Aesthetic Cognition
Aesthetic cognition, in Goodman's framework, is not a feeling or an emotional response but a form of genuine knowledge—understanding of the world achieved through the specific resources of aesthetic symbol systems that propositional systems cannot replicate. A Vermeer interior provides knowledge about the quality of light falling through a window—not the physics of light (the scientist handles that) but the perceptual, spatial, experiential dimension of how light constitutes an inhabited space. The knowledge is irreducible: it cannot be extracted from the painting and restated in words without becoming a different kind of knowledge, because the understanding is constituted by the painting's density, its exemplificational richness, its specific deployment of color and composition. Aesthetic cognition operates through sustained attention to the particular—the unrepeatable configuration of properties this work possesses. It demands what Goodman called saturation: the readiness to find significance in every discriminable feature, the alertness to minute variation that dense symbol systems require. The cognitive yield is understanding that generalizes poorly but deepens continuously—each encounter with the work reveals features that prior encounters missed, because the density of the symbol system exhausts attention without exhausting significance.
In the AI Story
Goodman's account of aesthetic cognition rejected both formalism (the doctrine that art is purely about formal properties) and emotivism (the doctrine that art is about feelings). Aesthetic works are about the world—they refer to objects, properties, events through the modes of denotation, exemplification, and expression. But they are about the world in a way that propositional systems cannot replicate, because the referential modes they employ (exemplification, expression) and the formal properties they exhibit (density, repleteness) provide understanding that is structurally unavailable through differentiated, propositional symbolic means. What a painting knows about light, a poem knows about grief, or a musical phrase knows about tension and resolution is genuine knowledge—worldmaking that organizes experience and yields insight—but it is knowledge of a kind that resists translation into the propositions that scientific or philosophical systems employ.
The cognitive value of aesthetic engagement is grounded in the attentional demand that dense symbol systems impose. A viewer cannot skim a Vermeer the way a reader can skim a news article, extracting the gist and moving on. The painting's density means that every perceivable feature is potentially significant—the specific luminosity of the wall, the precise angle at which light falls, the texture of the fabric, the compositional relationship between figure and ground. To understand what the painting knows requires attending to these features with the kind of saturated, sustained, discriminating attention that differentiated systems do not demand. The attention is cognitively costly. It is also cognitively productive—it builds the perceptual and conceptual resources through which the viewer sees not just this painting but light, space, and human presence generally.
AI-generated images can display formal properties that appear to invite the same attentional engagement. A diffusion model can produce compositions with subtle tonal variations, intricate spatial organizations, apparently deliberate emphases. The invitation is real—viewers may find themselves attending closely to AI-generated images, perceiving significance in the formal choices. But Goodman's framework asks: Is the significance there, or is it being projected by the viewer onto an output that was not configured to carry it? If the formal properties were generated by sampling from probability distributions rather than selected for their exemplificational value within a worldmaking project, then the properties may be present without being purposefully deployed. The viewer perceives significance because the viewer has been trained by engagement with human art to find significance in formal properties. The training leads the viewer to attribute to the AI output the kind of purposeful configuration that human works actually possess. The attribution may be mistaken—significance projected rather than perceived—and the mistake is undetectable from the inside, because projected significance and actual significance produce identical phenomenological experiences.
Origin
Goodman developed his account of aesthetic cognition across multiple works, particularly in 'Art and Inquiry' (1967), the final chapter of Languages of Art titled 'Art and Understanding,' and essays collected in Of Mind and Other Matters (1984). The framework built on his epistemological work showing that understanding is not passive reception but active construction, and that different symbol systems provide different, irreducible forms of understanding. The concept has influenced cognitive science's understanding of how perception and conception interact, how attention is educated through practice, and how aesthetic experience contributes to the development of general cognitive capacities.
Key Ideas
Art provides genuine knowledge. Aesthetic understanding is cognitive, not merely emotional—paintings, music, and literature organize experience and yield insights that propositional systems cannot replicate.
Understanding is symbol-system-specific. What a painting knows about light is available only through the painting's specific symbolic resources—verbal descriptions provide different, not better or worse, knowledge.
Density demands sustained attention. Aesthetic cognition requires the readiness to find significance in infinitesimal variation—an attentional discipline that dense symbol systems impose and that differentiated systems do not demand.
AI displays without configuring for significance. Outputs may possess formal properties that invite attentional engagement, but the properties were sampled, not selected—significance may be projected by the viewer rather than built into the work.
Appears in the Orange Pill Cycle
Further reading
- Nelson Goodman, Languages of Art, final chapter (Hackett, 1968)
- Catherine Z. Elgin, 'Relocating Aesthetics: Goodman's Epistemic Turn,' Revue Internationale de Philosophie 253 (2010)
- John Gibson, 'Cognitivism and the Arts,' Philosophy Compass 3/4 (2008)