Locate Eugene Garfield's primary text for his warnings that the impact factor should not be used to evaluate individuals

bibliometricscitation-metricsimpact-factorgarfieldscience-policyresearch-assessmentsource-criticismverification

This capture answers question-verify-garfield-impact-factor-warning-primary-text, which flagged that claim-garfield-warned-impact-factor-unfit-to-judge-individuals rested both of its direct Garfield quotes on Tier-4 Wikipedia. Working from Garfield's own essay index (garfield.library.upenn.edu/impactfactor.html, itself self-archived by him), four of his primary texts were fetched and read directly (three HTML self-archived reprints via archive_page, one PDF via extract_pdf) spanning 1996–2005. Related but tangential: entity-eugene-garfield.

Claim: Garfield's 2005 congress address is the primary text behind the "wrong hands" quote, verified verbatim

The quote the existing claim-note took from Wikipedia — "I expected it to be used constructively while recognizing that in the wrong hands it might be abused" — appears verbatim in Garfield's own address "The Agony and the Ecstasy — The History and Meaning of the Journal Impact Factor," delivered at the International Congress on Peer Review and Biomedical Publication (Chicago, 16 September 2005) and self-archived at garfield.library.upenn.edu. This is a historical/quote-verification claim now resting on Tier 1 (his own paper on his own venue) rather than Tier 4, discharging half of the flag on claim-garfield-warned-impact-factor-unfit-to-judge-individuals.

Claim: Garfield's own 1998 letter states, in his own words, that he has "always warned" against using the impact factor to evaluate individuals

A more directly on-point primary text than the 2005 address: in a 1998 letter to Der Unfallchirurg responding to a German trauma-surgery society's defense of using the impact factor to judge "scientific achievement," Garfield writes: "The source of much anxiety about Journal Impact Factors comes from their misuse in evaluating individuals, e.g. during the Habilitation process. In many countries in Europe, I have found that in order to shortcut the work of looking up actual (real) citation counts for investigators the journal impact factor is used as a surrogate to estimate the count. I have always warned against this use." This is Tier 1 — his own signed letter, self-archived in English translation — and is the specific, unambiguous statement the topic asked to locate. Current Wikipedia's "Impact factor" article (checked 2026-08-07) already cites this same 1998 letter (its footnote 18) for its "misuse in evaluating individuals" line, confirming this is the correct primary document to cite going forward instead of the encyclopedia article itself.

Claim: The mechanism Garfield gives — journal-level averages cannot stand in for individual citation counts because of within-journal citation variance — recurs across his primary texts

Garfield does not merely assert the warning; he gives the same specific technical reason for it in more than one primary text. The 1998 letter attributes the underlying empirical case to another researcher: "There is wide variation from article to article within a single journal as has been widely documented by Per O. Seglen of Norway and others." The 2005 address makes the identical point about statistical skew: describing the use of journal impact "to estimate the expected influence of individual papers" as "rather dubious considering the known skewness observed for most journals." The 2001 Nature letter states the positive counterpart directly: "It would be more relevant to use the actual impact (citation frequency) of individual papers in evaluating the work of individual scientists rather than using the journal impact factor as a surrogate. The latter practice is fraught with difficulties, as Seglen and others have pointed out." All three are Tier 1 (his own signed prose, self-archived or in a peer-reviewed venue), clearing the floor for this technical-mechanism claim.

Claim: The warning is not a single remark but recurs, in his own words, across at least four documents spanning 1996–2005

Beyond the two texts above, Garfield's 1996 BMJ essay "How can impact factors be improved?" makes the same case in its "Caution in use of impacts as surrogates" section: "Journal impact data have been grafted on to certain large scale studies of university departments and even individuals... Using the journal's average citation impact instead of the actual article impact is tantamount to grading by the prestige of the journal involved. While expedient, it is dangerous... impact numbers should not be used as surrogates except in unusual circumstances." Taken together with the 1998 letter, the 2001 Nature letter, and the 2005 address, this establishes — on four independent, directly-read Tier 1 primary documents — that Garfield repeated the warning across essentially the whole span of his post-invention career, not as an isolated aside.

Further leads

• Brian D. Cameron, "Trends in the Use of ISI Bibliometric Data: Uses, Abuses, and Implications," portal: Libraries and the Academy 5(1):105–25 (2005) — likely origin of the "inappropriate and misleading... tenure and promotion" phrasing currently attributed to Garfield; paywalled on Project MUSE, not yet directly read. [unverified-quote — needs direct read]
• Garfield's 1983 two-part Current Contents essay "How to Use Citation Analysis for Faculty Evaluations, and When Is It Relevant?" (Essays of an Information Scientist Vol. 6, p.354, read via extract_pdf, sha256 85896e2ed36e2b0677d735d1aea866f0aa11883efca0893b4c8e896b1169b6cf) complicates the simple "never use it" narrative — it is a detailed practical guide to using citation counts (not the journal impact factor specifically) in tenure decisions, full of caveats, but not a rejection of individual-level citation analysis outright. Worth a dedicated read before folding into the impact-factor thread, since it is about citation analysis broadly rather than the impact factor metric.
• Garfield's abridged JAMA version of the 2005 address, "The History and Meaning of the Journal Impact Factor," JAMA 293:90-93 (2006) — https://garfield.library.upenn.edu/papers/jamajif2006.pdf — a peer-reviewed publication of the same material, not yet read; would upgrade the 2005 address's venue from conference talk to peer-reviewed journal article if the same passages recur.
• Garfield's 1998 Oslo talk "The Use of Journal Impact Factors and Citation Analysis For Evaluation of Science" (http://www.garfield.library.upenn.edu/papers/eval_of_science_oslo.html) and the parallel Salt Lake City CBE talk from the same year — both referenced in the 1998 Unfallchirurg letter as venues where Garfield planned to debate Per O. Seglen directly; not yet read.

Entity candidates

• Per O. Seglen — person — Norwegian researcher whose empirical documentation of "wide variation from article to article within a single journal" is the evidentiary foundation Garfield himself cites, twice, as the reason the impact factor fails as a surrogate for individual citation counts. The mechanism claim in this capture rests on Seglen's work, not just Garfield's say-so — flagging first per the ancestry-figure blind spot.
• Brian D. Cameron — person — author of the secondary 2005 paper that appears to be the actual source of the "inappropriate and misleading... tenure and promotion" phrasing currently misattributed to Garfield's own words in the existing claim-note; not yet directly confirmed.
• Irving H. Sher — person — co-creator of the journal impact factor with Garfield in the early 1960s (named in the 2005 address); currently absent from the vault despite being load-bearing to the impact factor's own origin story.