Phoenix 24
Values attract attention, but saturation may push audiences away.
Santiago
A large scale study of approximately 1.62 million social media posts suggests that moral language can increase online engagement only up to a point. Researchers from Universidad del Desarrollo, the University of Massachusetts Amherst and Northwestern University found that messages containing ideas about justice, harm, duties, virtues and what is right or wrong often receive more reactions, but excessively dense moral vocabulary can produce the opposite effect.
The study distinguished between two related concepts. Moral loading measures how strongly a message is connected to moral ideas, while moral density measures how concentrated that vocabulary is within the text. That distinction proved important because posts could be highly relevant to moral issues without necessarily repeating moral terminology throughout every sentence.
Researchers analyzed individual posts from Twitter and Reddit, as well as grouped discussions from 8chan, covering 13 different topics and communities. A computational model converted words into numerical representations and compared their semantic relationship with moral concepts. Those results were then measured against reposts, replies and other forms of interaction.
The pattern was not linear. Higher moral relevance was consistently associated with greater engagement, but messages with very concentrated moral language tended to perform worse when compared with posts expressing similar moral relevance more moderately. Engagement peaked around an intermediate density level of 0.30 on the scale developed by the researchers, although the optimal range varied by platform and discussion context.
The findings challenge the idea that adding stronger moral vocabulary automatically makes a message more persuasive or shareable. Audiences may respond positively when a publication clearly communicates values, but repeated appeals to duty, virtue, justice or outrage can eventually make the message feel saturated, predictable or overly performative.
The study is observational, meaning it identifies association rather than causation. Researchers cannot yet conclude that excessive moral language directly causes lower engagement. Future experiments will be needed to determine why the relationship appears and whether similar effects influence trust, comprehension and sustained attention.
The findings may also become increasingly relevant as artificial intelligence generates more social media content. Automated systems can easily intensify emotionally and morally charged language, but optimization based solely on rhetorical force may reduce rather than increase audience response.
The broader lesson is subtle. Effective communication does not require abandoning values. It may require expressing them with enough clarity to be understood, but enough restraint to remain credible.
Narrative is power too.