More and more of our everyday conversations are taking place online, ranging from WhatsApp to Teams, and from e-mail to Twitter. We share our day with our extended family, we congratulate our colleague following the birth of her first child, we plan a get together with our friends, we share holiday pictures with our “followers,” we file a complaint to a customer service chatbot, and we discuss politics with strangers. But these online interactions, especially those about difficult topics, do not always fare well. Indeed, in the context of controversy, online interactions seem to be more prone to escalate into conflict and polarize opinions than discussions held face-to-face (e.g., Anderson et al., 2018). Consequently, a lot of research and theorizing has focused on why this is the case.

There are two prominent explanations for why online discussions escalate. First, it is suggested that people behave aggressively online because they feel anonymous, or at least distant from others (Suler, 2004). They think they can say anything they want without consequences, and become disinhibited as a result. A second commonly heard explanation for online conflict is that people tend to misinterpret each other’s messages because these are unclear or ambiguous due to a lack of non-verbal cues (Daft & Lengel, 1986). But do people really stop caring about others online and are online messages really that unclear?

“My dissertation project was aimed at understanding online escalation better by having a close look at what people actually do when they discuss online.”

My dissertation project (Roos, 2022) was aimed at understanding online escalation better by having a close look at what people actually do when they discuss online. We especially wanted to know how this compares to face-to-face discussions (because people are apparently better at harmoniously navigating disagreements when chatting in person). To this end, we asked small groups of unacquainted students to discuss controversial issues, such as the refugee crisis, via a text-based chat and face-to-face. We then analyzed their behavior during the interactions and followed-up after by asking them about their experiences.

We found that people in face-to-face conversations use several diplomatic skills to show that they are aiming for a harmonious and constructive discussion (Roos et al., 2020a, 2020b). First, people express their opinion tentatively or ambiguously by using disclaimers (e.g., “I do not know for sure”), hedges (e.g., “maybe,” “sort of”), and vocalizations that express doubt (e.g., a drawn out “hmmm”) or tentativeness (e.g., “uhm”). This tentativeness communicates tact and concern for the feelings of others. Second, people in face-to-face interactions respond with head nods, “yes” or “hmm” to the person speaking. This shows their continued attention to and understanding of what is being said. Online, however, we found that people expressed themselves less ambiguously, more clearly, and less responsively than in the face-to-face discussions. That is, people behave less diplomatically online.

“People in face-to-face conversations use several diplomatic skills to show that they are aiming for a harmonious and constructive discussion.”

This does not seem to result from anti-social intentions. Rather, the relative lack of diplomacy behavior seems to be inherent to the limitations of the medium: expression in text is relatively clear, and a lack of synchronicity encourages cross-talking. Nevertheless, these behaviors have social consequences.

We found that participants felt ignored online and thought that their interaction partners were disinhibited (Roos et al., 2020a). They thought that their partners were more concerned with clearly expressing their own opinion, of which they were strongly convinced, than with listening. As a consequence, participants thought they disagreed more than they actually did, and felt less close to each other when interacting online.

“People behave less diplomatically online, but this does not seem to result from anti-social intentions.”

What I think was most important is that we found no differences in actual disinhibition and actual disagreement between the online and face-to-face discussions (Roos et al., 2020a, 2020b). Participants did not consider themselves any more (or less) disinhibited online than when interacting face-to-face, nor did we find evidence for disinhibition in their conversation behaviors. (For example, we found no instances of aggressive language use.) Moreover, participants did not express more (or less) disagreement during the online discussions and did not show more disagreement in their private opinions after the discussions. This shows that people can perceive each other as disinhibited and experience disagreement, not because others are disinhibited nor because there is disagreement, but simply because the medium limits diplomacy (see also Roos et al., 2021).

“We found no differences in actual disinhibition and actual disagreement between the online and face-to-face discussions.”

This offers new insights into the commonly held assumptions about online escalation mentioned at the start of this article: 1) misunderstandings can arise online because people communicate too clearly, rather than too ambiguously, and 2) because the communication medium changes people’s behavior they can appear disinhibited while, in fact, there is no evidence that they are.

I would like to end with a practical take home message. Next time you notice a disagreement or misunderstanding with someone in WhatsApp, try to be responsive to the other person (for example, by typing “yes, I see your point. but,”) and try to formulate your opinion as an opinion (“I think that might not be true”) rather than as an established fact (“that is not true”). But, preferably, do not try to fix things online and continue discussing potential controversies face-to-face.

Note. Illustration of Rubin’s Vase by Milou de Groot.


Anderson, A. A., Yeo, S. K., Brossard, D., Scheufele, D. A., & Xenos, M.A. (2018). Toxic talk: How online incivility can undermine perceptions of media. International Journal of Public Opinion Research, 30(1), 156–168.

Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32(5), 554–571.

Roos, C.A., Koudenburg, N., & Postmes, T. (2020a). Online social regulation: When everyday diplomatic skills of harmonious disagreement break down. Journal of Computer-Mediated Communication, 25(6), 382-401.

Roos, C.A., Koudenburg, N., & Postmes, T. (2021). Dealing with Disagreement: The depolarizing effects of everyday diplomatic skills face-to-face and online. New Media & Society, 24(9), 2153-217.

Roos, C.A., Postmes, T., & Koudenburg, N. (2020b). The micro-dynamics of social regulation: Comparing the navigation of disagreement in text-based online and face-to-face discussions. Group Processes & Intergroup Relations, 23(6), 902-917.

Roos, C.A. (2022). Everyday Diplomacy: dealing with controversy online and face-to-face. University of Groningen.

Suler, J. (2004). The online disinhibition effect. Cyberpsychology & Behavior, 7(3), 321–326.