# Citing Objects, Lots, Archives and Databases > Auction lot URLs die after the sale and one in five scholarly articles already suffers reference rot. How to cite an object, a lot, an archive and a database so the citation survives. > The Study - 2026-08-02 > Canonical HTML: https://iconotheca.com/study/citing-objects-lots-and-databases/ ## Key points - Chicago's artwork citation ends with the museum's own object or accession number, and that number is the element that survives redesigns, retitling and reattribution. - Klein and colleagues found in 2014 that one in five of close to 400,000 scholarly articles suffered reference rot, rising to seven in ten among articles that cited web resources at all. - Christie's public results reach back only to 1998 and external access to its physical archive closed in February 2021, so a lot citation must carry the sale date and lot number rather than a link. - Getty's CONA gives objects persistent numeric identifiers but states that its coverage is still limited and growing, so ULAN identifiers and ARKs are the more widely available handles today. ## Cite the object by its number, because the number is the only stable part ## A lot is a bibliographic object with a date, a number and a currency ## In archives, cite the box, not the finding aid ## Databases and the identifier problem ## Reference rot is measured, and the remedy costs a minute Every part of a good art-historical citation is designed to work when the internet does not. The accession number, the sale date and lot number, the Lugt number, the box and folder, the ULAN identifier: each was created by an institution that intends to keep it, and each will still resolve long after the URL you copied has been retired in a platform migration. Put the identifiers in the note, hang the link on the end of them, and archive the page while you are already looking at it. That is roughly a minute per citation, and it is the difference between work that can be checked in thirty years and work that is decaying while it is still being marked. ## Questions **What do I do when an auction lot page has already disappeared?** Cite the sale: house, city, sale title, date and lot number, with the Lugt number if the sale falls between 1600 and 1925. Then find the catalogue itself in a digitised run, through Heidelberg's German Sales, the Wildenstein Plattner Institute or the Getty Provenance Index, and cite that copy naming the holding institution. The printed catalogue is the primary record; the web page was always a derivative of it. **Should I cite Wikidata?** Not as authority. A Q-number is genuinely useful as a crosswalk when you are reconciling names or objects across systems, so record it alongside the underlying institutional identifier rather than instead of it. The citation that carries weight is the museum accession number, the ULAN or CONA identifier, or the archival reference. **Does an accession number ever change?** Rarely. It is assigned when the object is accessioned and functions as the key to the curatorial file, so it is built to persist through retitling, reattribution and website rebuilds. Chicago's instruction is simply to use whatever term and number the institution itself uses, which is also the safest practice when a museum's conventions are unfamiliar. **Is Perma.cc available to a student?** It is run through a consortium of libraries, so check whether your university or law library is a member before assuming you need to pay. Where it is not available, the Internet Archive's Wayback Machine will archive a URL on request at no cost, and the archived snapshot can be cited alongside the live address. ## Sources - [1] Mississippi State University Libraries, 'Chicago Image and Artwork Citation' (artist, title, date, medium, dimensions, institution, object number, URL; worked example with object no. 162.1934). - [2] Columbia College Library, 'Chicago Citation Guide (18th Edition): Images, Artwork, and Maps' (use whatever term the museum uses for the accession number; omit unknown elements). - [3] California State University, Northridge, 'Chicago Manual of Style: Citing Archival Materials' (order of elements; box and folder; use of quotation marks). - [4] Klein M, Van de Sompel H, Sanderson R, Shankar H, Balakireva L, Zhou K, Tobin R, 'Scholarly Context Not Found: One in Five Articles Suffers from Reference Rot', PLOS ONE 9(12): e115253 (2014). - [5] 'Scholarly Context Adrift: Three out of Four URI References Lead to Changed Content', PLOS ONE (2016), on content drift. - [6] Perma.cc, 'About' (archiving at the moment of citation; library consortium model). - [7] Harvard Library Innovation Lab, 'Perma.cc' (origins; more than 150 law journals; over half of cited links in Supreme Court opinions no longer resolve). - [8] Harvard Law Review Forum, 'Perma: Scoping and Addressing the Problem of Link and Reference Rot in Legal Citations'. - [9] Getty Research Institute, 'Cultural Objects Name Authority (CONA)' (scope; persistent numeric IDs; limited but growing coverage updated monthly; links to AAT, TGN, ULAN, IA; ODC-By 1.0 for linked open data). - [10] Getty Research Institute Library, 'Getty Provenance Index Resources Relating to Sales and Auctions' (more than 2.3 million records; constituent datasets). - [11] ARK Alliance, home of the Archival Resource Key (some 15.3 billion ARKs by more than 1,700 organisations since 2001; stewardship at the California Digital Library). - [12] OCLC Research, Hanging Together, 'What is the value of persistent identifiers for cultural heritage objects?' (Smithsonian ARKs via N2T; Natural History Museum DOIs; ARK and Handle prevalence in Europeana). - [13] Courtauld Gallery Collection Online, 'What is IIIF?' (manifests as packages of image and metadata for reuse across systems). - [14] Artnet News, 'Christie's Has Closed Off Public Access to Its 255-Year-Old Auction Archive' (February 2021; contents of the King Street archive; online results from 1998). - [15] Brill, 'Lugt's Repertoire Online' (Lugt numbers as the standard reference for sales 1600-1925).