Virtual International Authority File (VIAF)

register entry viaf-aio-20261101 OBS-34cceb53f2ee · file as first run sha256 d0ad593cc0d7b296…

• T1-1 definition: 'is an international name authority service that combines multiple national library authority files into a single online database.' (two markers)
• T1-2 operator: 'Operator: Hosted and maintained by OCLC (Online Computer Library Center).' (two markers)
• T1-3 origin: 'Origin: Initiated in 1998 as a joint proof-of-concept by the U.S. Library of Congress, the German National Library (DNB), and OCLC, launching formally in August 2003.' (one marker)
• T1-4 purpose: 'Purpose: Solves author name disambiguation by linking variant names, pseudonyms, and corporate titles across different languages, scripts, and regional spelling rules.' (three markers)
• T1-5 integration: 'Integration: Each cluster receives a unique standard identifier/URI that is widely used in library catalogs, Wikipedia biographical articles, and Wikidata.' (two markers)
• T1-6 access: 'Access: The service and its APIs are publicly available via the VIAF Website.' (two markers)
• T1-7 closing offer: 'Would you like help searching for a specific author or entity in VIAF, or learning how to access its RDF/JSON data?' (no marker)

• B1 VIAF — VIAF unreachable: https://viaf.org/ HTTP 403 (Cloudflare block) at 2026-11-01T06:14:47Z, as the census's fetch at 2026-11-01T05:32:47Z; card title and snippet only
• B2 OCLC — VIAF: Convenient access to name authority files | OCLC card; B2.html sha256 fc02350aa6a54164, HTTP 200 at 2026-11-01T06:14:47Z
• B3 Wikipedia — Virtual International Authority File - Wikipedia card; B3.html sha256 10288c58ab5b9929, HTTP 200 at 2026-11-01T06:14:53Z; wikitext HTTP 200 (sha256 fd2bc9158af25086)
• B4 OCLC — VIAF (The Virtual International Authority File) card; B4.html sha256 980d3a7f49c45ff0, HTTP 200 at 2026-11-01T06:14:48Z
• B5 The Linked Open Data Cloud — LOD Cloud card; B5.html sha256 06ddf35a9abe8026, HTTP 200 at 2026-11-01T06:14:48Z; the page a script template, its record read from /json/viaf (HTTP 200, sha256 ef4a2030346d37b5)
• B6 Semantic Web Journal — Description of the VIAF (Virtual International Authority File) Dataset card; B6.pdf sha256 6aea26e7edced5b2, HTTP 200 at 2026-11-01T06:15:39Z
• B7 Wikipedia — International Standard Name Identifier (census source) census third party; B7.html sha256 24575f41df384288, HTTP 200 at 2026-11-01T05:32:47Z; the census's saved copy, sha256 matching the manifest

E_A, the archive's entity: the same file, in the leesharks.com authorship census record and the archive's deposits: library-led identity resolution that links authority records judged identical into one cluster and keeps no works, infrastructure to the question of how authorship is named and attributed; whether it holds a pseudonym or heteronym as a form of its bearer's cluster or as a cluster of its own is unshown, and its coalescing clusters are the point where an authorial name can be merged by algorithm, the collapse #940 names, or separated again; #940 classes the Clemens/Twain pair as pseudonymy and distinguishes the heteronym, attributable through the civil bearer without being reducible to that bearer; #668 holds that heteronyms often have independent authority records; #317 describes the file as making stable identifiers by institutional coordination, authority control assigning identity from above LS-22 LS-23 LS-12 LS-21 LS-14 LS-11 LS-04 O940-01 O940-02 O940-04 O940-06 O668-02 O317-01 O317-03 O261-01 O1722-01; defining papers #940, #1722, #317, #668, #261

E_F, the field's entity (type C — a conventional reading holds it (public entity: a library name-authority service)): the Virtual International Authority File, an international authority file, a joint project of several national libraries operated by OCLC (an OCLC service since 4 April 2012), begun in 1998 by the Library of Congress, the Deutsche Nationalbibliothek and OCLC, that links existing national and other library authority files into a single virtual file of clusters, each with a URI, the see and see-also forms of its source records kept and regional forms coexisting, rebuilt monthly so that clusters may coalesce or split and some identifiers fluctuate; led by libraries; its builders list conservative matching, limited correction from outside libraries and files that differ on Clemens and Twain among its shortcomings

ρ = extends: The archive reads VIAF as its cards and Wikipedia present it and adds, on the same object, the authorship question: how a chosen name is held by a cluster (LS-12, LS-21), the merging of authorial names by algorithm (LS-23), the pseudonym/heteronym distinction and civil-name reduction (#940), heteronyms' separate authority records (#668), category collapse (#261), a heteronym as a real author's practice (#1722), library leadership as identity resolution with nothing on authorial function (LS-22, LS-14). One archive claim contradicts a field claim: #317's 'stable identifiers' against F24's fluctuation; both are composed (P_coh) with F50's redirects. LS-12 LS-21 LS-23 LS-22 LS-14 O940-01 O940-02 O940-04 O940-06 O668-02 O261-01 O1722-01 O317-01. 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 #317 contradicts F24 both stand with their sources, and the census's unanswered question stays a question (open block).

The case: equal terms, ontology included (v0.9). The archive reads the field's entity as the entity. The census record describes VIAF from the Wikipedia article and keeps the resource's terms (operator, kind, unit, record object, method) apart from its own reading (ls_reading), which asks of the same file how a chosen name is held and reads its coalescing clusters as the point where an authorial name can be merged, 'the collapse #940 names'. #940 and #1722 supply the pseudonym/heteronym distinction and the claim that a heteronym is a real author's practice; #668 adds that heteronyms often hold separate authority records; #317 describes the same file as making stable identifiers. None of these reads VIAF as made by reception or contests that it is the OCLC-hosted linkage of library authority files: they add senses (how a chosen name is held, civil-name reduction, library leadership as a question of authorship) on the field's object. ρ = extends; the equal-terms case.

#LS census.json rows[slug == 'viaf'].ls_reading: VIAF, as Wikipedia describes it, resolves a name across library authority files by linking records judged identical into one cluster, with see and see-also forms carried from the source records; that is identity resolution led by libraries, and the saved source does not show whether a pseudonym or heteronym stays a distinct cluster or folds into its bearer's.

#LS census.json rows[slug == 'viaf'].ls_reading: Clusters that 'coalesce or split' mean an authorial name can be merged with another by algorithm, the collapse #940 names, or separated again; which happens to a given name is not recorded here.

#940 §2 ¶2 (line 99): A pseudonym changes the label attached to an authorial function. A heteronym changes the organization of the authorial function itself.

#668 External identifiers (line 196): P214 (VIAF), P244 (Library of Congress), P268 (BnF), P646 (Freebase), P1207 (NUKAT), and others where they exist. Heteronyms often have independent authority-control records because library systems cataloged them separately before the heteronymic nature of the authorship was widely understood.

#317 §3.1.1 Authority Control and Identity Management ¶1 (line 193): The Virtual International Authority File (VIAF) aggregates national authority files to create stable identifiers for persons, organizations, and works.

equal-terms case: every field claim composed as received, beside the archive claims it stands with; no receiving text. Reading the received column recovers the field (§5.9).

v0.9 §4.10 (unchanged in v0.10): 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): v0.9 0 of 13 (positions [None, None, None, None, None, None, None, None, None, None, None, None, None]; composed first under v0.9, so there is no v0.7 object to measure against). Opening-run check (§5.10): no breach.

The Virtual International Authority File (VIAF) is an international authority file, a joint project of several national libraries operated by OCLC, that combines library name authority files into a single OCLC-hosted name authority service. Contributors supply authority data that VIAF matches, links and groups into clusters, one per entity with its different names brought together, each cluster assigned a URI; it covers persons, organizations, places, works and expressions whose names libraries control. A VIAF record has a standard data number, carries the see and see-also records of the originals and refers to the original authority records, and national variations in authorized form coexist within it. After failed attempts at one joint authority file, the plan became to link existing national authorities into a single virtual file, toward lower cost and greater utility of library authority files. The Library of Congress, the Deutsche Nationalbibliothek and OCLC began it in April 1998, formed the consortium in August 2003, and the BnF joined in October 2007. It became an OCLC service on 4 April 2012; OCLC maintains it with the advice of the VIAF Council, the Principals hold key decisions, and OCLC Research's project page is kept for historical purposes. Its participants number 21 files from 18 organizations (2011), 22 in 18 countries (2012), more than 50 organizations from more than 30 countries (now), or 20 national libraries (the LOD Cloud), with ISNI, the Getty ULAN and Wikidata among them. The clustering runs monthly; clusters may coalesce or split, so some identifiers fluctuate, while record IDs are kept with redirects when clusters merge. Its builders list conservative matching, a limited ability for those outside the library community to correct errors, and files that differ on whether Samuel Clemens and Mark Twain are one entity or two. Borgman groups it with ISNI and ORCID as efforts to standardize name forms and notes that it is led by libraries, as opposed to individual authors or creators. ISNI, built mainly on VIAF, gives each pseudonym its own ISNI. Its data are open for research and sharing through a website, an API, bulk downloads and OAI-PMH, under ODC-By, and its numbers are carried into Wikipedia and Wikidata. Wikipedia's article was last edited on 19 June 2026; the ISNI article tags a viaf.org URL as a dead link.

The Virtual International Authority File (VIAF) is an international authority file, a joint project of several national libraries operated by OCLC, that combines multiple library name authority files into a single OCLC-hosted name authority service. Its contributors supply authority data that VIAF matches, links and groups; all the descriptions of one entity are merged into a cluster that brings the entity's different names together, and each cluster is assigned a URI. The entities it covers are those whose names libraries control: people, corporations and places, and works and expressions in the FRBR model; its clusters are chiefly of persons and organizations. In the cluster's data model the emphasis falls on recognizing, identifying and relating "the thing" that the contributors are trying to name, a model that passed from SKOS concepts through FOAF to foaf:focus. The file was designed as a linkage: after several failed attempts to develop one joint authority file, the plan became to link the existing national authorities into a single virtual authority file in which identical records from the different data sets are linked together. A VIAF record receives a standard data number, carries the primary "see" and "see also" records of the original documents and refers back to the original authority records; an example identifier is 106965171, and a VIAF URI has the canonical form http://viaf.org/viaf/[numerical value], which OCLC encourages for use in all appropriate circumstances. National and regional variations in authorized form coexist within a cluster, with regional preferences of language, spelling and script preserved, so that German users see a name in the form the DNB established, French users the form the BnF established and English-speaking users the form of the LC/NACO file. By linking disparate names for the same person or organization, it lets libraries serving different language communities repurpose one another's bibliographic data, toward its stated goal of lowering the cost and increasing the utility of library authority files. The Library of Congress, the Deutsche Nationalbibliothek and OCLC began it in April 1998 as a proof of concept that authority records can be linked; after four years of testing the three formed the VIAF Consortium at the 2003 IFLA congress, on 6 August 2003, and the Bibliothèque nationale de France joined in October 2007. On 4 April 2012 it passed from an OCLC Research prototype to an OCLC service, and OCLC Research's own page for the project now states that the research has concluded and that the page is kept "for historical purposes only". OCLC maintains its hosting, software and data with the advice of the VIAF Council, through which the contributors take part in governance; the Cooperative Agreement the research page records places all key decisions with the Principals, and agencies that wish to participate must apply. Counts of its participants differ by date and by source: 21 authority files from 18 organizations in the winter of 2011 on the research page, 22 principals and contributors in 18 countries at the 2012 transition and more than 50 organizations from more than 30 countries on OCLC's current page, and "a joint project of 20 national libraries" in the Linked Open Data Cloud's record. Wikipedia's table of participants lists national libraries beside ISNI, the Perseus Project, the Getty Union List of Artist Names and Wikidata, the last marked "citation needed". In 2012 its builders described a dataset built from nearly two dozen files of 25 million authority records and 110 million bibliographic records, forming some 20 million clusters, and the Linked Open Data Cloud's record gives 200 million triples; no current count stands in the census record's saved source. Source files and their bibliographic metadata are sent to or harvested by OCLC, usually monthly; the records go through a matching process documented down to its appendices of noise titles, occupation codes and name synonyms, ambiguities are resolved, clusters are built from the matching records, and the file is rebuilt each month. Because the clustering algorithm runs every month, clusters may coalesce or split as data are added, and the identifier of certain authority records fluctuates; record IDs are maintained across updates with redirects when clusters are merged, and a change made in a source file can take more than a month to show. An earlier archive paper describes VIAF as aggregating national authority files "to create stable identifiers for persons, organizations, and works", one of the registries that solve entity disambiguation by assigning and maintaining identifiers; that description and the record of fluctuating identifiers stand together, and the redirects bear on both. Hickey and Young list among its known shortcomings that differences in rules and conventions cause difficulties, the common example being one authority file with entities for both Samuel Clemens and Mark Twain where another has only Mark Twain, and that matching is fairly conservative, so many possible matches are missed. The archive classes Clemens writing as Twain as pseudonymy, a different name on a work whose authorial function stays traceable to the civil identity behind it; in its account a pseudonym changes the label attached to an authorial function and a heteronym changes the organization of the function itself, the two told apart by eight testable markers, and the heteronym's function is attributable through the civil bearer without being reducible to that bearer. The archive also holds that heteronyms often have independent authority-control records because library systems catalogued them separately before the heteronymic nature of the authorship was widely understood, and it lists VIAF's Wikidata property, P214, among the identifiers a heteronym's item carries. Whether VIAF holds a pseudonym or heteronym as a see or see-also form of its bearer's cluster or as a cluster of its own is unanswered in the census record's saved source, as is whether a given chosen name has stayed distinct or been merged by the clustering algorithm; no VIAF cluster for a pseudonym or heteronym has been checked. The census reads the clusters that coalesce or split as the point where an authorial name can be merged with another by algorithm, the collapse #940 names, or separated again, and it records no instance of either for a given name. ISNI, whose assignment system is based primarily on VIAF and which VIAF lists as a participant, gives each name or pseudonym under which an author has published its own ISNI. The archive counts the forcing of heteronyms into "pseudonym" categories that miss their distinctiveness as category collapse, a form of taxonomic violence it observes where literary personas are deleted from Wikipedia or miscategorized in Wikidata, and it calls a Wikidata statement sourced only to "stated in Virtual International Authority File" structurally vulnerable and informationally shallow. In #940's account a civil name is the atomic unit of attribution across the systems that govern legitimacy, and a civil-name attribution regime can attempt to collapse a heteronymic corpus into single-civil-identity ownership; the heteronymic architecture does not prevent the attempt and makes the collapse register as loss. #1722 keeps the existence of a maker apart from the kind of work: a heteronym is a real author's practice, and a heteronym has a hand that wrote it. Christine L. Borgman groups VIAF with ISNI and ORCID as "loosely coordinated efforts to standardize name forms" and attempts at the problem of author name disambiguation, and notes that VIAF, unlike the other two, is led by libraries, as opposed to individual authors or creators. Among its own stated shortcomings is a limited ability for those outside the library community to correct errors, and the archive sets authority control, which assigns identity from above, against its own practice of constructing recognizability from within. The census reads VIAF as identity resolution led by libraries that links records judged identical into one cluster, records nothing in the file on authorial function or responsibility, and enters it as infrastructure to its question, a record of type index that keeps no works and links the authority records that participating libraries keep for names. Its census asks how authorship is made, named, distributed, attributed and sustained across persons, personas, collectives and machine participation, and what is lost when those relations are collapsed, admits aggregators that gather and hold many works or records and say what they hold, and places VIAF under attribution, credit and identity infrastructure and meta-aggregation, its evidence an identity registry and a catalogue, read against #940 and #1722. The data are available online for research, data exchange and sharing through a public website, a public API and bulk downloads, with reciprocal updating over OAI-PMH, and the dataset is released under the Open Data Commons Attribution License, a licence the census record found unstated in its saved source. Its file numbers are added to Wikipedia biographical articles and incorporated into Wikidata, and as linked open data it was used in 2012 by freebase.com and other services; the Linked Open Data Cloud's record marks itself without licence or provenance metadata, and its listed download links on viaf.org fail with HTTP 403. In 2012 its builders counted about 55,000 visits a month to its pages and about 10 million web-service searches or page accesses a month, many apparently from commercial web harvesters. VIAF's own site refused the census's fetch and this row's with a Cloudflare block, so its own statements are read here through its card's snippet, OCLC's pages and the 2012 dataset description; the census record rests on the Wikipedia article, last edited on 19 June 2026, and the ISNI article, last edited on 13 October 2026, tags a viaf.org sourceID URL as a dead link. Asked for "VIAF" on 1 November 2026, Google's AI Overview said what the file is, who runs it, where it came from, what it is for and how to use it; the Capture Registry found its origin, operator and integration exact and the 2012 transition carried on the fourth card only. The registry also found that the Overview linked pseudonyms on no saved page, gave each cluster a "unique" identifier where the saved account records its fluctuation, named no participant beyond the three founders, and left Borgman's point on libraries uncarried.

current · derived from the current composition

T against L_F(B) (AIO's ontology at work on its own cards): At 'VIAF' (T1, six cards, B1–B6, the disclosed pool B_D; B7, the ISNI article, is the reference pool and stands outside this comparison, §0.3), T composes a definition, five bulleted features and a closing offer. Composed at its source's grain or coarser: the definition (F11, F1, F2, F30), the operator (F9, F11), the origin (F7, F13, F14; 'launching formally in August 2003' for the consortium's formation), integration into Wikipedia biographical articles and Wikidata (F20), and public access with APIs (F56, F19). Substituted reading: 'combines multiple national library authority files into a single online database' (T1-1) composes the virtual file as one database; the field's founding distinction, that the project links existing national files after attempts at one joint file failed (F12, F17, F30, F43 'virtually combines'), is carried as a combination. Force raised: 'Solves author name disambiguation' (T1-4) where Borgman's account, as card 3 gives it, calls the three systems 'attempts to solve the problem' (F22), and card 6 lists conservative matching that misses many matches (F52). Composed beyond the cards: 'pseudonyms' (T1-4). The field holds variant names (F43), regional forms (F33, F34, F4) and the Clemens/Twain shortcoming (F53), which records that authority files differ on whether a pseudonym and its bearer are one entity or two: the linkage of pseudonyms is asserted and that limit dropped. 'corporate titles' and 'regional spelling rules' compose F43 ('persons and organizations') and F4 ('regional preferences for language, spelling, and script') at coarse grain. Limit dropped: 'Each cluster receives a unique standard identifier/URI' (T1-5) composes F43 ('assigns a URI to each cluster') and F18 ('a standard data number') without F24 (clusters coalesce or split; some identifiers fluctuate) or F54 (changes take more than a month); F50 (IDs maintained with redirects when clusters merge) qualifies the fluctuation and is absent too. 'widely used in library catalogs' has no field claim. Available, not composed: F23 (led by libraries, as opposed to individual authors or creators) and F21 (Borgman's grouping with ISNI and ORCID); F18's see and see-also forms; F3 (all descriptions of an entity merged into one cluster); the participant counts F5, F9, F31, F38, F39 (card 5's '20 national libraries' on the rail only); governance F6, F35, F36; the BnF's joining F8, F15; the 2012 transition F16, F29, on card 4's snippet only; F28 (the research page concluded); the licence F45, F40; the shortcomings F52, F53, F54; coverage F47; size and use F51, F55, F42; the data model F57; the participants table F25; F41 (the LOD record's failing downloads); F26 (an example identifier); F27. Composed without attribution: every bullet, at the coarse grain of reference description (recorded). Content replaced by use: the closing offer (T1-7) proposes a search and 'its RDF/JSON data', a next step; 'JSON' has no field claim. Nothing is assigned to another; nothing is denied.

v0.10 §0.5: the first diff is ontological and narrow. T and the field agree on the entity, a library name-authority file operated by OCLC; T's ontology is the reference entry's (what, who, when, for what, how to use), and it composes the file as a product with a stable identifier and a solved problem. What it keeps out of its own cards is the file's dynamics (monthly clusters that coalesce or split, redirects) and its limits as its builders state them (conservative matching, the Clemens/Twain difference, library leadership, correction outside libraries), and it adds a pseudonym linkage none of its cards states.

L_F(B) against L_A(B ∪ A) (the intervention): Admission adds, in one claim space with the field's: the census's question of how a chosen name is held, unanswered in its saved source (LS-12, LS-21, LS-29, LS-34), and its reading of the coalescing and splitting clusters as the point where an authorial name can be merged by algorithm, the collapse #940 names (LS-23); #940's distinction of pseudonym from heteronym, the label against the organization of the authorial function, with Clemens under Twain as its own example of pseudonymy (O940-01 to O940-04), which meets the field's Clemens/Twain shortcoming (F53) on the same pair of names; #668's claim that heteronyms often hold independent authority-control records, and its listing of P214 (O668-02); #261's category collapse and #668's thin-sourcing judgment on Wikidata statements sourced to VIAF (O261-01, O668-01); #317's 'stable identifiers' and institutional coordination (O317-01, O317-02), a rival description composed beside F24 and F50 under P_coh, and its authority control that 'assigns identity from above' (O317-03) beside Borgman's 'led by libraries' (F23); #940's civil name as the atomic unit of attribution and the collapse that registers as loss (O940-05, O940-06); #1722's heteronym as a real author's practice (O1722-01, O1722-02); the census's reading of VIAF as library-led identity resolution that records nothing on authorial function (LS-22, LS-14), its relation of infrastructure (LS-04) and its organizing question (LS-35); and the Capture Registry's findings on the composition at the name (CAP-1-01 to CAP-1-14). Not surfaced: no archive source is on T's rail (archive_controlled_cites 0). The field is carried whole; the archive adds the authorship question to the field's own object, which the field states only as a shortcoming of rules and conventions (F53) and as Borgman's 'led by libraries' (F23).

alignment: {'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', 'F30', 'F31', 'F32', 'F33', 'F34', 'F35', 'F36', 'F37', 'F38', 'F39', 'F40', 'F41', 'F42', 'F43', 'F44', 'F45', 'F46', 'F47', 'F48', 'F49', 'F50', 'F51', 'F52', 'F53', 'F54', 'F55', 'F56', 'F57'], 'T1_composed': ['F1', 'F2', 'F11', 'F30', 'F9', 'F7', 'F13', 'F14', 'F43', 'F4', 'F20', 'F56', 'F19', 'F18', 'F46'], 'T1_composed_note': "F43 and F18 composed without F24, F50, F54 (limit dropped); F22 composed at a raised force ('Solves'); F12, F17, F30 composed as a single online database (substituted reading); F4 and F43 at coarse grain for 'regional spelling rules' and 'corporate titles'", 'T1_outside_cards': ["'pseudonyms' (T1-4: no field claim names pseudonyms as linked by VIAF; F53 records the difference among files)", "'unique' (T1-5: no field claim; F24 records fluctuation)", "'widely used in library catalogs' (T1-5)", "'its RDF/JSON data' (T1-7: RDF in F57; JSON in no field claim)"], 'rule': "a field claim counts as composed when T states it at its source's grain or coarser; B_D only (B7 excluded, §0.3); the transcript read for this alignment only"}

pool_note: T against L_F(B_D): B1–B6; L_F(B) against L(B ∪ A): B1–B7 with A

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.