An archive needs names and addresses before every item's origin, date or significance can be established. Those first descriptions make retrieval possible, but they also influence later interpretation. RawIndex is a Socioplastics operator for registering uncertain material without presenting provisional classifications as settled facts. Warburg shows how arrangements can propose historical relations; Otlet demonstrates the practical value and political risk of comprehensive retrieval systems; Briet explains how institutions make documents; Perec redirects attention to overlooked everyday evidence; and Eco distinguishes openness from unrestricted interpretation. The essay develops a practical requirement: preserve an item's location, provenance, competing descriptions and revision history, and identify when a provisional label acquires institutional authority. The aim is not to delay classification indefinitely. It is to make classification decisions visible, revisable and accountable.
Addressing Material Before Classifying It
An archive begins with a practical decision: how can a document be found again? A catalogue number, temporary title or box location makes retrieval possible before its meaning is settled. But even a preliminary description influences later research. A drawing filed as an architectural original may be given priority over workers' invoices or residents' testimony, although those other materials could answer questions the drawing cannot. RawIndex names a way of addressing materials while keeping their uncertain status explicit. The record identifies what is known, what is inferred and what remains disputed. It also records who supplied a description and when. The proposal is useful when a collection must remain searchable before agreement about its categories is possible. It fails if provisional labels become permanent without review, or if uncertainty is used as an excuse for missing provenance and poor documentation.
A mixed architectural collection makes this difficulty palpable. Imagine an inherited set of drawings, photographs, contractors' invoices, interviews with former inhabitants and fragments from a demolished exhibition. One curator might catalogue the drawings as originals and subordinate the others as supporting evidence. Another could place the interviews at the centre and treat technical drawings as partial claims made by a design institution. A third might reconstruct the maintenance history of the building through invoices and changes of material. None of these schemes is automatically wrong. The political difficulty arises when one classification becomes the sole condition through which all later questions can be asked. A RawIndex gives each fragment an address and a description of its documentary uncertainty before imposing a total hierarchy of relevance. This does not postpone interpretation indefinitely. It creates the possibility of testing rival interpretations against the same retrievable materials, preserving evidence of who introduced a category and when. The concept is only useful if provisionality is treated as accountable labour rather than an excuse for poor documentation.
Warburg and the Nonlinear Neighbour
Aby Warburg's unfinished Mnemosyne Atlas remains a compelling interlocutor because its photographic panels arranged relations that could not be reduced to either chronology or a conventional art-historical genre. Recurrence, gesture, migration and visual survival were made perceptible through adjacency. The panel operated as a provisional argument: a constellation could suggest a historical relationship while still requiring interpretation of the passages connecting its images. This is more demanding than celebrating associative freedom. The juxtaposition of two images is evidence of an editorial decision, not yet evidence of an actual historical transmission. To read Warburg productively for RawIndex is therefore to distinguish the generative power of arrangement from the responsibility to substantiate the relations it implies. The atlas shows that classification can be spatial, comparative and open to revision. It also shows why a plate's authority cannot be confused with the authority of the images it gathers.
The difference between an atlas and a raw index is equally important. Warburg's panels already advance a highly informed interpretive proposition, whereas RawIndex concerns the earlier capacity to register materials whose interpretive affiliations remain unstable. The two instruments are complementary rather than interchangeable. One could imagine a group of photographs first held in a provisional catalogue and then repeatedly assembled into competing panels, with the index preserving both their documentary histories and the succession of constellations. Such an infrastructure would allow an investigator to return from a persuasive visual claim to the fragment's source, custody and unresolved metadata. It would also make visible the labour of the editor, who is neither absent from the atlas nor identical with the actors represented in it. Warburg thus tests the ambition of Socioplastics: an extensive network of conceptual associations acquires historical force only when the architecture of retrieval can support the critique of its own arrangements.
Otlet and the Temptation of Universal Access
Paul Otlet's documentation projects sought to make knowledge systematically available through classifications, cards, bibliographic networks and an international intellectual infrastructure. Their relevance cannot be reduced to anticipation of the internet. They illuminate the ambition to construct a durable, shared apparatus of retrieval across media and institutional boundaries. Socioplastics inherits something of this architectural ambition in its numbered nodes, Books, Tomes and public index, but its relation to Otlet should remain critical. A universal system promises that knowledge becomes more accessible as its categories become more exhaustive. The promise can obscure the problem of who establishes those categories, which languages become defaults and how resistant practices are translated into admissible documentary units. Universal access does not necessarily produce equal interpretive authority. The documentary infrastructure may extend the reach of certain epistemologies while rendering others peripheral through apparently technical decisions.
RawIndex offers a way to hold this contradiction open. It preserves Otlet's insistence that material must have an address if it is to circulate beyond private custody, while resisting the conclusion that complete addressability requires completed conceptual agreement. A provisional descriptor is not a weak version of a definitive descriptor. It is a different kind of information, one that reports the state of an inquiry as well as a claim about an object. The researcher might register the date of description, the evidence supporting it, a confidence level and an alternative title proposed by another contributor. But such fields can themselves become rigid. Their value is conditional upon maintaining a route by which an item can contradict its own catalogue entry. Otlet consequently supplies the infrastructural scale against which RawIndex must defend its commitment to uncertainty. Without the former it risks locality without circulation; without the latter it risks global navigation purchased at the price of epistemic closure.
Briet and the Made Document
Suzanne Briet radically complicates the intuition that an archive merely receives pre-existing documents. Her reflections on documentation emphasise the institutional and cultural process through which an object comes to function as a documentary sign. The famous question of whether an animal can become a document matters less as a provocation than as a test of institutional mediation: the same entity acquires different evidentiary capacities when collected, catalogued, described and positioned inside an organised practice. RawIndex therefore cannot claim to operate before every act of classification. Assigning an address is already an intervention through which something begins to count as documentable. To call an index raw must not imply that it inhabits an innocent exterior to institutions. It signals that the relations established by those institutions remain open to challenge and have not yet been naturalised as a final ontology.
Consider an oral account of a building's undocumented alterations. A heritage office may reject it as unverifiable memory, while a local history group treats it as the principal description of lived use. The matter is not settled by entering it under the word testimony. An accountable index must preserve the account's conditions of production, the institutional reasons for its marginalisation and its possible relation to other evidence. Briet helps articulate this shift from the supposedly intrinsic document to the made and governed document. The RawIndex operator becomes stronger, not weaker, when it accepts the constructive power of addressing. Its task is to keep this construction inspectable. The relevant unit is no longer simply the catalogue entry but the history of its making, including rejected names, unresolved conflicts and the possibility that the institution's original description was itself a source of documentary distortion.
Perec, Eco and the Problem of Attention
Georges Perec's attention to the infraordinary unsettles the assumption that an archive's most important objects are those already recognised as events. The unnoticed detail, the repeated routine, the ordinary change of use and the overlooked description of a street can reveal structures concealed by spectacular evidence. Perec's lists are not merely quirky alternatives to canonical history; they invite scrutiny of how a research apparatus distributes attention before anyone begins to interpret its findings. An index designed around exceptional architectural monuments, for example, may contain a complete record of official renovations and almost nothing about cleaning, waiting, informal repair or seasonal occupation. RawIndex should therefore include mechanisms for locating not only missing information but missing genres of observation. It becomes an instrument for examining the archive's own habits of notice.
Umberto Eco introduces a necessary resistance to the celebration of open interpretation. His work on the open work and the limits of interpretation helps distinguish a text's capacity to support multiple readings from the claim that any reading is equally sustainable. A RawIndex cannot be a device for endlessly multiplying labels without evidential discipline. If a fragment is provisionally identified as a window component, a later reading as a political symbol may be illuminating, but the interpretive movement must still state what warrants the transition. Socioplastics can learn from the disagreement between Perec's expansion of attention and Eco's concern for interpretive constraints. The field should increase what is available to be considered while sharpening the procedures through which claims about those materials can be challenged. The interesting question is not how to keep everything open, but how to keep open precisely those features that have not yet earned closure.
A Test for Accountable Provisionality
The operator can be evaluated through an experiment in competing classification. Begin with a bounded, mixed collection—say one hundred items from an abandoned school, including plans, invoices, children's drawings, interviews and discarded teaching materials. Create a RawIndex recording basic location, material form, known custody and unanswered questions. Invite three groups with different relationships to the site to propose descriptive categories: conservation specialists, former users and spatial researchers. Preserve each category's author, rationale, evidence and revision date. Allow each group to retrieve the collection under the other's scheme and record which questions become possible or impossible. Finally, compare what is learned through the provisional apparatus with what would have emerged from a conventional catalogue composed exclusively by the institution. This would not prove the superiority of raw indexing in every setting. It would test whether accountable provisionality reveals disagreements that matter to the investigation and whether it does so without obstructing retrieval.
Three failures would be especially instructive. First, provisional descriptors may become permanent by administrative inertia, reproducing exactly the exclusion they were designed to postpone. Second, the burden of maintaining rival terms may fall disproportionately on already marginal participants, turning openness into unpaid interpretive labour. Third, endlessly recording uncertainty may become a strategy for avoiding consequential decisions about custody, restitution or access. The point is not that classification should never settle. It is that the transition from exploratory address to authoritative category should itself become a documented event. RawIndex differs from a merely incomplete catalogue because it makes this threshold visible and contestable. A mature documentary practice needs not only an index of what is held but an index of how its own descriptions became consequential.
A Searchable Archive That Can Be Reclassified
Socioplastics faces the same problem at greater scale. Numbering roughly eleven thousand nodes makes individual records retrievable, but a stable number should not imply a definitive interpretation. A document may be cited under different research questions while retaining a traceable origin. The public index should therefore distinguish addresses from conceptual classifications and retain records of important revisions. Warburg offers a method of visual comparison; Otlet offers a model of distributed retrieval; Briet makes documentary status a question of institutional practice; Perec challenges what the collection notices; and Eco demands constraints on interpretation. These are different contributions, not components of a single theory. RawIndex supplies a modest procedure joining them: make the material findable, record how it was described, and preserve the possibility of correcting the description.
Their intersection yields a practical, contestable proposition: an index becomes an epistemic infrastructure when it records not only objects and their assigned classes but also the disputes, revisions and interpretive thresholds through which those classes acquire authority. RawIndex names the interval where this responsibility can be designed. Its future application could extend from artistic archives to urban planning records, environmental observation and participatory documentation, provided each setting specifies the costs of provisionality and the conditions under which a decision must be reached. The task is not to preserve an innocent beginning that never existed. It is to make the first acts of ordering available for subsequent criticism, so that the archive's next reader can change its questions without having to destroy its memory.
Briet, S. (2006) What Is Documentation? English Translation of the Classic French Text. Lanham, MD: Scarecrow Press. (Original French publication, 1951).
Eco, U. (1989) The Open Work. Cambridge, MA: Harvard University Press.
Eco, U. (1990) The Limits of Interpretation. Bloomington: Indiana University Press.
Otlet, P. (1990) International Organisation and Dissemination of Knowledge: Selected Essays of Paul Otlet. Edited and translated by W.B. Rayward. Amsterdam: Elsevier.
Perec, G. (1997) Species of Spaces and Other Pieces. Edited and translated by J. Sturrock. London: Penguin.
Warburg, A. (2000) Der Bilderatlas Mnemosyne. Edited by M. Warnke and C. Brink. Berlin: Akademie Verlag.
Anto Lloveras is a Spanish architect, conceptual artist and curator trained at ETSAM Madrid and TU Delft. Founder of LAPIEZA, he develops Socioplastics as a long-term research framework connecting archives, exhibitions and the organisation of knowledge. His work examines how documents and artistic practices acquire new relationships through indexing and circulation.