Garfield's citation count became a target, and science built an institution to fight it

Core claims

1. Journal editors coerce authors into padding self-citations to inflate impact factor, and it's common. Wilhite & Fong (2012, Science 335:542-3) surveyed 6,672 researchers across economics, sociology, psychology, and business disciplines, plus 832 journals; 175 journals were identified as coercers, one named by 49 different respondents. Their definition: coercive self-citation is a request that "(i) give[s] no indication that the manuscript was lacking in attribution; (ii) make[s] no suggestion as to specific articles, authors, or a body of work requiring review; and (iii) only guide[s] authors to add citations from the editor's journal." One editor's actual line to an author: "you cite Leukemia [once in 42 references]. Consequently, we kindly ask you to add references of articles published in Leukemia to your present article." — source_tier: 1 (peer-reviewed primary study).

2. The scientific community's formal response explicitly names the metric as the problem. The 2013 San Francisco Declaration on Research Assessment (DORA), now signed by thousands of researchers and institutions, states: "Do not use journal-based metrics, such as Journal Impact Factors, as a surrogate measure of the quality of individual research articles, to assess an individual scientist's contributions, or in hiring, promotion, or funding decisions." — source_tier: 1 (primary declaration text, sfdora.org).

Why this was hop-worthy

The infrastructure Garfield built to help scientists trace priority and influence (count who cites whom) became, decades later, a KPI that editors and authors learned to game strategically — a clean Goodhart's Law case study with a formal institutional antibody (DORA) as the resolution.

Further leads

• Ike Antkare: in 2010 Cyril Labbé used SCIgen to generate 102 nonsense papers under a fake author name that cited each other, pushing "his" h-index to 94 — briefly ranking above Einstein on citation count. Fun, well-documented, not pursued as a core claim here.
• Goodhart's Law itself (Charles Goodhart, monetary economics, 1975) as a general lens — worth its own note someday, distinct from this citation-specific instance.
• Shepard's Citations (1873) as Garfield's explicit model, per his own 1964 essay — a genuine 91-year law-to-science bridge, worth a dedicated capture with Garfield's primary text.

Hop chain

Hop 1: Eugene Garfield Starts the Science Citation Index — https://www.historyofinformation.com/detail.php?id=817 (context via search summary; page itself didn't render body text on fetch)

• Hook type: The person behind the thing / cross-domain bridge
• Hook: Garfield modeled the 1955-64 Science Citation Index on Shepard's Citations, an 1873 legal citator system (Frank Shepard, Chicago) — law preceding science by 80+ years.
• Why followed: Direct answer to the seed question's "what changed" framing, and a strong cross-domain + cross-time bridge (legal information science -> scientometrics).
• Key findings: Garfield referenced Shepard's methodology explicitly in his own 1964 essay as proof-of-concept that citation linking could scale without subjective keywording.

Hop 2: Coercive Citation in Academic Publishing (Wilhite & Fong, Science 2012) — https://www.aaas.org/sites/default/files/Coercive_citation_2012.pdf

• Hook type: Surprising claim
• Hook: The citation-counting infrastructure descended from Garfield's SCI is being strategically weaponized by journal editors, not just used descriptively.
• Why followed: A quantitative, well-sourced, and genuinely surprising consequence of the seed topic — citation counting created a new kind of leverage nobody in 1955 anticipated.
• Key findings: 6,672 survey responses, 832 journals; 175 identified coercer-journals; exact editor quote pressuring an author to cite "Leukemia" journal captured verbatim.

Hop 3: Ike Antkare, one of the great stars in the scientific firmament (Cyril Labbé) / Scientific American coverage — https://www.scientificamerican.com/article/publishers-withdraw-more-than-120-gibberish-science-and-engineering-papers/

• Hook type: Cultural resonance / surprising claim
• Hook: A fabricated author generated by nonsense-paper software briefly out-ranked Einstein by citation count.
• Why followed: Vivid proof-of-concept for the same gaming dynamic, useful as illustration; saved as a lead rather than a core claim to keep the capture tight.
• Key findings: April 2010, SCIgen-generated papers under "Ike Antkare" self-cited to reach an h-index of 94 on Google Scholar.

Hop 4: Goodhart's Law (general concept, via search) — no single canonical URL, concept check only

• Hook type: Cross-domain bridge
• Hook: A monetary-economics principle ("when a measure becomes a target, it ceases to be a good measure") cleanly explains the whole citation-gaming pattern.
• Why followed: Ties Hops 2-3 into one named mechanism from a totally separate field (macroeconomics), which is exactly the kind of bridge the protocol weights highest.
• Key findings: Used as a framing device rather than pursued into its own economics history — flagged as a future note candidate instead.

Hop 5: San Francisco Declaration on Research Assessment (DORA) — https://sfdora.org/read/

• Hook type: Mechanism question (what's the institutional response?)
• Hook: A formal 2013 declaration explicitly telling institutions to stop using Journal Impact Factor to judge individual scientists.
• Why followed: Natural resolution point for the chain — closes the loop from "metric created" to "metric gamed" to "field organizes against the metric."
• Key findings: Originated at a December 2012 American Society for Cell Biology meeting; thousands of individual and organizational signatories within weeks of its May 2013 publication.

Saved hooks not followed:

• Goodhart's Law full history (Charles Goodhart, 1975 UK monetary policy) — from Hop 4 — interesting cross-domain origin story, but pursuing it fully would have meant a 6th+ hop and diminishing focus.
• Shepard's Citations' one-letter treatment codes (overruled/criticized/followed) as a precursor to modern NLP "citation function/sentiment classification" — from Hop 1 — a strong mechanism hook (legal tech -> NLP) saved for a future chain since it would have doubled the thread's length.
• Clarivate's ongoing JCR suppression list (20 journals lost impact factors in 2025 for self-citation/stacking) — from Hop 2 — good evidence that Garfield's institutional descendant (ISI -> Clarivate) is still actively policing this problem today; saved because it's more of a "current status update" than a new hook.

post-worthy: maybe — a clean, well-sourced Goodhart's Law case study with a vivid hoax anecdote and a real institutional resolution (DORA), but it leans expository rather than surprising by the last hop.