Psychoanalysis Beyond Pop Diagnosis: Reading Digital Culture Without Labeling People

Psychoanalysis & Culture · Sources checked 2026-07-24

A pattern on a screen can support an interpretation of a system. It cannot, by itself, authorize a clinical label for a stranger.

By Dr. RYU · NEWS TEXT

Prediction is not knowledge of a person

Recommendation systems learn from behavior, similarities among users, and features of available content. They can make useful predictions without knowing why a person clicked. A user may select something from desire, boredom, professional obligation, irritation, or lack of alternatives. The same observable act can belong to different stories.

Psychoanalytic thought is valuable here precisely because it distrusts the fantasy that desire is fully transparent. Desire is shaped through language, absence, repetition, prohibition, and the imagined wishes of others. Kehl's 2026 theoretical article examines platform algorithms as technical mediators that can intensify or reorganize conditions of repetition and desire. It is not an experiment proving a user's mental state.

The value of that framework is therefore caution, not mind-reading: it keeps observable behavior, platform-designed choice conditions, and a person's own account from collapsing into a single data profile.

Repetition can be analyzed without becoming a diagnosis

A feed may reward repeated viewing, rapidly supply similar material, and remove natural stopping points. Cultural analysis can ask how these features organize attention and how a platform converts repetition into value. It can compare interfaces, defaults, metrics, and public policy. None of this requires declaring that an identified user is addicted, compulsive, narcissistic, or disordered.

The boundary is both ethical and methodological. Clinical terms depend on more than one recorded behavior and belong to an appropriate assessment context. Public writing should analyze the text, system, and discourse it can actually observe, while leaving a person's mental state undecided.

Whose objective appears in the feed?

Users may want relevance, surprise, diversity, calm, or a quick answer. Platforms may optimize engagement, purchases, satisfaction, safety, or creator supply. Advertisers and publishers add further aims. Recent recommender-systems research examining how users perceive system objectives is a reminder that the purpose users imagine may differ from the objectives designers measure.

Saying ‘the algorithm wants’ is therefore a metaphor, not evidence of machine desire. The concrete questions are who selected the objective, which behavior stands in for success, which values are hard to quantify, and who bears the tradeoffs.

The personalized screen as a partial mirror

A feed resembles a mirror because it returns traces of prior action. But it is a partial and engineered mirror. It reflects available data, commercial and safety constraints, the eligible content pool, and the system's model of relevance. Unshown alternatives cannot become revealed preferences. Repeated exposure may also shape the next action, making prediction look self-confirming.

Flisfeder's scholarly monograph Algorithmic Desire reads social-media algorithms as part of a cultural and ideological organization of desire rather than as a transparent portrait of individual personality. That is a critical framework, not a clinical instrument or a universal explanation for every platform.

UNESCO's ethics recommendation links algorithmic provision of information with media and information literacy and asks institutions, not only individuals, to address impact. Responsible interpretation should therefore examine user agency and platform governance together.

A disciplined psychoanalytic reading

Begin with an observable feature: autoplay, an infinite feed, a recommendation label, an omission, a recurring genre. Describe the institutional setting and incentive. Use a psychoanalytic concept to generate a question, not to close the case. Then test alternative explanations and state what the available evidence cannot reveal.

Psychoanalysis is not a secret decoder ring for algorithms or users. It is most useful in public criticism when it disrupts easy identities: a click is not a whole desire, repetition is not a diagnosis, and prediction is not self-knowledge.