Available via license: CC BY 4.0
Content may be subject to copyright.
Network Science (2023), 1–30
doi:10.1017/nws.2022.39
RESEARCH ARTICLE
Searching for coherence in a fragmented field: Temporal
and keywords network analysis in political science
Dmitry G. Zaytsev1∗, Valentina V. Kuskova2, Gregory S. Khvatsky3and Anna A. Sokol3
1University of Notre Dame at Tantur, Jerusalem, Israel, 2Lucy Family Institute for Data & Society, University of Notre Dame,
1220 Waterway Blvd, #H248, Indianapolis, IN 46202, USA, and 3Lucy Family Institute for Data & Society, Department of
Computer Science and Engineering, University of Notre Dame, 384E Nieuwland Science Hall, Notre Dame, IN 46556, USA
∗Corresponding author. Email: zaytsevdi2@gmail.com
Guest Editor (Special Issue on Scientific Networks): Noshir Contractor
Abstract
In this paper, we answer the multiple calls for systematic analysis of paradigms and subdisciplines in
political science—the search for coherence within a fragmented field. We collected a large dataset of over
seven hundred thousand writings in political science from Web of Science since 1946. We found at least
two waves of political science development, from behaviorism to new institutionalism. Political science
appeared to be more fragmented than literature suggests—instead of ten subdisciplines, we found 66
islands. However, despite fragmentation, there is also a tendency for integration in contemporary polit-
ical science, as revealed by co-existence of several paradigms and coherent and interconnected topics of
the “canon of political science,” as revealed by the core-periphery structure of topic networks. This was
the first large-scale investigation of the entire political science field, possibly due to newly developed meth-
ods of bibliometric network analysis: temporal bibliometric analysis and island methods of clustering.
Methodological contribution of this work to network science is evaluation of islands method of network
clustering against a hierarchical cluster analysis for its ability to remove misleading information, allowing
for a more meaningful clustering of large weighted networks.
Keywords: bibliometrics; political science
1. Introduction
Historians of political science traced the field’s history back to ancient Greek philosophers
(Farr, 1988), with a complicated kaleidoscope of diverse political thoughts, theories, schools, and
approaches to study politics. Political science now carries wide and practical meaning in mul-
tiple areas of knowledge and skills. It became firmly associated with research-centered academic
discipline of “social sciences” sometimes by the mid-twentieth century, achieving the more profes-
sionally defined and narrow meaning (Adcock 2014).Historians are still debating the exact origins
of political science as a separate academic discipline (Adcock 2014: 216). Since Plato and Aristotle,
the discipline has undergone extensive transformations.
As a result, the field itself and the different disciplines within it were often described as lacking
integration and coherence (e.g., Goodin, 2009; Goodin & Klingemann, 1996; Heyse et al., 2003;
Horak, 2010), with frequent calls for more systematic and comprehensive depiction (e.g., Hydén
et al., 2002; Kouba et al., 2015). Goodin & Klingemann (1996) point out that there are “lamenta-
tions and jeremiads about the fragmentation of political science” (Goodin & Klingemann, 1996:
99) and that there is a question about “how all the subspecialties sit together to form a coherent
C
The Author(s), 2023. Published by Cambridge University Press. This is an Open Access article, distributed under the terms of the
Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and
reproduction, provided the original article is properly cited.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
2 D. G. Zaytsev et al.
larger whole” (Goodin & Klingemann, 1996: 7). Given that previous research thought of fragmen-
tation as a problem to be addressed, it appears that there is a need for a systematic evaluation of
the composition of political science as a field.
Multiple previous attempts at introspection in the field of political science (e.g., Ricci, 1984;
Parenti, 1983;Grant,2005) have revealed sharp internal divisions within (Garand, 2005). These
divisions were attributed to naturally occurring subfields (e.g., Almond, 1988;Smith,2002); sepa-
rate intellectual camps (Almond, 1988) or islands or communities (Shepsle, 1990); organizational
ties in the form of APSA memberships (Grant, 2005); methodological approaches (Garand &
Giles, 2003); and many other scientific disputes that make the field of political science, as many
scholars agree, rather fragmented (e.g., Underhill 2000; Udris 2019).
Yet, there is a lot to be gained from understanding of developments in the field. Academic
reasons for doing so, from research relevance (e.g., Stoker, 2015,Vecchi,2017) to epistemological
and methodological considerations (e.g., Wolin, 2009; Benoit & Laver, 2012), have been amply
discussed in the literature. In social sciences, however, the importance of coherent vision of the
field is amplified by a vital position that social sciences occupy not only in academia but also
in society and policy processes (Ceaser, 1985). Political scientists, economists, and sociologists
often take the roles of experts and consultants in applied research projects for governments and
corporations (e.g., Petracca, 1989;Lake,1989). Therefore, the field’s fragmentation, which could
lead to contradictive theories and explanations, may result in lower efficiency of proposed policies.
Understanding the fragmentation, on the other hand, can help unify the practical approaches used
by applied science.
The accumulation of knowledge about the social world and the efficiency of policymaking are
dependent on the understanding of longstanding development of social science disciplines. As
Fortunato et al. (2018) point out, research in the “science of science” domain could help iden-
tify “fundamental mechanisms responsible for scientific discovery” (Fortunato et al., 2018:1).
Therefore, it is crucial not only for academics and historians of science but also for everybody
to understand the past, present, and future of social sciences, in general, and political science in
particular.
Garand (2005) has noted that perceptions of the field’s fragmentation, though noted by many,
have not been subjected to systematic analysis. Partly, it was because very little systematic data
were available, and Garand called previous attempts at examining the field “largely experiential,
impressionistic, or anecdotal” (Garand, 2005: 982). The literature mostly splits into two parts, each
with its own limitations. One consists of comprehensive attempts to investigate the structure of
the whole field but is based mostly on qualitative research methods. A good example of research
output for this literature is handbooks (Goodin, 2009; Goodin & Klingemann, 1996). The second
part consists of research papers that employ modern quantitative methodology, but applies it to
limited subsets of political science, often just to one scientific journal from the field (e.g., Chi,
2012; Adams et al., 2016).
Garand’s (2005) study was one of the first major comprehensive and systematic attempts at
describing patterns in the fragmented discipline, focused on the patterns of evaluation of political
science journals by scientists. Around the same time, bibliometric analysis, first introduced in
the 1960s (Garfield, 1979), started to gain prominence and popularity as a tool for examining
political science. Many have recognized the value of citation analysis as it provided the networks of
scientific contacts between researchers (Lievrouw et al., 1987). Since then, many studies attempted
to evaluate the composition of political science using bibliometrics (e.g., Bjurström, 2011;Dale&
Goldfinch, 2005; Colman et al., 1995;Chi,2012;Pehl,2012). Again, the most notable limitation
of these studies is the specificity of the context: analysis was done by subfield, country, region,
etc.—not of the entire field of political science. Perhaps, it was because specific methodologies
for analysis of the entire field were not yet fully developed, or that computational technologies
were not sophisticated enough for such an undertaking, given the size of the field. Recently, an
opportunity to examine an entire field presented itself with methodological advances in networks,
applicable to bibliometric networks (e.g., Batagelj et al., 2019;Batagelj,2020).
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 3
As a result, the call for a comprehensive examination of the field of political science can now
be attempted to be answered by using bibliometric network approach. There are many questions
to answer and many ways in which such research can be conducted—all of them, clearly, beyond
the scope of one study. However, a good starting point appears to be the evaluation of the field
from the standpoint of main scientific schools. Such an analysis can provide an overview of the
structure of political science and its evolution over time, understanding how the scientific schools
are formed and why they change.
The most recognized division of the field in political science is the partition by time periods.
Almond (1996) defined at least three periods of political science development in the 20th cen-
tury: “Chicago school” in 1920–1940; “behavioral political science” after the World War II; and
“rational choice/methodological individualist approach.” However, historians of political science
still argue about the number and length of such periods, and even what to call them: scien-
tific traditions, schools, paradigms, scientific research programs, or approaches (Dryzek, 1986;
Ball, 1976; Rodman, 1980; Beardsley, 1974;Haas,1986). In this paper, we follow Adcock et al.
(2009) periodization, with three distinct periods. The first is the “old institutionalism” from late
nineteenth century until into the interwar years. The second—behaviorism from the 1950s until
the 1970s. The third and final—the “new institutionalism,” which became a new paradigm for a
range of “cutting-edge research agendas” into the 1980s (Adcock et al., 2009: 259). Throughout
the paper, we will refer to these periods as “paradigms” (Kuhn, 1970) or scientific research
programs (Lakatos, 1978)astheydescribethechangeofthedominant theoretic and method-
ological approach, not a total abolition of a paradigm. Competing approaches continue to co-exist
alongside the dominant.
Therefore, within this periodic division, research paradigms did not remain homogeneous.
Postbehavioral stage is characterized by the crisis of behaviorism in the beginning, with a rise
of diverse paradigms in political science (Farr & Seidelman, 1993). Before the new institution-
alism, there was a period of greater interest in the study of states and international regimes in
1975–1985—a “neostatism” (Adcock et al., 2009). The mid-1980s through the early 1990s were
characterized by the expansion of new institutionalism with its plurality of research programs.
Depending on the school, researchers recognize three or four major research programs within
new institutionalism: rational choice theory, sociological institutionalism, historical institution-
alism, and constructivism (also called discursive institutionalism, e.g., Wahlström & Sundberg,
2018; Bell, 2011) (Schmidt, 2010).
The new approach to institutions within organizational sociology, and more importantly, the
rational choice theory, which was an alternative to behaviorism since the 1980s, was introduced
by March and Olsen (1983). Historical institutionalism and constructivism emerged in the early
1990s as an alternative to rational choice (Adcock et al., 2009; Koelble, 1995,Hall&Taylor,1996).
Sociological institutionalism fully enters the field of political science in mid-1990s (Adcock et al.,
2009). By the term “constructivism” historians of political science also mean an alternative to new
institutionalism (Adcock et al., 2009), related to “interpretive turn” in political science (Blakely,
2013;Rhodes,2017).
The first objective of our study is to evaluate the presence, number, and change of paradigms
or research programs in political science. Currently, there is some skepticism in the field related
to this idea, the mere possibility to identify “a paradigm” within political science (e.g., Dryzek,
1986; Ball, 1976; Rodman, 1980; Beardsley, 1974;Haas,1986). Some, especially older, studies argue
that political science is in the pre-paradigm stage (Baum et al., 1976), so paradigms do not yet
exist. Others state that political science is multiparadigmatic (e.g., Della Porta & Keating, 2008),
and paradigms within the field cannot be clearly separated from each other as they compete and
interact with each other. However, in line with scholars who argue for the existence of paradigms,
we test for their presence in this exploratory study. So, our first research question is focused on
whether we can empirically identify paradigms in political science, quantify them, and evaluate
their change. The time-series nature of the dataset allows to answer these questions.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
4 D. G. Zaytsev et al.
Conventionally, there are several paradigms in political science: behaviorism (peaked in 1920–
1960), rational choice theory (1980s), historical institutionalism and constructivism (1990s), and
sociological institutionalism (mid-1990s) (Adcock et al., 2009). We also can evaluate the expan-
sion of studies in the research programs of Marxism (Almond, 1996; Farr & Seidelman, 1993),
neostatism, and computational political science (Lazer et al., 2009; Weber et al., 2013, Voinea,
2016; Wong et al., 2016). The latter is the emerging research program in political science, which
returned the focus back to quantitative methods. It was the advent of new approaches such as data
mining, machine learning, and other advanced methods of computational social sciences, which
allowed for testing hypotheses not possible previously.
The second objective of this study is to identify the subdisciplines or topics within the field.
Doing so may be useful for understanding the dynamics within the broad field. Goodin &
Klingemann (1996) proposed that the field of political science consists of ten subdisciplines:
political theory, political institutions, law and politics, political behavior, contextual political
analysis, comparative politics, international relations, political economy, public policy, and polit-
ical methodology. While our study is mostly exploratory, it is certainly possible to see whether
this proposition can withstand the scrutiny of an empirical exploration. Therefore, our second
research question here is concerned with whether we can empirically identify ten subdisciplines
listed above in political science, and if not—how many of them can be identified. To align the
terminology of “subdisciplines” with the terminology of the bibliometric community, we will use
words “clusters,” “islands,” and “topics” interchangeably.
To reach the study’s objectives, we use bibliometric network analysis with a dataset of articles
from the Web of Science (WoS), published from 1946. Using writings-to-words and keywords
co-occurrence networks’ analysis, we test the structure of the field computationally and measure
the levels of its fragmentation. To achieve a better understanding of the paradigms’ dynamics and
topic structure in political science, we analyzed words in different contexts: in time settings and
with co-occurrence with other words.
Next, we present our data and methodology in more detail, focusing on data collection, and
creating the temporal writings-to-word networks and keywords co-occurrence network. Then,
we provide the descriptive analysis of our networks and narrow the focus to temporal network
analysis. Next, we present the analysis of keywords co-occurrence, discovering the topics within
political science with link islands methodology (LIM). Finally, we perform a robustness check of
the LIM with hierarchical cluster analysis (HCA). We conclude our paper with discussion and
implications for the understanding of the field and further study.
2. Data and methodology
2.1. Data collection
In this paper, we follow the methodology of bibliometric analysis presented in detail by Maltseva
&Batagelj(2020). We extract our initial dataset from the Web of Science (WoS) Core Collection
database. WoS is known and trusted for its high indexing quality, and in some cases, better cov-
erage of historical works (Harzing, 2013) than Google Scholar (Prins et al., 2016). Specifically in
the field of political science, Bjurström (2011) also used it for bibliometric analysis. He noted that
while journal articles allow analysis of large research fields and their dynamics, monographs (not
all of which are indexed in WoS) often rely on these articles to make overarching conclusions. In
other words, according to Bjurström (2011), analyzing articles provides the same benefit as ana-
lyzing monograph references. For bibliometric analysis, especially extracting field dynamics over
time, and given its higher quality and trust that academic community places in it, WoS is a much
more defendable and reproducible choice than other bibliometric databases.
WoS stores writing metadata information in fields with two-letter names and allows its users to
run search queries using these fields. WoS fields relevant to this work are TI (title of the writing),
ID (Keywords Plus, keywords assigned to writings by the WoS platform), and DE (keywords
assigned to writings by their authors). In addition, WoS allows to use other pseudo-fields in its
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 5
search queries. For example, using the field TS (topic) allows to simultaneously search the title,
abstract, Keywords Plus, and author keywords of the writings for a specified scientific discipline.
For example, using the WoS query “TS =(Political Science)” we return publications from 1946 to
20191and the writings they cite within political science discipline. We retrieved 769,464 records,
from which we removed news, reports, etc. As a result, the initial dataset contained 736,006
records of writings (articles and books2).
Next, we used WoS2Pajek (Batagelj, 2017) to transform the initial dataset into the keyword
network WK (representing which writings used which words). The words were derived from the
ID, DE, and TI fields of the WoS records, by process of lemmatization and stopwords removal.
As a result, we produced network with sets of the following sizes: 3,824,982 writings, 111,258
words. Resulting network contains 676,603 writings with full descriptions and 3,148,379 cited-
only writings (listed as cited by papers with full description). We received less writings with full
description than records in the initial dataset because WoS2Pajek removes duplicates.
Next, we removed multiple links, which could happen as a by-product of keyword phrase entry
(e.g., “soft power” and “hard power” key phrases would connect the word “power” to the arti-
cle twice). Also, because the cited-only writings contain only partial information, we constructed
reduced networks containing only writings with complete descriptions. The final sizes of these
reduced networks (WKr) were as follows: 676,603 writings and 111,258 words. This reduction is
a well-known technique for bibliographic studies (Abouzid et al., 2021), as it allows to lessen the
computational complexity of further analysis.
To prepare for the analysis of changes of popularity of different words over time, we removed
writings that did not have information on the year they were published. We have also removed
articles from 2019, not only because we collected data until July 2019 but also because there is a
delay in indexing of articles in WoS, and not all articles from 2019 appear in the database even in
2020, as was suggested by other studies (Abouzid et al., 2021; AlGhamdi et al., 2021)toimprove
the quality of the data. After this step, networks (WKry) consisted of 671,699 writings and 111,258
words. As a final step, we removed words that were one character in size, contained numbers, or
punctuation. We also removed anonymous writings (writings for which author names were not
specified in WoS) and writings for which keywords and titles were not specified in WoS from the
network WKry (Franceschinietal.,2016). These preparation steps are used in existing literature,
as they allow to both remove writings with dubious contribution to the field and improve the
interpretability of the results of the analysis (Lei & Liu, 2019; Abouzid et al., 2021). The resulting
network (WKryx) consisted of 651,358 writings and 106,373 words.
For the analysis, we used two bimodal writings-to-words networks. The first was WK described
above. The second, WK2, was constructed only from keywords (ID and DE fields in WoS), without
using titles (TI field in WoS).
Specific reason for doing so was that in 1990, WoS indexing policy has changed, and it started
adding keywords (ID and DE fields, Figure 1). Using keywords only allows for a more accurate
analysis than using titles also, as approach to generating titles for articles has changed in the last
few decades. In the past, the titles were required to accurately reflect the content of the article
(Zeller & Farmer, 1999). Now authors have much more freedom, so titles are much more subject
to individual choice and author’s creativity (Jixian & Jun, 2015). Since the second goal of our
study was to identify subdisciplines, we did not want to clutter them unnecessarily with extra
words from catchy titles.
Of course, there is an issue of keyword designation and their accuracy for describing the study.
Journal practices differ with respect to keywords: some allow choices from a designated list;
some allow authors to generate their own. There are other criticisms of keyword analysis, amply
described in the literature, from linguistics (e.g., Egbert & Baker, 2016) to the discipline-specific
studies, including political science (e.g., Pérez-Paredes, 2019). However, there are studies that
rebuff this argument by stating that “keywords reflect the changes in knowledge and interests
in the discipline” (Yang, 2019),which is precisely what we were set out to find out. Therefore,
bibliometric analysis based on keywords only appears appropriate for our study purpose.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
6 D. G. Zaytsev et al.
Figure 1. Number of articles with keywords per year.
We then generated WK2rnetwork, from which we removed all writings published before 1990.
There were two reasons for doing so. First, it is a coincidence that WoS started using keywords
(ID and DE fields) only since 1990. We had a choice of the starting year for our study and could
have made a cut at a later year. However, modern political science is largely shaped by geopolitical
events of late 1980s-early 1990s. As Lowi (1992) pointed, out, “Regime changes throughout the
world since 1989 ought to give us a clearer perspective on some new sciences of politics” (Lowi,
1992: 6). No matter the approach to studying political science, the field today is a reflection of
research of most pressing political issues. Therefore, the subdisciplines within political science
reflect these changes as well, so the year 1990 appears appropriate for the start of our exploration.
In addition to the adjustments described above for WKryx and applied to WK2, we also
removed misleading keywords (multiple words glued together, words containing numbers, abbre-
viations, names of chemical compounds, or non-English language words, for example “iii” or “aa”)
by hand. This preprocessing step is widely used by bibliometric studies focused on keyword anal-
ysis (Rajasekharan et al., 2021; Ye et al., 2021; Yang et al., 2021). Resulting network (WK2ryx)
consisted of 115,378 writings and 25,355 keywords.
2.2. Temporal writings-to-words networks
Using the WKryx network as our starting point, we created a collection of 73 networks WKins,
each containing a full set of words as one mode and the writings only from a given year (from
the period 1946–2018) as the second mode. To construct temporal networks, we applied tempo-
ral quantities approach (Batagelj & Maltseva, 2020; Batagelj & Praprotnik, 2016)totheWKryx
network, using Python libraries Nets and TQ (Batagelj, 2014). We used temporal writings-to-
words networks (WKins) for further temporal analysis of paradigms change in political science.
In Table 1S of the Supplementary Materials, we provide a detailed step-by-step network trans-
formation from the WK network to WKins network with explanation of the steps and the
purpose.
2.3. Keywords co-occurrence network
To build keywords co-occurrence network, we used WK2ryx, since, as noted above, keywords
(without words from titles) more precisely determine the subdisciplines or topics in political
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 7
science. Keywords co-occurrence network nKKis a result of multiplication of normalized
reduced transposed writings-to-keywords network (nWK
2ryx)by itself.
nKK=nWK
2ryxT∗nWK
2ryx
Such approach to creating the keyword co-occurrence network can have some drawbacks. One
of them is overrating the contribution of writings with very large number of keywords. To deal
with such cases, we used the fractional approach (Batagelj & Cerinšek, 2013; Gauffriau et al., 2007),
which normalizes each article’s input to the resulting network to be equal to 1. It is called “penal-
ization” and was implemented to reduce the influence of writings that included a large number
of keywords. With too many keywords, the importance score (indegree) of an individual word
becomes lower. This means that scores of words that are typically used together with a large num-
ber of other words (e.g., the word “review” or “handbook,” which are used in writings describing
broad areas and topics) will be affected by this procedure the most.
In a regular network, the outdegree is equal to the number of keywords in the writing, and the
indegree is equal to the number of writings in which the same keywords are used. The normaliza-
tion creates the network where the weight of each arc is divided by the sum of weights of all arcs
having the same initial node as this arc (the outdegree of a node):
nWK2ryx w,k=WK2ryx w,k
max1, outdegree(w)
where wis a writing and kis a keyword. Here, the contribution of each writing is set to be equal
to 1. Note that this procedure assumes that all keywords are equally important.
The nWK2ryx network was reduced to the nWK
2ryx network by removing keywords with
weighted indegree less than 1 from the nWK2ryx. Indegree of less than 1 meant that none of the
keywords carried substantive meaning. Retention of such articles carried no substantive weight
but increased the computational complexity for construction of the keyword co-occurrence net-
work. Keywords with weighted indegree substantially less than 1 were either not related to political
science (e.g., “mycobacterium,” “veggie”) or contained combination of letters that were, perhaps,
some uncommon abbreviations (“pla,” “chp”). Removing such keywords did not alter the mean-
ing but simplified the analysis. Keyword co-occurrence network reduction is a common practice
in bibliographic studies aimed at reducing the computational complexity of the analysis, although
typically, studies use arbitrary threshold, such as keeping only top 10% of the studies’ entities (Kim
et al., 2020).
The minimum and maximum indegrees for the nWK2ryx networks were 0.012 (for words
“debreu” and “evolutionarily”) and 2,078.093 (for the word “policy”). The number of words that
were retained in the network nWK
2ryx was 5,564, the number of words that fell below the thresh-
old of 1 and were removed was 19,791. For further analysis, we normalized the WKryx network
using the same normalization technique. The minimum and maximum indegrees for the nWKryx
networks were 0.004 (for a group of German words and the word “chlorinate”) and 6,945.407
(for the word “politics”). We used keywords co-occurrence network (nKK), consisting of 5,564
keywords, for topics’ discovery in political science using island methods (Batagelj et al., 2014).
Summary statistics for all these networks is presented in Table 1; please also see Table 2S in
Supplementary materials for step-by-step description of the network transformation process.
3. Seventy-year evolution of paradigms in political science: Temporal
writings-to-words networks’ analysis
3.1. Descriptive network analysis
The distribution of the number of analyzed words in complete network, WKryx, containing words
derived from author and platform keywords (ID and DE fields) and writing titles (TI field in WoS),
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
8 D. G. Zaytsev et al.
Tab le 1. Descriptive statistics of the analyzed networks
WK2ryx nWK2ryx nWK
2ryx nKKWKryx nWKryx
Minimum indegree 1.000 0.012 1.000 0.050 1.000 0.004
........................................................................................................................................................................................................................................
Maximum indegree 1,6261.000 2,078.093 2,078.093 2,916.545 5,4302.000 6,945.407
........................................................................................................................................................................................................................................
Minimum outdegree 1.000 1.000 0.091 0.050 1.000 1.000
........................................................................................................................................................................................................................................
Maximum outdegree 85.000 1.000 1.000 2,916.545 250.000 1.000
........................................................................................................................................................................................................................................
Number of writings 115,378 115,378 115,288 N/A 651,358 651,358
........................................................................................................................................................................................................................................
Number of words 25,355 25,355 5,564 5,564 106,373 106,373
Figure 2. Logarithmic plots with distributions of the number of words per writings.
is presented in Figure 2. As is apparent from the figure, most studies contain 2–6 describing words.
After about 10 descriptive words, the number of articles containing them goes sharply down.
Figure 3shows the distribution of the number of words used in all writings in the WKryx net-
work. As is apparent from the graph, a large number of words are mentioned only once (50,550),
twice (14,036), or three times (7,401). The usage of these words is episodic. There are also words
which are used extensively, constructing the core concepts of the field.
The most frequent words are presented in Table 3S of Supplementary Materials. Not surpris-
ingly, the words “political” and “politics” are the most frequently used words (54,302 and 50,946,
respectively). The result that these words are used most frequently is certainly expected, since we
studied political science writings, but it also demonstrates the validity of the resulting datasets.
As a check for the fractional approach to network normalization, which was used in the same
manner for all of the networks, we compared top words in the normalized network (nWKryx)with
top words of the non-normalized network (WKryx). The network was normalized to remedy the
effect of writings with a large number of words on the analysis, similar to a TF-IDF approach to
text analysis. We did not find many differences. One the one hand, words “peace,” “life,” “reply,”
“future,” “French,” “Russia” (bold in Table 3S) which are present in the top-60 of the normalized
network, are absent from the top of non-normalized network. It means that these were the words
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 9
Figure 3. Logarithmic plots with distributions of the unique combinations of all words used in all writings.
that were “penalized” the least—in other words, they are used for designation of more narrow
topics. Given the meaning of these words, it is not surprising and appears to be accurate.
On the other hand, words “model,” “governance,” “civil,” “democratic,” “institution,” and “com-
munity” (underlined in Table 3S) are present in the top-60 of the non-normalized network but
are absent in the normalized. These words are penalized the most, which means they are used for
designation of wider topics. Such logic also appears to be accurate, given the meaning of these
words.
3.2. Temporal network analysis
To test our first assumption about paradigm changes in political science, we use temporal network
analysis. This method allows us to trace the longitudinal development of behaviorism, new insti-
tutionalism, and constructivism paradigms. Moreover, we can also attempt to find dynamics in
development of smaller traditions of Marxism, neostatism, and computational political science.
For temporal network analysis, we used writings-to-words networks created for each year from
1946 to 2018 (WKins). Figure 4presents temporal distributions of the number of all words used
in political science writings by year. These are not unique words—they are all words used by all
writings during the year.
The overall trend, of course, demonstrates a significant growth of the number of words used—
from 3,415 in 1946 to a peak of over 233,818 in 2016 (the drop in the numbers in 2017–2018
is due to delay in indexing). This growing trend is easily explained by the growth of political
science discipline and the subdisciplines within it, as well as the growth in the absolute number
of journals. In the decade prior to mid-1990s alone, over a hundred new journals were launched
(Almond, 1996). This appears to be a part of overall process that of development of the field of
political science. This includes, but is not limited to, more articles being published, more journals
becoming active, more topics coming into the view of political scientists. What is interesting,
however, is that this growth is not even—it occurs in spurts. Such spurts occurred in mid-1960s
and mid-1970s, continuing into the 2000s, coinciding with known changes in scientific debates.
The first spurt coincides with growth in behaviorism after the WWII. Then, the end of 1960s—
beginning of 1970s is a well-known period of debates around behaviorism and questions about
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
10 D. G. Zaytsev et al.
Figure 4. WKins: distribution based on words and writings.
its capacity to explain political phenomena (Farr & Seidelman, 1993). Some scholars even spoke
about the paradigm shift and the end of “normal science” in political studies (Ball, 1976;Beardsley,
1974). The substantial increase (almost double) from mid-1970s until 1980s corresponds to the
rise of rational choice theory tradition. The spurt around the 1980s–1990s is contemporaneous
with the two new versions of new institutionalism—sociological and historical. Finally, around
mid-2000s, we see the increase in constructivism, the expansion of advanced computational social
sciences methods, and the possible establishment of a multi-paradigm approach to political sci-
ence, first spoken about Ball (1987)andHaas(1986). Even if this timeline is not exact, the evidence
of a pattern consisting of periods of relative “calm” interleaved with periods of explosive growth
in the discipline development is clear.
A more detailed analysis of words for each specific paradigm allows to narrow down these
periods more precisely. We calculated the proportion of the number of appearances of each word
to the most frequent word for each year based on the WKins network (Batagelj & Praprotnik,
2016). This proportion normalizes the importance of a certain word over time from 0 to 100%.
The proportions for the words associated with each paradigm of political science over time are
presented in the Figures 5and 6.
For the first paradigm, behaviorism, the most indicative are the words “behavior,” “behavioral,”
and “psychology” (Adcock et al., 2009). If the use of the word “behavior” is stable over time,
then the usage distribution of the words “behavioral” and “psychology” is more interesting (see
the first line of Figure 5). From 1946 until the mid-1950s, there is a peak of the use of the word
“psychology,” closely related to behaviorism (Adcock et al., 2009). Another less pronounced and
shorter peak of the word is apparent after the mid-1960s, when the literature was full of debates
about the crisis of behaviorism. After that time, with exception of some barely noticeable peaks,
the popularity of this word steadily declines. We see a similar pattern in the dynamics of the word
“behavioral” with a more pronounced increase by mid-1960s and almost complete drop of interest
to the term by 1980s, when the dominance of the alternative traditions of new institutionalism is
firmly established.
The youngest approach within new institutional tradition is rational choice theory started to
grow in 1970s. We test its dynamics using such word markers as “rational” and “choice,” which
are obvious; “game” (as in “game theory”) is a synonym for this research program (Figure 5,
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 11
Figure 5. Distribution of proportion of keywords indicating paradigms (part 1).
second line) (Goodin & Klingemann, 1996;Hall&Taylor,1996). Indeed, if we look at the first two
words alone, the advent of this tradition occurs at the end of 1970s. However, the word “game”
reaches its first prolonged peak of popularity in 1970. In 1980s, at the time of new institutionalism
establishment, the growth of all three words reaches a near-plateau, with a sharp increase again in
the 1990s. The end of 1990s—beginning of 2000s—sees the highest popularity of this paradigm.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
12 D. G. Zaytsev et al.
Figure 6. Distribution of proportion of keywords indicating paradigms (part 2).
Perhaps, this is because popular in the U.S. approaches became popular elsewhere with a bit of a
delay. Nonetheless, from the mid-2000s, the popularity of rational choice theory is declining.
We started testing new institutionalism as a whole paradigm with words “institutionalism,”
“institutional,” “institution.” All three words show very similar dynamics of growth in usage from
1990s (Figure 5, line 5), indicating the advent and the growing dominance of this latest new
paradigm. We further distinguish two other approaches within it: sociological institutionalism
and historical institutionalism—based on their differences in the definition of “institution” and
approach to institutional change (Schmidt, 2010).
We tested sociological institutionalism development using such words as “sociological,”
“cultural,” “norm,” “rule,” and “frame,” historical institutionalism—using words “historical,”
“history,” “culture,” “path,” “dependency,” “critical,” and “juncture.” We observed interesting
dynamics for words “norm,” “rule,” and “frame” for sociological institutionalism, and “culture,”
“path,” and “juncture” for historical (Schmidt, 2010).
Apparently, historical institutionalism started to develop earlier than sociological institution-
alism. The word “culture” starts its long-term growth approximately in mid-1970s, reaches its first
peak in second part of 1980s, then continues its growth in the 1990s, and reaches the maximum
around the 2000s. It somewhat loses its popularity in recent years, but remains significant in the
literature, which means that this paradigm continues its dominance in political science. The word
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 13
“path” follows approximately the same trajectory. The word “juncture” is a much less pronounced,
with an advent in the 1990s and continued use in the 2000s (Figure 5, line 3) (Schmidt, 2010).
Judging by the dynamics of the words “norm,” “frame,” and “rule,” the stable long-term growth
of the sociological institutionalism paradigm started in the mid-1990s and continues this path
(Figure 5, line 4). An interesting dynamic is observed with the word “frame”: it came to use much
later than the other two words, but then sharply increased in popularity and remains popular until
now (Schmidt, 2010).
For constructivism or “discursive institutionalism,” we used the distinguishing framework
proposed by Schmidt (2010). Following her rationale, we tested the following markers of
constructivism: “constructivism,” “construct,” “ideas,” “ideational,” “discourses,” “discursive,”
“communication,” “communicative,” “memory,” “narrative,” “epistemic,” “deliberation,” “delib-
erative,” “advocacy,” “interpretation,” “interpretive.” They all share similar patterns—the start of
growth in the 1990s and proliferation in the 2000s (Figure 5, lines 6,7; and Figure 6, lines 1,2).
We also examined the dynamics of other, smaller, research programs in political science, not
all of which are accepted by historians of science as paradigms, but nonetheless, remain important
anchors in political science. It appears that the peak of interest to “Marx” and “Marxism,” which
some consider as a separate approach in political science (Bowman, 2007), was reached in the
1980s (Figure 6, line 3).
Neostatism (Adcock et al., 2009; Nettl, 1968) as a separate paradigm is not so clearly defined as
other traditions, because the word “state” is very popular in political science in general. However,
the other words that we tested—“statism,” “stateness,” “statehood,” “etatism”—are not. It seems
that the end of 1970s was marked by a rapid growth of the usage of the word “state,” with a
decade of dominance of this word from the mid-1980s to the mid-1990s (Figure 6, line 3). This
is evidence for the argument that neostatism was an alternative to new institutionalism after the
crisis of behaviorism. Another alternative to behaviorism, which is not so popular now, was struc-
tural functional approach (Varshney, 1978;Groth,1970; Mitchell, 1958). The usage of the base
words for this approach—“structure” and “function”—grew in the 1960s, peaked in the 1970s,
and declined in the 1980–1990s (Figure 6, line 4).
Words indicative of computational political science (Lazer et al., 2009; Weber et al.,
2013, Voinea, 2016; Wong et al., 2016)—“computational,” “centrality,” “mixed,” “machine,”
“Bayesian”—are not yet popular, relatively speaking, in political science yet (Figure 6, lines 6,7).
One exception may be the word “network.” However, this word has a much wider meaning: it
relates not only to “network analysis” but also to other areas of political science, not necessarily
studied with statistical network-analytic methods. The word “network” can also indicate forma-
tion of networks in public policy—policy networks; it could be related to a special management
type—network governance. Nonetheless, judging by the use of other related words in the aggre-
gate, we can conclude that computational social sciences are gaining popularity. While the trend is
optimistic, this paradigm is still in its infancy in political science. An implicit indication is the fact
that the literature is using the phrase “computational social science” as opposed to “computational
political science” (Lazer et al., 2009; Weber et al., 2013, Voinea, 2016; Wong et al., 2016).
4. Keywords co-occurrence network analysis: Topic discovery in political science
We next turn to the second objective our study: a finer examination of the paradigms for the
discovery of more narrow topics or subdisciplines. For this purpose, we use the nWK
2ryx and
nKKnetworks.
4.1. Descriptive network analysis
Figure 7(a) shows the distribution of the number of keywords per writing in the nWK2ryx net-
work. Figure 7(b) presents the same information for the nWK
2ryx network. Overall, most articles
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
14 D. G. Zaytsev et al.
Figure 7. (a) Logarithmic plots with distributions of the number of keywords per writing (nWK2ryx). (b) Logarithmic plots
with distributions of the number of keywords per writing (nWK
2ryx).
use from one to ten keywords. After ten, the number of articles using a higher number of keywords
starts to decline sharply.
Figure 8(a) and (b) show distribution of unique keywords used by writings. Figure 8(a) shows
distribution of all keywords used, from the nWK2ryx network. This plot demonstrates that there
are keyword combinations very large in size, used by a small number of writings. For example,
there is a combination of 10,000 keywords used only by one writing.
Figure 8(b) shows distribution of keywords from nWK’2ryx network, created after removal
of misleading keywords, and thus describing the field more accurately. The major difference
with Figure 8(a) is observed for the range of 10–100 keywords. There are many combinations of
10–100 keywords in the full network that are used by close to 100 writings. However, some of these
combinations, apparently, do not carry a lot of meaning: after misleading keywords are removed,
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 15
Figure 8. (a) Logarithmic plots with distributions of the number of unique combinations of all keywords used in all writ-
ings (nWK2ryx). (b) Logarithmic plots with distributions of unique combinations of keywords after removal of misleading
keywords, used in all writings (nWK
2ryx).
there are six keyword combinations left within the 10–100 range that are used by less than ten
writings. Otherwise, the distribution of keywords in nWK
2ryx network peaks at 184 individual
keyword combinations that are mentioned in 13 writings. Another 160 combinations are used by
12 writings each; 178—by 11 writings each, etc.
Creating a network without misleading keywords allowed us to focus on keywords that are
indicative of their respective topics or are used extensively, constructing the core concepts of
the field. Later analysis will show that with reduced network we have managed to identify both
keywords that specify the core concepts of political science as well as keywords that represent
individual subfields, this speaks for the validity of the data used in the analysis.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
16 D. G. Zaytsev et al.
An exploratory analysis showed that in the keywords co-occurrence network (nKK), the most
frequent keywords which are used together with other keywords are “political,” “state,” “poli-
tics,” “policy,” “social,” “public,” “democracy” and “power” (Table 4S, supplementary materials).
Comparative analysis of two sets of keywords: those used in the largest number of articles and
those used most frequently with other keywords, allows to separate keywords into generic and
specific. While the construction of the networks used for this comparison may have impacted the
scores themselves, it is the relative importance of different keywords that we consider for this
analysis. For this purpose, Table 4S contains top-60 keywords from two networks: the keywords
co-occurrence network (nKK) and writings-to-keywords network (nWK
2ryx).
The keywords that are underlined in Table 4S—“human,” “justice,” “community,” “American,”
“decision,” “strategy,” “culture” and “new”—are present in a relatively small number of articles,
but are connected with a large number of other keywords. These are the generic keywords, mean-
ing that by these keywords alone, it is not possible to understand the topic of the article. To place
the meaning correctly, other, more specific, keywords are required. These keywords can delineate
wider topics and be used in different contexts and areas. For example, the keyword “commu-
nity” may indicate a methodological topic—“community detection” in networks, but may also
be related to coalition formation in public policy—“policy communities,” or to a special policy
instrument—“community building.”
Keywords that are bolded in Table 4S—“representation,” “competition,” “integration,” “vot-
ing,” “trade,” “electoral,” “EU” and “growth”—are present in a relatively large number of articles,
but they are connected to a smaller number of keywords. Most likely, this means that they are
more specific. The topic of the article is understood from these keywords without additional expla-
nations. These keywords can delineate finer topics and be used in narrow contexts and areas. For
example, the keyword “EU” indicates a specific region of the world and the socioeconomic union
of countries within it. As another example, the keyword “growth” also has a rather narrow mean-
ing of “economic growth” in political science. Overall, again, this is a confirmation of the data
validity.
4.2. Cluster network analysis: Link islands method
We next use cluster network analysis to test our second assumption that political science con-
sists of ten subdisciplines or topics: political theory, political institutions, law and politics, political
behavior, contextual political analysis, comparative politics, international relations, political econ-
omy, public policy, and political methodology. We do so by using the link islands method (Nooy
et al., 2018; Batagelj et al., 2014), selected because it is now widely used in scientific networks and
has been shown to accurately detect scientific schools in many different disciplines (Doreian et al.,
2020). “Island” is defined (Batagelj et al., 2014) as a maximal connected subgraph, with values of
edges between the nodes of the subgraph higher than the values of edges between the subgraph
and the rest of the network. An important property of islands is that they “identify locally impor-
tant subnetworks at different levels. Therefore, they can be used to detect emerging groups of
phenomena” (Batagelj et al., 2014: 55). It allows to isolate not only large clusters but also smaller
islands—in this case, “coherent well-connected” (Batagelj et al., 2014: 54) subdisciplines in polit-
ical science. This is exactly what we need to grasp the meaning in the fragmented and separated
subfields in this discipline.
To determine the number of cuts that would allow us to delineate ten islands—the theoretical
number of subdisciplines in political science—we analyzed the plots of changes in the number
of islands against the maximum cut sizes [Figure 9(a) and (b)]. We selected the minimum cut
based on the distribution of the number of keywords per writing. As stated before, the majority of
writings contain between 1 and 10 keywords; therefore, for the minimum cut, we took 2, 3, 5, and
10 keywords. (We did not use one keyword because the minimum number of keywords required
to specify a topic is two.)
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 17
Figure 9. (a) Plots of changes in the number of islands against the maximum cut sizes (complete network nKK). (b) Plots
of changes in the number of islands against the maximum cut sizes (Main island (5,000 nodes) of the first cut of the network
nKK).
In order to achieve the theoretically specified number of islands (ten), we need a minimum cut
of 2 and maximum cuts in the range of 5,000–5,066 keywords; or minimum cut of 3 and maxi-
mum cuts in the range of 3,035–3,284 keywords. Cuts 5 and 10 do not provide us an opportunity
to receive 10 islands (Figure 9a). In bibliometric studies, thresholds of 2 or 3 keywords are typ-
ically used when analyzing keyword co-occurrence networks are often used (Schodl et al., 2017;
Maltseva & Batagelj, 2020; Žnidaršiˇ
c et al., 2021).
Therefore, the sizes of our subnetworks ranged from 2 to 5,000 nodes, allowing us to cluster our
keywords co-occurrence network (nKK) into 10 islands. The main island contains 5,000 nodes;
other smaller islands contain two keywords each (Table 5S, Supplementary Materials).
These smaller islands represent narrow, borderline topics in political science, perhaps cross-
ing into other disciplines. They can be described as follows: politics of Latin American countries
(“partido, nacional”), including the presidency of Juan Manuel Santos in Colombia (“manuel,
juan”); knowledge economy and developing society-academia partnerships (“triple, helix”); art
and political science (“avant, garde”); lobbying and interest representation in governing bodies
(“revolve, door”); human rights issues in liberal economies (“faire, laissez”); nuclear weapons pro-
liferation issues (“uranium, enrichment”); studies of human emotions and their connections to
policy (“gesture, imitation”); anarchy studies and public choice (“maghribi, trader”).
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
18 D. G. Zaytsev et al.
The main island of about 5,000 keywords, which we have obtained from the first cut, is too
large to be analyzed in its entirety and is likely to contain the core topics and subdisciplines of
political science. Therefore, using the island methodology (Batagelj et al., 2014) we partition it
again. To achieve theoretically relevant ten islands, we can partition this main island of the first
partition using 2 as minimum cut and any number in a range from 4,796 to 4,808 as a maximum
cut. Alternatively, we can choose 3 as a minimum cut and a number in the range from 3,035 to
3,284 for a maximum cut. Cuts 5 and 10 again do not provide us an opportunity to obtain ten
islands (Figure 9b). In order to achieve the theoretically specified number of islands (ten), we
need a minimum cut of 3 and maximum cuts in the range of 3,035–3,284 keywords.
The sizes of the subnetworks we obtained ranged from 3 to 3,100 nodes, allowing us to clus-
ter the main island of the keywords co-occurrence network (nKK) into 10 islands. The main
island contains 3,100 nodes; other islands contain 3–4 keywords each (Table 6S, supplementary
materials).
These smaller islands represent narrow, almost borderline topics in political science, perhaps
crossing into other disciplines. They can be described as follows: genetically modified organisms
and their consequences for policy (“genetically, modify, organism”); drones and political con-
sequences of their use (“aerial, unmanned, vehicle”); political philosophy of Gilles Deleuze and
its application to political research (“deleuze, gille, guattari”); neuroscience and politics (“cor-
tex, neuroscience, prefrontal”); conflicts and fate of African states (e.g., Côte d’Ivoire) (“cote,
divoire, ivoire”); pandemics (HIV/AIDS, SARS, and their consequences) (“immunodeficiency,
october, sar, virus”); conflict in Nagorny Karabach and surrounding regions (“armenia, azerbai-
jan, karabakh, o Nagorno”); introductions to special issues and volumes and analysis of editors’
and writers’ political identities, including written productions (“editor, introduction, writer”);
information cascade theory and its applications to the study of new media, new protests, policy
diffusion, etc. (“cascade, fad, informational”).
Despite the wide variety of topics, two partitions with main islands (5,000 and 3,100 nodes)
have similar nine smaller islands. They contain smaller islands with relatively new topics, small
number of writings’ citations, and small number of writings in general. These topics are new for
political science, but quite likely, very promising for future expansive research. This is because
they borderline other disciplines, such as sociology, neuroscience, medicine, engineering, philos-
ophy, psychology, philology, area studies, economics etc.—meaning, they are interdisciplinary or
cross-disciplinary. As calls for interdisciplinary research are becoming more urgent in social sci-
ences, these new, but already developing fields have the potential for quickly answering such calls
(Goodin, 2009).
The main island of 3,100 keywords we have obtained from the second cut is also large and,
similar to the first cut, likely to contain a mixture of topics. Therefore, we repeat the partitioning
procedure again (Batagelj et al., 2014), reapplying the same island methodology to the main island
of 3,100 nodes with limits from three to 300 keywords in an island. The resulting partition consists
of 18 islands (Table 6, supplementary materials). The first 17 islands are small, consisting of 3–5
keywords, with another big island of 300 nodes. They can be grouped into four categories associ-
ated with four subdisciplines: public policy, political theory, international relations, and political
methodology (Table 7S).
The largest group are the topics related to public policy issues. It includes citizen safety
and protection from fatal accidents fires, natural disasters, (“crowd, fatality, traffic”); motivated
reasoning, skepticism, and belief systems of politicians and citizens (“motivate, reasoning, skep-
ticism”); politics of equal opportunity and tolerance (“fetal, pain, sympathy”); collective political
trauma and trauma policy (“disorder, posttraumatic, stress”); sport policy (“football, professional,
sport, team”); indigenous people, youth policy, and what unites them—minority rights, transi-
tional justice, and other related theories (“indigenous, people, young”); analysis of financial and
other markets in periods of growth and decline (“automobile, boom, bust”); codes, standards, and
regulations (“code, conduct, operational”).
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 19
Figure 10. Main island of nKKnetwork: “the canon of political science”.
The second group are the studies related to theory testing. Not all the theories are popular in
political science, and the studies are often dedicated to their applicability to the field. These are the
theories of collectivism and individualism (“collectivism, individualism, methodological”); equity
theories and the works of Amartya Sen (“amartya, entitlement, sen”); Tullock’s rent-seeking game
and its applications (“efficient, rent, seek”); and the Condorcet jury theorem and its applications
to politics (“condorcet, jury, theorem”).
The third group are topics loosely related to international relations. It includes China and
its trade policies (“belt, road, silk”); colonialism and its consequences (“colonialism, Japanese,
settler”); problems related to natural resources and the tragedy of commons—resource curse, con-
flicts, and wars (“curse, natural, resource”). The final group is the group of methodological studies.
It includes the black box of causal inference in politics (“causal, inference, mechanism”) and
problems with survey methodology—nonresponse bias, etc. (“exchange, rate, response, survey,
telephone”).
The main island of the third cut or partitions represented in Figure 10. We argue that this island
represents the core topics in political science or the “the canon that practicing political scientists
need to master in order to have mastered the discipline” (Goodin, 2009: 29).
To have the opportunity to examine this island with more accuracy, following the same pro-
cedure of Batagelj et al. (2014), we again partitioned the islands with cuts of 2,5. This allowed us
to isolate 31 islands. The topics of islands corresponded most closely to the current state of affairs
in the field of political science, because they represent the nine of the coveted ten subdisciplines.
They are shown in Table 8S of Supplementary Materials.
“Public policy,” again, seems to be the topic leader, taking into consideration the number of
topics that were algorithmically extracted as the most connected by keywords related to this dis-
cipline. Public policy is comprised of seven islands out of 31. Each island is a separate topic in
the public policy subfield: civic and higher education (with keywords “civic, education, higher”);
health care policy (“care, health”); socioeconomic and political inequality (“income, inequality”);
issues of decision-making on individual, collective and state levels—the topic more related to pub-
lic administration at the intersection of decision sciences (“decision, make”); local governance
problems (“government, local”); global environmental policy and challenges including climate
change (“change, climate”); issues of gender policy and discrimination (“gender, woman”).
“International relations,” “political economy,” and “political behavior” are the next two sub-
disciplines that are related with several islands, four in each case. International relations are
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
20 D. G. Zaytsev et al.
represented by its namesake main island (“international, law, relation”). There is also a large
island representing an important topic for this subdiscipline—writings related to the analysis
of conflict resolution, wars, topics of peace and security (“civil, conflict, peace, society, war”).
Another island is related to the issues of transitional and international justice and law (“court,
justice, supreme, transitional”). The last island can also be attributed to the “law and politics”
subdiscipline, because it contains writings related to national courts and justice. The issues
of international self-determination, annexes, and secessions, are also important on agenda of
contemporary international relations (“determination, self”).
“Political economy” is represented by an island about financial crises (“crisis, financial”); eco-
nomic growth and development (“development, economic, growth”); the study of the markets,
including labor markets and the issues related to labor migration (“labor, market”). In addition,
there is an island related to the cyclicality of economic development and related political cycles
(“business, cycle”).
“Political behavior” is represented by an island related to collective action (“action, collec-
tive”) and social movements (“movement, social”). It also contains a more traditional subtopics of
electoral behavior (“turnout, voter”), and public opinion studies (“opinion, public”).
The already mentioned subdiscipline of “law and politics” is also represented by the writings
about relationships between national and international laws and the issues of priority of one over
the other (“court, justice, supreme, transitional”). It is also related to no less important topic of
human rights (“human, rights”). Both are separate islands.
The subdiscipline of “political institutions” is traditionally related to the study of institutes
of democracy, related institutes of elections and multiple parties, and also economic institutes,
often serving as the foundation of the subdiscipline. The keywords that represent this island
“democracy, economy, election, party, political.”
Two subdisciplines—“contextual political analysis” and “comparative politics”—do not have
their separate islands. More likely, they share topics with each other, studying them from different
angles. One such topic is regional studies, related to writings about specific world regions, such
as Eastern Europe (“eastern, Europe”); South Africa (“Africa, south”); Middle East and Southeast
Asia (“Asia, east, middle, southeast”); European Union (“European, integration, union”), Latin
America (“America, Latin”). They also share a common topic of country development (“country,
develop”). Each subdiscipline claims its own methodological approach to the study of politi-
cal phenomena. However, the literature also contains writings about the difficulties related to
establishing the identity of each discipline.
Two other subdisciplines—“political methodology” and “political theory”—are not well-
represented by extracted islands. However, some islands extracted during the third cut could be
attributed to each of them (Table 7S). Among the islands of the fourth cut, “political methodology”
also contains a topic of longitudinal analysis (“series, time”), which is quite popular and relevant
for the study of political processes. From the third cut, we also know that this subdiscipline is rep-
resented by writings related to survey methodology, methods of causal inference modeling, and
causal mechanism detection.
For “political theory,” we did not find any islands in the fourth partition, but they were present
in the third one. However, almost all islands of the fourth cut contained purely theoretical papers.
It is interesting to note that previous bibliometric studies found this subdiscipline to be less inte-
grated in political science (Goodin, 2009). It is also possible that we did not find a separate island
for “political theory” because theory scholars, for the most part, prefer to publish books. Also,
it is possible that there is a difference between theoretical topics and theory-oriented scholars.
The scholars are less integrated with other scientists, while the topics, on the contrary, are more
integrated and related to other studies that test them empirically.
Finally, there are three islands that are somewhat problematic—they cannot be related to any
one subdiscipline. The first island is the island that contains studies on “new” topics, but also New
Zealand (“new, Zealand”). The second is related to politics and news (“media, news”). The third
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 21
is related to the problems of foreign policy, state policy, and politics (“foreign, policy, politics,
state, unit”). If the first island is more logical but less relevant, the other two are quite interest-
ing. “Political communications” appears to be a legitimate subdiscipline, though Goodin (2009)
does not identify it with the main ten—only as a smaller subdiscipline, more prominent histori-
cally. It is recognized, however, by some other scholars of political science and media studies (e.g.,
Graber & Smith, 2005). The final island is a very interesting collection of writings on the inter-
section of at least two subdisciplines—“international relations” and “public policy”—and perhaps,
even contains a third—“political institutions.”
Altogether, we have extracted 66 islands, which is substantially more than ten theoretically
hypothesized subdisciplines. Relying on a team of subject matter experts, led by a professor with
over twenty years of work in the field of political science, we qualitatively interpreted them for
their correspondence with the ten theoretical subdisciplines. However, we were not always able to
do so: not all islands could be clearly assigned to one and only one subdiscipline. It appears that
ten theoretical subdisciplines are not distinctive; they are, more likely, formed by groups of islands
or clusters—not the exclusive topics. Therefore, a much more important conclusion of this study
is that the field of political science can be described by a clear core-periphery structure, which we
have observed repeatedly over several cuts across topics.
4.3. Cluster network analysis robustness check with HCA
To test the robustness of our conclusions, we implement the HCA method as an alternative to the
LIM. Specifically, we wanted to see if we would again obtain the core-periphery structure of the
topics, and whether we would find more than ten subdisciplines. HCA is a method that has been
widely used in bibliometric studies to analyze keyword co-occurrence networks along with the
link islands method (Sun et al., 2014; Bhuyan et al., 2021; Lee et al., 2010; Yang & Ning, 2021).In
implementation of the HCA approach, we used the same restriction on the number of clusters that
we have received from cuts of LIM method and followed the procedure of further clustering of the
largest island obtained by each partition (a total of four times). Table 9S of the Supplementary
Materials describes the partitions obtained by LIM and HCA.
Following this approach, we first converted the normalized keywords co-occurrence network
nKKto a dissimilarity matrix using the cosine distance measure. The use of the cosine measure
allowed us to group the keywords based on the similarity of contexts in which they were used
(Maldonado-Guerra & Emms, 2011; Zhiqiang et al., 2009). The use of this measure was suggested
by previous bibliometric studies (Bhuyan et al., 2021) focused on keyword co-occurrence analysis.
Then, we have applied the HCA procedure using complete linkage as the agglomeration
method. This method searches for tight, compact clusters, which allows the discovered topics to
be more interpretable (Sileshi & Gamback, 2009).
To evaluate the results received by applying the LIM, we partitioned the network
nKKinto 10 clusters using the HCA. In the first partition, the size of the largest
cluster was 5,521 keywords, which is comparable with the largest island of 5,000 key-
words received in the first LIM partition. In the HCA partition, the sizes of other clus-
ters were much smaller (from 1 to 11 keywords). This finding is also comparable to the
first LIM partition, where the sizes of all smaller islands were equal to two (Figure 11,
Table 9S in Supplementary Materials). Therefore, we can conclude that in first partitions by two
different methods we obtained the core-periphery network structure.
In line with the LIM, we further split the largest HCA cluster of the first partition (5,521 key-
words) into ten clusters. Resulting HCA partition had a large cluster of 5,395 keywords, and several
clusters consisted of 43 keywords or less (see Table 9S). In the second partition of LIM, the largest
partition was 3,100 nodes; the sizes of other islands ranged from three to four keywords. Again,
this structure of partitions we received from both methods (one giant cluster and several much
smaller ones) points to the presence of a core-periphery structure within the network.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
22 D. G. Zaytsev et al.
Figure 11. Comparison of partitions obtained with link island method and hierarchical clustering.
Again, following the LIM, we further split the largest cluster of the second HCA partition (5,395
keywords). For the third partition, we aimed to adjust the number of clusters in such a way that
that the size of the smallest cluster was at least three. In the third partition of LIM, the network was
separated into 18 islands, with 300-word largest island, and smaller islands ranging from three to
five keywords. In the third HCA partition, the network was also separated into 18 clusters, with
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 23
the largest cluster of size 5,054 and smaller clusters with four to 41 keywords (Figure 11,Table
9S). Therefore, we can again conclude that the network nKKhas a core-periphery structure.
Also, applying both methods—LIM and HCA—at this stage, when we do not limit the number of
output clusters with ten, we received more than ten clusters. The HCA, as an alternative method,
confirms our finding of existence of more than ten subdisciplines or topics in political science.
Following the final step of LIM, we once again extracted the largest cluster of 5,054 nodes from
the third HCA partition. With the fourth partition of LIM, we generated islands from two to five
keywords in size. We reproduce this with HCA, creating a partition with clusters of size two to five,
by adjusting the number of clusters. With the LIM, we have obtained 31 islands; with HCA, we
have found 75 clusters, with their sizes ranging from 1 to 548 keywords. Even though the number
of HCA clusters was much larger than with the LIM, we again supported our main theses that
political science is represented by much greater number of subdisciplines than ten.
However, these findings also point at an important difference between the link island method
and hierarchical clustering. The LIM not only allows for discovery of small, dense, cohesive sub-
groups within the network, but also automatically identifies keywords that do not belong to any
cohesive topic. Such keywords are removed from further analysis by being assigned to the zeroth
island which is not interpreted, leaving remaining subgroups to be more meaningful. In contrast,
the HCA method, which is requiring that all keywords in the analysis need to be assigned to a
cluster, gives rise to large, hard to interpret collections of keywords (e.g., larger than 500 keywords
in partition 4). In HCA, keywords that do not form cohesive topics are still added to the generated
clusters, thus sharply decreasing their interpretability.
The first three rounds of clustering with both methods show a very similar pattern of a single
giant cluster or island combined with several much smaller clusters or islands. This pattern sug-
gests that the normalized keyword co-occurrence network has a core-periphery structure, with
both methods peeling outer layers of the core of the network during each round of clustering.
Both HCA and LIM confirm the main findings of the previous section that there are more than
ten subdisciplines of political science and that the network nKKhas a core-periphery structure.
5. Discussion
This study had two major objectives. The first objective was to evaluate the presence, number,
and change of paradigms or research programs in political science; the second—to identify the
subdisciplines or topics within the field. Using a Web of Science (WoS) Core Collection dataset of
all writings in political science field from 1946, we used the LIM (Batagelj et al., 2014)andHCA
(Sun et al., 2014; Bhuyan et al., 2021; Lee et al., 2010; Yang & Ning, 2021) methods for extracting
related papers. To the best of our knowledge, this study was the first comprehensive bibliometric
study in the field of political science, not limited by subfield, region, or some other narrowing
criterion.
The study was mostly exploratory, though we relied on two assumptions to anchor our
exploration: the most recognized division of the field by time periods and presence of ten subdisci-
plines. Our results indicate that despite some skepticism surrounding the possibility of extracting
paradigms in political science, it is possible, in principle, to do so. We have found the evidence
of existing in the literature, but contested, periodization of two paradigms. The first is behavioral
studies from at least 1946 (the beginning of the study; actual advent of this approach happened,
perhaps, much earlier) until the 1970s. The second is the new institutionalism paradigm from
1980s until now. This was the first contribution of our study.
Another contribution of our study was that we have uncovered the dynamics of research pro-
grams inside the new institutionalism paradigm. The rational choice theory, one of the research
programs within this paradigm, appeared in the 1970s, but most likely was experiencing a decline
by the 2000s. Historical institutionalism, which reached its peak in the 1980s, continued to develop
well into the 1990s and 2000s, unlike rational choice. The latest research program within this
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
24 D. G. Zaytsev et al.
paradigm, sociological institutionalism, appeared in the 1990s and is also still on the rise. Overall,
new institutionalism, first associated with rational choice, continues its dominance due to the last
two research programs. Constructivism, which some call the fourth research program inside the
new institutionalism, does exist somewhat separately from the other three, but also started to rise
in the 1990s and continues to develop to present times.
To check yet another widely accepted notion of the field’s development—that the field is
multiparadigmatic—we examined the presence of smaller paradigms, such as Marxism, neo-
statism, and computational social sciences. We also observe the presence of new institutionalism
and other paradigms. In other words, multiple paradigms in the field do exist.
We also confirmed the presence of subdisciplines within the field, though their number is not
limited to ten major, as outlined by Goodin (2009). Altogether, there are 66 topics, though quite a
few of them are combined inside a major island or cluster, the “canon of political science.” From
the original ten that we’ve looked for, some subdisciplines, indeed, are more dominant, some are
more clearly delineated while others are fuzzy. There are also some that are, probably, on the rise as
the generators of new “hot” topics and methods and could form new subdisciplines in the future.
Overall, however, it is reasonable to conclude that “subdiscipline” is a construct: subdisciplines
do not exist in isolation, their borders are not clearly defined, and they often intersect with each
other. We have also identified the “borderline,” or interdisciplinary subfields, which cross over
many other fields of science. This was the third contribution of our study.
Another important contribution that logically flows from this idea is that despite the pro-
nounced fragmentation, political science is being integrated—but not only by scientists, as Goodin
(2009) suggested. His ideas were centered around the premise that with too much fragmentation
inside the field, we need “integrators”—scientists who could keep track of the field’s develop-
ment and provide the direction for a more coherent approach to research, applicable to many
diverse subdisciplines. Our results show that such “integrators” are not necessary: the field is being
“held together” by paradigms, and across all paradigms—by the “canon of political science.” As
paradigms develop, the dominant paradigm may be replaced by another. Previously dominant
paradigms do not disappear from the agenda, they continue to remain a part of the field, but to a
lesser degree. Paradigms develop over time, but at any given time, there is one dominant paradigm.
This dominant paradigm becomes the integrator of the political science field.
The same applies to the “canon” or the core of political science, formed by ten subdisciplines.
They remain more or less unchanged, but this does not preclude formation of new subdisciplines
on the periphery, or transfer of the new subdisciplines to the “core.” However, the core-periphery
structure of political science does not change and keeps the field from separating into separate
fields. So, in essence, fragmentation exists on the periphery of a solid, integrated foundation,
formed by “canon” of political science. Perhaps, this conclusion warrants further study.
Temporal analysis indicated that political science is a developing, evolving field, where many
studies are dedicated to specific events happening in the society. There are waves and periods
of dominance of certain paradigms or research programs, and future studies could examine, in
detail, what causes a certain paradigm or research program to appear.
There were some interesting results obtained with temporal analysis, which did not immedi-
ately fit inside the study’s main objectives: some interesting words did not form separate islands yet
were obviously present across multiple topics. The first was the word “American,” with very inter-
esting dynamics. It dominated until about mid-1950s, then continued a slow decline to present
times. This means that the political science was mostly American right after the WWII, but slowly
internationalized and expanded to other parts of the world. The word “democracy” also had an
interesting dynamic, with a sharp increase in interest in the 1990s that continues until now. These
two waves form two important topics in political science—“American politics” and “democracy
studies.” However, we did not find such islands in contrast with, for example, European politics.
It appears that words “American” and “democracy” are so popular that they are connected to too
many other words. They are the keywords of the wider context, creating their own discourse in
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 25
political science: everything in the field starts with American and democracy. This is very clear in
Figure 10, where these two words are clearly present in the busiest center of the network.
Our final contribution is to network science methodology: we validated the use of link island
methodology with the HCA, widely used for evaluation of large weighted networks (including
bibliometric). We have replicated the LIM procedure with HCA and obtained the same general
results: the presence of a core-periphery structure within the political science keyword network
(in the first three partitions) and the presence of more than ten topics in the field. Moreover, we
have verified that LIM provides for more interpretable results by removal of keywords that do
not belong to any cohesive topic. Unlike LIM, HCA does not have that ability and forces such key-
words to remain a part of the analysis. Therefore, LIM allows for a more meaningful interpretation
of the obtained islands.
6. Limitations and future work
This study is not without limitations. The first limitation is the use of the WoS database for our
study. Of course, like any database, it contains errors and omissions. It also limits the scope of
the study simply because it does not include all writings in the field of political science. To its
defense, WoS provides important indexing services that are much more reliable than those of
Google Scholar, for example, which is much more inclusive. There are standardized fields that
contain the necessary data, important article metadata, and many other benefits of using WoS
over collecting data manually—a rather daunting task for the project of this size.
Moreover, we have found that multiple studies were conducted that have examined WoS,
Scopus, Google Scholar, and other bibliometric databases with each other. For example, the time-
series study by Harzing & Alakangas (2016) showed that all three databases “provide sufficient
stability of coverage to be used for more detailed cross-disciplinary comparisons.” The study
showed that despite differences, the overall coverage of the disciplines is relatively consistent in
all databases. A field-specific study by Falagas et al. (2008), which also included PubMed in the
analysis, noted some differences in coverage, but most importantly, criticized Google Scholar for
some errors, omissions, and other “inadequacies” (p. 342). Of course, there is an issue of multi-
ple predatory journals that flood the net with “worthless” publications (Sharma & Verma, 2018),
which should not be representative of any science. Despite all the defenses of WoS as a database,
however, we do realize that some of the important writings could be missing from our analysis,
making it, at best, incomplete.
The second limitation of our study is the way in which we considered and treated words and
keywords. First, we treated author-generated keywords equally with WoS-generated keywords. It
is possible that author-generated keywords might have more weight. More important is the fact
that we treat title words the same as keywords (e.g., for longitudinal analysis). As we have stated in
the method section, we created two networks—one without titles—precisely because titles might
not be very representative of the paper topic. However, for examining the development of topics
over time, title words remain and are given equal weight with keywords. Presently, we do not have
a way to evaluate this possible difference in weights, so we leave it for future studies.
The third limitation, also related to our treatment of words, is the way we’ve reduced key
phrases and titles to single words to build networks of co-occurrences. We also removed dupli-
cates connecting words to writings (as would be the case with “soft power” and “hard power”
key phrases that connect the word “power” to the writing). There is a valid concern that doing
so might remove some of the richness that composite keywords provide and might lead to mis-
interpretation of the words that have different meaning based on context. This could also affect
the way that word combinations represent subdisciplines. We realize that this could be one of
the reasons for the “fuzziness” in the island definitions that we have observed. It could also be
the reason why some of the more common words in the field ended up together as opposed to
specific areas where they belong. While we certainly realize this limitation, we think that for this
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
26 D. G. Zaytsev et al.
first exploratory study our approach was sufficient for our goal. Moreover, we used the islands
methodology that isolate the words that are more often connected to each other and remove the
keywords that do not belong to any topic, segregating the fields based on word combinations that
separate them. In other words, LIM allows to obtain the phrases through connected keywords in
co-occurrence networks. Nonetheless, we suggest that future studies extend this methodology to
include composite key phrases to examine the field to a finer extent.
The fourth limitation is the starting year of our study, which was the year 1990. Despite two
compelling reasons for doing so—inclusion of keywords in WoS articles and geopolitical events
after the end of Cold War, which shaped the modern political science, inclusion of early writings
could have altered our results. We do not believe that the main findings of our study—presence
of a core-periphery structure and more than ten subdisciplines in the field—would have changed
by such inclusion. If anything, we expect our results would have been only more confirmed, given
the lower number of topics in the field in earlier years. However, this is a good starting point for
further exploration.
Overall, this study lays an important ground for further research. It was the first study to exam-
ine the entire field of political science, and it showed that such exploration is not only possible,
but also quite meaningful, with new and interesting insights. The most obvious extensions of
this study are further examination of temporal networks, exploration of co-evolution of topics
in academia and in practice and the role that academics play in influencing politics and society
in general. More work should be done on establishing the boundaries of the integrated core—the
“canon of political science”—and the fragmented periphery that forms that diverse subdisciplines.
Specific suggestions for future studies also include projections of what could be the “fields of
the future.”3Temporal network analysis could be extended to anticipate upcoming changes in the
field. Moreover, a finer analysis could be done on a smaller sample of journals to see whether the
same islands emerge there or if there is a different pattern of topics that are indicative of certain
journal editor preferences. Perhaps, some journals are responsible for the peripheral groups, while
others are developing the main core. Even on a finer scale, it would be interesting to see if pub-
lishing articles in certain topical fields would lead to tenure and promotion or if certain subjects
would dominate the agenda in graduate schools, preparing future scholars to work on specific
topics. In other words, the “science of science” (Fortunato et al., 2018) in political science field has
a lot of promise.
Competing interests. None.
Supplementary materials. For supplementary material for this article, please visit http://doi.org/10.1017/nws.2022.39
Notes
1The data were collected up to 19 July 2019.
2Even though the term “writings” is not very conventional in academic research, we are using it to denote the fact that not
just the articles were used in our analysis.
3We thank anonymous reviewers for this and other future research ideas in this paragraph.
References
Abouzid, M., Anna, K. G., & Marta, K. (2021). Trend research of vitamin D receptor: Bibliometric analysis. Health Informatics
Journal,27(4), 146045822110431. doi: 10.1177/14604582211043158.
Adams, W. C., Lind Infeld, D., Wikrent, K. L., & Bintou Cisse, O. (2016). Network bibliometrics of public policy journals.
Policy Studies Journal,44(S1), S133–S151.
Adcock, R. (2014). A disciplinary history of disciplinary histories: The case of political science. In R. E. Backhouse, &
P. Fontaine (Eds.), Ahistoriographyofthemodernsocialsciences(pp. 211–236). Cambridge: Cambridge University Press.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 27
Adcock, R., Bevir, M., & Stimson, S. C. (Eds.). (2009). Modern political science: Anglo-American exchanges since 1880.
Princeton, NJ: Princeton University Press.
AlGhamdi, S., Bret, H., Angel, C. C., Roland, R., Gwendolyn, J., & Tu˘
grul, U. D. (2021). Technology intelligence map: Fast
charging for electric vehicles. In Daim, T. U. (Ed.), Roadmapping future: Technologies, products and services (pp. 399–416).
Cham: Springer International Publishing.
Almond, G. A. (1988). Separate tables: Schools and sects in political science. PS: Political Science & Politics,21(4), 828–842.
Almond, G. A. (1996). Political science: The history of the discipline. In A new handbook of political science (pp. 75–82).
Oxford: Oxford University Press.
Ball, T. (1976). From paradigms to research programs: Toward a post-Kuhnian political science. American Journal of Political
Science,20(1), 151–177.
Ball, T. (1987). Is there progress in political science? In Idioms of inquiry: Critique and renewal in political science (pp. 13–44).
Albany: State University of New York Press.
Batagelj, V. (2014). Nets—Python package for network analysis. Retrieved from https://github.com/bavla/Nets/tree/
master/source
Batagelj, V. (2017). WoS2Pajek. Networks from Web of Science. Version 1.5. Retrieved from http://vladowiki.
fmf.uni-lj.si/doku.php?id=pajek:wos2pajek
Batagelj, V. (2020). On fractional approach to analysis of linked networks. Scientometrics,123(2), 621–633.
Batagelj, V., & Maltseva, D. (2020). Temporal bibliographic networks. Journal of Informetrics,14(1), 101006.
Batagelj, V., & Praprotnik, S. (2016). An algebraic approach to temporal network analysis based on temporal quantities. Social
Network Analysis and Mining,6(1), 1–22.
Batagelj, V., & Cerinšek, M. (2013). On bibliographic networks. Scientometrics,96(3), 845–864.
Batagelj, V., Doreian, P., Ferligoj, A., & Kejzar, N. (2014). Understanding large temporal networks and spatial networks:
Exploration, pattern searching, visualization and network evolution (Vol. 2). New York: John Wiley & Sons.
Batagelj, V., Ferligoj, A., & Doreian, P. (2019). Bibliometric analyses of the network clustering literature. In Advances in
network clustering and blockmodeling (pp. 11–64). New York: Wiley.
Bell, S. (2011). Do we really need a new constructivist institutionalism to explain institutional change? British Journal of
Political Science,41(4), 883–906.
Baum, W. C., Griffiths, G. N., Matthews, R., & Scherruble, D. (1976). American political science before the mirror: What our
journals reveal about the profession. The Journal of Politics,38(4), 895–917.
Beardsley, P. L. (1974). Political science: The case of the missing paradigm. Political Theory,2(1), 46–61.
Benoit, K., & Laver, M. (2012). The dimensionality of political space: Epistemological and methodological considerations.
European Union Politics,13(2), 194–218.
Bhuyan, A., Sanguri, K., & Sharma, H. (2021). Improving the keyword co-occurrence analysis: An integrated semantic sim-
ilarity approach. In IEEE International Conference on Industrial Engineering and Engineering Management (IEEM) (pp.
482–487). Singapore: IEEE.
Bjurström, P. (2011). On the use of bibliometrics for domain analysis—A study of the academic field of political science in
Europe. Uppsala University, Uppsala, Sweden.
Blakely, J. (2013). Returning to the interpretive turn: Charles Taylor and his critics. The Review of Politics,75(3), 383–406.
Bowman, P. (2007). Post-Marxism versus cultural studies: Theory, politics and intervention. Edinburgh: Edinburgh University
Press.
Ceaser, J. (1985). Alexis de Tocqueville on political science, political culture, and the role of the intellectual. American Political
Science Review,79(3), 656–672.
Chi, P. S. (2012). Bibliometric characteristics of political science research in Germany. Proceedings of the American Society for
Information Science and Technology,49(1), 1–6.
Colman, A. M., Dhillon, D., & Coulthard, B. (1995). A bibliometric evaluation of the research performance of British
university politics departments: Publications in leading journals. Scientometrics,32(1), 49–66.
Dale, T., & Goldfinch, S. (2005). Article citation rates and productivity of Australasian political science units 1995-2002.
Australian Journal of Political Science,40(3), 425–434.
Della Porta, D., & Keating, M. (2008). How many approaches in the social sciences? An epistemological introduction. In
Approaches and methodologies in the social sciences. A pluralist perspective. Cambridge: Cambridge University Press.
Doreian, P., Batagelj, V., & Ferligoj, A. (Eds.). (2020). Advances in network clustering and blockmodeling.NewYork:John
Wiley & Sons.
Dryzek, J. S. (1986). The progress of political science. The Journal of Politics,48(2), 301–320.
Egbert, J., & Baker, P. (2016). Research synthesis. In Triangulating methodological approaches in corpus linguistic research (pp.
195–220). London: Routledge.
Falagas, M. E., Pitsouni, E. I., Malietzis, G. A., & Pappas, G. (2008). Comparison of PubMed, Scopus, web of science, and
Google scholar: Strengths and weaknesses. The FASEB Journal,22(2), 338–342.
Farr, J., & Seidelman, R. (1993). Discipline and history: Political science in the United States. Ann Arbor, MI: University of
Michigan Press.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
28 D. G. Zaytsev et al.
Farr, J. (1988). Political science and the enlightenment of enthusiasm. American Political Science Review,82(1), 51–69.
Fortunato, S., Bergstrom, C. T., Börner, K., Evans, J. A., Helbing, D., Milojevi´
c, S., ... Barabási, A. L. (2018). Science of science.
Science,359(6379), eaao0185.
Franceschini, F., Maisano, D., & Mastrogiacomo, L. (2016). Empirical analysis and classification of database errors in Scopus
and Web of Science. Journal of Informetrics,10(4), 933–953.
Garand, J. C. (2005). Integration and fragmentation in political science: Exploring patterns of scholarly communication in a
divided discipline. The Journal of Politics,67(4), 979–1005.
Garand, J. C., & Giles, M. W. (2003). Journals in the discipline: A report on a new survey of American political scientists. PS:
Political Science & Politics,36(2), 293–308.
Garfield, E. (1979). Is citation analysis a legitimate evaluation tool? Scientometrics,1(4), 359–375.
Gauffriau, M., Larsen, P. O., Maye, I., Roulin-Perriard, A., & von Ins, M. (2007). Publication, cooperation and productivity
measures in scientific research. Scientometrics,73(2), 175–214.
Goodin, R. E., & Klingemann, H. D. (Eds.). (1996). A new handbook of political science. Oxford: Oxford University Press on
Demand.
Goodin, R. E. (Ed.). (2009). The Oxford handbook of political science (Vol. 11). Oxford: Oxford University Press.
Graber, D. A., & Smith, J. M. (2005). Political communication faces the 21st century. Journal of Communication,55(3),
479–507.
Grant, J. T. (2005). What divides us? The image and organization of political science. PS: Political Science & Politics,38(3),
379–386.
Groth, A. J. (1970). Structural functionalism and political development: Three problems. Western Political Quarterly,23(3),
485–499.
Haas, M. (1986). Metaphysics of paradigms in political science: Theories of urban unrest. The Review of Politics,48(4),
520–548.
Hall, P. A., & Taylor, R. C. (1996). Political science and the three new institutionalisms. Political Studies,44(5), 936–957.
Harzing, A.-W. (2013). A preliminary test of Google Scholar as a source for citation data: A longitudinal study of Nobel prize
winners. Scientometrics,94(3), 1057–1075.
Harzing, A.-W., & Alakangas, S. (2016). Google Scholar, Scopus and the Web of Science: A longitudinal and cross-disciplinary
comparison. Scientometrics,106(2), 787–804.
Heyse, L., Meyer, M., Mourey, D., & Wijkström, F. (2003). SWG 03: Organizing in and through Civil Society:
Perspectives, Issues, Challenges. EGOS – European Group for Organizational Studies. Retrieved from
https://www.egos.org/SWGs/SWG-03
Horak, M. (2010). Urban megaprojects in comparative perspective: Mapping the research terrain. In APSA 2010 annual
meeting paper.
Hydén, G., Immergut, E., Underdal, A., Ruin, O., & Öhngren, B. (2002). Swedish research in political science. Stockholm:
Vetenskapsrådet (The Swedish Research Council).
Jixian, P., & Jun, C. (2015). Comparing English titles of Chinese and international journal articles in applied linguistics.
Journal of Zhejiang University (Humanities and Social Sciences),1, 21.
Kim, M. C., Seojin, N., Fei, W., & Yongjun, Z. (2020). Mapping scientific landscapes in UMLS research: A scientometric
review. Journal of the American Medical Informatics Association,27(10), 1612–1624. doi: 10.1093/jamia/ocaa107.
Koelble, T. A. (1995). The new institutionalism in political science and sociology. Comparative Politics,27(2), 231–243.
Kouba, K., CíSAˇ
r, O., & Navrátil, J. (2015). The Czech political science: A slow march towards relevance. In Political science
in Europe at the beginning of the 21st century (pp. 63–85). Krakow: Jagiellonian University Press.
Kuhn, T. S. (1970). The structure of scientific revolutions (Vol. 111). Chicago: University of Chicago Press.
Lakatos, I. (1978). In: Worrall, J., & Currie, G. (Eds.), The methodology of scientific research programmes: Philosophical papers
1. Cambridge: Cambridge University Press.
Lake, C. (1989). Political consultants: Opening up a new system of political power. PS: Political Science & Politics,22(1),
26–29.
Lazer, D., Pentland, A., Adamic, L., Aral, S., Barabási, A. L., Brewer, D., ...Jebara, T. (2009). Computational social science.
Science,323(5915), 721–723.
Lee, S.-J., Choi, J.-H., & Kim, H.-W. (2010). Semantic network analysis on the MIS research keywords: APJIS and MIS
quarterly 2005∼2009. Asia Pacific Journal of Information Systems,20(4), 25–51.
Lei, Y., & Liu, Z. (2019). The development of artificial intelligence: A bibliometric analysis, 2007-2016. Journal of Physics:
Conference Series,1168, 022027. doi: 10.1088/1742-6596/1168/2/022027.
Lievrouw, L. A., Rogers, E. M., Lowe, C. U., & Nadel, E. (1987). Triangulation as a research strategy for identifying invisible
colleges among biomedical scientists. Social Networks,9(3), 217–248.
Lowi, T. J. (1992). The state in political science: How we became what we study. American Political Science Review,86(1), 1–7.
Maldonado-Guerra, A., & Emms, M. (2011). Measuring the compositionality of collocations via word co-occurrence vectors:
Shared task system description. In Proceedings of the Workshop on Distributional Semantics and Compositionality (pp.
48–53).
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
Network Science 29
Maltseva, D., & Batagelj, V. (2020). Towards a systematic description of the field using keywords analysis: Main topics in
social networks. Scientometrics,123, 1–26.
March, J. G., & Olsen, J. P. (1983). The new institutionalism: Organizational factors in political life. American Political Science
Review,78(3), 734–749.
Mitchell, W. C. (1958). The polity and society: A structural-functional analysis. Midwest Journal of Political Science,2(4),
403–420.
Nettl, J. P. (1968). The state as a conceptual variable. World politics,20(4), 559–592.
Nooy, W, Mrvar, A., & Batagelj, V. (2018). Exploratory social network analysis with Pajek (Revised and expanded edition for
updated software). Cambridge: Cambridge University Press.
Parenti, M. (1983). The state of the discipline: One interpretation of everyone’s favorite controversy. PS: Political Science &
Politics,16(2), 189–196.
Pehl, M. (2012). The study of politics in Germany: A bibliometric analysis of subfields and methods. European Political
Science,11(1), 54–70.
Pérez-Paredes, P. (2019). Little old UK voting Brexit and some Austrian friends: A corpus-driven analysis of the 2016 UK
right-wing tabloid discourse. In Populist discourse (pp. 152–171). London: Routledge.
Prins, A. A. M., Costas, R., van Leeuwen, T. N., & Wouters, P. F. (2016). Using Google Scholar in research evaluation of
humanities and social science programs: A comparison with Web of Science data. Research Evaluation,25(3), 264–270.
Petracca, M. P. (1989). Political consultants and democratic governance. PS: Political Science & Politics,22(1), 11–14.
Rajasekharan, P. K., Sankaran, K., Ramnarayan, K., & Nandan Prabhu, K. P. (2021). Research in higher education governance:
Past performance and an agenda for the future. Educational Review, 1–24. doi: 10.1080/00131911.2021.1984214.
Rhodes, R. A. (2017). Interpretive political science: Selected essays (Vol. 2). Oxford: Oxford University Press.
Ricci, D. M. (1984). The tragedy of political science: Politics, scholarship, and democracy. New Haven, CT: Yale University Press.
Rodman, J. (1980). Paradigm change in political science: An ecological perspective. American Behavioral Scientist,24(1),
49–78.
Sharma, H., & Verma, S. (2018). Predatory journals: The rise of worthless biomedical science. Journal of Postgraduate
Medicine,64(4), 226.
Schmidt, V. A. (2010). Taking ideas and discourse seriously: Explaining change through discursive institutionalism as the
fourth new institutionalism. European Political Science Review,2(1), 1–25.
Schodl, K., Klein, F., & Winckler, C. (2017). Mapping sustainability in pig farming research using keyword network analysis.
Livestock Science,196, 28–35. doi: 10.1016/j.livsci.2016.12.005.
Shepsle, K. (1990). Untitled Commentary. “The Nature of Contemporary Political Science: A Roundtable Discussion,” ed.
Monroe, Kristen et al.PS: Political Science and Politics,23, 40–42.
Sileshi, M., & Gamback, B. (2009). Evaluating clustering algorithms: Cluster quality and feature election in content-
based image clustering. In 2009 WRI World Congress on Computer Science and Information Engineering (pp. 435–441).
10.1109/CSIE.2009.1002.
Smith, R. M. (2002). Should we make political science more of a science or more about politics? PS: Political Science & Politics,
35(2), 199–201.
Stoker, G. (2015). Challenging three blockages to relevance and political science: The obvious, the avoidable and the thorny.
In The relevance of political science (pp. 19–35). New York: Palgrave Macmillan.
Sun, X., Tang, W., Ye, T., Zhang, Y., Wen, B., & Zhang, L. (2014). Integrated care: A comprehensive bibliometric analysis and
literature review. International Journal of Integrated Care,14, e017.
Udris, L. (2019). Political communication in and about crises. Potentials of a fragmented field. Studies in Communication
Sciences,19(1), 131–152.
Underhill, G. R. (2000). State, market, and global political economy: Genealogy of an (inter-?) discipline. International Affairs,
76(4), 805–824.
Varshney, A. (1978). Structural functional approach to politics: An evaluative over-view and relevance for research in
developing countries. The Indian Journal of Political Science,39(2), 306–318.
Vecchi, G. (2017, June 28–30). Ontologies, evidences, and policy making fostering the relevance of evaluation in the relations
with policy actors. In 3rd International Conference on Public Policy (ICPP3), Singapore. Retrieved from May 4, 2020, from
https://www.ippapublicpolicy.org/file/paper/594d294b248f2.pdf
Voinea, C. F. (2016). Political attitudes: Computational and simulation modelling. New York: John Wiley & Sons.
Wahlström, N., & Sundberg, D. (2018). Discursive institutionalism: Towards a framework for analysing the relation between
policy and curriculum. Journal of Education Policy,33(1), 163–183.
Weber, I., Popescu, A. M., & Pennacchiotti, M. (2013, February). Data-driven political science. In Proceedings of the sixth
ACM international conference on Web search and data mining (pp. 777–778).
Wolin, S. (2009). Paradigms and political theories. In Richter, W. L. (Ed.), Approaches to political thoughts (pp. 125–152).
Lanham, MD: Rowman & Littlefield Publishers.
Wong, F. M. F., Tan, C. W., Sen, S., & Chiang, M. (2016). Quantifying political leaning from tweets, retweets, and retweeters.
IEEE Transactions on Knowledge and Data Engineering,28(8), 2158–2172.
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press
30 D. G. Zaytsev et al.
Yang, C., Cui, H., & Tong, Z. (2021). Bibliometric analysis of AI policy evolution over ten years in China. In 16th International
Conference on Intelligent Systems and Knowledge Engineering (ISKE) (pp. 736–741).
Yang, W. (2019). A diachronic keyword analysis in research article titles and cited article titles in applied linguistics from 1990
to 2016. English Text Construction,12(1), 84–102.
Yang, W., & Ning, H. (2021). Knowledge graph technology application in Chinese SSCI: An example of short videos research.
Journal of Librarianship and Information Science, 096100062110632. doi: 10.1177/09610006211063201.
Ye, S., Jin, H., Zilong, L., & Man, L. (2021). Progress and research trends on Catha Edulis (Vahl) Endl. (Catha Edulis): A
review and bibliometric analysis. Frontiers in Pharmacology,12, 705376. doi: 10.3389/fphar.2021.705376.
Zeller, N., & Farmer, F. M. (1999). “Catchy, clever titles are not acceptable”: Style, APA, and qualitative reporting. International
Journal of Qualitative Studies in Education,12(1), 3–19.
Zhiqiang, L., Werimin, S., & Zhenhua, Y. (2009). Measuring semantic similarity between words using Wikipedia. In 2009
International Conference on Web Information Systems and Mining (pp. 251–255). 10.1109/WISM.2009.59.
Žnidaršiˇ
c, A., Maltseva, D., Brezavšˇ
cek, A., Maletiˇ
c, M., & Baggia, A. (2021). A bibliometric network analysis of green
information technology and green information systems research. Business Systems Research Journal,12(2), 17–45. doi:
10.2478/bsrj-2021-0017.
Cite this article: Zaytsev D. G., Kuskova V. V., Khvatsky G. S. and Sokol A. A. Searching for coherence in a fragmented field:
Temporal and keywords network analysis in political science. Network Science.https://doi.org/10.1017/nws.2022.39
https://doi.org/10.1017/nws.2022.39 Published online by Cambridge University Press