Active and Stored Self-Knowledge

Although individuals accumulate immeasurable amounts of knowledge over their lifespans, at any given moment they can access only a portion of that knowledge. The aspects of self-knowledge held in consciousness make up “active self-knowledge.” Other terms for active self-knowledge are the working self-concept (Markus & Kunda, 1986), the spontaneous self-concept (McGuire, McGuire, Child, & Fujioka, 1978), and the phenomenal self (Jones & Gerard, 1967). On the other hand, “stored self-knowledge” is information held in memory that one can access and retrieve but is not currently held in consciousness. Because different features of the self are active versus stored at different times depending on the demands of the situation, the self can be quite malleable without eliciting feelings of inconsistency or inauthenticity (Swann, Bosson, & Pelham, 2002).

Semantic and Episodic Representations of Self-Knowledge

People possess both episodic and semantic representations of themselves (Klein & Loftus, 1993; Tulving, 1983). Episodic self-representations refer to “behavioral exemplars” or relatively brief “cartoons in the head” involving one’s past life and experiences. For philosopher John Locke, the self was built of episodic memory. For some researchers interested in memory and identity, episodic memory has been of particular interest because it is thought to involve re-experiencing events from one’s past, providing a person with content through which to construct a personal narrative (see, e.g., Eakin, 2008; Fivush & Haden, 2003; Klein, 2001; Klein & Gangi, 2010). Recall of these episodic instances happens together with the conscious awareness that the events actually occurred in one’s life (e.g., Suddendorf & Corballis, 1997).

Episodic self-knowledge may shed light on the individual’s traits or preferences and how he or she will or should act in the future, but some aspects of self-knowledge do not require recalling any specific experiences. Semantic self-knowledge involves memories at a higher level abstraction. These self-related memories are based on either facts (e.g., I am 39 years old) or traits and do not necessitate remembering a specific event or experience (Klein & Lax, 2010; Klein, Robertson, Gangi, & Loftus, 2008). Thus, one may consider oneself intelligent (semantic self-knowledge) without recalling that he or she achieved stellar grades the previous term (episodic self-knowledge). In fact, Tulving (1972) suggested that the two types of knowledge may be structurally and functionally independent of each other. In support of this, case studies show that damage to the episodic self-knowledge system does not necessarily result in impairment of the semantic self-knowledge system. Evaluating semantic traits for self-descriptiveness is associated with activation in brain regions implicated in semantic, but not episodic, memory. In addition, priming a trait stored in semantic memory does not facilitate recall of corresponding episodic memories that exemplify the semantic self-knowledge (Klein, Loftus, Trafton, & Fuhrman, 1992). The tenuous relationship between episodic and semantic self-knowledge suggests that only a portion of semantic self-knowledge arises inductively from episodic self-knowledge (e.g., Kelley et al., 2002).

Recently some researchers have questioned the importance of memory’s role in creating a sense of identity. For example, at least when it comes to perceptions of others, people perceive a person’s identity to remain more intact after a neurodegenerative disease that affects their memory than one that affects their morality (Strohminger & Nichols, 2015).

Conscious and Nonconscious Self-Knowledge (Sometimes Confused With Explicit Versus Implicit)

Individuals may be conscious, or aware, of aspects of the self to varying degrees in different situations. Indeed sometimes it is adaptive to have self-awareness (Mandler, 1975) and other times it is not (Wegner, 2009). Nonconscious self-representations can influence behavior in that individuals may be unaware of the ways in which their self-representations affect their behavior (Greenwald & Banaji, 1995). Some researchers have even suggested that individuals can be unconscious of the contents of their self-representations (e.g., Devos & Banaji, 2003). It is important to remember that consciousness refers primarily to the level of awareness of a self-representation, rather than the automaticity of a given representation (i.e., whether the representation is retrieved in an unaware, unintentional, efficient, and uncontrolled manner) (Bargh, 1994).

A key ambiguity in recent work on implicit self-esteem is defining its criterial attributes. One view contends that the nonconscious and conscious self reflect fundamentally distinct knowledge systems that arise from different learning experiences and have independent effects on thought, emotion, and behavior (Epstein, 1994). Another perspective views the self as a singular construct that may nevertheless show diverging responses on direct and indirect measures of self due to factors such as the opportunity and motivation to control behavioral responses (Fazio & Twoles-Schwen, 1999). While indirect measures such as the Implicit Association Test do not require introspection and as a result may tap nonconscious representations, this is an assumption that should be supported with empirical evidence (Gawronski, Hofmann, & Wilbur, 2006).

Issues of direct and indirect measurement are a key consideration in research on the implicit and explicit self. Indirect measures of self allow researchers to infer an individual’s judgment about the self as a result of the speed or nature of their responses to stimuli that may be more or less self-related (De Houwer & Moors, 2010). Some researchers have argued that indirect measures of self-esteem are advantageous because they circumvent self-presentational issues (Farnham, Greenwald, & Banaji, 1999), but other researchers have questioned such claims (Gawronski, LeBel, & Peters, 2007) because self-presentational strivings can be automatized (Paulhus, 1993).

Recent findings have raised additional questions regarding the validity of some key assumptions regarding research inspired by interest in implicit self-esteem (for a more optimistic take on implicit self-esteem, see Dehart, Pelham, & Tennen, 2006). Although near-zero correlations between individuals’ scores on direct and indirect measures of self (e.g., Bosson, Swann, & Pennebaker, 2000) are often taken to mean that nonconscious and conscious self-representations are distinct, other factors, such as measurement error and lack of conceptual correspondence, can cause these low correlations (Gawronski et al., 2007). Some researchers have also taken evidence of negligible associations between measures of implicit self-esteem and theoretically related outcomes to mean that such measures may not measure self-esteem at all (Buhrmester, Blanton, & Swann, 2011). A prudent strategy is thus to consider that indirect measures reflect an activation of associations between the self and other stimuli in memory and that these associations do not require conscious validation of the association as accurate or inaccurate (Gawronski et al., 2007). Direct measures, on the other hand, do require validation processes (Strack & Deutsch, 2004; Swann, Hixon, Stein-Seroussi, & Gilbert, 1990).

Global and Specific Self-Knowledge

Self-views vary in scope (Hampson, John, & Goldberg, 1987). Global self-representations are generalized beliefs about the self (e.g., I am a worthwhile person) while specific self-representations pertain to a narrow domain (e.g., I am a nimble tennis player). Self-views can fall anywhere on a continuum between these two extremes. Generalized self-esteem may be thought of as a global self-representation at the top of a hierarchy with individual self-concepts nested underneath in specific domains such as academic, physical, and social (Marsh, 1986). Individual self-concepts, measured separately, combine statistically to form a superordinate global self-esteem factor (Marsh & Hattie, 1996). When trying to predict behavior it is important not to use a specific self-representation to predict a global behavior or a global self-representation to predict a specific behavior (e.g., Donnellan, Trzesniewski, Robins, Moffitt, & Caspi, 2005; Swann, Chang-Schneider, & McClarty, 2007; Trzesniewski et al., 2006).

Actual, Possible, Ideal, and Ought Selves

The self does not just include who a person is in the present but also includes past and future iterations of the self. In addition, people tend to hold “ought” or “ideal” beliefs about the self. The former includes one’s beliefs about who they should be according to their own and others’ standards while the latter includes beliefs about who they would like to be (Higgins, 1987). In a related vein, possible selves are the future-oriented positive or negative aspects of the self-concept, selves that one hopes to become or fears becoming (Markus & Nurius, 1986). Some research has even shown that distance between one’s feared self and actual self is a stronger predictor of life satisfaction than proximity between one’s ideal self and actual self (Ogilvie, 1987). Possible selves vary in how far in the future they are, how detailed they are, and how likely they are to become an actual self (Oyserman & James, 2008). Many researchers have studied the content of possible selves, which can be as idiosyncratic as a person’s imagination is. The method used to measure possible selves (close-ended versus open-ended questions) will affect which possible selves are revealed (Lee & Oyserman, 2009). The content of possible selves is also socially and contextually grounded. For example, as a person ages, career-focused possible selves become less important while health-related possible selves become increasingly important (Cross & Markus, 1991; Frazier, Hooker, Johnson, & Kaus, 2000).

Researchers have been interested in not just the content but the function of possible selves. Thinking about successful possible selves is mood enhancing (King, 2001) because it is a reminder that the current self can be improved. In addition, possible selves may play a role in self-regulation. By linking present and future selves, they may promote desired possible selves and avoid feared possible selves. Possible selves may be in competition with each other and with a person’s actual self. For example, someone may envision one possible self as an artist and another possible self as an airline pilot, and each of these possible selves might require the person to take different actions in the present moment. Goal striving requires employing limited resources and attention, so working toward one possible self may require shifting attention and resources away from another possible self (Fishbach, Dhar, & Zhang, 2006). Possible selves may have other implications as well. For example, Alesina and La Ferrara (2005) show how expected future income affects a person’s preferences for economic redistribution in the present.

Accuracy of Self-Knowledge and Feelings of Authenticity

Most individuals have had at least one encounter with an individual whose self-perception seemed at odds with “reality.” Perhaps it is a friend who believes himself to be a skilled singer but cannot understand why everyone within earshot grimaces when he starts singing. Or the boss who believes herself to be an inspiring leader but cannot motivate her workers. One potential explanation for inaccurate self-views is a disjunction between episodic and semantic memories; the image of grimacing listeners (episodic memory) may be quite independent of the conviction that one is a skilled singer (semantic memory). Of course, if self-knowledge is too disjunctive with reality it ceases to be adaptive; self-views must be moderately accurate to be useful in allowing people to predict and navigate their worlds. That said, some researchers have questioned the desirability of accurate self-views. For example, Taylor and Brown (1988) have argued that positive illusions about the self promote mental health. Similarly, Von Hippel and Trivers (2011) have argued that certain kinds of optimistic biases about the self are adaptive because they allow people to display more confidence than is warranted, consequently allowing them to reap the social rewards of that confidence.

Studying the accuracy of self-knowledge is challenging because objective criteria are often scarce. Put another way, there are only two vantage points from which to assess a person: self-perception and the other’s perception of the self. This is true even of supposedly “objective” measures of the self. An IQ test is still a measure of intelligence from the vantage point of the people who developed the test. Both vantage points can be subject to error. For example, self-perceptions may be biased to protect one’s self-image or due to self-comparison to an inappropriate referent. Others’ perceptions may be biased because of a lack of cross-situational information about the person in question or lack of insight into that person’s motives. Because of the advantages and disadvantages of each vantage point, self-reports may be better for assessing some traits (e.g., those low in observability, like neuroticism) while informant reports may be better for others (such as traits high in observability, like extraversion) (Vazire, 2010).

Furthermore, self-perceptions and others’ perceptions of the self may overlap to varying degrees. The “Johari window” provides a useful way of thinking about this (Luft & Ingham, 1955). The window’s first quadrant consists of things one knows about oneself that others also know about the self (arena). The second quadrant includes knowledge one has about the self that others do not have (façade). The third quadrant consists of knowledge one does not have about the self but others do have (blindspot). The fourth quadrant consists of information about the self that is not known to oneself or to others (unknown).

Which of these quadrants contains the “true self”? If I believe myself to be kind, but others do not, who is right? Which is a reflection of the “real me”? One set of attempts to answer this question has focused on perceptions of authenticity. The authentic self (Johnson & Boyd, 1995) is alternatively termed the “true self” (Newman, Bloom, & Knobe, 2014), “real self” (Rogers, 1961), “intrinsic self” (Schimel, Arndt, Pyszczynski, & Greenberg, 2001), “essential self” (Strohminger & Nichols, 2014), or “deep self” (Sripada, 2010). Recent research has addressed both what aspects of the self other people describe as belonging to a person’s true self and how individuals judge their own authenticity.

Though authenticity has long been the subject of philosophical thought, only recently have researchers begun addressing the topic empirically, and definitional ambiguities abound (Knoll, Meyer, Kroemer, & Schroeder-Abe, 2015). Some studies use unidimensional measures that equate authenticity to feeling close to one’s true self (e.g., Harter, Waters, & Whitesell, 1997). A more elaborate and philosophically grounded approach proposes four necessary factors for trait authenticity: awareness (the extent of one’s self-knowledge, motivation to expand it, and ability to trust in it), unbiased processing (the relative absence of interpretative distortions in processing self-relevant information), behavior (acting consistently with one’s needs, preferences, and values), and relational orientation (valuing and achieving openness in close relationships) (Kernis, 2003). Authenticity is related to feelings of self-alienation (Gino, Norton, & Ariely, 2010). Being authentic is also sometimes thought to be equivalent to low self-monitoring (Snyder & Gangestad, 1982)— someone who does not alter his or her behavior to accommodate changing social situations (Grant, 2016). Authenticity and self-monitoring, however, are orthogonal constructs; being sensitive to environmental cues can be compatible with acting in line with one’s true self.

Although some have argued that the ability to behave in a way that contradicts one’s feelings and mental states is a developmental accomplishment (Harter, Marold, Whitesell, & Cobbs, 1996), feelings of authenticity have been associated with many positive outcomes such as positive self-esteem, positive affect, and well-being (Goldman & Kernis, 2002; Wood, Linley, Maltby, Baliousis, & Joseph, 2008). One interesting line of research examines the interaction of authenticity, power, and well-being. Power can increase feelings of authenticity in social interactions (Kraus, Chen, & Keltner, 2011), and that increased authenticity in turn can result in higher well-being (Kifer, Heller, Peruvonic, & Galinsky, 2013). Another line of research examines the relationship between beliefs about authenticity (at least in the West) and morality. Gino, Kouchaki, and Galinsky (2015) suggest that dishonesty and inauthenticity share a similar source: dishonesty involves being untrue to others while inauthenticity involves being untrue to the self.

Valence and Importance of Self-Views

Self-knowledge may be positively or negatively valenced. Having more positive self-views and fewer negative ones are associated with having higher self-esteem (Brown, 1986). Both bottom-up and top-down theories have been used to explain this association. The bottom-up approach posits that the valence of specific self-knowledge drives the valence of one’s global self-views (e.g., Marsh, 1990). In this view, someone who has more positive self-views in specific domains (e.g., I am intelligent and attractive) should be more likely to develop high self-esteem overall (e.g., I am worthwhile). In contrast, the top-down perspective holds that the valence of global self-views drive the valence of specific self-views such that someone who thinks they are a worthwhile person is more likely to view him or herself as attractive and intelligent (e.g., Brown, Dutton, & Cook, 2001). The reasoning is grounded in the view that global self-esteem develops quite early in life and thus determines the later development of domain-specific self-views.

A domain-specific self-view can vary not only in its valence but also in its importance. Domain-specific self-views that one believes are important are more likely to affect global self-esteem than those self-views that one considers unimportant (Pelham, 1995; Pelham & Swann, 1989). As James wrote, “I, who for the time have staked my all on being a psychologist, am mortified if others know much more psychology than I. But I am contented to wallow in the grossest ignorance of Greek” (1890/1950, p. 310). Of course not all self-views matter to the same extent for all people. A professor of Greek studies is likely to place a great deal of importance on his knowledge of Greek. Furthermore, changes to features that are perceived to be more causally central than others are believed to be more disruptive to identity (Chen, Urminsky, & Bartels, 2016). Individuals try to protect their important self-views by, for example, surrounding themselves with people and environments who confirm those important self-views (Chen, Chen, & Shaw, 2004; Swann & Pelham, 2002) or distancing themselves from close friends who outperform them in these areas (Tesser, 1988).

Certainty and Clarity of Self-Views

Individuals may feel more or less certain about some self-views as compared to others. And just as they are motivated to protect important self-views, people are also motivated to protect the self-views of which they are certain. People are more likely to seek (Pelham, 1991) and receive (Pelham & Swann, 1994) feedback consistent with self-views that are highly certain than those about which they feel less certain. They also actively resist challenges to highly certain self-views (Swann & Ely, 1984).

Another construct related to certainty is self-concept clarity. Not only are people with high self-concept clarity confident and certain of their self-views, they also are clear, internally consistent, and stable in their convictions about who they are (Campbell et al., 1996). The causes of low self-concept clarity have been theorized to be due to a discrepancy between one’s current self-views and the social feedback one has received in childhood (Streamer & Seery, 2015). Both high self-concept certainty and self-concept clarity are associated with higher self-esteem (Campbell, 1990).

Stability of Self-Views

The self is constantly accommodating, assimilating, and immunizing itself against new self-relevant information (Diehl et al., 2011). In the end, the self may remain stable (i.e., spatio-temporally continuous; Parfit, 1971) in at least two ways. First, the self may be stable in one’s absolute position on a scale. Second, there may be stability in one’s rank ordering within a group of related others (Hampson & Goldberg, 2006).

The question of the self’s stability can only be answered in the context of a specified time horizon. For example, like personality traits, self-views may not be particularly good predictors of behavior at a given time slice (perhaps an indication of the self’s instability) but are good predictors of behavior over the long term (Epstein, 1979). Similarly, more than others, some people experience frequent, transient changes in state self-esteem (Kernis, Cornell, Sun, Berry, & Harlow, 1993). Furthermore, there is a difference between how people perceive the stability of their self-views and the actual stability of their self-knowledge. Though previous research has explored the benefits of perceived self-esteem stability (e.g., Kernis, Paradise, Whitaker, Wheatman, & Goldman, 2000; Kernis, Grannemann, & Barclay, 1989), a recent program of research on fixed versus growth mindsets explores the benefits of the malleability of self-views in a variety of domains (Dweck, 1999, 2006). For example, teaching adolescents to have a more malleable (i.e., incremental) theory of personality that “people change” led them to react less negatively to an immediate experience of social adversity, have lower overall stress and physical illness eight months later, and better academic performance over the school year (Yeager et al., 2014). Thus, believing that the self can be unstable can have positive effects in that one negative social interaction, or an instance of poor performance is not an indication that the self will always be that way, and this may, in turn, increase effort and persistence.

Organization of Self-Views

Though we have already touched on some aspects of the organization of self (e.g., specific self-views nested within global self-views), it is important to consider other aspects of organization including the fact that some self-views may be more assimilated within each other. Integration refers to the tendency to store both positive and negative self-views together, and is thought to promote resilience in the face of stress or adversity (Showers & Zeigler-Hill, 2007). Compartmentalization refers to the tendency to store positive and negative self-views separately. But both integration and compartmentalization can have positive and negative consequences and can interact with other metacognitive aspects of the self, like importance. For example, compartmentalization has been associated with higher self-esteem and less depression among people for whom positive components of the self are important (Showers, 1992). On the other hand, for those whose negative self-views are important, compartmentalization has been associated with lower self-esteem and higher depression.

Developmental Approaches

Psychologists have long been interested in when and how infants develop a sense of self. One very basic question is whether selfhood in infancy is comparable to selfhood in adulthood. The answer depends on definitions of selfhood. For example, some researchers have measured and defined selfhood in infants as the ability to self-regulate and self-organize, which even animals can do. This definition bears little resemblance to selfhood in adulthood. Nativist and constructivist debates within developmental psychology (and language development in particular) that grapple with the problem of the origins of knowledge also have implications for understanding origins of the self. A nativist account (e.g., Chomsky, 1975) that considers the human mind to be innately constrained to formulate a very small set of representations would suggest that the mind is designed to develop a self. Information from the environment may form specific self-representations, but a nativist account would posit that the structure of the self is intrinsic. A constructivist account would reject the notion that there are not enough environmental stimuli to explain the development of a construct like the self unless one invokes a specific innate cognitive structure. Rather, constructivists might suggest that a child develops a theory of self in the same way scientific theories are developed (Gopnik, 2003).

Because infants and children cannot self-report their mental states as adults can, psychologists must use other methods to study the self in childhood. One method involves studying the development of children’s use of the personal pronouns “me,” “mine,” and “I” (Harter, 1983; Hobson, 1990). Another method that is used cross-culturally, mirror self-recognition (Lewis & Brooks-Gunn, 1979; Lewis & Ramsay, 2004), has been associated with brain maturation (Lewis, 2003) and myelination of the left temporal region (Carmody & Lewis, 2006, 2010; Lewis & Carmody, 2008). Pretend play, which occurs between 15 and 24 months, is an indication that the self is developing because it requires the toddler’s ability to understand its own and others’ mental states (Lewis, 2011). The development of self-esteem has also historically been difficult to study due to a lack of self-esteem measures that can be used across the lifespan. Recently, psychologists have developed a lifespan self-esteem scale (Harris, Donnellan, & Trzesniewski, 2017) suitable for measuring global self-esteem from ages 5 to 93.

The development of theory of mind, the understanding that others have minds separate from one’s own, is also closely related to the development of the self. For example, people cannot make social comparisons until they have developed the required cognitive abilities, usually by middle childhood (Harter, 1999; Ruble, Boggiano, Feldman, & Loebl, 1980). Finally, developmental psychology is also useful in understanding the self beyond childhood and into adolescence and beyond. Adolescence is a time where goals of autonomy from parents and other adults become particularly salient (Bryan et al., 2016), adolescents experiment with different identities to see which fit best, and many long-term goals and personal aspirations are established (Crone & Dahl, 2012).

Biological Approaches

Biological approaches to understanding the origins of the self consider neurological, genetic, and hormonal underpinnings. These biological underpinnings are likely evolutionarily driven (Penke, Denissen, & Miller, 2007). Neuroscientists have debated the extent to which self-knowledge is “special,” or processed differently than other kinds of knowledge. What is clear is that no brain region by itself is responsible for our sense of self, but different aspects of the self-knowledge may be associated with different brain regions. Furthermore, the same region that is implicated in self-related processing can also be implicated in other types of processing (Ochsner et al., 2005; Saxe, Moran, Scholz, & Gabrieli, 2006). Specifically, the medial prefrontal cortex (mPFC) and the posterior cingulate cortex (PCC) have been associated with self-related processing (Northoff Heinzel, De Greck, Bermpohl, Dobrowolny, & Panksepp, 2006). But meta-analyses have found that the mPFC and PCC are recruited during the processing of both self-specific and familiar stimuli more generally (e.g., familiar others) (Qin & Northoff, 2011).

Twin studies of personality traits can shed light on the genetic bases of self. For example, genes account for about 40% to 60% of the population variance in self-reports of the Big Five personality factors (for a review, see Bouchard & Loehlin, 2001). Self-esteem levels also seem to be heritable, with 30–50% of population variance accounted for by genes (Kamakura, Ando, & Ono, 2007; Kendler, Gardner, & Prescott, 1998).

Finally, hormones are unlikely to be a cause of the self but may affect the expression of the self. For example, testosterone and cortisol levels interact with personality traits to predict different levels of aggression (Tackett et al., 2015). Differences in levels of and in utero exposure to certain hormones also affect gender identity (Berenbaum & Beltz, 2011; Reiner & Gearhart, 2004).

Intrapsychic Approaches

Internal processes, including self-perception and introspection, also influence the development of the self. One of the most obvious ways to develop knowledge about the self (especially when existing self-knowledge is weak) is to observe one’s own behavior across different situations and then make inferences about the aspects of the self that may have caused those behaviors (Bem, 1972). And just as judgments about others’ attributes are less certain when multiple possible causes exist for a given behavior, the same is true of one’s own behaviors and the amount of information they yield about the self (Kelley, 1971).

Conversely, introspection involves understanding the self from the inside outward rather than from the outside in. Though surprisingly little thought (only 8%) is expended on self-reflection (Csikszentmihalyi & Figurski, 1982), people buttress self-knowledge through introspection. For example, contemporary psychoanalysis can increase self-knowledge, even though an increase in self-knowledge on its own is unlikely to have therapeutic effects (Reppen, 2013). Writing is one form of introspection that does have psychological and physical therapeutic benefits (Pennebaker, 1997). Research shows that brief writing exercises can result in fewer physician visits (e.g., Francis & Pennebaker, 1992), and depressive episodes (Pennebaker, Kiecolt-Glaser, & Glaser, 1988), better immune function (e.g., Esterling, Kiecolt-Glaser, Bodnar, & Glaser, 1994), and higher grade point average (e.g., Cameron & Nicholls, 1998) and many other positive outcomes.

Experiencing the “subjective self” is yet another way that individuals gain self-knowledge. Unlike introspection, experiencing the subjective self involves outward engagement, a full engagement in the moment that draws attention away from the self (e.g., Csikszentmihalyi, 1990). Being attentive to one’s emotions and thoughts in the moment can reveal much about one’s preferences and values. Apparently, people rely more on their subjective experiences than on their overt behaviors when constructing self-knowledge (Andersen, 1984; Andersen & Ross, 1984).

Interpersonal Approaches

At the risk of stating the obvious, humans are social animals and thus the self is rarely cut off from others. In fact, many individuals would rather give themselves a mild electric shock than be alone with their thoughts (Wilson et al., 2014). The myth of finding oneself by eschewing society is dubious, and one of the most famous proponents of this tradition, transcendentalist Henry David Thoreau, actually regularly entertained visitors during his supposed seclusion at Walden (Schulz, 2015). As early as infancy, the reactions of others can lay the foundation for one’s self-views. Attachment theory (Bowlby, 1969; Hazan & Shaver, 1994) holds that children’s earliest interactions with their caregivers lead them to formulate schemas about their lovability and worth. This occurs outside of the infant’s awareness, and the schemas are based on the consistency and responsiveness of the care they receive. Highly consistent responsiveness to the infant’s needs provide the basis for the infant to develop feelings of self-worth (i.e., high global self-esteem) later in life. Though the mechanisms by which this occur are still being investigated, it may be that self-schemas developed during infancy provide the lens through which people interpret others’ reactions to them (e.g., Hazan & Shaver, 1987). Note, however, that early attachment relationships are in no way deterministic: 30–45% of people change their attachment style (i.e., their pattern of relating to others) across time (e.g., Cozzarelli, Karafa, Collins, & Tagler, 2003).

Early attachment relationships provide a working model for how an individual expects to be treated, which is associated with perceptions of self-worth. But others’ appraisals of the self are also a more direct source of self-knowledge. An extremely influential line of thought from sociology, symbolic interactionism (Cooley, 1902; Mead, 1934), emphasized the component of the self that James referred to as the social self. He wrote, “a man has as many social selves as there are individuals who recognize him and carry an image of him in their mind” (1890/1950, p. 294). The symbolic interactionists proposed that people come to know themselves not through introspection but rather through others’ reactions and perceptions of them. This “looking glass self” sees itself as others do (Yeung & Martin, 2003). People’s inferences about how others view them become internalized and guide their behavior. Thus the self is created socially and is sustained cyclically.

Research shows, however, that reflected appraisals may not tell the whole story. While it is clear that people’s self-views correlate strongly with how they believe others see them, self-views are not necessarily perfectly correlated with how people actually view them (Shrauger & Schoeneman, 1979). Further, people’s self-views may inform how they believe others see them rather than the other way around (Kenny & DePaulo, 1993). Lastly, individuals are better at knowing how people see them in general rather than knowing how specific others view them (Kenny & Albright, 1987).

Though others’ perceptions of the self are not an individual’s only source of self-knowledge, they are an important source, and in more than one way. For example, others’ provide a reference point for “social comparison.” According to Festinger’s social comparison theory (1954), people compare their own traits, preferences, abilities, and emotions to those of similar others, making both upward and downward comparisons. These comparisons tend to happen spontaneously and effortlessly. The direction of the comparison influences how one views and feels about the self. For example, comparing the self to someone worse off boosts self-esteem (e.g., Helgeson & Mickelson, 1995; Marsh & Parker, 1984). In addition to increasing self-knowledge, social comparisons are also motivating. For instance, those undergoing difficult or painful life events can cope better when they make downward comparisons (Wood, Taylor, & Lichtman, 1985). When motivated to improve the self in a given domain, however, people may make upward comparisons to idealized others (Blanton, Buunk, Gibbons, & Kuyper, 1999). Sometimes individuals make comparisons to inappropriate others, but they have the ability (with mental effort) to undo the changes made to the self-concept as a result of this comparison (Gilbert, Giesler, & Morris, 1995).

Others can influence the self not only through interactions and comparisons but also when an individual becomes very close to a significant other. In this case, according to self-expansion theory (Aron & Aron, 1996), as intimacy increases, people experience cognitive overlap between the self and the significant other. People can acquire novel self-knowledge as they subsume attributes of the close other into the self.

Finally, the origins and development of the self are interpersonally influenced to the extent that our identities are dependent on the social roles we occupy (e.g., as mother, student, friend, professional, etc.). This will be covered in greater detail in the section on “The Social Self.” Here it is important simply to recognize that as the social roles of an individual inevitably change over time, so too does their identity.

Cultural Approaches

Markus and Kitayama’s seminal paper (1991) on differences in expression of the self in Eastern and Western cultures spawned an incredible amount of work investigating the importance of culture on self-construals. Building on the foundational work of Triandis (1989) and others, this work proposed that people in Western cultures see themselves as autonomous individuals who value independence and uniqueness more so than connectedness and harmony with others. In contrast to this individualism, people in the East were thought to be more collectivist, valuing interdependence and fitting in. However, the theoretical relationship between self-construals and the continuous individualism-collectivism variable have been treated in several different ways in the literature. Some have described individualism and collectivism as the origins of differences in self-construals (e.g., Gudykunst et al., 1996; Kim, Aune, Hunter, Kim, & Kim, 2001; Singelis & Brown, 1995). Others have considered self-construals as synonymous with individualism and collectivism (Oyserman, Coon, & Kemmelmeier, 2002; Taras et al., 2014) or have used individualism-collectivism at the individual level as an analog of the variable at the cultural level (Smith, 2011).

However, in contrast to perspectives that treat individualism and collectivism as a unidimensional variable (e.g., Singelis, 1994), individualism and collectivism have also been theorized to be multifaceted “cultural syndromes” that include normative beliefs, values, and practices, as well as self-construals (Brewer & Chen, 2007; Triandis, 1993). In this view, there are many ways of being independent or collectivistic depending on the domain of functioning under consideration. For example, a person may be independent or interdependent when defining the self, experiencing the self, making decisions, looking after the self, moving between contexts, communicating with others, or dealing with conflicting interests (Vignoles et al., 2016). These domains of functioning are orthogonal such that being interdependent in one domain does not require being interdependent in another. This multidimensional picture of individual differences in individualism and collectivism is actually more similar to Markus and Kitayama’s (1991) initial treatment aiming to emphasize cultural diversity and contradicts the prevalent unidimensional approach to cultural differences that followed.

Recent research has pointed out other shortcomings of this dichotomous approach. There is a great deal of heterogeneity among the world’s cultures, so simplifying all culture to “Eastern” and “Western” or collectivistic versus individualistic types may be invalid. Vignoles and colleagues’ (2016) study of 16 nations supports this. They found that neither a contrast between Western and non-Western, nor between individualistic and collectivistic cultures, sufficiently captured the complexity of cross-cultural differences in selfhood. They conclude that “it is not useful to characterize any culture as ‘independent’ or ‘interdependent’ in a general sense” and rather advocate for research that identifies what kinds of independence and interdependence may be present in different contexts (2016, p. 991). In addition, there is a great deal of within culture heterogeneity in self-construals For example, even within an individualistic, Western culture like the United States, working-class people and ethnic minorities tend to be more interdependent (Markus, 2017), tempering the geographically based generalizations one might draw about self-construals.

Another line of recent research on the self in cultural context that has explored self-construals beyond the East-West dichotomy is the study of multiculturalism and individuals who are a member of multiple cultural groups (Benet-Martinez & Hong, 2014). People may relate to each of the cultures to which they belong in different ways, and this may in turn have important effects. For example, categorization, which involves viewing one cultural identity as dominant over the others, is associated positively with well-being but negatively with personal growth (Yampolsky, Amiot, & de la Sablonnière, 2013). Integration involves cohesively connecting multiple cultures within the self while compartmentalization requires keeping one’s various cultures isolated because they are seen to be in opposition. Each of these strategies has different consequences.

Finally, the influence of religion remains significant in many parts of the world (Georgas, van de Vijver, & Berry, 2004; Inglehart & Baker, 2000), and so religion is also an important source of differences in self-construal. These religious traditions provide answers to the question of how the self should relate to others. For example, Buddhism emphasizes the interdependence of all things and thus agency does not necessarily reside in individual actors. Moreover, for Buddhists the boundaries between the self and the other are insignificant, and in fact the self is thought to be impermanent (see Garfield, Nichols, Rai, Nichols, & Strohminger, 2015).

Need for Communion, Agency, and Coherence

Understanding what motivates people is one of social psychology’s core questions, and a variety of motives have been proposed. Three motives that are particularly important to self-processes are the need for communion (belonging and interpersonal connectedness), the need for agency (autonomy and competence), and the need for coherence (patterns and regularities). The needs for communion and agency are the foundations of many aspects of social behavior (Baumeister & Leary, 1995; Wiggins & Broughton, 1991). Among attitude researchers, constructs similar to communion and agency (i.e., warmth and competence) represent the two basic dimensions of attitudes (e.g., Abele & Wojciszke, 2007; Fiske, Cuddy, & Glick, 2007; Judd, James-Hawkins, Yzerbyt, & Kashima, 2005). Of even more relevance to the self, communion and agency correspond with the dual forms of self-esteem (e.g., Franks & Marolla, 1976; Gecas, 1971). That is, self-esteem can be broken down into two components: self-liking and self-competence (Tafarodi & Swann, 2001). Self-competence is an evaluation of one’s ability to bring about a desired outcome while the need for communion is an evaluation of one’s goodness, worth, and lovability. Each of these dimensions of self-esteem predicts unique outcomes (e.g., Bosson & Swann, 1999; Tafarodi & Vu, 1997).

Those who do not fulfill their communion needs have poorer physical outcomes such as relatively poor physical health, weakened immune functioning, and higher mortality rates (House, Landis, & Umberson, 1988; Uchino, Cacioppo, & Kiecolt-Glaser, 1996). As far as psychological outcomes, people who lack positive connections with others also experience greater loneliness (Archibald, Bartholomew, & Marx, 1995; Newcomb & Bentler, 1986), while those with rich social networks report higher levels of happiness and life satisfaction (Diener, Suh, Lucas, & Smith, 1999). People’s sense of autonomy also contributes to psychological well-being (Ryff, 1989) and encourages people to strive for high performance in domains they care about. Autonomy strivings can also be beneficial in that they contribute to people’s need for self-growth (e.g., Heine, Kitayama, & Lehman, 2001; Taylor, Neter, & Wayment, 1995).

Finally, a great deal of support exists for the notion that people have a fundamental need for psychological coherence or the need for regularity, predictability, meaning, and control (Guidano & Liotti, 1983; Heine, Proulx, & Vohs, 2006). Coherence is a distinct from consistency because it refers specifically to the consistency between a person’s enduring self-views and the other aspects of their psychological universe (English, Chen, & Swann, 2008). The coherence motive may be even more basic than the needs for communion and agency (Guidano & Liotti, 1983; Popper, 1963). That is, self-views serve as the lenses through which people perceive reality, and incoherence degrades the vision of reality that these lenses offer.

When people feel that their self-knowledge base is incoherent, they may not know how to act, and guiding action is thought to be the primary purpose of thinking in the first place (James, 1890/1950).

Self-Enhancement and Self-Verification Motives

Drawing on Prescott Lecky’s (1945) proposition that chronic self-views give people a strong sense of coherence, self-verification theory posits that people desire to be seen as they see themselves, even if their self-views are negative. Self-views can guide at least three stages of information processing: attention, recall, and interpretation. In addition, people act on the preference for self-confirmatory evaluations ensuring that their experiences reinforce their self-views. For example, just as those who see themselves as likable seek out and embrace others who evaluate them positively, so too do people who see themselves as dislikable seek out and embrace others who evaluate them negatively (e.g., Swann, Pelham, & Krull, 1989). The theory suggests that people both enter and leave relationships that fail to satisfy their self-verification strivings (Swann, De La Ronde, & Hixon, 1994), even divorcing people who they believe have overly positive appraisals of them (for a review, see Kwang & Swann, 2010). People may also communicate their identities visually through “identity cues” that enable others to understand and react accordingly to that identity (Gosling, 2008). People seek verification of their specific as well as global (self-views). They are especially inclined to seek self-verifying evaluations for self-views that are certain or important (Pelham & Swann, 1994; Swann & Pelham, 2002).

For the 70% of individuals with globally positive self-views (e.g., Diener & Diener, 1995), self-verification may look like self-enhancement strivings (Brown, 1986) in that it will compel people to seek and prefer positive feedback about the self. In fact, even people with negative self-views tend to self-enhance when they do not have the cognitive resources available to reflect on their self-views and compare it to the feedback available (Swann, Hixon, Stein-Seroussi, & Gilbert, 1990). In addition, people have a tendency to self-enhance before they self-verify (Swann et al., 1990). Other evidence for self-enhancement includes the tendency for people to view themselves as better than average, though this may be most likely for ambiguous traits that can describe a wide variety of behaviors because the evidence that people use to make self-evaluations is idiosyncratic (Dunning, Meyerowitz, & Holzberg, 1989).

It is important to remember in discussions of self-verification and self-enhancement that people do not seek to see themselves as they actually are but rather as they see themselves. As mentioned in the section on accuracy, this self-view may overlap to varying degrees with “reality” or others’ perceptions of the self.

Identity Negotiation

People’s self-views influence the kinds of relationships they will engage in, and people can take on numerous identities depending on the situation and relationship. Identity negotiation theory (Swann & Bosson, 2008) suggests that relationship partners establish “who is who” via ongoing, mutual, and reciprocal interactions. Once people establish a “working consensus” for what roles each person will take in the relationship (e.g., Swann & Bosson, 2008), their agreed-upon expectations help disconnected individuals collaborate toward common obligations and goals, with some commitment to each other. Identity negotiation processes help define relationships and serve as a foundation for organized social activity. The identities that people negotiate tend to align with their chronic self-views. People follow these identity-negotiating processes, albeit largely unintentionally, during each of several successive stages of social interaction. Identities only survive to the extent that they are nourished and confirmed by the social environment, so negotiating identities in relationships is one way an individual ensures the survival of their self-views.

Personal and Social Self-Knowledge

Researchers have historically distinguished between two types of identity: personal and social. Personal identity refers to those features of the self that distinguish us from others while social identity refers to features of the self that are a source of commonality with others, such as group memberships. Once formed, social identities have a powerful influence on thought and behavior (Tajfel, 1981; Tajfel & Turner, 1979). Social category memberships can influence a person’s self-definition as much or more than idiosyncratic personal attributes (Ray, Mackie, Rydell, & Smith, 2008). One version of social identity theory posits that people enter groups that they view as both positive and distinctive to bolster their self-views (e.g., Abrams & Hogg, 1988). Evidence shows that people display a strong ingroup bias, or tendency to favor their own group relative to outgroups (e.g., Brewer & Kramer, 1985; Tajfel, Billig, Bundy, & Flament, 1971). This bias, along with the outgroup homogeneity effect whereby people see outgroup members as more similar than ingroup members (Linville & Jones, 1980) facilitates people’s ability to dehumanize members of outgroups. Dehumanization, perceiving a person as lacking in human qualities, then allows for the justification and maintenance of intergroup prejudice and conflict (Cortes, Demoulin, Rodriguez, Rodriguez, & Leyens, 2005; Vaes, Paladino, Castelli, Leyens, & Giovanazzi, 2003).

Self-categorization theory, in contrast to emphasizing motivation as in social identity theory, stresses the perceptual processes that lead humans to categorize the world into “us” and “them” (Turner, 1985; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Other approaches argue that social identities reduce uncertainty (e.g., Hogg, 2007, 2012), make the world more coherent (e.g., Ellemers & Van Knippenberg, 1997), or protect people from the fear of death (Castano, Yzerbyt, Paladino, & Sacchi, 2002). Though these approaches emphasize cognitive aspects of group membership, group-related emotions are also an important component of social identity. For instance, intergroup emotions theory proposes that a person’s emotional reactions toward other social groups can change in response to situationally induced shifts in self-categorization (Mackie, Maitner, & Smith, 2009).

Whatever the nature of the motive that causes people to identify with groups, although group memberships are critical for survival, they can also place people in grave danger when they motivate extreme action on behalf of the group. Research on identity fusion, which occurs when the boundaries between one’s personal and social identities become porous, shows how strong alignment with a group can lead to fighting and dying for that group at great personal cost (Whitehouse, McQuinn, Buhrmester, & Swann, 2014). This occurs when people come to view members of their social group as family (Swann et al., 2014).

Some research has investigated how personal and social identities are cognitively structured (Reid & Deaux, 1996). The segregation model of identity assumes that social and personal attributes are distinct (Trafimow, Triandis, & Goto, 1991) while the integration model suggests that identities and attributes coexist in a limited set of cognitive structures. Deaux, Reid, Mizrahi, and Cotting (1999) suggest that what constitutes social versus personal identity should not be determined by the attribute itself but rather the function it is serving (i.e., connecting the self to other people or distinguishing the self from other people). Similarly, optimal distinctiveness theory (Brewer, 1991) argues that individuals have an inherent drive to identity with groups but an equally important drive to maintain their individuality. To cope, they strive to find a balance between these opposing forces by finding an identity that supports both the individual’s need for autonomy and affiliation.

Gender

For most people, gender and ethnicity are important social identities, and there is variation in the strength of people’s identification with these groups (Luhtanen & Crocker, 1992). In terms of how gender affects the expression of the self, girls are often socialized to prioritize the qualities that align them to others, while boys are taught to prioritize the qualities that distinguish and differentiate them from others (e.g., Spence, Deaux, & Helmreich, 1985). Moreover, women’s self-esteem tends to be connected more to their relational qualities, while men’s self-esteem is linked to their independent qualities (Josephs, Markus, & Tafarodi, 1992).

Though society has made great strides in allowing men and women to embrace identities of their own choosing (e.g., Cotter, Hermsen, & Vanneman, 2004), traditional social expectations about what it means to be a man or a women persist. For example, gender stereotypes have remained constant over the past thirty years even as women have made significant professional and political gains (Haines, Deaux, & Lofaro, 2016). These stereotypes remain entrenched for men as well. England (2010) argues that for the gender revolution to be complete, not only should traditionally male professions and domains be open to women but traditionally female domains should be increasingly occupied by men. This would help move society closer to attaining gender equality while signaling that traditionally female-dominated roles are equally valued.