Bursts! Theoretical Fashions in the Study of International Organizations - A Bibliometric Analysis

1

Bursts! Theoretical Fashions in the Study of International Organizations - A

Bibliometric Analysis

Felix S. Bethke

University of Greifswald

Christian Bueger

Cardiff University

Paper for presentation at the 55th

Annual Conference of the International Studies Association,

Toronto, Canada, March 2014.

DRAFT – Comments are welcome!

Abstract: What are the drivers of disciplinary change and intellectual innovation? How can

they be described? Frameworks from the philosophy of science tend to dominate the

discussion in IR and other social sciences. Neither Kuhnian paradigm shifts, nor Laktaosian

research programmes provide however a sufficient heuristic to capture the micro-shifts in

disciplinary developments or to understand drivers of change. In this paper we develop and

test an alternative approach to understand the mechanisms and drivers of scientific progress in

IR. More specifically, we examine one core mechanism of progress in social science, which

has not been thoroughly understood neither in science studies in general, nor in the study of

IR in specific, that is, the role of fashions in driving disciplinary developments. We develop

an approach that understands the progress of scientific disciplines as driven by sequences of

theoretical fashions. A discipline driven by theoretical fashions exhibits substantial variance in

the popularity of theories and concepts. Ideas rise to prominence, are incorporated into the

mainstream of scholarly research and fade away when alternative approaches gain

momentum. We test the appropriateness of the fashion approach for a subfield of IR

publications related to international organizations (IOs). We use burst detection analysis to

identify bursts of cited references in a sample of articles that deal with IOs as a topic. Our

results largely support our theoretical expectations about fashions in IO-research. Over time,

IO-studies experienced multiple reference bursts of particularly influential publications,

which can be linked to specific ideas representing theoretical fashions.

Keywords: Bibliometrics; Burst Detection; Sociology of the Discipline of International

Relations; International Organizations; Fashions;

2

1. Introduction

Discussions in International Relations (IR) on theoretical innovation, progress and the

organizational structures and mechanisms of the discipline are often related to frameworks of

the philosophy of the sciences. Scholars discuss the ideas of Karl Popper, Thomas Kuhn and

Imre Lakatos in order to advance a sociology of the discipline. However, given that most

frameworks in the philosophy of the sciences were developed on the empirical ground of such

“hard” sciences as Physics, Chemistry or Biology, it appears questionable to transfer these

approaches to social sciences such as IR. In this paper we develop and test an alternative

approach to understand the mechanisms and drivers of scientific progress in IR. More

specifically, we examine one core mechanism of progress in social science, which has not

been thoroughly understood neither in science studies in general, nor in the study of IR in

specific, that is, the role of fashions in driving disciplinary developments. Recognizing that

disciplines are primarily constituted by social practices (Rouse 1996, Bueger and Gadinger

2007, Bueger 2012), we develop an understanding of disciplines as driven by sequences of

theoretical fashions. A discipline driven by theoretical fashions exhibits substantial variance in

the popularity of theories and concepts. Ideas rise to prominence, are incorporated into the

mainstream of scholarly research and fade away when alternative approaches gain

momentum.

We test the appropriateness of the fashion approach for a subfield of IR publications

related to international organizations (IOs) using a sample of 2742 articles published in

academic journals between 1956 and 2012. To identify fashions in IO-research we rely on

bibliometric tools in general and the method of burst detection in specific. Bibliometrics

refers to the analysis of academic literature via quantitative methods (Broadus, 1987,

Pritchard, 1969). Bibliometrics has become a more and more established tool to empirically

describe intellectual developments within the discipline. We use burst detection analysis to

identify episodes of high popularity of scholarly work, as represented by the frequency of

cited references in our sample of articles on IOs.

Our results corroborate the argument made by previous studies on the sociology of the

discipline (Katzenstein et al., 1998:649) that social science in general and IR in specific

cannot be understood in terms of paradigm shifts and continuous change between periods of

normal science and periods of scientific revolution as described by Kuhn (1962). Instead, our

analysis reveals a novel perspective with regard to progress and innovation in the history of

IO research. We find that IO research was driven by six different theoretical fashions, namely

(1) functionalism (2) decision-making within IOs, (3) regimes, (4) interdependence, (5)

constructivism, and (6) IOs as part of the liberal peace. These ideas represent episodes of

theoretical fashions, where scholars focused their attention on a particular aspect of IOs for a

given time period.

The paper is organized as follows. In section two, we review related literature on the

sociology of IR as well as bibliometric studies related to IR. We build on these previous

studies to develop our own approach for the analysis of fashions in IO research, which we

describe in section three. In section four we describe the main aspects of our data as well as

the method of burst detection, which we use to identify fashions in IO research. The results of

3

our analysis are described in section five. In section six we discuss the results of the empirical

analysis and highlight areas of further research.

2. Related Literature: Sociology of IR and Bibliometrics

Discussions in IR on theoretical innovation, progress and the organizational structures and

mechanisms of the discipline were long dominated – with some minor exceptions (Hoffmann

1977, Platig 1967) – by references to frameworks of the philosophy of the sciences. Such

accounts tend to rely on a correspondence theory of truth and hence reduce the working of

disciplines to the discussion of the accuracy of disciplinary representations. They hence easily

slide into simplistic understanding of progress and innovation as a process of getting closer to

reality or strengthening correspondence between the world and the theory of it. The debate

over the status of realism in the discipline of the late 1990s and early 2000s made the theories

of science of Thomas Kuhn (1962) and Imre Lakatos (1978) influential as alternatives to

conceptualize progress and change (Elman and Elman 2002a, 2002b; Keohane 2001, Vasquez

1997, Hellmann 1999, Guzzini 1998). Although much of the interest in Kuhn and Lakatos, as

Molloy (2003:71) notes, primarily was to provide some “philosophy of science gloss” to the

debate, the discussion opened new ways of thinking about disciplinary structures and

progress. Thinking in terms of paradigms (Kuhn) or research programmes (Lakatos)

introduced the unit of ‘scientific collectives’ whose internal and external relations provide the

core structures of scientific disciplines. Thinking the discipline as constituted by collectives of

actual scientists (each having respective, theories, models, methods, vocabularies and puzzles)

was one important move for shaping today’s sociology of the discipline.

Three other related moves give the sociology of the discipline its contemporary contours.

This was firstly the discovery of IR as an object of intellectual history which led to new

accounts of the history and progress of IR – often challenging received wisdoms substantially,

such as the birth of the discipline and the first great debate as the locus where the narrative (or

myth) of International Relations begins (e.g. Schmidt 1994, 1998, 2002, Ashworth 2002,

Quirk and Vigneswaran 2005, Wilson 1998, Guilhot 2008). Secondly, further ideas from

contemporary sociology of science were introduced – a field that since Kuhn formed a

separate disciplinary identity and produced quite substantial empirical work on the structure

and practices of sciences (e.g. Zammito 2004). Waever (1998) was perhaps the first to point to

this body of literature, followed by a substantial range of further studies (e.g. Breitenbauch

and Wivel 2004, Bueger and Gadinger 2007, D’Aoust 2012). This included, aspects such as

the relations between scholars and funders, language and discursive structures or the

institutional environment in which academics operate. Thirdly, a substantial interest in IR

outside the North-American or Anglo-American context emerged. This interest was spurred

by a perceived hegemony of Anglo-American IR and the belief that a discourse of

emancipation from Anglo-Americanness was required. A growing range of national case

studies were produced initially on IR in European countries, but then increasingly on non-

Western IR (see among many others the chapters in Jorgensen and Knudsen 2006 or Waever

and Tickner 2009).

4

These four moves together formed the field of contemporary sociology of IR. Not the least

because of the intensifying dialogue between the sociology of science and IR and the

increasingly broader empirical evidence base gathered through case studies, the availability of

citation databases or surveys such as the Teaching, Research and International Policy (TRIP)

project (Peterson et al., 2005), the field today has considerably matured. If the sociology of IR

in the 1990s and early 2000s tended to be characterized by a lack of reflexivity, an often crude

form of eclecticism and the lack of accurate methods for the study of the discipline, this has

fundamentally changed.

If anything than the core outcome of the nascent field of the sociology of the discipline was

to reveal that IR, like other disciplines, is constituted by a rich pattern of various practices

(Bueger 2012) and that there is considerable variation of these practices across tempo-spatial

contexts. Such practices range from the external relations with IR’s funders, audiences and

neighboring disciplines (Bueger and Gadinger 2007) to practices such as researching and

interpreting, writing and publishing, teaching and learning, presenting and conferencing, or

interviewing and hiring. The community of IR scholars is constituted by such activities as

much as by a range of objects and technologies, such as books and databases, as well as a tacit

stock of knowledge and normative evaluations (concerning of what is IR, what is good and

bad research, what is publishable and what not). As Hagmann and Biersteker (2012) phrase it:

“IR is today led and governed by a multiplicity of differently configured material

and social, local, and transnational recognition and empowerment practices.

Higher education policies, financial resources, publication opportunities, citation

patterns, research infrastructures, hiring rationales, and career advancement rules

pose unevenly influential conditions to IR scholarship in different places”.

Hagmann and Biersteker (2012:5)

Notably the qualitative studies described above highlight such dimensions of the discipline

and provide a growing basis of evidence of what constitutes IR in different sites. If to some

degree qualitative research and single case studies have dominated the sociology of IR, there

is an increasing awareness that the discipline is best understood by mixing methods, and in

consequences there is a growing awareness of the importance of large-n quantitative analysis

to study the patterns of the discipline.

Measuring the Discipline: Bibliometrics and the Quantitative Approach in the Sociology of

IR

Two types of quantitative research approaches have been developed to study the discipline.

This is firstly survey research. Since 2005, the TRIP survey run by a team at the College of

William & Mary has become one major source of data about the discipline (Peterson, et al.,

2005). The survey develops data on the theory preferences of individual scholars, the status

and recognition of institutions and individuals (influence on the field, most interesting work,

etc.). Sub-sequentially, the TRIP-data has been used to address important questions for the

5

sociology of IR, namely the theoretical, methodological, and epistemological diversity of the

discipline (Maliniak et al., 2011), a potential gender citation gap in IR (Maliniak and Powers,

2013), the relationship between academics and policymakers in IR (Hundley et al., 2014,

Parks and Stern, 2014), or ideological bias of IR scholars (Rathbun, 2012).

The study of publication and citation patterns, which is generally known as bibliometrics,

provides the second major quantitative gateway for disciplinary sociology. Bibliometrics as a

scientific discipline is usually described as the analysis of academic literature using

quantitative methods (Broadus, 1987, Pritchard, 1969). If surveys provide data on individual

beliefs, preferences and evaluations, the study of publication and citation practices analyzes

inter-subjective and collective forms of meaning and actions that shape the discipline. Two

genres of publications have come in focus: (1) textbooks and syllabi and (2) journal articles.

The analysis of textbooks provides an understanding of how the discipline is taught and new

members are socialized into it. Nossal (2001), for instance, has focussed on 14 U.S. textbooks

to identify what visions of the world shape IR and concludes that these texts “portray the

world to their readers from a uniquely American point of view” (Nossal 2001:183). Biersteker

(2009) and Hagman and Biersteker (2012) investigate the theoretical and methodological

preferences as well as gender and language biases engrained in the core readings of IR

education programmes. Hagmann and Biersteker (2012) point out that the analysis of the

publications by which IR is taught provides access to a richer universe of IR including notably

those members of the discipline which do not engage with IR’s high ranking journals.

The analysis of journal articles has been a core mean for the first waves of the sociology of

the discipline. Holsti (1985), Wæver (1998), or Friedrichs (2004) drew on the study of

publication patterns. These early statistical analyses mainly investigated the authorship of

articles to reveal whether there is an US-American hegemony. Since these studies the use of

bibliometrics has significantly advanced and turned to the study of citations to understand

scholarly communication (Kristensen, 2012), gender dynamics (Maliniak et al., 2013,

Mitchell et al., 2013) or geographic representation in the discipline (Aydinli and Mathews,

2000, Kristensen, 2013, Tickner and Waever, 2009).

Another part of the literature investigates the bibliometric patterns of subfields of IR. For

instance, Jensen and Kristensen (2013) analyze bibliometric patterns in European Union (EU)

studies, which connects a diverse field of scholarship that deals with the EU as a topic. Here,

the bibliometric analysis focuses on the relationship between EU journals in terms of

citations, as well as the disciplinary and geographical clusters of the field. The results suggest

that although the EU as a topic connects numerous disciplines such as Law, History,

Economics and Political Science, the latter dominates the field. Furthermore, EU-studies is

dominated by two core journals and contributors are mostly located in North-America and the

EU member states. Liu, Hong and Liu (2012) use bibiometrics to identify the structure of

globalization research. They show the growth of this research field, describe main journals

and authors, the most countries that host authors, and draw on a keyword analysis to

document research trends. Reid and Chen (2007) develop a similar depiction of the social

structure of terrorism research drawing on a mapping approach, identifying core clusters of

researchers and themes. Gordon (2007) shows that terrorism research is characterized by a

very fluid structure. As he shows, the growing amount of literature in this field is produced

6

mostly “by one-timers who ‘visit’ the field, contribute one or two articles, and then move to

another subject area. This research pattern does not contribute to the regularity and constancy

of publication by which a scientific discipline is formed and theories and paradigms of the

field are created.” Sillanpää and Koivula (2010) study conflict research and identify the main

themes and gaps in the literature. As they show Democratic Peace Theory “constitutes a

powerful discursive core of contemporary conflict research, affecting most other discourses as

well.” Most surprisingly they show that “instead of systemic foci, contemporary conflict

research is dominated by the investigation of dyadic forms of interaction and that, somewhat

surprisingly, the substantive focus of the most frequently cited research has remained on

interstate war.” Bueger and Bethke (2014) draw on bibliometric tools to study the concept of

the failed states and the interactions through which the concept has achieved its prominence.

Moreover, they reveal that no actor or perspective clearly dominates the meaning of the

concept and there is a persistent struggle to define the concept.

In general, this literature review has shown that scholars are preeminently concerned with

using bibliometric methods either to reveal hidden power structures of the discipline and point

to imbalances in terms of gender or geographical representation or engage in a more

explorative bibliometric analysis of subfields or concepts within IR. There appears to be little

effort to study the more general questions about scientific progress and change, which is

addressed in qualitative studies within sociology of science. Meanwhile, bibliometric studies

outside of IR, have engaged more extensively into empirical work that analyzes general

assumptions about scientific progress (Bort and Kieser, 2011, Chen, 2006, Chen, 2004, Chen,

2012, Shiffrin and Börner, 2004). Chen (2012) developed an indicator to measure the

presence of normal science and scientific revolutions respectively. His approach combines

network measures and citation frequencies to identify paradigms shifts and turning points.

Bort and Kieser (2011) analyzed how the use of theoretical concepts in journal publications

within the field of organization studies is driven not solely by scientific research problems and

objective criteria but instead subject to fashions. In the next section, we build on these studies

to provide an elaboration of theoretical fashions as a mechanism of scientific progress in IO-

studies.

3. Theoretical Approach: Understanding Progress and

Innovation

The rise and fall of ideas has been a persistent theme in the philosophy, history and

sociology of science. For decades this implied a tripartite relation or a separation of labour in

which a philosophy of science is responsible for scientific content and concepts, the past and

the ‘reconstruction of progress’ is the meal for the historians, and what is left, or cannot be

explained otherwise, falls in the realm of the sociologist. The dominant picture of progress

that was developed, expressed in Popperian falsificationism (Popper, 1959) and Mertonian

normative structures (Merton, 1973), was that of scientific change as a rational, incremental

process of increasing certainty and getting closer to reality.

7

With the publication of Thomas Kuhn’s seminal Structure of Scientific Revolution the

tripatriate organization as well as the rational understanding of progress was successfully

challenged. Research in the sciences after Kuhn builds up on the assumption that time,

content and any social dimension of the sciences need to be integrated and that scientific

change is only poorly understood by reference to correspondence theories of truth. Kuhn’s

major contribution (and those following in his lines) was twofold. First, he successfully

challenged the idea that the development of science is a continuous process. In introducing the

concepts of normal science and revolutionary shifts, he demonstrated the discontinuities and

breaks that mark the history of science. As he argued, not linear progress, but crisis is the

driver of scientific change. Second, Kuhn emphasized the importance of the scientific

community. He argued that scientists form a closed community whose research draws on a

well-defined range of problems and who use methods adapted to this work. Instead of

determined by the rules as set out by philosophers of science, or the norms suggested by

Mertonians, scientific practices are attempts to solve concrete problems, regarded as

‘puzzles’. For Kuhn the problems stem from what he called ‘paradigm’ or ‘disciplinary

matrix’: A set of scientific achievements, theories and methods that are taken for granted by a

group of scientists. The actions of scientists are hence determined by paradigms, and the

community becomes indistinguishable from this paradigm. The term paradigm refers hence to

a certain shared knowledge or mind set, to shared practices of inquiry and to an actual

community of researchers (Rouse 2003).

Kuhn, as already sketched, was widely received in IR. While it is doubtful that anyone in

IR ever cared to appreciate his complex theorizing, notably his argument for recognizing the

importance of practice (Rouse 2003), his influence started to become problematic with the

inflationary use of the term ‘paradigm’, in which the concept gradually lost its Kuhnian

meaning (e.g. Legro and Moravcik 1999, Smith 1995). Rather than referring to actual

communities of IR researchers, and to practices of inquiry or publishing, the concept was used

to refer to abstract systems of thoughts, concepts and assertions. Moreover, IR was primarily

interested in Kuhn’s concept of incommensurability to mark boundaries between different IR

‘theories’ and provide justifications for misunderstandings and unproductive disciplinary

debates. Rather than understanding IR as a scientific community governed by the paradigm of

studying global politics and trans- or interstate relations as “international relations” – and not

as “macrosociology” or “international law” – the term paradigm was understood as referring

to isms and theories. While it was difficult to argue for revolutionary shifts in IR given the

continuing multiplicity of research approaches, when ‘realism’ lost the cold war, the ‘demise’

of ‘realism’, provided the opportunity to do so. Nonetheless realists are (fortunately) still with

us.

The move to Lakatos triggered by Kuhn, even worsened the situation. Lakatos’

Methodology was initially introduced as a more reflective, methodological way to appraise if

the “work is getting any better” (Elman and Elmann 2002:1) As Frank Gadinger (2002) has

shown in his examination of applications of Lakatos in IR, Methodology transformed into a

“discursive weapon”. Instead of gathering knowledge of the social processes and practice

constitutive of IR and ‘causing’ its development, studies were primarily interested in claiming

this or that ‘research programme’ to be ‘degenerative’ or ‘progressive’.

8

In contrast to the Kuhn-Lakatos exchange, the Kuhn-Bloor-exchange, so important for the

development of contemporary science studies, was never introduced to IR and its absence still

marks a major gap in the discipline. Bloor’s so called ‘strong programme’, sets up a critique

of Kuhn’s work and considerably extends its applicability. This tradition is associated with the

work of scholars conventionally referred to as the Edinburgh school following the writings of

David Bloor and Barry Barnes and the Bath group following Harry Collins. The seminal work

in this context is Bloor’s Knowledge and Social Imagery, first published in 1976. The merits

of Bloor, Barnes, Collins and followers lay in empirically showing the underdetermination of

any theory by data. As case studies exemplify, belief preferences, tactics of persuasion,

opportunistic strategies and local conditions such as equipment and procedures define to a

considerable degree the outcome of scientific practices. Further, controversies are rarely

solved by evidence or by rational means. Controversies come to be settled by diverse

strategies of boundary drawing and of persuasion or even by dishonest means. Hence in this

tradition scientific achievements are better explained by social ‘factors’. Moreover the

perspective shifts from the grand crisis that Kuhn saw as core drivers to the smaller scale and

often everyday controversies over the interpretation of evidence or applicability of concepts

methodologies and apparatuses. As Katzenstein et al. (1998) have argued IR hardly follows

the logic of scientific revolutions and normal science as proposed by Kuhn. Instead IR is, as

they argue, characterized by constant “debates between competing general theoretical

orientations and associated research programs” (Katzenstein, et al., 1998:649). Indeed, this

diagnosis points us to the importance of considering post-Kuhnian sociology of science to

understand the drivers of IR research. Drawing on post-Kuhnian sociology of science

moreover entails a shift away from the scientific community as core unit of analysis. As we

argue, a different way to study scientific progress for the social sciences is to perceive it as a

sequence of theoretical (or methodological) fashions.

Studying Scientific Progress as Sequence of Fashions

In common sense, a fashion is simply a practice, which is perceived as popular. Most of the

time people refer to a specific style of clothing or furniture as being fashionable but fashion

may also relate to a diverse set of other practices such as attitudes towards life (e.g. ecological

sustainability), political preferences (e.g. green parties), and use of technology (e.g. twitter)

just to name a few. Correspondingly, scientific methods, theories and concept may also be

perceived as fashionable. However, the notion of fashion generally contradicts the self-

conception of researchers of being solely devoted to scientific rigor and objectivity with the

aim to solve problems of scientific relevance. As Bort and Kieser (2011:657) put it “Science

Is Supposed to Be Fashion-Free”.

In the following, we provide an outline on how the concept of fashions can illuminate

scientific progress and change, and provide an understanding of how fashions foster

homogeneity through imitation as well as innovation and deviance. Our approach builds on

the general sociological work on fashions by Georg Simmel (1957 [1904]) and a more recent

application of the concept in organization studies by Bort and Kieser (2011).

9

For Simmel (1957 [1904]:543) the concept fashion is defined as “the imitation of a given

example and satisfies the demand for social adaptation”. When following fashions people

imitate someone or something. At the same time, however, Simmel sees fashion as a

mechanism of distinction. Fashions only come into being because they differentiate and

segregate (Simmel 1957 [1904]:545). This dual purpose of fashion makes it a paradox

(Czarniawska, 2011:601). However, the paradox can be resolved by looking at different

temporal phases of a fashion. In general, the evolution of a fashion can be seen as a sequence

of three phases: (1) an early innovator proposes a new practice or perspective, (2) he or she is

followed by a wave of imitators and (3) the fashion is incorporated into the mainstream,

which marks its end. If a fashion does not promise distinction anymore it fades away, which

does not mean that the respective practice disappears but it loses the attribute of being

fashionable (Simmel (1957 [1905]:547). Thus, fashions draw imitators because they promise

distinction, but the more imitators follow a fashion the less distinguishable they become from

the mainstream.

As Joseph Rouse (1996) has pointed out in his philosophy of scientific practices, science

faces a similar dilemma as fashionistas. As Rouse argues the most fundamental

epistemological issue is significance. In order for an innovative argument, empirical finding

or conceptual aparatus to have an effect on and change scientific discourse it needs to be

intelligible. That is it requires to rely on and relate to an existing body of literature. In Simmel

words, in order to innovate one needs to imitate. On the other hand, a scientific innovation

also needs to be a novel contribution, that is, it requires to reconfigure the current state of

affairs, fill a gap or add what hasn’t been there. Following Rouse’s understanding of

publishing as a social practice hence reveals that a publication always has to strike a balance

between significance and intelligibility, that is, having strong relations to the narrative field on

the one hand, and incoherence and innovation, that is, reconfiguring this field, on the other.

With such an outline Rouse points to the core conflict (or even paradox) of scientific

innovation and change. Scientific practices and achievements are intelligible if they have a

place within enacted narratives that constitute a developing field of knowledge, and they are

important to the extent that they develop or transform these narratives (Rouse 1996: 170). The

appeal of such an understanding of scientific practice as narrative reconstruction lies for

Rouse in being an alternative to standard view that scientific work “becomes intelligible and

important against a background of a research community’s shared belief and desires”. Such a

view is not plausible as it overstretches coherence, hence do not consider the interplay

between significance and incoherence and cannot cope with situations in which scientific

practice transforms a community’s prior commitments or changes what counts as the relevant

scientific community. Instead of what constitutes a scientific community and what is its

history and future is frequently at stake.1 What is hence in common among researchers “is a

field of interpretative conflict rather than any uncontested commitments about beliefs, values,

standards, or meanings” (Rouse 1996:172). To engage in one research project rather than

1 Rouse provides a telling example: Why are textbooks and state of the art articles continuously rewritten? If new

results have been produced, why not just publish occasional supplements? The answer is they are ongoing

reconfigurations and integrate new fashions and re-draw the history of the discipline after fashions have

deceased.

10

another is to (attempt to) reconfigure the story that would make sense of that project within its

historical situation.

Our approach to conceptualize scientific progress in terms of fashions also inhibits these

mechanisms of imitation and distinction. We can think of scholars as early innovators, if they

propose a new theoretical perspective, concept or method. If these innovators can gather

enough imitators a fashion comes into being, which dominates a discipline or subfield.

However, the more scholars adapt to the theoretical fashion the more likely the fashion dies,

because it is incorporated into the mainstream of the respective discipline. The death of a

fashion usually is accompanied by the birth of a new one. Therefore, a whole discipline driven

by fashions should exhibit substantial variance in the popularity of theories, concepts and

methods. Ideas gain prominence until they are incorporated into the mainstream of scholarly

research and fade away when alternative approaches gain momentum.

We argue that this characterization of scientific progress as sequence of theoretical fashions

more adequately describes social sciences in general and the discipline of IR in specific. What

Katzenstein et al. (1998:649) describe as “debates between competing general theoretical

orientations and associated research programs” could also be understood as sequence of

fashions. This approach would avoid some of the contradictions that occurred in the

application of Kuhn’s paradigms to the social sciences, as described above.

The difference between a fashion and a paradigm lies in the analytical scope of these

concepts. Whereas a paradigm basically structures how scientist see the world, a fashion only

highlights what part of the world is interesting for scientists. Kuhn essentially assumes that

during periods of normal science, scientists are mostly blind with regard to any abnormal

observations they make, which contradicts their paradigmatic world view. Only if

contradictions add up, scientists change their world view through a period of scientific

revolution until a new paradigm emerges. However, if we instead perceive scientific progress

as subject of fashions, we would observe theories, concepts and methods, which are popular

for some time period but not deterministically for the whole discipline in the sense that every

scientists conforms to the respective dominant fashion. The emphasis shifts from taking

(relatively stable) communities of scientist as the core unit of analysis to a relational

understanding that focuses on the shifts in which new communities of fashionistas are made,

and others decline.

The Practice of Citing or Fashions as Bursts of Cited References

Contrary to Hagmann and Biersteker (2012) we argue that publishing in leading journals

remains the most important practice of a scientific discipline, since publications are the core

symbolic capital in the negotiation of professional status and journal publications are the core

drivers of intellectual change. Yet, we agree that the discipline should not be reduced to the

“published discipline” and that if the intention is to provide general statements about the

structure of the discipline the pedagogical dimension requires consideration. However, for our

aim to study scientific progress and change as sequence of (theoretical) fashions, we consider

publications as an adequate unit of analysis, since publications are easier to observe than other

11

scientific practices and any progress or innovation should sooner or later find its way into

publications.

To identify fashions within a scholarly discipline or a subfield of a discipline, such as IO-

research, we rely on cited references as an indicator of fashions. Therefore, we build on

Small’s (1978) sociology of citations, which sees the main purpose of citing in the ascription

of meaning to documents. As stated by Small (1978):

“[…] when a scientist cites, he or she is creating a link between a concept,

procedure, or kind of data, and a document or documents. In some cases, the

association of idea and document is well established by uniform practice within

the community (Leach's 'standard symbol'). Recurring patterns of terminology

used by citing authors when referring to these documents show that they have

become standardized in their usage and meaning.”(Small, 1978:337)

When being cited, authors and their work are matched to a particular idea and the cited

reference becomes a symbol for that idea. This notion of citation may be illustrated with one

of the most cited articles in IR, namely Taking time seriously: Time-series-cross-section

analysis with a binary dependent variable by Beck et al. (1998). At the time of March 2014

this article has received 819 citations according to the Web of Science (WoS)-database, with

almost half of the citations coming from IR articles. According to Google Scholar, the article

was cited 1821 times. The reason for these high citation rates is that Beck et al. (1998)

introduced a methodological innovation for the study of intrastate conflicts, namely the

statistical modeling of the time dependency of events. Before the publication of the seminal

article by Beck at al. (1998) scholars largely neglected this issue and treated conflict events as

if they were independent. However, our argument is not about the use and misuses of

statistical modeling techniques but about how scholars used the article as a symbol of the idea

of time dependency. This process may be illustrated with a couple of quotes from the articles

that cite Beck et al. (1998):

“To account for temporal dependence in these models, we replace the lagged

dependent variables from the ordered models with a counter (years since last

“fix”) and cubic splines as suggested by Beck, Katz, and Tucker (1998).”

(Copelovitch and Pevehouse, 2013:393)

“Finally, to account for dependence between observations, that is, a history of

peace or conflict in a dyad, we include three splines and a peace years variable

that measure the number of years a dyad has been at peace (Beck, Katz &

Tucker, 1998).” (Lektzian and Souva, 2009:26)

12

“The analysis incorporates statistical corrections now common in pooled time-

series analysis: robust standard errors adjusted for clustering on dyads, and the

Beck, Katz, and Tucker (1998) spline correction for time-dependence among

observations.” (Kinsella and Russett, 2002:1055)

As shown in the quotes above, the article by Beck et al. (1998) is linked to the idea of time

dependency. It becomes a symbol for numerous aspects that are associated with this

phenomenon. Problems caused by time-dependency, solutions to these problems and

alternative approaches are all summarized under the cited reference Beck et al. (1998).

Now if cited references are understood as symbols for ideas, we can observe which ideas

are fashionable during some time interval and which are not depending on their frequency of

citations. We argue that theoretical fashions are associated with a burst of citations for one or

more articles that are linked to the respective idea. A burst of citations indicates that during

some time period a particular source is referenced substantially more often than other

(average) sources of the same kind. The idea is introduced by an early innovator and

experiences an increased citation rate because it attracts followers and imitators, which seek

to distinguish themselves from the mainstream of scholarly research. When enough scholars

follow, the once fashionable idea is incorporated into the mainstream and gives way to a new

theoretical fashion that attracts citations. Given the appropriateness of our approach to

conventionalize scientific progress within a discipline as sequence of (theoretical) fashions,

we formulate the following empirical implications, which we expect to observe in the

bibliometric analysis.

We should observe multiple bursts of citations for particular references

These bursts should last for a considerable long time period until they end and others

take their place

Those references that experience bursts should be linked to a particular idea,

representing a theoretical fashion

Over the entire time period of study, we should observe overlapping sequences of cited

reference bursts that are linked to different ideas/fashions

4. Research Design: Bursts in IO Research

In the following section, we seek to demonstrate the appropriateness of our approach with

a bibliometric study on theoretical fashions in IO-research. First, we outline our reasoning for

selecting IO-research as a case study. Second, we describe the data we use to analyze IO-

research. Third, we describe the method of burst detection, which we use to identify fashions

in IO-research.

13

Case Selection: The Sleeping Beauty of IO-Research

IO-research, along with International Security and International Political Economy, is one

of the core subfields of the discipline of International Relations (IR). If in times of the

discipline’s adolescence international organization (IO) research was a rather boring matter of

international lawyers discussing international treaties, IO research quickly became the

theoretical engine room of the IR discipline (Kratochwil and Ruggie 1986). IR’s core

theoretical debates of the 1970s and 1980s, such as the regime controversy or the entry of

constructivist perspectives, concerned IOs. In today’s discipline the debate on IOs is cast in

broader terms, often under the rather ambiguous header of ‘global governance’. Yet, IO

research maintains its status as the core subfield of IR and theoretical center stage. This is not

only reflected in the fact that the subfields’ journal International Organization is since

decades the top ranked disciplinary journal. The centrality of IO research is also confirmed by

the observation that in contrast to International Security and International Political Economy

there has hardly been a strive or perceived need to develop a sub-disciplinary identity. While

one would frequently refer to Security Studies as a sub-discipline which is somehow related

to IR, but different from it, we would hardly see such a claim for IO research. Given this

centrality of IO research for the discipline, it seems plausible to study it as a case to formulate

broader claims about how the discipline works. The sociology of the discipline of IR has

become a nascent field of research (Waever 1998, Waever and Tickner 2009, D’Aoust 2012).

Quite surprisingly IO research has not become a research object of it. Scholars investigate the

discipline at large, focus on Security Studies, or rather exotic sub-formations of the discipline.

IO research as a promising case for understanding how the discipline works remains an

untouched beauty. To understand how the field works, one is left with a range of excellent

state of the art articles (e.g. Kratochwil and Ruggie 1986, Katzenstein, Keohane and Krasner

1998, Simmons and Martin 2002), which provide a sense of history and direction, but hardly

any insight into the drivers of IO research. Therefore, we consider IO-research as a

representative and interesting case study to advance our argument to understand scientific

progress and innovation in the social science in general and the discipline of IR in particular

as driven by theoretical fashions.

Description of the Data: IO-Studies as Object of Bibliometric Research

Our data consists of all articles recorded in the WoS-database that deal with international

organizations. This refers to all articles that include the term “international organization” in

their title, abstract or among their keywords. The precise search query was:

Topic=("international organization") OR Topic=("international organizations") OR

Topic=("international organisation") OR Topic=("international organisations").

The timespan for the search query was set to 1956-2012. The data was retrieved in

December 2013. This query resulted in a sample of 2742 records, which we use as data in the

bibliometric analysis. Before analyzing this data with regard to the presence of fashions, we

14

evaluate some descriptive statistics of the data in order to gain an understanding of its

statistical properties. Each part of the subsequent data analysis was conducted with a

combination of bibliometric tools provided by the software packages BibExcel (Persson,

2006) and Sci2 (Sci

2 Team, 2009).

We begin our descriptive analysis of the data by evaluating how publications on IOs

evolved over time. Figure 1 shows the annual publication count of articles on IOs.

[Figure 1]

As described in figure 1, publications on IOs start in 1956, which is caused by the fact that

for the social sciences the WoS-database only records articles published after 1955. Until the

beginning of the 1990s publications counts per year on IOs were quite low, and varied

between five in 1961 and 34 in 1977. Beginning in the 1990s publication counts started to

rise. With few exceptions, the number of publications increased steadily until the year 2012.

Particularly large boosts were recorded in the years 1998, 2001, 2005, 2008 and 2012.

Though, this trend maybe a result of a general increase in articles recorded in the WoS-

database. Therefore, the identified pattern is most likely not a particularity of IO-research, but

instead IO-research is just part of the general growth rate of scientific publications (Larsen

and von Ins, 2010). However, as can be seen in Figure 1, there is sufficient data available to

draw conclusions about the history of IO-research.

Next, we look at the main scientific disciplines that are involved in IO-research and

examine, how heterogeneous the research field is with regard to the number of disciplines that

are involved. Furthermore, we look into representative articles from different disciplines, to

evaluate whether they share a common understanding of IOs as an object of investigation. We

examine the disciplinary structure of IO-research via the categories assigned to each article in

the WoS-database. Figure 2 displays the ten most frequent disciplinary categories recorded in

our sample of articles on IOs.2

[Figure 2]

Not surprisingly, most frequently articles are categorized with the label “International

Relations”, which supports our argument about the representativeness of IO-research for IR.

With a total of 833 records, almost one-third of the articles in the sample are labeled with this

category. Political Science, with a label count of 604, is represented the second most

frequently. Correspondingly, the scholarly work on political questions in general and IR-

2 We used the WoS-category WC (the WC field tag) to identify disciplinary involvement in IO research.

Alternatively, we could have used the SC field tag, which refers to an article’s “subject category”. The WoS-staff

assigns SC field tag based on a number of factors such as the journal’s title and its citation patterns (Leydesdorff

and Rafols, 2009:2-3). However, the different disciplinary labels of the SC field tag are kind of ambiguous. For

instance, the most frequent subject category for our sample of articles on IO-research was “Government & Law”,

which mixes the disciplines political science and law. The new WC label is more specific, with regard to the

identification of scientific disciplines.

15

topics in specific dominates IO-studies. Respective articles in our sample with these labels

tackle research topics such as how IOs come into being (e.g. Haftel, 2010), power-relations

within IOs (e.g. Chwieroth, 2008) and IOs as an actor of international politics (e.g. Barnett

and Finnemore, 1999).

Economics and Law are ranked third and fourth respectively, but exhibit considerably less

label counts than IR and political science. Articles with an economics-label mostly discuss

and present research on international financial organizations, such as the International

Monetary Fund (e.g. Reynaud and Vauday, 2009). Articles that are associated with the

discipline Law, engage with IOs as part of the creation and implementation of international

law (e.g. Anderson, 2009). The other disciplinary labels shown in figure 2 are studies that

mainly connect to IO-studies due to the relevance of specific international organizations in the

respective research fields. The World Health Organization is, for instance, particularly

relevant in articles associated with “Public, Environmental & Occupational Health” (Tang et

al., 2009), or the International Monetary Fund and the World Bank for “Planning &

Development”. Likewise IOs are a matter for “Information Science & Library Science” due to

their role in setting norms and standards for this field. Equal can be said for “Environmental

Studies” were IOs are central regimes for global environmental policies.

The descriptive analysis highlights some important aspects of the data. First, the analysis

of temporal variation in publication counts has shown that there is sufficient data available to

analyze fashions in IO-studies over a longer time period. However, it is important to note that

for early time periods in the sample, there is less data available than for later times, which

could bias the results of subsequent analyses. Second, the analysis of disciplinary labels has

shown that the sample exhibits a large degree of homogeneity, in the sense that although

numerous different disciplines are involved in IO-research, all articles share a common

understanding of the research topic and there appear to be no misconceptions about the

meaning of IOs across the different disciplines.3 Therefore, we assume that our data exhibits

sufficient face-validity to answer questions about fashions in IO-research.

Burst Detection to Identify Fashions

To identify theoretical fashions in IO-studies, with the data described above, we analyze

the cited references of the articles in the sample. As described above, we build on Small’s

(1978) conception of citations and argue that cited references are symbols for a particular

idea. Accordingly, the number of citations a document receives can be thought of as a measure

of importance of the corresponding idea. With this understanding of cited references as

symbols for ideas, we conceptualize theoretical fashions as time periods, where an idea,

represented by a particular document, experiences a considerable increase in being referenced.

We seek to identify ideas that rose to prominence in IO-studies as indicated by an active

discussion of the idea during a period in time.

3 In a similar bibliometric analysis of the concept “failed state” (Büger and Bethke, 2014), we had to do

additional data cleaning, because our search query identified a number of articles, which were unrelated to the

concept. For instance, some articles in the sample were from the scholarly discipline of computer science and

referred to the failed state of certain algorithms in their abstract. For our sample on IO-studies such additional

data cleaning appeared to be unnecessary.

16

For the analysis of fashions in IO-studies we proceed in two steps. First, we use a burst

detection algorithm developed by Kleinberg (2002) to identify highly cited references, which

may have the potential to represent a fashion. A burst is a period of increased activity, which

generally relates to our conception of fashion, described above. Second, we investigate how

the articles in our sample cite the respective bursty references. Here we seek to explore

whether the identified bursty references can in fact be linked to a particular idea that in turn

represents a (theoretical) fashion.

The respective burst detection algorithm that we use was initially developed to analyze

data streams over time in order to identify time intervals, where the frequency of occurrence

of the data changes significantly. Although, burst detection was originally used to analyze

arrival rates of e-mail messages, it can easily be applied to bibliographic records (e.g. Ord et

al., 2005).4

In order to apply the method, we structure our sample of articles as discrete time-series

data. Articles arrive in discrete batches, which means, each year a varying number of articles

on IOs was published and recorded in the WoS-database. Each of the different articles

contains a number of cited references. Thus, we analyze a “cited reference-stream”, which is

represented by the arrival rates of cited references each year. For each individual cited

reference the algorithm generates the fraction of occurrence in the batched arrival of all cited

references. The algorithm employs a probabilistic model to determine, whether each reference

is in the state of a burst or not. The non-burst state corresponds to the average rate of

appearance of each reference. A reference is in a burst-state if its occurrence during a year is

substantially above the average occurrence of cited references during that year. To be in a

burst state, a reference has to have at least twice the citation rate of the references in the non-

burst state. Using this approach, we can generate a list of the most significant reference-bursts

and the time intervals they occurred in the history of IO-studies. We expect to identify seminal

publications representing theoretical innovations for the analysis of IO as bursty for some

time period, when they were dominant.

5. Empirical Analysis: Detecting Bursts

The burst detection analysis of the sample described above identified 12 cited references to

be in a bursty state for some time period. A complete list of the burst periods along with the

full bibliographic reference of these publications is reported in the appendix of this paper

(table 1 and table 2). In order to better explore the results, we plotted the cited reference bursts

as a horizontal bar graph, which is shown in figure 3.

[Figure 3 about here]

4 Scholars of bibliometrics also highlight the potential of burst detection for identifying paradigm shifts, which

indicate the presence of a period of scientific revolution (Chen, 2004:5310). Furthermore, the method is applied

to study how information spreads over the internet, e.g. in blogs (Kumar et al., 2005).

17

As shown in figure 3, during the first thirteen years of the study, no reference burst is

detected, which may be due to a lack of fashions in IO-studies or lack of available data. As

shown in figure 1, for the early years of IO-research, there are only few articles recorded in

the WoS-database. The first reference burst occurs in 1969 and is linked to Haas (1964). What

also becomes evident from figure 3 is that there appears to be a considerable time lag between

the publication year of a source and the onset of a burst. None of the publications experienced

a burst in the same year as they were published. This intuitively makes sense, since the

scholarly community needs time to evaluate new ideas and the peer-review process also

causes a delay in scholarly response to emerging trends. The lag between publication and

burst onset also corresponds to our argument about the development of fashions. After

innovators have proposed a new idea, it takes some time until imitators follow and initiate the

fashion.

With regard to burst lengths we see considerable variance among the references. The

shortest burst-length in our sample is two years and the longest burst-length is 20 years. The

source with the longest burst-length is the book Power and Interdependence: World Politics

in Transition by Keohane and Nye (1977). The average length of reference-bursts in our

sample is 7.2 years. However, since two of the bursty publications are censored, which means

their burst period did not end at the end of the study time in 2012 the average burst lengths

may be not an appropriate measure of central tendency. Accordingly, we also calculated the

median survival time of reference bursts in the sample, which is given as six years. This

means half of bursts in the sample disappear after their sixths year of burstyness. The length

and distribution of reference bursts across time largely corresponds to our expectations. We

detected multiple reference bursts related to a variety of different publications. Furthermore,

these reference bursts survive for a considerable long time, but do not last for the whole time

period under study, which corresponds to our argument about the retention time of fashions.

Publications rose to prominence during the history of IO-studies for some time period until

the fade away and other publications take their place. During the early time period from we

see a sequence of overlapping reference bursts. However, an interesting feature of the data is

that over time the presence of contemporaneous reference bursts increases. From the 1950s

until 2000 only one or two bursts are present at the same time. Since 2003, however, three to

five contemporaneous burst can be identified. In order to examine whether these different

publications also relate to different ideas we need to investigate how these publications are

referenced by the articles in our sample.

Matching References Bursts to Theoretical Fashions

The first burst from 1969 to 1977 is Haas’ seminal contribution Beyond the Nation-State:

Functionalism and International Organization, which can easily be linked to a particular idea

that can be seen as a theoretical fashion. The articles in our sample that reference Haas

(1964), link this publication to the idea of (neo-)functionalism (e.g. Finkelstein, 1974:513,

Keohane, 1975:363, Kihl, 1971:338, Rochester, 1986:787), which became the dominant

explanation of political integration in IO-studies during that time period. Generally,

functionalism is an approach to explain global peace and international integration as the

consequence of cooperation between states in limited (functional) areas, which are often of

18

technical nature. Functional cooperation in turn fosters so called spill-over effects, which

means that cooperation (sometimes unintentionally) spills over from one functional area to

others and eventually supranational global governance can be established. Neofunctionalist

approaches, which were influenced by the process of European integration in the 1950s,

further emphasized the relevance of IOs to generate spill-over effects (Haas, 1964, Haas,

1958). How Haas’ publication is linked to the idea of functionalism may be illustrated by a

quote from Rochester (1986), who described the status of Haas in the following way,

“Influenced by Mitrany's functionalist thought-which called for supranational

institution building by first promoting international cooperation in relatively

technical, "apolitical" areas, later spilling over into more controversial areas

("federalism by installments")- Haas and so-called neofunctionalist theorists

accepted the functionalist premise that cooperation had to be learned rather than

imposed but argued that politics had to be restored to the equation if learning was

to eventuate in a larger political union.” (Rochester, 1986:788, emphasis added)5

Hence, for Rochester Haas’ work represents a whole school of thought, namely

neofunctionalism.

The next burst of Cox and Jacobson (1973) captures the growing interest of scholars to

look inside IOs. Together with other prominent scholars of the field at that time, Cox and

Jacobson (1973) studied the decision-making process in eight different IOs, using a common

theoretical framework of analysis. Their approach initiated a new fashion in IO-studies,

namely to focus on actors and bargaining within IOs. This fashion in IO-studies lasted for 13

years from 1974 to 1986. The articles in our sample that reference Cox and Jacobson (1973)

correspondingly attribute the idea of opening the black box of IOs to them, which essentially

means to look at the political process of decision-making within IOs and not on the formation

of IOs or the substantial implications of their policies (Finkelstein, 1974:491-92, Jonsson,

1986:42, Kratochwil and Ruggie, 1986:755-56, McCormick, 1982:544). For instance,

Kratochwil and Ruggie, (1986) cluster IO-studies into four analytical foci, namely (1) formal

institutions, (2) institutional processes, (3) organizational roles and (4) international regimes.

For the second analytical focus, they cite Cox and Jacobson (1973) as the “most

comprehensive work in this genre” (Kratochwil and Ruggie, 1986:756).

Keohane and Nye (1977) pioneered the ideas of interdependence and international regimes

which rose to prominence in IO-studies.6 Ranging from 1979 to 1998 their book produced the

longest burst in our sample of cited references. Many articles in our sample cite Keohane and

Nye for their definition of the regime concept (Jonsson, 1986:44, McCormick, 1982:544) or

more general discussions of IOs and international regimes (Gallarotti, 1991:190-91, Ness and

5 The quoted paragraph ends with a footnote where Haas (1964) is referenced.

6 At the same time, Ruggie (1975:570-73) also did essential work to introduce the concept international regimes

into IR and IO-studies. However, whereas Ruggie’s (1975) conceptualization represents a constructivist

understanding of the concept, Keohane and Nye (1977) worked within the rational choice framework (see

Katzenstein, et al., 1998:660).

19

Brechin, 1988:249, Shih, 1989:176). Similarly, but less frequently, articles refer to Keohane

and Nye (1977) when they discuss the idea of complex interdependence (e.g. Drury,

1998:503). The reason, for the lengths of the burst related to Keohane and Nye (1977) may be

that the authors managed to be credited not only for one but two ideas that became part of the

standard repertoire of IO-scholars. For instance, Drury (1998:503) cites Keohane and Nye

(1977) rather nonchalantly when he argues that “the world economy has become increasingly

interdependent”. However, becoming part of the standard repertoire also indicates the death of

even the longest fashion in our sample. Correspondingly, the year of Drury’s citation marks

the end of the reference burst for Keohane and Nye (1977).

Next on our list of bursts is a literature review of IO-studies by Kratochwil and Ruggie

(1986), which led to a reference burst from 1992 to 2000. Although literature reviews usually

tend not to propose innovative ideas in a research field, Kratochwil and Ruggie (1986)

discussion of the epistemological foundation of constructivist IO research was considered

path-breaking. When examining how the articles in our sample cite Kratochwil and Ruggie

(1986) it turned out that many articles refer to this publication in order to advance substantial

arguments like the idea of norms in IO research (e.g. Abbott and Snidal, 1998:8, Katzenstein,

et al., 1998:674) or a critique of positivist approaches to the study of IOs (e.g. Gale,

1998:260). Therefore, we consider the burst associated with Kratochwil and Ruggie (1986) as

representative for the first wave of constructivist scholarship in IO-studies.

Beginning in 2003, we identified one of the first series of reference bursts signifying an

intensification of bursts. All bursts relate to democratic peace theory and the liberal peace

discourse, namely Russett, Oneal and Davis (1998), Oneal and Russett (1999), Russett and

Oneal (2001), Maoz and Russett (1993), and Boehmer, Gartzke and Nordstrom (2004). These

publications advance the idea that IOs promote peace among states, arguing that if states share

IO membership they are less likely to engage in violent conflict with each other. Most articles

in our sample correspondingly cite these publications rather unspecific in the context of

empirical research on the liberal peace (e.g. Ashley Leeds and Mattes, 2007:195, Bearce,

2003:348, Hafner-Burton and Montgomery, 2008:118, Haftel, 2007:221, Hansen et al., 2008,

Ish-Shalom, 2008, Kim and Rousseau, 2005:524). Some examples from the articles in our

sample illustrate how the liberal peace emerges as a fashion in IO-studies.

“More recently, advocates of the normative explanation, such as Bruce Russett,

James Lee Ray, and John Oneal (Oneal and Ray, 1997; Oneal and Russett, 1997,

1999; Russett et al., 1998), have pointed to the interaction of democratic norms,

international institutions, and economic interdependence as the force behind

democratic peace.”(Ish-Shalom, 2008:286, emphasis added)

“IOs that are highly institutionalized, especially with respect to dispute resolution

mechanisms, will be more effective at promoting cooperation among members

and will have greater tools at their disposal for managing conflicts among member

states. IOs may also be more active conflict managers in world politics if their

membership is more democratic because democracies are amenable to using

20

peaceful and third-party methods of conflict resolution (Maoz and Russett 1993;

Dixon 1993, 1994; Raymond 1994; Mitchell 2002).” (Hansen, et al., 2008:296,

emphasis added)

At the end of the time period of our study, we detected reference bursts associated with a

new generation of constructivism in IO-studies. The bursts related to Barnett and Finnemore

(2004) and Finnemore and Sikkink (1998) are linked to a focus of IO-scholars on questions of

how IOs develop and promote norms of good behavior and legitimate actions and hence have

to considered as having autonomous agency (Barnett and Finnemore, 2004:7). Articles in our

sample reference these publication in order to advance constructivist approaches in IO-studies

in general or particular norm shaping aspects of IOs and bureaucratic culture in specific (e.g.

Gutner and Thompson, 2010:237, Lehtonen, 2009:390, McKeown, 2009:272, Van de Graaf

and Lesage, 2009:297). For instance, Lehtonen (2009) highlights:

“[…] processes whereby international organisations create international norms by

diffusing ideas, shaping a certain ‘repertoire’ of reform, and providing domestic

actors with arguments with which to legitimise their actions (Finnemore and

Sikkink, 1998 and March and Olsen, 1998).” (Lehtonen, 2009:390, emphasis

added)

Finally, the burst related to Weaver (2008) represents a recent focus of IO-scholars on

organizational culture, which reflects a distinct development related to constructivist

theorizing. Weaver (2008) pioneered the concept of hypocrisy traps to describe the process of

how IO-staff develops an independent agenda, which may deviate from the original purpose

of the organization. Since the beginning of 2010, IO-scholars frequently cite Weaver (2008)

when they make arguments about the organizational culture of IOs (e.g. Best, 2012:686,

Gutner and Thompson, 2010:238).

Sensitivity of the Results

The method of burst detection crucially depends on the specification of parameters, which

control the onset and lengths of bursts. Since there appears to be no general consensus on

parameter values, we reran the algorithm with different specifications to assess the robustness

of our results. To control the ease of burst onset, the algorithm uses a gamma parameter,

which represents a value that burst transition is proportional to. The higher the gamma

parameter, the fewer bursts are detected. In additional tests, we increased this parameter value

sequentially until no burst is detected anymore for the cited references in our sample. By

investigating which references drop from the detection list given higher gamma values, we

gain insights about the robustness of the identified bursts. Those references which were

identified as bursts given the highest gamma values are the most robust in our sample. Vice

versa those bursts which disappear with only slightly higher gamma values are more sensitive

to parameter specifications.

21

The most sensitive bursts with regard to parameter specifications were the ones linked to

Boehmer, Gartzke and Nordstrom (2004) and Weaver (2008). Once we slightly increase the

threshold for burst onset these references are not identified as bursty anymore, reducing the

total number of bursty references to ten. Only slightly more robust are the references bursts

for Maoz and Russett (1993), Kratochwil and Ruggie (1986) and Finnemore and Sikkink

(1998). The remaining references bursts appear to be less sensitive to changes in parameter

specification. Given another increase in the threshold value for burst detection, the bursts

related to Haas (1964) and Cox and Jacobson (1973) disappear. After the next increase, bursts

associated with Keohane and Nye (1977) and Oneal and Russett (1999) are not detected

anymore leaving only three burst references in the sample. The three reference bursts that

were the most robust in our sample were Russett and Oneal (2001), Russett, Oneal and Davis

(1998) and Barnett and Finnemore (2004), with the latter being the last reference that is

dropped from the detection list.

6. Discussion of the Results: Times of Fashion

Our burst detection analysis revealed 12 bursts of cited references, which represent

different theoretical fashions in IO-studies. We identified six of such fashions: (1)

functionalism as represented by Haas (1964), (2) the idea of analyzing decision-making

within IOs advanced by Cox and Jacobson (1973), the relevance of (3) regimes and (4)

interdependence for the study of IOs proposed by Keohane and Nye (1977), (5)

constructivism, which first developed in the 1980s and led to a burst related to Kratochwil and

Ruggie (1986). Furthermore, an even more significant fashion of constructivist research

occurred in recent years as represented by the burst of cited references for publications by

Barnett and Finnemore (2004), Finnemore and Sikkink (1998) and Weaver (2008), and (6) the

idea of IOs promoting peace among nation states as represented by reference bursts for

multiple papers associated with the liberal peace discourse (Boehmer et al., 2004, Maoz and

Russett, 1993, Oneal and Russett, 1999, Russett, 2001, Russett et al., 1998). Thus, we can

structure the history of IO-research along six episodes of theoretical fashions. Beginning in

1969 these were neofunctionalism, decision-making within IOs, regimes and interdependence,

constructivism, liberal peace and constructivism again.

In summary, these results of the burst detection analysis largely support our expectations

about a discipline driven by sequences of theoretical fashions. The algorithm identified

multiple bursts of citations for particular references, which lasted for a considerable long time

period. Furthermore, the identified episodes of reference bursts could be linked to a particular

idea, representing a theoretical fashion. More generally, the history of IO-research could be

described as overlapping sequences of cited reference bursts that are linked to different

ideas/fashions. However, the results also pointed out that there is a changing pattern in the

discipline regarding the frequency and duration of bursts. If early generations of research

were characterized by longer term bursts, and usually only one fashion prevailed at a single

point in time, intellectual debate has become more scattered and fast-paced in the past decade.

22

Our sensitivity analysis revealed that there is considerable variance with regard to the

robustness of the identified reference bursts. Correspondingly the associated theoretical

fashions also vary in terms of robustness. Given the results of our analysis, we can considerer

the second burst related to Barnett and Finnemore (2004) as the most robust in our sample,

which means the theoretical fashion of a constructivism oriented towards the analysis of

bureaucratic culture in IO-studies in the recent years appears to be quite significant.

Furthermore, the bursts associated with the regime- and interdependence fashion initiated by

Keohane and Nye (1977) as well as a couple of bursts related to the liberal peace discourse

also appear to be very robust. The ideas of functionalism and decision-making within IOs

appear to be less robust. With regard to the interpretation of these results, further research is

needed, which examines the causes for the variance in robustness. At this point, we can only

consider the second wave of constructivism, IOs within the liberal peace discourse and the

ideas of regimes and interdependence as theoretical fashions in the history of IO-research.

Finally, one major caveats of the analysis and area of further research should be

mentioned. Our analysis lacks an established baseline to interpret the results. Because of a

lack of comparable studies, we are not able to judge whether the number of bursts we

identified is high or low compared to other disciplines or subfields. Similar problems exist

when interpreting the results for the duration of bursty episodes and the parameter

specifications of the algorithm. Therefore, further research is needed, to compare our results

for instance with a sample of articles from IR as a whole or neighboring disciplines such as

Economics or Sociology.

23

7. Appendix

Table 1: “Bursty“ references

Reference Length Start End

Haas (1964) 9 1969 1977

Cox and Jacobson (1973) 13 1974 1986

Keohane and Nye (1977) 20 1979 1998

Kratochwil and Ruggie (1986) 9 1992 2000

Russett, Oneal and Davis (1998) 6 2003 2008

Oneal and Russett (1999) 6 2003 2008

Russett and Oneal (2001) 6 2003 2008

Maoz and Russett (1993) 4 2005 2008

Boehmer, Gartzke and Nordstrom (2004) 5 2006 2010

Finnemore and Sikkink (1998) 2 2008 2009

Weaver (2008) 3 2010 Censored

Barnett and Finnemore (2004) 3 2010 Censored

24

Table 2: Full description of publications identified in the burst detection analysis

Barnett, Michael, and Martha Finnemore. (2004) Rules for the World: International

Organizations in Global Politics. Ithaca, NY: Cornell University Press.

Boehmer, Charles, Erik Gartzke, and Timothy Nordstrom. (2004) Do Intergovernmental

Organizations Promote Peace? World Politics 57:1.

Cox, Robert W., and Harold K. Jacobson. (1973) The Anatomy of Influence: Decisionmaking

in International Organization. New Haven, Conn.: Yale University Press.

Finnemore, Martha, and Kathryn Sikkink. (1998) International Norm Dynamics and Political

Change. International Organization 52:887-917.

Haas, Ernst B. (1964) Beyond the Nation-State: Functionalism and International

Organization. Stanford, Calif.: Stanford University Press.

Keohane, Robert O., and Joseph S. Nye. (1977) Power and Interdependence: World Politics

in Transition. Boston: Little, Brown.

Kratochwil, Friedrich, and John Gerard Ruggie. (1986) International Organization: A State of

the Art on an Art of the State. International Organization 40:753-75.

Maoz, Zeev, and Bruce Russett. (1993) Normative and Structural Causes of Democratic

Peace, 1946-1986. American Political Science Review 87:624-38.

Oneal, John, and Bruce Russett. (1999) The Kantian Peace. World Politics 52:1-37.

Russett, Bruce M. Oneal John R. (2001) Triangulating Peace: Democracy, Interdependence,

and International Organizations. New York: Norton.

Russett, Bruce, John R. Oneal, and David R. Davis. (1998) The Third Leg of the Kantian

Tripod for Peace: International Organizations and Militarized Disputes, 1950-85.

International Organization 52:441-67.

Weaver, Catherine. (2008) Hypocrisy Trap the World Bank and the Poverty of Reform.

Princeton: Princeton University Press.

25

Figure 1: Number of IO-related articles per year

0

50

100

150

200

250

Nu

mbe

r o

f A

rtic

les

1956 1964 1972 1980 1988 1996 2004 2012

26

Figure 2: Most frequent disciplinary categories

0 200 400 600 800

Planning & Development

Sociology

Public, Environmental & Occupational Health

Environmental Studies

Social Sciences, Interdisciplinary

Information Science & Library Science

Law

Economics

Political Science

International Relations

27

Figure 3: Reference Bursts in IO-studies

28

8. References

Abbott, Kenneth W., and Duncan Snidal. (1998) Why States Act through Formal International

Organizations. Journal of Conflict Resolution 42:3-32.

Anderson, Kenneth. (2009) The Rise of International Criminal Law: Intended and Unintended

Consequences. European Journal of International Law 20:331-58.

Ashley Leeds, Brett, and Michaela Mattes. (2007) Alliance Politics During the Cold War:

Aberration, New World Order, or Continuation of History? Conflict Management and

Peace Science 24:183-99.

Aydinli, Ersel, and Julie Mathews. (2000) Are the Core and Periphery Irreconcilable? The

Curious World of Publishing in Contemporary International Relations. International

Studies Perspectives 1:289-303.

Barnett, Michael, and Martha Finnemore. (2004) Rules for the World: International

Organizations in Global Politics. Ithaca, NY: Cornell University Press.

Barnett, Michael N., and Martha Finnemore. (1999) The Politics, Power, and Pathologies of

International Organizations. International Organization 53:699-732.

Bearce, David H. (2003) Grasping the Commercial Institutional Peace. International Studies

Quarterly 47:347-70.

Beck, N., J. N. Katz, and R. Tucker. (1998) Taking Time Seriously: Time-Series-Cross-

Section Analysis with a Binary Dependent Variable. American Journal of Political

Science 42:1260-88.

Best, Jacqueline. (2012) Ambiguity and Uncertainty in International Organizations: A History

of Debating Imf Conditionality1. International Studies Quarterly 56:674-88.

Biersteker, Thomas J. 2009. The parochialism of hegemony: Challenges for ‘American’

International Relations, in Tickner, Arlene and Wæver, Ole, eds. International Relations

Scholarship around the World. New York: Routledge, 308–327.

Boehmer, Charles, Erik Gartzke, and Timothy Nordstrom. (2004) Do Intergovernmental

Organizations Promote Peace? World Politics 57:1.

Bort, Suleika, and Alfred Kieser. (2011) Fashion in Organization Theory: An Empirical

Analysis of the Diffusion of Theoretical Concepts. Organization Studies 32:655-81.

Breitenbauch, Henrik, and Anders Wivel. 2004. “Understanding National IR Disciplines

Outside the United States: Political Culture and the Construction of International

Relations in Denmark.” Journal of International Relations and Development 7 (4): 414–

443.

Broadus, Robert N. (1987) Toward a Definition of “Bibliometrics”. Scientometrics 12:373-79.

29

Bueger, Christian. 2012. “From Epistemology to Practice: A Sociology of Science for

International Relations.” Journal of International Relations and Development 15 (1): 97–

109.

Bueger, Christian, and Frank Gadinger. 2007. Reassembling and Dissecting: International

Relations Practice from a Science Studies Perspective. International Studies Perspectives

8 (1): 90-110. 2007.

Bueger, Christian, and Felix S. Bethke. (2014) Actor-Networking The "Failed State" An

Enquiry into the Life of Concepts. Journal of International Relations and

Development 17:30-60.

Chen, Chaomei. (2006) Citespace Ii: Detecting and Visualizing Emerging Trends and

Transient Patterns in Scientific Literature. Journal of the American Society for

Information Science and Technology 57:359-77.

———. (2004) Searching for Intellectual Turning Points: Progressive Knowledge Domain

Visualization. Proceedings of the National Academy of Sciences 101:5303-10.

———. (2012) Turning Points: The Nature of Creativity. Springer.

Chwieroth, Jeffrey M. (2008) Normative Change from Within: The International Monetary

Fund's Approach to Capital Account Liberalization. International Studies Quarterly

52:129-58.

Copelovitch, Mark S., and Jon C. W. Pevehouse. (2013) Ties That Bind? Preferential Trade

Agreements and Exchange Rate Policy Choice. International Studies Quarterly

57:385-99.

Czarniawska, Barbara. (2011) Introduction to the Special Themed Section: Fashion in

Research and in Management. Organization Studies 32:599-602.

D’Aoust, Anne-Marie. Ed, 2012. “Forum: Sociology/ies of International Relations.” Journal

of International Relations and Development 15 (1): 90–144.

Drury, A. Cooper. (1998) Revisiting Economic Sanctions Reconsidered. Journal of Peace

Research 35:497-509.

Elman, Colin, and Miriam Fendius Elman. Eds. 2002a. Progress in International Relations

Theory. Appraising the Field. Cambridge: MIT Press Press.

Elman, Colin, and Miriam Fendius Elman. 2002b. “How Not to Be Lakatos Intolerant:

Appraising Progress in IR Research.” International Studies Quarterly 46 (2): 231–262.

Finkelstein, Lawrence S. (1974) International Organizations and Change: The Past as

Prologue. International Studies Quarterly 18:485-520.

Finnemore, Martha, and Kathryn Sikkink. (1998) International Norm Dynamics and Political

Change. International Organization 52:887-917.

30

Gale, Fred. (1998) Cave 'Cave! Hic Dragones': A Neo-Gramscian Deconstruction and

Reconstruction of International Regime Theory. Review of International Political

Economy 5:252-83.

Gallarotti, Giulio M. (1991) The Limits of International Organization: Systematic Failure in

the Management of International Relations. International Organization 45:183-220.

Gordon, Avishag. 2007. Transient and continuant authors in a research field: The case of

terrorism. Scientometrics 72(2): 213-224.

Guilhot, Nicolas. 2008. “The Realist Gambit: Postwar American Political Science and the

Birth of IR Theory.” International Political Sociology 2: 281–304.

Gutner, Tamar, and Alexander Thompson. (2010) The Politics of Io Performance: A

Framework. The Review of International Organizations 5:227-48.

Guzzini, Stefano. 1998. Realism in International Relations and International Political

Economy: The Continuing Story of a Death Foretold. London: Routledge.

Haas, Ernst B. (1964) Beyond the Nation-State: Functionalism and International

Organization. Stanford, Calif.: Stanford University Press.

———. (1958) The Uniting of Europe. Stanford University Press Stanford, CA.

Hafner-Burton, Emilie M., and Alexander H. Montgomery. (2008) The Hegemon's Purse: No

Economic Peace between Democracies. Journal of Peace Research 45:111-20.

Haftel, Yoram Z. (2010) Conflict, Regional Cooperation, and Foreign Capital: Indonesian

Foreign Policy and the Formation of Asean. Foreign Policy Analysis 6:87-106.

———. (2007) Designing for Peace: Regional Integration Arrangements, Institutional

Variation, and Militarized Interstate Disputes. International Organization 61:217-37.

Hagmann, Jonas, and Thomas J Biersteker. 2012. “Beyond the Published Discipline: Toward

a Critical Pedagogy of International Studies.” European Journal of International

Relations, OnlineFirst, October 18th

, doi:10.1177/1354066112449879.

Hansen, Holley E., Sara McLaughlin Mitchell, and Stephen C. Nemeth. (2008) Io Mediation

of Interstate Conflicts: Moving Beyond the Global Versus Regional Dichotomy.

Journal of Conflict Resolution 52:295-325.

Hellmann, Gunther. Ed. 2000. “Brother, Can You Spare a Paradigm? (Or: Was Anybody Ever

A Realist?) Letter to the Editors.” International Security 25 (1): 169–174.

Hoffmann, Stanley. 1977. “An American Social Science: International Relations.” Daedalus

106 (1): 41–60.

Hundley, Lindsay, Benjamin Kenzer, and Susan Peterson. (2014) What Pivot? International

Relations Scholarship and the Study of East Asia. International Studies Perspectives

Forthcomming.

31

Ish-Shalom, Piki. (2008) The Rhetorical Capital of Theories: The Democratic Peace and the

Road to the Roadmap. International Political Science Review 29:281-301.

Jensen, Mads Dagnis, and Peter Marcus Kristensen. (2013) The Elephant in the Room:

Mapping the Latent Communication Pattern in European Union Studies. Journal of

European Public Policy 20:1-20.

Jonsson, Christer. (1986) Interorganization Theory and International Organization.

International Studies Quarterly 30:39-57.

Katzenstein, Peter J., Robert O. Keohane, and Stephen D. Krasner. (1998) International

Organization and the Study of World Politics. International Organization 52:645-85.

Keohane, Robert O. (1975) International Organization and the Crisis of Interdependence.

International Organization 29:357-65.

Keohane, Robert O., and Joseph S. Nye. (1977) Power and Interdependence : World Politics

in Transition. Boston: Little, Brown.

Keohane, Robert O. 2001. “Governance in a Partially Globalized World. Presidential

Address, American Political Science Association, 2000.” American Political Science

Review 95 (1): 1–13.

Kihl, Young W. (1971) Functional Performance and Member State Behavior in an

International Organization: Test and Evaluation. Journal of Politics 33:346.

Kim, Hyung Min, and David L. Rousseau. (2005) The Classical Liberals Were Half Right (or

Half Wrong): New Tests of The "Liberal Peace", 1960-88. Journal of Peace Research

42:523-43.

Kinsella, David, and Bruce Russett. (2002) Conflict Emergence and Escalation in Interactive

International Dyads. The Journal of Politics 64:1045-68.

Kleinberg, Jon. (2002) Bursty and Hierarchical Structure in Streams. In Proc. 8th ACM

SIGKDD Intl. Conf. on Knowledge Discovery and Data Mining.

Kratochwil, Friedrich, and John Gerard Ruggie. (1986) International Organization: A State of

the Art on an Art of the State. International Organization 40:753-75.

Kristensen, Peter M. (2012) Dividing Discipline: Structures of Communication in

International Relations. International Studies Review 14:32-50.

Kristensen, Peter Marcus. (2013) Revisiting the “American Social Science”—Mapping the

Geography of International Relations. International Studies Perspectives Online First.

Kuhn, Thomas S. (1962) The Structure of Scientific Revolutions. Chicago: University of

Chicago Press.

Kumar, Ravi, Jasmine Novak, Prabhakar Raghavan, and Andrew Tomkins. (2005) On the

Bursty Evolution of Blogspace. World Wide Web 8:159-78.

32

Lakatos, Imre. (1978) The Methodology of Scientific Research Programmes. Cambridge; New

York: Cambridge University Press.

Larsen, Peder Olesen, and Markus von Ins. (2010) The Rate of Growth in Scientific

Publication and the Decline in Coverage Provided by Science Citation Index.

Scientometrics 84:575-603.

Lehtonen, Markku. (2009) Oecd Organisational Discourse, Peer Reviews and Sustainable

Development: An Ecological-Institutionalist Perspective. Ecological Economics

69:389-97.

Lektzian, David, and Mark Souva. (2009) A Comparative Theory Test of Democratic Peace

Arguments, 1946-2000. Journal of Peace Research 46:17-37.

Maliniak, Daniel, Amy Oakes, Susan Peterson, and Michael J. Tierney. (2011) International

Relations in the Us Academy. International Studies Quarterly 55:437-64.

Maliniak, Daniel, and Ryan Powers. (2013) The Gender Citation Gap in International

Relations. International Organization 67:889-922.

Maliniak, Daniel, Ryan Powers, and Barbara F. Walter. (2013) The Gender Citation Gap in

International Relations. International Organization:Online First.

Maoz, Zeev, and Bruce Russett. (1993) Normative and Structural Causes of Democratic

Peace, 1946-1986. American Political Science Review 87:624-38.

McCormick, James M. (1982) Alternate Approaches to Evaluating International

Organizations: Some Research Directions. Polity 14:531-47.

McKeown, Timothy J. (2009) How Us Decision-Makers Assessed Their Control of

Multilateral Organizations, 1957-1982. The Review of International Organizations

4:269-91.

Merton, Robert K. (1973) The Normative Structure of Science. In The Sociology of Science:

Theoretical and Empirical Investigations, edited by Norman W. Storer and Robert K.

Merton. Chicago: The University of Chicago Press.

Mitchell, Sara McLaughlin, Samantha Lange, and Holly Brus. (2013) Gendered Citation

Patterns in International Relations Journals. International Studies Perspectives

14:485-92.

Molloy, Se¡N. (2003) Realism: A Problematic Paradigm. Security Dialogue 34:71-85.

Ness, Gayl D., and Steven R. Brechin. (1988) Bridging the Gap: International Organizations

as Organizations. International Organization 42:245-73.

Nossal, Kim Richard. 2001. “Tales That Textbooks Tell: Ethnocentricity and Diversity in

American Introductions to International Relations,” in Robert M.A. Crawford and Darryl

S.L. Jarvis, eds., International Relations—Still an American Social Science? Toward

Diversity in International Thought. Albany: State University of New York Press: 167-86.

Oneal, John, and Bruce Russett. (1999) The Kantian Peace. World Politics 52:1-37.

33

Ord, Terry J., Emilia P. Martins, Sidharth Thakur, Ketan K. Mane, and Katy Börner. (2005)

Trends in Animal Behaviour Research (1968-2002): Ethoinformatics and the Mining

of Library Databases. Animal Behaviour 69:1399-413.

Parks, Bradley C., and Alena Stern. (2014) In-and-Outers and Moonlighters: An Evaluation of

the Impact of Policy-Making Exposure on Ir Scholarship. International Studies

Perspectives 15:73-93.

Persson, Olle. (2006) Bibexcel: A Tool-Box Programme for Bibliometric Analysis.

http://www.umu.se/inforsk/Bibexcel/.

Peterson, Susan, Michael J. Tierney, and Daniel Maliniak. (2005) Teaching and Research

Practices, Views on the Discipline, and Policy Attitudes of International Relations

Faculty at Us Colleges and Universities.

Popper, Karl R. (1959) The Logic of Scientific Discovery. New York: Basic Books.

Platig, E. Raymond. 1967. International Relations Research. Problems of Evaluation and

Advancement. Santa Barbara, Cal.: Clio Press.

Pritchard, Alan. (1969) Statistical Bibliography or Bibliometrics. Journal of documentation

25:348-49.

Rathbun, Brian. (2012) Politics and Paradigm Preferences: The Implicit Ideology of

International Relations Scholars1. International Studies Quarterly 56:607-22.

Reid, Edna F., and Hsinchun Chen. (2007) Mapping the Contemporary Terrorism Research

Domain. International Journal of Human-Computer Studies 65:42-56.

Reynaud, Julien, and Julien Vauday. (2009) Geopolitics and International Organizations: An

Empirical Study on Imf Facilities. Journal of Development Economics 89:139-62.

Rochester, J. Martin. (1986) The Rise and Fall of International Organization as a Field of

Study. International Organization 40:777-813.

Rouse, Joseph. 1996. Engaging Science. How to Understand Its Practices Philosophically.

Ithaca: Cornell University Press.

Rouse, Joseph. 2003. “Kuhn’s Philosophy of Scientific Practice”, in T. Nickles (ed.), Thomas

Kuhn, Cambridge, Cambridge University Press, pp. 101-121

Ruggie, John Gerard. (1975) International Responses to Technology: Concepts and Trends.

International Organization 29:557-83.

Russett, Bruce M. Oneal John R. (2001) Triangulating Peace: Democracy, Interdependence,

and International Organizations. New York: Norton.

Russett, Bruce, John R. Oneal, and David R. Davis. (1998) The Third Leg of the Kantian

Tripod for Peace: International Organizations and Militarized Disputes, 1950-85.

International Organization 52:441-67.

34

Schmidt, Brian C. 1994. “The Historiography of Academic International Relations.” Review

of International Studies 20: 349–367.

Schmidt, Brian C. 1998. “The Political Discourse of Anarchy. A Disciplinary History of

International Relations.” International Studies Quarterly 42: 433–459.

Schmidt, Brian C. 2002. “Anarchy, World Politics and the Birth of a Discipline: American

International Relations, Pluralist Theory and the Myth of Interwar Idealism.”

International Relations 16 (1): 9–31.

Sci2 Team. (2009) Science of Science (Sci

2) Tool. Indiana University and SciTech Strategies.

Shiffrin, Richard M., and Katy Börner. (2004) Mapping Knowledge Domains. Proceedings of

the National Academy of Sciences of the United States of America 101:5183-85.

Shih, Chih-yu. (1989) A Cognitive Approach to International Organization: Perspective and

Application. Behavioral Science 34:176.

Sillanpää, Antti, and Tommi Koivula. (2010) Mapping Conflict Research: A Bibliometric

Study of Contemporary Scientific Discourses. International Studies Perspectives

11:148-71.

Simmel, Georg. 1957 [1904] “Fashion.” The American Journal of Sociology 62 (6): 541–558.

Small, Henry G. (1978) Cited Documents as Concept Symbols. Social studies of science

8:327-40.

Tang, Kwok-Cho, Don Nutbeam, Carmen Aldinger, Lawrence St Leger, Donald Bundy, Anna

Maria Hoffmann, Ekua Yankah, Doug McCall, Goof Buijs, and Said Arnaout. (2009)

Schools for Health, Education and Development: A Call for Action. Health Promotion

International 24:68-77.

Tickner, Arlene B., and Ole Waever. (2009) International Relations Scholarship around the

World. Routledge.

Van de Graaf, Thijs, and Dries Lesage. (2009) The International Energy Agency after 35

Years: Reform Needs and Institutional Adaptability. The Review of International

Organizations 4:293-317.

Wæver, Ole. (1998) The Sociology of a Not So International Discipline: American and

European Developments in International Relations. International Organization

52:687-727.

Weaver, Catherine. (2008) Hypocrisy Trap the World Bank and the Poverty of Reform.

Princeton: Princeton University Press.

Wæver, Ole, and Arlene B. Tickner. 2009. Global Scholarship in International Relations:

Worlding Beyond the West. Routledge.

Zammito, John H. 2004. A Nice Dearrangenent of Epistemes. Post-Positivism in the Study of

Science from Quine to Latour. Chicago and London: The University of Chicago Press.