Online communication takes the form of dialogue while producing almost none of it. What we call conversation — the posts, replies, threads, and reactions — is largely reactive performance. Understanding why requires looking not at the people inside these systems, but at the systems themselves.
What You’ll Learn
- What distinguishes the illusion of discourse from real dialogue
- How platform design — not human nature — drives reactive behavior
- Why outrage travels where nuance cannot, and what that costs
- The difference between presence and the performance of presence
- What a systems-level diagnosis means for how we engage
What Is the Illusion of Discourse?
The illusion of discourse is the condition in which communication mimics the structure of dialogue — a post, a reply, a thread of exchanges — while lacking its substance: mutual attention, genuine context, and the willingness to be changed by what another person says. Most online communication, as of 2025, operates under this condition across all major platforms.
The form is familiar enough to feel like conversation. Someone posts a thought. Someone replies. Others join in. But form and function are not the same thing. Real dialogue requires shared attention and a mutual investment in understanding. What platform design rewards is speed, emotional charge, and visibility. Those are the variables the algorithm measures. Understanding does not register.
As a general rule, what platforms optimize for is not what their users think they are optimizing for. Most people believe they are participating in conversation. The platform records them as a unit of engagement.
Why Do Online Arguments Escalate Instead of Resolve?
Online arguments escalate instead of resolve because the systems hosting them reward escalation. Content that generates outrage, shock, or righteous agreement produces more interactions per unit of time than thoughtful exchanges do. Platforms amplify what keeps users engaged — and conflict, as of early 2026, remains more efficient at that than understanding.
Research published in PNAS Nexus in 2025 (Milli et al.) found that engagement-optimized systems amplify divisive content. This is not a side effect. It is the designed output of systems that measure success in interactions, not in the quality of those interactions. A viral argument produces the same metric as a viral insight. The platform cannot tell the difference, and has no incentive to try.
The consequences compound. Online arguments rarely shift anyone’s position because shifting a position takes sustained attention over time. Platforms are designed to interrupt that attention — to pull the next piece of content into view before the current exchange has run its course. The argument appears to continue, but the participants have often moved on before it ends.
The most reliable indicator that a platform prioritizes engagement over dialogue: its escalation mechanics. Notifications that surface conflict, algorithms that surface controversial replies first, metrics that treat outrage and agreement identically.
How Does Platform Design Shape Communication Behavior?
Platform design shapes communication behavior by determining which behaviors are rewarded and which are penalized — often invisibly. Social media systems privilege brevity, emotional charge, and rapid response because those inputs produce the engagement metrics that drive revenue. The result is an environment in which complexity is structurally disadvantaged, not because it lacks value, but because the system has no way to measure it.
Research by Gligoric et al. (2019) found that shorter content consistently outperforms longer content in reach and engagement on social platforms. This creates a feedback loop: users who write at length reach fewer people, users who compress their thinking to fit the medium reach more. Over time, the medium teaches its audience what kind of thinking it rewards — and depth is not it.
This is not accidental, and it is not a neutral outcome. Frances Haugen’s 2021 congressional testimony documented how Meta’s own internal research showed the platform’s engagement systems amplified content that provoked strong emotional responses, including anger. The platform understood what it was doing. It continued, because the alternative would have cost engagement.
If you want to understand why online communication behaves the way it does, follow the incentive. The incentive is engagement. Everything else follows from that.
What Is the Difference Between Online Presence and Real Presence?
Online presence is the feeling of being in contact — notifications in real time, replies within seconds, a constant sense of activity. Real presence requires something different: shared attention, a mutual willingness to slow down, and the space for another person’s meaning to actually land before responding. Online platforms are designed for the former and structurally incompatible with the latter.
Asynchronous communication creates the sensation of immediacy without its conditions. A comment left in good faith may be read hours later in a different context, by someone who has already seen a dozen other posts since yours. Sarcasm gets misread. Empathy gets flattened. The absence of shared space and pacing removes the anchors that keep meaning stable in face-to-face conversation.
Research by Riordan and Trichtinger (2017) found that people significantly overestimate their ability to accurately interpret the emotional tone of text-based exchanges. We think we know what someone means. We are frequently wrong. And the platform moves fast enough that we rarely find out.
The most common failure mode here is optimizing posts for visibility instead of clarity. When the goal shifts from being understood to being seen, communication becomes signal-performance — the appearance of meaning without its function.
Why Does Outrage Travel Where Nuance Cannot?
Outrage travels because it is fast and binary: it triggers an immediate response and demands a position. Nuance is slower and more conditional — it requires the reader to hold multiple things true at once, which takes more time than a scroll-optimized environment allows. The asymmetry is not incidental. It is built into the infrastructure.
Emotionally charged content does not just perform better because of human psychology, though human psychology plays a role. It performs better because recommendation systems are designed to surface content that generates interactions, and outrage generates more interactions per post than careful analysis does. Research in Scientific Reports (Traberg et al., 2024) found that social cues — what others are visibly engaging with — amplify susceptibility to emotionally charged content, including misinformation. The system shapes what people see, and what people see shapes what they take to be true.
The deeper effect is what Zhao and colleagues described in a 2025 study as the “spiral of silence”: the phenomenon in which people who hold minority views, or views that feel socially risky to express, increasingly self-censor. Visibility determines perceived normalcy. What appears most frequently is taken to represent what is most common. Nuanced positions, underrepresented in algorithmic surfaces, become further underrepresented as people recognize the signal and adjust their public expression accordingly.
As a general rule, what goes unsaid online shapes public discourse as powerfully as what gets said. Silence is not neutral. It is itself a systemic output.
Is Online Dysfunction a User Problem or a Systems Problem?
Online dysfunction is a systems problem. The behaviors that make online discourse dysfunctional — reactive posting, escalation, shallow engagement, self-performance — are not moral failings. They are rational responses to the incentive structures of the environments in which they occur. The people inside these systems are mostly doing what the systems reward.
McLuhan’s argument — that the medium is the message — is rarely more legible than in the context of social media. The form of the platform does not merely carry content. It shapes what kind of content can be expressed, how it will be received, and what kind of communicator the user becomes through repeated exposure. You cannot participate extensively in engagement-optimized systems without being shaped by them.
This distinction matters because it changes what a useful response looks like. If dysfunction is a user problem, the answer is individual discipline: log off, be more thoughtful, resist. If dysfunction is a systems problem, the answer requires examining and redesigning the systems. Individual behavior change within a system is real but limited. Structural change has greater leverage.
The most reliable approach is to name the system before diagnosing the behavior. Behavior that looks irrational often makes perfect sense once you understand the incentive structure that produced it.
Conclusion
The illusion of discourse persists because the systems that host our communication are not built for understanding. They are built for engagement, and engagement — measured the way platforms measure it — has almost nothing to do with meaning.
Systems can be changed. What is designed can be redesigned. That is both a technical statement and a philosophical one: the environments we communicate in are not natural phenomena. They were built, and they can be built differently.
This essay opens a series that uses systems thinking to examine the hidden mechanics of digital communication. Rather than explaining online behavior as the product of individual failure, the series examines the structures that make those behaviors coherent, predictable, and — under different conditions — changeable. The next essay examines platform architecture: how the design of digital environments shapes the behavior of everyone inside them, and what it would mean to design for dialogue instead of reaction.
