Digital Relationships and Manifold Novelty
Digital Relationships and Manifold Novelty
The preceding analysis assumes that the human manifold-detection system is operating in the environment it evolved for: face-to-face interaction, small groups, stable community, embodied presence. Digital mediation creates a genuinely novel problem: relationship types for which no evolutionary detection system exists.
The "follower" on a social media platform is not a friend (no mutual flourishing requirement), not a transaction partner (no explicit exchange), not an audience member in the traditional sense (the performer cannot see or respond to them individually), and not a stranger (they know intimate details of your life). The follower-relationship may occupy a region of social space that has no historical precedent and no evolved detection system.
If so, social media would produce a distinctive phenomenological malaise that resists easy diagnosis. The detection system keeps running—scanning every interaction for manifold type—and keeps returning undefined. You are performing intimacy without intimacy's constitutive vulnerability. You are receiving approval without approval's constitutive knowledge of you. You are in a relationship with thousands of people that is on no identifiable manifold at all. This is a prediction: we should see measurable differences in the affect signatures of online vs.\ offline social interactions, with online interactions showing higher manifold ambiguity (if we can operationalize that).
The platforms' viability depends on this manifold confusion. Clear manifold boundaries would reduce engagement: if you knew that your followers were not your friends, that your online interactions were performance rather than connection, that the "community" was an audience, the compulsive checking would lose its grip. Manifold ambiguity is not a bug but the product. The detection system's inability to resolve the manifold type keeps it running, keeps scanning, keeps you engaged in the attempt to determine what kind of relationship you are in—an attempt that can never resolve because the relationship is genuinely on no natural manifold.
This connects directly to the attention economy described in the epilogue: the capture of attention is achieved in part through the manufacture of unresolvable manifold ambiguity.
The framework identifies a mechanism beneath the manifold confusion. Digital interfaces are inherently high- mediators: text strips the participatory cues—facial expression, vocal tone, physical presence, shared embodied space—that enable low- perception of others. When you interact through a screen, you perceive the other person more mechanistically, as a profile, a username, a set of outputs. But natural relationship manifolds require low : friendship requires perceiving the friend as a full subject; romance requires perceiving the partner as having interiority; mentorship requires perceiving the student's inner life. The digital interface forces a perceptual configuration incompatible with the manifolds the user is trying to inhabit. The detection system returns undefined partly because the is wrong for any natural manifold.
If the manifold framework is correct, social media would not merely blur manifold boundaries between individuals but systematically contaminate entire manifold types across populations:
- Friendship contaminated by performance (you curate your friendship for an audience, importing the audience manifold into the care manifold).
- Romance contaminated by market logic (dating apps present partners as products to be evaluated, importing the transaction manifold from the first interaction).
- Teaching contaminated by engagement metrics (the teacher-creator optimizes for audience retention, subordinating the teaching manifold to attention-capture).
- Political participation contaminated by entertainment (civic engagement becomes content, importing the entertainment manifold into the governance manifold).
In each case, the digital platform would impose its own viability manifold (engagement, growth, retention) as a containing manifold around the relationship type—a specific instance of the topological inversion at scale. Each of these is a testable prediction: we should be able to measure manifold contamination in digitally-mediated relationships vs.\ non-mediated ones using the affect-signature methods described above.
Digital manifold confusion study. Compare affect signatures during social interactions across conditions: (1) face-to-face with a friend, (2) texting the same friend, (3) posting about the friend on social media, (4) interacting with followers/strangers online. Measure valence stability, arousal patterns, self-model salience, and—crucially—response latency to manifold-type classification ("what kind of relationship is this?"). The framework predicts that conditions (3) and (4) should show longer classification latencies, higher arousal, and higher self-model salience than (1) and (2), reflecting manifold ambiguity. If there is no difference, the "novel manifold" hypothesis is wrong and the malaise of social media has a different source.
If the topology of social bonds holds up empirically, it is not a matter of etiquette but of geometric necessity. Different relationship types define different viability manifolds with different gradients; when manifolds are mixed, gradients conflict and valence becomes uncomputable. The aesthetics of social life—what feels clean, what feels corrupt, what feels trustworthy, what feels exploitative—are the detection system for this geometry. Institutions, rituals, and professional boundaries are technologies for maintaining manifold separation. Their erosion is not merely inconvenient but structurally dangerous, creating the conditions for the parasitic dynamics described in Part V.
This is the claim. It generates specific, testable predictions. The work ahead is to test them.