Why It’s Getting Easier to Misunderstand Each Other
Understanding Online Conflict and Relationship Strain
What digital communication does to the brain's capacity for understanding — and why the conflict that follows so often surprises everyone involved.
Published: 10 April 2026Updated: 3 days, 15 hours ago13 min read
ByNeuro
Levels of ScaleCommunity
LensEmpathyLanguage
Wellbeing DimensionInterpersonal
System of WellbeingThriving Communities
Wellbeing StrainOnline conflict and relationship strain
Regenerative Development GoalsRDG 17 - Interwoven Stewardship
Quick summary
Most of us have had an online exchange that went wrong in ways that felt disproportionate. A message that landed harder than intended. A thread that escalated in ways nobody quite planned. A friendship that felt different after an argument that played out in text. The conflict may have felt sharp while it was happening. In retrospect, many people find themselves wondering how it got there so quickly.
Online conflict and relationship strain describes the pattern of misunderstanding, escalation, and relational damage that arises specifically in digital communication environments because the medium itself strips away most of what the brain relies on to understand another person's meaning and intent. Text arrives without tone. Messages land without facial expression. Replies are composed under conditions of arousal that face-to-face conversation would naturally moderate.
This article explores what online conflict actually is, what is happening in the brain when digital communication goes wrong, why the conditions of online communication make conflict so much easier to stumble into than to resolve, and why the relational damage that follows is both understandable and, in important respects, not inevitable.
The argument that nobody quite meant to have — and the relationship that felt different afterwards
It tends to start small. A message typed in haste. A reply that reads as dismissive when the sender meant it as brisk. A comment on a post that someone takes as criticism when it was meant as observation. And then, before anything has had a chance to be clarified, a thread is running — faster and sharper than it would have been if the same exchange had happened in a room, face to face, where tone and expression and the simple physical presence of another person would have slowed things down and kept them human.
Many people recognise this pattern across different areas of their digital lives. In friendships, where a group chat develops a strange undercurrent after something ambiguous is said and not resolved. In workplaces, where an email exchange becomes tense in ways that require a face-to-face conversation to undo. In families where a message meant with affection is received as interference. In public online spaces, where a comment offered in good faith attracts responses that feel like an attack.
What makes online conflict worth understanding carefully is that it happens so much more easily, and that the conditions which allow it to escalate are built into the medium itself rather than being the product of any individual's bad intentions. Understanding those conditions can change how we move through digital communication — with a little more caution about what we assume we've understood, and a little more room for the possibility that what we received was not what was sent.
Text strips away most of what the brain uses to understand another person — and fills the gap with assumption
Human communication, in its evolved form, is an extraordinarily rich process. When we speak to someone face to face, we are simultaneously processing their words, their tone, their facial expression, their posture, their eye contact, and dozens of other signals that together allow us to form a reasonably accurate understanding of what they mean — what they are saying, how they feel about it, whether they are certain or uncertain, whether their words are literal or ironic. Most of this happens automatically, below the level of conscious awareness, and at great speed.
Digital text communication removes almost all of this. What remains is the words themselves — stripped of tone, expression, timing, and the embodied presence of another person. The brain, confronted with this poor signal, does what it always does when information is incomplete: it fills the gaps. It draws on its existing sense of the sender, its current emotional state, its habitual interpretive patterns — and generates a meaning. That generated meaning feels immediate and certain. It rarely announces itself as a guess (2).
People significantly overestimate how well their intended meaning is conveyed through email — believing their tone is clear to the recipient when it frequently is not (2). In some experiments where participants read email messages, they misidentified the intended emotional tone roughly half the time, even when the senders were close acquaintances (2). The fundamental problem is something called egocentrism: senders know what they meant, and that knowledge makes it difficult to accurately imagine how their message will land on someone who doesn't share that knowledge. The words feel clear from the inside. They arrive as something quite different on the outside (2).
There is a related phenomenon — the online disinhibition effect — that describes how the specific features of digital communication change how people behave toward each other (1). Anonymity reduces the felt accountability of what we say. Invisibility removes the physical presence of the other person that would naturally moderate our responses. Asynchronicity — the ability to send a message without witnessing its immediate impact — removes the real-time feedback that face-to-face conversation provides. The result is a communication environment in which people say things they would not say in person, in tones they would not use in person, to people whose full humanity can more easily be temporarily set aside (1).
When we read a message that could be interpreted as criticism, dismissal, or aggression, the nervous system's alarm response activates — sometimes before any conscious evaluation has taken place. Under conditions of emotional stimulation, the brain's interpretive tendencies shift: ambiguous information is more likely to be read as threatening, intentions are more likely to be attributed as hostile, and the impulse to respond defensively or aggressively becomes harder to override (11). The reply composed in that activated state tends to be sharper, less considered, and less accurate about the other person's actual intention — producing a response that itself reads as threatening, which triggers the same response in the recipient. Escalation, in online conflict, is rarely intentional. It is often the predictable output of two nervous systems each responding to what they've received as a threat (1, 11).
The same people behave very differentlyoffline andonline — and understanding why changes how we read the conflict
One of the most disorienting things about online conflict is the gap between who people are in person and who they sometimes appear to be in digital exchanges. Most people who have experienced the disinhibition effect — who have sent something in a message they would never have said face to face, or who have been on the receiving end of hostility from someone who is perfectly pleasant in person — are aware that something shifts in the translation from physical to digital space.
Some research on moral disengagement offers one frame. It describes how ordinary people come to behave in ways that contradict their own values through a set of psychological mechanisms — dehumanising language, diffusion of responsibility across a group, displacement of blame onto the other person, minimisation of consequences — that make harmful behaviour feel permissible in the moment (12). Digital environments facilitate several of these mechanisms simultaneously: the other person is reduced to a profile image and a string of text; the sense of social accountability that physical co-presence provides is absent; the responses of a crowd can feel like diffused rather than individual responsibility (12).
There is also the concept of relational aggression — harm that operates through exclusion, rumour, social manipulation, and undermining rather than through direct confrontation (4). Online environments are particularly well-suited to relational aggression because the tools for it (group chats, public posts, forwarding messages out of context, coordinated exclusion) are more readily available and less visible in their operation than in-person relational harm (4).
Difficult conversations usually contain multiple things happening at once — one about what happened and who is right, one about the emotional layer beneath the content, and one about what the exchange implies about who we are (5). Online conflict tends to compress all three into a single exchange without distinguishing them — producing a tangle in which people are arguing about facts while at the same time defending their emotional responses and protecting their sense of self, in a medium that provides none of the context that would allow any of those conversations to be heard clearly (5).
The single most reliable predictor of relational damage is the presence of contempt — the sense of being fundamentally superior to and dismissive of the other person (11). Contempt is easy to generate in text, easy to receive, and extremely difficult to recover from. The pattern of contempt, defensiveness, criticism, and withdrawal that predicts deteriorating relationships plays out in online exchanges with a speed that face-to-face interaction would rarely permit (11).
Digital environments are designed for reaction, rather than understanding — and that shapes the conflicts that happen within them
Online conflict arises in digital environments that have been designed — in ways both intentional and structural — to generate emotional engagement, rapid response, and the kind of shareable, attention-capturing content that conflict reliably produces.
Outrage and moral resentment are among the most powerful drivers of online engagement. Content that provokes these responses generates more reactions, more shares, more replies — and therefore more time on platform and more advertising revenue — than content that is nuanced, measured, or conciliatory (8). Platforms that optimise for engagement therefore have structural incentives to surface and amplify conflictual content, regardless of whether that amplification serves the wellbeing of the people involved. The result is an information environment in which the emotional temperature of online discourse is systematically raised by the design of the environments individuals inhabit (8).
Analysis of how people communicate online documents how the text-based, public, back-and-forth nature of many exchanges creates a default atmosphere of conflict — where disagreement is expected (6). Even messages sent without hostile intent are more likely to be received as hostile in this environment — because the interpretive conditions in which they land have been primed for conflict rather than connection (6).
Significant proportions of different community members report having been subjected to some form of online harassment, with rates significantly higher for women, people of colour, and LGBTQ+ individuals (9). There is also a pattern of systematic, coordinated harassment that goes well beyond individual conflict — with gendered abuse serving as a tool for silencing public participation and enforcing social hierarchies online (10). The spectrum of online conflict, in other words, runs from the everyday misunderstanding in a group chat to organised campaigns of abuse — and the mechanisms that produce the former also, at scale, enable the latter (13).
There is also the question of why arguments rarely resolve online. Confirmation bias — the tendency to seek out and favour information that confirms what we already believe — appears across all populations and is particularly sound for beliefs tied to who we feel we are (7). In an environment designed to show and amplify content that generates strong emotion, this tendency is systematically used rather than challenged. We deploy arguments not to discover what is true but to defend what we already believe — and we are disproportionately persuaded by information that confirms existing positions (7).
The gap between what we sent and what they received — and who bears the greatest cost of getting it wrong
There is a particular quality to the aftermath of online conflict. The sense of having been genuinely wronged by what someone said. The equally genuine certainty, on the other person's side, that they didn't mean it that way. Both people in the same exchange, holding incompatible but internally coherent accounts of what happened. Both of them right about their own experience. Both of them significantly wrong about the other person's.
When an online exchange touches something about how we see ourselves — our intelligence, our values, our belonging in a community — the stakes of the conflict expand well beyond the original subject. Defensiveness increases. The capacity for curiosity about the other person's perspective narrows. The conversation that might have been a genuine exchange of views becomes, instead, a protection of self (1, 5). And once the exchange has been frozen in text — re-readable, re-interpretable, potentially shareable — each return to it may add another layer of misinterpretation to an already distorted record.
There is also the question of who bears the greatest cost of online conflict. It is not evenly distributed. Targeted harassment, gendered abuse, and coordinated campaigns fall disproportionately on those who are already socially marginalised — and produce real psychological harm, including anxiety, hypervigilance, and withdrawal from the public online spaces where civic participation increasingly takes place (9, 10). Framing all of this as 'online conflict' — as though it were symmetrical between parties — risks obscuring the asymmetry of power and harm that characterises many of these exchanges.
The same platforms that generate conflict and enable abuse also deliver things that are genuinely valuable: connection, community, learning, the ability to find others who share your experience in a world where that might otherwise be impossible. Many people don't want to leave them. And for some — particularly those who are geographically isolated, socially marginalised, or whose communities of belonging exist primarily online — leaving isn't a meaningful option. The question of how to inhabit these environments without either abandoning them or absorbing their most damaging dynamics is one that individuals are currently left to navigate largely on their own.
Online conflict arises from structural conditions — and understanding those conditions is what changes how we move through it
The science of online miscommunication points toward something that feels simple but matters: the conditions that produce conflict are, in large part, the conditions of the medium itself. Text strips away tone, expression, and timing. The brain fills the gap with assumption. The assumptions are often wrong. And the activated state in which replies are composed tends to make the gap wider. Understanding this changes the interpretive frame that conflict arrives in. What looked like hostility may have been brevity. What landed as dismissal may have been haste (2).
A consistent finding across communication research is that people overestimate how clearly their meaning carries through text — and that this overconfidence is structural rather than careless (2). The sender knows what they meant. That knowledge makes it genuinely difficult to imagine how the words land on someone who doesn’t share it. Knowing this does make the assumption of hostile intent a little harder to reach automatically.
What some research on relationships points to — and what online environments tend to work against — is the importance of the overall relational climate in which conflict occurs (11). Conflict within relationships characterised by warmth, curiosity, and genuine regard tends to land differently from conflict in environments where those qualities are absent. Digital spaces that structurally amplify the negative and make the positive less visible are, in this sense, working against the conditions that allow disagreement to remain navigable.
What we owe each other in the spaces we share online
Online conflict and relationship strain is a story about human beings — equipped with brains built for rich, multi-signal, face-to-face communication — navigating environments that have stripped away most of what those brains rely on to understand each other. The misunderstanding that results is, in large part, a predictable consequence of the gap between what human communication evolved for and the medium through which so much of it now takes place.
The relational harm that online conflict produces is real, all the same. Friendships that cool after a misread message. Working relationships that carry a new wariness after a tense email exchange. Communities that fracture along lines that a different medium might never have drawn. The cost is carried in the quality of the relationships themselves — and in the gradual erosion of the trust and openness that genuine connection requires.
What the research points toward, at its deepest level, is something simple and difficult at the same time: that genuine understanding between people requires the conditions that genuine understanding has always required. Presence. Patience. The willingness to stay curious about what the other person actually meant before concluding that we know. The recognition that what we received may not be what was sent — and that the gap between the two is rarely evidence of malice.
The wider question — about what kinds of digital environments would support rather than undermine the relational commons of shared human life — belongs to designers, platform builders, regulators, and the cultures that use these tools. What we collectively inhabit online is a social environment — one that shapes the quality of human relationship at scale, and one that has so far been built with too little attention to what healthy human relationship actually requires. That is a design challenge, a governance challenge, and a cultural challenge. It is also, in a quiet but important sense, a question about what we owe each other in the spaces we share (13).
Schober MF, Roque A, Bloom JE. Lost in virality: how social media can amplify hidden misinterpretations. Technology,Mind, andBehavior. 2025 Dec;6(4). https://doi.org/10.1037/tmb0000174
Twenge JM, Martin GN, Campbell WK. Decreases in psychological well-being among American adolescents after 2012 and links to screen time during the rise of smartphone technology. Emotion. 2018;18(6):765–780. https://doi.org/10.1037/emo0000403
Vaillancourt T, Faris R, Mishna F. Cyberbullying in children and youth: implications for health and clinical practice. Canadian Journal of Psychiatry. 2016 Dec 16;62(6):368–373. https://doi.org/10.1177/0706743716684791
Stone D, Patton B, Heen S. Difficult conversations: how to discuss what matters most. New York: Penguin; 2000.
Gottman JM, Silver N. The seven principles for making marriage work. New York: Harmony; 1999
Nocera TR, Dahlen ER, Poor A, et al. Moral disengagement mechanisms predict cyber aggression among emerging adults. Cyberpsychology. 2022;16(1):1–18. https://doi.org/10.5817/cp2022-1-6
Harriman N, Shortland N, Su M, Cote T, Testa MA, Savoia E. Youth Exposure to Hate in the Online Space: An Exploratory analysis. International Journal of Environmental Research and Public Health. 2020 Nov 17;17(22):8531. https://pubmed.ncbi.nlm.nih.gov/33212999/
