Running Head: Strengths of the Authentic/Hubristic Model of Pride Word Count: 3970 Please address correspondence to:

The Authentic/Hubristic (or, A/H) Model of Pride has been empirically supported by dozens of studies drawing on thousands of participants. The model involves four distinct components, most central of which is the finding that pride is not a unitary construct but rather is comprised of two distinct facets, referred to as authentic and hubristic pride. In the present article, we review the four critical components of the A/H Model, and explain why Holbrook and colleagues’ critique raises questions for one part of one of these components only—the attribution distinction between the two prides. We then raise questions for Holbrook and colleagues’ alternative model of pride, and conclude by noting several convergences between the two perspectives. Strengths of the Authentic/Hubristic Model of Pride 3 Holbrook, Piazza, and Fessler (this issue) report results from four studies which, they argue, offer a conceptual and empirical challenge to the Authentic/Hubristic (or, A/H) Model of Pride. Here, we raise several concerns regarding their critique, present a challenge to their alternative model, and conclude with several points of convergence between our two perspectives. To begin, we must clarify what the A/H Model of pride is and does. There are at least four interrelated but distinct components of the model: (1) the empirical finding, originally emerging from seven studies of 2,399 participants (Tracy & Robins, 2007), that pride is not a unitary construct but rather involves two semantically and experientially distinguishable facets; (2) the claim that the scales we systematically developed across five of those seven studies, to measure the facets, adequately assess these two constructs; (3) the conceptual interpretation of these two facets as a psychologically adaptive (i.e., pro-social and socially desirable) and psychologically maladaptive (i.e., anti-social and low desirability) form of pride (i.e., “authentic” and “hubristic” pride, respectively, or, AP and HP); and (4) the overarching and multi-faceted theoretical conception of the two facets in terms of the diverse processes that elicit them; their connection to self-esteem, narcissism, and related self-evaluative processes; and their presumed cross-cultural universality and evolutionary functions. Based on our reading, Holbrook and colleagues’ challenge targets the second of these components—the validity of the scales—and one important aspect of the fourth component, the assertion that AP is elicited by attributions to effort and HP by attributions to ability. However, Holbrook and colleagues suggest that their critique is considerably broader, asserting that they question “the psychological validity of the dual facets” (p. 3), and demonstrate that the HP scale “does not measure feelings of pride at all, but rather acknowledgment that one has displayed Strengths of the Authentic/Hubristic Model of Pride 4 pride in an excessive manner” (Abstract). Given their implication that they are offering a challenge against our model in its entirety, we here examine the evidence presented against each of the model’s components. Component 1: Pride is not a unitary construct but rather is comprised of two distinguishable facets. None of the findings reported by Holbrook and colleagues address the question of whether pride is a unitary or two-faceted construct—the central claim of our model. In fact, to challenge this component of the model, researchers would need to begin not with new studies, but by offering an alternative explanation for the findings that emerged previously. This would mean critiquing or reinterpreting the results of our six studies, based on over 2,000 participants, that used factor analysis, cluster analysis, and the pathfinder algorithm (Schvaneveldt, 1990) to show that pride is comprised of two relatively independent facets. This two-facet structure emerged robustly across studies assessing participants’: (a) feelings during a moment of pride, (b) dispositional tendency to feel pride; and (c) ratings of the semantic similarity among many different pride-related feeling states (Tracy & Robins, 2007). The two-facet structure also held controlling for differences in valence, arousal, intensity, and temporality (i.e., whether words seem more trait-like or state-like), suggesting that the facets do not simply reflect differences in these dimensions (Tracy & Robins, 2007). These findings also held across studies examining both individualistic and collectivistic cultures, suggesting that the two-facet structure is not a uniquely Western conceptualization (Tracy & Robins, 2007; Yan et al., in prep.). Finally, this two-facet structure is also consistent with an independently derived theory of pride, which was developed prior to conducting any of these studies (see Tracy & Robins, 2004). Strengths of the Authentic/Hubristic Model of Pride 5 Given this body of work, it seems reasonable to conclude that the two facets of pride reflect a fundamental aspect of human emotional experience. Holbrook and colleagues’ challenge, however, makes little reference to these initial findings in support of our model. Instead, they focus on our subsequent findings, that each pride facet has a distinct and largely divergent set of personality correlates. Although these correlates are informative about the nomological network surrounding each facet, they are secondary to the two-facet structure, and have little bearing on the question of whether HP is a distinct pride experience. Component 2: The facets of pride are reliably and validly assessed by the two 7-item scales we developed. Turning to the second component of our model, the psychometric soundness of the AP and HP scales, Holbrook and colleagues assert that our HP scale does not, in fact, assess individual differences in the tendency to experience hubristic pride. There are at least three ways to support this claim. One could demonstrate that: (a) the HP-scale-items do not cohere to form an internally consistent scale (i.e., reliability); (b) the HP-scale items do not refer to actual pridefeeling states (i.e., content validity); and (c) the HP scale does not does not show a theoretically meaningful pattern of convergent and discriminant correlations with independent measures of the HP experience and other conceptually related and unrelated constructs (i.e., construct validity). Taking each of these in turn, Holbrook and colleagues concede (a), acknowledging the high internal consistency of the HP scale in all their studies (alphas ≥85). Turning to (b), in light of the extant evidence that literally thousands of research participants have had no problem rating the extent to which these items describe both their current and trait-like pride feelings (e.g., Ashton-James & Tracy, 2012; Cheng, Tracy, & Henrich, 2010; Carver, Johnson, & Sinclair, 2010; Damian & Robins, 2012a; 2012b; Dunlop & Tracy, 2012; Holbrook et al., this issue; Orth, Strengths of the Authentic/Hubristic Model of Pride 6 Robins, & Soto, 2011; Tracy, Cheng, Robins, & Trzesniewski, 2009; Tracy & Robins, 2007), it is not clear to us how this demonstration could be performed. In fact, given that all HP items initially came directly from participants who listed words that characterize their own pride feelings and the pride feelings displayed by others (Tracy & Robins, 2004a; 2007), these items necessarily capture what it means – at least to laypeople– to subjectively experience pride. More broadly, it is not clear how one could demonstrate that the HP scale does not measure a form of pride—that something that looks like a feeling, and is self-reported by laypeople as a feeling, is not, in fact, a feeling, but rather a cognitive evaluation of one’s arrogant self-presentation. Indeed, this issue applies to every self-report measure of emotion ever developed. We cannot know whether any measure corresponds to the actual subjective feeling assessed; this equally true of the HP scale we developed as it is of the PANAS (Watson, Clark, & Tellegen, 1988), a measure of emotional experience that has been used in thousands of studies. Short of neuroscientific evidence for distinct pride feelings in the brain, there are simply no methods to address this question. For this reason, point (c), the construction of a comprehensive nomological net around the scales, is crucial, as it is the closest we can come to demonstrating construct validity in the absence of a non-self-report measure. In fact, we and others have constructed such a network; based on a rough count, over 85 theoretically consistent personality correlates have been reported for the two facets; with traits, emotions, and behaviors; assessed via self-and informant-report measures, implicit measures, narrative coding, biographical measures, and performance in a behavioral task (e.g., Ashton-James & Tracy, 2012; Cheng et al., 2010; Carver et al., 2010; Damian & Robins, 2012a; 2012b; Dunlop & Tracy, 2012; Holbrook et al., this issue; Orth et al., 2011; Tracy, Cheng et al., 2009; Tracy & Robins, 2007). These findings have shown that the AP and HP scales have a coherent pattern of convergent and discriminant correlates. Strengths of the Authentic/Hubristic Model of Pride 7 Holbrook and colleagues question these findings, but we would argue that the correlations emerging from their three studies (which, in fact, are largely consistent with previous work) cannot offset the much larger prior literature. In sum, we agree that the AP and HP scales are vulnerable to the problems intrinsic to all self-report measures: there is no way to determine with complete certainty whether participants’ reports are veridical or reflect a more abstract conceptualization of the self as prone to unmerited displays of pride. Nonetheless, none of the findings reported by Holbrook and colleagues justify their claim that these scales are “fundamentally flawed” and should be “abandoned” (p. 14); this suggestion is particularly problematic given that these authors propose no alternative means of assessing pride. Components 3 and 4: AP has largely adaptive correlates and consequences, whereas HP has largely maladaptive correlates and consequences, and the facets are elicited by distinct cognitive appraisals. Although Holbrook and colleagues focus on the correlates of AP and HP, none of their findings counter our claim that AP is largely pro-social and psychologically adaptive, whereas HP is largely anti-social and maladaptive. This third component of our model reflects a longstanding theoretical distinction between pride as a “virtue” and pride as a “sin.” Indeed, one benefit of our model is that it provides a psychologically satisfying, explanatory account of an idea previously disseminated by philosophers, theologians, and novelists, for over a millennium. Although Holbrook and colleagues fail to replicate several of our previously found correlations, the results they present are largely consistent with the good/adaptive/virtuous vs. bad/maladaptive/sinful conceptualization. Strengths of the Authentic/Hubristic Model of Pride 8 However, Holbrook and colleagues’ largest concern regards the attribution distinction between the two facets; our argument that AP tends to be elicited by successes attributed to unstable, controllable causes, whereas HP tends to be elicited by successes attributed to stable, uncontrollable causes (Lewis, 2000; Tangney, Wagner, & Gramzow, 1989; Tracy & Robins, 2004; 2007). In contrast, Holbrook and colleagues argue that any kind of internal attribution promotes AP, whereas external attributions promote HP. We agree that there is reason to question the previously reported attribution patterns for AP and HP; this is an issue we initially raised in our article reporting these patterns (see Tracy & Robins, 2007, pp. 522-523). Indeed, as we explain in detail below, the correlational results that emerged from Holbrook and colleagues’ work (and from Studies 3 and 5 in Tracy & Robins, 2007), can be considered to provide at best weak, and, at worst, inconsistent, support for our appraisal model. However, the only experimental study conducted thus far does provide support for the appraisal component of our model (Tracy & Robins, 2007, Study 4). In this study, we manipulated effort vs. ability attributions using vignettes, and found relatively greater AP in response to effort, and relatively greater HP in response to ability. These were relative differences, so it remains likely that the effort/ability distinction is not the only distinction there is. However, Holbrook and colleagues’ only comment on these results is that our abilityattribution vignette manipulated “extraordinary confidence in one’s genuinely extraordinary abilities” (pp. 13-14). It is not clear what they mean by this; the manipulation prompt we used was, “You recently had an important exam and you didn’t bother studying much for it, but it still seemed very easy to you. You just found out that you did very well on the exam.” In our view, Strengths of the Authentic/Hubristic Model of Pride 9 this prompt describes a fairly commonplace experience, not one that suggests extraordinary confidence or abilities. It is unfortunate that Holbrook and colleagues did not experimentally test their appraisal model, particularly given their strong causal stance on the role of external attributions in eliciting HP. It would be informative to know whether manipulated external attributions for success do, as they predict, lead to greater HP compared to manipulated internal attributions for success; that is, whether people told to attribute a success to something outside the self (e.g., ease of the exam) report higher levels of arrogance and egotism (two HP-scale items) in response to that success than people told to attribute it to their abilities. In our minds, this prediction seems unlikely, but we agree that findings supporting it would offer a strong challenge to our appraisal model. Turning to the correlational results presented by Holbrook and colleagues, a close examination reveals that they do not present as clear-cut a challenge to our model as the authors suggest. In Studies 1a and 1b, correlations between attributions and pride scales were pooled across experimental conditions, including a condition that should elicit shame as well as pride— boasting more than is warranted. This kind of situation would likely promote both attributions to a range of causes (including external ones) and feelings of hubris, but any correlation between the two may be due to spurious factors (i.e., the situation of being told that you boasted more than was warranted suggests both that you were behaving arrogantly and, in at least some cases, that the success was not entirely caused by you). In Studies 2 and 3, correlations were reported between trait measures of pride and attributions for a particular success event. These analyses do not directly challenge our appraisal model, which primarily concerns the elicitors of pride experienced in response to a particular success (i.e., state, not trait pride). The finding that someone who tends to experience AP also tends to attribute positive events to his/her ability does Strengths of the Authentic/Hubristic Model of Pride 10 not mean that attributing a particular success to ability will promote AP in response to that success. For this reason, the results of these studies do not undermine our model of the appraisal antecedents of A/H pride. Similarly, the result that Holbrook and colleagues refer to as “the most glaring challenge to HP as a measure of pride” (p. 12) is their finding, from Study 3, of a weak correlation (r=.15) between trait HP and ratings of the extent to which a particular success was caused by external factors. This finding suggests that people who tend to feel HP are slightly likely to have attributed the success they wrote about during the study as due to external factors; this is consistent with the tendency of these people to report low self-esteem and demonstrate intrapsychic conflict (see Tracy, Cheng, Martens, & Robins, 2011), but does not mean that external attributions for a success promote HP in response. There is also a broader problem with Holbrook and colleagues’ claims about their appraisal findings. They suggest that AP is linked to a “dramatically self-enhanced appraisal style” (p. 10). However, there is no criterion for the truth value of these appraisals. AP is linked to an adaptive appraisal pattern (i.e., internal attributions for success), but there is no way of knowing whether this pattern is inaccurate (i.e., reflects self-enhancement), because people high in AP may in fact be particularly successful due to their own internal merits. Likewise, AP may be associated with self-assessments of high status because people high in AP in fact are high status (as would be consistent with both ours and Holbrook and colleagues’ evolutionary