Unhearing

In the AI Story

Adorno developed this analysis in the context of historical catastrophe. Weimar Germany's modernist culture—its art, philosophy, music, radical formal experiments—was not merely suppressed by National Socialism but unheared, rendered inaudible through reorganization of the categories of cultural reception. The administered world did not need to burn every modernist painting; it needed only to restructure perception so that difficulty, formal autonomy, and refusal to please no longer registered as art but as irrelevance. The paintings existed; the audience capable of receiving them had been dismantled.

The mechanism operates through what Segal calls the discourse's architecture. Platforms reward engagement, which favors clear positions ("This is amazing" or "This is terrifying") over qualified complexity ("I feel both simultaneously and cannot resolve the contradiction"). The silent middle—the largest and most accurate group—remains silent not because they lack voice but because the medium cannot carry their frequency. The algorithmic feed is tuned to amplify certainty; ambivalence produces no signal the algorithm can detect. The voices expressing irreducible loss are not censored—they are simply not amplified, which in an attention economy organized around amplification is a more total erasure.

Language models trained on this discourse inherit its unhearing. The models will generate enthusiasm about AI more fluently than grief about displacement because enthusiasm is more prevalent in the training data. The platforms that generated that data rewarded one frequency and did not reward the other. The model reproduces the asymmetry—not through bias in any simple sense but through the faithful statistical reproduction of a culture that has already structured what it hears and what it does not. The map of the model's generation is a map of the culture's hearing, and the blank spaces—the associations that never occur, the phrases that never appear together—are a map of what the culture unhears.

The critical task Adorno's framework identifies is not to amplify the unheared voices within the existing medium—that would merely integrate their grief into the system producing it—but to insist on the existence of frequencies the medium cannot carry. To maintain that what the system cannot hear nevertheless exists, matters, and must be reckoned with by any honest account. The Orange Pill performs this reckoning with visible effort, refusing to scroll past the elegists, insisting their loss is real. But the reckoning occurs within an architecture (the tower, the ascent) that exerts gravitational pull toward resolution. By the fifth floor, the argument has moved on. The frequency has been registered and then—structurally, not by authorial intention—left behind.

Origin

The concept is implicit throughout Adorno's work but receives explicit thematic development in Minima Moralia and the lectures on damaged life. It synthesizes Adorno's analysis of how the administered world operates (not through force but through categorical exclusion) with his diagnosis of the culture industry (which does not prohibit opposition but makes it structurally inaudible). The clearest single articulation appears in the fragment on "the splinter in your eye is the best magnifying glass"—the wound is the lens, the damage is the knowledge, and the system cannot hear the testimony of those it has damaged because their testimony speaks in frequencies it was designed not to process.

Key Ideas

Condition, not act. Unhearing is not a decision to suppress but a structural absence of the receptive apparatus—the system cannot process what is being said, so the message simply does not arrive within the system's reality.

More total than silencing. Suppression produces evidence of a suppressor and a thing suppressed; unhearing produces nothing—the voice dissipates, the speaker wonders if she spoke at all, and the culture continues without registering the absence.

Algorithmic architecture of unhearing. Platforms reward engagement-optimized content, rendering grief structurally disadvantaged—not censored but un-amplified, which in attention economies constitutes a categorical exclusion more effective than prohibition.

Training data inherits unhearing. Models trained on discourse that un-hears certain voices reproduce the asymmetry statistically—generating fluent enthusiasm and halting grief because the data distribution reflects the culture's existing hearing structures.

Insistence as resistance. The critical task is not amplifying the unheared within the medium but insisting the frequency exists outside it—maintaining that what the system cannot process nevertheless matters and must be reckoned with.