|
-- Setting the Proximity Frame
Setting the Proximity Frame:
Distance As an Affective Attribute in Reporting Terrorism Events
by
Kenneth C. Killebrew, Ph.D.
4202 E. Fowler Ave. CIS 1040
University of South Florida
Tampa, FL 33620
(813) 974-6795 - office
e-mail: [log in to unmask]
word count: 7,467
A paper presented to the Communication Theory and Methodology Division of the
Association for Education in Journalism and Mass Communications for presentation
at the annual meeting in New Orleans August 1999.
Setting the Proximity Frame:
Distance As an Affective Attribute in Reporting Terrorism Events
Introduction
Proximity is a long-held axiom of news. Along with the overwhelming or
commanding importance or prominence of a story, how close subjects are to a
story when it takes place is an important characteristic in whether the story is
considered news. A house fire in Polo, Illinois is generally of little interest
to the citizens of Atlanta, Georgia. But if that fire were a few miles down the
road in Dixon, Illinois and destroyed a childhood home of former President
Ronald Reagan, the nation would be told about the incident. Prominence now
takes the front row seat in the story.
Yet, should we now dismiss proximity as a factor in this story? How would
the reporting of the story differ from Dixon to Atlanta? If distance can
dictate whether a story is important enough to be considered news; might it also
be capable of shaping not only the scope, but the tenor of the news coverage?
Shaping the description of news events is a notion best conceptualized in
relationship to the framing process. Does this mean then that proximity is
potentially a contributing factor to our understanding of overarching framing
concepts? Ghanem suggests that "one of the weaknesses of most framing studies
is that the attributes of the issue or topic are not generalizable across
issues." Evaluating proximity, a basic tenet of news, should provide some
direction in the attempt to overcome this weakness. This research specifically
examines the value of proximity in the framing dimension of affective
attributes.[1]
Few studies have attempted to foster the notion of proximity outside of the
routine borders of the regional newspaper or television station. Fewer studies
still have examined distance in both a cross-national and framing manner. This
research then begins the process of moving proximity to a generalizable
attribute of framing.
English-speaking nations have both similarities and differences in their
cultural lives, and how they view the news. Ensuring, therefore, that the study
remains one of proximity differences rather than cultural differences is
difficult, but not impossible. To minimize those difficulties, the research
should identify the affective attributes of proximity through a common,
relatively uncluttered and well-defined manner.[2]
The selection of distant, but similar countries for a test of the strength
of proximity as an affective attribute frame in news coverage seems interesting.
Selecting the United States and Great Britain provides direct access to similar
print media reporting styles while establishing substantial physical distance in
order to test this proximity notion. A two-nation study on dynamic and
generally tragic events provides an understanding of the cross-national
influences of this framing.[3]
This research examines the portrayal of two specific critical events by
four elite newspapers, two in each nation, based on each newspaper's proximity
to the event. Both acts in this research are instances of political violence.
Events of a magnitude sufficient to warrant comparison between nations are
sometimes difficult to ascertain. However, in instances of political violence,
generally termed terrorism, the individual events often serve to provide
instances which are covered sufficiently in each nation. The two events
selected will be discussed in a moment.
How the proximity frame will manifest itself in both nearby and distant
reporting of the terrorist acts is the thrust of this research. Specifically,
it is expected that nearby reporting will tend to emphasize the criminal nature
of the act while the political implications of the act will be the focus of
reporting by distant media organizations. More simply: political talk and
criminal action talk differentiate or set the tone of the proximity frame. It
should be understood that the research does not compare political talk to
criminal action talk, but rather compares how the newspapers of each nation
choose to emphasize each category in their reporting. Measurement for this
question will be based on a computerized content analysis of all stories related
to the two specific events under study.
Background
Recent acts of terrorism in the United States have raised the awareness of
the American people concerning the nature of terrorism. Yet, it appears that
instances of terrorism in the U.S. are too infrequent for citizens to fear
becoming victims of terrorism. There is scant evidence on the issue. While
not inculcated into the thinking of Americans today, it is likely that continued
acts of political or quasi-political violence in the U.S. could serve to
heighten fears and thrust terrorism solidly onto the long term public agenda.
Today, we may portray terrorism as an issue on the rise. The Oklahoma City
bombing of the Murro Federal building and the capture and conviction of the
Unabomber have served to heighten the potential for increased salience of the
issue.[4]
For this study, two events have been selected and the United States and
Great Britain serve as investigative arenas. This comparative study then looks
at critical events in two nations and through the "eyes" of the elite mass
media. Content analysis allows researchers to examine an issue both from the
perspective of counting topical events and from the perspective of issue
attitude.
Comparing the United States and British Media
Friedland and Mengbai (1996) point out that "journalism is the first draft
of history" and that comparative studies aid in our interpretation of history by
giving us multiple frames of reference. The researchers point out that by
comparing news coverage from different sources, we may be able to draw
conclusions about the nature of news gathering and reporting in general.[5]
According to Blumler, Mcleod and Rosengren, there are three distinctive
contributions to knowledge supplied through comparative research. The first is
at the level of observation. "...comparative inquiry cosmopolitanizes, opening
our eyes to communication patterns and problems unnoticeable in our own spatial
and temporal milieux." The second is comparitivism's ability to "...overcome
space- and time-bound limitations on the generalizability of our theories,
assumptions and propositions." The third contribution is that only comparative
studies are capable of providing analyses to "...explore and reveal the
consequences of differences in how communication is organized at the macro
societal level."[6]
Blumler, et al suggest that more comparative studies are needed to ensure
that we produce theories which can move across the spatial and temporal
boundaries of traditional communication research.
Blumler, McLeod and Rosengren believe that "comparative research could be
called the communication field's 'extended and extendable frontier.'" They
report that as researcher's international contacts multiply in an ever-changing
but increasingly global environment, it is necessary for researchers to reach
across national and cultural borders to investigate communication phenomena.
This study is designed to advance this area of research. [7]
Dogan and Pelassy explain that the role of comparative studies is to find
commonality and isolate differences in discussions of the attributes of various
social actors, groups, and organizations. The comparativist seeks to gain
knowledge through reference beyond the limits of one environment. When
comparing nations, we look for reference frames which will allow for the common
explanation of actions/situations/ outcomes. This commonality, in turn,
provides meaning and understanding of the now enlarged environment. Knowledge
is extended.[8]
The media of different nations often cover the same issues differently.
Ethnocentrism is a common factor cited in national reporting differences. The
development of media historically will produce some cultural uniqueness within a
nation. This culturally-driven uniqueness generally spawns some ethnocentric
attitude which actually serves to increase the strength of the distance factor
in a nation's news reporting. Therefore, proximity has two dimensions. The
first dimension of proximity is geographic distance, the second dimension is a
nation's culturally-driven bias towards nationalism. Each dimension will be
explored in the research. Despite differences of geography and ethnocentrism,
there are significant similarities in how news stories are gathered, written and
disseminated.
How might we establish controls for this comparative study? The geographic
distance in a study is controlled by the selection of the news outlets reporting
the stories. Furthermore, controlling cultural differences is a process of
winnowing language and style to a central and common understanding. A common
language and the use of like-defined words and phrase syntax assist in this move
towards commonality. Once achieved, the research is free to examine the
geographic differences question more carefully.
Political Violence In this Study
Political violence has been relatively commonplace in Great Britain for
more than 25 years (some might say 50 years or more). Nearly 250 acts of
political violence were "experienced" by Great Britain and its environs between
1970 and 1992. Of those, 127 acts of violence occurred within Britain's island
borders.[9] Most active among those groups targeting the British have been the
Irish freedom/separatist groups known as PIRA (the Provisional Wing of the Irish
Republican Army), the INLA (Irish National Liberation Army), the Irish
Republican Army (IRA) and to a lesser extent, the IRA's "legal" arm, Sinn
Fein.[10] Throughout much of the 1970s and 1980s, and now into the 90s, the
three groups have fought a media war of one-upmanship in terrorist acts.
Second to the "Irish dimension" have been the Palestinians. Great Britain
has been an attractive haven for Middle-Eastern political moderates and
pro-westerners avoiding Palestinian retribution. The Palestinians have often
worked in British territory to carry out acts of violence against individuals
and groups who have appeared to be working to thwart the Palestinian goals of
organizational recognition and ultimately, nationalism.[11]
From the perspective of the United States, modern acts of terrorism have
been relatively few. Following the university bombings by the Students for a
Democratic Society (or fringe elements of the group like the Weathermen) during
the late 1960s and early 1970s, political violence within the borders of the
United States was nearly unheard of until 1993.[12] Small attacks against
foreign enclaves in the U.S., but not associated with the U.S. government, were
occasionally the focus of acts of political violence, but even those events were
rare. It was not until the bombing of the World Trade Center in 1993 that true
political violence again occurred against the U.S., its citizenry and
government. Since that time two more direct acts of political violence have
been aimed at the society in general, both in 1995. The April 1995 bombing of
the federal building in Oklahoma City and the October 1995 malicious destruction
of the rail lines near Santa Fe, New Mexico both stand as recent terrorist acts
against the citizens of the United States within the borders of the United
States.
These recent acts of political violence in the United States provide a new
backdrop for investigation of the media in both the U.S. and in Great Britain.
We are capable of comparing one nation with relatively recent acts of political
violence to a nation with a modern history of political violence. Both nations
share a common language and numerous cultural similarities, allowing media
differences to be more readily revealed and eliminated through investigation of
news coverage.
For this research the two instances of political violence were selected
because of their individual impact on the nation where the act took place and
because they could be easily labeled as political acts. The political violence
issues are; the 1993 bombing of New York City's World Trade Center, and the
1992 mortar shelling of #10 Downing Street in Great Britain aimed at
assassinating Prime Minister John Major. These events provide the basis for a
comparative study of the agenda-setting effects of political violence in the
United States and Great Britain.
Methodology
This content analysis was undertaken through the use of computer-generated
techniques. VBPro, developed by M. Mark Miller at the University of Tennessee,
is combined with SPSS for statistical analysis of the potential affective
attributes of proximity. News stories collected in archived databases are
examined via four elite national newspapers, two in the U.S., and two in Great
Britain. Full text archives of the stories were available through DIALOG. The
database information was compared to the publication indices to ensure all
stories were counted.
In order to set the affective attributes of proximity as a frame, two
hypotheses are set forth:
H1: Terrorist acts will be reported in criminal terms where the act is
proximal.
H2: Less proximal terrorist acts will be reported in more political terms.
Measurement
Four newspapers, two British and two U.S., were used for the examination.
They are the New York Times, the Washington Post, the Times of London, and The
Guardian (now in London, formerly the Manchester Guardian).
For this study full text accounts of the content of the four elite
newspapers were analyzed. These were used to identify, examine, determine and
track the type of discourse on terrorism by the newspapers.
The purpose is to determine whether proximity frames the issue of
terrorism by characterizing the event in either criminal or political terms. A
list of terms was developed to create the two categories and determine if the
terrorist event was defined in either criminal or political terms by the
selected media.
VBPro, a full-text document analysis computer program, was first used to
provide the raw data for statistical analysis. Content analysis, by definition,
is the quantification of meaning in documents. According to Babbie, "meaning
may be either manifest (obvious) or latent (implied)." In manifest content, the
words themselves define the manner in which the written material can be
assessed. Latent content analysis is more difficult to categorize since it
deals with underlying meanings attached to written material. For this study,
manifest content is used for our measurement data. It also is assumed that
manifest content would create the most direct correlation to affective
attributes of framing.[13]
VBPro is capable of fully counting all words in use for a particular story
or set of stories. Words are then listed in order of frequency in the news
copy. These lists were used supplementally to supply the terms used in the
analysis to differentiate between depictions of criminal/political actions.
Coding assistants were asked to categorize the words in a pretest to the final
statistical analysis.
For this study, the four newspapers were examined and their content
categorized by the level of discourse which focused on either political or
criminal talk. More than 8,000 pages of copy, extracted via DIALOG, were
initially part of the content analysis. The newspaper copy was carefully
screened and resulted in more than 4,000 pages for specific analysis.
The word list for coding was developed using results from a VBPro
alphabetized count of 15 stories from each newspaper. All were selected at
random. The alphabetized count was used to determine consistent/frequent use of
certain words by the newspapers. there was no attempt to differentiate between
the newspapers or the types of words. This was an aspect of screening out
cultural terms.[14]
Words of a general nature and common to most sentences were eliminated from
the search. Specifically eliminated were words which conveyed little or no
"value" meaning , but were generally described as objects or pronouns like we,
they, it, that, etc. Also eliminated were verbs which were forms of "to be,"
and other similar verb forms. After deleting words of little value for this
study, the alphabetized sorting resulted in more than 2,000 words for initial
examination. These words were then scrutinized for their potential content as
either political and criminal affective attributes. Those words without
potential content were eliminated because they contributed little to the
structure of political or criminal "talk" by journalists. Other words which
appeared, but infrequently, also were eliminated. From the total list, 77 words
were selected for potential use in the study.
Five university professors, all former print or broadcast journalists, then
coded the group of 77 words to determine whether those terms should be
considered political, criminal, neither or both. The coders were advised that
the nature of the study concerned political violence and that it was
specifically about newspaper coverage of the issue of terrorism. No other
information was given.
Terms which were considered neither or both by the majority of the coding
group were eliminated automatically. Terms which were considered political by
four of the five coders were included, as were terms which were considered
criminal by four of the five coders. This created two sets of terms. The coded
political search word set contained 26 words, while the criminal word set
contained 13 words.
The coders were at a high level of disagreement on the nature of certain
terms which could be used in a strictly criminal sense. For instance, the word
"kill," and its other word forms, killers, killing, etc., were labeled as "both"
by three of the coders. So, the terms were eliminated. Because this study is
about similarities and differences, it was felt that there would be adequate
copy studied to determine if the hypotheses were to be confirmed.
Two files, one for the 1993 World Trade Center coverage and another for the
1992 mortar attack on #10 Downing Street coverage were created. Each
newspaper's coverage of the event was tagged as a case. VBPro was used as the
initial analytical tool in the study to search for the terms from each list in
each newspaper in the two files. VBPro can be formatted to examine files by the
case, paragraph or sentence. In this instance, the cases were examined at the
paragraph and sentence levels. VBPro specifically looks for terms per
paragraph. Since it is probable that some paragraphs contain terms from both
lists, it is likely that some paragraphs are counted in each group. This is not
of consequence in this analysis because the questions sought to understand the
differences and similarities in coverage between two nations by word selection.
The paragraph results were consequently used for incorporation into SPSS for
analysis. Correlations were derived from both the individual newspapers and
through a cross-national examination of the terms.
DIALOG full-text citations were used as the source of story information for
the content analysis. As a test of VBPro, the DIALOG stories each were tagged
with a nonsense word and then counted through the search function of VBPro. A
test of 30 selected stories produced an exact count of 30 as a VBPro result.
Results
Content Analysis: World Trade Center
VBPro recognized 385 stories and 7,911 paragraphs in the New York Times
articles on the bombing of the World Trade Center. It also established that
1,709 paragraphs (21.6 percent) contained at least one of the selected political
terms and 990 paragraphs (12.51 percent) contained criminal terms. Specifically,
there were 2,622 hits from the list of political terms and 1,266 hits from the
list of criminal terms. These numbers, in and of themselves, provide only a
glimpse of the way in which the New York Times reported on the World Trade
Center bombing. A comparison with other news reports is needed to draw any
conclusions from the initial reports. The relationship will be examined after
reporting on the findings from each of the other newspapers.
For the Washington Post, VBPro recognized 179 stories with a total count
of 3,493 paragraphs. The program reported that 999 of those paragraphs (28.6
percent) contained at least one of the words from the political terms list. In
addition, there were 611 paragraphs (17.49 percent) in which criminal terms were
reported. Examining counts of the number of terms in each yielded results of
1,700 instances of the use of political words from the list and 815 criminal
words used in those same stories.
VBPro reported that the Times of London ran 51 stories on the World Trade
Center bombing. Those 51 stories contained 531 paragraphs. There were 227
paragraphs (42.75 percent) with political terms reported and 136 paragraphs
(25.61 percent) reporting the use of criminal terms from the word list. An
examination of the actual word count for political terms and criminal terms in
the Times of London yielded 339 political words and 186 criminal words.
The Guardian results from the VBPro analysis yielded an even smaller number
of stories reporting on the World Trade Center bombing, only 29. VBPro
recognized 445 paragraphs in the count of those stories. Of those, 155
paragraphs (34.83 percent) contained political terms and 75 paragraphs (16.85
percent) contained the use of criminal words in the selected list.
Table one (1) shows the results of each of the four newspapers based on
their results in the use of political words. Table two (2) is similar, but
lists the results from the criminal word term search. In each table, the
results have been organized so that we may begin to understand that there are
differences in the reporting of the terms between the nations.
Political Terms
World Trade Center Bombing
Elite Newspaper Coverage: U.S. & Great Britain
Newspaper
# of Stories
Mentions/
Table 1
[1] Ghanem, S. (1997). Filling in the Tapestry: The Second Level of Agenda
Setting. In Communication and Democracy (M. McCombs, D.L. Shaw and D. Weaver,
eds). Mahwah, NJ: Lawrence Erlbaum.
[2] Dogan, M. And Pelassy, D. (1990). How to Compare Nations: Strategies in
Comparative Politics. 2nd ed. Chatham, NJ: Chatham House Publishing.
[3] Richard A. Pride, "How Activists and Media Frame Social Problems: Critical
Events Versus Performance Trends for Schools," Political Communication, 12,
(1995), 6.
[4] For a discussion on issue salience see: S. Iyenger (1979). Television
News and Issue Salience. American Politics Quarterly, 7, 395-416. And D.P.
Demers, D. Craff, Y-H Choi and B.M. Pessin (1989). Issue Obtrusiveness and the
Agenda-Setting Effects of National Network News. Communication Research, 16,
6, 793-812.
[5] Lewis A. Friedland and Zhong Mengbai, "International Television Coverage
of Bejing Spring 1989: A Comparative Approach," Journalism and Mass
Communication
Monographs, 156 (April, 1996).
[6] Jay G. Blumler, Jack M. McLeod, and Karl E. Rosengren, eds. An Introduction
to Comparative Communication Research, in Comparatively Speaking:
Communication and Culture Across Time and Space (Newbury Park, CA: Sage, 1992)
3-4.
[7] Blumler, McLeod and Rosengren, 3-4.
[8] Ibid, Dogan & Pelassy, pp. 3-4.
[9] Bruce W. Warner. Great Britain and the Response to International
Terrorism in D.A. Charters (ed.), The Deadly Sin of Terrorism: Its Effect on
Democracy in Six Countries. (Westport, CT: Greenwood, 1994).
[10] While Sinn Fein is a proscribed (technically illegal) faction, the
British have kept channels of communication open with this group. They were
strongly involved in the negotiation process which temporarily created a
cease-fire in Northern Ireland in late 1995.
[11] Warner. 15.
[12] Some authors disagree as to the nature of political violence. In our
definition, only acts which specifically seek political power are under
consideration as political acts. By example, the actions of most rural militia
are against the nature of government itself, not
just a specific government policy and would be exempt from the study.
[13] Earl Babbie (1992). The Practice of Social Research. 6th ed. Belmont,
CA:
Wadsworth.
[14] Kerlinger, Fred N. (1986). Foundations of Behavioral Research, 3rd ed.,
Appendix C, pp. 637-644. Ft. Worth, TX: Holt, Rinehart and Winston, Inc.
|
