BLAISE CRONIN

THE ROLE AND SIGNIFICANCE OF

CITATIONS IN SCIENTIFIC
COMMUNICATION

TAYLOR GRAHAM
Copyright 0 Blaise Cronin 1984

First published 1984 by Taylor Graham,

500 Chesham House, 150 Regent St, London W1R 5FA.

All rights reserved. No part of this publication may be

photocopied, recorded or otherwise reproduced, stored
in a retrieval system or transmitted in any form or by any
electronic or mechanical means without the prior
permission of the copyright holder and publisher.

ISBN 0 947568 0 1 8

Typeset by Quadra Graphics, Oxford

Printed in Great Britain by I a + t e f w W
Contents

Introduction 1
Citation and the primary
communication system in science 11
Science as a social system 16
The need for a theory of citing 25
Context and content analysis of
citations 35
Perspectives on citation 50
Citation ‘in the round’ 60
Towards a synthesis 74
References 88
 Back to Contents Page

Introduction

This essay takes as its starting point the view that

science* needs to be looked at as a large social system. If
we are to understand how science works, then we have to
know something about the norms and values which
guide and constrain the actions of individual scientists.
In other words, it may be necessary to look beyond the
‘storybook’ image of science as a selfless and dispas-
sionate search after truth; a search in which the
individual scientist subjugates personal gain and vanity
to the greater communal good. Recent sociological
investigation has provided us with a ‘relativist’ account
of science derived from observed behaviours, which is far
removed from the ‘storybook’ or mythopoeic conception
(e.g. Knorr-Cetina and Mulkay, 1983).
Science is a social process. The actions and behaviours

‘The term science will be used throughout as a convenient shorthand for the
formalized and institutionalized process of systematic investigation.
knowledge creation and research dissemination, both in relation to the
natural (‘hard’) sciences and the social (’soft’) sciences. T h e undifferen-
tiated use of the term does not mean that some kind of unifying structure is
ascribed to science. As Ellis (1969) notes, science is a complex, fragmented
and amorphous activity - technically, politically and socially. There are,
in addition, different types of scientific research subsumed under the
rubric: basic and applied; theoretical and problem-solving; academic and
techno-commercial. T h e existence of different occupational roles and
motivations is therefore implicit in the use of the term science.

1
of scientists are context-dependent. To understand this
process we need to understand settings and personalities,
whether we are looking at the ways in which scientists
‘do science’, or the ways in which they communicate the
results of their professional endeavours.
The primary communication system is the principal
mechanism whereby the scientific establishment records
and ‘rubber stamps’ the individual and collective
achievements of its members. One function of the system
is to disseminate knowledge; a second to ensure the
preservation of standards; and a third to distribute
credits and recognition to those whose earlier work has
contributed to the development of ideas in different
fields. In some respects the scholarly journal is the
scientific establishment’s ledger of achievement and
roll of honour wrapped in one.
The most common means of bestowing credit and
recognition in science is via citations. With the emerg-
ence of commercially available citation indexes in the
course of the last two decades, the significance of citation
in the professional lives of career scientists has taken on a
new dimension. Citation indexing rests squarely on the
assumption that citations can be treated as units.
Aggregate these units, so the theory goes, and one can
arrive at a measure of an individual’s or group’s
contribution to the growth of knowledge in a particular
subject field. Implicit in this is the assumption that
authors’ citing habits display conformity and consistency.
In the ‘storybook’ account of how science works,
citation is seen as a serious acitivity, governed by a tacit
understanding of how and why authors should acknowl-
edge the works of others. That is to say: an author’s
reasons for citing in a particular way at a particular time
are controlled by an internalized set of norms. This essay
sets out to test and explore the validity of this assumption.
Writing in the 1950s, Merton commented wryly on the
fact that very few sociologists or historians of the
2
twentieth century could bring themselves to treat science
as ‘one of the great social institutions of the time’
(Merton, 1957 : 635). I n the intervening years the
situation has improved with the gradual emergence of
Social Studies of Science as a speciality in its own right
(MacLeod, 1977).
Perhaps the simplest and most economical way of
introducing and explaining the notion of science as a
social system is to quote Merton’s summing up from his
paper Priorities in scientzjic discovery: a chapter in the sociology
of science:
‘Like other social institutions, the institution of science has its
characteristic values, norms and organization. Among these, the
emphasis on the valuc of originality has a self-evident rationale,
for it is originality that does much to advance science. Like other
institutions also, science has its system of allocating rewards for
performance of roles. These rewards are largely honorific, since
even today, when science is largely professionalized, the pursuit
of science is culturally defined as being primarily a disinterested
search for truth and only secondarily a means of earning a
livelihood. In line with the value-emphasis, rewards are to be
meted out in accord with the measure of accomplishment. When
the institution operates effectively, the augmenting of knowlcdge
and the augmenting of personal fame go hand in hand; the
insitututional goal and the personal rewards are tied together.’
(Merton, 1957 : 659)

Several writers, scientists and sociologists of science

have used the ‘Game’ metaphor to describe the practice
of science and to explicate the underlying rules and
regularities which govern the players’ (sic scientists’)
actions (e.g. Mitroff, 1974A). In Mitroffs view:

‘Current written formal accounts of the Game (whether by

scientists in scientific papers or by philosophers of science in
treaties on the Game) bear as little relation to the actual conduct
(plays) of the Game as high school civics texts bear to the actual
workings of the political systcm. Personal clemcnts are integral
parts of the Game. As such they dcservc to be incorporated into
philosophical accounts of the Gamc to a far greater degree than
3
 More specifically, Lcopold ( 1973) identified tlw
‘Citation Index game’ as onc of t h c stratagems employcd
by scientists to incrcasc thcir visibility among thcir pccrs.
The ‘Game’ metaphor has usually been invoked to
counter the ‘storybook’ idca of science as a n idealised,
dispassionate and selfless quest after truth and know-
ledge, in which personal feelings and motivations are
held in check by institutional imperatives. The metaphor
is more than a gratuitous irreverence; its use is intended
to inject a measure of realism into the prevailing
conceptions of what science entails. The problem with
the ‘storybook’ version of how science operates is that it
does not easily accomodate the typical scientist’s striving
after peer group recognition - one of the trophies of the
‘Game’ (Mahoney, 1976). Taking the ‘Game’ metaphor
one step further, Gaston compared competition in
science with ‘a race between runners in the same track
and over the same distance at at the same time’ (Gaston,
1971 : 472). The difference between track events and
science, however, is that in the former there are medals
for the runners-up. In science this is not so: originality
alone is rewarded. Many run, but few are acclaimed.
Those who propound normative theories of science are
faced with the problem of reconciling the idealised
conception of science with the behavioural reality, and of
articulating the relationship between institutional and
personal norms. Merton, for example, has spoken in
terms of value clashes and internal contradictions
(Merton, 1957; hderton, 1963). Our aim will be to
consider the extent to which authors’ citation practices
are norm-regulated, and to what extent they are
influenced by extra-scientific factors. This objective
relates directly to ideas put forward by Kaplan in the
mid-l960s, when he proposed that citation should be
4
viewed as a social control mechanism (Kaplan, 1965).
Anecdotal evidence seems to favour the idea that
many authors cite in reflexive fashion, without necess-
arily dwelling on the implications of the practice. To put
it another way: authors may not be clear in their own
minds why it is that they cite the way they do, or how
ciatation relates to the ideology of science - ‘the norms
and values presupposed in the conduct of science’
( Tranay, 1980 : 191). Citation is therefore a potentially
useful starting point for an investigation into the degree
of congruence between the empirical reality of science
and the ‘storybook’ conception of science as a dynamic,
norm-regulated activity. To quote West:
‘It is commonly believed that persons engaged in scientific
research adhere to a set of moral values representing ideal types
of behavior which facilitate the production of new knowledge.
However, the many listings of these which may be found in the
literature range from the intuitive to, at best, speculative. Few
attempts have been made to discover whether a majority of
scientists actually holds them as they are stated.’ (West,
1960 : 54)
In a recent paper, Gilbert and Mulkay (1982) put
forward the thesis that the language used by scientists is
largely context-dependent. Focusing specifically on
formal research reports, they noted that when scientists
are in what might be termed systematic accounting
mode, the style and character of the language is
deliberately impersonal and restrained (‘storybook’
language). In a series of in-depth interviews they found a
wealth of evidence to support their hypothesis that
scientists move from informal to formal accounting as
experimentation leads to publishable findings. The
authors note that this stylistic transition is hallmarked
by the absence of reference to the dependence of
experimental observation on theoretical speculation, to
the degree to which experimenters are committed to
specific theoretical positions, and to the influence of
5
social relationships on scientists’ actions and beliecs.
Elsewhere, they characterise these two interpretative
repertoires as the ‘contingent’ and the ‘empiricist’.
respectively (Mulkay and Gilbert, 1981 ).
Citation, to use their terminology, is part of the formal
accounting process of science. Yet, like the language of
the scientific paper, it is contingent upon unrecorded and
unexpressed attitudes, relationships and experiences. In
their view, the actions and behaviours of scientists are
shaped by contexts, and can only be properly compre-
hended through a process of social accounting (Gilbert
and Mulkay, 1980; see also Mulkav, 1969).
It was Kaplan (1965) who pointed out that, although
many readers and writers of scientific papers had at least
partial awareness of the technical functions of citations,
little was known about either the operating norms of
actual practice, or the sociological functions of such
norms. It is no exaggeration to say that Kaplan’s early
speculations gave impetus and direction to a debate
which has matured greatly since then. Before considering
further the views of Kaplan, i t may be helpful to set his
remarks in their historical context.
It is quite conceivable that citation would not have
emerged as a serious ‘academic’ issue for sociologists and
historians of science had not the commercial develop-
ment of citation indexing proved so successful (Garfield,
1979; Hall, 1970; Narin, 1976)*. The practical utility of
citation-based information retrieval systems has meant
that citation has in a short time become a topic of great
8‘
A citation index is a n ordered list of cited articles, each accompanied by a
list of citing articles. T h e citing article is identified as a source, the cited
article as a reference. The Science Citation Index (SCI), published by the
Institute for Scientific Information is the only regularly issued citation
index in science. It is prepared by a computer and provides an index to the
contents o f e v e n issue published during ii calendar year of more than 2,000
selected journals. Journals covered by the index are chosen by advisory
boards of experts in each of the topics represented and by large-scale
citation analyses.’ (Garfield, 1970 : 669)

6
interest and importance to the scientific community at
large. Scientists have begun to look more closely and
critically at a practice which, previously, had hardly
been deemed worthy of serious attention. Prior to this,
citation was only mentioned when a serious breach of
etiquette or a priority dispute arose. -4s with many codes
of behaviour, or social rules, it is only when an
infringement occurs that the underlying issues and
assumptions are called into question. However, if
citations are to be used as aids to performance evaluation
and research policy formulation, then the scientific
community will naturally take a keener interest in the
matter. In fact, Garfield was at an early stage warning
against indiscriminate or unqualified use of quantitative
citation data for sociological evaluations, including
personnel and fellowship selection (Garfield, 1963).
An exchange of letters (one of but many over the
years) which appeared in the pages of Scientometrics
(Chubin, 1980; Garfield, 1980A) illustrates the nature of
the polarisation which has tended to characterise much
of the debate of the past two decades. In one corner, we
find Chubin, representing the views of the sociology
camp, who refers to ‘ambiguities in level and inter-
pretation of measurement’ and a ‘meaningless
numerology’ (Chubin, 1980 : 91). Basically Chubin’s
plea for greater caution is an echo of Kaplan’s (1965)
opinion that citation findings should be related to other
kinds of social data on science and scientists. In the other
corner, there is Garfield, concerned that citation
indexing should not be unfairly dismissed, simply
because of certain, recognised limitations inherent in the
raw data, or because of the irresponsible use made of
citation databases by some practitioners and policy
makers.
The launch of the Institute for Scientific Infor-
mation’s (ISI) Science Citation Index in the international
scientific community was accompanied by a flurry of
7
correspondence in journals such as Nature and Science,
concern and cynicism being mixed with cautious interest
and welcome (e.g. Goudsmit, 1974). It was as if the
scientific establishment had not previously recognised
the full import of one of its most frequently exercised
conventions - citation. The development of citation
indexes for science turned the spotlight on a little-
thought-about practice. It was as if a frisson had run
down the collective spine of the scientific establishment.
Authors were for the first time being obliged to reflect on
the deeper implications and significance of a practice
which had belonged to the penumbral world of profes-
sional ritual. Attention was now being focused on the
‘why’ as well as the ‘what’ of citation. Davies was
prompted to wonder whether reference-giving had not
become ‘an absurd convention which scientists unthink-
ingly adhere to’ (Davies, 1970 : 1356), while May ( 1967)
drew attention to the flawed and idiosyncratic manner in
which authors selected and applied citations.
Citation is a subject of interest to various constitu-
encies involved in the practice and management of
professional science. The quotation below, from a brief
but illuminating review of the central issues by Morman,
summarizes the impact of citation indexing upon various
communities engaged in the analysis and evaluation of
scientific activity.
‘Citation analysis in the history of science results from the
convergence of the needs of policy makers of the post-war era, the
maturation of bibliographic citation practice in the scientific
literature during this century, the availability of electronic data
processing, and a school of sociology concerned with the internal
dynamics of the scientific community. It prcsupposcs a model of
science which is simultaneously rational and literary; the
scientist is regarded above all as the producer of scientific
inscriptions, and he is cxpectrd to credit the source of each idea,
method or datum which he uses in consistent fashion.’ (Morman,
1981 : 12)

8
 Having provided the essential historical framework, he
then goes on to identify the central problem:
‘And since citation techniques share no common methodological
assumptions beyond acceptance of quantification and regard for
the citation as a usable primary datum, it is appropriate to look to
interests which reside outside the cognitive aspects of the
scholarly disciplines to which the techniques are applied, fix
much of the reason of their application.’ (Morman, 1981 : 12)
The key phrase here is ‘it is appropriate to look outside
the cognitive aspects of scholarly disciplines‘, a fact
which was first clearly grasped by Kaplan. The scaf-
folding within which much of the most illuminating
sociological research and discussion has subsequently
taken place was effectively erected by Kaplan. Recogni-
sing that citation could no longer be treated as an
inconsequential aspect of the scientific process, he
provided a conceptual framework within which to pose
certain fundamental questions about scientists’ habits
and motivations. At an early stage he concluded that
citation practices were in large measure a social device
for coping with the problems of property rights and
priority claims. This led him to think of citation as an
expression of the communistic ideal in science.
‘The citation is probably among the more important institutional
devices for coping with the maintenance of the imperative to
communicate one’s findings freely as a contribution to the
common property of science while protecting individual property
rights with respect to recognition and claims to priority (Kaplan,
1965 : 181)

In his view, and in the view of many sociologists who

have since been attracted by the subject, the most
important question concerns the degree to which citation
practices are reflective of elements of the normative and
value systems of science. This, of course, is precisely the
sort of question Mitroff (1974A) was raising in respect of
the relationship between the total corpus of scientific
9
 Back to Contents Page
To Next Chapter

norms and the existential reality of scientists’ daily lives.

But for Kaplan’s initial questioning, and the resultant
growth in cross-disciplinary interest, it is conceivable
that citation would have remained a peripheral issue,
and that citations would continue to be treated as
‘necessary trivialities’ (Kaplan, 1965 : 183). Kaplan’s
achievement was to set citation in its social context, and
to elevate the issue to one of serious sociological concern
for those interested in the internal dynamics of science.
To conclude this brief introductory statement of
intent, here is a first quotation from Ravetz’s seminal
work, Scientzjic knowledge and its social problems, which
warns against absolutism in analysing scientific activity:
‘Analyses of the social behaviour of scientists, and of external
influences on scientific research, have assumed the products of
that research to be absolute, and unconditioned by the peculiar
circumstances of their achievement. O n the other hand philo-
sophical analyses of the nature of scientific knowledge have either
been completely abstracted, or have invoked a model of a
working scientist isolated from his environment and traditions.
But a proper analysis of the social activity of science must be
based on understanding of the very special goals of the scientist’s
task; and an analysis of achieved scientific knowledge must
comprehend its character as a social possession, the product of an
historical process.’ (Ravetz, 1971 : 71)
I n Ravetz’s opinion, scientific knowledge is the
product of an historical process, and is shaped by the
‘peculiar circumstances’ of its achievement. The aim of
this essay is to explore the ‘peculiar circumstances’ of the
citation process.

10
 Back to Contents Page

Citation and the

primary 0

communication
system in science

The journal article with its accompanying list of

citations is, and is likely to remain, the universally
accepted medium whereby the scientific establishment
records and reports the results of its investigations. The
appearance of the electronic journal (e.g. Bezilla, 1982)
may herald the eventual demise of the print-on-paper
journal as it is known today, but even if the form is
changed it does not mean that the basic concept of the
refereed journal article will be any less central to the
publication system of the future.
T o understand the significance of citation it is first of
all necessary to understand the nature of scientists’
communication behaviour. The costs and efficiency of
journal publishing have been much discussed in recent
years (e.g. Meadows, 1979; Moravcsik, 1980). However,
despite attempts to introduce change, such as the
American Psychological Association’s (APA) proposal to
establish a multi-million dollar computerized system for
11
the exchange of pre-prints (Boffey, 1970), the appeal of
the traditional approach remains unweakened. The
attractiveness of the medium resides largely in the
system of peer review. Consequently, rather than the
scientific journal being the ruison d’&e of the peer review
system, it is almost as if the roles have been reversed and
the refereeing system has itself become the ruison d’2tre of
the primary publication process. Scientists may be less
than totally satisfied with the scholarly journal as a
dissemination mechanism, but they are deeply attached
to it as a means of preserving a faithful and reliable
account of scientific progress; as a repository of accepted
ideas and beliefs (Royal Society, 1981).
The principal shortcoming of the scientific journal is
its slowness in communicating news of the latest devel-
opments in science. It has long been recognised that for
the majority of scientists and researchers the primary
communication system is an inadequate current aware-
ness system (Garvey and Grifith, 1972). This is part-
icularly true as far as eminent or highly visible scientists
are concerned, as they, invariably, are ‘plugged into’ a
number of informal information exchange networks. A
recent review of the role and importance of ‘invisible
colleges’ within the social sciences noted that virtually all
attempts to take advantage of the news-switching
capability of inter-personal networks in order to build
improved dissemination systems had proved unsuccess-
ful (Cronin, 1982B). There is plenty of evidence to
suggest that most scientists are content to operate, and
participate in, a two-tier system, with informal networks
being used to channel preliminary notifications of
research findings, and the formal system being used as
an archive of findings which have been vetted and
approved by the scientific establishment.
The perceived importance of peer review is consider-
able. Its primary purpose is to ensure the preservation of
standards and to screen out (or at least delay the
12
introduction of) ideas which are antithetical to dominant
paradigms. Maverick ideas, or notions which are, scient-
ifically speaking, revolutionary, are thus effectively
debarred from the official record of science - the journal
archive. It is within this context that citation plays a
crucial role (see, for example, Ziman, 1968).
One of the most lucid expositions of the relationship
between citation and scientific communication practices
has been provided by Whitley:
‘By virtue of its peculiar links with the reward system in science,
the communication system plays a central part in the main-
tenance and growth of science. It is the means by which the
individual scientist relates to the social system: he publishes his
work to gain recognition, and reads the publications of others to
maintain his knowledge. The formal communication system also
forms the basis for the allocation of rewards: instrumental and
consumatory. Thus it is a means of exercising social control . . .
Publication of an article in an archival journal signifies a degree
of recognition for the author, while legitimising the object of
research and methodology.’ (Whitley, 1969 : 219)

The publication process combines reward and recog-

nition. The scientist is rewarded for his efforts by having
publication status conferred on his work, i.e. he receives
the seal of approval of the scientific establishment, and
those whose work he has cited in turn receive recognition
for the part they have played in the development and
furtherance of the citing author’s theories. On the basis
of this interpretation, Ravetz (197 1) introduced the idea
of intellectual property and intellectual property rights.
By describing briefly how the scientific journal came
into being (e.g. Knight, 1976; Houghton 1975), it should
become clear in what sense the notion of intellectual
property and its protection can be applied to the
products of individual scientists’ research efforts. A
short, but particularly good, account of its genesis has
been provided by Ravetz (1971), and is summarized
below.
13
 The development of a proprietary attitude to research
findings was first witnessed in the sixteenth century.
Prior to then, debates on the acceptability and admiss-
ability of atoms of knowledge tended to be conducted as
battles of Looks between conflicting traditions. In this
way, disputes were not seen as conflicts over intellectual
ownership rights, but as clashes between rooted and
favoured traditions. It was in the seventeenth century
that the scientific academies and societies came into
being (the formalization of the earliest ‘invisible col-
leges’), which in turn led to the precursor of the scholarly
journal as we know it today. U p until this development,
however, plagiarism was a very real and worrying
problem. In fact, it was not unknown for results to be
expressed anagrammatically in order to prevent intel-
lectual theft. It was only when communication of results
via Transactions became commonplace that there could
be a relaxing of caution over the protection of intellectual
property rights.
With the foundation of the earliest specialist journals
(which grew out of the German university tradition), the
process of authentication or quality control was com-
bined with that of claim-staking. The flourishing of the
scientific journal can be explained in terms of four
important functions it fulfilled for the scientific corn-
munity: (1) it provided a means of communicating with
interested colleagues; (2) it went some considerable way
towards guaranteeing the quality of results by instituting
the peer review system; (3) it also allowed individual
authors to demonstrate the originality and value of their
thinking; and (4) it facilitated the distribution of credits
among the scientific community.
This gradual institutionalization of scientific publish-
ing managed to combine utility (the promulgation of
ideas and information) with a social function (controlling
the professional conduct of the scientific fraternity)
(Zukerman and Merton, 1971). In Ravetz’s elegant
14
 Back to Contents Page

phrase, the scientific journal allows for the ‘harmonizing

of collective goals and private purpose’ (Ravetz,
1971 300)

15
 Back to Contents

Science as a
social system

Scientists are occasionally, and not without some justi-

fication, accused of working in ivory towers. But however
scientists choose to see their role, they cannot realisti-
cally claim to work in a social vacuum. To understand
the actions and motivations of scientists, as individuals
and as members of a socio-cultural elite, it is necessary to
think of science, not in purely operational terms (or as an
inert aggregation of intellectual accomplishments), but
as a social system, and of scientists, not as individual
actors on a public stage, but as members of a closed
community (subject, of course, to scrutiny and account-
ability), with certain shared prospects, attitudes, beliefs
and conceptual schema. As Trandy put it:

‘Scientific and other cognitive activity necessarily presupposes an

identifiable set of norms and values, an ideology, required to
guide and justify, to judge and evaluate such activity. It is
reasonable to demand that this set should be as far as possible
internally consistent.’ (Tran@y,1976 : 145)

As with any community, science has a more or less

coherent and articulated system of values, which define,
constrain and infuse both individual and group behav-
iours. These rules of association and conduct confer on
science its particular moral character.
16
 ‘The normative framework within which the scientist works
contains components pertaining to cognitive and technical
standards, to everyday practice, and to extra scientific dealings.’
(Robbins and Johnston, 1976 : 353)
That said, however, it seems not unreasonable to
suppose that individual scientists, if questioned on the
nature of science’s normative system, would differ in
their interpretations of how they, collectively and
individually, do and should behave. Merton is widely
recognised as having being the first to elaborate the
norms which underpin science. He identified four
guiding principles: Universalism; Organized Scepticism;
Communism; and Disinterestedness (Merton, 1973).
Universalism refers to the internationality and indepen-
dence of scientific findings. Organized Scepticism
requires that conscientious scientists check and question.
challenge, and hunt falsity wherever it exists. Com-
munism relates to the need for openness, honesty, and
the willingness to share the results of scientific research.
Disinterestedness can be summed up in the dictum
‘science for science’s sake’. Others, notably Barber
(1952) and Storer (1966), have commented on, refined,
or added to these norms. In fact, Merton (1957) himself
subsequently added the norms of Humility and Origin-
ality to his original quartet. The credit, however,
remains with Merton for opening up new vistas on the
social dynamics of science and effectively creating a new
sociological paradigm.
In an expansion of Merton’s initial list, Mitroff
(1974A) identified a corpus of eleven norms. Allowing for
the inevitable variations in nomenclature which arise in
any discussion of norms, this list probably consitutes a
comprehensive inventory of the norms which, in theory
or actuality, guide scientists’ actions. In enumerating the
normative foundations of ‘pure’ science, Mitroff was not
attempting to endorse the ‘storybook’ image of science.
Quite the reverse: his inventory is the necessary prelude
17
to a critical assessment of the logical status of science’s
norms. In his view the evidential bases from which these
norms have been inferred are suspect. Personal experi-
ence, folk wisdom and the collected testimonies of
scientists themselves, he seems to be saying, suggest that
an image of science predicated on these eleven criteria
would be less than accurate. In other words, the
stringency of the norms is such that their exemplification
in actual practice would be somewhat unlikely. They
constitute, in effect, an essentially Platonic conception of
how science works. Mitroff argues that we would be
better advised to regard these norms as inspirational,
rather than instrumental, in the modulation of an
individual’s behaviour. A similar view has been
expressed by Rothman ( 1972).
Storer’s explanation of how and why scientists abide
by the norms of science does not require that we see
scientists as a morally upright elite. He concedes that
scientists’ understanding of science’s normative frame-
work may be ill-formulated and based upon the principle
of self-interest.
‘Scientists subscribe to the norms of science first of all because of
their importance for the continued, adequate circulation of the
commodity in which they are mutually interested. That is not to
say that all scientists arc aware of this relation, but that through
a combination of pcrceiving in a gencral way the necessity for
such patterns of behavior, of training and mutual rcinforcemcnt,
and of their congeniality with thc personalities of many scientists
they have bccome institutionalized. It is the occasional rein-
forccmcnt givcn thesc norms by thc scientist’s awareness of their
relevancc to his work rathcr than to the general goal of science
which . . . accounts for thcir potency. . . ’ (Storer, 1966 : 84)

If the exchange system of science is to operate properly,

then, according to Storer, normative control is essential.
Without norms, science would suffer from a lack of
internal coherence. Just as Ravetz (1971) spoke of
6.
intellectual property’, Storer speaks here of knowledge
18
The norms of science (after MitroffJ 1974)
1. FAITH IN RATIONALITY
2. EMOTIONAL NEUTRALITY: an instrumental
condition for the achievement of rationality.
3. UNIVERSALISM: in science all men have morally
equal claims to the discovery and possession of
rational knowledge.
4. INDIVIDUALISM: expresses itself in science as
anti-authoritarianism.
5 . COMMUNITY: priority rights are reduced to
credit for priority of discovery; secrecy thus becomes
an immoral act.
6 . DISINTERESTEDNESS: men are expected to
achieve their self-interest in work satisfaction and
prestige through serving the community interest.
7. IMPARTIALITY: a scientist concerns himself only
with the production of new knowledge and not the
consequences of its use.
8. SUSPENSION OF JUDGEMENT: scientific
statements are made only on the basis of conclusive
evidence.
9. ABSENCE O F BIAS: the validity of a scientific
statement depends only on the operations by which
evidence for it was obtained and not with the person
who makes it.
10. GROUP LOYALTY: production of new knowledge
by research is the most important of all activities and
is to be supported as such.
1 1. FREEDOM: all restraint or control of scientific
investigation is to be resisted.
as a ‘commodity’, and of its place in the ‘exchange
system’ of science. T h c vommodity which scientists
traditionally exchange is knowledge or information, and
in drawing on the intellectual property of thcir pecrs.
scientists have to enter the exchange system aiid ’pay the
going rate’, so to speak. The currency, to maintain the
economic metaphor, is the ‘coin of recognition’ (Merton,
1968B : 56).
T h e exchange on which the social system hinges Is
information for recognition. T h e fornial record of thesc
transactions is the scientific establishment’s traditional
ledger, the scholarly journal. T h c most common form of‘
currency is the citation. If science appears to displa?
stable characteristics and to follow predictable patterm
in its internal workings, then, so the argument goes, this
is largely due to the existence of the primary com-
munication system. If the norms of science are to be
taken seriously, then, as both Storer and Ravetz seem to
be saying, it is only because the peer review system has
been built into the process whereby scientists oKer their
‘goods’ to the marketplace. This happy fusion of the
dissemination and quality control functions in a bingle
system seems to allow individual scientists to be at once
dispassionate and communistic, and at the same time to
be able to give vent to basic, though not necessarily base,
human urges such as the desire to be recognised and
acclaimed. T h e bonding effect of science’s normative
values has been summarized in the following manner by
Merton:
‘Science, as is the case with the profcssions in gcneral, includes
disintcrcstcdncss as a Iwsic institutional clcmcnt. Disintcrcsted-
ncss if not to be cquatcd \\.it11 ai1i.uisin ; 1 1 ~ l intercstcd aciioii \ \ i t 1 1
cgoism. Such equiva1riicc.s ronfusc institutional and moiiviitioiiiil
levels of analysis . . . For once the institution enjoins disin-
terested activity, i t is to thc interest of scientists to conform on
pain of sanctions and, in so far as the norm has been internalized,
on pain of psychological conflict.’ (Merton, 1968A : 612-613)
 Appealing though this interpretation is on the surface
(it does, after all, have a most plausible internal logic), it
does not satisfy all sociologists of science. In his book,
The subjective side of science, Mitroff ( 1974A) counters by
saying that normative versions of how science works fail
to give adequate attention to the question of normative
ambivalence; something, he argues, which observation
and inquiry would suggest more accurately characterizes
the scientific process. A much sharper distinction
between institutional and motivational levels of analysis
is required if the whole picture is to be painted. Further-
more, he sees emotional commitment (the antithesis of
disinterestedness) as being a ‘necessary condition for the
development of science’, and goes on to say that ‘if
disinterestedness deserves to be considered as a norm of
science, then commitment and bias play a positive role in
science and also deserve to be elevated to the regal status
of norms of science’ (Mitroff, 1974A : 73). His thesis is
that science (and indeed all social systems) are charac-
terized by ambivalences, by normative tensions. As far as
he is concerned, the popularly accepted norms of science
(e.g. universalism) can be matched, norm for norm, with
counter-norms, whose existence is no less conceptually or
empirically valid than the positive or pure norms. Those
who favour the positive norms (to the exclusion of all
others) are supporting an image of science, and of
scientists, in which scientific investigation consitutes the
apotheosis of emotional detachment and neutrality;
something which, in his view, is not borne out by the
evidence to hand (Mitroff, 1972; Barber, 1971).
Mitroffs critique takes in three aspects: (1) the
empirical or factual status of the norms; (2) the
‘regulative status’ of the norms as the embodiment of
ideal standards; and (3) the ‘epistemic status’ of the
norms. With regard to (1) he is concerned to establish
the degree of fit, or correspondence, between the norms-
as-assumed and the norms-in-use; with regard to (2) he
21
wishes to establish the extent to which the norms
describe a to-be-aimed-at level of group performance and
behaviour; while in (3) he is casting doubt on the
argumentative and evidential bases that are used to infer
the existential status of the norms. The third of these is
for him a particularly juicy bone of contention. As far as
he is concerned, the norms have been identified and
selected in the first instance by inductively generalizing
from the cases of individual scientists; from the highly
select writings of the rare, great scientists who were
psychologically motivated enough to write glowingly of
science. An essentially similar point has been made by
Law:

‘. . . there is a case for arguing that in our normative, and

conceptual studies of scientific knowledge, we have gone some
way to creating a self-validating methodological and theoretical
system. We look for norms, we choose certain types of data-
those where we expect to locate the norms, and we go on to
interpret that data normatively. If we fail to find shared norms we
take it that our methods are not good enough, or that the area has
not been institutionalized properly.’ (Law, 1974 : 168)

Mitroffs objective in exploring the subjective side of

science was to challenge the ‘bad metaphysics’ (norms
without empirical foundation) by marshalling couter-
vailing evidence in the form of the actual attitudes and
behaviours of ‘real’ scientists.
If the normative position is accepted without con-
ducting some sort of survey or field study among working
scientists, then it is of necessity based on assumptions, or
theoretical premises, which Mitroff would presumably
label ‘metaphysical’. In the absence of explicit and
universally recognised consistencies in individuals’
citation practices, it is difficult to see how citation can be
defined as a norm-regulated activity. Ravetz (1971) finds
himself in the position of having to employ very
‘unscientific’ terms such as ‘tacit’ and ‘unselfconscious’
22
to provide an explanation of why and how scientists cite
as they do.
‘Since citations must convey some very subtle messages by a very
crude device, the etiquette of each field will impose a code for
their interpretation, whereby the entries and their possibly brief
comments will convey the requisite meanings to those in the field.
Each such node will depend on the character of the problem in
the field, on the types of mutual dependence, and also on the
ruling conception of the right division of intellectual property. In
every case it will be a purely informal, perhaps tacit and
unselfconscious, craft knowledge shared by members of the field.
Thus, in the last resort, this aspect of the system of the protection
of property depends like others on an informal etiquette as well as
on a formal system of rules.’ (Ravetz, 1971 : 257)

It may be that on the surface things happen more or

less as Ravetz has chosen to describe them. Apart from
the occasional flare-up on a question of priority, it is
probably true to say that the majority of scientists see
citation as a low priority issue by comparison with the
many other professional and extra-professional matters
which are competing for their time and attention. Thus,
the views of Ravetz constitute a cosy and convenient
account of what happens, and why. If science, viewed as
a social system, works well as far as the majority of
scientists is concerned, for the majority of the time; and if
the profession has its own tailor-made system of
sanctions; and if there is some evidence that there exists
an inchoate code of citation conduct, then why, the
question might go, should the citation process be
subjected to such close scrutiny?
There is no one, best answer to this question. One
might, however, evince a desire to explicate what is
something of a muddled issue; to test theoretical as-
sumptions against concrete behaviours, or to identify
classes of activity or practice, which if in some way
altered, could contribute either to greater professional
understanding of what citation entails, or to the
23
 Back to Contents Page
To Next Chapter

increased effectiveness of the primary communication

process. Returning to a point made earlier, it is mildly
ironic that science, founded on traditions of quanti-
fication and verification, should be content with an
explanation of citation, an activity central to the
scientific process, which emits a whiff of the meta-
physical. I t may be that prevailing interpretations of
citation are the best that can be hoped for (or that are
required by the majority of scientists), but that is no
reason to discourage further investigation.
There is, in addition, another (rather more pragmatic)
reason why citation warrants further study as a social
phenomenon. This relates to the fact that citations (due
to the development of citation indexing systems) have
become an integral part of the information support
facilities used by scientists on a day-to-day basis. The
commercialization of the citation means that those who
generate citations (publishing authors), those who make
use of them (other scientists), those who process and
package them (the information industry), and those who
mediate in their delivery (librarians and information
scientists) need to have a sound grasp of what citation
entails and signifies. Discussing the quest for universals
in sociological research, Turner drew a distinction
between variables and concepts. As far as citation is
concerned, this distinction has become hopelessly
blurred, variables being treated as concepts.
‘A variable is any category which can be measured or identified
and correlated with something else. A concept is a variable which
is part of a theoretical system, implying causal relations. . .
correlations among variables, of themselves, do not provide a
basis for theory, or even for anticipating future correlations.’
( Turner, 1953 : 610)

24
 Back to Contents Page

The need for a theorv

-
of citing

Metaphorically speaking, citations are frozen footprints

on the landscape of scholarly achievement; footprints
which bear witness to the passage of ideas. From
footprints it is possible to deduce direction; from the
configuration and depth of the imprints it should be
possible to construct a picture of those who have passed
by, whilst the distribution and variety furnish clues as to
whether the advance was orderly and purposive. So it is
with citations in respect of the growth and development
of scientific knowledge; they give substantive expression
to the process of innovation, and, if properly marshalled,
can provide the researcher with an analytical tool of
seductive power and versatility.
The relatively objective data (Mulkay, 1974) provided
by citations have been imaginatively exploited in a
variety of applications. By treating citations as ‘a
quantitative and ‘computer manipulable’ measure of
something or other’ (Singleton, 1976 : 260) sociologists
and others have acquired what Lindsey (1978), some-
what disparagingly, describes as a quality sensor
machine, which can be used, with varying degrees of
confidence, to estimate the quality, impact, originality,
penetration or visibility of individual and corporate
25
performance within and across disciplines (see also:
Lawani and Bayer, 1983).
In any scientific field the existing ‘body of knowledge’
is an accumulation of distilled insight, theoretical
constructs, experimentally derived data and empirical
observations. The published literature of a subject field is
a selective, edited and approved inventory of that
knowledge, and if intelligently schematised it can display
the geneaology of achievement within the field. Citation
indexing, drawing on the inter-document linkages which
authors reveal in their choice of citations, has made this
possible in a way that has opened up new horizons in the
sociology of science. For the historian of science, citation
analysis can be employed to establish the pedigree of
ideas, and to unravel networks of scholarly interaction.
ISI’s multi-disciplinary citation indexes are the most
eloquent articulation of the principles of citation
analysis: they are to scientists what Debrett’s is to the
genealogist.
Citation indexing rests firmly on the premise that a
bibliographic citation is an expression of a relationship
between two documents, the citing and the cited. The
technique does not make explicit the nature of the inter-
document relationship, merely exploits it to index the
literature of science (Blackwell and Kochtanek, 1981) .
Apart from this inspired connection, citation indexing is
nothing other than a straightforward, though com-
mercially sophisticated, repackaging of publicly and
freely available data. The technique and all derivative
applications are based on the assumption that the
citation data can be treated quantitatively. Citation
indexing operates on the assumption that citations have
prima facie equal value (Voos and Dagaev, 1976). In
reality, of course, some citations are ‘more equal’ than
others. That is to say, the presence of a citation may
signify that author A has been influenced by the work of
author B, but it cannot, on its own, say anything about
26
the extent or strength of the influence (Martyn, 1964).
However, if the assumption (each citation is a micro unit
with a fixed value of one) is conceded, then the benefits
flow freely.
There is a considerable body of evidence to suggest
that citation counts correlate with a variety of subjective
and objective performance measures. Clark’s ( 1957)
study of American psychologists was one of the first to
use a battery of quality indicators which included
citations. The most important contributions have been
made by the Cole brothers, who showed that high
citation counts correlated positively with recognised
quality indicators such as honorific awards, Nobel
laureateships and reputational ability (Cole and Cole,
1967; Cole and Cole 1968; Cole and Cole, 1971). Support
for these findings came from Hagstrom (1971), who
correlated citation counts with such variables as quality
of graduate faculty and grants awarded to departments.
In the area of science policy, the National Science
Foundation (NSF) in the United States uses citations as
one indicator of the effectiveness of its funded research
programmes (e.g. National Science Foundation, 1981).
Yet, behind the practical benefits lie unresolved epistem-
ological and methodological questions, which are re-
sponsible for hairline cracks in the conceptual super-
structure (e.g. Edge, 1977; 1979).
The need for more secure epistemological foundations
has been touched on by a large number of writers (e.g.
Sullivan et al, 1977). More specifically, however, Crane
(1972) has argued that the. use of citation linkages
between scientific papers is an approximate rather than
an exact measure of intellectual debts; Porter (1977) that
the correct functional form of this linkage is unknown;
Whitley (1969) that it is difficult to say how much of the
difference in citation rates is due to the intrinsic quality
of the work and how much is due to other factors;
Hodges (1978) that what one knows about a writing,
27
simply because it cites or is cited by another, is almost
content-neutral in terms of the nature of the relationship;
Gilbert (1977) that we do not as yet have a clear idea
about what we are measuring when we analyse citation
data; Dieks and Chang (1976) that the impact of a paper
is not only determined by its scientific significance but
also by such extrinsic factors as locale of the author,
prestige of the journal in which the article is published,
and so on; Meadows (1974) that the evidence concerning
the selection of citations is inadequate; Chubin (1973)
that citation counts are not linear in relation to the
quality of an individual’s research; Dewitt et a1 (1980)
that uncritical use of citation data as a sole, or even
major, criterion could do a great deal of harm; and
Szava-Kovats ( 1982) that formal-numerical citation
analysis does not provide a true picture. More particu-
larly, Gottfredson and Garvey (1980) have noted that the
majority of studies seeking to establish correlations
between citedness and other quality indicators have
generalized from samples of eminent scientists to the
total population of scientists - a criticism endorsed by
Bensman ( 1982).
The welter of criticism has not gone unnoticed by the
Institute for Scientific Information (ISI). In a ‘reply’
paper, Garfield (1980B) indicated that IS1 might begin
to shift from counting citations to counting ‘authors
influenced by’. It remains to be seen whether public
opinion will be affected by the objections raised by
sociologists and concerned scientists, or whether the
considerable practical benefits of citation indexing will
continue to outweigh the residual misgivings concerning
the construct validity of the principle upon which the
system rests.
Ultimately, citation is a private process (Chubin and
Moitra, 1975), albeit a private process with a public face.
The essential subjectivity of the act of citing means that
the reasons why an author cites as he does must remain a
28
matter for conjecture. What is lacking, in Swanson’s
(1977) view, is a rapid and convenient measure for
discovering the nature of the relevance link which the
citing author has established. This conjectural element is
worth pursuing, precisely because the end-product of the
private process (the citation) acquires the status of a
public commodity. It is this privateness which invests
citation relationships with frequent biases (Zunde,
1971). Assumptions are made about the nature of this
commodity, despite the fact that its real significance is
obscured by the secretiveness of the production process.
Textual analysis of the citing paper cannot reveal why an
author has chosen to cite in a particular way, though it
may suggest very plausible reasons. In Mulkay’s (1974)
view, there has been no clear demonstration of the way in
which citations reflect the process of scientific influence.
Further, according to a review study commissioned by
the NSF, there has been little empirical work undertaken
to explore and understand the norms of scientists’
citation behaviour (Gottfredson et al, 1977).
The absence of a satisfactory theory of citing in part
explains why writers on the subject (see, for example, the
opening paragraph of this chapter) often resort to a
metaphor in an attempt to clarify what is not entirely
self-evident. But metaphor, like analogy, has an unfor-
tunate tendency to increase awareness at the expense of
understanding. Logically, the use of citations as a basis
for value judgements should imply that there is a
universally recognised convention among authors.
However, this convention, in so far as one can be said to
exist, displays a remarkable resistance to standard-
ization. It is for this reason that metaphor is so popular,
and useful, in citation exegesis. Ravetz (1971), as already
mentioned, chose to interpret citations as a form of
reward or income, while Gilbert (1977) has preferred to
view them as tools of persuasion. Small (1978), on the
other hand, speaks of citations as markers or symbols.
29
 Because the citation process is subjective and inhospi-
table to standardization it is necessary to rely on
ostensible reasons for citing or reasons which can be
adduced from the context of the citing work (Frost,
1979). It is no historical accident that studies of citation
have shied away from an analysis of the generative stage.
There are many reasons why authors cite the works of
others. In some cases it may well be possible to adduce
the motive, but this is an attributive exercise, and cannot
make the author’s intentions explicit. There will
invariably be a gap between why the author cited and
why we think the author cited. The reasons why an
author cites in a particular fashion may be serious or
frivolous. Weinstock (197 1) has identified fifteen specific
functions of citations, all of which could be classed as
‘serious’. These are listed below.
Reasonsfor citing: Weinstock’s list (1971)
1. Paying homage to pioneers
2. Giving credit for related work
3. Identifying methodology, equipment etc.
4. Providing background reading
5. Correcting one’s own work
6. Correcting the work of others
7. Criticizing previous work
8. Substantiating claims
9. Alerting researchers to forthcoming work
10. Providing leads to poorly disseminated, poorly
indexed, or uncited work
1 1. Authenticating data and classes of fact - physical
constants, etc.
12. Identifying original publications in which an idea or
concept was discussed
13. Identifying the original publication describing an
eponymic concept or term as, e.g., Hodgkin’s disease ...
14. Disclaiming work or ideas of others
15. Disputing priority claims of others
30
 The difficulty with attributing motivation using the
list of reasons drafted by Weinstock is that what at first
glance seem like a perfectly good explanation of an
author’s intentions may in fact be no more than a surface
explanation. At best, the attributive process can only
offer face validity. For instance, a particular citation in
the text may alert the reader to a forthcoming work (one
of Weinstock’s reasons), but unless the reader is wholly
familiar with the subject to which the paper relates, he
cannot know whether the cited paper was the most
appropriate choice bearing in mind the theme and
orientation of the citing paper. Furthermore, in most
cases he will not be able to tell whether personal bias has
played a part in the author’s selection process. In
citation, as in life generally, there are sins of omission
and commission. Full knowledge of the factors which
influence authors in their citation practices would
require omniscience on the reader’s part. For those
interested in the exploitation of citations for information
retrieval purposes this can, in the opinion of Grifith et al
(1977), be either an object of study or a bother. All too
often it is the latter.
Citation is coloured by a multitude of factors, not all of
which have to do with the accepted conventions of
scholarly publishing. Social and psychological factors
play a part, along with subconscious remembering and
forgetting (Aaronson, 1975). Then there are what might
be termed extrinsic factors: the target audience (more
precisely, the author’s perceptions of the readership’s
requirements, capabilities and expectations); the charac-
ter and status of the journal in which the article is to
appear; the scope, format, aims and length of the article
itselc the author’s knowledge of the area in which he is
writing; and his ability, not to say willingness, to use the
appropriate information services and sources associated
with the subject field. There is a bewildering array of
variables to be taken into account, and, as Aaronson
31
concedes, not all of these can be identified or quantified.
May (1967) challenged the orthodox view that cita-
tions give an accurate or exact picture of the intellectual
links between publications. He contends that there is a
considerable amount of ‘deviation’, which results from
‘memory failures, lack of self-awareness, carelessness,
plagiarism of other people’s citations without having
actually used them, the widespread custom of not citing
‘obvious’ sources, and many other causes - all conse-
quences of the simple fact that the author selects
citations to serve his scientific, political and personal
goals and not to describe his intellectual ancestry’ (May,
1967 : 890). Protagonists of the orthodox view, such as
Martino (1971), counter by saying that the force of the
statistical/empirical evidence in support of citation
analysis outweighs the objections. However, even
Garfield (1979) concedes that there are ambiguities
associated with the use of citation counts that prevent
them from being completely definitive.
Recognising the social significance of citation counts,
not least as arbiters in tenure or promotion disputes,
Weiner (1977) offered advice on how the aspiring author
could avoid the ‘dustbin of the uncited’, and noted that
his (not en tirely tongue-in-cheek) suggestions had
already been practiced with a vengeance by Eugene
Garfield (founder and president of ISI), whose self-
citation rate (self-citation being one of the ploys recom-
mended by Weiner) was calculated to be 79%. However,
the ‘footnote fetish’ (Weiner 1977) can perhaps be
excused as a not-too-serious manifestation of basic
human vanity. But, even if this is granted, there remains
the question of deliberate citation abuse.
An example of this is the so-called ‘Burt scandal’ in the
world of psychology. In launching his case against Burt,
Gillie ( 1978) raised numerous questions concerning the
probity of much of Burt’s published research, and in a
subsequent article posed the following questions.
32
 ‘As it is we must ask whether Burt’s pathological streak began
earlier. His attempts to steal credit from Spearman for inventing
factor analysis seem to go back to at least 1937. Burt’s earlier
work on IQ closely shadowed that of H. W. Winch although it
was inferior in quality - was Burt plagiarizing? Why does a 1951
bibliography approved by Burt contain references to non-existent
publications, supposedly published much earlier when he was
with LCC?’ (Gillie, 1980 : 15)

In attempting to answer Gillie’s question one has to

assume that Burt’s objective was to pre-empt criticism,
and to push his ‘knowledge claim’ (Gilbert, 1976) further
towards acceptance status. Burt, presumably, recognised
the psychological fact that citation, at least affirmative
citation, constitutes an appeal to disembodied authority.
The references to non-existent publications were
included to add spurious credibility and weight to his
argument. Had Burt’s work been scrutinised at the time
it was written, without there being any expectation or
suspicion of malpractice, then, presumably, it would
have been possible to classify the various citations using
Weinstock’s typology. This, of course, is an extreme
illustration of dishonest citation practice (but see also
Zirkle, 1954). However, the dividing line between deceit
and what Ravetz (1971 : 257) refers to as the ability ‘to
under-cite without stealing results, or over-cite with the
effect of inflating the value of the property of a colleague’,
or what Merton (1973) refers to as ‘cryptomnesia’
(unconscious plagiary), is somewhat blurred at times. It
is difficult to conceive of all bias and subjectivity being
removed from citation behaviour, although advances in
what is loosely referred to as Artificial Intelligence (AI)
research may ultimately make it unnecessary for authors
themselves to decide where and what to cite in writing
formal research papers. As early as 1964, Garfield had
discussed the possibility of citations being generated
automatically by computer without any involvement on
the part of the author (Garfield, 1965).
33
 Back to Contents Page

To Contents Page

For the present, however, the questions raised by

Jevons concerning the use of quantification in quality
measurement have still not been answered to everyone’s
complete satisfaction.
‘Numcrology . . . is interesting but should bc used only with the
utmost circumspection. In particular caws it could be wildly
rnislcading. Is it in fact likely to hc tnisuscd? Is therc a danger
that it might supplant hcttcr critcria for determining growth
rates, allocating grants and dcciding promotions? If so, should
work on thcse lines be discouraged? O r will it d o to adopt the
ethical neutralist position, doing the work just to get the
information and leaving it to others to decide how, if at all, it is to
bc used?’ Uevons, 1973 : 45)

Note: part of this chapter appeared as The need for a theory of citing, Journal
ofDocummtution, 37( I ) , 1981, 16-24.

34
 To Contents Page

Context and content

analysis of citations

Over the years a number of noteworthy attempts have

been made to peel away some of the ambiguity and
vagueness associated with citations and their use.
Basically, two approaches have been tried: context and
content analysis. In a comprehensive review, Small
(1983) noted that the development of citation classifi-
cation schemes had not been a cumulative endeavour,
with each classifier regarding his problem as unique. The
result has been as assortment of classification schemes,
each addressing the same problem yet designed in virtual
isolation. Nevertheless, Small did find that, despite the
range of approaches adopted and differences in the
literatures studied (from physics to German scholar-
ship), the results displayed certain regularities.
The studies assessed by Small had at least one
important feature in common: they all sought to
illuminate the inter-document relationship implied by
the presence of a citation by devising a classification or
taxonomy based on an analysis of the text surrounding a
citation. Hence his choice of the phrase ‘citation context
studies’. However, the specific objectives of these studies
differed in certain respects. In some cases the aim was to
improve retrieval performance by throwing extra light on
the actual functions of citations, while in others the goal
was to arrive at a deeper understanding of the citation
process in general. Broadly speaking, the studies were
35
either of the applied information science variety, or
sociological in character.
As it happens, both the earliest and most recent
studies in this area are characteristic of the information
science category. Lipetz’s (1965) work was an attempt to
improve the selectivity of existing citation indexes to the
scientific literature. One of the major drawbacks associ-
ated with citation indexes (then and now) has been what
information scientists refer to as the ‘noise’ factor. A
feature of citation-based information retrieval systems
(from the user’s point of view) is that a large number of
unproductive or irrelevant ‘leads’ are included in the
output of a search run on a citation index. Ironically,
citation indexing’s greatest advantage is also the cause of
its major shortcoming. Citation indexes are based on
connections between documents supplied by the citing
authors. There is no need for the intervention of trained
subject indexers. Citations are, in effect, used as descrip-
tors or indexing terms in their own right. It is the
primacy of these linkages which confers on citation
indexing its uniqueness over other secondary infor-
mation services. But, because authors cite in different
ways and for different reasons, not all the connections
prove to be useful. As was mentioned earlier, some
citations are ‘more equal’ and more useful than others.
Lipetz’s pioneering work was an attempt to compen-
sate for the variability in the quality of the output
generated by searches conducted on citation databases.
Being an information scientist Lipetz was keen to reduce
the ‘noise’ level, and (to continue with the jargon of
information science) to eliminate ‘false drops’. Noise
may, as Johansson (1976) feels, be an intrinsic factor of
referencing practice, but given the cost of using on-line
information retrieval systems (such as the Science Citation
Index), most information intermediaries and system users
would strongly support research designed to improve
system cost-effectiveness. This was precisely the thinking
36
behind Lipetz’s work. His aims were as follows:
‘To introduce the means for higher selectivity in a citation index
requires that the bare citing rcference be supplemented with
additional information which, in onc way or another, reflects thc
content of the citing rcfcrcncc as i t rclatcs to typical scarch
problems. In other words, each entry in the citation index must
be expanded to include one or more additional data categories.’
(Lipetz, 1965 : 82)
As others were to discover subsequently, it was often
necessary to include more than one qualifier (or re-
lational operator) to fully describe the function or
character of a given citation. The fact that there are in
Lipetz’s view no absolute and exclusive categories to
adequately and comprehensively describe the relation-
ship between two documents (linked via a citation)
means, in practical terms, that the enhancement effort is
not slight. In the event, Lipetz settled for four principal
relationship categories, with numerous sub-divisions.
His twenty-nine-item classification is reproduced below.
Lipetz actually applied these operators in the Continuity
Index of Information Science Abstracts, but because of the
labour-intensive nature of the work, his ideas were not
adopted by commercial database producers.

Relational Indicators: Lipetz’s list (1965)

GROUP 1. ORIGINAL SCIENTIFIC
CONTRIBUTION O R INTENT O F
CITING PAPER
1. Description of observed phenomena
2. Data transformation
3. Explanation
4. Hypothesis or theory
5 . Calculation from theory
6. Prediction
7. Definition or notation
8. Statement of experimental technique
37
GROUP 2. CONTRIBUTION O F CITING PAPER
OTHER THAN ORIGINAL
SCIENTIFIC CONTRIBUTION
9. Review article
10. Bibliography
1 1. Data cumulation
GROUP 3. IDENTITY O R CONTINUITY
RELATIONSHIP O F CITING PAPER
TO CITED PAPER
12. One or more authors in common
13. Same text
14. Abstract or condensation
15. Erratum
16. Continuation
17. Precursor
18. Inclusion
GROUP 4. DISPOSITION OF T H E SCIENTIFIC
CONTRIBUTION OF THE CITED
PAPER TO T H E CITING PAPER
19. Noted only
20. Distinguished
2 1. Reviewed or compared
22. Applied
23. Improved or modified
24. Replaced
25. Changed the precision (plus or minus)
26. Changed the scope of applicability (plus or
minus)
27. Questioned
28. Affirmed
29. Refuted

38
 Recognising from the outset that the need for intel-
lectual effort in applying the operators would be a very
real barrier to widespread implementation, he advocated
that authors should, ideally, be encouraged to include
the relationship indicators themselves, thereby not only
making the scheme more commercially attractive, but
also inducing authors to reflect rather more carefully on
their reasons for including citations in the first place.
An approach very similar to that employed by Lipetz
has been tried by a research team financed by the
Scottish Education Department (Duncan et al, 1981).
The aim of this project has been to establish a prototype
on-line qualified citation index to the literature of edu-
cational technology. The thinking behind the idea is
remarkably similar to that which motivated Lipetz. The
goal has been to create a citation index in which data
obtained from cited articles are amplified by including
descriptions of the reasons for citing, or of the relation-
ship between the cited and citing paper, with a view to
improving retrieval effectiveness. A longer term aim is to
explore the feasibility of context searching (i.e. using the
surrounding text in which citations are embedded to
provide additional information at the retrieval stage). In
passing, it is worth noting that O’Connor (1982) has,
independently, been exploring the possibility of using
computer-selected citing statements (i.e. using the text
immediately surrounding the reference to the cited
document) to improve retrieval performance.
The approach adopted by the Scottish team is
essentially the same as that used by Lipetz. First, a more
or less randomly assembled list of relational operators
was tested and refined for subsequent use in a series of
simulation exercises. A prime concern has been to come
up with a list of operators based on users’ needs and
perceptions, rather than those of the citing authors. At
the time of writing, feedback trials are being conducted
with a 500-item document collection and an initial
39
corpus of twenty-six qualifiers (see list below).
QualiJied Citation Zndex Project: list of operators (1981)
1. Paying homage 14. Description
2. Background reading 15. Current concerns
3. Historical 16. Development of ideas
4. Bibliographical leads 17. Disputing
5. Narrative 18. Criticism
6. Definition 19. Corroboration
7. Clarification 20. Disclaiming
8. Illustration 2 1. Substantiation
9. Example 22. Similar research
10. Experimental detail 23. Contradictory research
11. Theory 24. Further detail
12. Data 25. Same paper
13. Methodology 26. Statistics

Unlike the two classification attempts just described,

Frost’s (1979) study was centred on humanistic rather
than scientific scholarship. In selecting German literary
scholarship as her testbed, she was hoping to discover
whether citation usages in the humanities were affected
by the greater tendency to rely on non-empirical or
circumstantial evidence.
A scheme was developed in ‘an apriori way’, and, as
with the Scottish project, was then piloted to test its
reliability. Her final Classification (see below) comprised
three broad categories which reflected: (a) whether a
cited work was used as a primary or secondary source;
(b) whether the work was used as a basis for a statement
of fact or opinion; and (c) whether, in the case of
secondary works, the disposition of the cited work was
positive or negative. The scheme was applied to a small
sample of articles and monographs dealing with German
literary scholarship, but, as with the two previous
schemes, the task required considerable intellectual
effort in analysing the citing work and the nature of the
40
corpus of twenty-six qualifiers (see list below).
QualiJied Citation Zndex Project: list of operators (1981)
1. Paying homage 14. Description
2. Background reading 15. Current concerns
3. Historical 16. Development of ideas
4. Bibliographical leads 17. Disputing
5. Narrative 18. Criticism
6. Definition 19. Corroboration
7. Clarification 20. Disclaiming
8. Illustration 2 1. Substantiation
9. Example 22. Similar research
10. Experimental detail 23. Contradictory research
11. Theory 24. Further detail
12. Data 25. Same paper
13. Methodology 26. Statistics

Unlike the two classification attempts just described,

Frost’s (1979) study was centred on humanistic rather
than scientific scholarship. In selecting German literary
scholarship as her testbed, she was hoping to discover
whether citation usages in the humanities were affected
by the greater tendency to rely on non-empirical or
circumstantial evidence.
A scheme was developed in ‘an apriori way’, and, as
with the Scottish project, was then piloted to test its
reliability. Her final Classification (see below) comprised
three broad categories which reflected: (a) whether a
cited work was used as a primary or secondary source;
(b) whether the work was used as a basis for a statement
of fact or opinion; and (c) whether, in the case of
secondary works, the disposition of the cited work was
positive or negative. The scheme was applied to a small
sample of articles and monographs dealing with German
literary scholarship, but, as with the two previous
schemes, the task required considerable intellectual
effort in analysing the citing work and the nature of the
40
citation. Her findings confirmed the hypothesis that the
documentation of opinion is commonplace, and that
factual information is less heavily used than in the
scientific literature, No less interesting was her obser-
vation that two key reasons for using citations in the
scientific literature, to acknowledge previous work or
indebtedness, and to build on previous ideas, were
conspicuously lacking in the humanistic literature.

List of Citation Classajiers: Frost (1979)

A. DOCUMENTATION OF PRIMARY SOURCES
- REFERENCES TO LITERARY TEXTS,
LETTERS, ETC.
1. T o support an opinion or factual statement on the
specific literary author(s) or work(s) discussed in the
citing work
2. T o support an opinion outside the central topic of the
citing work
3. T o support a factual statement outside the central
topic of the citing work
B. DOCUMENTATION OF SECONDARY
SOURCES - REFERENCES TO PREVIOUS
SCHOLARSHIP
1. Independent of approval or disapproval of the citing
author
2. Representing the approval of the cited scholar
3. Representing the disapproval of the citing author
C. DOCUMENTATION O F SOURCES EITHER
PRIMARY O R SECONDARY
1. T o refer to further reading
2. T o provide bibliographical information on a specific
edition
One of the features common to the three classification
schemes described is the underlying (and apparently
unquestioned) assumption that all citations are of the
41
kind referred to earlier as ‘serious’. None of the available
classifications or operators is capable of indicating
whether the function of the citation is serious, essential,
discretionary or trivial in relation to its textual and
intellectual environment. Frost, at least, does give a nod
in this direction, before moving quickly on.
‘. . . while recognizing motivations for citation which are non-
scholarly in nature, such as window dressing, I will be concerned
either with the ostensible reasons for citation or with reasons
which can be adduced from the context of the citing work.’
(Frost, 1979 : 401)
This objection also applies to the classification scheme
devised by Hodges (1978). She ‘intuitively’ identified ten
relationship indicators (see below) on the basis of
analysing a small sample of papers from a range of
subject fields, but did not attempt to demonstrate the
reliability of her list by asking others to apply the
indicators. Another putative classification has been
devised by Peritz (1983) for the social sciences and
related fields, which is, essentially, an elaboration and
refinement of Hodges’ scheme.

Citation Classification Scheme: Hodges (1978)

1 . Evidential 6. Corroborative
2. General informational 7. Specific informational
3. Historical 8. Documentary
4. ‘Sibling’ 9. Methodological
5. Oppositional 10. Corrective

In a series of studies, Moravcsik and Murugesan

(1975; 1979; Murugesan and Moravcsik, 1978) at-
tempted to go one step further with this type of approach
by analysing the quality of and contexts in which citations
were made. As a physicist studying the literature of his
own field, Moravcsik felt that he would be better able to
understand the subtleties or organic nature of citation
42
than researchers with backgrounds in sociology or
library science. Their classification consisted of eight
paired categories (see below). A citation could belong to
more than one of the four groups, but not to both
categories in any one group.

Citation Categories: Moravcsik and Murugesan (1975)

Conceptual Organic
1. or 2. or
Operational Perfunctory
Evolutionary Confirmative
3. or 4. or
Juxtapositional Negational

The first of the four groups (Conceptual or Oper-

ational) specifies what was cited - whether a theory,
concept or idea, or, alternatively, a tool, method or
technique. As Small (1983) notes, this is more a content
than a relationship indicator. Group 2 (Organic or
Perfunctory) was included to distinguish between essen-
tial and non-essential citations - the aspect ignored by
the information science research community in its
classificatory efforts. With Group 3 (Evolutionary or
Juxtapositional) citations could be categorised in terms
of whether the citing paper was building on previous
ideas, or proposing an alternative viewpoint. Group 4
(Confirmative or Negational) focused on the citing
paper’s view of the ‘correctness’ of the cited work.
This conceptually more elastic, yet simpler, scheme
advanced thinking by highlighting the quality factor in
citation. For instance, the function of the Organic vs
Perfunctory classification is to help identify ‘also rans’ -
papers which get cited because of their mere existence
and not necessarily because they contribute to the
progress of science. Group 4,on the other hand, would
enable the reader to identify papers which are held to be
43
wrong, or whose status is disputed (Moravcsik and
Murugesan, 1975).
The authors also introduced the concept of redun-
dancy (e.g. several citations to papers all of which make
more or less the same point, but included to keep
everyone happy in the game of priority hunting).
Significantly, Moravcsik and Murugesan found that a
large number of citations fell into the Perfunctory
category - 41% in an analysis of thirty articles drawn
from Physical Review on theoretical high energy physics.
In their follow-up studies, the authors made a number of
modifications to the initial scheme, and carried out
additional research to determine the degree of inter-
classifier reliability for their scheme. A refinement of
their thinking on the nature of the distinction between
Redundant and Perfunctory citations is worth noting
here.

‘There is a difference between redundant and perfunctory

citations. The latter couldjust stand by itself, and still be judged
perfunctory because it does not contribute to the development of
the citing paper except that it stands in a group with other papers
cited, all of whom make the same contribution to the citing
paper.’ (Moravcsik and Murugesan, 1979 : 168)

The work of Chubin and Moitra (1975) was a more or

less direct response to Moravcsik and Murugesan’s 1975
paper. Although they recognised the value of an ap-
proach to citation analysis based on an inspection of
content and quality, they had certain reservations about
the eight category typology, and recommended that it
should be reworked as a set of mutually exclusive
categories, and then applied to a larger literature sample.
Their alternative was a six-class scheme which allowed
citations to be uniquely classified.

44
Citation Typology: Chubin and Moitra (1975)

Essential Supplementary
I I

Basic Subsidiarv Additional Pcrfunctorv

I
Type 1
1
Type2
I
Type 3
I
Type 4
I
Type 5
I
Type 6

Analysing citations in a sample of articles in high

energy physics, they found that citation practices varied
by form, content and outlet of the article (their sample
included journal articles and letters). More specifically,
they reported a 20% incidence of Type 4 citations
(Perfunctory: corresponding to the Group 2 category of
Moravcsik and Murugesan) . Despite the difference
between the two sets of figures (20% as opposed to
41%), Chubin and Moitra felt that they had replicated
the spirit, if not the letter, of the Moravcsik-Murugesan
analysis. Further support for these findings has come
from an analysis of the literature of business adminis-
tration by Prabha (1983), which found that less than one
third of the works cited by his sample were considered
(by the citing authors) to be essential to their works.
The two types of approach reviewed thus far (the
information science and the sociological/philosophical)
have features in common, but they have somewhat
different objectives. Lipetz and the Scottish project had
as their goal the improvement of citation indexing
systems as information retrieval tools, Moravcsik and
Murugesan and Chubin and Moitra, on the other hand,
wished to answer rather more fundamental questions
45
about the nature and purpose of the citation process.
This divergence of approach (which must not, however,
be over-played) is mirrored in the literature of this rather
specialised, and at times incestuous, field. At the risk of
oversimplifying, it may be useful to distinguish between
the Pragmatists (those whose principal aim is to improve
the quality of existing information retrieval systems) and
the Positivists (those, who, though, not necessarily
decrying the utility of citatidn indexes, would like to see
the fundamental questions relating to the construct
validity of citation tackled in thoroughgoing fashion.)
Apart from the studies already alluded to, there have
been a number of other attempts at citation classification
(e.g. Finney, 1979; Oppenheim and Renn, 1978). The
typology created by Oppenheim and Renn was devised
to help explain the reasons why certain old papers are
still heavily cited many years after their publication.
Although the authors acknowledged the usefulness of the
classification scheme proposed by Chubin and Moitra
(1975), they felt that there was a need for a typology
which would provide a separate classification for those
articles cited as historical background. For the purpose
of their study ‘old’ was defined as pre-1930. The
elaborated classification, which was applied to a sample
of 28 papers dealing with the literature of physics, is
reproduced below.

Classification of historical papers: Oppenheim and Renn (1978)

A. Historical background
B. Description of other relevant work
C. Supplying information or data, other than for
comparison
D. Supplying information or data for comparison
E. Use of theoretical equation
F. Use of methodology
G. Theory or method not applicable or the best one
46
 Of all the classification schemes mentioned, possibly
the most interesting is that produced by Finney (1979).
Her approach consisted of classifying the citations
attached to fifty-one medical articles (containing a total
of 1,115 citations) into seven groups (specified below) on
the basis of their lexical content and their location in the
text, and then searching for differences, variations and
regularities in patterns of use depending upon the type
and character of the citing paper.

Citation ClassiJiers: Finney (2979)

1. Assumed knowledge
2. Tentative
3. Methodological
4. Confirmation
5 . Negational
6. In terpre tation/developmen tal
7. Future research

As with earlier experiments from the information

science ‘stable’, Finney began with the assumption that
all citations were necessary to the author and to the
subject of the research being reported. In a sense,
therefore, her classification scheme represents a retro-
grade step: it ignores the reasons why authors cite some
papers in preference to others, and does not seek to
identify perfunctory or redundant citations. O n the
credit side, she feels that on the basis of her analysis of
the relationship between a citation and its lexical context
it should be possible (in principle) to automate the
process of assigning the seven categories.
Classification, though a popular avenue of explo-
ration, has not been the only approach tested. Herlach
(1978) attempted to find a way round the problem of the
heavy intellectual input required to assign qualifiers and
relational operators. Her research suggested that the
multiple mention of a citation is an indicator of a close or
serious relationship between the citing and cited docu-
ments. She also argued that this link could be mechan-
istically identified. Thus, if multiple mention of a citation
is used as a selection criterion for document retrieval,
there could be an increase in the precision level (i.e. a
reduction in the number of peripherally relevant items
retrieved), but a fall-off in the recall level (i.e. fewer items
will be identified in toto). This sting-in-the-tail effect is
explained by the simple fact that there may very often be
an appreciable yield of important and relevant papers
which have not been cited more than once in the course
of an article.
More recently, Bonzi experimented with another
technique in the hope of improving the predictive
capability of citation links. As before, the aim was to find
a way of selectively identifying strongly paired papers
using citations and structural features of the citing paper.
She likened her approach to a fishing expedition:
‘. . . an exploration of which characteristics of both cited and
citing works may lead us to the development of a reliable tool to
aid in the retrieval of relevant documents through citation
indexes. The hypothesis of the study, if there is one, is that
characteristics of the citing articles are generally better indicators
of the extent to which a citation adds information to the article
than are the characteristics of the cited work.’ (Bonzi, 1981 : 21 1 )

This in itself is not an unreasonable objective, and the

rationale is superficially appealing. However, the defi-
nition of citation relevance is based on the extent to
which the cited work is treated in the citing article. There
may be a positive correlation between the frequency
and/or extent to which a cited paper is referred and the
contribution or influence it has upon the development of
the citing author’s ideas, but this presupposition is not
explored. Instead, Bonzi selects four categories with
which to measure citation relevance (see below), and
then proceeds to link these to a variety of factors relating
48
 Back to Contents Page
To Next Chapter
to the characteristics of the journals, articles, authors
and citations.
Categories o f Citation Relevance: Bonzi (1982)
1. Not specifically mentioned in the text (e.g. ‘Several
studies have dealt with . . .’)
2. Barely mentioned in the text (e.g. ‘Smith has studied
the impact o f . . .’)
3. One quotation or discussion of one point in the text
(e.g. ‘Smith found that . . .’)
4. Two or more quotations or points discussed in the
text.
All of these studies (from Lipetz to Bonzi) have
essayed to increase our understanding of the relation-
ships which exist between citing and cited documents in
the scientific literature (and, in one case, the literature of
the humanities). In each case one aim has been to
improve the reliability of citation indexing, both as an
evaluative tool and as a conventional information
retrieval facility. Yet, each of the methodologies
employed depends on inference rather than motivational
analysis. None of the approaches mentioned is, or could
have been, capable of providing us with privileged
insights into the cognitive processes employed by citing
authors. Kaplan’s (1965) comment that little is known
about the norms and behaviour surrounding citation still
stands, despite the combined efforts of the researchers
just referred to. That, in itself, is not (nor is it intended to
be) a criticism of the quality or integrity of the research
carried out in the area of citation context and content
analysis, merely an observation on the intractability of
the problem. O n the basis of the findings outlined above,
it is difficult to see how we can go beyond saying that
citation is a presumptive indication of influence, or that
the work of a cited author has been influential inasmuch
as it has been noticed and reacted to (Bayer and Folger,
1966).
49
 Back to Contents

Perspectives on
citation

To understand why an author cites in a particular way at

a particular time we would need, to put it crudely, to step
inside that individual’s head. The complex of factors
which characterize an author’s approach to citation
belong to his phenomenal field, and not to the public
domain. Consequently, direct questioning might, at first
glance, appear to be the most sensible and productive
method of trying to gain access to this private world.
However, because an author’s reasons for citing may not
always be ‘pure’ or ‘scientific’ [Mitra (1970) refers to
malpractices of window dressing, padding and the
practice of sprinkling a few citations as an afterthought
as a means of enhancing the respectability of a paper]
this approach may prove to be self-serving (Chubin and
Moitra, 1975). Direct questioning and empathic under-
standing could, conceivably, result in a portfolio of
reasons, strategies and motivations about the citing
process, derived not from textual analysis, but from the
recorded experiences of publishing authors (Hedges,
1978).
Citation needs to be thought of in terms of the citing
author’s own constructs if the prevailing functionalist
interpretations are to be expanded and enlarged upon,
50
and our understanding of citation in the communication
process in science is to be improved. Edge (1977) has
argued that citation analyses of communication patterns
in science have to take as their starting point the
‘participants’ perspective’, because every decision (how
and what to cite) is particular, and because citation and
co-citation analysis, in striving to accumulate and
average, destroy the evidence we need to account for
individual variation. This conception of things is rather
different from the ‘storybook’ image (less uniform, less
glamourous) , which sees author citation practices as
being a controlled expression of a particular normative
position. Ravetz’s (1971) belief that citation is a profes-
sional etiquette maintained by ‘an aggregation of self-
interested policies’ is therefore unlikely to entirely satisfy
those who choose not to assume cognitive consensus (e.g.
Law, 1974).
If our understanding of the citation process is to be
broadened, then it will be necessary to find out more
about what Bavelas (1978) calls the social psychology of
citation. This is an important area for investigation
because it could, ultimately, lead to the development of
more efficient and economical citation indexes. As has
been stated, many writers in the literature of information
science have commented on the ‘noise’ factor in citation-
based information retrieval, which, at heart, is a direct
by-product of authors’ tendencies to include trivial,
perfunctory, redundant, or wayward citations. If authors
can be educated as to the social and informational role of
citations, and can be encouraged to show greater
restraint and care in their choice and placement of
citations, then it may be possible to arrive at a point
where increased standardization and consistency in
citation habits will result in better quality information
retrieval systems.
Despite the wealth of literature on citation there have
been few convincing attempts to explore citation ‘in the
51
round’. Sociologists have for some time questioned the
usefulness and validity of citations as definitive measures
of influence and connectedness in the scientific com-
munication process, and in the past few years there has
been a perceptible shift from the ‘aggregationist’ ap-
proach to citation analysis (i.e. treating citations as
things-in-themselves or units which can be combined or
totalled to produce a faithful picture of impact and
scholarly interaction) to integrated social accounting
(i.e. viewing citation as but one element or activity in the
social construction of scientific facts). The sociology
camp would argue that citation can only be compre-
hended by examining the social conditions which pre-
dispose publishing scientists to cite as they do. Ergo, to
understand the significance of citation it is necessary to
understand the social reality of citation.
In their highly original anthropological account of the
operations, functions, structure and inter-personal
dynamics of a large research laboratory, Latour and
Woolgar put forward the notion that scientific activity
should be equated with a continuous process of literary
inscription.
‘A laboratory is constantly performing operations on statements;
adding modalities, citing, enhancing, diminishing, borrowing,
and proposing new combinations. Each of these operations can
result in a statement which is either different or merely qualified.
Each statement in turn provides the focus for similar operations
in other laboratories. Thus members of our laboratory regularly
noticed how their own assertions were rejected, borrowed
quoted, ignored, ‘confirmed, or dissolved by others. Some
laboratories were seen to be engaged in the frequent manipu-
lation of statements while elsewhere there was thought to be little
activity. The problem for participants was to persuade readers of
papers (and constituent diagrams and figures) that its statements
should be accepted as fact. To this end rats had been bled and
beheaded, frogs had been flayed, chemicals consumed, time
spent, careers made or broken. and inscription devices had been
manufactured and accumulated within the laboratory.’ (Latour
and Woolgar, 1979: 86-87)
52
 In this interpretation of scientific activity and purpose,
the published paper with its quota of citations is the final
stage in establishing the ‘facticity’ of the statements and
ideas forged in the laboratory. In essence, the micro-
sociological view is that citations do not exist in uacuo,
and that a proper comprehension of the citation phenom-
enon and its surface manifestations will only be achieved
by moving the critical gaze from the formal communi-
cation mechanisms (the superstructure) to the social
reality (the infrastructure) which supports the primary
communications system. Hence, the relatively new-
found interest in the ‘work-a-day’ life and personal
motivations of individual scientists. Citation indexing
(applied aggregationism) is of indisputable practical
value to a great many professional scientists, but the
nature of the technique is such that it cannot take
account of this othcr (social) reality. As Martyn (1965)
observed, citation is not a unit, but an event. What
sociologists demand is a fuller understanding of this
event.
Despite the apparent naturalness with which authors
cope with the matter of citation, the fact remains that
authors do not cite in a standardized fashion. Always,
there remains the question: ‘Well, why did so-and-so cite
so-and-so’s paper in such a way at such-and-such a point
in the text?’. As we have seen, functionalist inter-
pretations lack the flexibility to account for the human
element, while textual analysis (i.e. looking at citations
in relation to the surrounding text), though illuminating
in certain respects, does not help close the interpretative
gap.
Recently, Small (1978; 1980) have been advancing the
idea that citations are markers or symbols which denote
particular theories, concepts, proofs, ideas or method-
ologies. The great attraction of this approach is its
pragmatism, but the limitation, in some people’s eyes, is
that it still does not account for an individual’s motiv-
53
ations. To come to grips with the social psychology of
citation there is a need for wide-angled investigation.
Chubin and Moitra (1975) have talked in terms of a
phenomenology of citation, but no research along these
lines has been reported.
Writers such as Merton and Ravetz seem to favour the
notion that citation behaviour is norm-regulated; that
the scientific community adheres to an implicit code of
professional conduct, which guides individuals in the
crucial and delicate matter of dispensing credits. This
communistic interpretation is not to everyone’s liking.
Many sociologists are reluctant to accept the normative
viewpoint - echoes of the ‘ghost in the machine’ (Ryle,
1949) - because, as Cozzens (1981) has observed, the
mechanisms involved in linking actions to particular
positions are difficult to lay bare. Is there a middle
ground between the pragmatism of the aggregationists’
position and the stringency of the microsociological
view? In the final analysis, the answer may be ‘no’, but it
may nonetheless prove instructive to look again at the
various strands of the debate before abandoning the idea
that reconciliation is possible. Indeed, it may also be
instructive to break out of a mode of thinking which
encourages an ‘either-or’ option. A proper (pluralistic)
explanation of what citation entails may mean that we
accept aspects of both (or all) perspectives. It may,
therefore, be counterproductive to think in terms of
‘competing’ theories or perspectives.
As indicated, it is possible to introduce a somewhat
artificial, though revealing, dichotomy into the volumin-
ous literature on the subject of citation. O n the one side
there are those whose attitude and approach can best be
described as pragmatic, or naively rationalistic. The
social reality of citation is not a topic they dwell upon, for
the good reason that they feel questions pertaining to
underlying motives and needs are unlikely to result in
answers or insights which can be readily converted into
54
practical improvements in existing citation indexing
systems. O n the other side of the divide are those (the
positivists) who are unwilling to turn a blind eye to the
more fundamental questions, even if the answers they
seek are not immediately forthcoming.
The way in which the pragmatists deal with (side-
step?) the niggling question of intention is worth noting,
as is the language used. Smith, for example, summarizes
the epistemological problem in the following manner:
‘Although citation linkages do not necessarily reflect social
contacts, it is probable that there is a certain amount of congruence
between documental and social structures.’ (Smith, 1981 : 95)
(italics added).

And Small, though echoing these sentiments:

‘The reasons and motivations for citing appear to be as subtle
and as varied as scientific thought itself, but most references do
establish valid conceptual links between scientific documents.’
(Small, 1976 : 67) (italics added).
attempts to show that the difficulty has been exaggerated
by arguing that citations operate as symbols representing
identifiable and traceable concepts in the literature of a
given field.

‘The concept symbol interpretation of citation practice does not

contradict the functional, social or political interpretations, but is
complementary to them. Whether the motive for citing a work is
politically conditioned or merely haphazard (for example, adding
references to a paper after it is written, where they ‘fit in’), the
work must be associated with specific language in the text and
cannot be appended without some explicit or implicit context’.
(Small, 1978 : 337)

In his most recent work Small claims that co-citation

links can be ‘rewritten in sentence-like form’ with the
result that citation-derived networks can be ‘translated
into patterns of interlocking sentences’ (Small and
Greenlee, 1980 : 300; see also Cozzens, 1982). This idea
55
has been taken up by McInnis, who see the citation as a
metaphor:
‘A bibliographic citation, as a symbol for a concept, functions as a
metaphor for the cognitive content of a specific publication. That
is, the relationship between the cited document and the concept it
symbolizes is metaphoric.’ (McInnis, 1982 : 56)
If particular citations are associated with identifiable
concepts, ideas or methods, then it is possible, in Small’s
view, for authors to engage in a meaningful dialogue
because the cited works, within the context of a
particular paradigm, have acquired a standard or
conventional interpretation. It is the putative existence
of such group vocabularies, or semaphores, which gives
citation its objective character. Nevertheless, even
though this approach gets close to eliminating some of
the long-standing uncertanties surrounding citation, it
does not, as Morman has noted, resolve the central issue:
‘Small presents the concept symbol thesis explicitly in response to
demands for a theory of citing behaviour. But by dismissing other
possible explanations for choice of citation (e.g. the desire to
persuade, to curry favour, to publicize, or to avoid offending) as
inadequate Small avoids dealing with interactions between
symbolization and other factors. He is thus left with a wide gap in
his explanation of citing behaviour - why do particular authors
choose particular cited works as symbols for particular concepts?’
(Morman, 1981 : 11)
The ‘queries and caveats’, to use Porter’s (1977 : 257)
phrase, remain unanswered, despite Small’s attempt to
root citation in a theory of interactive symbolism. In
Porter’s view, citations are:
‘. . . measurable indicators, logically linked to interesting theor-
etical variables (e.g. scientific productivity, communication units
or whatever), but the correct functional form of this linkage is
unknown. At best, the errors in measurement thus introduced
may not be serious, but at worst the indicator may appear
inappropriate - ‘construct validity’ is lacking.’ (Porter,
1977 : 263)

56
It is this lack of construct validity which has united a
substantial body of sociologists of science to contest and
probe some of the claims made on behalf of citation
indexing. The split between the naive rationalists and
the positivists is quite starkly revealed in the writings of
two of the principal protagonists already mentioned,
Small and Edge, respectively.
At the heart of the interpretative rift is the question of
unknown motivations. Citation is not a transparent
activity: the process is not amenable to scrutiny. To
quote Miller (1962 : 71): ‘It is the result of thinking, not
the process that appears spontaneously in conscious-
ness’. And so it may be with authors when it comes to
citing the works of others.
Interpetation is impeded by our virtual ignorance of
authors’ motivations. Also, if we accept Miller’s thesis,
then attempts to expose personal motivations are likely
to founder, for the simple reason that it is the results, not
the process leading up to citation selection, which
authors are likely to recall if questioned directly on their
practices. We are, in fact, forced into a position where
we have to accept Wittgenstein’s aphorism: ‘What we
cannot speak about, we must pass over in silence’
(1961 : 151). In some respects this appears to be the
conclusion reached by Ravetz (1971), who was forced to
view citation as an instance of tacit knowledge - a craft
skill osmotically acquired. It may be, as we said earlier,
that this is the most satisfying solution available, and in
this respect the following quotation from Hudson is
worth bearing in mind.
‘Far from being a porridge-like and irrational mess, the mental
processes of which we are unaware often prove to possess a high
degree of structure. Indeed, some of the most precise thinking we
do seems to be thinking to which we have little access, and over
which we exert a minimum of control. Secondly, the distinction
between conscious and unconscious is in any case far from clear-
cut. There exist massively influential forms of thinking that are

57
 strictly neither conscious nor unconscious, but tacit.’ (Hudson,
1975 : 72)
The dichotomy referred to, naive realism versus
positivism, does not only refer to participants’ percep-
tions of citation. By implication, it reflects the different
conceptions of science held by the two groups. For the
naive realists science is likely to be thought of in terms of
institutionalized standards and purposes, whilst for the
positivists science is more likely to be conceived of in )
terms of particularistic accounts of behaviour. Krohn has
captured the essentials of the alternative Weltanschauungen
most effectively:
‘Several authors have used the economic metaphor to describe
current science. At least one finds the use of economic metaphor
frequent among scientists themselves. ‘Production and repro-
duction’, ‘investment of resources’, ‘symbolic capital’ etc. have
made insights available into relations among resources, the
exchange and conversion from one kind of research resource,
such as scientific ‘credit’, to another, research or publication in
journals, etc. That is, the economic language has allowed the
juxtaposition and interpretation of elements otherwise segregated
into ‘intellectual’, ‘professional’, ‘institutional’, ‘cognitive’, etc.,
categories. Less explicitly, economic language has allowed the
transition of the semi-mystified or sacred language of institution-
alized science into a secular language. Thus behind noble
motives, ‘love of truth’, or ‘curiosity’ can be detected the down-
to-earth motives of seeking recognition, prestige and career
advancement.’ (Krohn, 1980 : xviii)

Looking at the issue from a slightly different angle,

how is it that the subjective, intensely personal activity of
creative science can be transmuted into objective,
impersonal knowledge? How is it that citation (a private
process) can be accepted as science’s principal account-
ing mechanism, upon which so much depends? The most
successful attempts to answer questions of this kind have
been made by a number of the ‘new wave’ sociologists of
science (e.g. Edge, Mulkay, Gilbert, Latour and
58
 Back to Contents Page
To Next Chapter
Woolgar, to mention some whose work we have already
referred to). The microsociological approach breaks with
scientometric and quantitative traditions by preferring to
focus on the quotidian reality of science as it is practised,
Science, in other words, is taken off its pedestal, and the
process of knowledge creation viewed, less as a mechan-
istic, well-oiled sequence of actions, than as a fortuitous
harmonizing of individual and institutional objectives.
The new conception of science has been described in the
following way.
‘A high/distant/formidable image is replaced by a low/close/
everyday one. An abstractlspirituallperfect image is replaced by
a physicallimperfect image. And an impersonal/universal/
permanent image is replaced by an image of personal work
carrying the marks of a craftsman and his time and location of
work, which is assumed to be of temporary utility and value.’
(Krohn, 1980 : xii)
Although an interpretative dichotomy exists, some
degree of harmonization may, nonetheless, be possible.
As Cozzens (1981) says, there are a number of perspec-
tives which can be adopted (she speaks of the normative,
the interpretative and the symbolic), but there is no
necessary reason why one should be preferred at the
expense of the others. It is her contention that the
various theories need not be viewed as being in compe-
tition, and that in the course of writing a paper a
scientist’s actions may be consistent with aspects of the
various perspectives. Furthermore, to refer again to
Cozzens (1981 : lo), there is a need for research into
citation which ‘cuts across the traditional theoretical
orientations of the sociology of science and involves
cooperation among people with different methodological
inclinations’.

59
 Back to Contents Page

Citation ‘in the round’

Citation needs to be considered in relation to the

habits, attitudes, experiences and expectations of four
loosely defined groups: the Quality Controllers; the
Educators; the Consumers; and the Producers. Together,
these four groups shape, define and exemplify through
practice the norms of citation, in so far as these can be
said to exist. Yet, citation tends not to be discussed in
terms of the behaviours and perceptions of these four
constituencies. Instead, citations are spoken of as if they
were things in themselves (Dinge an sich). Impercep-
tibly, the humble citation has undergone a process of
reification. The four-fold categorisation proposed above
may prove helpful in developing a more rounded and
balanced appreciation of what citation entails.
The Quality Controllers are those who act as the
scientific community’s invigilators, i.e. journal editors,
referees and editorial board advisers. It is this select
group which vets submitted manuscripts and, by exten-
sion, the attached lists of citations. The attitudes and
practices of this group vis li vis citation are little known.
Both Garfield (1977) and Price (1964) have commented
on the need for, and desirability of, greater control and
consistency in refereeing practices as far as citation is
concerned.
It may be that citation cannot be defined more
60
satisfactorily other than in terms of tacit knowledge (a
skill which is acquired and exercised largely without
conscious reflection), and it may be that fledgling
authors model their citation behaviour on that of
established authors in their particular field, but this
‘learning through modelling’ theory remains untested.
Likewise, the role played by university teachers and
mentors in socializing students into the conventions of
the field is something of a ‘black box’. The Educators
are, therefore, an important group to consider in any
analysis of citation.
The Consumers are those at whom citations are
directed; in other words, the readers of the scientific
literature and members of the attentive scientific com-
munity. In some respects, little is known about their
requirements, expectations or reactions to citations.
The fourth group, the Producers, are those who are
actively involved in the scholarly publication process.
The practices and habits of this group effectively define,
shape and exemplify the prevailing (and presumably
accepted) standards in a given domain.
It is not for a moment being suggested that these four
groups constitute mutually exclusive populations;
membership of one does not necessarily preclude
membership of another, and in many cases members of
one group are defacto members of a second or third, or
may after successful apprenticeship become so. Never-
theless, this crude classification provides an articulated
structure for a more sensitive analysis of the citation
process.
In an article entitled The outlook ofjournal editors and
referees on the normative criteria of scientzjic craftsmanship,
Lindsey and Lindsey (1978) examined the criteria used
by editorial board members in a number of social science
disciplines in appraising submitted manuscripts. Their
aim was to lay bare the technical and artistic norms
against which authors’ submissions were assessed. Res-
61
pondents were asked to react to a twelve-item list of
criteria (see below) so that the relative importance of the
various dimensions could be established. Citation was
not one of the criteria used, though it could perhaps be
swept under ‘scholarship’ or ‘empirical evidence’.
Journal editors’ normative criteria: Lindsey & Lindsey (1978)
Value of findings Grasp of design
Theoretical relevance Scholarship
Creativity of ideas Empirical evidence
Sophistication of methods Relevance to journal
Ethical sense Value to social life
Entertainment quality Reputation of author
In reviewing the role and evolution of the scholarly
journal, we made the point that the conventions and
procedures for reporting research findings were so
designed as to minimize the possibility of suspect or
fraudulent data being accepted into the journal archive
and dispersed throughout the scientific community.
Within this scheme of things journal editors and the
supporting ranks of advisors (‘the institutionalized
interpreters of normative criteria’ Lindsey, 1978 : 17)
operate as quality controllers. Senior members of the
scientific establishment undertake quality control work
because they see this as a necessary consequence of their
belief in the norm of communality. In Meadows’ view:
‘This will bind them to the negative task of ensuring that a paper
contains no demonstrable deviations from normal scientific
standards. . .’ (Meadows, 1974 : 41)
But he also states that delailed criticism over and above
this is somewhat rarer. Indeed, the refereeing process by
its nature is not entirely objective or consistent. An
ingenious study by Peters and Cecci (1981) of inter-
referee reliability, which involved the re-submission of
accepted articles to the same journals, clearly demon-
strated the problems associated with manuscript eval-
62
uation, and called into question some of the basic
assumptions made about the journal refereeing system in
general (see also Whitley, 1970).
It may therefore be legitimate to see citation as the
‘Achilles’ heel’ of the journal refereeing system: one
particular area in which an author’s judgement and
habits are perhaps not always subjected to close scrutiny.
Price (1964), for instance, recommended that journal
editors and referees should summarily reject bibli-
ographies that are either insufficient or padded. Others,
notably Cleverdon (1970) and Garfield ( 1977), have
expressed essentially similar views. In practice, matters
are neither so simple nor so easily remedied. One very
good reason for this is that the literature, even in a
narrowly specialized field, is too voluminous for a single
author to utilize fully (Kochen and Perkel, 1978). If this
is true for authors, then it is even more likely to apply to
editors and referees. Lawani (1977A) also makes the
point that it can be a problem finding-enough good
referees.
One of the sharpest criticisms of citation practices was
delivered by Thorne (1977), an academic with thirty
years editorial and publishing experience in psychology.
He concluded his article, The Citation Index: another case of
spurious validity with the following comments:

‘In my tenure of editorial ofice, I came to the conclusion that

reference citations were idiosyncratic and non-evidential. I
deliberately limited authors to citing only a few directly relevant
citations in reference to the specific points at issue. Such editorial
standards cannot avoid influencing citation rates, and such
factors should be considered in evaluation studies.’ (Thorne,
1977 : 1161)

His experience suggested that authors employed a

variety of stratagems to manipulate citation frequencies.
These are summarized overleaf:

63
Authors ’ citation stratagems: Thorne’s list (1977)
1. Serial publication (division of a single research
project into many parts, each reported separately)
2. Multiple publications (minor variations of a project
report submitted to differentjournals)
3. Hat-tipping citations (acknowledgement of eminent
figures)
4. Over-detailed citations
5. Over-elaborate reporting
6. Evidentiary validity (citations can be selected to
support any point of view)
7. Self-serving citations
8. Deliberate premeditation (conscious playing of the
citation game) t
9. Searching out grant funding (identifying currently
popular research trends)
10. Funding support for publications (the publication of
luxurious research reports to attract attention)
11. Editorial preferences (authors seek to identify pre-
ferred topics and styles ofjournals to which they
submit)
12. Citations as projective behaviors (citations as reflec-
tion of author biases)
13. Conspiratorial cross-referencing (the ‘you scratch
my back and 1’11 scratch yours’ syndrome applied to
citation)
14. Pandering to pressures (citing works because it is felt
that the reading public requires, or expects, them to
be cited)
15. Editorial publication policies (discriminatory biases
in editorial policies re. selection and rejection)
16. Non-recognition of new authors
17. Intra-professional feuding
18. Obsolete citations
19. Political considerations (citing the ‘party line’)

64
It is instructive to think of Thorne’s list as the obverse of
that produced by Weinstock (1971). Both lists offer a
variety of explanations as to why authors cite as they do,
but whereas one (Weinstock’s) is closely aligned with the
‘storybook’ image of science, the other (Thorne’s) is less
reverential and closer to the thinking of the micro-
sociologists.
In an effort to explore the views of Quality Controllers,
Cronin (1982A) carried out a survey of journal editors
and editorial board members in a sample of English-
language psychology journals. The results obtained tend
to support Cozzens’ view that the different perspectives
on citation (e.g. the normative and the interpretative)
are artificial. Respondents’ replies to Cronin’s question-
naire did not reveal an uncomplicated split along this
notional binary line. To some extent this lack of
consensus suggests that editors have widely differing
outlooks on the significance, importance and functions of
citations. In part this can be explained by the lack of
explicit guidelines for authors on how and why they
should cite; it may also, to some degree, be explained by
the fact that citation, alongside other factors, such as the
presence of demonstrable errors, insignificant results,
lack of originality, or lack of clarity, plays a relatively
small part in manuscript evaluation (Gordon, 1979).
Although there has been some discussion of ways to
improve manuscript evaluation (e.g. Bowen et al, 1972;
Wolff, 1973; Scott, 1974) the question of citation
standardization has received little explicit attention.
Entrants to the world of career science and research
have to learn to accept that there are certain ways of
doing things (such as citing the works of others), and
that certain rules, traditions, etiquettes and codes of
conduct determine the limits of acceptability for indi-
vidual actions. The newcomer is a little like the traveller
in a strange land who has to adjust to different cultural
sets and expectations. ‘Doing science’ is not just a matter
65
of having an enquiring mind or high academic ability: it
also has to do with playing the ‘Game’ according to the
written and unwritten rules. Training and preparation
for a career in science goes beyond the acquisition of
technical skills and competencies; it requires that a
student be socialized to the mores and behaviours of the
discipline he is preparing to enter. There are, so to speak,
certain rites of passage to be undergone before the
transition from apprentice to craftsman can be con-
sidered completed. T o quote Fisch:
‘Becoming a scientist involves not only an educational process
and an accumulation of knowledge and skills but also in part a
filtering process through which one goes in the course of
professional development.’ (Fisch, 1977 : 289)
Much of the socialization process, what Bourdieu
(1972 : 45) terms ‘pre-reflexive adherence to the tacit
pre-suppositions of the field’, will take place on-the-job,
but university provides an early and extended intro-
duction to the academic and research ethos. The extent
to which university prepares neophytes and encourages
enculturation will depend on the attitude of faculty, and
how they see their role as socializers.
This process has been most aptly described by
Hagstrom in his book The scientzjic community:

‘The effects of scientific socialization are reinforced by a highly

selective system of recruitment. Of the fraction of the population
who enter college, only fractions of those interested are permitted
to graduate in the exact sciences and enroll in graduate school.
Attrition in graduate school tends to be high, and only the more
competent and highly motivated students obtain the doctorate.
Among those who do obtain doctorates in science, only a fraction
are permitted to enter careers in basic research; the rest become
teachers, administrators, and applied scientists. Basic scientists,
then, are a highly selected and socialized elite group. The entire
socialization and selection process tends to produce scientists
who are ‘self-starting’ and ‘self-controlling’. A common view of
the organization of science, held implicitly or explicitly by most

66
 scientists, is that these individual characteristics are suficient to
account for conformity to scientific values and norms.'
(Hagstrom, 1965 : 11)
However, he also notes that the effectiveness of the
socialization process will depend, to some extent, on the
concreteness o f the norms which operate in a particular
community of students and scholars. In his view, there is
less likelihood of deviation from norms when those norms
are specified for a concrete set of practices. Thus, we
would expect physical scientists to be less likely to
deviate from the norms of science and scholarship than
social scientists or humanists.
Cronin (1983) has reported a study of current citation
training practices in British and American university
psychology departments. His findings suggest that the
approaches used differ considerably from department to
courses on research methods and the ethics of citation) to
the highly informal (e.g, co-authorship involving faculty
and students). Overall, however, Educators appear to be
concerned to instill an appreciation of correct biblio-
grapic format, but less concerned with explaining the
underlying principles. Cronin concluded that much more
could be done by Educators to improve the general
quality and consistency of students' citation practices, a
view echoed by a number of other writers (e.g. Basefsky,
1982; Lawani, 1977A) including Voverene (1981) who
advocated that the ethics of scientific work (in particular
the underpinning rationale of citation) should be taught
as a matter of course in scientific training programmes.
We do not know why an author cites in a particular
way at any given moment. We may be able to guess
(even guess correctly), but we cannot be certain that we
have adduced the real reason or motivation. It may be a
truism to state that citation is private event, but the point
needs to be made, if only because it serves as an antidote
to the infectious appeal of the normative position, which,
if accepted at face value, would commit us to the view
67
that the actions, behaviours and statements of scientists
are governed and directed by a (largely unseen and
unspoken) set of universalistic and communistic prin-
ciples. In an ideal world, citations would be taken as
objective markers or symbols; would be seen as being
employed in consistent fashion, and would be susceptible
to consistent analysis and interpretation. As we have
tried to show, this view is, if not suspect, at least
incomplete.
Theoretically, if citation were dependent solely upon
the character of the citing paper and its objective
relationship to existing literature and scholarship, and in
no way dependent upon the perceptions, needs, atti-
tudes, prejudices, background and erudition of the
author, then the occurrence of citations within a paper
should be predictable. Recognising this fact, Garfield
( 1965) discussed the possibility of citations being gen-
erated automatically by computer without any direct
involvement on the part of the citing author. He was, in
effect, speculating on how the citation process might be
‘desubjectivized’. The intuitive hypothesis suggests that
even if two equally well-informed individuals were
presented with a citation-less paper and asked to ‘dress’
or ‘prime’ it with citations, that the two sets of
recommendations would differ. Even if only asked to
indicate where citations were required, without specifying
what they should be, it is highly improbable that there
would be a perfect match between the two lists.
However, even if this supposition could be demon-
strated, it would not necessarily invalidate the normative
position. The normative view does not have to be
rejected if anything less than 100% adherence to the
supposed norms is registered. As Mitroff (1974A)
observed, the norms of science can have regulative
status, insofar as they embody the ideal standards of
rationality. The fact that prevailing standards fall short
of the ideal does not mean that we cannot, or should not,
68
continue to aim for the ideal. The standards remain the
rational ideal, even if our behaviours do not always
exemplify them.
Bibliographic citations have been described variously
as symbols, markers, metaphors and signposts. Gen-
erically, citations act as signs: they denote particular
works and seem to imply some sort of relationship
between the citing and cited work. The study of signs is
known as semiotics, and can be thought of in terms of
three levels (Cherry, 1978 : 223).
Syntactics: signs and their relation to other signs
Semantics: signs and their relations to the outside
world
Pragmatics: signs and their relations to users
As a rule, citation analyses are concerned with the
first of these; that is to say, they explore and articulate
patterns of connectedness between signs (citing and cited
document elements). Small’s (1978) concept marker
theory is perhaps the closest to the second of Cherry’s
categories, in that it seeks to establish equivalence
between a sign and a theory, model, concept or
methodology. Cronin (1983) has described a number of
experiments which attempted to explore the relationship
between signs (citations) and users (readers) of the
literature - what Cherry calls pragmatics.
Cronin’s strategy was to distribute unpublished
journal articles denuded of their original citations to
carefully selected samples of readers, asking them to
suggest where citations were required. The innovative
feature of this approach lies in the fact that it shifts the
focus of interest from the author (producer) to the reader
(consumer). Traditionally, citation studies have taken as
their starting point the lists of citations attached to
published papers and proceeded to establish networks
and connections based on such data sets. This means, in
practice, that little attention is given to what the reader
69
expects or requires. The result, inevitably, is generally a
selectively biased picture of the citation process. And yet,
citations are a public commodity (and scholarly journals
their common carrier) used by various communities of
interest.
The results of Cronin’s infill experiments suggest that
the ways in which different groups perceive the need for
citations display an underlying regularity. Although the
response levels in his study were poor, there did
nonetheless appear to be some evidence of a shared
understanding as to how and where citations should be
affixed to a scholarly journal article. However, Cronin
was unable to show that certain trigger words or kernel
phrases were unambiguously associated in the readers’
minds with the need for citation (the idea put forward by
Finney, 1979. Had this been demonstrated, then it would
have been possible to think of citations being generated
automatically without any direct involvement on the
part of author (Garfield, 1965).
The great mass of the literature on citation is
producer-oriented, i.e. it consists of quantitative analyses
of the citation lists generated by individual authors, by
particular research communities or networks of scholars,
or of the citation lists attached to particular journals or
clusters of journals (Hjerppe, 1980). Viewed as a social
system, science is international and cosmopolitan.
Adherents of the normative view would argue that,
despite the many differences between individual nations
and individual scientists, the actions of scientists in citing
are guided by supra-national and supra-personal con-
siderations.
However, Cronin’s (1981) study of transatlantic
citation patterns in educational psychology suggests that
this may not be the case. Even though two or more
groups of readers of a particular paper might propose the
same number of citations, and agree on the locations for
those citations in the text, it would not follow that their
70
choice of works (and the nationality of those works)
would necessarily be the same. For instance, British
authordreadem might propose a preponderance of
British citations, and American scholars a pre-
ponderance of American works, or individuals within
either group might show favouritism towards colleagues
or fellow ideologues.
Citation analyses of communication within psychology
(Cronin’s field of study) are not uncommon. Since Cason
and Lubotsky’s (1936) pioneering study of journal
interactions there have been several attempts to evaluate
journal impact and connectedness, using both subjective
and quantitative approaches (e.g. Jakobovits and
Osgood, 1967). One of the best known quantitative
studies was that of Xhignesse and Osgood (1967), which
established a matrix representation of flow and inter-
dependence within a 21 journal network. More recently,
Pinski and Narin (1979) refined Xhignesse and Osgood’s
matrix display idea to produce a set of citation influence
measures for a variety of psychological sub-fields. This
did not, however, include educational psychology, the
area studied by Cronin. Cronin analysed citations linked
to papers appearing in two educational psychology
journals (one British, the other American) and found
that the profiles of the two journals differed markedly,
with US authors displaying a heavily ingrown citation
tendency (i.e. on average US authors cited 95%
American works, while British authors cited rougly 40%
American authors).
There are many reasons why authors display inward-
looking citation tendencies: proximity or sense of loyalty
to local colleagues; institutional affiliation; ease of access
to home-produced literature; vague political or cultural
pressures; linguistic isolation. However, as far as edu-
cational psychology is concerned, there are in the main
shared concepts, experimental practices and a common
knowledge base, all of which would be expected to
71
facilitate healthy interaction across national boundaries.
Cronin’s findings of citation insularity (based, though
they were, on a small sample) are not idiosyncratic.
Inhaber and Alvo (1978) found that the US literature
attracted by far the greatest number of citations from
within and without. They found, inter alia, that US
journals were ar .roximately seven times as likely to cite
themselves, or o.ner US journals, than journals from the
UK, while UK journals, by way of contrast, divided
their attention almost equally between the US and the
UK literature. Further evidence of the global dominance
of US literature in science has been provided by the
American Psychological Association ( 1965), Lawani
(1977B), Bath University (1979) and Jagodizinski-
Sigogneau et a2 ( 1982).
T o allow that bias is possible is not quite the same
thing as imputing base or mischevious intent to authors
whose citation practices deviate from the expected norm
(Broadus, 1983). However, recognition of the fact that
citation may be coloured by extra-scientific factors is a
useful caution to the extremes of the normative position.
In his article The disinterested scientist: fact or_fiction, Mitroff
(1973 : 765) argues persuasively that science derives
great strength from its ‘contaminating subjective
elements’. T o support his case, he provides a wide
variety of quotations from eminent scientists reflecting
on the presence of, and need for, subjective biases in
science. Contrary to what might be expected, many
scientists seem opposed to the simplified interpretation of
scientific behaviour implied in the normative view. The
following comment from one of the Apollo moon
scientists interviewed by Mitroff should illustrate his
point.
‘The disinterested scientist is a myth. Even if there were such a
being, he probably wouldn’t be worth much as a scientist. I still
think you can be objective in spite of having strong interests and
bias. Ifyou make neutral statements, nobody really listens to you.
72
 Back to Contents Page
To Contents Page
You have to stick your neck out. The statements you make in
public are actually stronger than you believe in. You have to get
people to remember that you represent a point of view even if for
you it’s just a possibility. I t takes commitment to be a scientist.
One thing that spurs a scientist on is competition, warding off
attacks against what you’ve published.’ (Mitroff, 1973 : 765)
In Mitroffs view (1973; 1974B), bias and commitment
are not necessarily negative features. Without them, he
maintains, science would be an impoverished under-
taking. If Mitroff is right, then we have to consider the
implications as far as authors’ citation practices are
concerned. We have, to use Merton’s stylish phrasing
(1968B : 271), to take account of the interaction between
normative imperatives and ‘ethnocentric particularism’.

73
 Back to Contents Page

Towards a synthesis

We began this essay by presenting two broad approaches

to the analysis of science and the professional activities of
scientists. Traditionally, the analysis and interpretation
of science has favoured a normative view of how science
is organised and how scientists’ conduct is shaped and
controlled. This perspective has been described by
Mulkay (1977) as having a plausible descriptive rhetoric.
In certain respects this view was confirmed and encour-
aged by the writings of sociologists and historians of
science such as Merton, Barber and Storer. The great
appeal of the normative view is its apparent ability to
reconcile individual ambition with collective concerns.
Thus, acceptance of the normative position made it
possible to preserve the ‘storybook’ image of science,
even when the totality of available evidence might have
pointed to the need for theory modification or revision.
In Mulkay’s view, however, the lack of direct data on
scientists’ commitment to social norms is nothing less
than astonishing.
With the emergence of a new wave of interest in the
social processes of science came a challenge to the
hegemony of the normative view. The iconoclasm of the
interpretative approach to the sociological analysis of
science required that the taken-for-granted assumptions
upon which the normative view of science was posited
74
be re-examined. In some quarters the static view of
scientific behaviour was replaced by an on-going process
of empirical investigation and cautious theory construc-
tion. Science, it was contended, did not in fact display
the imagined or hoped-for regularities ascribed to it by
Merton and his followers: instead it was seen as being
subject to continual negotiation. Interpretative sociolo-
gists preferred not to think in terms of a high degree of
cognitive and normative consensus, but in terms of
negotiated and fluid interactions between individuals.
That is not to say that the interpretative sociologists
rejected out of hand the idea that science was in part
normatively regulated, merely that this assumption
should only be accepted if empirically validated in terms
of the behaviours, actions and testimonies of individual
scientists.
Thinking specifically in terms of citation it is not hard
to see why the normative view should have proved so
popular. An all-embracing normative framework meant
that it was possible to interpret citation as yet another
instance of scientists’ shared appreciation of how and
why they should act in a manner conducive to the
common good. Consequently, there was no need to seek
a fundamentally fresh or original explanation of citation
behaviour in general, nor any need to undertake detailed
and laborious analyses of individuals’ approaches to
citation. Citation was seen as being locked into the
prevailing view of science as a normatively governed
system, in which individuality played second fiddle to
communism. The distinguishing feature of interpretative
sociology was its reluctance to accept generalizations
which were not behaviourally or experimentally derived,
however convenient the alternative. It preferred to take
as its starting point the often mundane, but nonetheless
relevant, daily realities of the working life and relations
of individuals, and from these to fashion models (perhaps
narratives would be a better term) of scientific behaviour.
75
 By far the greatest attraction of the normative
position, as far as citation is concerned, is that it offers an
all-in-one account of a complex phenomenon. Within the
normative framework the strength of the norms and the
power of the associated sanctions are deemed sufficient
to check idiosyncratic and individualistic behaviour. In a
sense, the normati1-e position ‘straitjackets’ behaviour in
a coat of rationality and enlightened group concern.
‘The normative view has been described as ‘bad
mctaphysics’ (Mitroff, 1974B). This may be something
of an overstatement, but in the writings of Polanyi
(1966A; 1966B) there is a tendency to veer towards a
rnctap1i)kdI account. It may be difficult to resist the
arguments put forward by Polanyi, if only because of the
subtle way in which he shepherds us towards the desired
conclusion. In preparing the reader for the unorthodoxy
of his view (citation as an instance of ‘tacit’ knowledge)
Polyani engages in a process of low-key desensitization.
The following two extracts illustrate the approach and
the language used.
‘The process of formalizing all knowledge to the exclusion of any
tacit knowledge is self-defeating.’ (Polanyi, 1966A : 20)
‘. . . an unbridlcd lucidity can destroy our understanding of
complex matters.’ (Polanyi, 1966A : 18)
That being the case, one can easily begin to see the
attractiveness of placing citation among that class of
activities whose exercise a n d acquisition arc not readily
amenable to exposure and analysis. If we were to adopt
Polanyi’s position, then we might choose to think of
citation as a skill like bicycling or swimming.
‘If I know how to ride a hicyclc or how to swim, this does not
mean that I can tell how I managc to keep my balance on a
bicyclc, or krrp afloat when swimming. I may not have thc
slightrst idea of how 1 d o this, or even an entirely wrong or
grossly imperfect idea of it, and vet go on cycling or swimming
merrily. Nor can it be said that I know how to bicyclc or swim

76
 and yet do not know how to coordinate the complex pattern of
muscular acts by which I do my cycling or swimming. I both
know how to carry out these performances as a whole and also
how to carry out the elementary acts which constitute them,
though I cannot tell what these acts are. This is due to the fact
that I am only subsidiarly aware of these things and our
subsidiary awareness of a thing may not suffice to make it
identifiable.’ (Polanyi, 1966B : 4)

As with many analogies, its usefulness is skin-deep. To

view citation in this way would mean glossing over a
number of critical differences (between citation and
swimming). Swimming is a motor activity - one is
either able to swim or not (to put it crudely). For the
would-be swimmer there is immediate, unequivocal
performance feedback. If he sinks, then he has not
suceeded in swimming; if he remains afloat in swimming
mode, then he has been successful. These are baseline
measures (sinking or staying afloat) against which to
assess performance. As measures, they are both objective
and definitive. Such performance measures do not exist
so far as the evaluation of citation behaviour is con-
cerned. Secondly, to rely on an analogy of this kind
would be to overlook the matter of intentionality. The
notion of subsidiary awareness which Polanyi proposes
as an explanation of how citation is correctly practiced
does not take into account the potentially enormous
variability of individuals’ motivations for citing in a
particular fashion. In some respects, this echoes
Merton’s ( 1972) distinction between ‘acquaintance with’
and ‘knowledge about’. Thus, we might think of authors
as being acquainted with the protocols of citation, but
not always or necessarily knowledgeable about its signifi-
cance. What both Polanyi and Merton, in their different
ways, seem to be saying is that there are levels of
knowledge and appreciation.
Perhaps we need to think of citation along similar
lines: a process which is residually subjective, but which,
77
through the application of a set of more or less agreed
techniques, aspires to be a vehicle of universal com-
munication within the scientific information exchange
process.
A failing of normativism is that it takes no heed of the
‘soft underbelly’ of science (Edge, 1979 : 115). The
deficiency of Polanyi’s position is that it eliminates the
motivational component in citation. A frustrating aspect
of the interpretative perspective proposed by contem-
porary sociologists of science is that it denies us a simple
and satisfying account or theory of citation generally.
What we now need to do is consider the extent to which
it is possible to develop an integrated account of citation
behaviour from the evidence to hand.
We began by saying that science should be thought of
as a social system. The participants in this social system,
career scientists and researchers, enter of their own
volition. I n so doing they presumably recognise that
there will be constraints on individual actions, and
responsibilities towards others operating within the
system. The stability of the social system is maintained
by a delicate process of complementary interaction. To
quote Storer:

‘The patterns of interaction among two or more participants in a

social system are maintained because the reaction of the
participants to one another’s behavior is mutually rewarding.
Implied here is the assumption that each participant has some
standards by which to judge whether the other’s actions are
appropriate or not, and also that the other participant knows of
these standards so that he can choose the behavior that will elicit
-
rewarding behavior from the first.’ (Storcr, 1966 : 32)

In reality, things are rather more harried; rather more

complicated. The normative view of science tends to play
down, or gloss over, the more robust and venial aspects
of scientific life. Even in 1902, Rutherford was writing to
78
his mother saying:
‘I have to keep going, as there are always people on my track. I
have to publish my present work as rapidly as possible in order to
keep in the race.’ (quoted in: Crowther, 1952 : 54)
In the intervening eighty or so years the pressures and
stresses associated with science have increased greatly.
The competitiveness of science has attracted investi-
gation, and there is evidence to suggest that the pressures
on scientists substantially affect their social interactions
(e.g. Gaston, 1971; Watson, 1970).
Science is institutionalized competition, and scientific
knowledge is a cultural artifact (Collins, 1975). In
academic life (particularly) the main reward or intangi-
ble reinforcer is peer group recognition. Mitroff s
(1974A) deep-felt conviction of this led him to propose a
set of counter-norms to explain the actions of scientists.
As a result, much modern sociological investigation has
concentrated on stripping away the romance associated
with professional science. Bernal (Goldsmith, 1980) feels
that the scientific paper functions as a sketch map to the
truth, while Medawar is of the opinion that it is a fraud:
‘The scientific paper is a fraud in the sense that it does give a
totally misleading narrative of the process of thought that goes
into the making of scientific discoveries.’ (Medawar, 1964 : 43)
According to Meadows, this fraudulence has been
cultivated intentionally:
‘One motive behind this development has clearly been the desire
to transmit research results in a stylized and impersonal form so
as to transfer both information and priority claims with the
greatest possible efficiency.’ (Meadows, 1974 : 82)

If the scientific paper is a fraud, then one needs to ask

how this might affect authors’ approaches to citation.
Gilbert (1977) has argued that citation should be
thought of as a process of persuasion. How, in that case,
does an author maintain the necessary ‘delicate balance’
79
(Kaplan, 1965) between the desire to make priority
claims and the requirement that recognition and credits
be distributed equitably?
Perhaps the single most important criticism of citation
analysis is that it ignores this tension. As Cozzens (1981)
noted, many citation analysts view their measures as
direct manifestations of certain social constructs, without
visualizing at the same time the scientists who create the
patterns. Citation is part of the social process of science,
but this fact is over-looked in the majority of applied
citation analysis studies.
Citation is a complex issue. Earlier, we outlined a
number of interpretative perspectives which could be
applied to citation, but it may be, as Cozzens (1981) has
proposed, that there is no need to settle for one of these at
the expense of others. In her view, the action of scientists
in writing a scholarly paper are consistent with the
various perspectives. I t is also consonant with Maslow’s
perceptions of how professional scientists behave.
‘It is possible for healthy scientists to enjoy not only the beauties
of precision but also the pleasures of sloppiness, casualness, and
tight and/or loose, sensible and/or crazy, sober and/or playful,
seems to be a characteristic not only of psychological health but
also of scientific creativeness.’ (Maslow, 1966 : 31)

If Maslow is correct in his assessment, then it is only to

be expected that sloppiness, casualness and veniality will
be exhibited occasionally in the citation practices of
scientific authors.
In a thought-provoking paper, Wartofsky (1982)
proposed a social model of medical knowledge, i.e. a
model of medical knowledge as a social product and of
medical practice as a social activity. In doing so he
argued that neither the product nor the activity could be
fully comprehended other than in terms of social
relations and the ideological framework within which
members of the medical community operated.
80
 In defining the nature of medical knowledge, and the
nature of possession of this knowledge, he invoked the
knowledge-as-property metaphor. This metaphor is also
used in relation to citation, the latter being seen as one
means of protecting intellectual property rights in the
scientific marketplace. Wartofsky’s idea is noteworthy
because it posits three forms of property right. These
forms he defines in the following manner:
‘The first form of possession may be seen as akin to slave systems
in which the slave has no possession . . . The second form is that
of communal or feudal property in which individual use . . . is on
the condition of certain obligation to the commune or to the
feudal lord, and the right to use of one’s possession is a fiduciary
right. . . The third form of private or individual ownership with
the concomitant right to use. . . is in fact the classical model of
the free-exchange system of capitalism, where property exchanges
for property through the medium of money, by the consent of
exchangers. . . knowledge or rather disposition over its use in
this context is a commodity, exchangeable for a fee, in the
market-place.’ (Wartofsky, 1982 : 126)

However, having identified three forms, or evolutionary

phases, he found that no one seemed to apply to medical
knowledge.
‘The peculiar thing about medical practice is that it is caught up
in all of these three forms of property-right and represents a
complex historical palimpest of these modes.’ (Wartofsky,
1982 : 126)

Now, as we have attempted to show, there is no single,

all-embracing theory of citation. We have also seen that
the appropriateness of the various analytical perspectives
is situationally or contextually dependent. Citation,
however we care to view it, ostensibly involves the
expression of a relationship between two participants in
the social process of science - the citing and cited
authors. The coordinates of this relationship are the
atoms of knowledge (intellectual property) defined by
the citing and cited works. Using Wartofsky’s model we
81
might like to think of citation in relation to the three
forms of property possession mentioned above. Better
still, we should perhaps think of citation in terms of a
‘complex historical palimpest’. The advantage of such a
move would be to rescue us from the invidious, if not
impossible, position of having to explain citation in terms
of one, and only one, of the available interpretations. In
Wartofskian terms the ‘logic’ of citation is etched on a
complex palimpest, which seeks to acknowledge the
three forms of knowledge possession.
The first of these, the slave image, corresponds to the
normative view of science (and by extension citation) as
a socially responsible activity devoid of personal or
selfish considerations. Knowledge is universalistic and
scientists are fired by the principle of communism.
Ownership and intellectual property rights are therefore
antithetical to this world view.
The second, or feudal, form of knowledge possession is
a somewhat less severe version of the first. The indi-
vidual scientist or researcher, though having to acknowl-
edge the ultimate authority of the scientific establish-
ment (and the general public, for that matter), is
nevertheless entitled to some degree of recognition and
some claim-staking, provided personal interest does not
get in the way of the greater scientific and communal
good. There is, after all, a difference between possession
and possessiveness.
The third form corresponds in many respects to the
interpretative approach espoused by the microsociol-
ogists. In other words, the actions of scientists are
expressions of personal motivations and ambitions (not
least the search for peer group recognition and acclaim),
which may on occasion be moderated by ‘purer’
considerations. I n this view of things scientists see their
atoms of personally discovered knowledge as belonging
unequivocally to themselves. These they can exchange
for a particular sort of currency - citations from other
82
members of the scientific community. Knowledge is thus
a property which can be exchanged or traded for public
recognition. Citation behaviour has to be seen in relation
to an individual’s background, professional training and
experience, and belief system. Differences in ideology or
in the processes of socialization to the norms of scientific
conduct; or differences in the practices, structures and
social dynamics of subject fields and research com-
munities, will also effect citation behaviour.
Because citation is a private act, and because there are
many shades of interpretation as to what constitutes
correct or acceptable citation, it is particularly difficult, if
not impossible, to speak of a theory of citation. At both
the individual and group level differences in approach to
citation are widespread. Such inconsistency does not
facilitate the construction of predictive theories. And yet,
one senses (and the word is used advisedly) that authors
do have some shared perceptions and understanding of
what is called for and why (Cronin, 1983). It may be this
vague feeling that the practice is not entirely haphazard
or whimsical, which explains why writers on the subject
have resorted to terms such as ‘osmosis7,‘tacit’, ‘cultural
transmission’, and ‘craft skill’. Hence, too, the appeal of
the notion of ‘self-coordination by mutual justification’
(Polanyi, 1966A : 71), which conveniently sweeps the
interpretative difficulties under a metaphysical carpet.
Citation needs to be thought of as a process. The
outcomes of this process (on a recurring basis) are lists of
citations attached to scholarly papers. The character and
composition of the lists reflect authors’ personalities and
professional milieux. The elements in the chemistry of
citation are almost infinite, and it is this fact which
necessitates particularistic accounts of citation. We need,
in effect, to move into the ‘unexamined psychology of
science’ (Mahoney, 1976) if further progress is to be
made.
One of the issues which we have alluded to, though not
83
attempted to resolve, is that of the relationship between
cognitive style and an individual’s personality (Mitroff,
1977). Future studies will have to devise novel methods
of defining and articulating the nature of this relation-
ship, if our understanding of what citation behaviour
entails is to be expanded. There is no single theory of
citation capable of telling us why authors cite in a
particular fashion. Singer has provocatively suggested
that, if asked for an explanation of the ultimate ruison
d’2tre of science, few scientists would be able to reply with
confidence and assurance. O n the other hand, ‘they are
fairly sure they know one when they see one’ (Singer,
1971 : 1012). Perhaps the same applies for citation.
We cannot say that citation is an activity governed by
adherence to a specific and universally recognised set of
norms. By the same token, the evidence does not permit
us to conclude that the practice is characterised by
randomness and inconsistency. The interplay between
institutional norms (even if only vaguely grasped by
authors) and personal considerations is extremely com-
plicated. We may not be able to champion the normative
view, but by the same token we are under no obligation
to subscribe wholly and uniquely to the literalism of the
interpretative approach. As Mitroff says of science in
general:
‘It is too simple to say that the process or system of science is
either totally objective or totally subjective.’ (Mitroff, 1972 : 615)

So indeed, it may well be as far as citation is concerned.

T o seek and expect to find a single, theoretically
integrated and undimensional account of authors’
citation behaviour may be as unprofitable an activity as
it is unrealistic. The ‘storybook’ account of science may
be a fairytale (Mitroff, 1972), but the ‘warts-and-all’
version would itself appear to veer occasionally into the
realms of fantasy.
Kuhn has argued that there are two sorts of history of
84
science, the internalist and the externalist (see below) ,
and it may be that the same can loosely be said to apply
to citation.
‘. . . there seems at times to bc two distinct sorts of history of
science, occasionally appearing between the same covers but
rarely making firm or fruitful contact. The still dominant form,
often called the ‘internal approach’, is concerned with the
substance of science as knowledge. Its newer rival, often called
the ‘externalist approach’, is concerned with the activities of
scientists as a social group within a larger culture. Putting the
two together is perhaps the greatest challcnge now faced by the
profession . . .’ (Kuhn, 1968 : 76)

That is to say: most citation analyses have been

internalist rather than externalist in character, in that
they have concentrated on quantities and frequency
distributions rather than on the contexts within which,
and processes by which, authors employ citations. The
reservations which exist about the use and validity of
citations as indicators of impact and performance are
unlikely to evaporate simply because further quantitative
studies are carried out, or because more refined citation
mapping and clustering techniques are developed.
Citation analysis requires the sort of firm and fruitful
contact between these two approaches which Kuhn feels
is needed in the history of science.
Although his research was not specifically concerned
with scientists’ appproaches to citation, Mitroff (1973;
1974A; 1974B) has perhaps succeeded in furnishing us
with a richer and more plausible interpretation of how
scientists view their own actions and attitudes, and those
of their peers. If novel insights on citation are to be
mined, then an approach similar to that used by Latour
and Woolgar (1979) in their ‘anthropological’ account of
life in the laboratories of the Salk Institute will be
required. Investigators will need to mix with and
participate in the work-a-day life of authors and scientists
if they are to grasp fully the significance of citation in the
85
social process of science. As Latour and Woolgar state:
‘The production of papers is acknowledged by participants as the
main objective of their activity. The realisation of this objective
necessitates a chain of writing operations from a result first
scribbled on a sheet of paper and enthusiastically communicated
to colleagues, to the final registering of published literature in the
laboratory archives. The many intermediary stages (such as talks
with slides, circulation of preprints, and so on) all concern
literary production of one kind or another. It is thus necessary
carefully to study the various processes of literary production
which lead to the out-put of papers.’ (Latour and Woolgar,
1979 : 71)

Citation is not something which happens in a void,

and citations are not separable from the contexts and
conditions of their generation. Citation, to use the words
of Latour and Woolgar, is part of the various processes of
literary production. If we are to comprehend the
significance of citation, then we need to lay bare these
processes.
Future studies of citation should therefore concentrate
of the content of citations, and the conditions of their
creation and application. As far as Law and French are
concerned:
If science is viewed as a social process, then there is as much
sociological reason to be interested in it from the point of view of
the laboratory technician or the student as there is from the
research scientist or journal editor. Another area of interest is the
production of scientific papers. These are written in situations
that are peopled by such significant others as administrators,
professors, anticipated audiences, recalcritant research assist-
ants, typists, colleagues, husbands and wives. These situations
refer to laboratories, promotions, salaries, research grants,
equipment, computer time, and mortgages. Thus, for example,
any study which uses scientific papers as data should take
cognisance of the situations in which they are written.’ (Law and
French, 1974 : 589)

This view is widely held by contemporary sociologists

of science (e.g. Knorr-Cetina and Mulkay, 1983;
86
 To Contents Page
Next to References
Woolgar, 1980), but it has not penetrated the thinking of
the information science community. There is, arguably,
a need to explore what Star (1983) calls the ‘dialectical
relationship’ between resource constraints and research
results. In her study of the ‘simplification’ process in
scientific work she identified several types of constraints:
(1) intersection; (2) clinical; (3) technical; (4) conclusion
pressures; (5) formatting; ( 6 ) editing rules; (7) special-
ization. Along with these constraints there are social,
institutional and ideological pressures bearing on scien-
tists’ behaviours and actions. If we are to claim to be able
to understand the purpose and significance of citation in
the social process of science, then some attention has to
be given to these pressures and constraints. As Star says,
scientific work involves the representation of chaos in an
orderly fashion. If we accept this interpretation, then we
are obliged to consider the part played by citations in
conferring orderliness and acceptability on published
research findings. I t is to be hopcd that future studies of
citation will at least take note of the microsociological
viewpoint, and use it to enrich our appreciation of what
citation signifies in the knowledge construction and
dissemination process.

87
A-B C D-E-F-G H-I-J- K-L-M N-O-P-Q

R-S T-U-V- W-X-Y-Z

References

AARONSON, S. (1975). The footnotes of science.

Mosaic, 6 ( 2), 22-2 7.
AMERICAN PSYCHOLOGICAL ASSOCIATION
( 1965). Scientific information exchange in psychology.
vol. 2. Washington: APA.
BARBER, B. (1952). Science and the social order.
Glencoe: Free Press.
BARBER, B. (1971). Resistance by scientists to scientific
discovery. Science, 134, 596-602.
BASEFSKY, S. (1982). Citation training needs. American
Libraries, April, 226 (letter to the editor).
BATH UNIVERSITY (1979). The design of infor-
mation systems in the social sciences: the structure of
social science literature as shown by citations. Bath:
Bath University (Research Reports Series A no. 3).
BAVELAS, J.B. (1978). The social psychology of
citation. Canadian Psychological Review, 16, 158-163.
BAYER, A.E. and FOLGER, J. (1966). Some correlates
of citation as a measure of productivity. Sociology of
E ~ U C U ~39(4),
~ O T Z381-390.
,
BENSMAN, S.J. (1982). Bibliometric laws and library
usage as a social phenomenon. Library Research, 4,
279-3 12.
88
BEZILLA, S.J. (1982). Online messages, files, text and
publishing. Online, 6(2), 5 1-55.
BLACKWELL, P.K. and KOCHTANEK, T.R. (1981).
An iterative technique for document retrieval using
descriptors and relations. Proceedings of the 44th
American Society for Information Science Annual
Meeting. Washington: ASIS, 2 15-2 17.
BOFFEY, P.M. ( 1970). Psychology: apprehension over
a new communications system. Science, 167, 1228-1230.
BONZI, S. (1982). Characteristics of a literature as
predictors of relatedness between cited and citing works.
Journal of the American Society for Information Science, 33(4),
208-2 16.
BOURDIEU, P. (1972). The specificity of the scientific
field and the social conditions of the progress of reason.
Social Science Information, 14(6), 19-47.
BOWEN, D.D. et al. (1972). Improving manuscript
evaluation procedures. American Psychologist, March,
22 1-225.
BROADUS, R.N. (1983). An investigation of the
validity of bibliographic citations. Journal of the American
Society f o r Information Science, 34(2), 132-1 35.
CASON, H. and LUBOTSKY, M. (1936). The influence
and dependence of psychological journals on each other.
Psychological Bulletin, 33, 95-103.
CHERRY, C. (1978). On human communication: a
review, a survey and a criticism. London: MIT, 3rd ed.
CHUBIN, D.E. (1973). The use of the Science Citation
Index in sociology. American Sociologist, 8, 187-191.
CHUBIN, D.E. (1980). Is citation analysis a legitimate
evaluation tool? Scientometrics, 3 (2), 9 1-92 (letter to the
editor).
CHUBIN, D.E. and MOITRA, S.D. (1975). Content
analysis of references: adjunct or alternative to citation
counting? Social Studies of Science, 5 , 423441.
89
CLARK, K. (1957). America’s psychologists: survey of a
growing profession. Washington: APA.
CLEVERDON, C.W. (1970). Letter to the editor.
Nature, 228, 1356.
COLE, J.R. andCOLE, S. (1971). Measuring the quality
of sociological research: problems in the use of the
Science Citation Index. American Sociologist, 6, 23-29.
COLE, S. and COLE, J.R. (1967). Scientific output and
recognition: a study in the operation of the reward
system in science. American Sociological Review, 32(3),
377-390.
COLE, S. and COLE, J.R. (1968). Visibility and the
structural basis of awareness of scientific research.
American Sociological Review, 33, 397-41 3.
COLLINS H.M. (1975). The seven sexes: a study in
the sociology of a phenomenon, or the replication of
experiments in physics. Sociology, 9, 205-224.
COZZENS, S.E. ( 1981) . Taking the measure of science.
Philadelphia: IS1 (unpublished paper).
COZZENS, S. (1982). Split citation identity: a case
study from economics. Journal of the American Society _for
Information Science, 33(4), 233-236.
CRANE, D. (1972). Invisible colleges: diffusion of
knowledge in scientific communities. Chicago:
Univerisity of Chicago Press.
CRONIN, B. (1981). Transatlantic citation patterns in
educational psychology. Education Libraries Bulletin, 24(2),
48-5 1.
CRONIN, B. (1982A). Norms and functions in citation:
the view of journal editors and referees in psychology.
Social Science Infirmation Studies, 2(2), 65-78.
CRONIN, B. (1982B). Invisible colleges and infor-
mation transfer: a review and commentary with particu-
lar reference to the social sciences. Journal of
Documentation, 38(3), 2 12-236.
90
CRONIN, B. (1983). Citation and the social process of
science: an assesment of citation practice with particular
reference to the journal literature of psychology. Belfast:
The Queen7sUniversity, PhD Thesis.
CROWTHER, J.G. (1952). British scientists of the
twentieth century. London: Routledge and Kegan Paul.
DAVIES, D. (1970). Letter to the editor. Nature, 228,
1356.
DIEKS, D. and CHANG, H. (1976). Differences in
impact of scientific publications: some indices derived
from a citation analysis. Social Studies of Science, 6,
247-26 7.
DEWITT, T.W. et al. Science Citation Index. Sciento-
metrics, 2(4), 265-275.
DUNCAN, E.B. et al. (1981). Qualified citation
indexing: its relevance to educational technology. in
Duncan, E.B. and McAleese, R. (eds.). Information
retrieval in educational technology. Proceedings of the
First Symposium on Information Retrieval in
Educational Technology, 1st April 1981, Aberdeen.
Aberdeen: University of Aberdeen, 70-79.
EDGE, D. (1977). Why I am not a co-citationist. Society
for the Social Studies of Science Newsletter, 2, 13-19.
EDGE, D. (1979). Quantitative measures of com-
munication in science: a critical review. Histov of Science,
17, 102-134.
ELLIS, N.D. (1969). The occupation of science.
Technology and Society, 5 ( l ) , 3 3 4 1 .
FINNEY, B. (1979). The reference characteristics of
scientific texts. London: City University, Centre for
Information Science (MSc dissertation).
FISCH. R. (1977). Psychology of science. in Spiegel-
Rasing, I. and Price, D.J. de Solla (eds.). Science,
technology and society: a cross-disciplinary perspective.
London: Sage, 227-3 18.
91
FROST, C. (1979). The use of citations in literary
research: preliminary classification of citation functions.
Library Quarterly, 49(4), 399-414.
GOLDSMITH, M. (1980). Sage: a life of J.D. Bernal.
London: Hutchinson, 143-144.
GARFIELD, E. (1963). Citation indexes in sociological
research. American Documentation, 14(4), 289-291.
GARFIELDy E. (1965). Can citation indexing be
automated? in Stevens, M.E. et al (eds.). Statistical
association methods for mechanized documentation.
Washington: National Bureau of Standards, 189-1 92.
GARFIELD, E. (1970). Citation indexing for studying
science. Nature, 227, 669-67 1.
GARFIELD, E. (1977). Essays of an information
scientist. vol. 1, Philadelphia: IS1 Press.
GARFIELD, E. (1979). Citation indexing: its theory and
application in science. New York: Wiley.
GARFIELD, E. (1980A). Is citation analysis a legit-
imate evaluation tool? Scientometrics, 3(2), 92-93 (letter to
the editor).
GARFIELD, E. (1980B). Is citation indexing in the arts
and humanities inherently different from that in science?
T h e effects that ISI’s Citation Index for the Arts and
Humanities is expected to have on future scholarship.
Library Quarterly, 50( l ) , 40-57.
GARVEY, W.D. and GRIFFITH, B.C. (1972). Com-
munication and information processing within scientific
disciplines: empirical findings for psychology. Information
Storage and Retrieval, 8(3), 123-1 36.
GASTON, J. (1971). Secretiveness and competition for
priority of dicovery in physics. Minerua, 9(4), 472-492.
GILBERT, G.N. (1976). The transformation of research
findings into scientific knowledge. Social Studies o f Science,
6,281-306.

92
GILBERT, G.N. (1977). Referencing as persuasion.
Social Studies of Science, 7, 1 13- 122.
GILBERT, G.N. and MULKAY, M. (1980). Contexts of
scientific discourse: social accounting in experimental
papers. in Knorr, K.D. et al. (eds.). The social process of
scientific investigation. Dordrecht: Reidel, 269-294.
GILBERT, G.N. and MULKAY, M. (1982). War-
ranting scientific belief, Social Studies o f Science, 12(3),
383-408.
GILLIE, 0. (1978). Sir Cyril Burt and the great I Q
fraud. New Statesman, 96, 688-694.
GILLIE, 0. (1980). Burt: the scandal and the cover-up.
Bulletin of the British Psychological Society, 33,9-16.
GORDON, M.D. (1979). Deficiencies of scientific infor-
mation and access in less developed countries. Journal of
the American Society for Information Science, 36(3), 340-342.
GOTTFREDSON, S.D. et al. (1977) Quality indicators
in the scientific journal publication process. Baltimore:
John Hopkins University, Center for Research in
Scientific Communication.
GOTTFREDSON, S.D. and GARVEY, W.D. (1980).
Review of ‘Citation indexing’ by Eugene Garfield.
Behavioral and Social Sciences Librarian, l(4). 289-294.
GOUDSMIT, S.A. (1974). Letter to the editor. Science,
183(4120),28.
GRIFFITH, B.C. et al. (1977). On the use of citations in
studying scientific achievement and communication.
Society for the Social Studies of Science Newsletter, 2, 9-1 3.
HAGSTROM, W.O. (1965). The scientific community.
New York: Basic Books.
HAGSTROM, W.O. (1971). Inputs, outputs and the
prestige of university science departments. Sociology of
Education, 44 (4), 375-39 7.
HALL, A.M. (1970). The use and value of citations: a
93
state-of-the-art report. London: Institute of Electrical
Engineers.
HERLACH, G. ( 1978). Can retrieval of information
from citation indexes be simplified?Journal ofthe American
Son'etyfor Infonnation Science, 29 (6), 308-3 10.
HJERPPE, R. (1980). A bibliography of bibliometrics
and citation indexing and analysis. Stockholm: RIT.
HODGES, T. (1978). Citation indexing: its potential for
bibliographic control. Berkeley: University of California
(PhD thesis).
HOUGHTON, B. (1975). Scientific periodicals: their
historical development, characteristics and control.
London: Bingley.
HUDSON, L. (1975). Human beings: an introduction to
the psychology of human experience. London: Cape.
INHABER, H. and ALVO, M. (1978). World science as
an input-output system. Scientometrics, 1( l ) , 43-64.
JACODZINSKI-SIGOGNEAU, M et al. (1982). How
to measure the degree of independence of a research
system. Scientometrics, 4(2), 119-133,
JAKOBOVITS, L.A. and OSGOOD, C.E. (1976).
Connotations of twenty psychological journals to their
readers. American Psychologist, 22, 792-800.
JEVONS, F.R. (1973). Science observed: science as a
social and intellectual activity. London: Allen & Unwin.
JOHANSSON, C.M.A. (1976). Citation behaviour is
chemistry. London: City University, Centre for Infor-
mation Science (MSc dissertation).
KAPLAN, N. (1965). The norms of citation behavior:
prolegomena to the footnote. American Documentation,
16(3), 179-184.
KNIGHT, D. (1976). The nature of science: the history
of science in western culture since 1600. London:
Deutsch.
94
KNORR-CETINA, K.D. and MULKAY, M. (1983).
(eds.). Science observed: perspectives on the social study
of science. London: Sage.
KOCHEN, M. and PERKEL, B. (1978). Improving
referee-selection and manuscript evaluation. in Balban,
M. (ed.). Scientific information transfer: the editor's role.
Dordrecht: Reidel, 203-229.
KROHN, R. (1980). Introduction: towards the empiri-
cal study of scientific practice. in Knorr, K.D. et al (eds.).
The 'social process o f scientific investigation. Dordrecht:
Reidel, vii-xxv.
KUHN, T.S. (1968). The history of science. International
Encyclopedia of the Social Sciences, 14, 74-83.
LATOUR, B. and WOOLGAR, S. (1979). Laboratory
life: the social construction of scientific facts. Beverley
Hills: Sage.
LAW, J. (1974). Theories and methods in the sociology
of science: an interpretative approach. Sociology of Science,
13,163-172.
LAW, J. and FRENCH, D. (1974). Normative and
interpretative sociologies of science. Sociological Review,
22,581-595.
LAWANI, S.M. (1977A). Citation analysis and the
quality of scientific productivity. Bioscience, 27( l ) , 26-31.
LAWANI, S.M. (1977B). The professional literature
used by American and French agronomists and the
implications for agronomic education. Journal o f Agrono-
mic Education, 6,41-46.
LAWANI, S.M. and BAYER, A.E. (1983). Validity of
citation criteria for assessing the influence of scientific
publications: new evidence with peer assessment. Journal
of the American Society for Information Science, 34( l ) , 59-74.
LEOPOLD, A.C. (1973). Games scientists play. Bio-
science, October 23rd, 590.

95
LINDSEY, D. (1978). The scientific publication system
in social science. San Francisco: Jossey-Bass.
LINDSEY, D. and LINDSEY, T . (1978) the outlook of
journal editors and referees on the normative criteria of
scientific carftsmanship: viewpoints from psychology,
social work and sociology. Quality and Quantity, 12,4562.
LIPETZ, B.A. (1965). Improvement of the selectivity of
citation indexes to science literature through the inclu-
sion of citation relationship indicators. American Documen-
tation, 16(2), 81-90.
MacLEOD, R. (1977). Changing perspectives in the
social history of science. in Spiegel-Rosing, I. and Price,
D.J. de Solla (eds.). London: Sage, 149-174.
McINNIS, R. (1982). Do metaphors make good sense in
teaching research strategy? in Oberman, C. and Strauch,
K. (eds.). Theories of bibliographic education: designs
for teaching. New York: Bowker, 47-74.
MAHONEY, M.J. (1976). The scientist as subject: the
psychological imperative. Cambridge, Mass.: Ballinger.
MARTINO, J.P. ( 1971). Citation indexing for research
and development management. IEEE Transactions on
Engineering Management, EM-18(4), 146-151.
MARTYN, J. (1964). Bibliographic coupling. Journal of
Documentation, 20(4), 236.
MARTYN, J. (1965). An examination of citation
indexes. Aslib Proceedings, 17(6), 184-196.
MASLOW, A.W. (1966). The psychology of science.
New York: Harper & Row.
MAY, K.O. (1967). Abuses of citation indexing. Science,
156, 890-892 (letter to the editor).
MEDAWAR, P.B. ( 1964). Is the scientific paper fraud-
ulent? Saturday Review, August lst, 4 2 4 3 .
MEADOWS, A.J. (1974). Communication in science.
London: Butterworth.
96
MEADOWS, A.J. (ed.). (1979). The scientific journal.
London: Aslib.
MERTON, R.K. (1957). Priorities in scientific
discovery: a chapter in the sociology of science. American
Sociological Review, 22(6), 635-659.
MERTON, R.K. (1963). The ambivalence of scientists.
Bulletin of the John Hopkins University, 112(2), 77-97.
MERTON R.K. (1968A). Social theory and social
structure. New York: Free Press.
MERTON, R.K. (1968B). The Matthew effect in
science: the reward and communication systems of
science are considered. Science, 159, 56-63.
MERTON, R.K. (1972). Insiders and outsiders: a
chapter in the sociology of science. American Journal of
Sociology, 78( 1), 9-47.
MERTON, R.K. (1973). The sociology of science:
theoretical and empirical investigations. Chicago:
Chicago University Press (ed.). Storer, N.
MILLER, G.A. (1962). Psychology: the science of
mental life. New York: Harper & Row.
MITROFF, 1.1. (1972). The myth of subjectivity or why
science needs a new psychology of science. Management
Science, 18, 6 13-6 18.
MITROFF, 1.1. (1973). The disinterested scientist: fact
or fiction? Social Education, 37, 761-765.
MITROFF, 1.1. (1974A). The subjective side of science:
a philosophical inquiry into the psychology of the Apollo
moon scientists. Amsterdam: Elsevier.
MITROFF, 1.1. (1974B). Norms and counter norms in a
select group of Apollo moon scientists: a case study of the
ambivalence of scientists. American Sociological Review, 39,
579-595.
MITROFF, 1.1. (1977). Some unresolved issues in the
psychology of science: a research agenda. Societyfor the
97
Social Studies of Science Newsletter, 2 (3), 20-22.
MITRA, A.C. (1970). The bibliographic reference: a
review of its role. Annals of Libra? Science, 17(3/4),
117-123.
MORAVCSIK, M.J. and MURUGESAN, P. (1975).
Some results on the function and quality of citations.
Social Studies of Science, 5, 86-92.
MORAVCSIK, M.J. and MURUGESAN, P. (1979).
Citation patterns in scientific revolutions. Scientiometrics,
1(2), 161-169.
MORAVCSIK, M.J. (1980). How to grow science. New
York: Universe Books.
MORMAN, E.T. (1981). Citation analysis and the
current debate over quantitative methods in the social
studies of science. Society f o r the Social Studies of Science
Newsletter, 5 (3), 7-1 3.
MULKAY, M. (1969). Some aspects of cultural growth
in the natural sciences. Social Research, 36( l ) , 22-52.
MULKAY, M. (1974). Methodology in the sociology of
science: some reflections on the study of radio astro-
nomy. Social Science Information, 13(2), 107-1 19.
MULKAY, M. (1977). Sociology of the scientific
research community. in Spiegel-Rosing, I. and Price,
D.J. de Solla (eds.). Science, technology and society: a
cross-disciplinary perspective. London: Sage, 93-148.
MULKAY, M. and GILBERT, G.N. (1981). The truth
will out: a device which scientists use to resolve potential
contradictions between interpretative repertoires.
Unpublished paper.
MURUGESAN, P. and MORAVCSIK, M.J. (1978).
Variations on the nature of citation measures with
journals and scientific specialties. Journal of the American
Society f o r Information Science, 29(3), 141-147.
NARIN, F. (1976). Evaluative bibliometrics: the use of
98
publication and citation analysis in the evaluation of
scientific activity. Cherry Hill, New Jersey: Computer
Horizons Inc.
NATIONAL SCIENCE FOUNDATION (1981).
Science indicators 1980. Washington: NSF.
O’CONNOR, J. (1982). Citing statements: computer
recognition and its use to information retrieval. Infor-
mation Processing 63 Management, 18(3), 125-131.
OPPENHEIM, C. and RENN, S.P. (1978). Highly cited
old papers and the reasons why they continue to be cited.
Journal o f the American Society for Information Science, 29(5),
225-23 1.
PERITZ, B.C. (1983) A classification of citation roles for
the social sciences and related fields. Scientometrics, 5(5),
303-312.
PETERS, D.P. and CECCI, S.J. (1982). Peer review
practices of psychological journals: the fate of accepted,
published articles submitted again. The Behavioral and
Brain Sciences, 5 (2), 187-255.
PINSKI, G. and NARIN, F. (1979). Structure of the
psychological literature. Journal o f the American Society for
Information Science, 30(3), 161-168.
POLANYI, M. (1966A). The tacit dimension. London:
Routledge & Kegan Paul.
POLANYI, M. (1966B). The logic of tacit inference.
Journal of the Royal Institute o f Philosophy, 41 (155), 1-18.
PORTER, A.L. (1977). Citation analysis: queries and
caveats. Social Studies of Science, 7, 257-267.
PRABHA, C.G. (1983). Some aspects of citation
behavior: a pilot study in business administration.
Journal of the American Society for Information Science, 34(3),
202-206.
PRICE, D.J. de Solla (1964). Ethics of scientific
publishing. Science, 144, 655-657.
99
RAVETZ, J.R. (1971). Scientific knowledge and its
social problems. Harmondsworth: Penguin.
ROBBINS, D. and JOHNSTON, R. (1976). The role of
cognitive and occupational differentiation in scientific
controversies. Social Studies of Science, 6, 349-368.
ROTHMAN, R.A. (1972). A dissenting view on the
scientific ethos. British Journal of Sociology, 23, 102-108.
ROYAL SOCIETY (1981). A study of the scientific
information system in the United Kingdom. London:
The Society.
RYLE, G. (1949). The concept of mind. London:
Hutchinson.
SCOTT, W.A. (1974). Inter-referee agreement of some
characteristics of manuscripts submitted to the Journai
of Personality and Social Psychology. American Psycho-
logist, 29, 698-702.
SINGER, B.F. (1971). Toward a psychology of science.
American Psychologist, 26, 1010-1 0 15.
SINGLETON, A. ( 1976). Journal ranking and selec-
tion: a review in physics. Journal of Documentation, 32(4),
258-289.
SMALL, H. ( 1976). Structural dynamics of scientific
literature. International Clusszfication, 3 (2), 67-74.
SMALL, H.G. (1978). Cited documents as concept
symbols. Social Studies of Science, 8, 327-340.
SMALL, H.G. (1980). Co-citation context analysis and
the structure of paradigms. Journal of Documentation,
36(3), 183-196.
SMALL, H.G. (1982). Citation context and content
analysis. in Dervin, B.J. and Voigt, M.J. (eds.). Progress
in communication sciences. Norwood, N.J.: Ablex, vol.
3, 287-310.
SMALL, H. and GREENLEE, E. (1980). Citation
context analysis of a co-citation cluster: recombinant-
100
DNA, Scientometrics, 2(4), 277-301.
SMITH, L.C. (1981). Citation analysis. Library Treizds,
3, 83-106.
STARR, S.L. (1983). Simplification in scientific work:
an example from neuroscience research. Social Studies of
Science, 13(2), 205-228.
STORER, N.W. (1966). The social system of science.
New York: Holt Rinehart & Winston.
SULLIVAN, D. et al. (1977). Co-citation analysis of
science: an evaluation. Social Studies of Science, 7, 223-240.
SWANSON, D. (1977). Information retrieval as a trial-
and-error process. Library Quarter&, 47( 2), 180-205.
SZAVA-KOVATS, E. ( 1982). Confutation of the
Ortega Hypothesis by citation analysis: a case study of
citation practices. Tudomanyos es Musazaki Tajekottatas,
29( 12), 522-523 (English language summary).
THORNE, F.C. (1977). The citation index: another case
of spurious validity. Journal of Clinical Psychology, 33,
1157-1 161.
TRANIZIY, K.E. (1976) The foundations of cognitive
activity: an historical and systematic sketch. Inquiry, 19,
131-1 50.
TRAN0Y, K.E. (1980). Norms of inquiry: rationality,
consistency requirements and normative conflict. in
Hilpinen, R. (ed.). Rationality in science. Dordrecht:
Reidel, 191-202.
TURNER, R.H. (1953). The quest for universals in
sociological research. American Sociological Review, 18,
604-61 1.
VOOS, H. and DAGAEV, K.S. (1976). Are all citations
equal? or did we op.cit. your idem? Journal o f Academic
Librarianship, 1( 6 ) , 19-2 1.
VOVERENE, 0.1. (1981). Scientists’ attitude towards
bibliographic references. Nauchno- Tekhnicheskaya Infor-
101
matsiya, Seriya I, 10, 8-12, cover-to-cover translation in
Scientzjic and Technical Information Processing, 5 , 1981)
WARTOFSKY, M.W. (1982). Medical knowledge as a
social product: rights, risks and responsibilities. in
Bondeson, W.B. et al. (eds.). New knowledge in the
biomedical sciences: some moral implications of its
acquisition, profession and use. Dordrecht: Reidel,
1 13-1 30.
WATSON, J.D. (1970). The double helix. Harmonds-
worth: Penguin.
WEINER, J. (1977). The footnote fetish. Telos, 31,
172-1 77.
WEINSTOCK, M. ( 1971). Citation indexes. Encyclo-
paedia of Librav €Y Information Science, 5 , 1 6 4 0 .
WEST, S.S. (1960). The ideology of academic scientists.
I R E Transactions on Engineering Management, EM-7 (2),
54-62.
WHITLEY, R.D. (1969). Communication nets in
science: status and citation patterns in animal physio-
logy. Sociological Review, 17 (2), 2 19-233.
WHITLEY, R.D. (1970). The formal communication
system of science: a study of the organisation of British
social science journals. Sociological Review, 16, 163-1 79.
WITTGENSTEIN, L. (1961). Tractatus logico-
philosophicus. London: Routledge & Kegan Paul.
WOLFF, W.M. (1973). Publication problems in psycho-
logy and an explicit evaluation schema for manuscripts.
American Psychologist, March, 257-26 1.
WOOLGAR, S. (1980). Discovery: logic and sequence
in a scientific text (I). in Knorr, K.D. et al. (eds.). The
social process of scientific investigation. Dordrecht:
Reidel, 239-268.
XHIGNESSE, L.V. and OSGOOD, C.E. (1967). Biblio-
graphic citation characteristics of the psychological
102
journal network in 1950 and 1960. American Psychologist,
22, 778-791.
ZIMAN, J.M. (1968). Public knowledge: an essay
concerning the social dimension of science. Cambridge:
Cambridge University Press.
ZIRKLE, C. (1954). Citation of fraudulent data. Science,
120, 189-190 (letter to the editor).
ZUCKERMAN, H. and MERTON, R.K. (1971).
Patterns of evaluation in science: institutionalisation,
structure and functions of the referee system. Minerva,
9( l ) , 66-100.
ZUNDE, P. (1971). Structural models of complex
information sources. Information Storage €Y Retrieval, 7,
1-18.

103