Praise and condemnation



Running head: PRAISE AND CONDEMNATION

Do Those Who Throw Praise Also Throw Stones?: Prescriptive and Proscriptive Moralities in Judgment of Moral Character

Word Count: 3,999

Address correspondence to the first author at wiltermuth_scott@gsb.stanford.edu

Abstract

Our studies examined whether the centrality of morality to the self leads individuals both to praise moral behaviors and condemn immoral ones. Across three studies, factor analyses revealed that the tendency to praise others for positive behaviors was orthogonal to the tendency to condemn others for negative behaviors. Further, we found that these distinct tendencies can be assessed using existing morality scales. Our results enable us to refute the alternative hypothesis that simply caring deeply about morality leads individuals to both praise moral behaviors and condemn immoral ones. The research instead suggests that individuals who are most praising of positive behavior are not necessarily those who are most condemning of negative behavior, as orthogonal conceptions of morality influence each type of judgment.

Keywords: Morality, Moral Identity, Judgment, Praise, Condemnation

Do Those Who Throw Praise Also Throw Stones?: Prescriptive and Proscriptive Moralities in Judgment of Moral Character

Does caring deeply about morality lead people both to condemn others for performing immoral behaviors and praise others for performing moral ones? Do people even give others moral credit for their good deeds? While many people would agree that willfully inflicting harm makes you immoral, people agree much less about whether performing good deeds makes you moral. What drives this disagreement? We propose that the tendency to give moral credit for pro-social behavior is distinct and independent from the tendency to assign moral blame for bad deeds. We therefore argue that the individuals most praising of positive behavior are not necessarily those who are most condemning of negative behavior. We further propose that the inclusion of these distinct dimensions in individuals’ conceptions of morality can be assessed using existing measures of moral identity, and that their study can enrich our understanding of moral judgment.

Morality of Duty vs. Morality of Aspiration

Philosopher Fuller (1969) distinguishes between a morality of duty, focused on meeting minimally acceptable standards of behavior, and a morality of aspiration, focused instead on maximizing virtue. In a morality of duty, punishment befalls people who violate minimal standards, but reward is not issued for those exceeding those standards (Hamilton, Blumenfeld, & Kushler, 1988). In contrast, in a morality of aspiration, one can be more moral by going beyond minimal standards. Table 1 presents related distinctions. This distinction resonates with Kant’s (1785) perfect and imperfect duties. The Catholic concept of supererogation (Flescher, 1994; Urmson, 1958), i.e., conduct that is “morally good but not (strictly) required,” would only be meaningful in a morality of aspiration, and not in a strict morality of duty. Janoff-Bulman, Sheikh and Hepp (2009) recently made a similar distinction between proscriptive and prescriptive morality as the two sides of moral regulation, showing, for example, that engaging in proscribed behavior is condemned more than not engaging in prescribed behavior, and, conversely, that engaging in prescribed behavior is praised more than not engaging in proscribed behavior.

Reactions to Positive vs. Negative Moral Behavior

How do positive and negative behaviors affect judgments of moral character? Research suggests that people weigh negative behaviors more heavily than positive behaviors when forming impressions of others (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001; Birnbaum, 1972; 1973; Kanouse & Hanson, 1972; Reeder & Brewer, 1979; Skowronski & Carlston, 1987; 1989; Ybarra, 2001). Indeed, people often discount positive behavior as reflective of normative pressures and social desirability rather than true virtue (Ybarra, 2002; Ybarra & Stephan, 1996; 1999).

Despite this focus on negative behavior, individuals sometimes value positive behavior. Firefighters were praised for their efforts on September 11th, Warren Buffet was acclaimed for donating billions of dollars to charitable foundations, and community “do-gooders” are regularly described in glowing terms. The cynicism about good deeds exposed by social psychologists does not seem universally held. The model presented here casts light on individual differences in giving moral credit.

Orthogonality in Moral Judgment

What explains these contrasting moral judgments of positive behaviors? We contend that only some individuals conceive of morality in terms of aspiration and that it is these individuals, who see virtue in positive behaviors, who praise others for good deeds. We also propose that the more one subscribes to a conception of moral character including negative behavior, the more one should condemn those violating moral norms. Because any individual could conceive of morality in terms of aspiration, duty, or both aspiration and duty, these dimensions are uncorrelated. We therefore propose that the extent to which one is impressed with good deeds is unrelated to the harshness with which one condemns bad ones.

A reasonable (and more parsimonious) alternative would be that one either cares about morality, or one does not – people who describe positive behavior as very moral should be the same who describe negative behavior as very immoral. Such a one-dimensional focus on morality would predict extremity in judgments at both ends of the continuum for an individual who considers morality central to his/her identity (e.g., moral chronics, see Lapsley & Narvaez, 2004). If that were so, how much individuals praise others for good deeds would be related to how much they condemn others for bad deeds. Our model predicts instead that distinct and orthogonal theories of moral character lead individuals to be differentially influenced by good and bad deeds when judging others. We therefore propose that the extent to which one condemns bad deeds is orthogonal to the extent to which one praises good deeds.

Two Dimensions of Moral Identity

Aquino and Reed’s (2002) moral identity scales may capture the degree to which one includes positive and negative behaviors in one’s conception of morality, or subscribes to a morality of duty or aspiration[1]. This ten-item instrument yields two scales which measure how much one cares about being a good person (internalization) and how much one believes that one’s own (positive) behavior reflects that one is a good person (symbolization). The questionnaire asks respondents to imagine a person who possesses moral characteristics such as compassion, fairness, or honesty, and then to indicate the extent to which they value possessing these traits (internalization; e.g., “It would make me feel good to be a person who has these characteristics”) and the extent to which their activities and possessions reflect that they possess these traits (symbolization; e.g., “I am actively involved in activities that communicate to others that I have these characteristics”).

We propose that symbolization taps into a conception of moral character in which positive behavior reflects morality, or a morality of aspiration. If individuals believe that their positive actions communicate that they possess moral traits, they must first think that positive actions can communicate underlying morality rather than merely compliance with social pressure. Thus, high scorers on the symbolization dimension should be more willing to give moral credit for good deeds than low scorers. We therefore predict that the extent to which one praises good deeds is associated with the symbolization scale. Supporting this view, symbolization predicts engagement in positive moral behavior, such as willingness to volunteer for community service (Aquino & Reed, 2002), intrinsic satisfaction and psychological involvement with volunteer activities (Aquino & Reed, 2002) and preferences for giving time rather than money to charitable causes (Reed, Aquino & Levy, 2007).

In contrast, we believe that internalization corresponds to a morality of duty. The scale asks about the importance respondents place on possessing moral traits. The attribution literature suggests that when people consider whether they possess moral traits, they focus foremost on negative behaviors (e.g., Reeder & Brewer, 1979). Therefore internalization should be associated with attention to failures of duty and negative behavior. Indeed, internalization predicts reduced engagement in negative behavior, such as lying in negotiations (Aquino, Ray & Reed, 2003), unethical workplace behavior (Aquino, Reed & Lim, 2004), support for revenge against those causing personal harm (Reed & Aquino, 2003), or for socially unapproved behavior to achieve goals (Aquino & Reed, 2002). Although internalization does sometimes correlate with positive behaviors (Aquino & Reed, 2002; Reed & Aquino, 2003; Reed et al., 2007), in the only example in which both positive and negative behaviors were included, internalization was associated with reduced antisocial behavior but not with increased pro-social behavior (Sage, Kavussanu, & Duda, 2006).

Overview of Studies

Other authors (e.g., Janoff-Bulman et al., 2009) have demonstrated that individuals treat positive and negative behavior differently. The studies presented here go beyond this distinction between types of behavior to show that (a) individuals differ in the extent to which they condemn negative behavior and praise positive behavior, (b) these two dimensions are orthogonal, and (c) they can be predicted by distinct measures of moral identity (Aquino & Reed, 2002). Study 1 does this by utilizing EFA and CFA to analyze the structure of morality ratings of actors in hypothetical vignettes. Study 2 confirms this structure with CFA and a new set of behaviors, controlling for related measures. Study 3 extends this analysis to show that ratings of real-world exemplars and the trait-words people use to describe them are affected by these two distinct conceptions of morality.

Study 1

We predicted that the extent to which one condemns bad deeds would be orthogonal to the extent one praises good deeds, and that people’s ratings of each type of behavior would relate to their conceptions of morality as measured by the internalization and symbolization scales. While we offer clear predictions for how people might differ in judging people performing good deeds and people performing bad deeds, it is less clear how people would differ in judging people who do not perform bad deeds. Two possibilities arise: 1) the fact that one does not cheat or lie could impress people who care about cheating or lying, or 2) not doing bad things (following the attribution literature) may be seen as meaningful only for people who put credence in good deeds. To explore this question we presented participants with one type of negative behavior but two types of positive behaviors, proactive and inhibitive. In other words, we tested whether people who “always tell the truth” are judged differently than those who “never lie.” This study includes two samples, enabling us to use both EFA and CFA.

Method

Participants and procedure

In Sample 1, 114 participants (60% female, age: M = 20) rated the morality of 21 actors in one-sentence vignettes (see Table 2). Seven vignettes depicted proactive positive behavior (α = .85), seven vignettes depicted negative behavior (α = .94), and seven vignettes depicted inhibitive behavior (i.e. not performing negative behavior) (α = .78). Each category of behavior included one vignette related to each of the seven traits or trait pairs that Aquino and Reed (2001) identified as related to morality. We asked participants, “How moral is [the vignette character] for acting like this?” and instructed them to “rate the morality from -10 (extremely immoral) to 10 (extremely moral).”

Results

Consistent with the independence hypothesis, ratings of immoral behaviors (M = -3.32, SD = 3.94) did not correlate significantly with either proactive positive behaviors (M = 6.69, SD = 2.00; r=.01, prep = .52) or inhibitive behaviors (M = 5.60, SD = 2.17); r = .13, prep = .84). Proactive positive behaviors correlated strongly with ratings of inhibitive behaviors (M = 6.77, SD = 2.13), r = .76, prep = .99). A principal component analysis using a Varimax rotation revealed that all vignettes depicting proactive positive behavior and inhibitive behavior loaded highly on the first factor but not on the second factor. As Table 2 displays, vignettes depicting negative behavior loaded on the second factor but not the first factor. The factors respectively explained 32% and 27% of the variance.

We performed a confirmatory factor analysis using structural equation modeling on a second sample of 96 on-line participants (68% female, age: M = 34). A structure with two uncorrelated factors based on the first two dimensions identified in the exploratory analysis provided a significantly better fit to the data (RMSEA = .079, 90% CI = .067 – .103) (weighing both fit and parsimony) than did a one-factor structure (RMSEA = .148, 90% CI = .132 – .163), χ2 (1) = 218.3, prep >.97. Using the three-factor and four-factor structures generated by the EFA produced negative error variances and did not improve the fit of the model. Similarly, allowing factors in the two-factor model to correlate did not improve the fit of the model, χ2 (1) = 3.6, prep .99) and were therefore combined into one category, termed positive behaviors (α =.90).

To determine whether separate conceptions of morality drove ratings of positive and negative behaviors, participants also completed both the Internalization (α =.70) and Symbolization (α =.91, r =.21, p >.92 with internalization) subscales of Aquino & Reed’s (2001) Moral Identity Scale. We controlled for known sources of variation in moral judgment, including gender (Reimer, 1984; Gilligan, 1977), years of education (Haidt, Koller & Dias, 1993), religiosity (1 = Not at all, 7 = Very much), and conservatism (1 = Extremely liberal, 7 = Extremely conservative).

As Table 3 displays, symbolization predicted ratings only of positive behaviors. In contrast, internalization significantly predicted ratings only of negative behaviors. Men rated negative behaviors more negatively, and years of education attenuated ratings of positive behavior.

Discussion

The results suggest that the tendency to praise others for positive behaviors is orthogonal to the tendency to condemn others for negative behaviors. The results also suggest that Aquino and Reed’s (2002) scales measure more than just the importance of morality, but instead also measure also how much individuals subscribe to a morality of virtue and a morality of aspiration.

Study 2

In Study 2 we introduced several scales that directly assess the importance of morality as a criterion for self-worth and an organizing dimension of social perception. We also used vignettes illustrating a broader set of moral issues (e.g., prejudice, environment) to go beyond the traditional domains of morality used in Study 1.

Methods

Participants

One hundred nineteen participants (65% female, age: M = 34.0) in a web-based research pool completed this study for a $7 gift certificate.

Procedures and measures

Participants completed the internalization (α = .65) and symbolization (α =.79, r =.10, prep =.77 with internalization) subscales of Aquino & Reed’s (2001) instrument and the virtue subscale of Crocker et al.’s (2003) Contingency of Self-Worth scale (α =.87). Participants also completed Lapsley and Narvaez’s (2004) chronicity of morality scale, which assesses the extent to which individuals view their social world through the lens of morality: Individuals think of someone they like, someone they dislike, someone they seek out, and someone they avoid, and then list six characteristics of each person. To be classified as moral chronics, participants need to list three or more prototypical moral characteristics for at least three of the five targets.

Participants then rated the morality exhibited in 24 vignettes (available from authors upon request). We included a broader sample of moral issues: sexual propriety, bodily purity (i.e. drugs), civic duty, the environment, women’s rights, caring for less privileged others, honesty, and helping. Eight vignettes depicted proactive positive behavior (α =.85), eight vignettes depicted negative behavior (α =.94), and eight vignettes depicted inhibitive behavior (α =.73). Vignette characters were rated on a 21-point scale ranging from -10 (extremely immoral) to +10 (extremely moral).

Results

As predicted, ratings of immoral behaviors (M = -2.23, SD = 3.85) did not correlate significantly with either proactive positive behaviors (M = 7.39, SD = 1.97; r = .15, prep = .86) or inhibitive behaviors (M = 7.36, SD = 2.41); r = .01, prep = .53. Proactive positive behaviors correlated strongly with mean ratings of inhibitive behaviors (M = 6.77, SD = 2.13), r = .80, prep = .99. To demonstrate that ratings of these vignettes would group along the same moral/immoral dimensions found in Study 1, we conducted a confirmatory factor analysis. This analysis revealed that a model with one factor predicting ratings of positive and inhibitive behaviors and another uncorrelated factor predicting ratings of immoral behaviors provided a reasonable fit to the data (RMSEA = .079, 90% CI = .066 – .092), and provided a better fit than did a one-factor model (RMSEA = .166, 90% CI = .155 – .176), χ2(1) = 580.4, prep > .97. Allowing the two factors to correlate did not improve the fit of the model, χ2(1) = .466, prep = .77.

As Table 3 displays, symbolization positively predicted ratings of positive behavior controlling for other measures of centrality of morality to the self; it did not, however, significantly predict ratings of negative behavior. In contrast, internalization significantly predicted ratings only of negative behavior. Men rated negative behaviors more negatively, and years of education attenuated ratings of positive behavior. People who were “moral chronics” rated negative behaviors as more immoral but did not rate positive behavior as more moral. In fact, moral chronicity was the only measure of moral centrality aside from internalization have a significant zero-order correlation with ratings of negative behaviors (r = -.22, prep > .92). No measure of moral centrality aside from internalization significantly predicted ratings of negative behaviors. The Contingent-Self Worth Virtue subscale correlated significantly with internalization (r = .32, prep > .99) but not symbolization (r = .18, prep < .9), whereas self-reported religiosity correlated significantly with symbolization (r = .32, prep > .99) but not internalization (r = .16, prep < .9). Moral chronicity did not correlate significantly with either internalization or symbolization.

Discussion

Using vignettes addressing a wider range of values, Study 2 replicated the Study 1 finding that distinct conceptions of morality, as captured by Aquino and Reed’s scale, can influence how severely one judges the morality of positive and negative behaviors. Results supported the independence hypothesis, even when controlling for the contingent self-worth one places on virtue and the moral chronicity of the participant. To our knowledge, this is also the first time that self-importance, chronicity and CSW-Virtue are measured in the same participants. We present the correlations between these measures in the supplementary materials online for readers interested in the divergent and convergent validity of these scales.

Study 3

In Studies 1 and 2, participants rated the morality of behaviors performed by hypothetical characters in vignettes. In Study 3, we test whether the tendency to praise real-world exemplars of positive morality, such as the firefighters who entered the World Trade Center to save people on September 11th, 2001, is independent from the tendency to condemn real-world exemplars of immorality, such as those convicted of fraud in the Enron case. We also test whether differences in conceptions of morality predict individuals’ choice of trait-words to describe these real world actors.

Method

Participants and Procedure

One hundred twelve participants (77% female, age: M = 34) in a web-based research pool received a $7 gift certificate for completing the study. After completing the internalization (α = .76) and symbolization (α = .83) scales, participants completed the Lapsley and Narvaez (2004) measure of moral chronicity and Crocker et al.’s (2003) Contingencies of Self Worth – Virtue subscale (α = .86). They then rated the morality of individuals recently discussed in the news, and selected from a list of adjectives those that they believed accurately described these individuals. The individuals had recently made headlines for either a positive act (α = .81) or an immoral act (α = .88) (see Table 4). Adjectives connoted morality (admirable, noble, heroic, and saintly), immorality (evil, cowardly, sleazy, and unprincipled), or neither (intelligent, determined, and misguided). Participants also had the option to choose none of the above. Finally, they provided demographic information.

Results and Discussion

Numerical Ratings

We examined whether numerical ratings of positive actors correlated with ratings of negative actors. Consistent with the independence hypothesis, ratings of negative behaviors (M = -5.79, SD = 3.94) did not correlate significantly with positive behaviors (M = 7.67, SD = 2.14; r= -.14, prep = .85). We also conducted a principal component analysis using Varimax rotation to see if the patterns emerging from ratings of real-life actors would match those emerging from the vignettes of hypothetical behavior. Two principal factors explained, respectively, 36% and 27% of the variance. As Table 4 displays, ratings of real-world actors committing positive actions loaded highly on the first factor but not on the second factor. Ratings of real-world actors committing negative actions loaded highly only on the second factor.

Did different conceptions of morality drive individuals’ tendencies to praise and condemn? Symbolization positively predicted ratings of positive actors but did not predict ratings of negative actors (see Table 3). Internalization did not significantly predict ratings of positive actors, but did predict ratings of negative actors.

Adjective Use

We counted the frequency with which people rating the positive actors chose to check the adjectives “heroic” and/or “saintly,” which are associated with extremely high levels of moral praise (Flescher, 1994; Urmson, 1958). As predicted, people higher on symbolization used the strong adjectives “heroic” or “saintly” marginally more frequently to describe moral actors, ß = .17, t(109) = 1.86, prep = .90, whereas internalization had no effect, prep < .72. Adjectives typically associated with lower levels of moral praise (i.e. “admirable” and “noble”) did not significantly correlate with symbolization.

People scoring high on internalization were more condemnatory in the labels they used to characterize the negative exemplars. They more frequently used the adjectives “sleazy”, “cowardly”, and “unprincipled” (ß = 0.24, t(109) = 2.54, prep = .96) to describe immoral actors. They were not more likely to use the most extreme word, “evil”, (prep .55 | | |

-----------------------

[1] Janoff-Bulman et al. (2009, Study 4) demonstrate that Carver & Scheier’s BIS and BAS scales predict praise, respectively, for resisting proscribed behavior and engaging in prescribed behavior. Given the confounding of type of behavior and omission/commission in that study, however, it is very likely that these correlations reflect the well-documented links between BIS and inhibition and BAS and action, and thus say little about whether the BIS/BAS dimensions relate to prescriptive and proscriptive morality per se. We believe that Aquino & Reed’s scales are conceptually closer to the issues at hand and more likely to predict reactions to the performance of positive and negative behavior.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download