Institutional Frame

In the AI Story

Duchamp's 1917 Fountain episode was the empirical demonstration that forced the institutional question. The Society of Independent Artists rejected the urinal not on aesthetic grounds—their charter precluded such judgment—but on categorical grounds: this is not art. The rejection proved Duchamp's point more effectively than acceptance would have. The institution's refusal revealed the institution's power—the power to say what counts and what is excluded, the power whose exercise is normally concealed by the presentation of categories as natural. When AI-generated output enters institutions whose categories were built for human production—academic journals, publishing houses, patent offices, professional licensing boards—the categorical stress produces visible institutional failure: prohibition, disclosure requirements, defensive gatekeeping disguised as quality assurance.

The art world's adaptation to Minimalism, Conceptual art, and the Pictures generation required institutional innovation—new galleries (alternative spaces, artist-run venues), new journals (October, Artforum's theoretical turn), new curatorial practices (thematic rather than medium-based exhibitions), new degree programs (visual studies, critical theory). The adaptation took decades and was contentious throughout. Established critics dismissed the new work, museums resisted exhibiting it, collectors refused to buy it. The work persisted because practitioners kept making it and because small, provisional institutions emerged to frame what the established ones could not see. AI production is following the identical pattern at compressed speed—computational creativity conferences, AI art exhibitions, new publishing imprints—but the gap between production volume and institutional capacity is orders of magnitude wider.

The evaluative frameworks that new institutions construct are never neutral. They encode the values, assumptions, and interests of their builders. When October framed postmodern appropriation art, it privileged criticality—the precision of the theoretical intervention—over the aesthetic pleasure that formalist critics had valued. The choice was not arbitrary; it reflected October's post-structuralist commitments and its political orientation. AI governance institutions are making analogous choices now—privileging safety, or democratization, or performativity, or human autonomy—and each choice produces a different frame that foregrounds different outputs and obscures others.

Krauss's structural insight is that the frame cannot be eliminated. Even the refusal to frame is a framing choice—the decision to leave output unevaluated is itself an institutional position that produces consequences (the excellent and the mediocre become indistinguishable, trust collapses, quality standards erode). The question is never whether to frame but who builds the frame, what values it encodes, and whether it develops quickly enough to serve the populations depending on it. The institutional lag Segal identifies—the widening gap between AI capability and institutional response—is, in Krauss's framework, the period during which the culture operates without adequate frames, and the cost of that period is borne by the practitioners and audiences who cannot tell what deserves attention in the undifferentiated flood.

Origin

George Dickie developed the institutional theory of art in a series of essays and books from the late 1960s through the 1980s, most influentially in Art and the Aesthetic (1974). Krauss engaged critically with Dickie's framework—finding it too sociological, insufficiently attentive to the formal properties of artworks—but acknowledged its core insight: that the art world's institutional apparatus is not external to art but constitutive of it. This insight, combined with her own formalist training, produced the distinctive Kraussian method: structural analysis of positions within a field whose boundaries are drawn by institutions.

Key Ideas

Categories are institutional constructions. What counts as art (or code, or authorship) is determined by framing institutions, not by intrinsic properties of objects.

Frame is constitutive, not decorative. The gallery wall does not merely present the artwork—it makes the work visible as art rather than as object.

Authority conceals contingency. Institutions present their categories as natural, concealing that the boundaries were drawn by interested parties and could be drawn differently.

New positions require new institutions. When material practice outpaces institutional categories, new framing apparatuses emerge—slowly, contentiously, with significant casualties.

Speed of adaptation determines cost. The gap between production and institutional recognition is the space where quality collapses into quantity because no evaluative frame exists to distinguish them.