussed wanting to emulate their twin’s successful behaviors so they could also excel:

When I’m like, feeling insecure, I’ll think about, “Hey, like what would Nora do?” And try to, like emulate that, especially in like areas where I feel like I could grow. I guess in terms of like stress and like schoolwork, I’m like, “Okay, like think like Nora. Like, what would Nora do right now?” [Theresa, MZ, FF]

Some twins (33%) acknowledged the immense comfort they found in being similar to their sibling and did not feel ready to separate.

I feel like, for all major life decisions I've always thought about him because of how similar our lives have been… So we’ve very much done life together and I didn't really feel like I was at a point where I was ready to do life on my own just yet. [Max, MZ, MM]

We always like had each other, no matter what we were doing. Like the first week [of college], just like meeting people, we’d always be together. So it was kind of like a good crutch to have like, to rely on someone to be with. [Logan, MZ, MM]

Similarly, 28% expressed pursuing similar experiences as a way to sustain closeness in their relationship, even though they understood this could mean delaying their personal autonomy and, perhaps, reinforce their dependency:

I didn't want to be separated from her because she's like my best friend. So I kind of wanted to go to the same school as her so and we both really liked this one so. I guess we just decided mutually to come here because then we could be together. [Michelle, MZ, FF]

Interestingly, 28% of the participants voiced being so invested in sustaining their close relationship that they would make decisions (even sacrifices) to honor their twin’s preferences even when a different choice might be in their own best interest:

I’m always putting myself second to her. Like I always like prioritize like her well-being like, I just like- I don't know, she's so important, and like I don't want anything to like happen to her, um, kind of thing. So like, if I can prevent it or like, do something to help, like, I would do it in a heartbeat. Even if, like I suffered from it or like it didn't help me out in any way. [Penelope, MZ, FF]

Whereas DZ twins were more likely to offer “better together” as a key reason for making similar choices (14% MZ, 83% DZ), MZ twins were more likely to discuss factors such as enjoying the fun they shared (55% MZ, 33% DZ), a desire to sustain closeness (36% MZ, 17% DZ), and a reluctance to give up the comfort and reliance that comes with sibling similarity (41% MZ, 22% DZ). Whereas 23% of MZ twins expressed their willingness to sacrifice their personal interests to follow the same path as their sibling, none of the DZ twins did.

Steps taken to seek similarity

The most common steps taken to seek similarity were developing a similar friend network (63%); engaging in activities together and joining the same clubs (58%); making intentional choices to be geographically close (45%); attending the same college (45%); coordinating their decisions and plans (e.g., taking internships in the same location; 40%); and emulating their twin to build personal success (30%).

Overall, MZ twins were more likely than DZ twins to engage in similar activities (73% MZ, 39% DZ), seek proximity (59% MZ, 28% DZ), collaborate on decisions and plans (50% MZ, 28% DZ), and build cohesion (36% MZ, 17% DZ). MZ twins were also more likely than DZ twins to attend the same college (78% MZ, 22% DZ).

Deidentification: Seeking differentiation

Twins who emphasized deidentification as their primary mode of sibling influence indicated a need to branch out on their own, develop their unique identities and interests, and increase their sense of autonomy, independence, and self-reliance. Many twins spoke of “being tired of having overlapping lives,” expressed concerns about an over-reliance on their sibling for support (“I need to be my own person”), and viewed college as an ideal time to increase their differentiation:

We are almost, like, interlinked, and so I want to be able to explore myself and be able to find who I am without him weighing in on that and pressuring me. And so I'm going to go to a different college than him. [Kyle, MZ, MM]

I feel like a very strong need to be like very separate right now. Not in terms of like how much we talk but in terms of like how connected our lives are. [Mia, MZ, FF]

Domains of differentiation

The most common domains in which twins sought differentiation were academic and intellectual pursuits (98%); personality, behavior, and conduct (95%); social relationships (68%), interests, hobbies, and extracurricular activities (60%); career objectives (60%); family relationships (48%); and athletics and sports (45%). Less frequently mentioned domains of differentiation included their physical appearance (25%), artistic expressions (23%), and gender identity and sexuality (5%).

DZ participants were more likely to seek distinct social relationships (55% MZ, 83% DZ) and interests, hobbies, and extracurriculars (45% MZ, 78% DZ).

Rationale for differentiation

The most common rationale given for wanting to differentiate was a desire to establish one’s own identity and build one’s sense of self (83%):

[Being seen as an individual is] very important to me... It was even my New Year’s resolution for this year (laugh). [Tessa, DZ, FF]

I feel like neither of us want to live in like the other one’s shadow or like put ourselves in a position to where we would. [Adam, DZ, MF]

A related rationale was irritation over others treating them as a unit or as interchangeable parts and wanting to demonstrate their individuality (75%):

We were kind of compared to each other a lot growing up. And even in high school… people grouped us as the same person. So it was kind of nice to kind of be like, ‘oh, we actually get to form who we are and like have other people know us for ourselves’… It was definitely like a transition. But it wasn't like a conversation of like, ‘oh, we want to be different people.’ It's like, we are different people. [Hana, DZ, F/NB]

Many spoke of the immense pressure put on them by other people who were “always comparing them” (75%)— which contributed to feelings of self-doubt, and at times, inadequacy.

Because the internal [pressure] was like, I inherently and genetically should be as good as her at everything, so I need to make sure that’s true. And then externally, it’s like people are expecting us to be equal and that would be like, just embarrassing and shameful and feel like a failure if I can’t reach that standard. [Scarlett, MZ, FF]

One MZ dyad stressed the pain they felt when they were sometimes confused with or mistaken for their twin, making it challenging to develop meaningful relationships:

I feel like, especially as twins ... one of us will have like a deep conversation with someone or like, like, some sort of emotional connection with someone, and then, the next day, like, they reference it to the other person, like clearly thinking that they had the talk with like me instead of her, or her instead of me... so disorienting. It's like… I thought I just made this connection with someone, and... it feels completely invalidated. [Mia, MZ, FF]

This concern about forming social relationships with people who aren’t sure “which one you are” was echoed by her sister:

When people meet us… they’d just be trying to get to know us in the context of who we were in comparison to the other… [shakes head] I really never felt like an individual. [Scarlett, MZ, FF]

Some twins expressed wanting to differentiate to follow one’s own path in life (70%) and pursue different interests and goals (58%):

We had different plans in mind. Like she's nursing, I'm business… It was important for us to [each] like follow what we wanted to do. And then it's nice to like, have people know me and not know us as like the twins and stuff. Like having our own, like, identities and like… It felt like, very weird at first, but it's like, so nice to like, have some separation. [Aria, MZ, FF]

Interestingly, 40% of participants expressed concerns of a strong dependency or “codependency,” being aware that they may be too comfortable in their reliance on one another in ways that are “holding them back,” for example, by relieving the social pressure of making new friends in college:

I feel like I could have, would have, like, developed better skills of like being independent and learning how to make friends and like try new things if I didn't like having her at my side… I like obviously loved it and everything, but there are times where I'm like… if I didn't do everything with her, I probably would have gotten better at being able to like going out and do stuff like for myself or like being more independent. [Layla, DZ, FF]

Overall, more DZ than MZ twins voiced pronounced interests in pursuing distinct goals (41% MZ, 78% DZ). In contrast, MZ twins were more likely than DZ twins to express wanting to differentiate due to dependency concerns (50% MZ, 28% DZ).

Steps taken to deidentify

The most common steps taken to differentiate reported were: (1) choosing different majors (68%); (2) different colleges (58%); (3) creating a unique social network (48%); (4) making independent decisions (45%); (5) separating geographically (45%); (6) stressing one’s individuality to others by directly telling them they are distinct (28%); (7) setting rules about how similar they can be (25%); (8) seeking different extracurriculars and activities (25%); and (9) changing one’s appearance to display their uniqueness and to help others tell them apart (25%).

Some of the steps twins reported taking to differentiate did not seem to the coders to be very different at all (e.g., “We both were in track but did different events” or “We’re both microbio majors but I’m into research and she’s into healthcare”). This “superficial differentiation” appeared to enable twins to experience a taste of being unique, perhaps to prepare for more substantial forms of differentiation in the future.

MZ twins were slightly more likely than DZ twins to differentiate by changing their appearance (36% MZ, 11% DZ) and developing rules for how similar they can be (32% MZ, 17% DZ). One female MZ dyad created an elaborate process to ensure they selected different colleges, enlisting a friend to pass only essential information about the admissions process back and forth. The twins subsequently enrolled in universities on opposite coasts. All went well until one sister wanted to transfer to a university in the same city as her twin; this was perceived as a violation of their agreement.

In summary, all participants described ways in which they would like to both be more like, and different from, their twin. However, more twins prioritized differentiation. Those prioritizing similarity (all MZ) desired to sustain closeness and fun in their relationship, saw distinct advantages for working as a team, but did reveal concerns about over-reliance and dependency. Those prioritizing differentiation expressed a desire to establish a unique identity, demonstrate their uniqueness to those who see them as interchangeable, and become more self-reliant.

RQ #2: Is deidentification a developmental process?

Most dyads who prioritized differentiation were able to pinpoint a time when they intentionally began to choose distinct identities or pathways, such as during high school (20%) or even earlier (25% elementary; 10% middle school):

I think I first started kind of differentiating myself when I was like 7 or 8 because, you know, like whenever you're a baby, they put you in the same Halloween costumes, everything like that…When I was eight, I started like reading and I read a lot of books… like getting into books as like a hobby. And my brother does not pick up books at all, ever. [Carly, DZ, MF]

We've always known, like, there was no possible way that we were ever going to do the same thing. [Mia, MZ, FF]

Thirty percent reported that college was the right opportunity to explore “separate lives” with some articulating that by college graduation, they “should” have a clear, unique identity and be self-reliant—able to do things for themselves with minimal support from their twin:

…people get us confused all the time. And like we were so tired of it. Like, it felt like college was the natural time to… make ourselves be our own people without the influence of the other. [Mia, MZ, FF]

Many of those who did feel they took steps to differentiate (45%) reported that it felt “different,” “weird,” or “odd,” especially when they were physically separated; 25% reported discomfort and anxiety at these moments. In contrast, 30% described themselves as thriving— enjoying being seen, and experiencing social relationships, as a unique individual:

I like gained a lot of maturity because I had to think really deeply about like what aspects of, like, myself as an entire human being like had been reliant on her, and, and how I needed to improve. And it just made me like think, really critically, in a way that I think, like maybe other people my age haven't been like forced into a situation where they had to do that yet. [Scarlett, MZ, FF]

Interestingly, several strongly differentiated pairs anticipated that as they pursued different paths, became established in their own careers, or became “different enough,” they would feel less of a need to differentiate.

College was going to be the first opportunity we had to like, um... like choose where we wanted to live and what we wanted to do. And, um, we just knew that we needed to place ourselves in separate environments. But like with a mutual understanding that, like we're going to live together again someday, or near each other. [Mia, MZ, FF]

Many envisioned futures in which they and their twin would live nearby and be heavily involved in each other’s lives:

I feel like she's gonna be the person I pick up the phone and call first forever. And then it will be fun to have sleepovers as adults… We can drink wine and talk crap about our husbands and stuff. And like, I can kick him to the couch and like, the two of us can sleep in the bed together. [Aria, MZ, FF]

DZ twins tended to begin the differentiation earlier than MZ twins, with 89% embarking on this process by high school in contrast to 27% of MZ twins.

Resisting differentiation

As described above, the 5 MZ dyads who sought similarity as their primary mode appeared to resist or wish to delay differentiation. Two of these dyads reported previously attempting differentiation but abandoned this track, realizing they prioritized closeness and similarity over seeking independence and unique identities. Two dyads indicated that they understood the value in ultimately differentiating to help them succeed in other personal and professional relationships (and felt it was expected of them) but wished to delay the process as long as possible:

Sometimes I worry about moving on, as like, this is my kind of last 2 or 3 years to be so close with him, um, and that I'll really have to forge my own path later. Uh, so sometimes I worry about that I don't know what that's gonna look like. [Max, MZ, MM]

In summary, the twins’ reports suggest a developmental progression in which differentiation becomes increasingly important as a way to solidify one’s personal identity. DZ twins tended to begin this progression earlier than MZ twins, who mostly waited until college to pursue different activities, friend networks, and identities. Once a comfortable level of differentiation was achieved (being “different enough”), the twins anticipated being integral parts of one another’s lives in the future.

RQ #3: Links with sibling relationship quality

We next explored whether twins who report greater warmth and emotional closeness and less conflict also tend to seek similarity more often and differentiate less often. Given the small sample size, statistical tests were not conducted; however, some interesting patterns were observed that may inspire future hypothesis testing.

Twins who placed high importance on seeking similarity tended to report slightly more warmth (M = 3.69, SD = 0.31) on the ASRQ-VSF than those who valued differentiation (M = 3.31, SD = .66). Dyads who sought similarity were also rated by researchers as experiencing greater closeness (M = 6.88, SD = .18) than those emphasizing differentiation (M = 5.39, SD = 1.11). No patterns were found with respect to sibling conflict. Whereas those individuals who prioritized seeking similarity reported relatively low levels of conflict (M = 2.08, SD = .36), so did those who valued differentiation (M = 2.08, SD = .80).

The qualitative responses reflected a strong and consistent desire to maintain sibling closeness— even if geographically separated— as the relationship was viewed as a critical source of emotional and tangible forms of support.

I know people out there don’t have someone that understands them the way Penelope does… Even if I don’t tell her, or like don’t even know myself, if I’m upset, she’ll know. She’s like, “What’s wrong?” I’m like, “Nothing.” She’s like, “I can tell.” [Riley, MZ, FF]

I think that for the benefit of everybody and for my life I need to have a strong relationship with my brother…That's where a lot of my happiness and a lot of my memories come from, and, um, I'm just not willing to, um, remove that from my life for it, for anybody else, at least right now. [Max, MZ, MM]

Sibling support was considered as instrumental for facilitating their adjustment to college:

I think it did [bring us closer] first year, like that's when we started realizing that our relationship changed a bit. Because, um, out of all my siblings, um, I struggle with my mental health. So going to college was, my first year was really hard, so. We started like talking and sharing how we felt. [Amelia, DZ, FF]

He really pushes me to be a better student and keeps me accountable. I think that accountability is a big thing that he's, um, kind of put on to me… I would not have the grades that I have right now with, without his influence on me. [Max, MZ, MM]

With emotional closeness came a strong sense of trust:

And I think like ultimately it just comes down to like... we trust each other, I trust her more than anyone else…She's just, like, my partner. [Mia, MZ, FF]

Trusting their twin’s judgment extended to respecting their opinions about other life decisions, such as choosing majors, internships, and even potential romantic partners.

This is very important to me… with any, like, romantic partners… if she doesn’t like them, then I probably shouldn’t either. I generally trust her to suss out bad characters. [James, DZ, MF]

I tell the people that are significant in my life, um, like that's one of the big things I say at the very beginning is, you need to be close to my brother. I need you to work on that relationship, um, because… I will always pick my brother. [Max, MZ, MM]

Disagreements and conflicts were described as brief and readily resolved as it made no sense to stay upset with someone who was essentially “an extension of oneself.” Whereas all twins reported comparing themselves in at least one domain, competition was experienced as a process that drove both to excel, rather than to jealousy or a malevolent rivalry:

[We compare everything.] Always GPA compared to GPA. Then even now we compare resumes all the time. Um. We compare points in sports, how many points we get… We can basically compare everything… It makes me so much better… Always competition for motivation, but not for the end result. [Cody, MZ, MM]

In summary, whereas dyads seeking similarity perceived slightly more warmth in their twinship than those prioritizing differentiation, almost all highly valued the emotional closeness and support they derived from their twinship.

Gender constellation

The predominantly female (72.5%) sample hindered our ability to detect systematic patterns in the interview or ASRQ-VSF responses in accord with gender constellation. However, the 5 dyads who placed high priority on seeking similarity were same-sexed, with 2 MZ female and 3 MZ male. Twins in mixed-sexed dyads reported less of a need to differentiate:

I don't think I've ever like felt like I need to be more similar to him, or more different from him, especially because we're different genders, which I'm really grateful for... because I'm like there's never that comparison of like of who's prettier, or who's more athletic… Like you're gonna look at me and my brother and ask me “who's prettier?” Like I'm gonna say, “It's me.” It is gonna be me! [Sofia, DZ, MF]

Discussion

For many twins, the transition to college may be the first time they consider pursuing different paths and identities. The objective of the current research was to give voice to twins as they reflect upon their experiences and mutual influences during emerging adulthood when they are forming key aspects of their identities and making plans for their futures. Following social learning (Bandura, 1969) and deidentification (Ansbacher & Ansbacher, 1956; Schacter et al., 1976, 1978) theories, we sought to understand how and why twins seek similarity and/or deidentify (Her et al., 2021) and whether one process is more characteristic of twins than the other.

As the twins in this sample illustrate, the decision about which college to attend is not simply a choice about finding a university that best meets their academic interests, but it is a pivotal statement about the nature of their relationship, their personal identities, and their respective desires to embark on this stage of their life together or solo. Although this decision was dependent, in part, on both twins applying and being admitted to the same colleges, we heard that their priorities for following similar versus divergent paths reflected the importance they placed on establishing their own identities, achieving greater self-reliance, and retaining emotional closeness.

Instead of inferring processes of sibling influence, as many previous studies have done, we asked twins directly about how they are forging their identities and making decisions vis a vis their sibling. This qualitative approach enabled us to construct a representation of their experiences in a manner that honored their distinct voices. Unique from many past studies that included only one sibling in a family, we interviewed both members of the twin dyad, enabling us to coordinate their responses and discover overarching patterns of sibling influence (Eisikovits & Koren, 2010). This dyadic approach was particularly informative for gauging whether the twins shared similar perceptions of influence processes and were attuned to the other’s needs and interests. As one research assistant recorded in a memo, “it’s like reading a book with alternating chapters—it gives you more context for their relationship, and a deeper understanding, as you alternate between each sibling’s perspective.”

Results from the qualitative analyses were rich and indicated that 75% of the twins we studied placed importance on differentiation. However, even twins who highly valued differentiation also sought similarity in some domains (and vice versa); hence, as Cassinat and Jensen (2020) and Whiteman et al. (2014) have also observed, social learning and deidentification are likely not mutually exclusive processes.

Social learning aptly described the primary orientations of 5 (25%) MZ dyads who found the interdependence in their relationship to be comforting and supportive, especially valuable during the transition to college. Some expressed that pursuing greater “independence” could be helpful but pointed to cogent factors that drew them to seek similarity and stay firmly connected to their twin. This orientation led them to select the same college, with some even sacrificing personal goals to maintain proximity and closeness. These MZ dyads particularly valued their siblings as someone they could trust and rely on—a “life partner” who fully understood them, was always ready to help, brought fun, and made their lives easier, for example, by relieving the social pressure of making new friends in college. “Better together,” they enjoyed a pronounced closeness that Segal (1999) named, “friendship extraordinaire” (p. 97). These twins also reported slightly higher levels of sibling warmth and emotional closeness than those who sought to differentiate.

The majority of our sample (75%) prioritized differentiation, particularly around domains of personality characteristics, academics and career pursuits, and social relationships. Steps taken to deidentify ranged from the minor and superficial (e.g., dying a streak of green in one’s hair, choosing to major in anthropology rather than their twin’s sociology) to the more extreme (e.g., setting rules about how similar and how geographically proximal they could be in college). The desire to differentiate was intensely expressed by more than half of the MZ sample who sought to demonstrate they were unique individuals and not interchangeable parts of a single unit (“the twins”). All DZ twins prioritized differentiation, and many reported having begun to differentiate early in development, before most MZ twins. Because DZ twins are generally less similar in appearance and behavioral characteristics than MZ twins (Segal, 1999), they may see themselves as distinct, and encounter environments that are more conducive to individuation, earlier in development.

Social learning and deidentification as complementary developmental processes

Whereas many studies of sibling influence processes have positioned social learning and deidentification processes as opposing processes, the findings of the current study suggests that this is not necessarily the case as twins were likely to draw from both types of influence processes. For example, they may seek similarity in academic achievement yet aim to establish distinct social networks. As shown in this study, as well as in others (Campione-Barr et al., 2013; Lindell et al., 2014), understanding the domains in which twins seek similarity and difference, as well as the rationale they provide to explain their significance, are important for clarifying how processes of imitation and differentiation operate and contribute to identity formation. In the current study, we did not predetermine what these domains might be— rather, as participants shared their thoughts, we recorded the domains that were most salient to them. Interestingly, twins who emphasized similarity and differentiation mentioned essentially the same set of domains—personality, academics and career pursuits, and social relationships. However, the steps they took to achieve similarities and differences within these domains differed.

In accord with developmental and life course perspectives, our findings suggest that deidentification, at least among twins, may best be considered as a developmental process (Erikson, 1968) in which twins progress from identification, seeking similarity and imitation early in life, to differentiation and deidentification in late adolescence and emerging adulthood. Following Bowen’s (1978) theory of differentiation of the self, most twins understood that to be successful in future personal and professional relationships, it was important—or expected— that they ultimately establish their own identity and become self-reliant. Most twins in the current study described a progression in which they increasingly differentiated themselves in their identities, personality, academic pursuits, and social relationships. Many, predominantly DZ twins, reported beginning to differentiate as early as elementary, middle, or high school, wanting to be known for their unique characteristics and not one of “the twins.” Some (mostly MZ twins) only began to differentiate as they entered college. And some expressed wanting to delay the process as long as possible, anticipating differentiation in the future when the circumstances of adult life, professions, and romantic relationships demanded greater distance (geographically and emotionally) from their twin. Interestingly, twins who did feel they had achieved some level of differentiation described less of a need to continue differentiating. Feeling more secure in their unique identities, they welcomed opportunities to enjoy experiences with their twin and, potentially, follow similar paths. These observations are in line with Feinberg et al.’s (2003) suggestion that moderate degrees of differentiation might be optimal for enabling siblings to balance their needs for individuation and closeness and that too little and too much differentiation could be associated with more difficult sibling relationships. As Erikson and Bowen theorized, and our results support, striving towards a balance of individuation and closeness is an important developmental challenge during emerging adulthood, especially for twins.

Regardless of zygosity, the twins in this study generally perceived their relationships to be high in emotional closeness and sibling warmth. This was particularly true for dyads who prioritized sibling similarity. However, even dyads who prioritized differentiation were invested in maintaining closeness in their relationship, for example, by envisioning a future in which they played meaningful roles in one another’s lives. For many, the transition to college brought a sense of gratitude for having a person in their lives who deeply understood them, whom they could fully trust and rely on. Disagreements and conflict were described as brief, infrequent, and not emotionally intense. As one participant said, “it hardly makes sense to fight with someone who is essentially like you.” These reports of emotional closeness and limited conflict may also relate to the fact that, as same-aged relationships, twins may have a more egalitarian, rather than hierarchical structure, than mixed-aged sibling relationships.

Does deidentification protect twins from conflict and competition?

Adler (Ansbacher & Ansbacher, 1956), Schacter (Schacter et al., 1976, 1978), and others posited that siblings consciously or unconsciously deidentify to protect themselves from conflict, rivalry, and competition. Whereas the twins in this study frequently discussed differentiation as being important for reducing unwelcome comparisons from others, and undo competition among themselves, none of the twins in the current study articulated a wish to be different specifically to reduce sibling conflict. This finding may be due to the low rates of conflict reported by this sample. However, competition and social comparison were noted by many dyads. Interestingly, competition was understood as an important motivational tool that kept each sibling on track towards achieving personal goals— not as a form of mutual opposition, like conflict. Twins described themselves as competing—about almost anything— in a friendly, but determined fashion, with the goal of promoting both of their success. Serving to keep them connected and aware of each other’s progress, this type of competition (trying to outdo the other but not by too much) may represent one way that twins begin to experiment with differentiation.

Evidence to support the protective functions of deidentification against conflict and rivalry has been reported in studies of non-twin siblings in childhood or adolescence (Schacter et al., 1978); however, most of these studies inferred the presence of deidentification. Whiteman et al. (2009), who directly assessed adolescents’ perceptions of social learning and deidentification, also did not find evidence that deidentification functioned to prevent conflict or rivalry. Future research should further examine the linkages between sibling influence processes and diverse forms of positive and negative sibling engagement among twin and non-twin siblings in emerging adulthood using direct assessment methods, such as interviews and observations.

Implications for practice

As the twins in this study placed high importance on maintaining a strong tie to their self- described “life partner” while also pursuing new opportunities and establishing their unique identity, psychoeducational approaches that incorporate family systems theory constructs (Bowen, 1978; Minuchin, 1974) could be designed to help them take steps towards achieving this balance. Such approaches could foster direct communication between siblings about their respective needs for both independence and closeness in specific domains. Particular attention should be devoted to assisting those who wish to pursue distinct personal goals, but may feel reluctant to do so due, perhaps, to an “unhealthy loyalty” (Lewis, 2023), concern about disappointing or hurting their sibling, or apprehension about pursuing their own path without their sibling’s support.

Whereas the majority of participants described their parents as supporting their individuality since childhood, and advocating for comparable treatment by others, this is likely not the case for all twins. Again, following family systems theory, psychoeducational approaches can be designed to help parents, grandparents, teachers, coaches, and others support the unique characteristics, abilities, and interests of multiples. Similarly, it will be important to help twins advocate for themselves by practicing effective ways to respond to unwelcome comparisons by others who fail to support their unique identities.

Limitations

It is important to recognize the limitations of this study. Although larger than many qualitative studies, the sample was modest, limiting the statistical analysis of quantitative data. The sample was predominantly white, female, and well-educated and may not reflect the experiences of racially and ethnically underrepresented twins and those who do not attend college. We hope that future research will focus on twins’ sociocultural and ethnic-racial identity formation in relation to sibling influence processes (Padilla et al., 2021), which exceeded the scope of this study. Additionally, participants provided retrospective reports of events that may have occurred two or three years prior. Future research should ideally take a prospective approach and begin to follow high schoolers as they begin the application process (or even earlier) to gain a more complete understanding of sibling influence processes throughout the transition to college. Finally, we did not ask participants to report their sexual orientation or disability status and so we encourage future research to examine sibling similarity and differentiation in these respects.

Conclusion

In summary, processes of sibling influence on identity development are likely multi-faceted for twins during emerging adulthood. Evidence was found to support the occurrence of both social learning and deidentification as notable processes of sibling influence; Schacter et al.’s (1978) notion that individuals form their identities, in part, in reaction to how they perceive their siblings appears apt. Results further support the premise that deidentification is a developmental process in which twins experience a growing need in emerging adulthood to establish autonomy and forge unique identities yet remain close and connected. How to help twins establish their autonomy and personal distinctiveness, while retaining emotional closeness, is an important goal for future research and practice.

Supplemental Material

Supplemental Material - Forging one’s identity as a twin: Balancing sibling cohesion and deidentification

Supplemental Material for Forging one’s identity as a twin: Balancing sibling cohesion and deidentification by Laurie Kramer, Cassandra M. Dean, Alexis A. Dean, and Melissa S. Morgan in Journal of Social and Personal Relationships