Manuel Cardona and Werner Marx "Confused, forgotten, rediscovered - References, the deficient memory of science" Physik Journal 3 (11): 27-29 November 2004

Eugene Garfield garfield at CODEX.CIS.UPENN.EDU
Thu Dec 23 15:10:42 EST 2004

Werner Marx :  W.Marx at

Full text available at:,7058,2-3-0-0-1-phy_toc_content_page-0-21221,00.html

Title      : Confused, forgotten, rediscovered - References, the deficient
             memory of science

Authors    : Manuel Cardona and Werner Marx

Address    : Max Planck Institute for Solid State Research, Heisenbergstr.
1, D-70569 Stuttgart (Germany)

Journal    : Physik Journal 3 (11): 27-29 November 2004

Publisher: Wiley-VCH (Weinheim)
Document type: Article
Language: German
Cited References: 19

Abstract: Incorrect references are more than annoying scribal errors but
serious faults in the scientific archives. Due to lack of
standardization of journal titles sometimes papers are cited in hundreds
of variants. Analyzing the reference variants of highly-cited papers
shows that about 5 percent of all citations are erroneous with respect
to the reference numerical data. Several highly-cited papers were found
to be cited to about 90 percent with incorrect volume and/or starting
page. Such incorrect citations seem to have been copied from prominent
early citing papers causing or amplifying the errors. The earlier the
cited papers the higher the probability that reference “mutations”
occur. Sometimes different early papers of the same authors are
confused. Occasionally, the author names of scholars were replaced by
those of their prominent teachers. Specific errors allow an estimation
of the percentage of cited papers actually read by the citing authors.
Non-English 19th-century journals like “Annalen der Physik” are
difficult to acquire and to read for many present day scientists. Thus,
confused papers are sometimes not rectified, if at all, until decades
after the errors were introduced. Some early papers are now seen by
experts as seminal works. However, when their disciplines started to
boom they obviously were forgotten. Other papers from the end of the
19th or the beginning of the 20th century were rediscovered and are
increasingly and highly cited until present. Such papers obviously need
to be originally cited by prominent carrier papers to be reintroduced
into the citation network. Typical examples are given for each of the
error variants mentioned.


More information about the SIGMETRICS mailing list