Overuse of Moral Language Dampens Content Engagement on Social Media

Abstract

Online social-media platforms are a vital arena in which socio-political perspectives are put forth, debated, and spread. Prior work suggests that certain moral-emotional language drives online contagion, but theoretical and empirical findings remain debated, inhibiting constructive interventions. We substantially advance this ongoing debate using a diverse range of topics and both mainstream and extremist social media platforms. We find two countervailing dynamics predict the rise and fall of posting engagement online. First, we confirm that content infused with a greater number of moral words is associated with increased engagement; still, contrasting with prior work, we find no evidence that it is driven specifically by moral-emotional words as opposed to the more general and larger lexicon of moral language. Second, we identify a striking reversal in the relationship between moral language and engagement, a phenomenon we call “moral penalty.” We find that as the ratio of moral to non-moral words surpasses a threshold, the process of engagement around a post reverses with marked decreases in engagement and online diffusion. These findings help clarify links between morality and online engagement and contagion: infusing messages with moral language increases their spread to a point after which embedding posts in morality may backfire.

Contextualized Construct Representation: Leveraging Psychometric Scales to Advance Theory-Driven Text Analysis

“…The results of this study showed that applying CCR to textual data from social media returns substantially more accurate text-based measures of moral psychological constructs. These results are particularly important since much of prior work in moral text analysis has used dictionary-based methods (e.g., Brady et al, 2017;Buttrick et al, 2020) and DDR (e.g., Candia et al, 2022;Wang & Inbar, 2021). The power of contextual language models in two domains can explain the power of CCR in outperforming its predecessors: (a) quantifying the nuances of moral constructs that are captured in psychometric scales, but absent in dictionaries; and (b) quantifying the contextual information present in social-media language such as Facebook updates.…”

“…In the last decade, a large number of social scientists have turned to social-media data (e.g., Twitter, Facebook) to examine psychological theories in an ecologically valid environment. Since many of these efforts are theory-driven, they have primarily relied on top-down methods such as dictionaries (e.g., Brady et al, 2017;Burton et al, 2021;Simchon et al, 2020), DDR (e.g., Candia et al, 2022;Wang & Inbar, 2021), CONTEXTUALIZED CONSTRUCT REPRESENTATION 30 and human annotations (e.g., Hoover et al, 2020;Mooijman et al, 2018). Since we show that CCR can outperform word-counting and DDR methods, future theory-driven work can rely on CCR as a more robust technique that captures the complexities of social-media language.…”

The (moral) language of hate

“…Indeed, findings from Study 3 provide support for this interpretation. Individuals posting prejudicial language on social media have obvious communicative goals, and thus might imbue their language with moral rhetoric and framing for higher engagement and reach ( 40 ); however, the expression of hatred might implicitly draw on the moral concerns of the speaker beyond the post’s immediate wording. In conclusion, morality is used in the communication of hatred, but it is also related to hatred at a more fundamental level ( 23 , 25 ).…”

The paucity of morality in everyday talk

“…In social-media platforms like Twitter, politicians and organizers of social movements often express moral concerns in an effort to increase online engagement and to influence perceived norms within social networks. In such contexts, using moral language in a post has been found to stimulate moral engagement and sharing, whereas using "too much" moral language reduces engagement and sharing, a phenomenon termed "moral penalty" 56 . In the present research, we were primarily interested in the frequency of moral language in daily communication, not in the use of moral rhetoric in public persuasion.…”