What is learning if not perpetual doubt? Pyrrho, the first Greek skeptic philosopher, advised sophists never to freeze their judgment because of the “acataleptic” nature of knowledge, that is, the impossibility to reach absolute certainty in knowledge. Many famous developmental and social psychologists were influenced by these Pyrrhonian conceptions (e.g., Berlyne, 1960; Kruglanski, 2004; Piaget, 1936). Notable amongst them are Mugny and Doise, who proposed that socio-cognitive conflict played a central role in cognitive development (Mugny & Doise, 1978). For four decades, Mugny and his colleagues have shown that receiving conflicting information from another—when not freezing the epistemic process—has the potential to influence the receiver, lead him/her to integrate the two pieces of knowledge, and eventually result in learning (for reviews, see Doise & Mugny, 1984; Pérez & Mugny, 1993; Mugny, Butera, Quiamzade, Dragulescu & Tomei, 2003; Quiamzade, Mugny & Butera, 2013). Despite these benefits, conflicts are generally prevented in educational settings (DeCecco & Richards, 1974) and educators tend to avoid them (Uline, Tschannen-Moran & Perez, 2003). According to Johnson and Johnson (1985), this can be explained by the fact that “conflict is perceived negatively in our society and in our schools” (p. 354). In this research, we argue that the perceived value of conflict in laypeople may not be as uniform as generally supposed. Specifically, we use the so-called judge paradigm (Dubois, 1994) to determine how (target) individuals regulating conflict in different ways may be viewed and valued differently.

A Taxonomy of Socio-Cognitive Conflict Regulations

A situation of confrontation with a disagreeing other on a learning task has traditionally been referred to as socio-cognitive conflict. This term derives from the fact that disagreement on a learning task includes a cognitive component (related to doubt about the answer of the task) and a social component (related to doubt about the one’s and the interactant’s relative competence; for a short historical account of the development of this term, see Quiamzade et al., 2013). Given these two components, there are two main ways in which conflict can be regulated: epistemic and relational (Butera, Darnon & Mugny, 2010).

First, when the climate of an interaction is not threatening, the cognitive component of conflict overcomes its social component. In such a case, conflict regulation behavior can be described as both focused on the self (the individual’s position) and the interactant (the other’s position; for such a conceptualization, see Blake & Mouton, 1964; Pruitt & Rubin, 1986; Thomas & Kilmann, 1974). In other words, conflicting individuals question the validity of each standpoint regardless of the source of the information. They use task-related criteria to distinguish the correct (or more valid) discursive elements from the incorrect (or less valid) ones (De Dreu, Evers, Beersma, Kluwer & Nauta, 2001). Early findings of Mugny and his colleagues revealed that during this decentering process (Quiamzade, Mugny & Darnon, 2009) not only do individuals consider their own point of view, but they also put themselves in the shoes of the disagreeing other and are likely to progress (Mugny, De Paolis & Carugati, 1984; Mugny, Doise & Perret-Clermont, 1975–1976; Mugny, Giroud & Doise, 1978–1979). This is why such conflict regulation has been designated as epistemic regulation (Quiamzade et al., 2013). It is also known as task-related conflict (in Organizational Psychology, Jehn, 1995) or co-constructive critical argumentation (in Educational Psychology, Asterhan & Schwarz, 2009).

Second, when the climate is threatening, the social component of the conflict overcomes the cognitive component. In such a case, individuals display a focus on their relative standing compared with the interactant, and engage in what has been termed as relational conflict regulation (Quiamzade et al., 2013). Recent research showed that relational conflict regulation entails two possibilities: (1) When individuals perceive that they are sufficiently competent to cope with the threat elicited by the disagreement, conflict regulation behavior can be described as more focused on the self than on the other (favoring the validity of one’s position); (2) When individuals perceive that they are insufficiently competent, conflict regulation behavior can be described as more focused on the other than on the self (favoring the validity of the interactant’s position; for such a conceptualization, see Sommet, Darnon & Butera, 2015).

In the former case, one confirms his/her own point of view to the detriment of the disagreeing interactant, that is, displays a dominant, active, and appetitive behavioral pattern (Butera, Gardair, Maggi & Mugny, 1998, Study 1; Butera, Mugny, Legrenzi, & Pérez, 1996; Butera & Mugny, 2001, Study 7). This particular mode of relational conflict regulation is called competitive conflict regulation (Darnon, Doll & Butera, 2007). In other research traditions, it is also known as contending tactics (De Dreu & Van Vianen, 2001) or adversarial argumentation (Asterhan, 2013).

In the latter case, one complies with the other’s position to the detriment of one’s own, that is, one displays a submissive, passive, and aversive behavioral pattern (Carugati, De Paolis & Mugny, 1980–1981; Chaiken, 1987; Mugny et al., 1978–1979). This particular mode of relational conflict regulation is called protective regulation (Sommet et al., 2014) and is also known as conceding tactics (De Dreu & Van Vianen, 2001) or quick consensus seeking (Asterhan, 2013). Neither competitive relational regulation nor protective relational regulation leads to further elaboration or cognitive progress (Buchs, Butera, Mugny & Darnon, 2004).

Disagreements are often assumed to be a difficult experience, and some managerial techniques even aim at limiting their appearance (e.g., Nominal Group Technique, Van de Ven & Delbecq, 1974, in Amason, 1996). However, as we have just seen, conflict is multifaceted and—depending on the way it is regulated—it might not always be perceived as undesirable or useless. In order to predict when and how a particular form of conflict regulation (i.e., epistemic, relational-competitive, and relational-protective) may be attributed some social value, we apply to this matter an important distinction introduced by the literature on social judgment.

A Bi-dimensional Structure of Social Value

Beauvois (2003; Beauvois & Dubois, 2009) showed that the social value of persons or objects is organized along two orthogonal dimensions: social desirability and social utility (for another theoretical framework making a similar distinction between warmth and competence, see Abele, Cuddy, Judd & Yzerbyt, 2008). On the one hand, perception of behaviors varies along an unlikability-to-likability vertical axis. This axis corresponds to the perceived ability of the social target to gain social approval, to adjust oneself to the motivation of the group to which s/he belongs. It pertains to affective traits like cold/warm, unreliable/good-natured, or antisocial/prosocial. On the other hand, perception of behaviors varies along an incompetence-to-competence horizontal axis. This axis corresponds to the perceived ability of the social target to reach social success, to adjust oneself to the social functioning of his/her environment. It pertains to adaptive traits like skilled/unskilled, ambitious/passive, or assertive/indecisive.

Social desirability is akin to concepts like expressiveness (how a person assumes an other-oriented role; Parsons & Bales, 1956), other-profitability (how advantageous the attribute of a person is for others; Peeters, 1992), or communality (how a person “gets along”; Wiggins, 2003). Socially desirable behaviors are typically those described as focused on the others, that is, as taking into account the perspective of others and involving benevolent interaction styles (for a similar proposition, see Abele & Wojciszke, 2007). Examples of socially desirable behaviors are organizational citizenship behavior at school (showing respect for peers and teachers; Jouffre, Esnard & Taillandier-Schmitt, 2012), consideration-oriented leadership in organization (caring about subordinates’ well-being; Dubois, 2010), or even modesty (Matteucci, 2014).

Social utility is akin to concepts like instrumentality (how a person takes a leadership role), self-profitability (how advantageous the attribute of a person is for him/her), or agency (how a person “gets ahead”). Socially useful behaviors are typically those described as focused on the self, that is, as sustaining the effective pursuit of one’s goals and involving the interest of the self. For instance, individualism (notably self-sufficiency, Dubois & Beauvois, 2005), structure-oriented leadership (being focused on goal achievement; Dubois, 2010) or comparative optimism (expecting good outcomes for the self; Milhabet, Le Barbenchon, Molina, Cambon & Steiner, 2012) are all perceived as socially useful.

Importantly, some social behaviors could be described as focused on both the others and the self. For instance, individuals pursuing cooperative goals (i.e., mastery-approach goals) show more positive attitudes towards helping others (Poortvliet & Darnon, 2014) and higher performance attainment (Van Yperen, Blaga & Postmes, 2014); accordingly, it is not surprising that they are judged as both socially desirable and useful (Darnon, Dompnier, Delmas, Pulfrey & Butera, 2009).

Hypotheses

In this study, we investigate the possibility that conflict is not uniformly viewed as negative by laypeople. In the first section of this introduction, we have seen that individuals facing a disagreeing interactant could display regulation behaviors described as focused on both the self and the other (epistemic conflict regulation), more on the self (the competitive mode of relational conflict regulation) or more on the other (the protective mode of relational conflict regulation). In the second section, we have seen that behaviors focused on the others are typically high in social desirability, whereas behaviors focused on the self are typically high in social utility. Hence, we formulated two hypotheses: (1) Epistemic and protective regulations should be perceived as higher in social desirability than competitive regulation; (2) Epistemic and competitive regulations should be perceived as higher in social utility than protective regulation. Questionnaires, raw data, and (Stata and SPSS) syntax files are available through FigShare (https://figshare.com/s/5ab9e7425297aec0b37c; DOI: https://doi.org/10.6084/m9.figshare.4768126).

Method

Participants

A convenience sample of 120 first-year psychology students from a French-speaking Swiss university filled out a questionnaire at the beginning of a lecture. All the students present that day participated in the study. One participant was removed due to missing data. The final sample was comprised of 101 women and 18 men, with a mean age of M = 21.76 (SD = 5.01).

Procedure and variables

First, participants were provided with the description of a fictitious previous study in which dyads of students discussed the etiology of bipolar disorder. They were told that one of the two students thought the cause of the mental trouble to be nature-based (i.e., genetic determinants), whereas the other thought it to be nurture-based (social determinants).1

Participants were then provided with three sheets, each containing a nine-item socio-cognitive conflict questionnaire allegedly completed by students from the previous study using a 1 = not at all to 7 = completely scale (Darnon, Muller, Schrager, Pannuzzo & Butera, 2006; Sommet et al., 2014). Specifically, participants were told that they would “evaluate the responses from three different [students]” collected following “situations of disagreement.” In the first bogus questionnaire, the alleged student from the previous study had answered 6 or 7 on the three items measuring epistemic regulation (e.g., “Did you try to examine the condition under which each point of view could help you understand?”). In the second questionnaire, another student had answered 6 or 7 on the three items measuring competitive regulation (e.g., “Did you try to resist by maintaining your initial position?”). In the third questionnaire, one last student had answered 6 or 7 on the three items measuring protective regulation (e.g., “Did you comply with your partner’s proposition?”). On each sheet, the score on the other six items of the questionnaire was 1 or 2 (for the example of a bogus questionnaire, see Appendix). The order of the sheets was counterbalanced, to account for order effects in the within-participant design.

For each questionnaire, participants were asked to evaluate the student “according to his/her responses” on seven traits using a 1 = not at all to 7 = completely scale. The measure was adapted from Darnon et al. (2009): Three of the traits pertained to social desirability (“pleasant,” “likable,” and “nice”) and four others to social utility (“likely to succeed,” “competent,” “gifted,” and “smart”). Descriptive statistics, reliability coefficients, and a correlation matrix are reported in Table 1.

Table 1

Descriptive variables, coefficients of reliability and inter-variable correlations for the social value of conflict regulation.

Conflict regulation Descriptive statistics
Correlation matrix
α M SD 1 2 3 4 5 6

Social desirability Epistemic (1) 0.90 4.53 1.18 1.00
Competitive (2) 0.92 2.95 1.22 –0.03 1.00
Protective (3) 0.90 4.32 1.14 0.21* 0.07 1.00
Social utility Epistemic (4) 0.90 4.31 1.10 0.53*** 0.19* 0.18 1.00
Competitive (5) 0.89 3.90 1.09 0.11 0.54*** 0.28** 0.11 1.00
Protective (6) 0.90 3.42 1.11 0.01 0.22* 0.48*** 0.14 0.09 1.00

Note. ***p < 0.001, **p < 0.01, *p < 0.05, p < 0.1.

Results

Social desirability

As a specific pattern of results was expected, a contrast analysis approach was used (Rosnow & Rosenthal, 1991). The planned contrast aimed at testing Hypothesis 1, namely that epistemic and protective regulations should be perceived as higher in social desirability than competitive regulation (respective weights: +1/3, +1/3, –2/3). The orthogonal contrast aimed at ensuring that the scores for epistemic and protective regulations were not different (weights: –1/2, +1/2, 0). A complete analysis of covariance was conducted in a preliminary stage, in order to test the effects of sex, age, and response sheet order. Age was found to interact with our planned contrast, B = –0.06, 95% CI [–0.12, –0.01],2F(1, 115) = 5.02, p = 0.027, η2p = 0.04. However, including this term did not change the effect and was not theoretically relevant. Thus, none of the covariates were kept in the analysis.

The planned contrast revealed a difference congruent with Hypothesis 1, B = 1.48 [1.20, 1.75], F(1, 118) = 114.70, p < 0.001, η2p = 0.49. As it can be seen in Figure 1 (white bars), the social desirability of epistemic regulation, M = 4.53 [4.31, 4.74], and protective regulation, M = 4.32 [4.11, 4.52], was nearly one and a half point (out of seven) higher than the social desirability of competitive regulation, M = 2.95 [2.72, 3.17]. The orthogonal contrast was not different from zero, B = –0.21 [–0.48, 0.05], F(1, 118) = 2.53, p = 0.115, η2p = 0.02, suggesting that the predicted pattern best accounted for the observed data.

Figure 1 

Social desirability and social utility of epistemic regulation (critical treatment of one’s and other’s views), relational competitive regulation (confirmation of one’s views to the detriment of the other’s views), and relational protective regulation (complying with other’s views to the detriment of one’s own views). Note. Error bars represent 95% confidence intervals (CI).

Social utility

This time, the planned contrast aimed at testing Hypothesis 2, namely that epistemic and competitive regulations should be perceived as higher in social utility than protective regulation (respective weights: +1/3, +1/3, –2/3). The orthogonal contrast compared the scores for epistemic and competitive regulations (weights: –1/2, +1/2, 0). Preliminary analysis did not show any effect of sex, age, or response sheet order. Thus, none of the covariates were kept in the analysis.

The planned contrast revealed a difference congruent with Hypothesis 2, B = 0.68 [0.45, 0.91], F(1, 118) = 34.34, p < 0.001, η2p = 0.23. As it can be seen in Figure 1 (grey bars), the social utility of epistemic regulation, M = 4.31 [4.11, 4.51], and competitive regulation, M = 3.90 [3.70, 4.09], was between half a point and one point higher than the social utility of protective regulation, M = 3.42 [3.22, 3.62]. However, the orthogonal contrast was found to be significant, B = –0.41 [–0.68, –0.15], F(1, 118) = 9.55, p = 0.003, η2p = 0.07, albeit having a descriptively smaller effect size. This indicated that the social utility of epistemic regulation was unexpectedly slightly superior to that of competitive regulation.

Discussion

In the present study, the perceived social value of conflict was found to differ as a function of its mode of regulation. In line with Hypothesis 1, regulating a conflict in an epistemic (combining one and another’s views) or protective way (conforming to another’s views) was judged as more socially desirable than regulating conflict in a competitive way (confirming one’s views). This is consistent with the fact that social desirability is linked with other-profitability, that is, communal attributes which are directly profitable to others (e.g., altruism, Esnard & Jouffre, 2008). As noted in the Introduction, epistemic and relational protective regulation behaviors may be described as including a focus on the other, whereas relational competitive regulation may not, and indeed the present results show that they are perceived by laypeople as more socially desirable.

In line with Hypothesis 2, regulating a conflict in an epistemic or competitive way was judged as more socially useful than regulating conflict in a protective way. This is consistent with the fact that social utility is linked with self-profitability, that is, agentic attributes which are profitable for the self (e.g., self-sufficiency, Dubois & Beauvois, 2005). As noted in the Introduction, epistemic and relational competitive regulation behaviors may be described as including a focus on the self, whereas relational protective regulation may not, and indeed the present results show that they are perceived by lay people as more socially useful.

Given that epistemic conflict regulation is perceived as both socially desirable and useful, one may wonder why conflict is viewed negatively by educators and seldom implemented. This may be explained by two reasons. First, the general public—and even professionals such as teachers, social workers and managers—view conflict as being primarily a matter of diverging interests rather than a confrontation of different ideas. Conflicts between individuals are often interpreted as relationship rather than task conflicts, as involving an emotional rather than intellectual component, and in which disputants are seeking to win rather than a compromise (Pinkley, 1990). Second, individuals often experience conflicts as anxiety-provoking, stressful, and uncomfortable (Narayanan, Menon & Spector, 1999). Conflict-avoidance is a common behavioral response to conflict (e.g., among teachers, Morris-Rothschild & Brassard, 2006). Thus, emphasizing the possibility for conflicts to be regulated in an epistemic way may change supervisors, perceived social value of conflict and encourage them to try to use conflicts as a tool for learning or problem solving (e.g., using a constructive controversy procedure, Tichy, Johnson, Johnson & Roseth, 2010; using conflict-related interventions, Afzalur Rahim, 2002).

Three limitations should be acknowledged. First, the present research is a single-study report. Regardless of the large effect sizes (η2p s > .14; Richardson, 2011), the study should be replicated before definite conclusions can be drawn. In a further experiment, one could use the so-called self-presentation paradigm to confirm the social desirability of conflict regulation (e.g., Guignard, Bertoldo, Goula & Apostolidis, 2015, Study 1). When instructed to present oneself in a positive way, participants should report more epistemic and protective regulations and less competitive regulation than when asked to make a negative impression.

Second, the orthogonal contrast for social utility was small but significant, indicating that epistemic regulation was unexpectedly judged as being more socially useful than competitive regulation. As focus on the self is a continuous rather than a dichotomous variable, epistemic regulation may be perceived as more self-profitable than competitive regulation, accounting for this observed residual difference.

Third, we did not assess the extent to which conflict regulation behaviors are perceived by participants as focused on the self and/or the other. We have pointed out that the set of relations between conflict regulations and focus on the self and/or the other laid out in the introductory section are based on existing definitions and results (Quiamzade et al., 2013; 2014): Epistemic and protective regulation behaviors are, in essence, oriented toward the other (showing a behavioral tendency to take into account the other’s perspective), whereas epistemic and competitive regulation behaviors are, in essence, oriented toward the self (showing a behavioral tendency to take into account one’s own perspective). In future research, it would be interesting to test whether the relationship between epistemic or protective regulation behaviors and social desirability is mediated by the perception of the other-based nature of such behaviors; and whether the relationship between epistemic or competitive regulation behaviors and social utility is mediated by the perception of the self-based nature of such behaviors.

Despite these limitations, the present research show that—contrary to what is generally assumed—lay perceptions of disagreement are not always negative. In particular, when regulated in an epistemic way, conflict is valued as both desirable and useful. This is an important finding because, with such a regulation, socio-cognitive conflict is not a zero-sum game in which there would be a winner and a looser: both interactants can learn and progress (Ames & Murray, 1982). Our findings show that, if procedures aiming at structuring conflict positively are used, disagreements might be socially accepted and useful.

Additional File

The additional file for this article can be found as follows:

Appendix

Example of scanned questionnaire. DOI: https://doi.org/10.5334/irsp.88.s1