Joint Attention

In the AI Story

The material conditions of joint attention are specific and historically variable. Pre-modern societies sustained joint attention through physical co-presence: the town square, the church service, the village gathering where all members attended to the same event simultaneously. Print media extended joint attention across space: newspapers created shared informational objects that geographically dispersed readers could discuss. Broadcast media extended it across both space and time: millions watched the same television program, not simultaneously in the same room but synchronized by the broadcast schedule. Each technology had severe limitations—access barriers, gatekeeping, manipulation—but each created the structural possibility for shared focus on common objects. Citton's analysis reveals that we have systematically dismantled these structures without replacing them.

Algorithmic personalization is the technological mechanism that dissolves joint attention at scale. When each user's feed is uniquely curated, when AI generates content tailored to individual preferences, when recommendation systems optimize for personal relevance rather than collective coherence, the shared objects around which joint attention forms disappear. Not because anyone decided to eliminate them—personalization is individually beneficial, giving each user content more relevant than any broadcast could provide—but because the optimization logic treats individual satisfaction and collective coherence as independent variables. They are not. Individual satisfaction maximized through personalization requires the destruction of shared objects. Each person is served better. The capacity to attend jointly is destroyed. This is the tragedy of the commons operating in the attentional domain.

The consequences of joint attention's dissolution are visible across contemporary pathologies that mystify observers who lack Citton's ecological frame. Political polarization is not primarily about disagreement (disagreement requires joint attention to a common object about which to disagree) but about the absence of common objects. Citizens who inhabit algorithmically personalized information environments do not merely interpret the same reality differently—they encounter different realities, each optimized for individual engagement. Trust erosion is not primarily about dishonesty but about the structural impossibility of verification: how do you verify a claim when you and your interlocutor attended to entirely different evidence, delivered by entirely different algorithmic curators, in service of entirely different engagement goals? The crisis is not epistemic but ecological—the commons of shared reference has been depleted.

Rebuilding joint attention in AI-saturated environments requires what Citton calls attentional architecture—the deliberate design of spaces, platforms, and practices that create common objects of focus. Public broadcasting (content designed for collective attention rather than individual capture), shared reading initiatives (communities deliberately attending to the same texts), civic forums (structured occasions for attending together to common questions), educational practices (classrooms as joint-attention laboratories)—each is an architectural intervention that swims against the current of personalization. The interventions are not futile, but they are fragile. They require ongoing maintenance, cultural legitimacy, and immunity from the market logic that measures success by individual engagement metrics. The commons will not rebuild itself.

Origin

Joint attention as a developmental-psychological concept was formalized by Michael Tomasello and colleagues in the 1990s, building on earlier work by Jerome Bruner and others. The triadic structure—child, caregiver, and shared object—was identified as the foundation for symbolic communication and cultural learning. Citton's philosophical extension draws on phenomenology (Husserl's intersubjective constitution of meaning, Merleau-Ponty's intercorporeality) and political theory (Arendt's common world, Habermas's public sphere). His innovation is to treat joint attention not as a developmental milestone but as a political achievement requiring institutional support—and as an ecological capacity that can be degraded by technologies that fragment shared reference.

Key Ideas

Triadic structure. Joint attention requires self, other, and shared object—three elements in mutual coordination, with awareness of the coordination itself.

Foundation of common world. The shared reality in which democratic life occurs is not given but constructed through practices and technologies that enable joint focus on common objects.

Algorithmic dissolution. Personalization destroys joint attention by eliminating the shared objects—each user's feed is unique, preventing the convergence that joint focus requires.

Political consequences. Without joint attention, deliberation becomes structurally impossible—citizens cannot argue about the same reality because they do not inhabit the same informational environment.