5.0068 Citations (cont.) (3/85)
Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Thu, 16 May 91 16:50:33 EDT
Humanist Discussion Group, Vol. 5, No. 0068. Thursday, 16 May 1991.
(1) Date: Thu, 16 May 91 09:27:59 BST (20 lines)
From: stephen clark <AP01@liverpool.ac.uk>
Subject: Re: 5.0055 Citation Statistics ...
(2) Date: Thu, 16 May 91 14:50:07 BST (15 lines)
From: DEL2@phoenix.cambridge.ac.uk
Subject: Re: 5.0056: Citation and Scholarship
(3) Date: Thu, 16 May 91 09:53:59 CDT (50 lines)
From: Michael Sperberg-McQueen <U35395@UICVM>
Subject: citation rates in the humanities and sciences
(1) --------------------------------------------------------------------
Date: Thu, 16 May 91 09:27:59 BST
From: stephen clark <AP01@liverpool.ac.uk>
Subject: Re: 5.0055 Citation Statistics ...
I agree with Judy Koren's surprise and horror at the suggestion that
we humanists only cite what we agree with. Of course we cite people from
other schools and with other opinions. Unless we're literary critics, I
suppose :-).
One other factor - apart from the ones mentioned - in the low citation
rate is that a good many journals don't care for lengthy bibliographies
and footnotes. In my own discipline (philosophy) there was even a fashion
for pretending that every article was a new beginning, owing nothing much
to the past, or else a contribution to a face-to-face conversation where
every reader already knew who was being attacked. Personally, I like
citing people, and even quoting them - but I've found that reviewers
of that kind then complain that there are too many quotations....
Stephen Clark
Liverpool University UK
(2) --------------------------------------------------------------25----
Date: Thu, 16 May 91 14:50:07 BST
From: DEL2@phoenix.cambridge.ac.uk
Subject: Re: 5.0056: Citation and Scholarship
Michael Hart replying to Larry Hurtado moves our discussion on
citations and scholarship towards an assessment of methodology.
Paul Feyerabend argues vehemently that there is *no* algorithm/
methodology for productive research; and produces (like Michael)
a string of key examples (not including Feynman, though) to back
up his assertion. But is the case comparable in the humanities?
Where a primary objective seems to be assessing all the arguments
then probably reading (and citing) a large volume of other works,
whether thorough, clear, careful usw or not, is probably
inevitable. Or do we challenge that objective?
Douglas de Lacey, Cambridge.
(3) --------------------------------------------------------------58----
Date: Thu, 16 May 91 09:53:59 CDT
From: Michael Sperberg-McQueen <U35395@UICVM>
Subject: citation rates in the humanities and sciences
Some readers of this list may be interested to know what was actually
claimed and measured in the studies cited at third-hand by Peter Denning
in CACM.
Denning refers to two 'News & Comment' pieces in Science by David P.
Hamilton, who summarizes yet-unpublished work done at the Institute for
Scientific Information by David Pendlebury. The figures given by ISI to
Hamilton and Science are for the percentage of source items published in
1984 and indexed by ISI remained uncited by other items in ISI's indexed
journals during the first five years after publication. So the
'uncitedness' rate quoted is indeed based on the Arts and Humanities
Citation Index (AHCI) database, and not upon the Science Citation Index
data, as has been suggested here.
In a letter to Science (22 March 1991) Pendlebury points out that much
of the discussion of these observations ignores several basic facts.
First, since ISI indexes everything in the publications covered, the
source items counted include letters, meeting abstracts, editorials,
obituaries, and reviews; in natural-scientific journals these make up
about a quarter of the items; in the AHCI journals they make up about
70% of the items. It may be unsurprising to most of us that reviews of
books and of theatrical performances are only rarely cited in later
literature. Ditto for letters to the editor (with the notable exception
of Edsger Dijkstra's famous 'GOTO considered harmful.' Citation rates
for journal articles only are naturally somewhat higher than for all
'source items'. (Preliminary figures show that only 93%, not 98%, of
articles in the AHCI database were still uncited after five years.)
Two other factors also complicate the direct comparison of figures for
natural science and humanities, as well as the interpretation of the
absolute figures: books are more important in the humanities (70-80% of
the citations in AHCI are to books, not to other articles, according to
Pendlebury), and humanities citations do not share the age pattern of
natural-science citations. In the natural sciences, the median age of a
citation is (if I remember aright) between 1 and 4 years; in the
humanities, the median age of citations is more apt to be measured in
decades, even if one dates cited works according to the date given in
the citation (which will give us linguists citing articles by Kant
written in the 1970s).
Pendlebury does not observe, but I will, that many of the journals in
AHCI are not in fact as clearly academic and research-oriented as
are their counterparts in SCI.
C.M. Sperberg-McQueen