Difficulty: Intermediate to Advanced | Prerequisites: Parts 1 and 2 (schemas, priming, self-concept, self-esteem).
This unit explores why we think the way we do, not just how. It covers motivated reasoning (lay epistemics), the major self-related motives (consistency, enhancement, presentation, assessment), the complexities of implicit versus explicit self-esteem, implicit processes in close relationships and the social construction of reality through language, culture and shared meaning. This is where the course moves from cognitive mechanics to the bigger question of what thinking is for.
Our thinking is not neutral information processing; it serves motives for closure, accuracy and desired conclusions. Self-cognition is shaped by drives for consistency, enhancement and social presentation. Self-esteem has implicit and explicit components that can diverge, with important consequences for narcissism and aggression. Close relationships operate through implicit expectancies and strategies. Reality itself is socially constructed through shared meaning, language and culture.
Lay epistemics (Kruglanski)
A framework identifying three motivational forces that shape how we process information: need for structure/closure, desire for validity and motivation for specific conclusions.
Need for closure/structure
The desire to reach a definite answer rather than remain in uncertainty. Leads to seizing on available information and freezing on it. Can be driven by time pressure or personality.
Desire for validity
The motivation to be accurate, leading to more systematic and effortful processing. Increased by outcome dependency (when your fate depends on understanding someone correctly).
Cognitive dissonance (Festinger)
Psychological tension arising from holding inconsistent cognitions (e.g. "I believe smoking causes cancer" and "I smoke"). Motivates resolution through attitude or belief change.
Self-justification model
The revision of dissonance theory arguing that inconsistency only produces tension when it threatens a positive self-image. We change attitudes to maintain self-enhancement, not merely consistency.
Self-affirmation (Steele)
When people are given a chance to affirm a valued aspect of the self (even unrelated to the dissonant behaviour), the attitude-change effect of dissonance disappears, supporting the self-justification interpretation.
Defensive high self-esteem
High explicit self-esteem that is unstable, fragile and contingent on continued success. Distinguished from secure high self-esteem, which is stable and not easily threatened.
Implicit self-esteem
Automatic, overlearned evaluative associations with the self, measured by the IAT or the name-letter task. Often does not correlate well with explicit self-esteem measures.
Narcissism
An inflated, self-centred view of the self characterised by grandiosity and entitlement. Research links it to the combination of low implicit self-esteem and high explicit self-esteem (a defensive compensation).
Shared reality
The motivation to achieve mutual understanding with others, driven by both interpersonal needs (getting along) and epistemic needs (knowing the world is meaningful and predictable).
Linguistic relativity (Whorfian hypothesis)
The idea that the language we speak shapes how we perceive and think about reality. The strong version (language determines thought) is overstated, but weaker effects are well documented.
Gergen's three frameworks
Romantic (deep interior, passion, character), Modern (reason, functionality, science) and Postmodern (awareness of social construction, multiple selves, irony). All three coexist in contemporary culture and can create internal conflict.
Time pressure increases reliance on stereotypes and priming. In a study with Israeli participants, stereotyping of Ashkenazi versus Sephardi targets occurred only under time pressure.
People high in dispositional need for structure were more influenced by reckless/adventurous primes when judging Donald. Low-need-for-structure people were relatively unaffected.
The mechanism: seizing on available information and freezing, rather than continuing to collect and evaluate.
Operates in the opposite direction: when motivated to be accurate, people process more systematically.
Heuristic versus systematic processing (Chaiken): when a policy change was low-relevance (implemented after graduation), students were persuaded by the likeable speaker with weak arguments. When high-relevance (implemented next year), they were persuaded by the unlikeable speaker with strong arguments.
Outcome dependency increases attention to schema-inconsistent information. Participants who expected to work with a confederate spent more time examining information that contradicted the confederate's claimed traits.
Sometimes we want a particular answer. We believe life is fair, we are invulnerable, we are good people.
Study on partner selection: participants had to identify which of two people was more intelligent based on a sequence of test items. When the objectively more intelligent person was also pleasant, it took 9 items to reach a conclusion. When the less intelligent person was unpleasant, participants concluded after only 6 items, seizing on evidence that confirmed the conclusion they wanted.
Kruglanski's integration: need for closure maps to self-consistency motives, desire for validity to self-assessment and specific conclusions to self-enhancement.
Four broad motives shape how we think about ourselves, illustrated through counter-attitudinal behaviour (getting people to argue for a position they disagree with):
We want our cognitions to fit together. If I freely chose to argue for tuition increases but actually oppose them, the inconsistency produces dissonance, motivating attitude change toward the argued position.
This only works under free choice. If paid to argue, there is sufficient justification and no dissonance.
Among smokers after the Surgeon General's report, 40% questioned the science, versus only 10% of non-smokers. The smokers could not change the behaviour, so they changed the belief.
Inconsistency only bothers us when it threatens our positive self-image. The issue is not that ideas do not fit but that behaving hypocritically makes us feel bad about ourselves.
Steele's self-affirmation finding: if people affirm a valued self-aspect (writing about a personal success) before the counter-attitudinal task, the attitude-change effect disappears. The threat to the positive self-image has already been neutralised.
We manage others' impressions of us, like actors in a play (Goffman). Costumes, props and scripts all serve impression management.
Making the counter-attitudinal process public produces attitude change even in the low-choice condition, suggesting people will shift their expressed attitude to appear consistent in front of others.
We are motivated to perceive ourselves accurately. Self-perception theory is an example: we observe our own behaviour and draw conclusions, just as an outside observer would.
Bem's interpersonal simulations showed that outside observers attributed attitudes to hypothetical participants in the same pattern as the participants themselves, supporting a cool, rational self-perception account over the hot, tension-driven dissonance account.
Resolution: dissonance (with its negative affect) operates when behaviour falls in the "latitude of rejection" (a position the person finds clearly unacceptable). Self-perception operates when behaviour falls in the "latitude of acceptance" (a position close enough to the person's own view). Both processes are real; they just apply in different zones.
Deci measured SE stability over time alongside trait hostility. People with high but unstable SE reported the most anger in response to provocative scenarios, because everyday events threaten their fragile positive self-view.
Measured by the IAT (pairing self + good vs. self + bad) or the name-letter task (people like their own initials).
Implicit and explicit SE correlate weakly (r = 0 to .25), suggesting they tap different constructs.
They predict different outcomes: explicit SE predicted self-reported anxiety in a stressful interview; implicit SE predicted how anxious the person appeared to an outside observer.
Possible explanation: implicit SE reflects overlearned, habitual self-evaluation. If someone spent years thinking "I am a loser," that association persists automatically even after conscious efforts to think positively.
People highest in narcissism score high on explicit SE but low on implicit SE, consistent with the theory that narcissism is a defensive inflation compensating for underlying feelings of inferiority.
This discrepancy also predicts hostile responses to out-group members. In a study on punishment for a bar fight, defensive high-SE participants gave harsher sentences when the perpetrator had an Aboriginal-sounding name.
A computer game repeatedly paired the participant's own name with smiling faces (acceptance) while other names were paired with frowns. For participants who started with low implicit SE, the game boosted their IAT scores and reduced aggressiveness in a subsequent noise-blast task.
This is essentially Pavlovian conditioning of the associative link between self and social acceptance.
An IAT pairing relationship words (commitment, share, together) with positive/negative words correlated with explicit relationship measures. Use of plural pronouns (we, our, us) when listing thoughts about the relationship predicted increased commitment six weeks later, beyond what the explicit measure captured.
Barriers to exit (arranged marriage, sunk investment) weakened the correlation between implicit and explicit attitudes: people who felt trapped rated their relationship more positively on explicit measures than their implicit reactions suggested.
Women who implicitly associated their boyfriend with a superhero-rescuer fantasy (measured by IAT) reported lower aspirations for social power, education and leadership. The explicit version of the same measure predicted nothing. Men showed no such effect.
Avoidantly attached people were faster to identify "hurt" as a word after reading "If I trust my partner...," indicating an automatic expectancy that trust leads to pain.
Both anxious and avoidant individuals were faster to identify negative outcomes in the context of anger, while secure individuals showed no such speeding.
Safe haven strategy: after being primed with failure words, participants were faster on a Stroop task when their loved one's name appeared (attention captured by the safe haven). An avatar task showed people physically moved toward a significant other's name faster after writing about failure.
Commitment maintenance: women in relationships automatically avoided attractive male faces in a virtual-reality balloon task. Men did not show this avoidance unless they practised implementation intentions for defending the relationship ("If an attractive woman approaches, I will...").
Sherif's autokinetic effect: alone, people's estimates of light movement varied widely. In groups, they converged on a norm, and individuals stuck with the group norm even when later tested alone. Often cited as conformity, but it also demonstrates that we construct our sense of reality together.
Intersubjectivity: the child and parent reach shared understanding through mutual attention and emotional echoing. Failed intersubjectivity (misunderstanding what someone implies) is immediately noticeable.
Definition of the situation: in any social encounter, participants must agree on what kind of situation they are in (lecture, date, emergency). Bystander intervention fails partly because people look to others to define the situation, and if no one reacts, the situation is defined as non-threatening.
Participants told a confederate liked or disliked a target person shaped their descriptions accordingly. A week later, their own memories of the target were distorted in the direction of their communication. Labelling someone for an audience changes your own memory.
The Whorfian hypothesis: language shapes thought. The strong version is overstated, but real effects exist.
Bilingual study: English-Chinese bilinguals read descriptions of people. When thinking in English, they used English personality categories (e.g. "artistic type") and made predictions consistent with those categories (e.g. "would take hallucinogenic drugs"). When thinking in Chinese, different categories (e.g. "shi gu," a concept with no English equivalent) shaped memory and prediction differently.
Miller's cross-cultural study: American participants became increasingly dispositional in their attributions with age ("she's just like that"), while Indian participants became increasingly contextual ("she has a sick mother"). Young children in both cultures started similarly.
The fundamental attribution error (over-attributing behaviour to disposition) is partly a western cultural product.
Romantic (1800s, persists today): deep interior, passion, character, soul. Depth psychology (Freud, Jung). Soap operas. The self is discovered through introspection.
Modern (1700s, resurgent 1950s): reason, functionality, science. Behaviorism, cognitive therapy. Walmart, Sherlock Holmes, MacGyver. The self is understood through beliefs, attitudes, traits and goals.
Postmodern (late 20th century): awareness of social construction, multiple selves, irony, identity flexibility. Seinfeld, Warhol, sampling. The self is performed and negotiated, not discovered.
All three coexist in contemporary culture. The "head versus heart" conflict (modern vs. romantic) in career decisions is one example. Freud's work on internal conflict can be partly understood as tension between these frameworks.
"Cognitive dissonance and self-perception theory are competing, so only one can be right." They apply in different zones. Dissonance (with negative affect) operates when behaviour is clearly rejected. Self-perception (cool, rational inference) operates when behaviour is close to one's own position.
"Narcissists truly believe they are superior." Research suggests narcissism is often a compensatory inflation: high explicit SE masking low implicit SE. The fragility is the point.
"Shared reality means conformity." Sherif's autokinetic study is often simplified to "people conform," but the deeper point is that we construct our sense of what is real through social interaction, and those constructed norms persist even in private.
⚠️ Be able to explain Kruglanski's lay epistemics framework and map its three components onto the self-related motives (closure → consistency, validity → assessment, specific conclusions → enhancement).
⚠️ Know the difference between Festinger's original dissonance theory and the self-justification revision, and be able to describe Steele's self-affirmation study as the critical test.
⚠️ Understand how implicit and explicit SE are measured, how they diverge and what the discrepancy pattern predicts (narcissism, hostility, discrimination).
⚠️ The glass slipper effect, the commitment-maintenance virtual-reality study and the safe-haven Stroop study are good examples of implicit processes in relationships. Be able to describe one in detail.
⚠️ Gergen's three frameworks are testable as a cultural analysis. Be prepared to give an example of each and explain how they can create internal conflict.
True or False: Steele's self-affirmation study supports Festinger's original consistency-based theory of dissonance. (False – it supports the self-justification/enhancement revision.)
Fill in the blank: People highest in narcissism tend to show ______ implicit SE combined with ______ explicit SE. (Low; high.)
True or False: In Miller's cross-cultural study, young children in the US and India started with similar attribution patterns. (True – the cultures diverged with age.)
Fill in the blank: The "saying is believing" study showed that communicating a biased description to an audience distorted the communicator's own ______ a week later. (Memory.)
Q: A student argues that cognitive dissonance is simply people lying to experimenters to look consistent. How would you respond using the distinction between self-presentation and self-justification accounts?
A: The self-presentation account does suggest people change their stated attitude to appear consistent publicly. However, the self-justification account goes deeper: attitude change occurs even in anonymous conditions when there is free choice, which rules out pure impression management. Steele's self-affirmation study is the critical evidence. If dissonance were just about looking good to the experimenter, affirming an unrelated positive self-aspect should not eliminate the attitude change, yet it does. This shows the mechanism is internal (protecting a positive self-image) rather than purely performative. That said, when the process is made fully public, attitude change increases even in low-choice conditions, indicating self-presentation can layer on top of the internal process.
Q: Explain how the combination of implicit and explicit self-esteem scores maps onto narcissism and hostile out-group behaviour. Why does the discrepancy matter more than either score alone?
A: People highest in narcissism show high explicit SE ("I am extraordinary") combined with low implicit SE (slow to pair self with positive on the IAT). The theory is that narcissism is a defensive inflation: the person's gut-level self-evaluation is negative, and the conscious self-view is inflated to compensate. This combination is fragile because any threat that penetrates the explicit shield exposes the implicit negativity. The discrepancy predicts hostile responses because these individuals are motivated to denigrate others (especially out-group members) to shore up their fragile self-view. In the bar-fight study, it was specifically the defensive high-SE participants who gave harsher punishments when the perpetrator appeared to be Aboriginal. Neither high explicit SE alone nor low implicit SE alone predicted this pattern as strongly as the combination.
Q: Describe the "saying is believing" study and explain what it demonstrates about the social construction of memory.
A: Participants read an ambiguous description of a target person and then described that person to a confederate they believed either liked or disliked the target. Participants tailored their descriptions accordingly (more positive when the confederate liked the target). A week later, their own memories of the original description were distorted in the direction of what they had communicated. This demonstrates that the act of communicating for an audience does not merely filter what is expressed; it actively reshapes the communicator's own stored representation. Reality, in this case one's own memory of another person, is constructed through the process of achieving shared understanding with others.
Lay epistemics connects back to dual-process models in Part 1 (automatic vs. controlled) and forward to terror management in Part 4 (where need for closure and worldview defence are central). The implicit/explicit SE discrepancy connects to Fazio's implicit prejudice paradigm from Part 1 and to the sociometer model in Part 2. Social construction and Gergen's frameworks provide the cultural backdrop for understanding why self-evaluation standards differ across individuals and cultures, linking to the attachment work in Part 2 and the existential approach in Part 4.
Motivated cognition, lay epistemics, Kruglanski, need for closure, need for structure, seizing and freezing, desire for validity, heuristic versus systematic processing, Chaiken, outcome dependency, motivation for specific conclusions, cognitive dissonance, Festinger, self-consistency, self-justification, self-enhancement, self-affirmation, Steele, self-presentation, Goffman, impression management, self-assessment, self-perception theory, Bem, latitude of rejection, latitude of acceptance, counter-attitudinal behaviour, defensive high self-esteem, secure high self-esteem, unstable self-esteem, implicit self-esteem, IAT, name-letter task, narcissism, implicit cognition in relationships, plural pronouns, barriers to exit, glass slipper effect, attachment expectancies, safe haven, commitment maintenance, implementation intentions, shared reality, intersubjectivity, autokinetic effect, Sherif, saying is believing, linguistic relativity, Whorf, bilingual cognition, culture and attribution, fundamental attribution error, Miller, Gergen, romantic framework, modern framework, postmodern framework, PSY 320, social psychology