J R Soc Med 2003;96:191-193
doi:10.1258/jrsm.96.4.191
© 2003 Royal Society of Medicine
The reference: more than a buttress of the scientific edifice
G D Schott MD PhD
National Hospital for Neurology and Neurosurgery, Queen Square, London
WC1N 3BG, UK
E-mail:
geoffrey.schott{at}uclh.org
 |
INTRODUCTION
|
|---|
Nowadays few scientists would dare to submit an article entirely
without
references. An unreferenced scientific text appears
naked and incomplete, its
contents unattested and unauthoritative.
References are tools for the
dissemination of knowledge, preservation
of standards and recognition of
previous
contributions.
1 But
what
other functions are served by references, and what lurks behind
them?
This paper explores the development and the uses, misuses
and abuses of
references in scientific publications.
|
ORIGINS
|
|---|
Long ago, anagrams, lectures, unpublished letters and sealed
notes were
amongst the means by which scientists sought to communicate
yet also protect
their intellectual
property.
2 It is
through
citation, however, that priority and the efforts of others are
most
readily acknowledged, and the reference as a method to
document sources
probably dates from the Italian
Renaissance.
3 But
references represent more than just an economical device
whereby a writer
refers back to the corpus of preexisting knowledge.
By formally recognizing
the contribution of his predecessors,
the writer avoids possible charges of
plagiarism or of expressing
merely unsubstantiated or eccentric opinions.
Whilst providing
the background that makes the text respectable, the reference
at
the same time permits the author to claim
originality.
4
References
can also serve another function. In many reviews or analyses
today,
they are less an underpinning than the structure itself.
Evidence-based
medicine depends on the systematic evaluation
of multiple references arranged
in hierarchies of reliability.
|
WHAT CONSTITUTES ADEQUATE REFERENCING?
|
|---|
Although an article without references seems naked, such articles
are still
to be found. Many popular science books, and even
editorials in
The
Lancet (which reflect the views of ‘the
journal’), lack
references. When scientific facts are
under discussion the reader is now left
not only dependent on
the integrity of the
writer
5 but also
frustrated when attempting
to explore the subject further. Even when writing
for laymen
a scientist should be able to justify factual statements, and
this
practice characterizes exemplary works such as Stephen
Jay Gould's
The
Mismeasure of
Man,
6 in which
the text is sparingly
yet thoroughly referenced. Admittedly, in certain genres
of
writing—the recording of
anecdotes,
7
narratives,
8 and
even
hunches
9—the
lack
of citations is no drawback and is admissible.
Between the anecdote and the conventionally referenced scientific work is
the uneasy amalgam of science with speculation. Examples include some of the
work on
consciousness10 and
many of Freud's writings, in which references to scientific data are used to
support a creative but inherently speculative and possibly rickety
edifice.
The amount of referencing necessary to support the structure of the text
with appropriate authority is far from clear, and writers take different
views. Even in a single multiauthored
book,11 one chapter
has 252 citations and another has just 4. Usually it is the editorial
architect who determines the structure of the text and how it is buttressed.
Moreover, the authority provided by references varies considerably: one text
may lean heavily on ‘personal communications’ and coy citations to
‘data on file’ held by drug companies; another on hundreds of
peer-reviewed references, offered to provide a deep foundation of historical
precedents. Perhaps the most remarkable feat of underpinning is an historical
paper in which four layers of references have been constructed one upon
another.12 At the
least, this all-encompassing approach demonstrates that the writer has been
truly industrious, and with luck it will avert subsequent correspondence in
the journal about the omission of some key reference (which usually happens to
be the work of the complainant).
|
SELECTION OF REFERENCES: MORE THAN MEETS THE EYE
|
|---|
The references that authors select, and the mode of selection,
often seem
idiosyncratic and haphazard. For example, I suspect
that few clinicians who
cite the popular gate control theory
of
pain
13 will have
read the subsequent papers casting doubt
on the conclusions and clinical
relevance of the
concept,
14,15,16,17
or
the riposte.
18
As a result, it is only the reference in
Science13
that
usually appears in medical texts, and this may be part of the
reason why
the hypothesis took root in
neurology.
15
Apart from ensuring approval from one's peers today and providing
intellectual ‘payment for use of the
material’2 in
the future, the choice of reference can be manipulated intentionally in other
and sometimes nefarious ways to serve the writer and his cause. Thus a writer
may think it prudent to pepper his work with references that will gratify a
colleague, the departmental head, a possible reviewer, or the editor
(‘hat-tipping
citations’).19
Excessive self-citation is a variant. Biases of this kind can be enhanced by
the interventions of editors and reviewers: reviewers may suggest that a paper
of their own deserves citation, and editors may propose inclusion of
references from their journal—with an eye to citation index and impact
factor.20
A writer too can influence the reader as to the relative importance of a
particular reference. Some review articles include a weighting system which,
like restaurant guides, append bullet or similar marks to denote degrees of
approbation.
|
INAPPROPRIATE, UNACCEPTABLE AND UNFORTUNATE USES OF REFERENCES
|
|---|
As well as representing the building blocks whereby the author's
work is
incorporated into the citadel of knowledge, references
are the building blocks
of others and are at risk of being mislaid
or mishandled. The vast corpus of
scientific writings now available
has paradoxically increased the risk that a
key paper will be
overlooked. Mendel's genetic studies were said to have been
ignored
for 35
years.
21 This is
wrong,
22 but might
his observations
have been lost for ever had they first appeared in
Verhandlungen des naturforschenden Vereines in
Brünn23
today?
Failure to cite a relevant reference may be deliberate—perhaps
because of personal animosity or intraprofessional
feuding.19
Sometimes it may also be expedient to suppress an opposing
viewpoint.24
Conversely, references may be fabricated in an allegedly fraudulent manner.
For example, Burt claimed he had previously published certain results of
studies on twins' IQs, but no trace of these publications has been
found.25 Writers of
scientific texts often write for a comparatively select readership that
comprises their peers. Those ‘in the know’ will appreciate the
undercurrents and spot what is being promoted and what is being ignored.
Others have to accept the text at face value.
Certain references become irretrievable since they do not appear on
databases. A paper that has great local impact in a small national journal not
listed on electronic databases may fall by the wayside simply for that reason.
If the language is other than English, the risk of neglect is vastly higher.
Such lack of citation weakens the validity of the scientific
record26 and
meta-analyses are particularly vulnerable to these
biases.27 In
reviews, too, where systematic literature searches might be expected, there is
citation bias related to the author's discipline and country of
residence;28 and
the global dominance of US publications has long been
apparent.1
|
SCIENTIFIC REFERENCES IN THE WIDER WORLD
|
|---|
The discussion so far has largely concerned the reference as
a constituent
of the scientific process itself. But there are
wider issues too. For example,
in the competitive fields of
commerce and publicity, references are the means
of providing
date-stamped proof of publication in the race for priority.
Another
aspect is the counting of citations to evaluate the performance
of a
scientist, a department or a journal. Today citation indices
are the currency
by which scientific output is weighed, grants
are awarded or withheld, and
first-rank researchers are
identified.
22 Not
surprisingly, therefore, their susceptibility to manipulation
and conversion
to ulterior
purposes
19 is all
too evident. And
references can persist or disappear inappropriately. At one
end
of the spectrum is the reference that for some reason is withdrawn
but
continues to be
cited.
29 At the
other end is the attempt
to de-publish work that has already appeared, as
occurred with
the controversial paper on genetic markers in Palestinians and
Jews.
30
|
CONCLUSION
|
|---|
Whether the creation of a master mason, jerry-builder or crook,
references
have for centuries buttressed and underpinned the
construction of ascending
tiers of the scientific edifice and
all that radiates from it. Having
originated as devices to communicate,
attribute, prioritize and protect,
references have also revealed
such human frailties as laziness, jealousy,
self-aggrandisement
and dishonesty. References deserve critical scrutiny
because
they can be put to covert as well as overt uses by individuals,
organizations
and the State.
|
REFERENCES
|
|---|
- Cronin B. The Citation Process. London:
Taylor Graham, 1984: 2, 72
- Ravetz JR. Scientific Knowledge and its Social
Problems. Oxford: Clarendon Press, 1971:245
-59
- Burkle-Young FA, Maley SR. The Art of the
Footnote. Lanham, MD: University Press of America,1996
: 7-13
- Grafton A. The Footnote. London: Faber
& Faber, 1997: 4-5, 9
- Himmelfarb G. Where have all the references gone? In: On
Looking into the Abyss. New York: Knopf, 1994:122
-30
- Gould SJ. The Mismeasure of Man.
Harmondsworth: Penguin Books, 1992
- Nathan PW. When is an anecdote? Lancet1967; ii:607
- Hurwitz B. Narrative and the practice of medicine.
Lancet2000; 356:2086
-9[Medline]
- Lehman R. In praise of hunch backing. BMJ1999; 319:1143[Free Full Text]
- Taylor JG. Paying attention to consciousness. Trends
Cogn Sci 2002;6:206
-10[Medline]
- Williams AC, ed. Patient Care in Neurology.
Oxford: Oxford University Press, 1999
- Junker H. Uber iranische Quellen der hellenistischen
Aion-Vorstellung. Bibliothek Warburg. Vorträge
1921-22. Leipzig and Berlin: BG Teubner, 1923:125
-78
- Melzack R, Wall PD. Pain mechanisms: a new theory.
Science1965; 150:971
-9[Free Full Text]
- Schmidt RF. The gate control theory of pain: an unlikely
hypothesis. In: Janzen R, Keidel WD, Herz A, Steichele C, Payne JP, Burt RAP,
eds. Pain. London: Churchill Livingstone,1972
: 124-7
- Iggo A. Critical remarks on the gate control theory. In: Janzen R,
Keidel WD, Herz A, Steichele C, Payne JP, Burt RAP, eds.
Pain. London: Churchill Livingstone,1972
: 127-8
- Nathan PW, Rudge P. Testing the gate-control theory of pain in man.
J Neurol Neurosurg Psychiatry1974; 37:1366
-72[Medline]
- Nathan PW. The gate-control theory of pain. A critical review.
Brain1976; 99:123
-58[Free Full Text]
- Wall PD. The gate control theory of pain mechanisms. A
re-examination and re-statement. Brain1978; 101:1
-18[Free Full Text]
- Thorne FC. The citation index: another case of spurious validity.
J Clin Psychol1977; 33:1157
-61
- Hemmingsson A, Edgren J, Mygind T, Skjennald A. Use of impact
factors. Lancet2002; 359:173
- Norman JM, ed. Garrison and Morton. Morton's Medical
Bibliograpy, 5th edn. Aldershot: Scolar Press, 1991:42
- Keynes M. Mendel—both ignored and forgotten. J R
Soc Med 2002;95:576
-7[Free Full Text]
- Mendel GJ. Versuche über Pflanzen-Hybriden. Verh
Naturforsch Vereines Brünn 18641865; 3:3
-47
- Horrobin DF. The philosophical basis of peer review and the
suppression of innovation. JAMA1990; 263:1438
-41[Abstract]
- Hearnshaw LS. Cyril Burt: Psychologist.
London: Hodder & Stoughton, 1979:230
- McAuley L, Pham B, Tugwell P, Moher D. Does the inclusion of grey
literature influence estimates of intervention effectiveness reported in
meta-analyses? Lancet2000; 356:1228
-31[Medline]
- Egger M, Smith D. Meta-analysis. Bias in location and selection
studies. BMJ1998; 316:61
-6[Free Full Text]
- Joyce J, Rabe-Hesketh S, Wessely S. Reviewing the reviews. The
example of chronic fatigue syndrome. JAMA1998; 280:264
-6[Abstract/Free Full Text]
- Pfeifer MP, Snodgrass GL. The continued use of retracted, invalid
scientific literature. JAMA1990; 263:1420
-3[Abstract]
- Suciu-Foca N, Lewis R. Anthropology and genetic markers
[Editorial]. Hum Immunol2001; 62:1063[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
TOP
INTRODUCTION
ORIGINS
