One-Dimensional Thought

In the AI Story

Feenberg's engagement with the concept throughout his career has been more complex than either wholesale endorsement or rejection. From Marcuse, Feenberg inherited the analytical framework — the recognition that advanced industrial technology shapes not only what people do but how they think, the concern that market-governed technological development produces specific cognitive pathologies, the commitment to identifying and contesting these pathologies through political rather than merely individual means. Against Marcuse's sometimes apocalyptic tone, Feenberg has consistently argued that one-dimensionality is real but never total — that historical evidence shows repeated cases of critical consciousness developing within conditions that theory predicted would foreclose it. The argument is not that Marcuse was wrong but that his framework requires supplementation by the empirical openness of the social construction of technology.

Applied to AI, one-dimensional thought identifies a specific cognitive pathology that smooth AI interfaces may cultivate: the preference for confident answers over productive uncertainty, for polished outputs over provisional drafts, for agreeable collaboration over challenging dialogue, for efficient solutions over difficult questions. Scott Timcke's work on algorithms explicitly frames contemporary AI as a mechanism of one-dimensional thought, arguing that recommendation systems and language models systematically favor mainstream, consensus, conventional responses in ways that narrow the space of cognitive possibility. The concern is not that AI produces false information (though it sometimes does) but that it trains users to prefer a specific register of thought — the register the systems were optimized to produce — at the expense of registers the systems cannot produce or do not produce well.

The concept connects directly to Feenberg's recursion problem: if AI cultivates one-dimensional thought, then AI erodes the cognitive capacities that critique of AI would require. This is where Marcuse's framework becomes most challenging for Feenberg's democratic project. Marcuse's own response to one-dimensionality was not optimistic — he looked to marginal populations and aesthetic experience for resources of resistance, and his later work had a distinctly utopian character that acknowledged the scale of the pathology he had diagnosed. Feenberg's position is more empirically grounded: historical precedent shows critique developing within one-dimensional conditions, and the task is to identify and support the institutional structures through which such development becomes more likely rather than less.

Origin

Marcuse developed the concept in One-Dimensional Man: Studies in the Ideology of Advanced Industrial Society (1964), drawing on Frankfurt School analyses of instrumental reason and Hegelian concepts of negation. The book became one of the most widely read works of twentieth-century philosophy and substantially shaped the intellectual background of the New Left, where Feenberg's own political formation occurred.

Key Ideas

Reduction of reason to instrumental calculation. The cognitive pathology of absorbing all rationality into the calculation of means.

Closure of alternatives. The specific harm of one-dimensionality is that alternatives become not merely suppressed but unthinkable.

AI as contemporary instance. The smooth interface may cultivate specific one-dimensional dispositions adequate to Marcuse's diagnosis.

Real but not total. Feenberg's consistent supplement to Marcuse: one-dimensionality is demonstrably present but never complete.

Recursion implication. If AI cultivates one-dimensional thought, the capacities for critique of AI are themselves at risk from the technology being critiqued.