The Grab Bag Self

Vincent van Gogh, Self-Portrait

Vincent van Gogh, Self-Portrait

What is the “self”? Such a question has had a multitude of answers from philosophers and psychologists throughout history. Although there is an immediate understanding of what I refer to when I say “I”, upon reflection that clarity vanishes. Do I refer to my physical body? That changes over the course of my life as my cells are replaced one by one. (If I have my arm amputated, am I still the same person?) Is it my consciousness? Then I am conceivably a different person when asleep or drunk then when awake or sober. Is it my memories and experiences? Psychology has demonstrated that recalled memories are largely a reconstruction of the brain rather than a true recollection. And what happens if I get amnesia or Alzheimer’s?

All these questions make it difficult to truly pin down what the self entails. We have some sense of continuity over time, but that continuity can be easily broken. So I’d like to take some time to examine, from a psychological perspective, just what it means to have a “self” and to have a sense of self-identity. In the process, I’d like to advance a theory of the self that suggests that at least some of the continuity we experience is illusory. Instead of being a coherent structure, the self is constantly being assembled and reassembled by our minds. So with that said, hang on to your hats, and let’s begin.

Culture and Connection

Independent and interdependent selves

Diagram for independent and interdependent self-schemas (click to enlarge)

The first thing to acknowledge is that there are cultural differences in the way the self is felt. In other words, not everyone has the same sense of self-identity that you do. There are (at least) two patterns of how the self can be organized: independent and interdependent.1 When an independent sense of self exists, the primary referent is the individual’s own thoughts, feelings, and actions. The self is seen as fundamentally distinct or separate from other individuals, and is derived from a sense of stable internal attributes. This pattern is most common in Western cultures. In contrast, when an interdependent sense of self exists, the primary referent is the thoughts, feelings, and actions of others with whom the individual is in relationship. The self is seen as fundamentally connected with and related to others, and is thus contingent on the individual’s position relative to others. This pattern is most common in Eastern cultures (and indeed, in much of the world).

Note that when I say that the self is distinct from others or connected to others, I don’t mean to say that Western people are just “selfish”. Rather, the nature of how one acts selfishly can differ in an independent context. A Westerner would be selfish by acting in his or her own interest at the expense of other individuals. An Easterner would be selfish by acting in the interest of his or her family or work group at the expense of other individuals outside the group. For an interdependent self, the line between the individual and others is blurred—it is not just that the individual is more kind to others or more giving to those around him or her. Independent and interdependent selves are composed in fundamentally different ways.

The fact that the self can be construed in these very distinct ways means that one’s own sense of self is a product of the culture one lives in. What is most interesting is that these patterns of self-organization are stable; both independent and interdependent selves are able to exist and thrive, and there seems to be no obvious advantage favouring one over the other (besides cultural fit).

The Permanence of Personality

Change Over Time

Personality, become more annoyingAnother way that the self might be conceptualized is as a system of enduring traits collectively making up one’s personality. It is important to note that this, in itself, is a very independent idea of the self (based on internal attributes rather than external relationships). Nevertheless, people generally have some sense that their patterns of behaviour remain relatively stable over time, and this constitutes their personality. Given this idea, one can ask the question, “Just how stable is personality?”

Numerous psychologists have attempted to answer this question by giving people personality scales at more than one time interval (longitudinal, in other words) and looking for changes. However, what makes it more difficult is that there are multiple methods to assess changes in personality. One method, rank-order stability, looks at whether individuals maintain their relative placement to each other on different traits over time. In other words, if I am the most extroverted person in a group of people, do I stay at #1 or do I shift downward?

One influential meta-analysis (combining the results of 152 other studies) found that rank-order stability peaked at age 50-59, and then leveled off.2 More specifically, it increased in a steplike pattern at three different points in time: between infancy/toddlerhood (0-2.9) and preschool (3-5.9), between college (18-21.9) and early adulthood (22-29), and between early middle age (40-49) and later middle age (50-59). In addition to this, the peak point at age 50-59 was still well below complete consistency (r = .75), indicating that there was substantial consistency but that personality traits did not stop changing even into middle age. This study and others have also found that the greater the time interval between personality tests, the less rank-order stability is found.

Age and personality change

Personality change across the lifespan (click to enlarge)

In contrast to rank-order stability, however, is mean-level change. This method looks at whether traits in groups of individuals increase or decrease over time. There have been numerous studies using this approach, but one excellent one in particular examined a German sample of 14,718 participants (representative of the German population), ranging from ages 16 to 82 and assessed twice over four years.3 This study found that all five factors of the well-validated Big Five personality scale showed mean-level changes over the four years, and four of these changes were influenced by age: older people were more likely to decrease in extraversion, openness, agreeableness, and conscientiousness, with no influences of age on emotional stability.4 They also found evidence for a reciprocal relationship between personality and life events. Some personality traits made an individual more likely to experience a life event: for example, highly extraverted people were more likely to move in with their partner during the four years between testing. Life events, however, also made individuals more likely to shift in personality: for example, individuals became more conscientious after getting divorced and less conscientious both after having a baby and after retiring. Given the relatively short testing period of four years, of course, it is possible that people who experience important life events only temporarily shift in personality. However, given that the average time between the life event and the second testing time would likely be about two years, this still means that these events can be associated with significant changes in personality that persist for at least that long.

Given this evidence using both rank-order and mean-level methods of assessing change in personality, we can arrive at three conclusions:

  1. personality is generally stable yet still malleable, even in adulthood;
  2. personality is affected by age (or time between testing, more generally) such that general trends in personality change can be found for individuals across the lifespan; and
  3. significant life events can cause (and be the cause of) changes in personality.

Change Across Situations

Personality changesBut more broad than the idea of personality changing over time is the question of whether personality is more fluid than researchers typically believe. Certainly, people generally have some sense of continuity in what the “self” consists of over time and across situations. Then again, people often remember things in ways that are consistent with their preconceived beliefs about how things are “supposed” to be. Do we really have a consistent personality across situations? One relevant factor here is that of independence or interdependence, as mentioned above. Individuals with an interdependent sense of self feel less of a need to act consistently from situation to situation. Their behaviour is more dependent on the context and the roles in which they are placed than on a sense of internal traits or attributes that guide their actions. Thus, we might expect that interdependent people taking a personality test at work and at home might show greater differences than independent people.

However, this idea reveals a bias in the way personality is conceptualized to begin with. By and large, most research in personality is done using trait-based measures, which is based on quite a Westernized view of the self. Personality traits are defined as inherently stable dispositions. One psychologist, Steven Quackenbush, has argued that personality research assumes traits are “transcontextual,” being stable across a variety of situations.5 Based on this assumption, failure to find high levels of stability in personality must be viewed as measurement error of the personality tests themselves rather than an indication of meaningful differences. Indeed, personality tests themselves are often phrased in terms that imply transcendence of context: How does one answer their agreement or disagreement with the item “I see myself as someone who is full of energy” when one’s level of energy depends on the situation, on the people one is with, on the level of sleep one has had the night before, and so on? Personality tests like these force individuals to think in broad terms that ignore contextual differences, and thus can at best only serve as a general average of one’s behaviour across a large variety of situations.

We all have such experiences that come to mind readily. We act differently around our friends than we do around our grandmothers. We understand that people’s behaviours change in interpretable ways as a result of life events: “Since her boyfriend broke up with her, Sandra has been rather withdrawn.” And psychologists know that by reminding individuals of characteristics that they hold, they can change how they act. Something as simple as reminding people of a value they hold strongly can have a meaningful effect on their ability to buffer against failure and negative feedback, on their persistence at difficult tasks, and on their willingness to change negative behaviours such as smoking and unsafe sex.6 Given these details, it is evident that our personality is much more fluid than most of us would care to admit. That is not to say that there is no consistency, but rather that most of the consistency in our personality may instead come from consistency in the situations we are in (same job, same school, same family) rather than from the attributes inside us.

 The “Grab Bag Self”

Grab bagSo what has all this discussion about culture and personality been leading up to? My theory is that the self is a fluid construct that dynamically shapes itself and recreates itself throughout time. It is not a solid construction made of stable attributes that express themselves across situations and time. Instead, it is a collection of bits and pieces: characteristics, identity statements, roles, theories, expectations, and so on. These bits and pieces are continually assembled and reassembled as the self continues through life, depending on what is most salient in the environment at the time. When the situation indicates that it is time to study hard for a test, the self musters up its bits and pieces regarding academic goals, learning strategies, identity statements about being a “good student”, and so on. (Alternately, it might decide that eating chocolate is actually the more salient concern and instead pull out all the eating-related characteristics so that you end up sitting on the couch eating Reese’s Pieces. It will also pull out the “expert procrastinator” identity to match.) It assembles the self according to what is needed at the time based on its own theories and expectations about how the world operates. But more interestingly, these theories and goals and identities need not be internally consistent with each other. When studying we might think about how being a “good student” is central to who we are, but after failing the test we might reassure ourselves by thinking about how school is not really important to us and instead emphasize the centrality of our friends and family to our lives.

I call this the “grab bag self”. It’s a description that probably does not line up with the lived experiences of most people. Again, people generally have a sense of internal consistency over time, but I am arguing that this sense is largely an illusion. However, I believe that this description of the self I advocate lines up quite well with the things that psychology has come to understand about how the human mind works. Humans are simply not the rational, consistent creatures that philosophers and economists have long considered us to be. Instead, we are creatures that operate using heuristics providing us with a reasonably good method for navigating through the world. But these heuristics are not internally consistent (just think of the equally popular sayings, “birds of a feather flock together” and “opposites attract”), and only get activated in situations in which they seem relevant. This means that we use whatever strategies and theories seem to work best in the situation, and at the same time the ones that don’t work get inhibited.

How does this work in practice? Well, since we were infants, we have been developing implicit (largely unconscious) theories about how the world works, how people are supposed to behave, and so on.7 These theories are sometimes modified by new experiences, and sometimes the theories themselves shape how new experiences are perceived. As we grow, our theories become more complex, and we develop ideas about who we are (identities) and how we should think and act (personality). Both these theories and the demands of the environment lead us to develop goals to survive and thrive. But crucially, given that our environments are typically variable and complex, we can begin to hold inconsistent ideas about the self at the same time. When reminded of the clothes on the floor in our room, we say, “I’m such a disorganized person!” On a separate occasion, when reminded of our colour-coded calendar at work, we say, “I’m such an organized person!” The different contexts elicit different theories of the self, and pull for different behaviours and goals. This is not a bad thing; it helps us to adapt to changing environments and switch between different roles when necessary. But crucially, when we are thinking about ourselves as disorganized, we are not as likely to remember the cases that would suggest the opposite!8 This is what allows us to generally live what feels to us to be an internally consistent life. The theory of the self that is active at the time also activates a whole host of beliefs, goals, memories, identities, and characteristics that are relatively consistent with that theory.

With this said, of course, it is true that some theories and identities are more chronically active than others. This could be the case for a variety of reasons: continual reminders from the environment (a woman engineer may be more aware of her gender when continually being surrounded by men), more consistent situations (like a son who works on the family farm vs. a son who works in the city), less complex situations (fewer roles to juggle), greater consistency of past experiences, and so on. There is also the issue that some theories are so strongly ingrained that most or all situations are interpreted in light of that theory, leading to a feedback loop that leads to chronic activation. But in general, I am willing to argue that the consistency of self that we experience is, to a larger portion than most would expect, illusory. It’s just that when we recollect our past experiences, we are also interpreting them in terms of whatever theories and identities are activated at the current time. Of course things seem more consistent with our theories when we interpret them according to those same theories!

The Importance of Being Incremental

Failure is the best way to learnSo why is this important? Why does it matter whether the self is an established, consistent construct or a fluid, dynamic system of bits and pieces? Well, these concepts in themselves can become implicit theories about how the self is “supposed” to operate. And when that happens, it can affect how well we are able to adapt to our changing environment. Carol Dweck and others have identified two common implicit theories that people can have about intelligence: an entity theory, which views intelligence as a fixed quantity that cannot be changed much by learning and effort, and an incremental theory, which views intelligence as malleable and expandable. Although these theories are commonly framed in terms of intelligence, I believe they can be fruitfully expanded to cover theories of the self more broadly. Some people view the self as largely unchangeable, defined by static internal qualities; and other people view the self as dynamic and changeable through effort.

Research on these two implicit theories has shown numerous benefits for an incremental view of intelligence. Robins and Pals (2002) found that college students with an entity view were more likely to set performance goals (e.g., “get good grades”) while those with an incremental view were more likely to set learning goals (e.g., “understand and be able to apply the material”).9 Students with an entity view were also more likely to respond in a helpless way after failure (attributing failure to lack of ability; being ashamed, distressed, and upset; and giving up), whereas students with an incremental view were more likely to respond with a mastery-oriented pattern (feeling determined, excited, and inspired; and trying even harder). Nussbaum and Dweck (2008) did similar research, where students were given a difficult comprehension task on which they did poorly and then were given the chance to examine the strategies that other students (representing a range of scores) used on the task.10 They found that students with an entity view were more likely to defensively repair their self-esteem by examining the strategies of others who did more poorly than they did (downward social comparison), whereas students with an incremental view who did poorly were more likely to opt for self-improvement by examining the strategies of others who did better than they did (upward social comparison). And these implicit theories can affect those around us, as well. Rattan, Good, and Dweck (2012) found that teachers who held an entity theory of math intelligence (vs. an incremental theory) were more likely to try to comfort a student who did poorly on the first test by using strategies that could reduce engagement (e.g., explaining that not everyone is a “math person”, assigning less homework, talking about them dropping the class).11 They found that these attempts to comfort students communicated entity views to them, and in turn lowered their own motivation and expectations for their performance in the class.

Given these findings and others, it is evident that an entity view of intelligence can have detrimental effects for thriving in the world. People with such views react poorly to failure because they believe they do not have, and are incapable of attaining, the proper skills for achieving success. Those who have an incremental view of intelligence, however, are able to understand that flexibly expending effort and adapting to the task at hand can produce mastery and success. This provides them with incredible motivation to achieve those tasks which are most important to them.

The process should work quite similarly for theories about the self in general. An entity view can serve to “reify” the self, turning it into a concrete entity and emphasizing its stability over time. As long as a person’s environment stays consistent, this strategy can be successful. However, given that virtually no one has a completely consistent environment (especially in a modern society that is getting more and more complex), an incremental view of the self can lead to more flexible adaptation and shifting of resources to navigate the current situation. Accepting that one need not always act consistently, especially if the situation calls for something different, can be of huge benefit. It can lead to greater feelings of success and a better sense of “fit” in a wider variety of situations. More than this, an incremental view of the self can help develop a better understanding of how the self truly operates (as a “grab bag” of bits and pieces assembled and reassembled). Knowledge of how the self operates allows us to wield it more effectively. It’s not that people with an entity view can’t be successful and satisfied, but an incremental view allows us to cope with whatever life throws at us. That, to me at least, seems like a beneficial thing.


So let me wrap this up. My argument is that the self is a dynamic system that assembles and reassembles itself according to current goals and cues from the environment. This argument is rooted in basic facts about the psychology of the human mind, and reveals itself in cultural differences and in the changes in personality that people can undergo over time and across situations. Most importantly, holding such a view about the self should have positive benefits for flexibly adapting to situations in their innumerable and beautiful diversity. What can hold us back is not our inability to change, but our existing beliefs about whether change is possible.


  1. Markus, H.R., & Kitayama, S. (2010). Cultures and selves: A cycle of mutual constitution. Perspectives on Psychological Science, 5(4), 420-430. The article is available here. []
  2. Roberts, B.W., & DelVecchio, W.F. (2000). The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126(1), 3-25. The article is available here. []
  3. Specht, J., Egloff, B., & Schmukle, S.C. (2011). Stability and change of personality across the life course: The impact of age and major life events on mean-level and rank-order stability of the Big Five. Journal of Personality and Social Psychology, 101(4), 862-882. The article is available here. []
  4. Please note that the names of some of these personality traits are used differently by psychologists than by the general public. To understand more precisely what is meant by these trait names, the Wikipedia page on the Big Five provides a good overview. []
  5. Quackenbush, S.W. (2001). Trait stability as a noncontingent truth: A pre-empirical critique of McCrae and Costa’s stability thesis. Theory & Psychology, 11(6), 818-836. The article is available here. []
  6. What I am describing here is self-affirmation, a technique that has been used by social psychologists in a variety of domains with incredible results. But more broadly, half the work that social and cognitive psychologists do involves priming people with certain concepts, thoughts, beliefs, or attitudes, and then seeing how that impacts their behaviour. Often these effects are short-lived (which is important in a research context if the primed thought is negative!), but if a priming effect creates a feedback loop that changes how people interpret future events, it can result in meaningful, long-lasting change. In one prominent example, a 15-minute writing exercise about a personally important value was able to reduce the academic gap between seventh-grade White and Black students by 40% up to two years later (Cohen, Garcia, Purdie-Vaughns, Apfel, & Brzustowski, 2009). []
  7. Note that some of these may have an evolutionary basis and thus be relatively hard-coded in the brain. For the purposes of this article, however, the origin of the implicit theories is not really important. []
  8. This is called confirmation bias; our mind seeks to confirm, rather than disconfirm, theories. []
  9. Robins, R.W., & Pals, J.L. (2002). Implicit self-theories in the academic domain: Implications for goal orientation, attributions, affect, and self-esteem change. Self and Identity, 1, 313-336. The article is available here. []
  10. Nussbaum, A.D., & Dweck, C.S. (2008). Defensiveness versus remediation: Self-theories and modes of self-esteem maintenance. Personality and Social Psychology Bulletin, 34(5), 599-612. The article is available here. []
  11. Rattan, A., Good, C., & Dweck, C.S. (2012). “It’s ok — Not everyone can be good at math”: Instructors with an entity theory comfort (and demotivate) students. Journal of Experimental Social Psychology, 48, 731-737. The article is available here. []

4 responses to “The Grab Bag Self”


My discontent with the points made here is similar to my discontent with the Myers Briggs personality test. You focus on relatively minor expressions of oneself across a few contexts
instead of focusing on our internal schema, our “life goals” or our virtues (however we wish to define them). Studying hard in one moment while procrastinating in another does not alternate personalities make. Procrastination may just as well be a tactic for managing our internal schema that we are ‘hard working’ or a ‘good student’. Individuals who consider themselves to be good students can oftentimes become overwhelmed with the pressure to succeed and manage their anxiety by postponing working on an assignment.

Furthermore, believing you oftentimes express a particular behaviour (i.e. high energy, laid back, serious) and actually expressing that behaviour are not the same phenomenon. An individual who switches from prioritizing school to friends and family after failing a test may not be expressing alternate personalities but a consistent trait of a noncommittal personality type.

Although I agree with many of the points made here, your conclusions make assumptions that require further support.


Hi Ariel,

I will fully agree with you that some goals can be more central or fundamental than others (whether you want to call them “life goals” or something else), but everyone has many, many goals that are constantly being juggled, some of which may conflict, and some of which will be more salient at any given moment. I’m trying to point out two things.

First, I’m saying that situations pull for different goals. The university kid at her part-time job, in the moment, will not be focused on “being a good student.” That’s the last thing on her mind unless she steps back and reflects on past events, future plans, etc. Our ability to self-reflect makes it possible for us to bring goals back into salience, but that doesn’t change the fact that for most people there is not one central goal that defines every action they take. Multiple goals compete for attention, and the situation can very easily tip the scales toward one or the other. That is what psychologists do all the time with primes. Priming you with food cues increases the activation of the goal to eat. Situations can do the same.

Second, I’m saying that goals can conflict. Two goals can be inconsistent with each other, yet still be held by the same person. We can see this very clearly with people with low self-esteem, who want to feel love and belonging (very central goal), but also want to be right about how the world works (very central goal). Because of their negative self-concept, they can end up thwarting their goal for belonging by convincing themselves that the person of interest must not love them because they are not worthy of love. The two goals are in conflict, and how it plays out is dependent on a lot of complex factors. Most people, of course, don’t have low self-esteem, but everyone has a vast set of goal structures (plural) that help them navigate their lives successfully. That these goals can often come into conflict is glossed over by self-theories about consistency and selective re-interpreting of past behaviour.

So yes, you’re right that seemingly inconsistent behaviours don’t necessarily indicate alternate personalities. (I’m not trying to argue for alternate personalities anyway! At the root of it, I’m basically just arguing that the self is complex.) But as far as I’m concerned, having done research in and taken classes about motivation, trying to root the self in motivation is an even trickier subject than trying to root it in “traits”. Goals are fundamentally dynamic, which means even more fluidity of the self than a trait-based model would suggest.

Anyway, I know you’re not wanting to argue about this, and I really did not write this with the intent to prove you wrong. I am just trying to lay out my ideas about the self in more detail. Thanks for sparking the idea to write about it in the first place, and thanks for being willing to entertain my notions even if you don’t agree with them 🙂