The Universal and the Contextual of Media Systems: Research Design, Epistemology, and the Production of Comparative Knowledge

Abstract

Current discussions about the state of comparative research in journalism studies and political communication suggests the field is characterized by a methodological imbalance (i.e., many quantitative studies, few qualitative ones). This paper suggests the problem is better understood as an epistemological imbalance. We suggest that one epistemology—we call it “universalism”—underpins much comparative scholarship. While this approach produces numerous comparative insights, it also struggles to adequately account for the diversity of contexts it studies. We therefore describe an alternative epistemology, which we term “contextualism.” This approach aims to identify the mechanisms or principles that unify or differentiate cases across contexts. We suggest that progress in the field depends in part on the coexistence of multiple epistemologies, each with careful awareness of its strengths and limitations.

Full text

https://doi.org/10.1177/1940161218771899
The International Journal of Press/Politics
1 –18
© The Author(s) 2018
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1940161218771899
journals.sagepub.com/home/hij
Research Article
The Universal and the
Contextual of Media
Systems: Research Design,
Epistemology, and the
Production of Comparative
Knowledge
Matthew Powers1 and Sandra Vera-Zambrano2
Abstract
Current discussions about the state of comparative research in journalism studies
and political communication suggests the field is characterized by a methodological
imbalance (i.e., many quantitative studies, few qualitative ones). This paper suggests
the problem is better understood as an epistemological imbalance. We suggest
that one epistemology—we call it “universalism”—underpins much comparative
scholarship. While this approach produces numerous comparative insights, it also
struggles to adequately account for the diversity of contexts it studies. We therefore
describe an alternative epistemology, which we term “contextualism.” This approach
aims to identify the mechanisms or principles that unify or differentiate cases across
contexts. We suggest that progress in the field depends in part on the coexistence of
multiple epistemologies, each with careful awareness of its strengths and limitations.
Keywords
comparative research, press systems, global news, journalism
Toward the end of a recent article reviewing comparative research on journalism and
political communication, Daniel Hallin and Paolo Mancini (2017: 165–66) issue a
warning. While the field has “advanced dramatically” in developing quantitative
1University of Washington, Seattle, WA, USA
2Universidad Iberoamericana, Mexico City, Mexico
Corresponding Author:
Matthew Powers, Assistant Professor, Department of Communication, University of Washington, Box
353740, Seattle, WA 98195, USA.
Email: mjpowers@uw.edu
771899HIJXXX10.1177/1940161218771899The International Journal of Press/PoliticsPowers and Vera-Zambrano
research-article2018

2 The International Journal of Press/Politics 00(0)
indicators for core concepts, they write they are “troubled by the imbalance in recent
comparative research, which has been heavily dominated by quantitative approaches
with limited development of other approaches.” This imbalance, they suggest, threat-
ens to create an “academic monoculture” in the field, whose effect is to produce a
“very limited picture” of real-world developments in journalism and political com-
munication. They call for a greater number of case studies, field-based methods and
synthetic approaches that can provide the knowledge necessary for assessing the valid-
ity of quantitative measurements, as well as process tracing that can link causes and
consequences and, thus, move the comparative enterprise “beyond recognizing pat-
terns to explain them.”
The imbalance that Hallin and Mancini identify is an important one, and their call
to diversify methods is one that we wholeheartedly support. At the same time, we
believe there is more to the current imbalance than simply the dominance of certain
methodological techniques over others. In our view, the dominance of quantitative
methods is linked to an epistemological perspective that sees comparative knowledge
advanced by standardizing core concepts. While this approach has produced numerous
comparative insights, it struggles recurrently to account for the diverse contexts in
which those concepts are used. It is this epistemological imbalance—rather than the
preponderance of quantitative methods per se—that creates a limited picture of real-
world developments, and which hinders the overall progress of knowledge in the field.
This paper has three aims. First, we seek to recast discussions about the state of
comparative scholarship on journalism and political communication. Rather than
emphasize the need for diversity in methodological techniques per se, we argue for
paying closer attention to the epistemologies that shape the very ways comparative
research on journalism and political communication is produced. Second, we describe
the epistemological perspective that underpins much comparative research.
Universalism, as we call it, produces comparative knowledge by analyzing variables
assumed to have similar (i.e., universal) meanings across the cases analyzed. The very
effort to standardize these variables, which universalists see as crucial for building
cumulative knowledge, also abstracts away from the contexts in which media systems
are embedded, and sometimes leads researchers to misread the phenomena they study.
Third, and given universalism’s limitations, we describe an alternative approach: con-
textualism. This approach aims to understand the meaning of an idea or practice in its
context and uses comparison to examine the mechanisms or principles that unify or
differentiate cases.
Our argument is based on a review of comparative cross-national research pub-
lished in leading journals, as well as a careful reading of key statements by leading
scholars in the field.1 It is also based on our experience as comparativists seeking to
produce research that draws from an epistemological perspective that is infrequently
encountered, rarely formalized in explicit terms, and frequently misunderstood as
either “unscientific” or nonexplanatory in the extant literature. We do not aim to show
that one approach is better or truer than another. Instead, we depict two ways of doing
comparison that can each be rigorous and explanatory on their own terms, and which
provide different perspectives on similar research issues. In doing so, we seek to move

Powers and Vera-Zambrano 3
beyond general agreement about the importance of multiple approaches, and to pro-
vide a formal statement of the strengths and limitations of two actually existing
approaches, which scholars can use when designing their own comparative projects.
Bringing Epistemology to the Fore
It would be difficult to overstate the energy and attention leading scholars have devoted
to standardizing core concepts in the comparative study of journalism and political
communication. Pippa Norris (2011: 369) has written that the field is “hampered by
poor conceptualization and measurement.” Frank Esser et al. (2012: 142), in their roles
as leaders of a group of European political communication researchers, produced a
special issue of Journalism that aimed to clarify concepts and standardize core vari-
ables with the aim of “increasing comparability and cumulativity.” A recent edited
volume describes the empirical results of their efforts to achieve such aims (de Vreese
et al. 2017; see also Albæk et al. 2014). Elsewhere, David Weaver and Lars Willnat
(2012: 547) rue the lack of standardized questionnaires or representative samples in
comparative research, which “makes it difficult to evaluate journalistic norms and
values in a broader, international context” and Folker Hanusch and Thomas Hanitzsch
(2017: 526) report that the Worlds of Journalism project was founded “in response to
these limitations and calls for actions.” Recurrent discussions of validity and reliabil-
ity subtend these specific projects, as do calls for shared research infrastructures
among scholars (de Vreese 2017; Worth and Kolb 2012).
There is plenty to applaud in these efforts. Such efforts help clarify what is being
studied, how it is being studied, and the effects of methodological decisions on the
findings of a given study (see Hallin and Mancini 2017 on this point). As the number
of comparative studies proliferate, such clarity is necessary for scholars seeking to
integrate the findings of multiple studies into larger theoretical accounts about the
state of the field. Moreover, efforts to identify comparable indicators across cases
helps extend, refine, and test theories pertaining to various aspects of journalism and
political communication. One core contribution of the creation of comparable indica-
tors has been to show just how atypical the U.S. media system—long, and for too
many still, the common frame of reference—is on various dimensions (e.g., level of
commercialism; see Aalberg et al. 2010). Finally, the collaborations these efforts to
improve research design occasion also lead to innovative ways of thinking about aca-
demic research partnerships. Given real-world limitations of resources (money, knowl-
edge of distinct media systems, etc.), it is important to think about how to structure
collaborative work (see Hanusch and Hanitzsch 2017 for a thoughtful discussion on
this issue).
At the same time, it is important to see these efforts as just one way of producing
comparative knowledge about journalism and political communication. Standardizing
core concepts is certainly one way to produce comparative knowledge; it is hardly the
only way. Moreover, and as we discuss below, the focus on standardizing core con-
cepts is not unproblematic on its own terms, as it often fails to account for the diverse
real-world contexts in which those concepts are deployed.

4 The International Journal of Press/Politics 00(0)
Critics of the effort to measure progress in the field by the degree to which concepts
are standardized typically couch their concerns in terms of methodological techniques
(Hallin and Mancini 2017). To be sure, there is an empirical basis for such an associa-
tion. The epistemology we describe below as universalist often goes hand-in-hand
with quantitative techniques (e.g., surveys, content analysis), and calls for qualitative
techniques often arise because of their perceived epistemic value (and not simply a
desire to use different methods). But a greater number of qualitative studies will not on
their own address the “academic monoculture” about which Hallin and Mancini rightly
warn. Such studies can be descriptive in nature, thus offering limited import beyond
the cases they study. Moreover, qualitative research can and sometimes does rely on
similar forms of standardization and, in such cases, shares the strengths and limitations
associated with quantitative versions of the approach.2
We therefore suggest that the imbalance can be better grasped as epistemological
rather than one of methodological technique. By epistemology, we refer to theories
about how one gains knowledge of the social world, which is itself premised on a
researcher’s underlying theories of the social world (Bachelard 1938). In framing the
issue in this way, we underscore the basic point that researchers’ epistemologies are
best viewed as perspectives on the social world. The point of a perspective is not to
provide all the answers to all the questions. Instead, a perspective gives a vantage point
from which one can understand a problem and provide partial answers to it. Thus,
rather than see the standardization of core concepts as the primary way to improve
comparative knowledge, an emphasis on epistemology begins from the premise that
such a perspective has strengths and limitations. On this view, progress in the field is
made possible in part by researchers explicating the strengths and limitations of their
respective approaches.
Of course, comparative scholars routinely state the importance of having multiple
approaches.3 Yet there is less certainty about what alternative approaches in the com-
parative study of journalism and political communication might look like. In some
instances, scholars simply deem anything that does not proceed by way of standardiza-
tion as lacking in rigor. This is the basis of Pippa Norris’s (2009: 334) well-known
criticism of Hallin and Mancini’s synthetic approach, which she terms “fuzzy, impres-
sionistic, and unscientific.” In other instances, scholars deem alternative approaches
important but suggest they lack some capacity that their approach fulfils. Thus, Frank
Esser (2013: 116) writes that an interpretive and historically grounded approach pro-
vides “structured, accurate, and detailed descriptions” but “otherwise offers little by
way of explanation or theory.” We think this accurately describes a fair amount of
contemporary comparative research (quantitative and qualitative), but we think it is
crucial to describe ways of doing comparison that are rigorous and explanatory but do
not assume that standardizing variables is the only way to do so.
Even among comparative media scholars who advocate for “contextually-specific”
accounts, there is rarely explicit discussion of how such research can be conducted
comparatively.4 At times, these discussions focus on single-nation case studies, which
are certainly important for knowledge production but are not explicitly comparative in
nature (and thus do not face standard comparative questions regarding equivalence

Powers and Vera-Zambrano 5
across cases).5 At other times, scholars simply call for more context but offer limited
guidance in terms of how, exactly, one should go about providing it (Hallin and
Mancini 2012). Thus, to advance alternative ways of producing comparative knowl-
edge, we seek to explicate more fully what this approach looks like and how it differs
from the dominant approach in comparative scholarship.
To be sure, there are many ways to study context, and comparative scholars of vary-
ing epistemological traditions insist that context matters. Our aim here is to show why
universalist accounts struggle to account for context, and to describe one way of being
contextual and comparative. Our own approach—described below—draws on
Bourdieu’s field theory. We recognize that other theoretical traditions (e.g., grounded
theory, interpretivist traditions, critical realism) share many aspects of the epistemol-
ogy we describe but draw on different theoretical toolkits to analyze social life. Our
specific use of Bourdieu can, thus, be seen as illustrating one way—our way, hardly
the only one—of being contextual.
Universalism and Its Limits
We suggest that the dominant perspective in comparative research on journalism and
political communication can be understood as sharing a “universalist” epistemology.
We define universalism as an approach that examines similarities and differences in
journalism and political communication by analyzing variables (e.g., professionalism,
commercialism) that are assumed to have similar (i.e., universal) meanings across the
cases analyzed. This approach utilizes different methodological techniques (including
qualitative ones) and is integrated into different theoretical perspectives. What unites
research under the universalist label is the emphasis on standardizing core concepts
and variables. This approach provides several important insights but also has limits
that are not frequently recognized. Therefore, one condition of progress for the field is
a clearer understanding of these strengths and limitations.
As the name implies, the universalist approach aims to produce universal claims
about certain types of communication phenomena. A paper by Rinke (2016) provides
a nice illustration of an advantage gained through this perspective. It explores whether
shorter television sound bites reduce the probability that speakers face of justifying—
rather than merely presenting—their political opinions. Because the theoretical ratio-
nale “is expected to hold universally across different contexts,” Rinke draws on data
from three highly different media systems: evening television news in Germany,
Russia, and the United States (p. 631). His results “indicate that shrinking utterance
durations universally diminish the amount of justification for political opinions
appearing in television news largely irrespective of social context” (p. 639). This sug-
gests the universally problematic nature of this journalistic format (from the perspec-
tive of various strands of democratic theory), and the decision to study highly different
cases allows him to convincingly deduce the generalizability of his findings.
Another strength of the universalist approach is that it permits the establishment of
classifications and typologies, which in turn allow for the formulation and testing of
hypotheses. As Esser (2013: 116) notes in an overview of the field, classifications

6 The International Journal of Press/Politics 00(0)
“seek to reduce the complexity of the world by grouping cases into distinct categories
with identifiable shared characteristics.” This is certainly the most common use of
Hallin and Mancini’s (2004) landmark book, as scholars draw from the authors’ clas-
sifications and typologies to test the degree to which certain variables (e.g., media
markets, political parallelism) might lead to certain types of news content, or to
develop indicators so that such studies can be conducted (Brüggemann et al. 2014).
Among other things, research in various national settings (usually European or North
American) consistently demonstrates that certain variables (e.g., strong public service
broadcasting, moderate levels of commercialism, high levels of journalistic profes-
sionalism) are more likely to be associated with certain types of news content (e.g.,
more public affairs) (Aalberg et al. 2010; Albæk et al. 2014).6 Such research is crucial
for building “cumulative” knowledge about certain types of communication
phenomena.
The universalist approach also has certain limits. The approach begins from the
assumption that concepts and variables have similar meanings across contexts (and
that context comes from the addition of other variables, see Esser 2013: 117). This
assumption—which Blumler and Gurevitch (1995: 75) once termed naïve universal-
ism—can and does encourage a view of journalism and political communication in
which one’s own group is the basis of evaluation, and all others are scaled in reference
to it. In such cases, what is generalized is the researcher’s framework for journalism
(what it is and should be), rather than the empirical analysis of journalism across
contexts.
Consider, for instance, the fairly common item found on surveys that ask journalists
to rate “potential sources of influence” on their daily work (Eberwein et al. 2011;
Hanitzsch et al. 2011). This list typically includes items such as peers, editorial super-
visors, management, owners, government officials, and so forth. The list of influences
are assumptions about influences that stem from the people who design the survey
(almost always U.S. or Western European researchers) and are derived from reviews
of the available literature. Journalists rate the influence on a Likert scale.
Now imagine the moment such a question is asked to a journalist in Turkey. There,
most media outlets are owned by individuals and groups with backgrounds in non-
media industries (energy, urban development, health care) and use media holdings to
curry favor with government officials. Media owners appoint managers and editors
that can oversee not only the work of staff but also ensure their reporting does not run
afoul of government objections (Yeşil 2016). How should a journalist in Turkey
respond? If she or he reports being heavily influenced by news editors, managers,
bosses, and politicians, they are not referring to distinct sources of influence as the
question assumes. Instead, they are saying—or trying to say—that each of these influ-
ences are at root a function of the same phenomenon: namely, the interpenetration of
state and business power. But the survey question is not designed to hear such a
statement.
Of course, these issues can be limited and controlled for—but only to a degree.
Doing so requires enormous sophistication in the research protocol, and such sophis-
tication is rarely evident. (The Worlds of Journalism study is certainly the most

Powers and Vera-Zambrano 7
ambitious in this respect). But what is crucial to note is that the effort to address the
problem is discussed in terms of standardizing the research design, when the issue has
to do with the diversity of real-world experiences. Survey protocols, for example, are
translated into multiple languages, and then back-translated by researchers in those
countries to ensure that everyone is asking the same questions. Results are regressed
to produce cluster analyses, thus verifying that journalists share seemingly similar
perceptions of influence. But what the researchers are equalizing across the cases are
the questions and answers, not the distinct social realities being analyzed. Addressing
these distinctive social realities would require multiple protocols, as concepts depend
on their contexts, but doing so would sacrifice the universalizing aims of the survey.
Another issue associated with the universalist perspective is what Bourdieu and his
collaborators (1991: 15), drawing on Durkheim, call “artificialism.” This refers to the
search for superficial similarities across cases, while ignoring the underlying processes
or functions that shape them. Such artificialism can be seen in recently published
research on personalization (Van Aelst et al. 2016). This concept examines the extent
to which political news focuses increasingly on individual politicians rather than polit-
ical institutions. While this phenomenon has received prior scholarly attention, the
extent to which political news is becoming “personalized” across Western Europe is
unknown, and the causes that drive this development are likewise unclear. The authors,
therefore, develop a series of testable hypotheses: personalization may be associated
with different types of media organizations (e.g., the assumption being that television
on the whole is more likely to personalize than newspapers, and commercial broad-
casters are more likely to personalize than public broadcasters); it may correspond to
different media system variables (e.g., with more competitive media markets assumed
to foster greater degrees of personalization); it may also be a function of distinctive
political system characteristics (e.g., majoritarian systems may be more likely to foster
personalization than systems with proportional representation). They then operational-
ize the concept of personalization and code for its presence in political news in sixteen
countries.
At the descriptive level, the authors find that Italy and the United Kingdom have the
most personalized news. In both countries, approximately 75 percent of political actors
in the news are individual politicians (rather than spokespersons for political parties or
government institutions). However, as Paolo Mancini (2017) points out in a review of
the book, the data misunderstands the Italian case, where legislation requires news
outlets to permit equal time to politicians of different parties in the news. The large
number of individual politicians in Italian media is, thus, not evidence of personaliza-
tion as the authors define it—that is, the extent to which political news focuses increas-
ingly on individual politicians rather than political institutions. Instead, the individual
politicians in the news are party members “who appear on the screen because the news
outlets, because of legislative impositions, need to give voice to politicians of different
parties” (Mancini 2017: 390). Thus, the similarity between the Italian and British
cases is an artificial one that conceals the distinctive dynamics shaping the presence of
political actors in the news.

8 The International Journal of Press/Politics 00(0)
As Mancini notes, this miscomparison does not stem from lack of methodological
rigor. Indeed, the larger book, from which the chapter is drawn, is admirable in part for
the sophistication of the research design, which includes careful discussions of sam-
pling strategies, data collection, and methods of analysis. Nor, we would add, does the
problem stem from lack of awareness that things may be more complex than their
analysis reveals, as the authors suggest at the end of the chapter. But the problem does
arise, we suggest, in part because the authors pay less attention to what the concept of
personalization means in distinctive national contexts. By abstracting away from those
contexts, the authors are able to test for the presence or absence of universal causal
factors (e.g., the extent to which television everywhere is likely to produce personal-
ized news coverage), but they do so by sacrificing a more basic understanding of the
phenomenon under analysis.
As with surveys, researchers exhibit greater and lesser degrees of awareness of the
challenges involved in content analysis. Our aim here is not to identify better and
worse resolutions to these problems. Instead, it is to point out that it is a problem that
derives from a desire to standardize the research design to the greatest degree possi-
ble.7 This desire is reasonable but has costs not often acknowledged. In the above
example, for instance, artificial similarities belie deeper differences in contexts and a
misreading of real-world developments. Not only does this make it hard to understand
the phenomenon under analysis, it also suggests that research is unlikely to be “cumu-
lative” in any simple sense (as different cases cannot simply add up to a larger finding,
because distinctive contexts need to be accounted for).
One of the core stated aims of most comparative researchers is to shed light on their
own national cases (de Vreese 2017; Esser and Hanitzsch 2012). The implication of
such statements is that such light can force researchers to rethink their assumptions
and challenge some of their core beliefs about what journalism and political commu-
nication are or should be. To be sure, universalist approaches shed light on a variety of
issues, but on our reading, their assumptions about what constitutes good journalism
is rarely challenged. The Northern European model of public broadcasting, media
accountability, and strong journalistic professionalism serve as both the starting point
of empirical investigations and the normative standard by which results are evaluated.
We do not disagree with the normative standard per se, but the tendency to generalize
their visions of good journalism and political communication to other cases actually
maintains, rather than challenges, extant assumptions. Challenging those assumptions,
thus, requires a different epistemological approach.
Contextualism as an Alternative Epistemology
Contextualism begins with a simple question: What does an idea or practice mean in
its context? This premise requires explicit concern with two issues not strongly present
in universalist accounts. First, it requires a historical understanding of what a given
variable or phenomenon means in a particular case (Humphreys 1996),8 and second, it
demands local definitions, representations, and ideas that exist in specific contexts be
taken seriously (Voltmer 2012).9 This approach certainly can provide hypotheses to be

Powers and Vera-Zambrano 9
tested by researchers operating from universalist perspectives. However, it does not do
so by standardizing core concepts in the sense identified above. Instead, a contextual
approach aims to explain the mechanisms or principles that underlie the phenomena
being observed across cases, and this happens through the study of particular events
and specific individuals. Rather than relying primarily on methods to equalize cases,
the contextual approach delegates an important role to theory as a way to give order to
results.
Our own theoretical framework for being contextual is Bourdieu’s (1996) field
theory. This theory conceives of social life as occurring largely in semiautonomous
social fields, each characterized by distinct logics that cannot be reduced to external
causes. These fields shape and are shaped by the actions taken by actors within them,
and the task of the researcher is to identify the specific logic or logics that shape action
in a given field (i.e., in a specific context).10 This happens partly by taking seriously
actors’ local definitions (i.e., knowing what they mean by using specific terms) and
partly by understanding that these definitions are, themselves, the product of historical
struggles among actors in the field.
One example of this type of Bourdieusian contextualism can be found in the work
of Benson (2013), and the example shows that the issue is best conceived as one of
epistemology rather than just method. For his study of immigration news in France
and the United States, Benson uses content analysis to examine, among other things,
whether French or American journalism most closely accord with normative aims of
“multiperspectivalism.” To answer the question, Benson does not pick strictly similar
news outlets (e.g., one national newspaper, one regional newspaper). Rather, he con-
structs his sample by thinking about the specific role that individual newspapers play
in each country’s journalistic field (i.e., its contexts, and the local ideas associated with
different newspapers). Among other things, this leads him to include L’Humanite, a
communist newspaper, in the French sample, but not the U.S.’s People’s Weekly World,
given the marginal place of the far-left in American politics and media.
This contextual approach runs through the entirety of Benson’s field analysis.
Where most cross-national content analyses use the news article as the unit of analysis,
Benson uses instead the “ensemble”—that is, the range of voices and viewpoints on
immigration that emerge in a single edition of a newspaper, rather than at the story
level. Why? In the United States, journalists seek to provide a range of perspectives
and ideas within a single news article. In France, though, journalists seek to achieve
diversity at the level of the news page as a whole. Thus, to evaluate the extent to which
American or French journalists fulfill “multiperspectival” norms, Benson first has to
understand historically how the norm of pluralism has been fulfilled in the two cases,
and to build a content analysis that accurately captures those two cases.
This approach allows Benson to challenge various assumptions about what consti-
tutes good journalism in the two cases, and to identify potential ways that good jour-
nalism can be achieved. To take the U.S. case as an example: Benson shows the
potentially positive role played by the state—through subsidies to newspapers—in
shaping pluralism, something that remains anathema to most American journalists and
to many American scholars of journalism. It also questions “narrative” forms of

10 The International Journal of Press/Politics 00(0)
storytelling, which emphasize dramatic individual accounts. Such formats are lauded
by American journalists as providing empathy, engaging audiences, and showing the
real-world impact of government policies. Yet Benson shows that such forms typically
preclude attention to structural complexities, power dynamics, or diverse perspectives,
and that alternative formats like the “debate ensemble” format—more common in
France—actually do a better job providing explanations that incorporate diverse per-
spectives. While Benson suggests that lessons can be learned from the French case, he
also stresses that French media policies cannot simply be placed mechanically into the
American media system. Instead, one needs to think about the field dynamics that
shape how journalists do their work.
Another example, based on our own research, can help illustrate the idea of identi-
fying the mechanism or principle that links the results across cases (Powers and Vera-
Zambrano 2017; see also Christin 2018). It begins from prior research, which finds
that American and French journalists maintain different relationships to their audi-
ences. Stated schematically, we can describe these differences as a tendency among
French regional journalists to largely ignore their audiences in their everyday work,
while American journalists talk about—and sometimes even engage with—them regu-
larly.11 On the surface, it seems “audiences” are a unique comparative variable, and
different results are found in the two countries. To the extent that comparative scholars
have analyzed this variable, they tend to deductively attribute the results to “system-
level” variables like professionalism, with lower levels of journalistic professionalism
in France hypothesized as disincentivizing audience engagement (Engesser and
Humprecht 2015).
Closer inspection (made possible by interviews with journalists), however, reveals
that local journalists in France and the United States simply have different conceptions
of what audiences are (i.e., different local definitions and representations of the con-
cept). Here again, stating the findings schematically, we can describe these concep-
tions as a tendency among American journalists (especially those working in regional
news media) to think about news consumers as their audience, while French journal-
ists think of their peers as their audience (Powers and Vera-Zambrano 2017). To under-
stand this difference, one must first understand the role played by audiences in securing
revenues for news organizations in the two countries. In France, dependence on the
audience for an organization’s stability is less strong, due in large part to the law of
pluralism, which stipulates that the state has a role to play in guaranteeing the exis-
tence of news organizations. In the United States, news organizations are heavily reli-
ant on audiences, historically as a basis for securing advertising revenues (itself
premised on audience size) and increasingly as a direct source of payment for news
consumption (e.g., via subscriptions). These differences are also shaped by historical
differences in the division of journalistic labor. At French regional news outlets, media
organizations commonly hire people with the explicit aim of boosting readership, and
historically, these individuals are not journalists. As a result, attention to audiences is
typically evaluated negatively. In the United States, such evaluations are generally
much less common, because the responsibility for getting readership has been devolved
in large part to journalists.

Powers and Vera-Zambrano 11
This finding does not invalidate prior research. Indeed, it is true that journalists in
the two countries have different relationships to their audience. Rather, our approach
aims to explain the mechanisms or principles that underlie the phenomena under
observation. In the case above, the principle that unites the French and American cases
is the idea of recognition, and the mechanisms of market dynamics and journalistic
divisions of labor shape how it is pursued. In both countries, journalists see audiences
as a way to generate recognition in their respective fields. How that happens, and
indeed what audiences mean to journalists, depends entirely on the dynamics of the
specific field. Regional journalists in the United States pay attention to audiences in
large part to advance their careers or demonstrate efforts to address economic prob-
lems by their news organizations. French journalists, instead, think of audiences as
primarily composing their peers, in part because they are not driven by the same eco-
nomic necessity and in part because the division of labor devolves those functions to
others.
This mechanism or principle can certainly be integrated into universalist hypothe-
ses. At the same time, the explanation of anything is always necessarily local, because
causal mechanisms can only be studied through specific individuals or events. Rather,
an important role for theory—field theory in our examples—is to integrate different
cases into a larger narrative that gives order to the results. This flips the emphasis
found in much comparative research, away from standardized research design and
toward theory. Rather than standardizing measures for hypothesis testing, this approach
seeks first to understand the principle or mechanism that connects the cases. Identifying
such a principle or mechanism is, in our view, a precondition for figuring out what we
might want to measure, and how we might go about measuring it.
As with any perspective, contextualism has limits. Almost by definition, the
approach favors—in some sense, it requires—a small-N comparison. Researchers
need to have detailed knowledge of the cases being studied. Human limitations mean
that one can only speak so many languages and know so many countries. We would
also suggest that the political economy of comparative scholarship limits the potential
for this research in important ways. Funding bodies give grants to researchers who
produce results in a relatively short time period, and these time scales are often unre-
alistic for contextual accounts. Moreover, the social networks that grow out of these
grants also foster collaboration on issues of research design, which are presumably
easier to garner cross-national agreement on than the messy realities of media systems,
or sharing a theoretical perspective from which to analyze such systems. These factors
contribute to built-in limits to contextualism that favor a small-N approach to
comparison.
There also seems to us a potential danger of exoticizing other countries’ media
systems. If the universalist can be criticized for implicit ethnocentrism, contextualists
often run the danger of romanticizing or glamorizing the media systems of others.
Western European scholars, for instance, sometimes criticize American scholars for
glamorizing the benefits of public service broadcasters while ignoring their weak-
nesses (see, for example, Berry 2016). This is problematic on its own terms, but it is
also worrisome because exoticism limits the extent to which a comparativist sheds

12 The International Journal of Press/Politics 00(0)
light on each of the cases a researcher is analyzing. Media systems are complex enti-
ties, each potentially with things for a comparativist to appreciate in their own terms.
The goal ought to be to shed light on all of the cases under analysis.
Conclusion
The philosopher Abraham Kaplan (1964: 11) once told a story “of a drunkard search-
ing under a street lamp for his house key” even though he had dropped the key else-
where. “Asked why he didn’t look where he had dropped it, he replied, ‘It’s lighter
here!’” In telling the story, Kaplan sought to caution researchers against the temptation
of using a single epistemology to understand a wide range of research problems. While
comparative scholarship on journalism and political communication has made—and
continues to make—great strides, the field faces a similar concern. By focusing so
heavily on standardizing core concepts, comparative scholarship certainly gains
insights into a range of important comparative questions. But it struggles to provide
theoretical insights that take into account the diverse contexts it studies.
In this paper, we have suggested that it would be useful to reframe the debate about
“imbalances” in the field (Hallin and Mancini 2017). Where scholars often suggest the
issue is one of methodological technique (i.e., too many quantitative studies, not
enough qualitative scholarship), we believe the issue is better—and more broadly—
understood as an epistemological imbalance. Comparative research on journalism and
political communication has dedicated considerable energy to issues of standardized
research design, with the aim of improving generalizability and cumulativity. We have
sought to show the strengths and limitations of this approach, while introducing an
alternative epistemological perspective on the production of comparative knowledge.
This approach, which we call contextualism, seeks to explain the mechanisms or prin-
ciples that underlie the phenomena being observed across cases.
Several implications follow from our analysis. One implication is that universalist
approaches need to be more explicit about their own limits. The drive to standardize
concepts and operationalize terms is in many respects laudable, but it also needs to be
accompanied by careful considerations about the limitations of such efforts. Such con-
siderations would consider, among other things, ways to address potential biases intro-
duced by ethnocentrism and artificialism. A second implication is that the growth of
qualitatively driven studies will not on their own fix the imbalance that Hallin and
Mancini (2017) identify. Such studies can and sometimes do conform to universalist
aims of standardization. Moreover, given the general preponderance of descriptive
scholarship in journalism and political communication (Hallin and Mancini 2012:
217), it is entirely possible to envision a growth in largely descriptive studies that do
little to move the field forward theoretically.
A final implication is that evaluations of different epistemological approaches to
comparison are best seen not as questions of “truth value” per se (i.e., which perspec-
tive provides the correct answer), but instead ought to focus on the light they shed on
a problem. Here, too, scholars seem to agree in principle. Yet reviews of the state of
the field often paint an evolutionary portrait of comparative scholarship that belie a

Powers and Vera-Zambrano 13
preference for one perspective. In their introductory chapter of The Handbook of
Comparative Communication Research, for example, Frank Esser and Thomas
Hanitzsch (2012: 7–8) describe “early attempts” at comparison that consisted largely
of “nation-by-chapter reporting.” These efforts were surpassed by a “second develop-
mental step in the field’s evolution” that included “two-country comparisons.” Yet
“because of the limited ability to generalize from two-country studies, they are increas-
ingly being replaced with medium-N and large-N studies.” We suggest that a better
approach is simply to think about the epistemological bases of these different studies
and to carefully understand the terms on which each attempt to be explanatory and
rigorous.
While we highlight differences in the two epistemological approaches discussed
here, we want to stress there are ways in which the two can be mutually informative.
Contextual accounts certainly help validate and interpret the findings of universalist
approaches; they can also certainly be used to help refine the hypotheses of universal-
ist approaches. Similarly, universalist approaches can examine the extent to which
processes identified in one setting exist elsewhere, and the classifications and typolo-
gies such research produces is and continues to be a useful measuring stick by which
to evaluate the extent to which media systems are and are not converging over time. At
the same time, we think there are many knowledge gains to be made through a contex-
tualist approach. Understanding the logics and mechanisms that underlie the phenom-
ena being analyzed comparatively are a crucial first step to problematizing the very
research questions that scholars pose. This cannot be achieved by a field that focuses
overwhelmingly on universalist approaches.
In identifying the strengths and limitations of various approaches, we have sought
to formalize two distinctive ways of producing comparative knowledge. This approach
pays less attention to the social organization of comparative scholarship on journalism
and political communication, which is certainly important for shaping the production
of knowledge in the field. Put simply, the turn toward comparison in scholarship favors
some researchers and research styles more than others. While a full accounting of this
state of affairs is beyond the scope of this paper, we would suggest that universalist
approaches are better adapted to the current institutional environment than are contex-
tualist approaches (e.g., funding agencies privilege large-N studies, academic journals
require explicit discussions of methods rather than theories, etc.). A discussion of epis-
temological approaches is an important way of recognizing these imbalances; how-
ever, epistemology alone cannot address all the issues we discuss here. The social
organization of comparative scholarship also needs to be addressed if comparative
scholarship on journalism and political communication is to become truly embracing
of multiple approaches.
We want to stress that there are multiple ways of being contextual. Our examples
here draw on field theory because it is the framework within which we have done
much of our comparative work. But we recognize a range of other theoretical frame-
works that also aim at providing context in comparative perspective. These include but
are not limited to grounded theory, interpretivism, and critical realist traditions.

14 The International Journal of Press/Politics 00(0)
In sum, while comparative research has grown tremendously, progress in the field
depends in part on moving beyond the current emphasis on standardized research
design. We believe that a focus on epistemology is important, and at present underem-
phasized, in the field. Knowing how to produce better comparative knowledge is a
prerequisite not only for better scholarship but also for providing the basis for under-
standing potential transformations of actually existing media systems.
Authors’ Note
Both authors contributed equally to this manuscript; names listed alphabetically.
Acknowledgments
The authors thank Olivier Baisnée, Rodney Benson, Lucas Graves, Risto Kunelius, Adrienne
Russell, Annika Sehl, three anonymous reviewers, and Press/Politics editor Rasmus Kleis
Nielsen for thoughtful comments and criticisms. Thanks also to the participants in the Reuters
Institute for the Study of Journalism Summer School on Comparative Qualitative Research in
Journalism and Political Communication for discussing an earlier version of this paper, as well
as attendees of a University of Washington Department of Communication colloquium.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship,
and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this
article.
Notes
1. We examined the prevalence of comparative research in four major scholarly journals:
Journal of Communication, the flagship publication of the International Communication
Association (ICA); Political Communication, the joint publication of the Political
Communication Division at ICA and the American Political Science Association;
International Journal of Press/Politics, which was created in 1996 with the explicit mis-
sion of fostering comparative and internationally oriented political communication schol-
arship; and Journalism: Theory, Practice and Criticism, which was launched in 2000 with
the aim of providing a scholarly home for research on the news media, often including
political communication scholarship. In each journal’s website, we entered the term com-
parative and culled through the list of results to only include those specifically focused on
comparison as defined by Esser and Hanitzsch (2012: 5): research that involves compari-
sons “between a minimum of two macro-level units (systems, cultures, markets, or their
subelements) with respect to at least one object of investigation relevant to communication
research.” We use this data not to describe the state of the field but to shape our discussion
of how comparative knowledge is produced. We address both journalism studies and politi-
cal communication because comparative scholarship in these cognate fields face similar
epistemological issues.

Powers and Vera-Zambrano 15
2. See footnote 7 for an example.
3. Tellingly, though, the calls are almost always framed in terms of “methodological plural-
ism” with limited attention to the underlying epistemological bases of the research in ques-
tion (see, for example, de Vreese 2017; Esser 2013; Hallin and Mancini 2012).
4. The situation is different in other areas of comparative scholarship, where scholars have
with varying degrees of success grappled with ways to produce contextually specific com-
parative knowledge (e.g., Collier 2011; Ragin 2014; Steinmetz 2004).
5. Some scholars—generally rooted more in the humanities—claim comparison is inherently
problematic due to the uniqueness of events (see Steinmetz 2004 for an overview of these
arguments). We are sympathetic to the criticisms they raise regarding commensurability
but seek in this paper to identify a form of comparison that can account for individual con-
texts while making comparisons across them. Humphreys (1996) provides one important
example of the potential to do scholarship that is both historical and comparative.
6. It is worth noting that the findings often do not hold outside the geographic settings that
Hallin and Mancini used as the basis for their own study (Waisbord 2000), and that Hallin
and Mancini (2012) are themselves very clear about the need to rework these variables in
contexts beyond Western Europe and North America.
7. This desire to standardize can also be seen in qualitative approaches, like “qualitative con-
tent analysis” or QCA. This method—increasingly utilized in comparative research—typi-
cally begins with an outcome whose differential cross-national salience a researcher seeks
to explain. Yet this approach requires assuming that the meaning of the outcome is constant
across cases. Downey and Stanyer (2013) provide a useful example. Their sophisticated
QCA examines the differential prevalence of sex scandals across eight countries without
questioning whether a sex scandal has the same function in all cases. This allows them to
read their results as indicative of cross-national differences in what is considered public
and private. Whatever the validity of the finding, it ignores the specific contexts in which
such scandals arise. In France, for instance, sex scandals typically only become news when
they affect the functioning of politics, whereas in the United States, sex scandals are often
used to discredit political actors. It is precisely this type of contextual understanding that
universalist approaches—including those using qualitative methods—struggle to achieve.
8. Developing a historical understanding is not possible without historical scholarship.
Therefore, even when not directly utilizing historical research methods, contextual
approaches nonetheless rely heavily on them to form the background understanding of the
phenomena analyzed.
9. As Voltmer (2012: 233) notes, the concepts used by comparative scholars “are not as unan-
imous as they appear to be.” Ideas like press freedom do not have a “fixed meaning that
could claim validity outside time and space.”
10. It is worth noting that Bourdieu made extensive use of both quantitative (e.g., surveys) and
qualitative (e.g., interviews) techniques.
11. This finding is specifically linked to regional journalists in the two countries. As Christin
(2018) shows, the dynamics are different when basing the comparison on journalists work-
ing in media capitals such as Paris and New York.
References
Aalberg, Toril, Peter van Aelst, and James Curran. 2010. “Media Systems and the Political
Information Environment: A Cross-National Comparison.” International Journal of Press/
Politics 15 (3): 255–71.

16 The International Journal of Press/Politics 00(0)
Albæk, Erik, Arjen van Dalen, Nael Jebril, and Claes de Vreese. 2014. Political Journalism in
Comparative Perspective. New York: Cambridge University Press.
Bachelard, Gaston. 1938. The Formation of the Scientific Mind. Paris: Clinamen Press.
Benson, Rodney. 2013. Making Immigration News: A French-American Comparison. New
York: Cambridge University Press.
Berry, Mike. 2016. “No Alternative to Austerity: How BBC Broadcast News Reported the
Deficit Debate.” Media, Culture & Society 38 (6): 844–63.
Blumler, Jay, and Michael Gurevitch. 1995. The Crisis of Public Communication. London:
Routledge.
Bourdieu, Pierre. 1996. The Rules of Art. Stanford: Stanford University Press.
Bourdieu, Pierre, Jean-Claude Chamboredon, and Jean-Claude Passeron. 1991. The Craft of
Sociology: Epistemological Preliminaries. New York: Walter de Gruyter.
Brüggemann, Michael, Sven Engesser, Florin Büchel, Edda Humprecht, and Laia Castro. 2014.
“Hallin and Mancini Revisited: Four Empirical Types of Western Media Systems.” Journal
of Communication 64 (6): 1037–65.
Christin, Angèle. 2018. “Counting Clicks: Quantification and Variation in Web Journalism in
the United States and France.” American Journal of Sociology 123 (5): 1382–415.
Collier, David. 2011. “Understanding Process Tracing.” PS: Political Science & Politics 44
(4): 823–30.
de Vreese, Claes. 2017. “Comparative Political Communication Research.” The Oxford Handbook
of Political Communication, ed. Kate Kenski and Kathleen Hall Jamieson, 287–300. Oxford:
Oxford University Press.
de Vreese, Claes, Frank Esser, and David Nicolas Hopmann, eds. 2017. Comparing Political
Journalism. London: Routledge.
Downey, John, and James Stanyer. 2013. “Exposing Politicians’ Peccadilloes in Comparative
Context: Explaining the Frequency of Political Sex Scandals in Eight Democracies Using
Fuzzy Set Qualitative Comparative Analysis.” Political Communication 30 (3): 495–509.
Eberwein, Tobias, Susanne Fengler, Epp Lauk, and Tanja Leppik-Bork, eds. 2011. Mapping
Media Accountability—In Europe and beyond. Köln: Herbert von Halem Verlag.
Engesser, Sven, and Edda Humprecht. 2015. “Frequency or Skillfulness: How Professional
News Media Use Twitter in Five Western Countries.” Journalism Studies 16 (4): 513–29.
Esser, Frank. 2013. “The Emerging Paradigm of Comparative Communication Enquiry:
Advancing Cross-National Research in Times of Globalization.” International Journal of
Communication 7:113–28.
Esser, Frank, and Thomas Hanitzsch. 2012. “On the Why and How of Comparative Inquiry
in Communication Studies.” In Handbook of Comparative Communication Research, ed.
Frank Esser and Thomas Hanitzsch, 3–22. London: Routledge.
Esser, Frank, Jesper Strömbäck, and Claes de Vreese. 2012. “Reviewing Key Concepts in
Research on Political News Journalism: Conceptualizations, Operationalizations and
Propositions for Future Research.” Journalism 13 (2): 139–43.
Hallin, Daniel, and Paolo Mancini. 2004. Comparing Media Systems: Three Models of Media
and Politics. New York: Cambridge University Press.
Hallin, Daniel, and Paolo Mancini. 2012. “Comparing Media Systems: A Response to Critics.”
In Handbook of Comparative Communication Research, ed. Frank Esser and Thomas
Hanitzsch, 207–220. London: Routledge.
Hallin, Daniel, and Paolo Mancini. 2017. “Ten Years after Comparing Media Systems: What
Have We Learned?” Political Communication 34 (2): 155–71.

Powers and Vera-Zambrano 17
Hanitzsch, Thomas, Folker Hanusch, Claudia Mellado, Maria Anikina, Rosa Berganza, Incilay
Cangoz, Mihai Coman, Basyouni Hamada, María Elena Hernández, Christopher D.
Karadjov, Sonia Virginia Moreira, Peter G. Mwesige, Patrick Lee Plaisance, Zvi Reich,
Josef Seethaler, Elizabeth A. Skewes, and Dani Vardiansyah. 2011. “Mapping Journalism
Cultures across Nations: A Comparative Study of 18 Countries.” Journalism Studies 12
(3): 273–93.
Hanusch, Folker, and Thomas Hanitzsch. 2017. “Comparing Journalistic Cultures across
Nations: What We Can Learn from the Worlds of Journalism Study.” Journalism Studies
18 (5): 525–35.
Humphreys, Peter. 1996. Mass Media and Media Policy in Western Europe. Manchester:
Manchester University Press.
Kaplan, Abraham. 1964. The Conduct of Inquiry. New Brunswick, NJ: Transaction
Publishers.
Mancini, Paolo. 2017. “Review of Comparing Political Journalism.” Journalism & Mass
Communication Quarterly 94 (1): 389–91.
Norris, Pippa. 2009. “Comparative Political Communications: Common Frameworks or
Babelian Confusion?” Government & Opposition 44 (3): 321–40.
Norris, Pippa. 2011. “Political Communication.” In Comparative Politics, ed. Daniele Caramani,
352–370. Oxford: Oxford University Press.
Powers, Matthew, and Sandra Vera-Zambrano. 2017. “How Journalists Use Social
Media in France and the United States: Analyzing Technology Use across
Journalistic Fields.” New Media & Society. Published electronically September 15.
doi:10.1177/1461444817731566
Ragin, Charles. 2014. The Comparative Method: Moving beyond Qualitative and Quantitative
Strategies. Berkeley: University of California Press.
Rinke, Eike. 2016. “The Impact of Sound Bite Journalism on Public Argument.” Journal of
Communication 66 (4): 625–45.
Steinmetz, George. 2004. “Odious Comparisons: Incommensurability, the Case Study, and
‘Small N’s’ in Sociology.” Sociological Theory 22 (3): 371–400.
Van Aelst, Peter, Tamir Sheafer, Nicolas Hubé, and Stylianos Papathanassopoulos. 2016.
“Personalization.” In Comparing Political Journalism, ed. Claes De Vreese, Frank Esser
and David Nicolas Hopmann, 112–130. London: Routledge.
Voltmer, Katrin. 2012. “How Far Can Media Systems Travel?” In Comparing Media Systems
beyond the Western World, ed. Daniel Hallin and Paolo Mancini, 224–245. New York:
Cambridge University Press.
Waisbord, Silvio. 2000. “Media in South America: Between the Rock of the State and the Hard
Place of the Market.” In De-Westernizing Media Studies, ed. James Curran and Myung-Jin
Park. London: Routledge.
Weaver, David, and Lars Willnat. 2012. The Global Journalist in the 21St Century. New York:
Routledge.
Worth, Wernet, and Steffen Kolb. 2012. “Securing Equivalence: Problems and Solutions.”
In Handbook of Comparative Communication Research, ed. Frank Esser and Thomas
Hanitzsch. London: Routledge.
Yeşil, Bilge. 2016. Media in New Turkey: The Prigins of an Authoritarian Neoliberal State.
Urbana: University of Illinois Press.

18 The International Journal of Press/Politics 00(0)
Author Biographies
Matthew Powers is an assistant professor in the Department of Communication at the University
of Washington in Seattle. His academic writings explore emergent forms of journalism and
political communication, and they have been published in Journal of Communication, and
International Journal of Press/Politics, among others.
Sandra Vera-Zambrano is an assistant professor at Universidad Iberoamericana in Mexico
City. Her academic writings explore journalism and political communication, and they have
been published in Journal of Communication and Sur Le Journalisme, among others.