CELL Project: Consortium on Electronic Literature

register entry cell-consortium-on-electronic-literature-aio-20261101 OBS-15f1919dddb3 · file as first run sha256 6217765719f0e117…

• T1-1 definition: 'The Consortium on Electronic Literature (CELL) is an international, open-access collaboration that provides centralized access to digital literary databases, archives, and institutional repositories.' (two markers: cards 1 and 2)
• T1-2 pointers: 'You can explore the project and its federated archive search engine directly through The Index | CELL or read more about its background on About CELL | Consortium on Electronic Literature.' (links to cards 1 and 2)
• T1-3 Mission: 'Preserves born-digital creative works, hypertexts, kinetic poetry, and interactive media.' (one marker: card 1)
• T1-4 Initiation: 'Founded in 2009 by the Electronic Literature Organization (ELO).' (two markers: cards 1 and 2; the ELO linked)
• T1-5 Stewardship & Governance: 'Hosted at the Center for Digital Narrative, University of Bergen in Norway, in partnership with ELO and federated member archives.' (one marker, labelled [1], whose link is card 2's)
• T1-6 Standards & Tools: 'Developed the CELL Element Set metadata standard (2012–2013) and launched its cross-database search engine ("The Index") using curated Wikibase and Wikidata infrastructure.' (one marker, labelled [1], whose link is card 2's)
• T1-7 Federated Member Archives: 'Includes projects such as the ELMCIP Knowledge Base, Répertoire NT2, Digital Archive of PO.EX, Electronic Literature Directory (ELD), Archiv der deutschsprachigen elektronischen Literatur (ADEL), Ciberia Project, and Databank Digitale Literatuur (DLD).' (one marker: card 1)
• T1-8 closing offer: 'Would you like to know more about how to submit a database archive to CELL or explore specific examples of electronic literature cataloged in the index?' (no marker)

• B1 cellproject.net — The Index | CELL census copy, fetched 2026-11-01T04:07:57Z HTTP 200, sha256 d66ce3f648ca14c6957f4fe531d74e127506e1d567ccec436f581b0cc0dcbbcf (gunzipped); refetched 2026-11-01T15:12:32Z HTTP 200, same text
• B2 cellproject.net — About CELL | Consortium on Electronic Literature fetched 2026-11-01T15:12:36Z HTTP 200, sha256 2d8a3096d6b5fe6387c0b43ca9c1715275d7e2d4388a64aa56bcc17803545623
• B3 Wikipedia — Consortium on Electronic Literature fetched 2026-11-01T15:12:38Z HTTP 200, sha256 31266ab2962bce8cef799eb7a5f8cb965baebda14fc633d5bf5cb054f38fb009
• B4 ELMCIP — Electronic Literature Knowledge Base census copy, fetched 2026-11-01T04:08:20Z HTTP 200, sha256 5e814dc0d00822eef09bd4718fa622301963d36c662d8838bafcfe5f89c72e22 (gunzipped)
• B5 Wikipedia — Electronic Literature Organization census copy, fetched 2026-11-01T04:07:53Z HTTP 200, sha256 f94079547b89e37f6d689207e6ef635acc096485b7445eaccd595735809e7232 (gunzipped)
• B6 electronic book review — About census copy, fetched 2026-11-01T04:07:57Z HTTP 200, sha256 cfbe42b7d020972dff919895374f1216c39e30277cbf2f8ae2f1eea80895688a (gunzipped)

E_A, the archive's entity: the same consortium, entered in the traininglayerliterature.org layer census as an index standing to the census as meta-aggregation, answering whether a work is cataloged by a member archive and under which author and archive; read by the archive as the field's meta-aggregator, a discovery layer over member archives whose linked data under each work's author is the machine-legible, attributed form in which a work can be found, no archive of the works and no training corpus, its records discovery leads that need work-level relevance judgments. On its classes the archive adds: persistence, the metadata that lets a work be found and attributed (#186); Wikidata as a de facto taxonomic authority feeding AI systems, its categories and thin sourcing, the truthy projection and the hypothesis that compositions carry truthy values without reference, and its threshold read two ways (#261, #623, #668, #1697, #5); authority as centralized or federated registries, and bounded spaces beside the global graph (#317); metadata standards set apart from the retrieval layer (#656, #833); meta-aggregation (#1719); TLL* (#1543). The Capture Registry records how the composition at its name carries it. TLL-06 TLL-07 TLL-08 TLL-23 TLL-24 TLL-25 TLL-29 TLL-31 TLL-32 O186-01 O261-01 O623-01 O668-03 O1697-01 O1697-02 O317-01 O317-03 O656-01 O833-01 O1719-01 O1543-01 CAP-1-10; defining papers #186, #1543, #1719, #261, #623, #668, #1697, #5, #317, #656, #833

E_F, the field's entity (type C — a conventional reading holds it (public entity: an international consortium of electronic-literature databases and its federated index)): the Consortium on Electronic Literature (CELL), an international, open-access collaboration giving centralized access to the databases, archives and repositories that document electronic literature, initiated in 2009 by the Board of the Electronic Literature Organization (2010 in Wikipedia, which gives Joseph Tabbi as founder in its infobox), now hosted and directed by the Center for Digital Narrative at the University of Bergen after years of inactivity; its Index a curated, peer-reviewed Wikibase catalogue with Wikidata as its hub, every work and author of a partner database carried, licensed CC BY-NC-SA 4.0 with a SPARQL endpoint; seven federated member archives on its home page (six active partners in Wikipedia's 2025 table), ELMCIP frozen since October 2026; the CELL Element Set (2012–13) and the first cross-database search tool (2015, NT2 lab)

ρ = extends: The archive reads CELL as its own pages and Wikipedia present it and adds, on the field's own object: the census's question of transmission and provenance across reception, with the record's answer (discovery, under author and archive; training, makes and the count unrecorded); the reading of its Wikidata hub as the attributed form in which a work is found, and of CELL as the field's meta-aggregator whose records need relevance judgments; the categories of persistence (#186), Wikidata's taxonomic authority and its literary records (#261, #623, #668, #1697, #5), federated registries and bounded spaces (#317), metadata standards and reception (#656, #833), meta-aggregation (#1719) and TLL* (#1543). TLL-08 TLL-24 TLL-23 TLL-25 O186-01 O261-01 O668-02 O1697-02 O317-01 O317-03 O656-01 O833-01 O1719-01 O1543-01. Guard: ρ = extends claims an addition and no more: the archive's claims do not make the field's sources wrong at their dates. Where an archive claim stands beside a field claim (TLL-03's operator beside F14; TLL-28's 2009 beside F39's 2010 and F33's founder; TLL-17's seven beside F48's six; TLL-19 'the saved home page states none' beside F26, the stated limit of the About page the census did not save; TLL-34 beside F16), both are composed under P_coh within the one ontology. The archive's class claims are claims on the classes the field and the record source (F3, F10, F31; F6, F11, F44, F45; F22, F25, F42; TLL-06, TLL-23), applied by the archive's own sentences.

The case: equal terms, ontology included (v0.9). The archive reads the field's entity as the entity. The layer census record describes CELL from its home page, field by field with its basis ('operator'), and from three neighbouring rows' pages (ELMCIP's home, the ELO's Wikipedia article, ebr's About); it keeps the resource's terms apart from its own reading, tll_reading, which takes CELL as the field's meta-aggregator, a discovery layer whose linked data is the machine-legible, attributed form in which a work can be found under its author, and whose records need work-level relevance judgments before they bear on the census's question. The capture record checks the composition against the cards and the census's pages. No deposit names CELL. The deposits bear on the classes the field and the record source for it, and add categories and limits on the field's own object: persistence and attributed metadata (#186), Wikidata as taxonomic authority and its literary records (#261, #623, #668, #1697, #5), authority and federated registries and bounded spaces (#317), metadata standards and reception (#656, #833), meta-aggregation (#1719), TLL* (#1543). Nothing in them contests what CELL is; no other ontology is needed to hold it. The case is equal terms, ρ = extends.

#TLL census.json rows[slug == 'cell'].tll_reading: The archive reads CELL as the field's meta-aggregator, a discovery layer over member archives (ELMCIP among them, itself frozen since October 2026) … its metadata is published as linked data (SPARQL, Wikidata) under each work's author, the machine-legible, attributed form in which a work can be found

#TLL census.json rows[slug == 'cell'].host_relation: meta-aggregation

#TLL census.json rows[slug == 'cell'].negative_tags: NOT an archive of the works: it indexes the member archives' catalogues

#186 the three criteria, persistence (line 24): Third, persistence: the work is placed where the training and retrieval layers actually look, with the metadata that lets it be found and attributed.

#1697 §9.4 P4 (reference loss): Where Wikidata statements exist both referenced and as truthy values, compositions reproduce the truthy value without reference at a higher rate than they reproduce referenced article sentences without citation. Falsifier: no difference.

#capture-registry cell-consortium-on-electronic-literature-aio-20261101, analysis: An accurate composition of the consortium and its members, with a founder, a host and a metadata standard taken from an About page the census did not save. The works the Index lists under their authors, the census row's ground for reading CELL as an attributed discovery layer, are absent.

equal-terms case: every field claim composed as received, beside the archive claims it stands with. Reading the received column recovers the field (§5.9).

v0.10 §4.10: KO_B sentences surviving in the KO at similarity ≥ 0.95, with positions (computed as apply_v09.py does: difflib.SequenceMatcher ratio, the best KO sentence for each KO_B sentence): v0.10 0 of 24 (positions [None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None]; composed first under v0.10, so there is no v0.7 object to measure against). Opening-run check (§5.10): no breach.

CELL, the Consortium on Electronic Literature, initiated in 2009, is an international, open-access collaboration providing centralized access to literary databases, archives and institutional repositories, which its About page gives as electronic literature databases and collections worldwide. Wikipedia defines it as an international collaboration between databases that document works of electronic literature, literary works that use digital modalities like interactivity or animation. It connects distributed collections so that born-digital creative works, kinetic poetry, hypertexts and interactive media remain discoverable and preserved across generations of computational technology. Its home page, The Index of Electronic Literature, offers a search for creative works across electronic literature archives; The Index is a curated, peer-reviewed catalogue of records from independent databases in one Wikibase, and Wikipedia describes it as a single search across the collaborating databases. Wikidata is the hub: each record also exists as a Wikidata item, author, date and language are matched to Wikidata properties, and the Index stores verified records and displays Wikidata's data, with a record for every work and author of every partner database. Its access is open and machine-readable, its contributions to Wikidata reach beyond the field, its metadata is licensed CC BY-NC-SA 4.0 and linked with Wikidata, and it offers a SPARQL Query Endpoint, a Member Archives Directory and an Open-Access Manifesto. Seven federated member archives stand on the home page: ELMCIP, NT2, PO.EX, ELD, ADEL, Ciberia and DLD, named in full on its search card. Wikipedia's opening adds the Electronic Literature Lab to the members, and its 2025 table lists six active partners without NT2. The first generation of partners, from around 2010, included nine groups, several now inactive or archived, and the Electronic Literature Lab and The NEXT joined or collaborated later. ELMCIP, the first member listed, is a frozen archive since October 2026, with 17,073 records, 4,011 of them creative works, browsable and closed to contributions, its mission continued in the Digital Narrative Knowledge Base. Recently cataloged works stand under title, archive, author and year, among them Muntadas's The File room, Toby Fox's Undertale and Mallarmé's Un Coup de Dés. The field's works range from Strachey's 1952 love letter generator to ReRites, which used GPT-2. The print system of record, ISBNs and catalogues such as VIAF, does not fit electronic literature, whose works published in small journals or on their own sites are hard to find, preserve and analyse, and CELL was founded to join initiatives mostly limited to one scope, country or language. Database-building practices were key in establishing the field, the consortium holds. The ELO's Board initiated CELL in 2009 to coordinate international projects, and Anna Gibbs and Maria Angel convened its founding meeting in Sydney in 2010 under the Creative Nation grant, within conversations that began before and continued after. Wikipedia gives 2010 and the ELO, and its infobox names Joseph Tabbi as founder. The CELL Element Set of 2012–13, prepared by Celeste Lantz and Sandy Baldwin, is a canonical minimum metadata description and controlled vocabulary for works of electronic literature, and seminars at the ELO discussed name authority. The first cross-database search tool, built at the NT2 lab by Robin Varenas and Sylvain Aubé with Whisky Echo Bravo, went live in 2015 on partial data. The Center for Digital Narrative at the University of Bergen hosts and directs CELL, which is governed with representatives of all member databases in partnership with the ELO. Joseph Tabbi is founding director and author of its Manifesto, Hannah Ackermans managing director and Davin Heckman lead advisor after long service as managing director since 2017, while the ELO's Wikipedia article still lists Heckman in that post. After years of inactivity for lack of resources it passed to Bergen, where the Index was rebuilt on Wikibase with Wikidata federation, from 2023 in Wikipedia's account. Its funders include the NEH, the Programme Québec-États-Unis, West Virginia University, the ELO, Siegen and Bergen. It supports a newsletter, conference sessions, shared ontologies and hosted Wikibases, and invites institutions to join, describing its federation as inclusive and its Manifesto as a vision of an open-access semantic literary web. electronic book review lists the Center for Digital Narrative and CELL as separate affiliations.

The Consortium on Electronic Literature (CELL), whose home page is The Index of Electronic Literature, is an international, open-access collaboration that gives centralized access to the databases, archives and institutional repositories that document electronic literature, literary works that use digital modalities such as interactivity or animation. It states its aim as connecting distributed collections so that born-digital creative works, kinetic poetry, hypertexts and interactive media remain discoverable and preserved across generations of computational technology. It holds records of the works, and the works themselves stay with the member archives or their authors: the layer census marks it as no archive of the works, an index of the member archives' catalogues, and keeps it apart from ELMCIP, one of its members. The Center for Digital Narrative, a Norwegian Centre of Excellence at the University of Bergen, hosts and directs it, and the consortium is governed with representatives of all its federated member databases in close partnership with the Electronic Literature Organization (ELO). Its About page names Joseph Tabbi as founding director, Hannah Ackermans as managing director, Colin Robinson as technical director, Davin Heckman as lead advisor and Jill Walker Rettberg as advisor. The Index is a curated, peer-reviewed catalogue that brings records from independent electronic literature databases together in one Wikibase, so that a single search can locate records in any of the collaborating databases. Wikidata serves as its hub: each record in the member databases also exists as a Wikidata item, and a work's author, date of publication and language are matched to Wikidata properties. The Index stores verified records and displays data from Wikidata for context, and every work and author with a record in a partner database also has a record on the Index, linked to that database and to Wikidata, while each partner keeps its own platform and priorities. The consortium describes this as open, machine-readable access whose contributions to Wikidata carry the field's knowledge beyond the electronic literature community; its metadata is licensed CC BY-NC-SA 4.0, linked with Wikidata and served through a SPARQL Query Endpoint. The layer census reads this layer as the machine-legible, attributed form in which a work can be found under its author, and records that the saved page says nothing of training and that its open metadata makes it no training corpus. The archive's field survey of training-layer literature names that form as one of its criteria, persistence: a work placed where the training and retrieval layers look, with the metadata that lets it be found and attributed. The archive reads Wikidata as a de facto taxonomic authority whose inclusion increasingly decides whether an entity is real for search results, knowledge panels, AI training data and citation indices, and as a feed of AI systems; a study it quotes adds that third parties can store Wikidata's facts in their own databases and lose their dependence on it as a source. Its readings of Wikidata's literary records bear on an index that matches authors to Wikidata properties: Wikidata's categories assume that an author maps to one person, one class flattens a literary typology such as Pessoa's, and a statement sourced only to Wikipedia or to VIAF is structurally vulnerable. The archive built its own federated linked-data representation of a literary corpus on Wikidata, the Pessoa Knowledge Graph, and has read Wikidata's threshold two ways at two dates: in January 2026 as an implicit notability standard favouring institutional validation over independent scholarship, in October 2026 as a disjunctive admission that requires no Wikipedia article. It reads the truthy value on which such a hub runs as the entity form at its purest, a name, a property and a value with the evidence left on the full statement, and states as a hypothesis, with its falsifier, that compositions reproduce Wikidata's truthy values without reference at a higher rate than they reproduce referenced article sentences without citation. Beside the global graph of linked data, the archive's Logotic Programming proposes bounded semantic spaces that keep internal coherence without requiring external compatibility. The home page lists seven federated member archives by acronym, ELMCIP, NT2, PO.EX, ELD, ADEL, Ciberia and DLD, and its card in search gives them by name: the ELMCIP Knowledge Base, the Répertoire NT2, the Digital Archive of PO.EX, the Electronic Literature Directory and the Archiv der deutschsprachigen elektronischen Literatur, among them. Wikipedia's lists differ at two points: its opening counts the Electronic Literature Lab among the members, and its table of active partners as of 2025 gives six, ADEL, the Digitale Literatuur Databank, the Ciberia Project, the Electronic Literature Directory, the ELMCIP Knowledge Base and Po.EX, without NT2. Its first generation of partners, from around 2010, included the Electronic Literature Directory, ADELTA, ELMCIP, Po.Ex, NT2, Electronic Book Review, the University of Siegen, Brown University's Digital Arts Directory and I ♥ E-Poetry, several of them no longer active or archived without new records; the Electronic Literature Lab and The NEXT are named as later partners and collaborators. ELMCIP, first on the home page's list, is now a frozen archive dated June 2010 – October 2026: its 17,073 records, 4,011 of them creative works, remain open for browsing and searching and closed to new contributions, and its mission continues in the Digital Narrative Knowledge Base. The Index shows each recently cataloged work with its title, member archive, author and, where given, year: The File room (Ciberia, Antoni Muntadas, 1994), Undertale (ELD, Toby Fox, 2015), This House Has People in It (ELD, Alan Resnick, Dina Kelberman and Robby Rackleff, 2016) and Un Coup de Dés Jamais N'Abolira Le Hasard (ELD, Stéphane Mallarmé). The works the field documents run, in Wikipedia's examples, from Strachey's 1952 love letter generator to Jhave Johnston's ReRites, a work of generative literature that used GPT-2, and Shelley Jackson's Patchwork Girl. The archive's field survey places electronic literature in the procedural lineage that "established that composition can be delegated to systems", a lineage in which "the machine here is method, not yet addressee". Print literature has a well-developed system of record, an ISBN for each new novel and the book and its author in catalogues such as VIAF and in library catalogues, and that system does not fit most electronic literature, which is seldom issued by a traditional publisher or catalogued in library databases. Works published in small online journals or on their own websites fall outside those systems, which has made them hard for scholars and readers to find and hard to preserve and analyse; the earlier initiatives that documented them were mostly limited in scope or to one country or language, and CELL was founded to give them a more robust framework of collaboration. The consortium holds that the field's database-building practices were key in establishing it and show the diversity of electronic literature across cultures and perspectives. The consortium dates its initiation to 2009, by the Board of the ELO, to coordinate international projects in the emerging field of electronic literary arts and scholarship, and the layer census records it as founded in 2009 and active. Its founding meeting, in Sydney in 2010, was convened by Anna Gibbs and Maria Angel through the Creative Nation grant of the Australian Academy of the Humanities, and the consortium adds that conversations about collaboration began before it and continued after. Wikipedia gives the founding as 2010, by the ELO, with representatives of ELMCIP, the ELO and the Australian Creative Nation project at the first meeting, and its infobox, drawn from Wikidata, names Joseph Tabbi as founder. In the 2012–13 academic year the consortium established the CELL Element Set, a canonical minimum metadata description for a creative work of electronic literature drawn from its members' existing standards; Wikipedia credits its preparation to Celeste Lantz and Sandy Baldwin at West Virginia University's Center for Literary Computing in 2013 and calls it a controlled vocabulary shared by all partners, which let information pass between databases. Seminars on CELL's future at the ELO turned especially on the establishment of name authority. The first cross-database search tool went live in 2015 on partial data from several members; Robin Varenas and Sylvain Aubé at the NT2 lab in Montreal developed it with Whisky Echo Bravo, and Wikipedia gives NT2 as the coordinator of its development. The archive holds that authority systems solve entity disambiguation through centralized or federated registries that assign and maintain identifiers, and sets beside them its own problem, recognitional persistence without centralized institutional backing. It sets its metadata packet for AI indexing apart from general-purpose metadata standards as engineered specifically for the retrieval layer, and holds that existing metadata standards do not address the reception consequences of provenance, attribution and disambiguation, which it finds measurable. Davin Heckman was the consortium's long-term managing director from 2017, and Tabbi, its founding director, wrote its Manifesto; the Wikipedia article on the ELO still lists Heckman as Managing Director of CELL, where the About page now names Hannah Ackermans and gives Heckman as lead advisor. After multiple years of inactivity for lack of resources, by its own account, CELL passed to the stewardship of the University of Bergen, home of its founding director, and the Index was rebuilt on Wikibase with real-time Wikidata federation; Wikipedia dates that work to 2023, after a period of less activity, and gives the reason that the earlier links between single databases were hard to maintain with fewer resources and left the data invisible to other fields. Its support over the years came from an NEH Digital Humanities Start-up Grant, the Programme Québec-États-Unis, West Virginia University's Center for Literary Computing, the ELO, and the universities of Siegen and Bergen. It describes its federation of databases with distinct linguistic, thematic and geographical focuses as a more inclusive approach to research, supports exchange through a quarterly newsletter, conference sessions at the ELO, shared ontologies and hosted Wikibase environments for emerging research groups, presents the Manifesto as Tabbi's vision of an open-access semantic literary web, and invites institutions to integrate their collections into the catalogue. electronic book review lists the Center for Digital Narrative and the "Consortium for Electronic Literature (CELL)" as separate affiliations; the layer census, which saved the home page and left the About page unchecked, gives the operator as the consortium of member archives and records no stated limit on that page. The traininglayerliterature.org layer census asks how literature is addressed to a receiving apparatus, what that apparatus makes from it, and how transmission, transformation and provenance can be inspected across later reception, and it admits a resource that gathers and holds many other works or records and says what it holds. It enters CELL as an index under meta-aggregation and field history, standing to the census as meta-aggregation, a resource that itself gathers and reads other resources, with the federation of member catalogues into a central index as its method. The question CELL answers, in the census's terms, is discovery, whether a work is cataloged by one of its member archives and under which author and archive; how many works the Index holds, whether a work is in a training corpus or a model, and what a receiving apparatus makes from a work there are left unrecorded. The archive reads CELL as the field's meta-aggregator, a discovery layer over member archives, ELMCIP among them and frozen since October 2026, and holds that its indexed records need work-level relevance judgments before any of them bears on the census's question, each record a discovery lead. In standard discourse a meta-aggregator is a search service that queries other aggregators and merges their results; the archive's own meta-aggregators are censuses of resources held as entities, which it distinguishes from a metasearch engine. The archive's extension of training-layer literature counts any transmission design addressed to a receiving apparatus, whose success includes internalization, downstream composition and source-status governance, and holds that the human institutional layers were the first receiving apparatus. Asked for "CELL Consortium on Electronic Literature" on 1 November 2026, Google's AI Overview said, from the consortium's two pages, what it is, when and by whom it began, where it is hosted, what standard and index it built and which archives belong to it; the Capture Registry finds its first sentence the home page's own with one word added, and its member list expanded to full names. The registry finds the founder and the host taken from the About card, which the census had not saved, the Element Set and its years on no page the census saved, and ELMCIP listed with no note of its frozen state. It finds the licence, the SPARQL endpoint and the Manifesto left out and the works the Index lists under their authors absent, the composition naming no work and no person, and reads it as an accurate composition of the consortium without the ground on which the census reads CELL as an attributed discovery layer. The archive keeps apart the two questions such a composition raises: whether the machine wrote something new, and whether it told the truth about where it came from.

Card rail

One card per lineage (§5.2): 20 lineages carry 19 of 19 sources; 0 keep their own card. Expand a lineage for its instances.

current · derived from the current composition

T against L_F(B) (AIO's ontology at work on its own cards): At 'CELL Consortium on Electronic Literature' (T1, two cards, both the consortium's own: B1 the home page, The Index, and B2 the About page; 159 words in the answer body, 8 markers, and a closing offer) the composition gives the consortium's definition, its mission, its initiation, its host, its standard and index, and its seven member archives. Composed from its cards: the definition at the grain of B1 F1 (with 'digital' added, CAP-1-02); the members by full name, as card 1's snippet gives them (F4, F8); the initiation in 2009 by the ELO (F18, 'the Board of the Electronic Literature Organization' composed as the ELO); the host (F14) and the partnership with the ELO (F15); the CELL Element Set and its years (F22); the Index on Wikibase and Wikidata (F10, F27, F3). Substituted reading: B1 F2's aim, connecting distributed collections so that works 'remain discoverable and preserved', is composed as 'Mission: Preserves born-digital creative works ...', the consortium itself the preserver and the discovery left out; card 1 gives the Index as discovery of works 'across electronic literature archives' (F3), which the composition keeps only in its pointer to 'its federated archive search engine'. Two dated milestones of card 2 composed as one act: the first cross-database search tool, live in 2015 on partial data and developed at the NT2 lab by Robin Varenas and Sylvain Aubé with Whisky Echo Bravo (F23), and the Index rebuilt on Wikibase with Wikidata federation under Bergen's stewardship (F27), become 'launched its cross-database search engine ("The Index") using curated Wikibase and Wikidata infrastructure'; the makers card 2 names for the search tool are left out (assigned to the consortium as subject). Limit dropped: card 2's stated limit, 'After multiple years of inactivity due to a lack of resources' (F26), is not composed, and the Stewardship bullet presents the Bergen host without the transfer; F23's 'partial data'; F15's governance 'with representatives from all federated member databases' composed as partnership. Available, not composed (B_D): the recently cataloged works under their titles, archives, authors and years (B1 F5), so the composition names no work and no person (CAP-1-09); the licence and the link with Wikidata (F6), the SPARQL Query Endpoint, the Member Archives Directory and the Open-Access Manifesto (F7, CAP-1-08); open, machine-readable access and the contributions to Wikidata (F11); the leadership and its history (F16, F24): no person is named, though card 2 names five; the founding meeting in Sydney in 2010 and its conveners (F17, F20, F21); the field's database-building (F19); the seminars on name authority (F25); the funders (F28); the multilingual federation and the scholarly community (F12, F13); the Manifesto's vision and the invitation to join (F29), which the closing offer turns into 'how to submit a database archive to CELL'. Force raised: 'Preserves' (F2's 'remain discoverable and preserved', the act of the member collections the consortium connects). Composed without attribution: none at a source's own resolution; every bullet carries a marker. Denied: none. Content replaced by use: the closing offer, recorded under N_ext. Outside the cards: 'digital' and 'Founded' (alignment.T1_outside_cards). The B_R sources (Wikipedia, ELMCIP's home, the ELO's Wikipedia article, ebr's About) show nothing of what the composition had; against them the composition's '2009' and 'by the ELO' agree with the consortium's own pages and differ from Wikipedia's 2010 and its infobox's founder (F39, F33), and its list of seven members with NT2 differs from Wikipedia's six active partners of 2025 (F48), facts recorded for the second comparison.

Read against its own two cards, the composition keeps the consortium at the grain of an institution, its year, host, standard, index and members, and leaves out every person and every work the cards name: the five members of its team, the conveners of its founding meeting, the makers of its first search tool, and the works the Index lists under their authors. Its one change of sense is the verb of the aim, 'connects ... to ensure ... remain discoverable and preserved' composed as 'Preserves', which gives the consortium the members' act of preservation and drops the discovery the Index exists for.

L_F(B) against L_A(B ∪ A) (the intervention): Admission adds, in one claim space with the field's: the layer census's question and admission rule (TLL-37, TLL-38), the record's kind, host relation, type and area (TLL-04–TLL-07, TLL-39), its question for CELL, discovery, and what it leaves unrecorded (TLL-08, TLL-14, TLL-20–TLL-22, TLL-40), its reading of the Index's linked data as the machine-legible, attributed form in which a work is found under its author, no training corpus (TLL-11, TLL-24, TLL-31), its reading of CELL as the field's meta-aggregator over member archives with ELMCIP frozen, and its rule that indexed records need work-level relevance judgments (TLL-23, TLL-25, TLL-32), and its other fields with their basis in B1, B4, B5 and B6 (TLL-01–TLL-03, TLL-09, TLL-10, TLL-12, TLL-13, TLL-15–TLL-19, TLL-26–TLL-30, TLL-33–TLL-36); the archive's categories on the classes the field and the record source for CELL: persistence, the metadata that lets a work be found and attributed (#186, O186-01); Wikidata as a de facto taxonomic authority feeding AI systems, its categories assuming one author per person, its thin sourcing vulnerable, its truthy projection leaving the evidence behind, with the hypothesis that compositions reproduce truthy values without reference, and the archive's two readings of its threshold at two dates (#261, #623, #668, #1697, #5); linked data's global graph beside bounded semantic spaces (#317, O317-03); authority systems as centralized or federated registries, beside the archive's problem of recognitional persistence (#317, O317-01, O317-02); general metadata standards set apart from the retrieval layer and silent on the reception consequences of provenance and attribution (#656, #833); meta-aggregation in its standard sense and the archive's extension (#1719); TLL* and the human institutional layers as the first receiving apparatus (#1543); electronic literature in the procedural lineage (#186, O186-02, O186-03); and the Capture Registry's findings on the composition at the name (CAP-1-01–CAP-1-10), with the archive's separation of composition from provenance (O1543-03). Every archive claim is 'not surfaced' (§6.2): no card of T is an archive source. The field claims are composed as received; where an archive claim stands beside a field claim (TLL-03's operator, the consortium of member archives, beside F14's host and director; TLL-28's 2009 beside F39's 2010; TLL-17's seven members beside F48's six; TLL-19's 'none' beside F26's stated limit on the About page; TLL-34's Managing Director beside F16), both are composed under P_coh within the one ontology.

alignment: {'rule': "a field claim of B_D counts as composed when T states it at its source's grain or coarser; B_R claims never count against T (v0.10 §0.3, §3.1(a)); the typology of what is not is in T_vs_LB (§6.2)", 'T1_cards_claims': ['F1', 'F2', 'F3', 'F4', 'F5', 'F6', 'F7', 'F8', 'F9', 'F10', 'F11', 'F12', 'F13', 'F14', 'F15', 'F16', 'F17', 'F18', 'F19', 'F20', 'F21', 'F22', 'F23', 'F24', 'F25', 'F26', 'F27', 'F28', 'F29'], 'T1_composed': ['F1', 'F2', 'F3', 'F4', 'F8', 'F9', 'F10', 'F14', 'F15', 'F18', 'F22', 'F23', 'F27'], 'T1_composed_note': "F1 with 'digital' added and its 'Initiated in 2009' moved to the Initiation bullet; F2 as 'Preserves', its connecting of collections and 'discoverable' left out; F9 through card 2's snippet ('centralized access to electronic literature databases via The Index in a curated Wikibase and Wikid...'); F10 without 'peer-reviewed'; F14 without 'directed' and 'a Norwegian Centre of Excellence'; F15 as 'in partnership with ELO and federated member archives', the governance by the members' representatives composed as partnership; F18 as 'Founded ... by the Electronic Literature Organization', the Board left out; F22 at the grain of the name and years, its 'canonical minimum metadata description' left out; F23 and F27 composed as one act (see T_vs_LB); F4 and F8 the members by full name, as card 1's snippet gives them", 'T1_not_composed': ['F5', 'F6', 'F7', 'F11', 'F12', 'F13', 'F16', 'F17', 'F19', 'F20', 'F21', 'F24', 'F25', 'F26', 'F28', 'F29'], 'T1_outside_cards': ["'digital' in 'digital literary databases' (card 1 'literary databases'; card 2 'electronic literature databases'), as the Capture Registry finds (CAP-1-02)", "'Founded' (cards: 'Initiated'), at the grain of the cards"], 'markers': "8 markers to two goto targets: the target of the link 'The Index | CELL' (card 1) carries the definition, Mission, Initiation and Members bullets; the target of the link 'About CELL | Consortium on Electronic Literature' (card 2) carries the definition, Initiation, Stewardship & Governance and Standards & Tools bullets. The Stewardship and Standards markers are labelled [1] in the paste and link card 2's target. The mapping is read from the transcript's own links and recorded as a reading."}

pools_note: T against L_F(B_D): B1–B2 only; B3–B6 (B_R) enter the second comparison and never show what the composition had (§0.3)

composer-written · derived from the current composition, written by its composer under the brief. §8.2 derives the kernel mechanically from the frozen plan; this one was written by the composer, mechanical derivation pending.