ARCHIVE GAPS AND THE DISCIPLINE OF NOT KNOWING

Full report

On this page

1. Executive Summary

The digital historical record is defined as much by its silences as by its preserved contents. As the velocity of information sharing has accelerated globally, so too has the rate of information loss, precipitating a pervasive epistemological crisis across disciplines. Researchers, journalists, legal investigators, and the general public routinely encounter digital gaps: broken hyperlinks, missing web pages, excluded domains, and unresolvable scholarly citations. In the absence of a visible record, human cognition tends toward pattern matching and motivated reasoning, frequently filling the void with assumptions of malfeasance, censorship, or institutional conspiracy. This report serves as the foundational theoretical and operational document for THE GAP MAP, a framework designed to standardize the evaluation of absent digital and physical evidence. The framework establishes rigorous analytical boundaries for determining what can and cannot be inferred when a document, webpage, image, claim, or organization is absent from archives for a specified period. The architecture of the web, the limitations of web crawling technology, the legal constraints of intellectual property, and the natural decay of data infrastructures produce a landscape where absence is the statistical norm, not the exception1. The framework is governed by a singular, immutable central law:

NOT FOUND is not automatically REMOVED, RETRACTED, SUPPRESSED, DESTROYED, or SECRETLY MAINTAINED.

Adherence to this law requires dismantling the false binary between absolute preservation and intentional destruction. To navigate this landscape, researchers must master the discipline of not knowing—restricting their conclusions strictly to what surviving metadata and affirmative evidence can support. By operationalizing the concepts of statistical missingness, archival silence, and digital decay, THE GAP MAP provides a mechanism to classify absent evidence without resorting to conspiratorial overreach.

2. Theoretical Foundations

The analysis of missing data necessitates a multi-disciplinary theoretical foundation, drawing upon the philosophy of science, statistics, historiography, and network architecture.

Absence of Evidence versus Evidence of Absence

The philosophical distinction between the absence of evidence and evidence of absence is the cornerstone of gap analysis. The philosopher Elliott Sober’s "Probabilistic View" posits that the absence of evidence only serves as evidence of absence when the discovery of evidence is highly expected4. In Bayesian terms, if a hypothesis strongly predicts the existence of observable evidence, then a failure to observe that evidence significantly lowers the probability of the hypothesis5. Sober utilizes the analogy of a firing squad: if a prisoner survives a firing squad, the absence of bullet wounds is strong evidence that the squad missed or fired blanks, because the expectation of wounds was nearly absolute6. Translated to web archiving, if a highly trafficked homepage is crawled by the Internet Archive every ten minutes, the absence of a major headline during a six-hour window constitutes strong evidence that the headline was not published during that interval. However, if a heavily nested subpage on a dynamic, JavaScript-heavy portal is not found in an archive, the expectation of initial capture is extraordinarily low7. Under what theorists term the "Pragmatic View," the absence of evidence in low-expectation environments cannot be used to infer nonexistence; rather, it serves as a scaffold for investigating auxiliary hypotheses, such as crawl exclusions or technical barriers4.

Statistical Missingness

To rigorously evaluate why data is absent, one must apply the statistical classifications of missingness to the behavior of web crawlers and archival systems:

• Missing Completely at Random (MCAR): The probability of a page being missing is unrelated to the page's content or any other observed variable. In archiving, this occurs during random server downtimes, transient Domain Name System (DNS) resolution failures, or random packet loss during a wide crawl3. The absence reveals nothing about the artifact itself.
• Missing at Random (MAR): The probability of a page being missing is related to an observable variable, but not the specific content of the page itself. For instance, a domain might be completely excluded because its host server's robots.txt file blocks all automated crawlers12. The exclusion is systemic and observable, applying equally regardless of whether a specific page contains innocuous data or controversial claims.
• Missing Not at Random (MNAR): The probability of missingness is directly related to the unobserved value or content itself. This occurs when a specific article is targeted with a Digital Millennium Copyright Act (DMCA) takedown notice, a Right to Be Forgotten request, or a court order specifically because of the claims it contains13.

Archival Silence and the Production of History

The anthropologist Michel-Rolph Trouillot theorized that historical narratives are built upon layers of silence, which are actively manufactured rather than passively inherited. Trouillot identified four crucial moments where silences enter the production of history: the moment of fact creation (the making of sources), the moment of fact assembly (the making of archives), the moment of fact retrieval (the making of narratives), and the moment of retrospective significance (the making of history)16. In the digital realm, Trouillot’s four moments map flawlessly to the archival lifecycle. First, at the moment of fact creation, a webpage may be rendered dynamically via client-side JavaScript, inherently resisting static capture19. Second, at the moment of fact assembly, an archiving crawler might encounter a payload limit, a robots.txt exclusion, or a paywall, ensuring the page is not written into the Web ARChive (WARC) container file12. Third, during fact retrieval, a researcher might query a CDX index server but use the wrong Uniform Resource Locator (URL) syntax, failing to retrieve an existing Memento22. Finally, at the moment of retrospective significance, a pundit may weaponize the resulting gap, falsely claiming the missing page proves a conspiracy25.

Bias and Censoring in the Archive

Archival records are systematically distorted by overlapping biases. Preservation bias occurs when institutional policies favor the capture of certain domains—such as government (.gov) or academic (.edu) sites—over marginalized, counter-cultural, or ephemeral platforms27. Survivorship bias leads researchers to study only the web structures that successfully withstood time and technical transitions, mistakenly assuming the surviving network is representative of the whole2. Selection bias is introduced by human curators who designate specific URL seeds for high-frequency crawling while ignoring the broader ecosystem. Furthermore, digital decay is subject to strict statistical censoring mechanisms:

• Right Censoring: A URL is captured by an archive in 2021, and the crawler never returns. The page may still exist on the live web today, or it may have been deleted in 2022, but the archive's observable timeline permanently ends in 2021\.
• Left Censoring: A webpage was demonstrably created in 2014, but the first archival snapshot was not successfully captured until 2017\. The artifact's initial three years of existence, including potential early edits, are permanently undocumented.
• Interval Censoring: A webpage was archived on January 1 and again on December 31\. Any alterations, temporary removals, retractions, or defacements that occurred between those two dates are lost to history, rendering the continuity of the interval entirely opaque29.

Digital Decay: Link Rot, Content Drift, and Platform Decay

The phrase "digital dark age" refers to the accelerating loss of cultural memory due to format obsolescence and infrastructure decay. This decay manifests primarily as "reference rot," which comprises two distinct but equally destructive phenomena:

1. Link Rot: The resource identified by a Uniform Resource Identifier (URI) ceases to exist, resulting in an HTTP 404 Not Found error, a 410 Gone error, or a total DNS failure10. Studies reveal that 38% of webpages existing in 2013 were inaccessible by 2023, and over 65% of URLs spanning a 26-year longitudinal sample are dead on the live web1. Link rot is particularly devastating in legal and scholarly citations, where foundational texts simply vanish31.

2. Content Drift: The URI continues to resolve successfully (HTTP 200), but the content hosted at the endpoint has fundamentally changed, rendering historical citations invalid30. A URL that once hosted a vital research paper may now host an expired domain parking page or unrelated marketing material.

Combined with platform decay—where structural changes to a host platform, such as a shift from static HTML to dynamic API-driven rendering, break legacy links and orphan thousands of pages—these forces ensure that the baseline state of the web is one of constant, grinding erasure29.

3. Missingness Taxonomy

When an analyst encounters an archival gap, they must systematically eliminate mundane technical and administrative causes before considering targeted removal or suppression. The following taxonomy comprehensively outlines why evidence disappears or fails to appear, categorizing absences into Ontological, Technical, Administrative, and Epistemic domains.

This taxonomy demonstrates the vast statistical probability that a missing document is the result of structural friction rather than targeted suppression. The burden of proof always rests on the analyst to eliminate technical and administrative failures before declaring an artifact suppressed.

4. Archive-Platform Comparison

Navigating digital gaps requires an intimate understanding of the differing technical architectures, legal jurisdictions, and collection policies of various archives. The Open Archival Information System (OAIS) ISO 14721 standard outlines the functional requirements of long-term preservation—Ingest, Archival Storage, Data Management, Administration, Preservation Planning, and Access—but platforms implement these functions with extreme variance43. Furthermore, the underlying data is typically stored in Web ARChive (WARC) files, standardized under ISO 28500, which bundle HTTP requests, responses, headers, and payloads into a single verifiable file20. Indexes of these WARC files are served via CDX servers, allowing researchers to query captures by URL and timestamp22. Understanding the strengths and weaknesses of different crawling architectures is critical. Traditional crawlers like Heritrix (used historically by the Internet Archive) excel at high-volume, continuous crawling but struggle heavily with JavaScript, dynamic content, and API-driven interfaces19. Conversely, Headless Browsers (like Brozzler and Browsertrix) execute client-side JavaScript and record the network traffic. They capture high-fidelity representations of modern web apps but are computationally expensive and therefore cover significantly less total volume7. The following table compares the primary preservation platforms researchers utilize when hunting for absent evidence:

5. Gap-State Definitions

To prevent analysts from defaulting to "suppressed" when encountering an HTTP 404 error, THE GAP MAP requires every absent artifact to be assigned a strictly defined qualitative state. This standardizes reporting and forces the acknowledgment of epistemic limits.

Crucial Directive: "Suppression documented" and "Destruction documented" cannot be inferred from a gap. They must be proven by a positive artifact (e.g., an email ordering the deletion, a Lumen Database entry showing a DMCA notice, or a public lawsuit)14. Silence alone is never proof of suppression.

6. Case Studies in Archival Absence

The following eight cases demonstrate how archival silence can be misunderstood, and how rigorous gap analysis ultimately resolves the epistemic crisis by adhering to the principles of THE GAP MAP.

6.1. A Known Technical Crawl Failure: Twitter's UI Migration

• The Gap: In 2020, researchers noticed that the Internet Archive's "Save Page Now" feature was failing to capture Twitter profiles accurately. Instead of the profile content, the archive displayed a "Browser is no longer supported" error or a "Sorry, that page doesn’t exist\!" error60.
• Initial Misinterpretation: A naive observer might conclude that specific accounts were being aggressively censored or deleted by Twitter.
• The Reality: Twitter rolled out a new user interface reliant on dynamic API calls to api.twitter.com. The Internet Archive's crawlers were hitting severe rate limits (HTTP 429 Too Many Requests), causing the capture to fail. When the Memento was replayed by a user, it tried to fetch missing JSON components, resulting in a "temporal violation" where the archived page displayed an error that never existed on the live web29.
• Gap State: Capture inaccessible (Technical).

6.2. A Known Domain Migration: Gawker Media to Kinja

• The Gap: Following a devastating lawsuit funded by billionaire Peter Thiel, Gawker Media was forced into Chapter 11 bankruptcy. Its assets were sold to Univision, but the flagship domain, Gawker.com, was left behind34.
• The Threat: As Gawker's former sister sites migrated to the Kinja platform under new ownership, legacy URL structures risked being broken. Furthermore, Thiel explicitly attempted to purchase the Gawker estate's assets, leading to widespread fears that he intended to permanently eradicate the archive of over 200,000 articles34.
• The Reality: While platform migrations routinely cause massive link rot, public outcry and the intervention of archivists ensured that static captures of Gawker were prioritized and saved before domain control could result in erasure.
• Gap State: Platform migrated / Reason disputed (Legal threat vs. Administrative decay).

6.3. A Genuine Documented Removal: CDC Airborne Guidance

• The Gap: In September 2020, the US Centers for Disease Control and Prevention (CDC) updated its official COVID-19 guidance page to state that the virus could spread through the air via aerosols. Days later, the page reverted to its previous language, quietly removing the references to airborne transmission61.
• Initial Misinterpretation: The sudden disappearance of the text fueled immediate allegations of political interference and scientific suppression.
• The Reality: Web archives successfully captured both the update and the subsequent removal, effectively eliminating interval censoring62. The CDC later claimed the draft was posted in error before final technical review. While the motive remains heavily debated, the removal itself is a verified fact, documented through timestamped Mementos.
• Gap State: Page known removed.

6.4. A Documented Retraction: Surgisphere and The Lancet

• The Gap: In 2020, a highly influential study on hydroxychloroquine published in The Lancet was abruptly removed from the active scientific consensus58.
• Initial Misinterpretation: Proponents of the drug claimed the medical establishment was suppressing valid science to harm political opponents.
• The Reality: Independent researchers discovered that the data provider, a tiny company named Surgisphere, lacked the infrastructure to possess the data it claimed to have. The Lancet and the New England Journal of Medicine issued formal Expressions of Concern, followed by full retractions requested by the papers' own co-authors when Surgisphere refused to allow an independent data audit58.
• Gap State: Retraction documented.

6.5. A Physical Artifact with No Early Digital Presence: Empty Fields

• The Gap: The Museum of Anatolia College in Merzifon, Turkey, possessed a vast natural science collection curated by an Armenian scientist before World War I. For decades, there was virtually no digital footprint or academic archive of this collection16.
• Initial Misinterpretation: If one relied solely on early web archives or modern Turkish state records, one might assume the institution and its collections never existed.
• The Reality: The silence surrounding the museum was a direct consequence of the Armenian genocide and subsequent state denial. The physical archive was destroyed or dispersed, and the silence was enforced politically16. The 2016 exhibition Empty Fields reconstructed the gap using a surviving handwritten inventory catalog, proving the limitations of digital-only research in the face of historical atrocity16.
• Gap State: Physical copy survives / Suppression documented.

6.6. A Falsely Claimed Suppression: Substack NFT Plagiarism

• The Gap: A cryptocurrency critic writing on Substack, Mike Burgersburg, had his entire body of work taken down by Substack following DMCA copyright notices filed by a company called Mevrex on behalf of a site called "UNFT News"13.
• The Claim: UNFT News claimed Burgersburg stole their articles, presenting backdated publishing timestamps on their live website as proof of original authorship13.
• The Reality: Using the Wayback Machine's CDX server, researchers proved that the UNFT News domain was not even registered, and its website did not exist, on the dates they claimed to have published the articles13. The archive proved the DMCA takedown was a fraudulent abuse of copyright law to enact censorship.
• Gap State: Falsely claimed suppression (Reversed by archival evidence).

6.7. Fragmented Text Reconstructed from Citations: The Epic Cycle

• The Gap: The "Epic Cycle," a collection of Ancient Greek epic poems relating to the Trojan War, is entirely lost to history. No physical manuscript, papyrus fragment, or digital surrogate of the original texts exists.
• Initial Misinterpretation: The texts and their specific narrative details are permanently erased.
• The Reality: The narratives survive almost entirely through secondary citations, fragments quoted by later authors, and a prose summary contained in the Chrestomathy attributed to Proclus35. The absence of the primary source does not equate to the total absence of the data; the shape of the missing text is outlined by the surrounding literature.
• Gap State: Citation survives without source.

6.8. A Genuinely Unresolved Gap: Missing True Crime Files

• The Gap: In historical true crime and unresolved mysteries (e.g., the disappearance of Johnny Gosch), specific police reports, witness statements, or contemporary news broadcasts are occasionally discovered to be missing from local archives or digital databases65.
• Initial Misinterpretation: Amateur sleuths frequently cite missing files as definitive proof of a cover-up by authorities or a conspiracy to protect powerful individuals.
• The Reality: While intentional destruction by corrupt actors is historically possible, it is statistically dwarfed by the banality of administrative failure: misfiling, flood or fire damage in paper archives, server migrations, and routine retention schedule purges. Without affirmative proof of a cover-up, the gap remains entirely unresolved and cannot be used as evidence of a conspiracy.
• Gap State: Reason unknown.

7. Required Public Language: The Rewrite Table

When communicating the results of a gap analysis to the public, precision of language is critical. Anthropomorphic verbs and assumptions of intent must be rigorously stripped from the reporting. Analysts utilizing THE GAP MAP must utilize the following safe replacements for common overclaims.

8. A Model Gap Record

A valid entry into THE GAP MAP must follow a strict schema, separating technical metadata from qualitative analysis. The following is a model record based on the CDC case study. SUBJECT: CDC Airborne Transmission Guidance Update URI-R (Original): https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/how-covid-spreads.html INTERVAL: September 18, 2020 – September 21, 2020\. CDX STATUS: 14 captures present across Internet Archive and Archive.today within the interval. GAP STATE: Page known removed. EVIDENCE OF EXISTENCE: Memento-Datetime: 2020-09-19T10:15:30Z confirms the text "viruses... can remain suspended in the air and be breathed in by others" was present62. EVIDENCE OF ABSENCE/REMOVAL: Memento-Datetime: 2020-09-21T14:22:10Z confirms the text was reverted62. MOTIVE/MECHANISM: Reason disputed. Agency claimed the draft was posted in error; external critics allege political pressure. CONCLUSION: Continuity of publication was broken. Motive cannot be proven exclusively via archival metadata.

9. WHERE THE GAP MAP BREAKS

The framework approaches its limits when confronting the inherent opacity of certain digital architectures and the threat of sophisticated manipulation. THE GAP MAP breaks down in the following scenarios:

1. Deep Web and Database Silos: Content dynamically generated from proprietary, authenticated databases cannot be archived by external crawlers. If a record is purged from an internal corporate database (e.g., a hospital record system or an intelligence database), no public CDX server will ever hold a trace of its existence.

2. Encrypted Peer-to-Peer Networks: Ephemeral messaging platforms (e.g., Signal, Telegram secret chats, WhatsApp) are explicitly designed to resist archiving. Silence in these domains is the intended operational state, making historical gap analysis mathematically impossible.

3. The "Save Page Now" Poisoning: Malicious actors can use tools like the Wayback Machine's "Save Page Now" to archive a generated 404 error for a specific URL, establishing a false historical baseline. This suggests a page did not exist at a time when it actually might have, simply by querying a mistyped or modified URL37.

4. State-Level Archive Manipulation: If a highly sophisticated threat actor gains access to the physical servers of a web archive, they could theoretically tamper with the WARC files directly. While cryptographic hashing makes this exceedingly difficult, it represents the absolute epistemic limit of digital trust8.

10. Disclosure Language

Any research product, publication, or investigative report utilizing THE GAP MAP methodology must prominently append the following disclosure:

This analysis relies on publicly available archival telemetry, CDX indexes, and heuristic observation. Digital preservation is an inherently lossy process subject to technical limitations, platform decay, and random data loss. Conclusions regarding the absence of data are limited strictly to observable metadata. Absence from a specific archive does not constitute proof of global nonexistence, nor does removal constitute proof of malicious suppression.

11. Verification Checklist

Before finalizing a gap state and publishing findings, an analyst must complete the following rigorous verification steps:

• \[ \] Query the exact URL against the Internet Archive CDX API.
• \[ \] Query URL variations (HTTP vs HTTPS, www vs non-www, trailing slashes, parameterized tracking codes).
• \[ \] Query Memento Aggregators (e.g., TimeTravel) to check national and secondary archives.
• \[ \] Test the URL in Archive.today to bypass aggressive robots.txt or JS blockers.
• \[ \] Check the domain's historical robots.txt file for the date in question to rule out crawler exclusion.
• \[ \] Check search engine caches and scholarly citation databases (e.g., Crossref, DOI resolvers).
• \[ \] Verify if a platform migration or CMS upgrade occurred around the time of the gap.
• \[ \] Search for affirmative evidence of suppression (e.g., lawsuits, DMCA notices via the Lumen Database).

12. Source Audit Appendix

This report is synthesized from a rigorous audit of interdisciplinary research spanning web science, digital preservation architecture, and the philosophy of knowledge. The evaluation of digital decay relies heavily on longitudinal studies of the web. Research originating from the Web Science and Digital Libraries (WS-DL) Research Group at Old Dominion University forms the backbone of our current statistical understanding of link rot and URI redirection. Notably, their study analyzing 27.3 million URLs from the Wayback Machine demonstrated that 65% of historic URLs are dead on the live web1. The Hiberlink project and the Chesapeake Digital Preservation Group provide critical, terrifying data on reference rot within scientific (STM) and legal literature, respectively, demonstrating the fragility of the academic record30. The technical mechanics of the archive are derived from the foundational standards of the field: ISO 28500 (WARC file format) and ISO 14721 (OAIS Reference Model)20. Understanding how crawlers like Heritrix and Brozzler negotiate with the live web clarifies why technical capture exclusions occur19. The mechanics of datetime negotiation and the CDX server index are rooted in RFC 7089 (The Memento Protocol)22. Finally, the epistemological boundaries of this report—determining what silence actually means—are informed by Michel-Rolph Trouillot's historiographical masterwork, Silencing the Past, which deconstructs the power dynamics inherent in the making of archives17. This qualitative framework is balanced by the analytical logic of philosopher Elliott Sober, whose work on the transitivity of evidence dictates the probabilistic rules for when absence of evidence truly constitutes evidence of absence4.

13. Full Bibliography

• Alam, S., Nelson, M. L., & Weigle, M. C. (Various Years). Research publications from the Web Science and Digital Libraries (WS-DL) Group. Old Dominion University.
• Chesapeake Digital Preservation Group. (2012–2014). Annual Link Rot Reports.
• Consultative Committee for Space Data Systems (CCSDS). (2012). Reference Model for an Open Archival Information System (OAIS). ISO 14721:2012.
• Garg, K., Alam, S., Ayala, D., Weigle, M. C., & Nelson, M. L. (2025). Not here, go there: Analyzing redirection patterns on the web. Old Dominion University.
• International Internet Preservation Consortium (IIPC). (2017). WARC file format. ISO 28500:2017.
• Klein, M., Sompel, H. V., et al. (2014). Scholarly Context Not Found: One in Five Articles Suffers from Reference Rot. PLoS ONE.
• Lumen Database. (2026). Archive of DMCA and Legal Takedown Notices.
• Pew Research Center. (2024). When Online Content Disappears.
• Sober, E. (2009). Absence of Evidence and Evidence of Absence: Evidential Transitivity in Connection with Fossils, Fishing, Fine-tuning, and Firing squads. Philosophical Studies, 143: 63-90.
• Trouillot, M.-R. (1995). Silencing the Past: Power and the Production of History. Boston: Beacon Press.
• Van de Sompel, H., Nelson, M., & Sanderson, R. (2013). HTTP Framework for Time-Based Access to Resource States \-- Memento. IETF RFC 7089\.

THE ARCHIVE MAY BE SILENT FOR MANY REASONS. SILENCE MUST NOT BE FORCED TO CONFESS.

Works cited

1. Gone but Not Forgotten: Recovering the Dead Web | Internet Archive Blogs, https://blog.archive.org/2026/04/23/gone-but-not-forgotten-recovering-the-dead-web/

2. Longitudinal Sampling of URLs From the Wayback Machine \- arXiv, https://arxiv.org/pdf/2507.14752

3. Not Here, Go There: Analyzing Redirection Patterns on the Web, https://digitalcommons.odu.edu/cgi/viewcontent.cgi?article=1385\&context=computerscience\_fac\_pubs

4. When Should Absence of Evidence Be Evidence of Absence? A Case Study from Paleogeology | Philosophy of Science \- Cambridge University Press & Assessment, https://www.cambridge.org/core/journals/philosophy-of-science/article/when-should-absence-of-evidence-be-evidence-of-absence-a-case-study-from-paleogeology/D6C43B5292B649163DE577B634CCB1AE

5. (PDF) Is absence of evidence of pain ever evidence of absence? \- ResearchGate, https://www.researchgate.net/publication/348252440\_Is\_absence\_of\_evidence\_of\_pain\_ever\_evidence\_of\_absence

6. Cosmological Fine-Tuning Arguments: What (If Anything) Should We Infer From the Fine-Tuning of Our Universe for Life? | Reviews, https://ndpr.nd.edu/reviews/cosmological-fine-tuning-arguments-what-if-anything-should-we-infer-from-the-fine-tuning-of-our-universe-for-life-2/

7. Archival Crawlers and JavaScript: Discover More Stuff but Crawl More Slowly | Request PDF, https://www.researchgate.net/publication/318752105\_Archival\_Crawlers\_and\_JavaScript\_Discover\_More\_Stuff\_but\_Crawl\_More\_Slowly

8. Hashes are not suitable to verify fixity of the public archived web \- PMC \- NIH, https://pmc.ncbi.nlm.nih.gov/articles/PMC10256179/

9. When Should Absence of Evidence Be Evidence of Absence? A Case Study from Paleogeology \- ResearchGate, https://www.researchgate.net/publication/396657037\_When\_Should\_Absence\_of\_Evidence\_Be\_Evidence\_of\_Absence\_A\_Case\_Study\_from\_Paleogeology

10. Google Sports Data, https://support.google.com/knowledgepanel/answer/9787176

11. Abstracts \- IIPC \- International Internet Preservation Consortium, https://netpreserve.org/ga2024/abstracts/

12. Intermediaries stories at Techdirt., https://www.techdirt.com/tag/intermediaries/

13. How fraudulent DMCA takedowns censored a prominent cryptocurrency critic on Substack, https://sea.mashable.com/tech/21055/how-fraudulent-dmca-takedowns-censored-a-prominent-cryptocurrency-critic-on-substack

14. Lumen Database, https://lumendatabase.org/

15. Website removal from search engines due to copyright violation \- Emerald Insight, https://www.emerald.com/ajim/article/71/1/54/60589/Website-removal-from-search-engines-due-to

16. Empty Fields Revisited | The Art Institute of Chicago, https://www.artic.edu/digital-publications/36/perspectives-on-instability/19/empty-fields-revisited

17. Silence in the Archives: France, the Algerian War, and National Identity – Archivaria \- Érudit, https://www.erudit.org/en/journals/archivaria/2025-n99-archivaria010099/1118508ar/

18. (PDF) Silence, Power, and the Archive: Reading Marisa J. Fuentes and Michel-Rolph Trouillot \- ResearchGate, https://www.researchgate.net/publication/398932315\_Silence\_Power\_and\_the\_Archive\_Reading\_Marisa\_J\_Fuentes\_and\_Michel-Rolph\_Trouillot

19. Strategies for Preserving Digital Scholarship / Humanities Projects \- The Code4Lib Journal, https://journal.code4lib.org/articles/16370

20. INTERNATIONAL STANDARD ISO 28500, https://cdn.standards.iteh.ai/samples/68004/2f853716b6c94ed39193e8701b1a571a/ISO-28500-2017.pdf

21. The WARC Format \- IIPC Community Resources, https://iipc.github.io/warc-specifications/specifications/warc-format/warc-1.1/

22. wayback/wayback-cdx-server/README.md at master \- GitHub, https://github.com/internetarchive/wayback/blob/master/wayback-cdx-server/README.md

23. Access Archive-It's Wayback index with the CDX/C API, https://support.archive-it.org/hc/en-us/articles/115001790023-Access-Archive-It-s-Wayback-index-with-the-CDX-C-API

24. Longitudinal Sampling of URLs From the Wayback Machine \- arXiv, https://arxiv.org/html/2507.14752v1

25. Twitter Files \- Wikipedia, https://en.wikipedia.org/wiki/Twitter\_Files

26. archival silence \- SAA Dictionary, https://dictionary.archivists.org/entry/archival-silence.html

27. The Future(s) of Web Archive Research Across Ireland, https://mural.maynoothuniversity.ie/id/eprint/17558/1/HealySC-2023-FuturesOfWebArchiveResearch-Ireland.pdf

28. Profiling Web Archive Coverage for Top-Level Domain and Content Language, https://www.researchgate.net/publication/256606471\_Profiling\_Web\_Archive\_Coverage\_for\_Top-Level\_Domain\_and\_Content\_Language

29. 2020-11-04: New Twitter UI: Replaying Archived Twitter Pages That Never Existed \- WS-DL, https://ws-dl.blogspot.com/2020/11/2020-11-04-new-twitter-ui-replaying.html

30. Scholarly Context Not Found: One in Five Articles Suffers from Reference Rot, https://www.researchgate.net/publication/270001492\_Scholarly\_Context\_Not\_Found\_One\_in\_Five\_Articles\_Suffers\_from\_Reference\_Rot

31. Death To "Link Rot": Here's Where The Internet Goes To Live Forever \- Fast Company, https://www.fastcompany.com/3028321/death-to-link-rot-heres-where-the-internet-goes-to-live-forever

32. Scholarly Context Not Found: One in Five Articles Suffers from Reference Rot \- PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC4277367/

33. Perma: Scoping and Addressing the Problem of Link and Reference Rot in Legal Citations, https://harvardlawreview.org/forum/vol-127/perma-scoping-and-addressing-the-problem-of-link-and-reference-rot-in-legal-citations/

34. Erasing history \- Columbia Journalism Review, https://www.cjr.org/special\_report/microfilm-newspapers-media-digital.php

35. Rescuing the Bod(ies): Thinking about the Epic Cycle, Neoanalysis, and Introducing Iliad17, https://sententiaeantiquae.com/2026/03/31/rescuing-the-bodies-thinking-about-the-epic-cycle-neoanalysis-and-introducing-iliad17/

36. Wayback Machine \- Wikipedia, https://en.wikipedia.org/wiki/Wayback\_Machine

37. Wayback Machine: Digital Web Archive | PDF | Internet \- Scribd, https://www.scribd.com/document/970251978/Wayback-Machine

38. The European Union general data protection regulation: what it is and what it means, https://www.tandfonline.com/doi/abs/10.1080/13600834.2019.1573501

39. How to fix broken links: a practical agency guide \- wpcto, https://wpcto.net/wordpress-insights/how-to-fix-broken-links/

40. See You in Court, Harvard\! \- The New Inquiry, https://thenewinquiry.com/see-you-in-court-harvard/

41. Abstracts \- IIPC \- International Internet Preservation Consortium, https://netpreserve.org/ga2025/abstracts/

42. Chilling Effects On Chilling Effects As DMCA Archive Deletes Self From Google | Techdirt, https://www.techdirt.com/2015/01/12/chilling-effects-chilling-effects-as-dmca-archive-deletes-self-google/

43. OAIS Reference Model (ISO 14721): Home, http://www.oais.info/

44. Open Archival Information System \- SAA Dictionary, https://dictionary.archivists.org/entry/open-archival-information-system.html

45. ISO 14721 A Comprehensive Guide to Digital Preservation and Long-Term Archiving, https://certbetter.com/blog/iso-14721-a-comprehensive-guide-to-digital-preservation-and-long-term-archiving

46. Open Archival Information System \- Wikipedia, https://en.wikipedia.org/wiki/Open\_Archival\_Information\_System

47. Web ARChive Format (WARC) \- Metadata Standards Index \- Dublin Core, https://msi.dublincore.org/standards/warc

48. What Is a CDX File and How It Helps You Work with Archive.org \- Smartial, https://smartial.net/what-is-a-cdx-file-and-how-it-helps-you-work-with-archive-org/

49. WAC Programme \- \- International Internet Preservation Consortium, https://netpreserve.org/ga2018/wac/

50. Archiving web sites \- LWN.net, https://lwn.net/Articles/766374/

51. Introduction to the Memento Protocol, https://mementoweb.org/guide/quick-intro/

52. List of web archiving initiatives \- Wikipedia, https://en.wikipedia.org/wiki/List\_of\_web\_archiving\_initiatives

53. GitHub \- lanl/invenio-memento-proxy: Memento protocol support for the Invenio Repository: TimeGate and TimeMap services, https://github.com/lanl/invenio-memento-proxy

54. Memento Time Travel | Thought Splinters \- Peter Baumgartner, https://notes.peter-baumgartner.net/2021/05/25/memento-time-travel/

55. Infrastructure/Memento \- W3C Wiki, https://www.w3.org/wiki/Infrastructure/Memento

56. Perma.cc, https://perma.cc/

57. 2024-03-27: If the First Archived Copy by the Wayback Machine is 404, When did the URL First Exist? \- WS-DL, https://ws-dl.blogspot.com/2024/03/2024-03-27-if-first-archived-copy-by.html

58. Surgisphere: governments and WHO changed Covid-19 policy based on suspect data from tiny US company | Medical research | The Guardian, https://www.theguardian.com/world/2020/jun/03/covid-19-surgisphere-who-world-health-organization-hydroxychloroquine

59. Andrew Wakefield's fraudulent paper on vaccines and autism has been cited more than a thousand times. These researchers tried to figure out why. \- Retraction Watch, https://retractionwatch.com/2019/11/18/andrew-wakefields-fraudulent-paper-on-vaccines-and-autism-has-been-cited-more-than-a-thousand-times-these-researchers-tried-to-figure-out-why/

60. 2020-07-15: Twitter Was Already Difficult To Archive, Now It's Worse\! \- WS-DL, https://ws-dl.blogspot.com/2020/07/2020-07-15-twitter-was-already.html

61. Scientific Brief: SARS-CoV-2 Transmission \- CDC Archive, https://archive.cdc.gov/www\_cdc\_gov/coronavirus/2019-ncov/science/science-briefs/sars-cov-2-transmission.html

62. Updated CDC guidance acknowledges coronavirus can spread through the air \- dpaq.de, https://dpaq.de/KTTT3

63. Live updates, June 5-7: No new cases for 16th day straight | The Spinoff, https://thespinoff.co.nz/society/07-06-2020/live-updates-june-5-rugby-and-netball-take-lions-share-of-support-funds

64. Ivan Oransky – Page 48 \- Retraction Watch, https://retractionwatch.com/author/ivanoransky/page/48/

65. \[Unresolved Disappearance\] Taking a Crack at Dissecting the Claims Made in 'Who Took Johnny' by Noreen Gosch \- Reddit, https://www.reddit.com/r/UnresolvedMysteries/comments/8707vy/unresolved\_disappearance\_taking\_a\_crack\_at/

66. A Framework for Verifying the Fixity of Archived Web Resources \- ODU Digital Commons, https://digitalcommons.odu.edu/context/computerscience\_etds/article/1125/viewcontent/Aturban\_a\_framework\_for\_verifying\_the\_fixity.pdf

67. Fifth Annual Link Rot Report of the Chesapeake Digital Preservation, https://www.slaw.ca/2012/05/03/fifth-annual-link-rot-report-of-the-chesapeake-digital-preservation-group/