Paper 1

Introduction: Morality Within Intimate Bonds

• Core Research Question: The research investigates how individuals react when romantic partners, friends, or family members behave unethically, exploring the conflict between personal moral values and the maintenance of cherished relationships.

• Existing Research Gap: Previous studies on moral judgment have focused almost exclusively on responses to transgressions committed by strangers in a social vacuum (e.g., Gray & Wegner, $2009$; Greene et al., $2009$).

• Proposed Theory: The authors suggest that moral perception is not isolated but is fundamentally affected by social relationships. When a close other transgresses, it generates a complex, ambivalent response characterized by leniency toward the transgressor but harshness toward the observer's self-concept.

• Key Findings Summary:

  • Perceptions of Others: Participants reported fewer other-critical emotions (anger, contempt, disgust), more lenient moral evaluations, and reduced desire to punish close others compared to strangers.

  • Perceptions of Self: Participants reported more self-conscious emotions (shame, guilt, embarrassment) and harsher moral self-evaluations when close others transgressed.

  • Relationship Impact: Transgressions had smaller consequences for relationship quality with close others compared to acquaintances.

Theoretical Framework: The Case for Leniency vs. Harshness

• The Case for Leniency:

  • Relational Needs: Maintaining close relationships fulfills fundamental needs and contributes to identity (Baumeister & Leary, $1995$).

  • Cognitive Dissonance: Behaving unethically creates an inconsistency in beliefs about a loved one. To resolve this without exiting the relationship (which is costly), observers may rationalize the behavior.

  • Rationalization Strategies:

    • Act Minimization: Viewing the act as less unethical, harmful, or consequential.

    • Transgressor Decoupling: Separating the behavior from the person's overall moral character, often by using situational justifications or blaming external factors.

• The Case for Harshness toward the Transgressor (The Black Sheep Effect): Previous work on group membership suggests in-group members might be judged more severely to protect group integrity (Marques et al., $1988$). The authors test whether this applies to dyadic intimate bonds.

• The Case for Harshness toward the Self:

  • Moral Contagion: Individuals may feel "contaminated" by the morally reprehensible actions of those they are close to (Eskine et al., $2013$).

  • Shared Identity: A close other's behavior reflects upon the observer's sense of self, potentially leading to vicarious shame or felt responsibility.

• Alternative Explanation: Background Information: The authors test if leniency is simply due to having more positive background information about close others. They compare close others to acquaintances and test relationship length as a proxy for knowledge.

Study 1: Hypothetical Wrongdoings

• Design: A within-subjects design involving $207$ participants ($101$ male, $106$ female) from Amazon Mechanical Turk (MTurk).

• Methodology:

  • Participants thought of a romantic partner and a close friend.

  • Average relationship length: $7.68$ years ($SD = 7.98$) for partners; $11.64$ years ($SD = 8.37$) for friends.

  • Scenarios: Participants read three randomized hypothetical acts paired with different transgressors (partner, friend, or stranger named Adam/Alice):

    1. Stealing a dollar from a donation jar for the homeless to buy a candy bar ($M = 4.37, SD = .92$).

    2. Lying about a hotel room number to charge drinks to a stranger's bill ($M = 4.48, SD = .80$).

    3. Spreading a false rumor that a coworker cheated on his wife ($M = 4.61, SD = .66$).

• Results:

  • Transgressor: Close others were rated as more moral ($t(282.11) = 2.60, p = .01$) and elicited lower other-critical emotions ($t(205.14) = 2.94, p = .004$).

  • Self: Participants reported higher self-conscious emotions for close others than strangers (t(204.83) = 11.42, p < .001).

  • Mechanisms: Act minimization and transgressor decoupling mediated other-critical emotions. Relationship length did not predict judgments.

Study 2: Recalled Unethical Events

• Design: A $2 \times 2$ between-subjects design ($n = 434$) recruited from Prolific Academic.

• Variables: Relationship type (close other vs. stranger) and Instructions (rationalization vs. control).

• Methodology:

  • Recall Task: Participants recalled a real situation where someone they knew (partner, friend, family) or a stranger did something unethical.

  • Rationalization Prompt: Half were guided to select three reasons (e.g., "behavior wasn't that bad," "social pressures") and write sentences justifying the behavior.

  • Coding: Independent raters coded act severity from $0$ (not bad) to $3$ (very bad). Reliability was high ($ICC = .91$).

• Results:

  • Leniency: Participants felt less anger/contempt/disgust (F(1, 423) = 30.96, p < .001, \eta_p^2 = .07) and judged close others as more moral ($F(1, 423) = 11.83, p = .001$).

  • Self: High self-conscious emotions ($F(1, 423) = 5.32, p = .02$) and harsher self-evaluations ($F(1, 423) = 4.23, p = .04$) occurred for close others.

  • Rationalization Manipulation: Directing participants to rationalize reduced other-critical emotions across the board but did not interact with relationship type, suggesting close others rationalize spontaneously.

Study 3: Daily Diary Study (Preregistered)

• Design: A $15$-day experience sampling study ($n = 113$ participants reporting $735$ incidents).

• Methodology:

  • Participants reported daily if they witnessed a partner, friend, family member, coworker, or stranger act unethically.

  • Baseline measures included the Inclusion of Other in Self (IOS) scale and the Investment Model Scale (commitment, satisfaction, investment).

• Key Data Points:

  • Incident Breakdown: Stranger ($30.60\%$), Coworker ($21.40\%$), Family ($17.10\%$), Friend ($15.80\%$), Partner ($15.10\%$).

  • Severity: Acts by close others were coded as less severe than those by strangers (t(113.72) = 3.67, p < .001).

• Results:

  • Transgressor: Close others were judged more moral and observers had lower desire to punish them compared to strangers (t(145.73) = 5.55, p < .001).

  • Relationship: Observers reported higher commitment and closeness even after transgressions for close others compared to coworkers (t(81.89) = 4.28, p < .001).

  • Self: Participants felt less moral on days when they witnessed a close other transgress (restricted to single-incident days; $t(114.89) = 2.23, p = .03$).

Study 4: Novel Immoral Behavior (Laboratory Study)

• Design: Between-subjects lab experiment with dyads ($n = 271$ participants; $146$ dyads).

• Methodology:

  • Participants brought a partner or friend or were paired with a stranger.

  • Phase 1: Participants disclosed neutral info to a stranger (creating an acquaintance condition).

  • Phase 2 (Manipulation): Researchers provided pre-filled false responses ostensibly from the study partner. These indicated the partner lied every day, plagiarized an essay, and would keep all money ($10$) in a dictator game.

• Results:

  • Ratings: Close others were rated as more moral than acquaintances (t(108.54) = 4.86, p < .001).

  • Emotions: Contrary to other studies, participants felt higher other-critical emotions toward close others, potentially because the behavior was presented brazenly without explanation.

  • Self: Self-conscious emotions were higher for close others ($t(100.58) = 2.11, p = .04$).

General Discussion and Implications

• Ecological Validity: The high prevalence of observed transgressions in daily life (approx. $2.5$ acts per week) underscores the importance of contextualizing morality.

• The Irony of Protection: In shielding a close other from harsh judgment, the self absorbs the moral burden, experiencing shame and damaged self-evaluation.

• Social Policing: Leniency toward loved ones might lead to the normalization of unethical behavior and a failure to police moral norms within society.

• Relationship Specifics: Effects were largely consistent across romantic partners, friends, and family, though romantic partners often elicited the most leniency and the strongest self-impact due to high interdependence.

• Limitations:

  • Focus on low-to-moderate severity acts; results may not apply to extreme crimes (e.g., sexual assault).

  • Reliance on recall and crowdsourced samples (MTurk, Prolific).

  • Causal evidence for rationalization as the primary mechanism remains suggestive rather than definitive.