Content uploaded by Jesper Strömbäck
Author content
All content in this area was uploaded by Jesper Strömbäck on Jul 28, 2014
Content may be subject to copyright.
http://jou.sagepub.com/
Journalism
http://jou.sagepub.com/content/13/2/144
The online version of this article can be found at:
DOI: 10.1177/1464884911427797
2012 13: 144 originally published online 11 November 2011Journalism Susana Salgado and Jesper Strömbäck
findings
Interpretive journalism: A review of concepts, operationalizations and key
Published by:
http://www.sagepublications.com
can be found at:JournalismAdditional services and information for
http://jou.sagepub.com/cgi/alertsEmail Alerts:
http://jou.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://jou.sagepub.com/content/13/2/144.refs.htmlCitations:
What is This?
- Nov 11, 2011Proof - Jan 11, 2012Version of Record >>
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Journalism
13(2) 144 –161
© The Author(s) 2011
Reprints and permission: sagepub.
co.uk/journalsPermissions.nav
DOI: 10.1177/1464884911427797
jou.sagepub.com
Interpretive journalism:
A review of concepts,
operationalizations and key
findings
Susana Salgado
Foundation for Science and Technology, Portugal
Jesper Strömbäck
Mid Sweden University, Sweden
Abstract
The overall purpose of this article is to review theory and research on interpretive
journalism, one of the key concepts in research on the style and character of news
journalism. While it is often claimed that news journalism over recent decades has
changed from being predominantly descriptive to becoming increasingly interpretive,
our review suggests that there is a lack of systematic research in this area. The literature
is furthermore characterized by different conceptualizations and operationalizations of
interpretive journalism, as well as by different normative assumptions. Taken together,
this suggests not only insufficient conceptual clarity, but also problems related to the
comparability and cumulativity of findings. To remedy this, and based on our review of
how interpretive journalism has been conceptualized and operationalized, this article
suggests how interpretive journalism should be conceptualized and operationalized
in order to increase conceptual clarity, comparability across studies, and ultimately
research cumulativity.
Keywords
interpretive journalism, journalism of opinion, journalistic objectivity, political news
journalism
Corresponding author:
Jesper Strömbäck, Department of Media and Communication, Mid Sweden University, 851 70 Sundsvall,
Sweden
Email: jesper.stromback@miun.se
427797JOU13210.1177/1464884911427797Salgado and StrömbäckJournalism
Article
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 145
Introduction
According to many observers, the rise of a more interpretive style of journalism is one of
the most significant changes over recent decades. Djerf-Pierre and Weibull (2008: 209),
for example, characterize the rise of interpretive journalism as ‘the most significant
change in political journalism’, while Patterson (1993: 67) notes, ‘Today, facts and inter-
pretation are freely intermixed in election reporting. Interpretation provides the theme,
and the facts illustrate it. The theme is primary; the facts are secondary.’
The notion that journalism has become more interpretive is so pervasive that fre-
quently it is taken for granted without being systematically investigated. Hence, the
extent to which assertions of a more interpretive style of journalism are empirically
supported is not clear. In addition, a cursory reading of the literature reveals different
conceptualizations and operationalizations of interpretive journalism. This suggests
not only insufficient conceptual clarity, but also problems related to the comparability
of findings.
Against this background, the purpose of this article is twofold: first, to review how
interpretive journalism has been conceptualized and operationalized, and to identify
some key findings; second, to suggest how interpretive journalism should be conceptual-
ized and operationalized to increase conceptual clarity and comparability across studies.
The review focuses on studies based on quantitative content analysis, focusing on the
period since the 1960s.
In the first four sections of this article, we review different conceptualizations and
operationalizations of interpretive journalism together with some key findings of the
most relevant studies. Based on this review, in the following sections we suggest a
synthesis and how interpretive journalism should be conceptualized. We also propose
how interpretive journalism should be operationalized in studies based on quantita-
tive content analysis to increase conceptual clarity as well as comparability across
studies.
Different theoretical and conceptual approaches
The concept of interpretive journalism is both important and complex. It is important
since it is often alleged that journalism has become more interpretive, which, if true,
would signify changes in both journalism and in the relationship between journalists and
the sources and events they cover. It is also complex as it touches on a range of episte-
mological and ontological questions, as there are different journalistic styles, norms and
values in different countries, and as journalism has changed over time (Benson and
Hallin, 2007; Weaver and Willnat, 2012). Journalism may, for example, follow an objec-
tive, an impartial, a literary, or a partisan approach.
Still, while a review of the literature shows that there are different conceptualizations
of interpretive journalism, virtually all conceive interpretive journalism as somehow
opposed to or going beyond descriptive, fact-based journalism, and some emphasize a
shift of emphasis related to the five Ws of journalism: What, Where, When, Who and
Why. Generally speaking, interpretive journalism is conceptualized as entailing a greater
emphasis on the ‘meaning’ of news beyond the facts and statements of sources, and a
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
146 Journalism 13(2)
greater emphasis on the Why of the five Ws (Patterson, 1993). It is characterized by a
disbelief in the notion that it is possible to separate facts from values (Schudson, 1978).
One critic of interpretive journalism is Patterson (1980, 1993, 2000a). According to
his conceptualization, interpretive journalism is journalism driven by themes, where
facts are used mainly to illustrate the theme chosen by the journalist. This conceptualiza-
tion should be understood within the larger context of the relationship between journal-
ists and their sources. A journalism that focuses on the What, When, Where, and Who
allows sources to set the news agenda and news frames, reducing journalists to carriers
and amplifiers of sources’ messages. In contrast:
The interpretive style empowers journalists by giving them more control over the news
message. […] The descriptive style places the journalist in the role of an observer. The
interpretive style requires the journalist to act also as an analyst. The journalist is thus positioned
to give shape to the news in a way the descriptive style does not allow. (Patterson, 2000a: 250)
This is significant, as the rise of interpretive journalism has been linked to increasing
media negativity or cynicism towards politicians (Djerf-Pierre and Weibull, 2008;
Farnsworth and Lichter, 2011; Kovach and Rosenstiel, 2001), and to an increasing focus
on the framing of politics as a strategic game (Patterson, 1993, 2000a).
Benson and Hallin (2007), on the other hand, distinguish between four main journal-
istic functions: reporting current facts or statements, giving background information,
giving interpretation, and giving opinion. To them, ‘interpretation is a kind of empirical
discourse, but goes beyond current facts, setting or historical context to speculate on such
things as significance, outcomes and motives’, while ‘giving opinion’ refers to journal-
ists’ ‘exercise of judgment, either normative (what is good or bad) or empirical (what is
true or false)’ (2007: 32).
Djerf-Pierre and Weibull have offered a third conceptualization of interpretive jour-
nalism. According to them:
Interpretive journalism is characterized by four entwined features: critical expertise, speculation,
advocacy, and meta-journalism. Journalists appear as experts in studio interviews, or as
commentators interpreting political reality to their audience. Taking on the role of ombudsmen
of the public, journalists advocate the presumed interests and needs of the public/audience.
(2008: 209)
This conceptualization also stresses the critical component of interpretive journalism,
and the greater journalistic control over news content.
The same holds true for how Barnhurst and colleagues have conceptualized interpre-
tive reporting as part of the ‘new long journalism’ (Barnhurst, 2003; Barnhurst and Mutz,
1997; Steele and Barnhurst, 1996). In their view, the decline of event-centered reporting
and the rise of interpretive news are two sides of the same coin: ‘Across differences in
study context, time frame, and medium, closely related terms describe the shift from
descriptive to analytical journalism, from event-centered to interpretive reporting, or
from episodic to thematic coverage’ (Barnhurst and Mutz, 1997: 28). Related to this and
other conceptualizations is a greater emphasis on the journalistic voice, which may include
overt commentary, accompanied by, for example, shorter sound bites by politicians.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 147
However, in contrast to other scholars, Barnhurst and Mutz do not conceptualize inter-
pretive journalism as opposed to a focus on What, When, Where, and Who. While they
note that interpretive journalism places greater emphasis on Why than descriptive jour-
nalism, they also note that journalism may be interpretive in how it deals with What,
When, Where, and Who (1997: 29–30).
This brief review suggests that the concept of interpretive journalism is complex and
that there are several, albeit overlapping, conceptualizations. In many cases, most nota-
bly in Patterson’s analysis (1993), interpretive journalism carries negative connotations.
From our perspective, interpretive journalism should, however, be considered a non-
normative concept. If interpretive journalism focuses on journalistic interpretations and
analyses of current events, including overt commentary, these interpretations and analy-
ses can be well informed as well as uninformed, critical as well as uncritical, and provid-
ing context as well as distractions. This is, however, not a matter of interpretive journalism
as a concept. It is an empirical not a conceptual matter. Normative assessments should
hence be kept apart from the conceptualization of interpretive journalism.
Conceptual and empirical definitions
One of the influential scholars who have empirically investigated the concept of interpre-
tive journalism is Patterson (1993, 1996). According to him, facts and interpretations
have become increasingly intermingled in contemporary journalism. His study of how
election campaigns were covered by the New York Times shows that the share of pre-
dominantly interpretive news stories increased tenfold between 1960 and 1992. In his
study, news stories were coded as interpretive if they mainly focused on Why instead of
What (Patterson, 1993: 81–3). The same holds true of a follow-up study on the coverage
of the US Congress (Patterson, 1996).
Simultaneously, other scholars were also working on this subject but using differ-
ent terminology and empirical approaches. In Seeing the Newspaper, Barnhurst
(1994) suggests the concept of ‘new long journalism’ to describe trends and changes
in newspapers’ content. The new long journalism hypothesis suggests that newspaper
stories have become longer, that the emphasis in reporting has moved from event-
centered coverage to interpretive news analysis, and that the tone of reports has
become more negative.
In a follow-up study on the journalism of opinion in US presidential campaigns,
Steele and Barnhurst also work mainly with the concepts of the ‘new long journalism’
and the ‘journalism of opinion’ (1996). The goal of this study was to provide a frame
of descriptive statistics on the role of journalists in political coverage and to investi-
gate whether the more journalist-centered newscasts serve to enhance the journalists’
own prestige, influence and authority. The results show that by different measures,
journalists expanded their presence (1996: 189), and that network news has adopted a
more conversational tone (1996: 197). In order to provide context to this argument,
Steele and Barnhurst cite studies which describe speakers in a discussion as dominant
if they exhibit certain behaviors such as faster pace of communication, longer and
more proficient talks, or speaking more often, and claim (1996: 197) that journalists
exhibit precisely these traits.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
148 Journalism 13(2)
In another study, Barnhurst and Mutz (1997) investigated trends in reporting using
samples of news stories about crime, jobs and accidents. These results showed that what
used to be the main goal of journalism, ‘the report of events and their novelty to the
hearer’, gradually gave place to more analytical and interpretive reporting. These schol-
ars also conceptualized the new long journalism and interpretive reporting in relation –
albeit not opposed – to the five Ws of journalism:
In general, we expected to find a greater emphasis on news analysis and the ‘Why’ part of the
traditional five Ws formula. ‘Where’ might incorporate a broader context or broader range of
locales than would event-centered coverage; ‘When’ might show greater emphasis on times
other than the present; ‘What’ might schedule a larger number of events in single stories; and
‘Who’ might deemphasize individuals, so that people become identified less often by name and
more often by demography or group affiliation.
In the European context, there are also several empirical studies on the concept of
interpretive journalism. De Vreese (2001), for example, investigated how Dutch public
television covered the 1998 national election and the 1999 European election. The frame-
work for this study was based on the concepts of sacerdotal vs pragmatic and agenda-
sending vs agenda-setting. With respect to sacerdotal and pragmatic approaches the
distinction was first made by Blumler and Gurevitch (1975), and refers to the inherent
importance attributed to politics by the media. The purpose of distinguishing ‘agenda-
sending’ from ‘agenda-setting’ was to determine the weight the journalist attached to the
campaign agendas:
Agenda-sending refers to the conviction that the role of newsmakers is to send and cover the
agendas of the political parties whereas agenda-setting refers to news-makers’ attempt to set
their own agenda, based on audience preferences, editorial policies and in house expertise. The
agenda-setting notion of newsmakers is that they play an active role in setting the campaign agenda
by turning to themes and issues, which are not already on the established political agenda.
(1975: 160–1)
Interpretive journalism was thus conceptualized as journalism driven by the journal-
ists rather than their sources.
A more explicit focus on journalistic style can be found in a series of cross-national
studies conducted by Strömbäck and colleagues (Strömbäck and Aalberg, 2008;
Strömbäck and Dimitrova, 2006; Strömbäck and Luengo, 2008; Strömbäck and Shehata,
2007). In these studies, the ‘dominant journalistic style’ was investigated, building on the
distinction between a mainly descriptive and a mainly interpretive journalistic style. In
these studies, the journalistic style was conceptualized as descriptive when the news
story mainly told what had happened in a rather straightforward style, and as mainly
interpretive when the journalist analyzed, evaluated or explained a situation while also
describing it. Strömbäck (2008) made use of the same conceptualization in a study com-
paring Swedish election news between 1998 and 2006, in which he also investigated the
prominence of so-called news analyses, that is, commentary pieces written by journalists
with some kind of label (such as ‘analysis’) that signals a more subjective approach.
Interpretive journalism is conceptualized as a greater journalistic control over the news,
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 149
a greater focus on explanations and analyses by, and the voice of, journalists, and as
going beyond a descriptive style of journalism.
In another comparative study, Benson and Hallin (2007) investigated the extent to
which journalists in US and French elite press in the 1960s and 1990s were ‘giving inter-
pretation and giving opinion’, as opposed to reporting facts or giving background infor-
mation. Interpretive journalism was conceptualized as going ‘beyond current facts,
setting or historical context to speculate on such things as significance, outcomes and
motives’. Giving opinion was conceptualized as journalists’ ‘exercise of judgment, either
normative (what is good or bad) or empirical (what is true or false)’ (2007: 32).
Wilke and Reinemann (2001; see also Reinemann and Wilke, 2007) investigated
trends in campaign coverage in Germany between 1949 and 1998. As part of this study,
they investigated whether there was ‘a tendency towards a more interpretive character
of coverage’ (2001: 293). Conceptually, they link interpretive journalism with the
degree to which the news coverage is mediated and journalist centered (2001: 295).
More specifically, an interpretive journalistic style was conceptualized as a journalistic
style focused more on the journalists and their analyses and opinions, and less on the
candidates and their statements. This is similar to a study of Semetko and Schoenbach
(2003) that focused on the coverage of German elections between 1990 and 2002 in the
newspaper Bild.
Aside from the studies above, there are others that refer to an interpretive journalistic
style, and some that investigate it but without providing a clear conceptualization of what
entails an interpretive journalistic style. There are also studies using similar indicators as
those reviewed above – for example, the length of politicians’ sound bites – without linking
it conceptually to interpretive journalism (Farnsworth and Lichter, 2011; Hallin, 1992). As
the main focus of these studies is not on interpretive journalism per se, or as they do not
provide clear conceptualizations of interpretive journalism, we have left them out.
To sum up: although there are differences related to, for example, whether interpretive
journalism is conceptualized as opposed to or just related to the five Ws of journalism,
there are also similarities in how interpretive journalism has been conceptualized. In
most cases interpretive journalism is conceptualized as opposed to or going beyond
descriptive, fact-based or source-driven journalism; entailing greater journalistic control
over news content and a more prominent journalistic voice, which may include overt
commentary by journalists; entailing a stronger emphasis on the theme chosen by the
journalist; and as entailing a focus on the meaning and Why of news rather than on the
Who, What, Where and When.
Despite these conceptual similarities, the breadth and depth of the conceptualizations
and the choice of empirical indicators differ across studies. In the next section, we ana-
lyze how interpretive journalism has been operationalized in more detail.
Operationalizations of interpretive journalism
When reviewing how interpretive journalism has been operationalized, one key distinc-
tion is between content and format-related operationalizations. Most studies focus on the
content and characteristics of news, but there are some that also include format-related
measures.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
150 Journalism 13(2)
To study the journalistic style Patterson (2000b: 26), for example, distinguishes
between a descriptive and an interpretive/analytical/evaluative style, coded as either the
one or the other. The descriptive style is defined as ‘tells “what” happened in a rather
straightforward, descriptive way’, while the interpretive/analytical/evaluative style is
defined as ‘analyzes, evaluates, or explains a situation while also describing parts of it’.
In this study the content rather than the format is at the center.
With the goal of illustrating how the new long journalism trend manifested itself in
newspaper content, Barnhurst and Mutz (1997) investigated the specific components of
the new long journalism. For each story, coders answered several questions about indi-
vidual actors or groups in the text, the number of events reported, when these events took
place and where the events happened. They also identified whether the story reported
any information about background context, implications, interpretations, or recommen-
dations, and whether the story reported any causes, general problems, collective social
issues, or themes. On a 10-point scale, they rated the emphasis from highly specific,
event-centered coverage (1) to very general news analysis (10). This is highly similar to
the approach taken by Barnhurst (2003) in his study of how US public radio covered
elections between 1980 and 2000, and in both cases the operationalizations target the
content rather than the format of news.
Focusing on the idea that distinct types of journalistic discourse establish different
kinds of relationships between press and citizenry, and considering that the political
voice of the journalist increases in prominence from the lowest to the highest category
(opinion), Benson and Hallin (2007) coded each paragraph according to its primary func-
tion: reporting current facts or statements, giving background information, giving inter-
pretation, or giving opinion. Their sample included articles on domestic politics in US
and French elite press, randomly selected from editions from the mid 1960s and from the
mid 1990s.
A different type of approach was taken by De Vreese in his investigation on how
Dutch public broadcasting (2001) covered the 1998 national and the 1999 European
elections. The analysis was based on newsroom observation, interviews and content
analysis of election news coverage. The indicator most directly related to the journalistic
style, and investigated through content analysis, was whether stories were adapted from
the campaigns’ agendas, originated from the campaigns’ agendas but with a significant
and identifiable spin by NOS Journaal, or initiated purely by NOS Journaal.
Strömbäck and colleagues used content analysis to compare how major newspapers
covered national elections in countries such as Sweden on the one hand, and Britain,
Norway, Spain and the USA on the other. The operationalization of interpretive journal-
ism was consistent across these studies, and determined based on the dominant journal-
istic style. The content was thus decisive, not the format. Briefly, coders were instructed
to code descriptive if the news story tells what happened in a straightforward way, and
interpretive if ‘analyzes, evaluates, or explains a situation while also describing aspects
of it’. The journalistic use of overt commentary and value-laden terms was used as indi-
cators of an interpretive style. Lengauer et al. (2007; Plasser and Lengauer, 2009)
followed a similar approach, distinguishing between a mainly interpretive and a mainly
descriptive journalistic style, although they also used the intermediate category ‘ambiva-
lent’. The same operationalization was used in research on Swedish election news
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 151
(Strömbäck, 2008; Strömbäck and Nord, 2006), in addition to a variable measuring the
format of the news story, distinguishing between ‘straight news’ and ‘news analysis’.
The operationalization of ‘news analysis’ was ‘news stories that the media explicitly give
a label that sends the signal to readers/listeners that journalistic interpretations are to be
expected. In broadcast news, that includes segments where one journalist interviews
another journalist.’
Also considering both content and format in their operationalization, Wilke and
Reinemann’s study (2001) covered the media coverage of German national elections
between 1949 and 1998. Operationally, the indicators used to analyze whether campaign
coverage had become more interpretive and whether de-authentication had taken place
were: format (hard news, opinion, feature, other); and length of the statements of the
candidates (in lines). Similarly, Semetko and Schoenbach (2003) used two operational
indicators relevant with respect to interpretive journalism in their study of how the Bild
covered German elections between 1990 and 2002: the type of news story (commentary
or non-commentary); and the way political actors were evaluated (positively or nega-
tively), with a particular focus on negative evaluations.
Salgado (2007, 2010) adopted a different approach to investigate how interpretive the
election coverage is in Portugal. Besides coding news according to the format, she also
coded news stories according to whether they were predominantly retrospective or pro-
spective. In cases where they were coded as predominantly prospective, focusing on
future possibilities or interpretations about the future or the consequences of events or
facts, this was taken as an indicator of interpretive journalism.
Partly following Hallin’s framework (1992) and focused on the content, Steele and
Barnhurst (1996) used several operational indicators of the ‘new long journalism’ and
interpretive journalism. More specifically, they investigated: How long did each journal-
ist’s speech go on? How often did journalists speak in each report? How much time in
total did journalists spend talking in each report? To understand if more journalistic pres-
ence translated into more influence in their political reporting, the authors classified the
journalists’ sound bites in different communication tasks. The purpose of this was to
classify not what facts or content were discussed but how journalists spoke. The results
showed the highest rate of activity in ‘giving opinion’ for each year with the second most
frequent incidence being ‘giving information’. Steele and Barnhurst (1996) also meas-
ured the dynamic tendencies of speakers. With the objective of determining who controls
the report, they compared the relative contribution of two tasks: giving and asking for
information. These indexes showed a clear pattern: journalists’ increased control over the
news, and in giving opinions and expressing judgments, while asking fewer questions in
the process of reporting on elections.
This review of how interpretive journalism has been operationalized shows that dif-
ferent operationalizations have been used. Some use only a few variables as indicators of
interpretive journalism, while others use several variables. Some make use of direct
measures, while others make use of indirect measures. Some measure the journalistic
style on the basis of the dominant journalistic style, while others do not. And while most
focus on the content of news, some focus on the format or combine content and format-
related operationalizations. This should be kept in mind when the findings across studies
are assessed.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
152 Journalism 13(2)
Key findings: The rise of interpretive journalism?
Most of the empirical studies about interpretive journalism that were reviewed for this
article found at least some evidence of an increasingly prominent and interpretive role of
journalists in the news coverage, not only with respect to election or political news, but
also with respect to current affairs coverage. Most authors link interpretive journalism
with other trends in journalistic coverage and, in some cases, present explanations on
how this form of journalism has other implications, such as less neutral and more nega-
tive coverage.
One of the most prominent critics of the trend towards interpretive journalism is
Patterson (1980, 1993, 2000a). His research shows that the share of news stories on the
New York Times’ front pages that are predominantly interpretive as opposed to descriptive
increased from 8 percent in 1960 to 80 percent in 1992. The consequence of this rise of
interpretive journalism, he claims, is that news stories have become increasingly negative,
less focused on issues, and more focused on the political game and strategies (1993: 240).
Steele and Barnhurst (1996) also discuss the divergent opinions on whether journal-
ists ought to be transmitting their judgments. Aside from these normative discussions,
comparing data from 1968 and 1988, they found that ‘when journalists spoke at the end
of a report, they gave factual information much less often, instead making judgments
about the election events grew to almost 90 percent of these final remarks. Journalists
made their opinions the last word, the final say’ (1996: 22). Barnhurst and Mutz (1997)
also showed that reporting moved in the general direction of analysis and interpretation.
More specifically, their results show that over time, newspaper stories grew longer, with
more interpretation, less focus on particular geographic locales and more focus on
broader regions, and with a stronger focus on thematic summaries of events and state-
ments at the expense of episodic stories on particular events or statements by sources.
Similar results were found in Barnhurst’s study of National Public Radio (2003: 21):
While political reports grew longer, following the trend at newspapers, sound bites shrank,
following the trend for television. Citizens spoke more often but also more briefly and
emotionally and had less to say about politics than about other matters. Politicians took a
double hit, speaking less often and in shorter sound bites. […] The packaging and pacing of
speech imply a fundamental change in the professional role of journalists, who put themselves
forward as interpreters of the political world.
Journalists are ‘makers of meaning’, and not mere ‘mechanics of event transmission’.
They also tended to act more often as expert sources, interpreting politics by looking at
the politicians’ hidden agendas and by engaging in a critique of policies.
Benson and Hallin (2007) confirmed that, during the 1990s, the French press offered
both more interpretation (12%) and opinion (5.9%) than the US press (5.4% and 0.7%,
respectively). Comparing the 1960s and the 1990s, they also found that the French press
became slightly less interpretive and opinionated (19.8% to 17.9%) while the US press
became slightly more so (4.7% to 6.1%).
In the European context, De Vreese (2001) showed that election news in the NOS
Journaal in 1998 was more analytical and had a more pronounced journalistic spin on
election stories compared to earlier elections. The time devoted to politicians decreased
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 153
while the time devoted to journalistic commentary increased, with the journalistic role as
a commentator more prominent than before. Wilke and Reinemann (2001) also found a
trend towards more interpretive news coverage in their study of German election news
between 1949 and 1991. The dominance of hard news decreased significantly over time
while opinion and interpreting formats became more and more important. The share of
hard news, which accounted for more than 80 percent of all articles between 1949 and
1961, fell below 60 percent in the 1990s. On the other hand, the share of opinion-related
and interpreting formats represented about one in 10 articles during the first elections,
but doubled in 1965 and remained at that level in the following elections. In a follow-up
study, Reinemann and Wilke (2007: 108) confirmed that these trends continued in the
2002 and 2005 German elections. Not least, the introduction of televised debates caused
changes in the character of media coverage about the German chancellor candidates: the
number of evaluative statements exploded and the criteria by which the candidates were
evaluated changed, increasing the level of personalization.
Also in the German context, Semetko and Schoenbach (2003) showed that the
number of commentary journalistic pieces increased over time, further confirming a
trend towards more interpretive journalism. Their results also suggest that this trend
contributed to an increased volume of news stories that contained negative evalua-
tions of political actors.
Turning to the studies by Strömbäck and colleagues, most of these focus on compari-
sons across countries and not across time. While they do not show whether or not interpre-
tive journalism has become more common, they suggest that there are differences across
countries (Strömbäck and Aalberg, 2008; Strömbäck and Dimitrova, 2006; Strömbäck and
Luengo, 2008; Strömbäck and Shehata, 2007). Research comparing television news in the
USA, Germany, Austria and Italy similarly found differences across countries (Plasser and
Lengauer, 2009). Studies on interpretive journalism in Sweden has shown that close to 10
percent of all news stories are explicit ‘news analyses’ while around 40 percent of all sto-
ries are predominantly interpretive. However, during the last 10 years there does not appear
to be any trend towards more interpretive journalism (Strömbäck, 2008). The same appears
to hold true for Austria (Lengauer et al., 2007).
In the Portuguese case, finally, Salgado (2007, 2010) found that approximately 20
percent of all election news stories in daily newspapers were prospective – that is, dealt
with interpretations of future possibilities or consequences of current events or facts.
Taken together, this review suggests that (a) there are different although overlapping
conceptualizations of interpretive journalism, (b) several different operationalizations of
interpretive journalism have been used, (c) but that despite this, most studies show that
interpretive journalism has become more – or is – common, (d) although there are differ-
ences across countries.
However, the review also suggests that it is difficult to compare the results, not least
because different operationalizations have been used. Often, the exact variables and cod-
ing instructions are not provided, making it even more difficult to compare the studies
and results. From the perspective of conceptual clarity as well as research comparability
and cumulativity, this is problematic. The next sections thus suggest how interpretive
journalism should be conceptualized to increase conceptual clarity, and operationalized
to increase comparability and cumulativity.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
154 Journalism 13(2)
Towards conceptual clarity
To be meaningful, any concept should ideally have precise boundaries, delineating it
from other, related concepts. To the extent possible, ambiguities should be avoided. The
conceptual definition should furthermore be grounded in established scholarly under-
standings, and in cases where similar terms are used in everyday life, resonate with popu-
lar understandings. In addition, the concepts as such should not be normative. What is
and what should be are two separate questions, and should not be conflated in how the
concepts are defined.
As shown by this review, most scholars conceive interpretive journalism as opposed
to or going beyond descriptive and fact-based journalism. Descriptive journalism is
conceived as focusing mainly on What, Where, When and Who, and on verifiable
facts. In contrast, interpretive journalism focuses mainly on the Why, and on the mean-
ing or context of verifiable facts. However, the other Ws can also be the subject of an
interpretive angle, as happens with When and What in cases of speculation about the
future or the past.
In a sense, interpretive journalism aims to find out the truth behind the verifiable
facts. One reason for this is the disbelief in value-free facts or in the notion that facts
speak for themselves. Another reason is that descriptive journalism lends the objects of
news reports the power to define what the news is about. For journalists believing that
journalism should function as a critical watchdog of those in power, this is anathema.
While journalists frequently cannot overtly question what political news sources say,
without being perceived of being biased, they can decide on the thematic framework
within which the actions and words of political news sources are placed, and analyze the
meaning of the What, Where, When and Who. They can explain and evaluate what the
news is about by contextualizing the news and by linking different events or statements
to each other within a specific, thematic framework. A third reason is that facts may need
context to carry meaning, and interpretive journalism can provide such context.
How, then, should interpretive journalism be conceptualized? Based on this review
and as a synthesis of how it has been conceptualized in previous research, we propose the
following definition of interpretive journalism:
Interpretive journalism is opposed to or going beyond descriptive, fact-focused and source-
driven journalism. On the story-level of analysis, interpretive journalism is characterized by a
prominent journalistic voice; and by journalistic explanations, evaluations, contextualizations,
or speculations going beyond verifiable facts or statements by sources. It may, but does not
have to, also be characterized by a theme chosen by the journalist, use of value-laden terms, or
overt commentary.
When interpretive journalism includes the use of value-laden terms or overt commen-
tary it comes very close to what Benson and Hallin (2007: 32) conceptualize as ‘journal-
ism of opinion’, where the journalist explicitly makes use of his or her own judgment in
relation to what is good or bad, true or false. This suggests that is important to acknowl-
edge that there are different types of interpretive journalism. While in some cases, a
public service orientation guides the journalistic work, in others the journalist’s personal
style is more accentuated, or the news story is based on a more speculative approach.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 155
Hence, it may be possible to identify different kinds of interpretive journalism such as
‘public service or public responsibility journalism’, ‘commentary journalism’, or ‘criti-
cal and adversarial journalism’. As the assessment of different kinds of interpretive jour-
nalism may vary, this is another argument why it is important that the concept per se is
not normative.
Towards increasing comparability and cumulativity
Based on the conceptualization of interpretive journalism as suggested, the question is
how to operationalize the concept. Before providing our answer to this question, three
caveats should be noted. First, our suggested operationalizations are developed for quan-
titative content analyses. Second, the unit of analysis is conceived of as full news stories.
Third, the operationalizations should work for both print and broadcast media.
As in all quantitative content analyses, it is important that the operationalizations
increase the likelihood of high intercoder reliability. The operationalizations should
therefore, to the extent possible, target the manifest content. Second, the operationaliza-
tions should be as unambiguous as possible. Third, to increase the overall reliability and
validity of the findings, multiple indicators are preferable.
The most obvious indicator of whether the journalistic style is interpretive or descrip-
tive refers to the format of the news story, where several scholars make a distinction
between categories such as hard or straight news, commentary, opinion or news analysis,
and feature. Since the terminologies may vary both across media and across countries, to
capture the share of news stories that explicitly signal the presence of interpretive jour-
nalism, we suggest the following variable: (1) Does the news story through labeling
explicitly signal to audiences that they can expect interpretive journalism?
Considering that a descriptive news story tells what happened and an interpretive
news story contains journalistic explanations, contextualizations, evaluations, or specu-
lations, we also suggest a variable addressing the overall salience or amount of journal-
istic interpretation or explanation: (2) What is the overall amount of journalistic
interpretations and explanations: a) Primary amount; b) Secondary amount; c) Peripheral
amount; d) No amount?
The downside with variables such as this is that it may be difficult to distinguish
between different amounts of journalistic interpretation and explanation, and hence to
achieve strong intercoder reliability. To supplement this variable and to address the fact
that there might be different forms of interpretive journalism, we suggest a set of varia-
bles coded on a present-absent basis, which could then be used to construct an additive
index of interpretive journalism: (3) Does the news story include journalistic explana-
tions or interpretations of the reasons behind events or actions? (4) Does the news story
include journalistic speculations about future consequences of events? (5) Does the news
story include a journalistic contextualization of events or actions? (6) Does the journalist
include overt commentary when covering events and actions?
Reviewing previous research, perhaps one of the most important indicators of inter-
pretive journalism is the prominence of the journalistic voice – or, conversely, the lack of
prominence of politicians’ voices. Thus, several studies investigate the length of politicians’
sound bites, inferring that the decreasing length of these sound bites suggests more inter-
pretive journalism (Farnsworth and Lichter, 2011; Hallin, 1992). Other studies target
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
156 Journalism 13(2)
journalists’ sound bites directly (Steele and Barnhurst, 1996). The advantage of such
variables is not only that they have strong conceptual linkages, but also that they often
yield high intercoder reliability. The downside is that they need to be adjusted for content
analyses of print media. A better approach that would work for both print and broadcast
media would be to measure the share of the news story taken by the spoken or written
words of the journalist, excluding statements from sources. Hence, to measure the promi-
nence of the journalistic voice, we suggest the variable: (7) What is the share of the news
story taken by journalists’ words?
In Table 1, we summarize the variables we suggest should be used to investigate the
extent to which journalism is interpretive. For details and coding instructions please see
the Appendix.
These seven variables may not cover all facets of interpretive journalism, and other
variables are certainly conceivable. Nevertheless, we believe they cover the conceptually
most important aspects of interpretive journalism, and that, taken together, studies using
these operationalizations should yield both valid and reliable results regarding the extent
to which news journalism is, or has become, interpretive.
Discussion
Although there may be many who are critical towards journalism practice, there is rather
wide consensus that one of journalism’s most important democratic functions is to pro-
vide people with the kind of information they need to be free and self-governing (Kovach
and Rosenstiel, 2007). The consensus quickly breaks down, however, as soon as the
discussion focuses on what, exactly, constitutes the kind of information that people need.
For those who defend interpretive journalism, the answer may be that people need ana-
lytical and contextualized information – information that makes it easier for people to
make sense of facts and what is happening. For those who are critical towards
Table 1. Variables for measuring the extent to which journalism is interpretive
Variable Values
1 Does the news story through labeling explicitly signal to
audiences that they may expect interpretive journalism?
Yes/No
2 What is the overall amount of journalistic
interpretations and explanations?
Primary/Secondary/Peripheral/
No amount
3 Does the news story include journalistic explanations
or interpretations of the reasons behind events or
actions?
Yes/No
4 Does the news story include journalistic speculations
about future consequences of events?
Yes/No
5 Does the news story include a journalistic
contextualization of events or actions?
Yes/No
6 Does the journalist include overt commentary when
covering events and actions?
Yes/No
7 What is the share of the news story taken by
journalists’ words?
Percentage
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 157
interpretive journalism, the answer is rather that people need information unfiltered by
journalists – information that comes as directly as possible from the original sources.
Having different perspectives on the desirability of interpretive journalism is of course
legitimate. At the same time, what is and what should be are two different issues that
should always be kept separate.
To this end, this article has suggested not only a non-normative conceptualization of
interpretive journalism, but also variables that can be used to investigate the extent to
which news journalism is interpretive. We hope that these will appeal to scholars inter-
ested in investigating interpretive journalism, thereby yielding results that are more com-
parable across studies and hence increasing research cumulativity.
In addition, our hope is that these will free interpretive journalism from any normative
connotations and thereby help to open up new research that does not assume but that empiri-
cally investigates the extent to which there are correlations between interpretive journalism
on the one hand, and other features of journalism such as the framing of politics, or the
effects on audiences, on the other. It may very well be a fact that interpretive journalism is
linked to particular ways of framing politics or to media negativity, to take just two exam-
ples, or that interpretive journalism has adverse effects on people’s political attitudes. It may
also be the case that interpretive journalism is linked to more in-depth coverage of issues and
that it contributes positively to people’s political attitudes. There might also be differences
between interpretive journalism in different media or in different countries in terms of ana-
lytical styles, linkages with other features of journalism, and effects on audiences.
These are important questions to investigate, but they can only be investigated if the
concept of interpretive journalism is non-normative and decoupled from other concepts.
References
Barnhurst KG (1994) Seeing the Newspaper. New York: St Martin’s Press.
Barnhurst KG (2003) The makers of meaning: National public radio and the new long journalism,
1980–2000. Political Communication 20(1): 1–22.
Barnhurst KG and Mutz D (1997) American journalism and the decline in event-centered reporting.
Journal of Communication 47(4): 27–53.
Benson R and Hallin D (2007) How states, markets and globalization shape the news: The French
and US national press, 1965–97. European Journal of Communication 22: 27–48.
Blumler JG and Gurevitch M (1975) Towards a comparative framework for political commu-
nication research. In: Chaffee SH (ed.) Political Communication. Issues and Strategies for
Research. Beverly Hills, CA: SAGE, 165–193.
De Vreese CH (2001) Election coverage – new directions for public broadcasting. The Netherlands
and beyond. European Journal of Communication 16(2): 155–180.
Djerf-Pierre M and Weibull L (2008) From public educator to interpreting ombudsman. Regimes
of political journalism in Swedish public service broadcasting, 1925–2005. In: Strömbäck
J, Ørsten M and Aalberg T (eds) Communicating Politics. Political Communication in the
Nordic Countries. Gothenburg: Nordicom, 195–214.
Farnsworth SJ and Lichter SR (2011) The Nightly News Nightmare. Media Coverage of US
Presidential Elections,1988–2008, 3rd edn. Lanham, MD: Rowman & Littlefield.
Hallin DC (1992) Sound bite news: Television coverage of elections, 1968–1988. Journal of
Communication 42(2): 5–24.
Kovach B and Rosenstiel T (2001) The Elements of Journalism. New York: Crown.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
158 Journalism 13(2)
Kovach B and Rosenstiel T (2007) The Elements of Journalism. What Newspeople Should Know
and the Public Should Expect, 2nd edition. New York: Crown.
Lengauer G, Pallaver G and Pig C (2007) Redaktionelle Politikvermittlung in Österreichischen
Wahlkämpfen, 1999–2006. In: Plasser F (ed.) Wechselwahlen. Analysen zur Nationalratswahl
2006. Vienna: Facultas, 103–151.
Patterson TE (1980) The Mass Media Election. How Americans Choose Their President. New York: Praeger.
Patterson TE (1993) Out of Order. New York: Knopf.
Patterson TE (1996) Bad news, bad governance. The Annals of the American Academy of Political
and Social Science 546(1): 97–108.
Patterson TE (2000a) The United States: News in a free-market society. In: Gunther R and Mughan
A (eds) Democracy and the Media. A Comparative Perspective. New York: Cambridge
University Press, 241–265.
Patterson TE (2000b) Doing Well and Doing Good. How Soft News and Critical Journalism
Are Shrinking the News Audience and Weakening Democracy – And What News Outlets
Can Do About It. Cambridge, MA: Joan Shorenstein Center on the Press, Politics and
Public Policy.
Plasser F and Lengauer G (2009) Wie ‘amerikanisch’ sind europäische Fernsehwahlkämpfe?
In: Kaspar H, Schoen H, Schumann S and Winkler JR (eds) Politik–Wissenschaft– Medien.
Festschrift für Jurgen W Falter zum 65. Geburtstag. Wiesbaden: VS Verlag, 323–346.
Reinemann C and Wilke J (2007) It’s the debates, stupid! How the introduction of televised
debates changed the portrayal of chancellor candidates in the German Press, 1949–2005. The
Harvard International Journal of Press/Politics 12(4): 12–92.
Salgado S (2007) As Presidenciais de 2006: reflexões sobre a interpretação da política nos jornais
[The 2006 Portuguese presidential election: Some thoughts on the journalistic interpretation of
politics]. Studies in Communication Review 1: 232–249.
Salgado S (2010) Os Candidatos Presidenciais: Construção de Imagens e Discursos nos Media
[Portuguese presidential candidates: Construction of images and discourses in the media].
Coimbra: MinervaCoimbra.
Schudson M (1978) Discovering the News. A Social History of American Newspapers. New York:
Basic Books.
Semetko HA and Schoenbach K (2003) News and elections. German Bundestag campaigns in the
Bild, 1990–2002. The Harvard International Journal of Press/Politics 8(3): 54–69.
Steele CA and Barnhurst KG (1996) The journalism of opinion: Network coverage in US presiden-
tial campaigns, 1968–1988. Critical Studies in Mass Communication 13(3): 187–209.
Strömbäck J (2008) Swedish election news coverage: Towards increasing mediatization. In:
Strömbäck J and Kaid LL (eds) The Handbook of Election News Coverage around the World.
New York: Routledge, 160–174.
Strömbäck J and Aalberg T (2008) Election news coverage in democratic corporatist countries:
A comparative study of Sweden and Norway. Scandinavian Political Studies 31(1): 91–106.
Strömbäck J and Dimitrova DV (2006) Political and media systems matter: A comparison of elec-
tion news coverage in Sweden and the United States. Harvard International Journal of Press/
Politics 11(4): 131–147.
Strömbäck J and Luengo ÓG (2008) Polarized pluralist and democratic corporatist models. A com-
parison of election news coverage in Spain and Sweden. The International Communication
Gazette 70(6): 547–562.
Strömbäck J and Nord LW (2006) Do politicians lead the tango? A study of the relationship
between Swedish journalists and their political sources in the context of election campaigns.
European Journal of Communication 21(2): 147–164.
Strömbäck J and Shehata A (2007) Structural biases in British and Swedish election news cover-
age. Journalism Studies 8(5): 798–812.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 159
Weaver DH and Willnat L (eds) (2012) The Global Journalist in the 21st Century. New York:
Routledge.
Wilke J and Reinemann C (2001) Do the candidates matter? Long-term trends of campaign cover-
age – a study of the German Press since 1949. European Journal of Communication 16(3):
291–314.
Variable Coding instructions
(1) Does the news story
through labeling explicitly
signal to audiences that they
may expect interpretive
journalism?
This variable has two codes: 0 = no, 1 = yes.
Coders should type 0 if the news story does
not have any labels or visual markers signaling
to readers that they may expect interpretive
journalism, and/or if the news story does not
include any segments where the news anchor
interviews a journalist or news commentator or
senior correspondent for the news program; and
1 if the news story is presented as ‘news analysis’,
‘analysis’, ‘commentary’ or other functionally
equivalent labels or visual markers signaling that
readers may expect interpretive journalism, and/
or if the news story includes as segment where
the news anchor interviews a journalist, news
commentator or senior correspondent working
for the news program.
(2) What is the overall amount
of journalistic interpretations
and explanations?
This variable addresses the level of salience or
amount of journalistic interpretation within the
news story and has four codes:
0 = No journalistic interpretations
1 = Peripheral amount of journalistic interpretations
2 = Secondary amount of journalistic interpretations
3 = Primary amount of journalistic interpretations.
‘Journalistic explanations or interpretations’ refer
to explanations and interpretations put forward
by the journalist without explicit support from
verifiable facts or statements by news sources.
Primary amount is meant to classify news stories
in which interpretation is the main purpose; in
other words, if the news story is mainly focused
on journalists’ interpretations or explanations,
including their statements and/or opinions.
Secondary amount refers to those news stories
where a balance between facts/events and
journalists’ interpretations and explanations is
provided. Statements and/or opinions are provided
as an illustration of a fact/idea; Peripheral amount
suggests that there is mainly a description and
a small portion of journalists’ interpretation or
explanations is included. If there are no journalistic
interpretations, coders should type 0.
Appendix
(Continued)
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
160 Journalism 13(2)
Variable Coding instructions
(3) Does the news story include
journalistic explanations or
interpretations of the reasons
behind events or actions?
This variable has two codes: 0 = no, 1 = yes.
‘Journalistic explanations or interpretations’ refer
to explanations and interpretations put forward
by the journalist without explicit support from
verifiable facts or statements by news sources.
The coder should type 0 if the news story does
not include any explanations or interpretations
targeting the reasons behind events or actions
by the journalist covering the story, and 1 if the
news story includes at least one explanation
or interpretation of the reasons behind events
or actions by the journalist covering the story.
Particular attention should be paid to whether
the journalist in the news story attempts to
provide an answer to why something has
happened or why a social or political actor has
acted or said what he or she did or said, that is,
retrospective speculations.
(4) Does the news story include
journalistic speculations
about future consequences of
events?
This variable has two codes: 0 = no, 1 = yes. The
coder should type 0 if the news story does not
include any journalistic speculations about future
consequences of events or actions; and 1 if the
news story includes at least one speculation by
the journalist about future consequences of events
or happenings. Any speculation by a journalist
about things that may happen in the future as a
consequence of something should be counted.
(5) Does the news story
include a journalistic
contextualization of events
or actions?
This variable has two codes: 0 = no, 1 = yes.
The coder should type 0 if the news story does
not include any attempts by the journalist to add
context to events or actions covered in a news
story; and 1 if the news story includes at least one
attempt by the journalist to add context to the
event or action that is covered in a news story.
(6) Does the journalist include
overt commentary when
covering events and actions?
This variable has two codes: 0 = no, 1 = yes.
The coder should type 0 if the news story does
not include any overt commentary from the
journalist covering a story; and 1 if the news
story includes at least one overt commentary,
including clearly value-laden terms, from the
journalist covering the story. Value-laden terms
refer to terms that are clearly subjective or carry
connotations that cannot be considered neutral.
Overt commentary refers to expressions of
opinions by the journalist, for example, whether
something is good or bad, true or false, without
support from verifiable facts.
Appendix. (Continued)
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
Salgado and Strömbäck 161
Variable Coding instructions
(7) What is the share of
the news story taken by
journalists’ words?
Coders should measure the share of the news
story taken by journalists’ words. For broadcast
media, the share of the news story taken by
journalists’ words is measured by speaking time
taken by journalists – the sum of journalists’ sound
bites divided by the total length of the news story.
For print media, the share of the news story taken
by journalists’ words is measured by the number of
words taken by journalists – the sum of all words
written by the journalist – thus excluding any
quotes from sources – divided by the total number
of words in the news story.
Appendix. (Continued)
Biographical notes
Susana Salgado is Professor of Strategic and Political Communication and Post-Doctoral Researcher
in Media and Democracy at the Foundation for Science and Technology, Portugal.
Jesper Strömbäck is Professor in Media and Communication, and Professor and Chair in
Journalism, at Mid Sweden University. He is also Research Director at the Centre for Political
Communication Research at Mid Sweden University.
at Mittuniversitetet on January 11, 2012jou.sagepub.comDownloaded from
