One in five scientific articles suffers reference rot; 7-in-10 web-citing have compromised context

An analysis of 3.5 million scientific articles and approximately one million Uniform Resource Identifiers found:

• One in five scientific articles suffers reference rot.
• Among articles that cite web content, seven in ten have compromised scholarly context (S-0053).

The figures measure the cost of reference rot at the scholarly-record level. Reference rot here is not the failure of a citation in isolation; it is the failure of citations to communicate the context in which the citing claim was made.

When 70% of web-citing articles have compromised context, the cumulative effect on the scholarly record is that the chain from claim to evidence breaks at a substantial rate even within published, peer-reviewed literature. Future researchers can read the citing article, see the web URL, and find that the URL no longer points to what it once did — or to anything at all.

The case is one of the strongest available inputs for C-0023. The architectural fix is content addressing of cited material; the W3C Robust Link standards, perma.cc, Internet Archive, and Tier 3 distributed archives each address it through different mechanisms.