Moral convergence, which refers to sharing the same moral principles with a group, is likely to occur in online settings where communities are formed based on these principles. New research published in Social Psychological and Personality Science found that across 5 studies moral convergence is associated with increased likelihood of radicalism.
“Moral emotions such as anger, disgust, and hatred toward the outgroup are typically influential in ‘us vs. them’ contexts, creating vehement tribal bonds with one’s ingroup and extreme hostility toward out- group members who may be dehumanized and seen as deserving harm and persecution,” wrote study author Mohammad Atari and colleagues.
“Once an issue becomes moralized, people are more likely to act upon it to ensure their group prevails by nearly any means necessary, including persuasion, lying, cheating, protesting, or—in extreme cases—violence.”
Considering the popularity of online communities where people with similar ideologies can gather, such as Reddit, researchers were interested in looking at moral convergence in these communities and how it might relate to radical ideologies.
For the current studies, researchers focused on five moral domains that have reliably predicted attitude and behavioral differences between ideologically opposed groups (e.g., Republicans and Democrats). These domains consist of Loyalty, Authority, Purity – typically referred to as the “binding” moral values – and Care and Fairness – referred to as the “individualizing” moral values. Researchers conducted a series of 5 studies to assess this research question.
In the first study, researchers focused on the website Gab, as the site has attracted a large audience of far-right identifying people. Gab was also where much of the facilitation of the attempted insurrection at the U.S. Capitol in January 2021 occurred. To test for moral convergence, researchers analyzed millions of posts for hate speech and moral language from thousands of randomly selected Gab users. Results indicate that moral convergence on Loyalty, Authority, and Purity (binding values) was related to increased rate of posts containing hate-speech. Convergence on Care and Fairness (individualizing values) was not related to hate-speech posting.
In the second study, researchers sought to replicate the analysis done in Study 1 but with a different online community: Incels, a subreddit (i.e., online community) formed for people with openly misogynistic attitudes (incels = “involuntary celibates”). Unlike Study 1, results indicate that moral convergence on all five moral values was associated with increased likelihood of hate-speech posts. Thus, these results suggest this phenomenon is not unique to Gab. “The higher the convergence between a user and their community in terms of expressed moral concerns, the more likely it is for them to use radicalized, hateful language,” concluded the researchers.
In Study 3, researchers conducted an experiment to test the effects of moral convergence on radicalism in a controlled setting. They recruited a sample of 333 adult participants from Mechanical Turk, an online platform. Participants were presented with the five moral concerns and asked to pick one that was most important to them. They were then randomly assigned to either a convergence or a nonconvergence condition. In the convergence condition, participants were presented with an invite to Facebook group of like-minded individuals based on the moral concern they selected. The nonconvergence condition was the opposite (i.e., invited to a Facebook group with people that do not share your concern). Participants were then measured for activism and radical intentions. Results show that those in the convergence condition scored higher on radicalism and pro-group activism than those in the nonconvergence condition.
In Study 4, researchers sought to replicate and expand on the findings from Study 3. They recruited a final sample of 510 American adults through Qualtrics Panels, another online research platform. Importantly, this sample was generally evenly split on political affiliation. The procedure was identical to Study 3 except for the inclusion of identity fusion and sense of power scales. Results are consistent with Study 3 in that those in the convergence condition scored higher on radicalism. However, follow-up analyses suggest that convergence does not directly impact radicalism scores. Importantly, identity fusion mediates (or explains) this relationship. In other words, convergence likely led participants to feel more oneness (identity fusion) with the presented Facebook group, which then led to their increased radicalism. This effect was not observed for feelings of power.
Study 5 was conducted to address limitations of the previous two studies. “Specifically, we aim to show that (a) the effect on radicalism is specific to ‘‘moral’’ convergence, not other types of amoral homogeneity in attitudes; (b) the effect can be observed in ‘‘real’’ groups rather than hypothetical groups; and (c) the effect is not limited to radicalism intentions and can be generalized to outcomes related to extreme pro-group acts, such as the willingness to fight and die for the ingroup.”
Thus, participants were randomly assigned to be asked about five moral concerns vs. five non-moral concerns and then randomly assigned to either the convergence or nonconvergence condition. These conditions differed from the previous studies in that rather than presenting the participant with a Facebook group invitation, they were presented with statistics from a national database (i.e., information about a “real” group). In the convergence condition, they were told most Americans share their moral/non-moral concern and in the nonconvergence condition, they were told only a small number of Americans shared their concern.
Results showed a significant interaction between morality and convergence on radicalism. Specifically, radicalism intentions were stronger in the moral convergence condition compared to the non-moral nonconvergence, non-moral convergence, and moral nonconvergence conditions. These results suggest that radicalism influences are specific to moral convergence, not general attitude convergence.
“Our results highlight the importance of moral diversity in online social networks to avoid affective polarization and creation of moral echo chambers that could contribute to radicalization through formation of cult-like identities to which individuals get vehemently attached,” the researchers wrote.
The authors do caution some limitations to their work such as the exclusive use of Western participants and American definitions of political ideology.
The study, “Morally Homogeneous Networks and Radicalism“, was authored by Mohammad Atari, Aida Mostafazadeh Davani, Drew Kogon, Brendan Kennedy, Nripsuta Ani Saxena, Ian Anderson, and Morteza Dehghani.
