Curation
Variant slug: curation
Version: v1.0
State: Canonical
Reviewed against: derivation-protocol v1.0
Philosophical grounding: Hume / Moore / Groys / O'Doherty
Opening¶
Curation is a profession organised around a sentence. The sentence is simultaneously description ("this work concerns memory and exile"), evaluation ("this is the strongest instance of the concern in the exhibition"), and prescription ("therefore it has been placed at the entry wall"). The three moves are delivered in a single utterance whose genre has not changed in a century. The wall text is the compressed form. The catalogue essay is the long form. The press release is the market form. The curator's talk is the oral form. The exhibition design — lighting, plinths, adjacencies, title placement — is the non-verbal form. Each performs the same fusion. A visitor, a reviewer, a funder, and a later historian all receive a bundled object in which description, thesis, and verdict cannot be located as separate moves.
This is Split-Domain Cognition's failure mode operating at the scale of a cultural institution. The curatorial sentence smuggles evaluative and prescriptive work into what presents itself as descriptive reporting. Unlike many other failures of the same kind, the curatorial failure has an institutional body: the museum, the biennale, the commercial gallery, the art fair, the auction house. The institutional body converts the fused sentence into market valuation, reputational sorting, and historical record. The collapse is therefore not only a rhetorical failure; it is the mechanism by which contemporary art history is produced.
The variant that follows describes an alternative kind of curation — one in which the three layers are held apart as separate artefacts, contestable independently, recoverable across time. The alternative is not new. It has been practised by individual curators and by occasional institutional programmes whenever the genre's default fusion was refused. The contribution of this document is to name the move, give it a structural account, and make it transferable.
The language work in this domain¶
The language work of curation is the literal attention to each work on its own terms. What the work is made of, what the artist has said about it, what prior reception has recorded, what the materials do, what the scale is, what the work does in time, what the viewer's body has to do to encounter it. This record is the descriptive artefact the rest of the curation can be traced to. It is sometimes called a condition report in conservation contexts, a catalogue entry in collection-management contexts, or an object file in archival contexts. What is distinctive about its curatorial form is that it precedes and is distinct from the thesis the curator will later apply.
Layer 1 in this domain has a particular posture: it holds works and observations open without forcing them into the evaluative register. A descriptive entry that already encodes the verdict the thesis will later issue is not a Layer 1 artefact; it is a Layer 3 artefact in disguise. The discipline of Layer 1 is to produce descriptions that stand independently of the curatorial argument — not as neutral reporting (no description is neutral) but as descriptions the thesis will be required to connect to, rather than descriptions whose shape has already been carved to fit the thesis.
The language work also includes the artist's own voice. The artist has said things about the work — in interviews, in studio visits, in notes, in earlier exhibitions — that are part of the record the curator will later interpret. Preserving the artist's own language at Layer 1 protects the exhibition from being the curator's sentence alone. The visitor should enter the exhibition with their own voice audible, not drowned by the curator's; this requires that the curator's voice be recognisable as a voice (Layer 2 declared) and not mistaken for the voice of the works themselves (Layer 1 preserved).
Language work alone is insufficient. This must be stated plainly because the profession's default objection to any descriptive-only discipline is that it produces nothing to act on. A full condition record, a catalogue of observations, a thick archival account — each of these by itself does not close a curatorial question. The exhibition still has to open; the wall text still has to be written; the works still have to be placed; the catalogue still has to make an argument. Layer 1 is the material the thesis can be traced to, not a substitute for the thesis. SDC in curation does not argue that curators should only describe. It argues that descriptive work should be preserved as a distinct artefact before the thesis is applied, so that the thesis and the description can be contested independently when the exhibition opens.
The judgement work in this domain¶
The judgement work is the curatorial thesis. This is what the exhibition is arguing, what it includes and excludes under that argument, what it positions centrally and peripherally, what it frames and what it refuses to frame. The thesis is the Layer 2 artefact: the rubric against which the descriptive material is assessed and from which the inclusion-and-positioning decisions follow.
A thesis can be issued invisibly or explicitly. The curatorial profession's default is invisibility: the thesis operates through the selection and arrangement of works, through the voice of the wall text, through the catalogue essay's rhetorical authority, and through the exhibition design, but is rarely stated as a separate artefact the visitor can read, argue with, or cite. The thesis is what everyone knows the curator has decided; it is not what the curator writes down as a thesis.
The alternative is to declare the thesis explicitly. A curator working under SDC publishes a thesis document as its own artefact: a statement that says what the exhibition is arguing, marks the statement as a pronouncement rather than a finding, and makes itself available to be contested as a position rather than absorbed as observation. The reader can refuse the pronouncement — the document is not pretending to coerce — but cannot mistake it for a report. The thesis is visible as a thesis, and the rest of the exhibition operates under its declaration.
The institutional layer carries the same requirement. Galleries, museums, biennials, and other curatorial bodies operate under framework commitments that shape what thesis is possible, what descriptions are admissible, and what narration is permitted. These commitments should themselves be declared as institutional framework — what the body is, what it can and cannot do, who funds it, whose interests it serves, what it refuses to programme — and not absorbed into an implicit "we" that lets the framework's work happen invisibly.
The judgement work in curation has a specific further commitment that does not appear in every variant of SDC: the commitment to revisability. A rubric that cannot be revised is a rubric that is not being held as a rubric; it is being held as a position the rubric-author is now committed to defending against all comers. The curatorial thesis must be revisable in public acts — published reconsiderations, follow-up exhibitions, corrigenda, retrospective essays — or the exhibition's closing becomes the thesis's permanent state. A practice that refuses to finish — that holds the thesis open across years rather than closing it at the exhibition's run — is the architectural enforcement of revisability over time.
The collapse, with examples¶
The collapse in contemporary curation is extensive. The inventory below names the forms the fusion takes in ordinary curatorial practice. None of them is a bad faith act on the part of individual curators. All of them are genre conventions that curators inherit, practise, and teach.
The wall text as compressed verdict. Two hundred words that carry description of the work, thesis of the exhibition, evaluative framing of the work within the thesis, and positioning cue for the visitor's response. The four moves are fused at the level of sentence. The visitor who disagrees with the framing cannot locate the disagreement because the framing has been absorbed into the description.
"The artist is exploring …" A descriptive-sounding phrase that does evaluative work. The curator has decided what the artist is exploring; the artist has not necessarily said this in the interview; the verdict is delivered as observation. "Exploring" is one of several verbs that perform the collapse — investigating, interrogating, foregrounding, grappling with, addressing, confronting, questioning — each presents the curator's reading of the work as the work's own project.
The catalogue essay as monologue. Catalogue essays routinely fuse descriptive presentation of works with evaluative framing, without marking which move is being made when. A reader who disagrees with the framing cannot locate the disagreement — the essay's rhetoric ensures that disagreement reads as aesthetic ignorance rather than a position against the thesis. The fusion is architectural: the essay genre does not have a separate place for description and argument because the convention is that they are the same document.
"Important artist", "important show", "important work". Important is a rubric-verdict in adjective form. It declares that the framework that measures importance has produced a positive reading, without naming the framework and without describing what produced the reading. It is the verbal form of a market signal.
The biennale as compressed thesis. Biennials announce themes and select works under the themes. The theme is Layer 2; the selection is its application to Layer 1; the catalogue narration is Layer 3. In practice all three are folded into the curator's essay, which typically runs 3,000 to 5,000 words and contains all three moves. Visitors receive a unified curatorial statement and cannot examine the thesis independently of the works or the works independently of the thesis. Biennial criticism — the journalistic reviews that follow — re-performs the fusion at the next layer up: the reviewer evaluates the exhibition by restating the thesis and asking whether the works met it, rather than evaluating the thesis itself and evaluating the works against their own terms.
The curator as figure. In the commercial gallery and biennale circuits, the curator's name has become a reputational object that travels independently of the exhibition's content. "Curated by X" is a warrant for the works; the works' warrant is the curator's reputation; the reputation was earned by prior exhibitions whose thesis-rubric and work-descriptions were similarly fused. The curator-as-figure converts the fusion into a recursive market mechanism: each generation of exhibitions consolidates the next generation's reputational objects.
The press release as fused object. The press release is a six-hundred-word artefact produced at the exhibition's opening and distributed to journalists and institutions. It compresses description, thesis, positioning, and market cue into a form designed to be reproduced verbatim in coverage. Reviews that follow a press release reproduce its fusion; reviews that deviate from it are usually deviating in the reviewer's direction (a second fused object layered on top of the first), not into the separation the SDC architecture would make available.
The exhibition design as silent thesis. Adjacencies, lighting, plinth heights, wall colours, label typography, title placement, entry-wall selection — each is a design decision that carries the thesis without speaking it. A visitor encounters the thesis through the design before they read the wall text. The non-verbal thesis cannot be argued with because it has not been stated; it can only be felt. Exhibition design is the curatorial layer most completely outside the visitor's capacity to examine, because it does not present itself as a layer at all.
The "context statement" as smuggled rubric. A recent genre: the curator prefaces the exhibition with a contextual essay naming the political, social, or historical circumstances under which the exhibition is occurring. The context statement appears descriptive. It is rubric-work — it establishes what the works will be read against, which framings are legitimate, which critiques are admissible. The genre has become standard in institutional shows since roughly 2017 and now operates as a stealth Layer 2 that looks like Layer 1.
The "what is not being shown" move. A newer genre: the curator names what has been excluded from the exhibition, ostensibly to make the selection criteria visible. In practice, the exclusion statement performs the fusion again — exclusions are described as though the framework had observed them, rather than as framework-decisions the framework is making. The statement reads as transparency while performing the same compression the essay would have performed.
The refusal of the refusal-to-finish. An exhibition closes. Its thesis closes with it. The catalogue is printed; the images are archived; the photographs circulate. Revisability does not outlive the run. The architectural alternative — holding the thesis open across years rather than closing it at the exhibition's run — would require an institutional commitment that most exhibitions cannot make; the absence of the commitment is the fusion's institutional form.
The cost of the collapse¶
The costs of accepting the fusion as curation's default vocabulary are specific, cumulative, and recognisable to the profession when pressed.
Audiences cannot re-enter the argument. A visitor who disagrees with an exhibition's premise has no place to stand. The premise has been absorbed into the descriptions of each work, so objection to the premise reads as objection to the works. The visitor is offered a binary — accept the show or refuse its aesthetics — when what was needed was a triad (accept the thesis, accept the descriptions, accept the inferences from one to the other) in which disagreement at any layer could be articulated without collapsing into dismissal.
Curators become authorities rather than authors. The framework-level claim is hidden, which means the claim's authorship is hidden, which means the curator speaks from a position rather than a commitment. Position is defended against critique; commitment is open to revision. The institutional incentives favour position because position produces reputational stability; commitment produces reputational risk. Over time the profession consolidates toward position, and the authorial voice — the curator as someone willing to say what they are arguing and revise it — atrophies.
Markets respond to fused verdicts, not to separated judgements. When important means valorised by the curation, and curation is fused with description, market valuation follows the fusion. Curatorial capital becomes financial capital through a mechanism that cannot be examined because its intermediate steps have been absorbed into a single rhetorical object. The fusion is not only an epistemic failure; it is a market mechanism, and the auction houses have a direct interest in keeping it opaque.
Pedagogy is impossible. Curatorial teaching becomes apprenticeship to the fused voice — trainees absorb a style rather than learning the three distinct practices the form could make teachable. A student who tried to write a curatorial text that separated description from thesis from narration would be told the text "didn't work as curation" — the voice is what is being assessed, and the separated voice sounds wrong. The implicit assessment of the genre-voice produces reproduction of the genre, which produces reproduction of the fusion.
Institutional drift is invisible. A museum collecting over decades under implicit rubrics drifts toward the tastes of successive curators without a rubric-level conversation. The drift is invisible because the rubric was never explicit enough to be checked. Decades later, when the collection's orientation is examined, the examiners can read tendencies but cannot reconstruct decisions. The institution has no record of its own thinking.
Criticism cannot distinguish exhibition-failure from thesis-failure. A review that finds an exhibition unconvincing cannot easily say whether the works failed, the selection failed, the thesis failed, or the narration failed. Criticism therefore operates at the same compressed level as curation — evaluative sentences applied to fused objects — and loses its capacity to contribute to revision.
Audience learning does not accumulate. A visitor who has seen fifty exhibitions under the fused genre has fifty experiences of exhibitions; they do not have a fifty-exhibition education in how to read a thesis, evaluate a framework, locate a disagreement, or follow a lineage. The genre's structural opacity prevents the viewing capacity from developing, because what would be taught (how to read the layers) is never made available.
The refusal-to-finish disappears as an option. An exhibition that closes becomes a verdict. A verdict issued by the institutional body of the museum or biennale circulates as permanent. The refusal is architecturally available — a practice can hold its thesis open across years — but the institutional form does not permit running exhibitions indefinitely. The fusion's institutional embodiment forecloses a move the architecture could otherwise make.
The separation, in curatorial language¶
Applying SDC in this domain means producing three artefacts where the current genre produces one.
Layer 1 — the descriptive artefact. Condition reports for each work, artist's own statements preserved verbatim, materials and provenance, prior exhibition history, photographs taken from the work's own terms rather than from the exhibition's, the ethnographic record of the work's making where available. Organised so that each work is readable without first reading the thesis. The test: if the exhibition's thesis were replaced with a different thesis, would the Layer 1 artefact still be usable? If yes, Layer 1 is descriptive. If no, Layer 1 has already absorbed the thesis and is not a distinct artefact.
Layer 2 — the thesis artefact, signed and revisable. A document that states what the exhibition is arguing, what it includes and excludes under that argument, what criteria are being applied, and what the curator has decided to refuse. The document is signed — the thesis has an author, not only an institution. The document is revisable — the thesis can be updated in a public act. The document is prior to the exhibition, in the sense that the visitor can read it before, during, or after the works and trace the works' presence to the argument.
Layer 3 — the narration that connects the first two. Wall text, catalogue essay, exhibition design, public programme, gallery talk, press release. Each makes visible how the inclusion and positioning of works follows from the thesis applied to the descriptions. The narration is explicitly narration, not description; it reports a verdict that has already been issued at Layer 2 against material preserved at Layer 1. The narration therefore does not issue verdicts at the moment of the visitor's reading — it shows the visitor where the verdict was issued and how to re-enter the argument. The reader (visitor) must do their own judgement work, and the narration's job is to make the Layer 1 and Layer 2 material available for that work, not to pre-empt it.
Layer 1 alone is insufficient — without a thesis, it is archival material that closes nothing. Layer 2 alone is a slogan — without descriptions, a thesis is a stance. Layer 3 alone is the current genre — compressed narration without separated layers, delivering a verdict the visitor cannot re-enter. The three together are the separation. Each layer is contestable independently. A visitor can accept the thesis and reject its application to a specific work; accept the descriptions and reject the thesis; accept the narration's account of how the two connect and reject the connection. The exhibition becomes a structure the visitor can enter rather than a verdict the visitor must absorb.
One further architectural commitment is specific to curation: the institutional layer should also be visible. A museum is not neutral; a biennale is not neutral; a commercial gallery is not neutral. The institutional body shapes what thesis is possible, what descriptions are admissible, and what narration is permitted. A curation under SDC would include, as a fourth artefact, an institutional statement: what the institution is, what it can and cannot do, who funds it, whose interests it serves, what it refuses to programme. This is not a political appendix; it is the background against which the other three layers become legible.
Philosophical grounding¶
The descriptive–normative distinction has been available since Hume. G. E. Moore formalised it at the level of the ethical proposition. Boris Groys has argued that curation is always evaluation-in-selection, which concedes the fusion as the profession's self-image. Brian O'Doherty named the white cube as the curatorial apparatus whose neutrality was itself a thesis. None of this literature is new; what SDC adds to it is architectural enforceability — the insistence that the distinction does not maintain itself in prose or in exhibition design and must be held up by form.
The curatorial-specific philosophical ground is the refusal of the tastemaker. A tastemaker is a figure who issues verdicts that accumulate into reputation through sheer repetition, without the framework against which the verdicts are being issued ever being made public. The tastemaker model is structurally closed: the framework is not revisable because it is not visible; the reputation is not contestable because it is not argued; the taste accumulates without accountability. The refusal of the tastemaker is the refusal to be the fused object that the tastemaker model requires. A curator under SDC issues verdicts; the verdicts are traceable; the framework is revisable; the authority is authored.
The work of holding description and thesis apart, of keeping the thesis revisable, of refusing the tastemaker model, is volunteered labour. Institutional incentives do not produce it; market incentives actively oppose it; audience expectations are calibrated to receive the fused object and find the separated practice uncomfortable. A curator who practises SDC does so as a volunteered commitment, not as a role-compliant activity. This is the ethical temperature of the variant and is why the curation-as-SDC move is most clearly seen in individual curatorial practices, and is rarer at the institutional scale.
A structural bridge to the worth argument is also operative. Hierarchical rankings of worth across domains fail at the unit level — domains of capacity do not share a unit in which a global weighted sum could be computed. The same structural observation applies to curatorial ranking. "This is the most important work in the show" is a statement that presupposes a common unit across works whose excellences are incommensurable. The curation of "important works" in a thematic exhibition routinely performs a global ranking across works whose yardsticks do not reduce to a common unit. The worth argument's structural observation — that the global ranking is arbitrary in the strict sense — licenses a curatorial practice that refuses to issue global importance-verdicts in favour of domain-specific readings. The refusal to assemble fragments into a finished hierarchy is this position in form.
Closing — what the variant makes possible¶
Three things become possible when the variant is written down and placed in the canon.
A curatorial teaching becomes teachable. The layers are nameable. A student can be assigned to produce a Layer 1 artefact, a Layer 2 artefact, and a Layer 3 artefact as separate documents, and the student can be assessed against each independently. The genre of the fused voice becomes one pedagogical option among several rather than the implicit default. Apprentice curators can learn practices rather than a voice.
A curatorial criticism becomes locatable. A reviewer can distinguish a weak thesis from a poor selection from a clumsy narration. The review can address the layer where the failure occurred rather than the compressed object. Criticism becomes a contribution to revision rather than an alternative verdict at the same compressed level.
A the audience is given a role. Under the fused genre, the visitor is a receiver of verdicts. Under SDC-curation, the visitor is a judgement layer that completes the exhibition in their own reading. This is not a democratic gesture — it is an architectural one. The visitor's judgement is not added to the exhibition; the exhibition is built so that the visitor's judgement is where the exhibition completes. The refusal to close at the curator's thesis is the gift.
The variant does not argue that the fused genre is wrong. It argues that the fused genre is one mode of curation, that the separated mode is another, and that the separated mode is available to anyone willing to refuse the fusion as the profession's default. Naming the move is what this variant does. What curators do with it is their own practice.
How to follow this variant¶
- Reference — cite the variant in framing. See Adoption — Reference.
- Alignment — your artefact performs the separation. See Adoption — Alignment.
- Derivation is for new domains; this slug is canonical, so derivation does not apply here. See Adoption — Derivation.
The badge URL pattern for this variant:
https://splitdomaincognition.org/badge/curation-reference.svg
https://splitdomaincognition.org/badge/curation-reference-dark.svg
https://splitdomaincognition.org/badge/curation-alignment.svg
https://splitdomaincognition.org/badge/curation-alignment-dark.svg
To declare an adoption of this variant, use the chooser on the Declare page.