Although nearly all reviews of Gulf War coverage recognize that media were driven by a desire for images, explicit attempts to systematically analyze the universe of depiction that came to represent "Desert Storm" have not appeared.1 Of course, there has been a great deal of commentary con­cerning the unprecedented role played by CNN in establishing an aura of "live" coverage. Television, especially in the initial days of the war, created the impression that the conflict would unfold before us on the screen, and prompted expectations for continuous visual coverage that, to a large extent, were never realized. Several scholarly reports, articles, and books since the war have examined the implications of impromptu Gulf War reporting for issues of information control, and journalistic and editorial practice.2 The best analyses at least implicitly examine the visual components of media reporting, and a few provide perceptive discussions of media visualizations of the war.3

The Gulf War presents a fascinating and ironic case study in this regard for it was touted as a war the public could witness firsthand at the same time that it was considered one of the most effectively censored and controlled media events in modern history. What images of the conflict did the public actually see? The press frequently used terms like "techno- imagery" and "nintendo warfare" to characterize photojournalistic cover­age.4 But to what extent did these sensational "nintendo" images represent the media's depiction of the war? As with some of the most famous images of the Vietnam War (Eddy Adam's photograph of the Saigon police chief, Brig. Gen. Nguyen Ngoc Loan, summarily executing a suspected Vietcong operative in the street, 1968; Ronald Haberle's photograph of My Lai vic­tims, 1968; Nick Ut's photograph of a young girl running naked along a road, her skin flayed by napalm, 1972), one must question whether certain images have come to symbolize a conflict because they epitomize typical character­istics of the event and its coverage, or whether they ascend to symbolic prominence because they represent some facet of the war that is unique or sensational.5

In an attempt to fashion one concrete indicator of the way the war was visualized by the news media, we conducted a systematic analysis of all 1,104 Gulf War-related pictures published in the three largest circulation U.S. newsmagazines during the course of the war. This included every issue of Time, Newsweek, and U.S. News and World Report dated 21 January 1991 through 18 March 1991, a complete population 7 of newsmagazine photos published during "Desert Storm." 

These three magazines offered a week-by-week compendium of war news and featured prominently scaled, color photographs from the Gulf War region throughout their reports. Because newsmagazines hit the stands more than a week after the events they report on, they serve as a kind of news digest ‒ compressing, recapitulating, elaborating upon, and even critiquing the television and newspaper reports of a previous week. In the case of the Gulf War, the photographs in these magazines seemed to serve a parallel function, offering a set of visual "highlights" that reiterated the news images of each week's events. For these reasons, the weekly newsmagazines provide useful site for examining the way the conflict was pictured for U.S. audiences.

The Analysis of Newsmagazine Pictures. We categorized news pictures not just by their most manifest content,  but also in terms of visual style, pictorial genre, and event context. That is, we not only identified picture subjects per se (e.g., U.S. soldiers, Iraqui soldiers, F-16 fighter-bombers) but accounted for the genre of photograph (e.g., weapons catalog illustration vs. an on-location action shot) and the contexts and events in which subjects appear (e.g., soldiers pictured in combat zones vs. soldiers appearing "be­hind the lines" in bivouac or in exercises and preparations). Relevant to this is the use of current on-location photographs vs. file or archival photographs. We formulated coding categories in which we attempt to account for these distinctions (see below). 

An important part of the analysis includes registering information from accompanying written text, captions, and bylines concerning the source of the picture and any claims made about the pictured subject matter (e.g., its location or its status as an event). A picture of soldiers in battle gear struggling over a sand dune might be a photo taken during actual combat in Saudi Arabia, Kuwait, or Iraq. Or it might be a shot of soldiers engaged in training exercises in southern California. Often, photographs accompanying reports of the ground war appeared to be pictures of U.S. soldiers moving against Iraqi troops, but information in the text or small print in the caption identified the picture as a file photo from an earlier military exercise. 

All coding was done by the two authors, who formulated a compre­hensive set of forty types of photographs one might encounter in Gulf War coverage. These forty categories were based on:

(1) our review of collections of war photography and documentary film from the coverage of World War II, Korea, and Vietnam, and our survey of the literature on war photojournalism;

(2) preparatory surveys of news coverage from the Iraqi invasion of Kuwait in August 1990 though the cessation of "Desert Storm" at the end of February 1991 (this included reviewing scholarly analyses of Gulf War coverage, taking note of the patterns they identified, and precoding pictures of Persian Gulf coverage prior to Desert Storm in the November and December issues of each newsmagazine);

(3) our own expectations of what might logically appear in coverage of United States involvement in a war in the Middle East.

The authors precoded pictures from two issues of each magazine together, clarifying the pertinence and application of coding categories and resolving any unanticipated ambiguities for specific coding units. Then each author coded the entire population of 1,104 newsmagazine photos independently. A post-test of intercoder reliability showed a +97.8% agreement between coders. The number of disagreements was small enough that discrepancies were re-examined and resolved.

The use of forty coding categories produced a highly detailed analysis that allowed us to identify the presence and absence of a variety of pertinent photographs. After accounting for a larger set of distinctions, we were able to drop two and collapse others in our discussion of major content patterns.

The popular perception that the public had witnessed the Gulf War first-hand motivated our formulation of coding categories. A primary interest was to see how much of the pictorial coverage actually depicted wartime events and combat-related military activity. Categories were designed to discriminate between depictions of actual warfare and depictions of noncombat army life, training exercises, and catalogues of weaponry. Thus, a central research question concerned whether the photojoumalistic coverage in U.S. newsmagazines was characterized more by candid, on-the-scene, visual reporting of events, or by pre-existing, staged, or symbolic representations of nations, political actors, and military power. Were the photographs motivated by breaking news, by specific events unfolding in the Gulf, or do they represent the kind of generic, conventional illustration that might have been taken at a number of locations or times? These questions relate directly to current epistemological debates about the reliability of photojoumalistic "records",'6 and to the growing concern for understanding the role of visual communication in media literacy.7 Do news photographs provide reliable first-hand information about specific places and current events, or do they more often serve as symbols for enduring concepts (the desert, Arab culture, the president, American military technology, etc.).8 Gauging the frequency of on-the-scene combat pictures also provides a means of checking on journalists' complaints about military censorship and limited access to war zones.

A second concern was to compare the coverage of U.S. and Iraqi military forces and political leaders. Because the media were accused by many of promoting the war, we tried to devise categories that would help track the emphases of pictorial coverage. Were pictures used to exaggerate the strength of Iraq's army, to symbolize the war as a personal conflict between Saddam Hussein and George Bush, or to emphasize the superiority of American forces? To what extent were other coalition members and Middle Eastern nations made visible? Was the pictorial coverage strictly United States centered, or was there some attempt to illustrate historical or contextual factors affecting the conflict? 

Finally, given the legacy of Vietnam so frequently invoked by military spokesmen and media commentators, we also wanted to examine the atten­tion given to civilian casualties and damage and the frequency with which pro-war and anti-war demonstrations were pictured.

The distribution of pictures is consistent enough across the three magazines to make an aggregate analysis worthwhile (the categories "Cata­loguing the Arsenal," "U.S. Troops," and "U.S. Political Leaders" placed 1, 2, and 3 in Time and Newsweek, and 1, 2, and 4 in U.S. News and World Report).However, following our report of the aggregate data, we compare the three magazines, noting points of similarity and difference in pictorial emphasis. 

After coding depictions in each of the three newsmagazines, we found that 6 categories constitute more than half of all the 1,104 pictures coded and that of the 36 final categories account for 88% of the total (91% in U.S. News, 90% in Newsweek, and 82% in Time).9 The 12 major categories (comprising 76% of the total) are displayed in Table 1. The remaining 24 low frequency categories are listed in the endnotes.10 

The dominance of a narrow range of picture types is even more striking when one notes that the top 18 categories, created when we were attempting to elaborate an exhaustive set of coding categories, constitute only nine or ten different genres of pictorial content and style (one genre, "Political Leaders" for instance, comprises four coding categories detailing Western, Iraqi, non-Iraqi Arab, and Other leaders pic­tured). The three most numerous genres ‒ pictures of military hardware, noncombat scenes of troops, and photos of political leaders ‒ make up more than half of all the pictures published in these magazines. A summary of the major picture genres follows.

The most numerous types of pictures in all three magazines are photographic records of military hardware. All non­combat illustrations of planes, ships, tanks, missiles, or other weapons systems were included in this category. (Depictions of weapons and military hardware being used in scenes of actual combat were coded "Combat.") "Arsenal" pictures include Department of Defense and photo agency file photos, illustrations provided by arms manufacturers (McDonnell Douglas, Raytheon, etc.) and photos reproduced from arms catalogues (Janes Fighting Ships, Janes Fighting Aircraft, etc.) 

These catalogue-style pictures make up 26% of the total. In Newsweek (32%) and U.S. News and World Report (31%), almost one third of all pictures were of this type. Time ran fewer of these arsenal pictures (15%). However, the 25 February 1991 issue of Time included a pull-out "war map" cata­loguing 110 types of aircraft, missiles, tanks, armored vehicles, and artillery. Since they did not appear as "news photos," these illustrations were not coded.

TABLE 1 22
Gulf War Pictures in Newsmagazines (Number of Pictures)

Category	Time	Newsweek	U.S. News and World Report	TOTAL
Arsenal (U.S.-Allied)	46 (13%)	112 (27%)	91 (27%)	249 (23%)
Troops (U.S.-Allied)	39 (11%)	73 (18%)	46 (14%)	158 (14%)
Political Leaders (U.S.: Bush)	31 (9%)	26 (6%)	23 (7%)	80 (7%)
Military Leaders (U.S.)	10 (3%)	16 (4%)	33 (10%)	59 (5%)
Political Leaders (Iraq: Hussein)	14 (4%)	19 (5%)	14 (4%)	47 (4%)
Arsenal (Iraq)	7 (2%)	19 (5%)	14 (4%)	40 (4%)
Combat (All nations)	10 (3%)	14 (3%)	14 (4%)	38 (3%)
Destruction (Iraq)	16 (4%)	14 (3%)	7 (2%)	37 (3%)
Media (U.S.)	21 (6%)	8 (2%)	7 (2%)	36 (3%)
Civilian life (U.S.)	20 (5%)	5 (1%)	11 (3%)	36 (3%)
POWs (U.S.-Allied)	12 (3%)	12 (3%)	8 (2%)	32 (3%)
Historical Photos (All nations)	12 (3%)	9 (2%)	10 (3%)	31 (3%)
Total of 24 Low-frequency
Categories*	120 (34%)	81 (20%)	60 (18%)	261 (24%)
Total	358 (100%)	408 (100%)	338 (100%)	1,104 (100%)

Spearman rank-order Correlation: Time vs. Newsweek +.874 / Time vs. U.S. News +.820 / Newsweekv s.U.S. News +.834. 

*For Other categories, see endnote 10.

In many cases, all three magazines reproduced the same pictures from the same sources. The 28 January 1991 issues of Time, Newsweek, and U.S. News and World Report, for example, each contain an almost identical sequence of photo inserts with the same roster of planes and missiles (Newsweek pp. 18‒22, 30‒31; Time pp. 30‒31; U.S. News and World Report pp. 30, 33, 38) 

The prevalence of weapons illustrations in the war coverage has been noted by several observers of Gulf War reporting; Douglas Kellner describes it as part of the media's preoccupation with "images of techno-war";11 George Gerbner identifies it as part of a "Gulf War movie" constructed by the media.12In his synopsis of the coverage, Elihu Katz writes, "We saw portraits of the technology - advertisements for smart planes, tanks, missiles, and other equipment in dress rehearsals of what they are supposed to do in combat, but we rarely, if ever, saw them in action. Indeed, it was as if there was no other side."13

Our analysis supports these observations. For every picture of actual combat (3% of the total), the newsmagazines printed about nine noncombat photographs of American military hardware. In place of images of warfare, there are images of American military technology and power. This pattern becomes even more pronounced when we include the analysis of the second most numerous category, U.S. Troops. 

Troops. Initially, we created categories called "Troops (Allied)" and "Troops (Iraqi)," under the assumption that press coverage of the war would not restrict itself to the U.S. military but include material on the Iraqi army as well as other members of the multi-national coalition (the "allies"). We made this assumption in light of the fact that the Bush administration's response to Iraq's invasion of Kuwait was predicated both on United Nations resolutions and multi-lateral participation in the military action.14 Our coding results, however, reveal virtually no pictures of troops other than American. Of the 163 pictures of "troops" (that is, soldiers in noncombat situations), 154 (95%) show U.S. soldiers, 4 (2%) show Saudi or Kuwaiti pilots or soldiers, and 5 (3%) show Iraqi soldiers. Because the "Allied Troops" pictured are almost exclusively American, we routinely refer to this category as "U.S. Troops." 

Pictures of U.S. troops constitute 14% of all pictures, the second largest category, varying from 11% in Time, to 14% in U.S. News and World Report, to 18% in Newsweek. When combined with pictures of Iraqi and other troops, the percentage for all "troops" pictures climbs only slightly, to 15%. Two aspects of the troops pictures are most salient:

1. The wholesale substitution of photos of soldiers in noncombat situations for any actual combat coverage. Based on our precoding of coverage of the troop build-up in the November and December 1990 issues of the three magazines, it appears that the newsmagazine photographs of soldiers taken during Desert Storm rarely differed in content and style from the photographs appearing before the start of the war. In fact, it is apparent that some of the pictures used to illustrate stories during the war are file photos taken during maneuvers prior to the war.
2. The completely one-sided focus on U.S. forces. Our expectation for pictures of "allied" troops was not realized, and pictures of Iraqi troops are virtually nonexistentfjhov are liter­ally nonexistent in 17.S. News and World Report).

Political and Military Leaders. The third most frequent category, "West­ern political leaders" (7% of the total), again consists almost exclusively of pictures of U.S. politicians, with photos of George Bush making up more than half (45 of 81). Similarly, the category "Iraqi political leaders" is almost exclusively pictures of Saddam Hussein (42 of 47). 

Pictures of George Bush (45) and Saddam Hussein (42) are reproduced on a regular basis and with remarkably equal frequency. Out of twenty seven total magazine issues, only one (the 4 February 1991 issue of Time) includes no pictures of Bush or Saddam. On several occasions pictures of these leaders are positioned opposite one another across the same page, crea­ting an explicit visual confrontation. Such cases clearly serve to personalize the conflict. But whereas Saddam is virtually the only Iraqi political figure to be pictured (the one exception being Ambassador Tariq Aziz), the 45 pictures of Bush join 36 pictures of other American diplomats and politicians, including several pictures of Secretary of State James Baker and pictures of a number of congressional leaders. Thus, the image of a solitary dictator contrasted to a more pluralistic image of American democratic consensus. 

There is an even sharper contrast in the representations of military leaders. Fifty-nine pictures of U.S. generals and military commanders were published but not one picture of an Iraqi military leader, nor any pictures of Saudi, Egyptian, French, British, or other coalition commanders.

Among the three magazines, Time (17%) published slightly more pictures of political figures than Newsweek or U. S. News and World Report (both 13%), and ran more pictures of political figures from nations other than the U.S. and Iraq. Newsweek and U.S. News and World Report depicted a narrower range of political figures more heavily dominated by pictures of George Bush and Saddam Hussein.

U.S. News and World Report differed noticeably from Time and Newsweek in its heavier emphasis on pictures of U.S. military leaders. It is the only one of the three to publish more pictures of U.S. military commanders than of U.S. political figures.15
Thus, the three predominant genres of Gulf War pictures (military hardware, unengaged troops, and political and military leaders) constitute 57% of all the pictures published in the three newsmagazines and present a pattern of visualization in which:

1. The cataloguing of a massive, high-tech, U.S. arsenal is juxtaposed to a much smaller catalogue of Iraqi weaponry (249 pictures to 40 pictures).
2. There are 158 pictures of noncombat U.S. troops plus 38 pictures of U.S. forces engaged in combat activities (196 pictures of U.S. soldiers in all) compared to 5 pictures of Iraqi forces (Iraq remains a largely "unseen enemy").
3. Pictures of U.S. political figures and military leaders (139 in all) are not paralleled by pictures of Iraqi, or even other Arab or Middle-Eastern political and military leaders. With the exception of Saddam Hussein, who is paralleled to George Bush, nations and leaders in the Gulf region remain obscure. 

Seemingly important genres of war photojournalism whose frequency was surprisingly low include scenes of battle and wartime destruction, and pictures of the human costs of war. 

Combat and Destruction

Pictures of actual combat activity (troops, tanks, or armored vehicles advancing in combat zones, artillery or naval guns firing, planes taking off from aircraft carrier decks or Saudi airstrips, "smart bomb" images, anti-aircraft fire, the launching of missiles, soldiers forming defensive lines, retreating from battle, or evacuating wounded) constitute only 3% of newsmagazine pictures during the war. Depictions of wartime destruction (bombed out buildings, pieces of scorched scud missile debris, burning oil wells, etc.) are 6% of the total: 37 in Iraq (3%), 27 in Kuwait and Saudi Arabia (2%), and 7 in Israel (0.6%)

Casualties

One of the most interesting picture categories involves depictions of human casualties. During the war, we read reports of the U.S. 
Government's desire to prevent images of war victims from reaching the television screens at home, and we learned of their policy to bar the media from Dover Air Force Base, where transport planes carrying dead American soldiers arrived back in the States. We also heard photojoumalists complain that their photographs of Iraqi corpses and U.S. bombing damage had been denied circulation in the U.S. by news publications and wire services.16

The results of this analysis confirm that pictures of war casualties were not often reproduced in U.S. newsmagazines. Only 27 of 1,104 pictures (or 2% of the total) show any signs of wounded or killed American soldiers (there was 1 picture of a killed Saudi soldier), and very few of these pictures were taken in the Gulf region. Nearly all are either flag-draped coffins in U.S. home-town funeral ceremonies or portraits taken of soldiers prior to deploy­ment in the Middle East. Although Time published more pictures of killed or wounded U.S. and Iraqi soldiers (21,6%) than Newsweek(9,2%) and U.S. News and World Report (4,1%), this is skewed by the fact that 12 of these occur in one issue as a collection of military file portraits.17 David Tumley's award- winning photo of a wounded U.S. soldier, grief-stricken over the dead comrade beside him in an evacuation helicopter, is the only example of the kind of dramatic imagery of human sacrifice that previous war photojour­nalism has led us to expect.

The relative absence of images of U.S. casualties might be attributable to the extremely low number of actual casualties incurred by U.S. forces. But the relatively large number of Iraqi casualties is even more invisible (there were 2 pictures in each magazine, 6 out of 1,104). And the number of civilian casualties pictured from Iraq, Kuwait, Saudi Arabia, and Israel totals only 19 (1.7%). Photographs revealing any sign of military or civilian casualties or funerals anywhere total 53, or less than 5% of the pictures published. This lends support to the charge made by many commentators that the U.S. media cooperated with the military to eschew images of bloodshed and present the American public with a "clean war." Pictures of media personnel - photos of reporters, photographers, TV camera crews, and newscasters on TV screens ‒ actually outnumber pictures of military casualties.

The predominance of a few pictorial genres logically implies a scarcity of many others. One of the most important results of the study was identifying facets of war that remained invisible. Categories for which we found no pictures in any of the three magazines included Iraqi Military Leaders and Individual Iraqi Soldiers. We also found that Iraqi POWs (a category separate from Troops) were only shown in large groups. The conflict was personalized in the visual juxtapositions of Bush and Saddam, but all other Iraqis remained part of an anonymous, or even unseen, collective.

Of course, this is consistent with traditional wartime depictions of the enemy, and presumably compatible with U.S. government and military agendas for characterizing Saddam and the opposing forces, but it is at odds with the claims made by American media that they were providing unprec­edented, comprehensive coverage of the war. Other types of pictures virtually absent in the newsmagazines included:

• Pictures of non-U.S. Allied military casualties. Time carried one picture of a killed Saudi soldier; otherwise, losses among coalition forces were not depicted.
• Public demonstrations against the war in coalition member nations. Despite widespread protests against the war in Britain, Egypt, France, Germany, Italy, Morocco, etc., Time printed one small picture of a demonstration in Paris, and Newsweek and U.S. News and World Report carried no images of demonstrations in nations aligned with the U.S.
• Pictures of civilian wartime life in Saudi Arabia. Al­though roughly 500,000 coalition troops were stationed in Saudi Arabia, U.S. News and World Report and Newsweek each carried only 1 picture of Saudi citizens in January issues and none appeared thereafter.
• Iraqi Troops. In total, Time ran 3 pictures of Iraqi troops, Newsweek ran 2, and U.S. News and World Report ran none.

Other categories containing less than one-half of one percent of the total include: Iraqi Military Causalities, Iraqi Civilian Casualties, and Civil­ian Casualties in Nations Attacked by Iraq (Kuwait, Saudi Arabia, Israel).

With regard to the research questions posed at the outset, the following conclusions are indicated by our analysis.

Remarkably little of the photojoumalistic coverage of the war pre­sented in these newsmagazines depicts actual combat activity of any kind, especially considering the claims that were made by television, newspapers, and magazines about their comprehensive and on-the-spot coverage. Cata­logues of military hardware and generic military illustrations stand in for images of actual events, and the massing of a powerful U.S. military machine becomes the central signifying image of the war.18 More than 500 "backstage" and file photos were published and only 38 of current combat activity. The number of historical photographs alone nearly matches the number of combat pictures.

The photographic portrayal of the war seems to promote American military and technological superiority as a central theme. The overall image of the war that emerges from the analysis is extremely U.S. centered. The majority of pictures appearing in each of the magazines are pictures of U.S. troops and technology; the military forces of Iraq and other coalition nations are almost completely absent, and, with the exception of Saddam Hussein, the political and military leaders of other nations do not appear. Even taking into account the fact that these magazines were addressing an American audience during a time of national military mobilization, the pattern of depiction seems extremely narrow.

The focus of photographic coverage neglects several important as­pects of the war, particularly those involving the human cost of the conflict. There are very few images of casualties or human suffering due to the war. There are very few images of political activity related to the war (discussion in the United Nations, war-related demonstrations in Europe or North America, pro-Iraqi demonstrations in parts of North Africa and the Middle- East). There are no portrayals of the cultural contexts surrounding the conflict (no images of cultural ways of life in Iraq, Kuwait, Saudi Arabia, the Arab Emirates, Jordan, Israel). There are no pictures related to the historical or geopolitical factors drawing Western Europe and North America into the conflict (the historical photos that appear are nearly all emblematic scenes from previous wars, especially Vietnam and the Iran-Iraq war).

The photographs of the newsmagazines, rather than offering perti­nent pictorial information about specific developments in the Gulf War, reproduce traditional patterns of war imagery much like those identified by Perlmutter in his study of high school history textbooks.19 Perlmutter found that the selection of war illustrations appearing in textbooks was overwhelm­ingly U.S. centered, that pictures were selected with a preference for "action" but that the action was usually contrived and not authentic. He found little, if any, visual portrayal of the politics of war, and he found that images of death, maiming, and funerals were avoided. 

Although it is clear that a number of factors contributed to constrain­ing photo opportunities in the Gulf ‒ including the press pool system, limited access to areas of military activity, the censorship imposed by the military, and the self-regulation maintained by news media - the relative uniformity of Gulf War depictions across the three newsmagazines suggests that pho­tographers, photo agencies, and newsmagazine editors operated from the outset within parameters and expectations that conform to long-standing conventions of war illustration.20 

One aspect of the pictorial coverage, however, seems unprecedented: the tremendous emphasis on cataloguing the high-tech arsenal. This may be a reflection of the expanding role played by weapons technology and commercial arms manufacturers in modem warfare, warfare that is more than ever calculated and controlled by a professional-technical-military elite. Or it may be a direct result of effective public relations efforts and media control on the part of the Pentagon, efforts that promoted U.S. military technology while masking the human costs. More comparative research is needed to address this issue. However, the present findings suggest that a narrowly limited inventory of pictures was used by the newsmagazines to represent the Gulf War, images shaped more by established conventions of military illustration than by specific events in the Gulf. As in historical illustrations of past American wars, there is a central preoccupation with U.S. military power and military leaders. But more than in the past, the newsmagazine pictures emphasized military technology and hardware at the expense of the human side of war. 

1. This may be due to the inherent practical difficulties of large-scale content analyses or the fact that too few researchers are trained in methods of visual analysis.
2. Anna Banks, "Frontstage/Backstage: Loss of Control in Real-time Coverage of the War in the Gulf," Communication 13 (1992): 111‒19; Everette E. Dennis, David Stebenne, John Pavlik, Mark Thalhimer, Craig LaMay, Dirk Smillie, Martha FitzSimon, Shirley Gazsi, and Seth Rachlin, The Media at War: The Press and the Persian Gulf Conflict (NY: Gannett Foundation, 1991); Michael Griffin and Simon Kagan, "CNN and National Broadcasting in Israel During the Gulf War" (paper presented at the MacBride Round Table, Honolulu, January 1994); Elihu Katz, "The End of Journalism? Notes on Watching the War," Journal of Communication 42 (summer 1992): 5-13; Dou­glas Kellner, The Persian Gulf TV War (Boulder, CO: Westview Press, 1992); Hamid Mowlana, George Gerbner, Herbert Schiller, eds. The Triumph of the Image: The Media's War in the Persian Gulf ‒ a Global Perspective (Boulder, CO: Westview Press, 1992); William J. Small, "A Report ‒ The Gulf War and Television News: Past, Future and Present," Mass Communication Review 19 (1992): 3‒13; Barbie Zelizer, "CNN, the Gulf War, and Journalistic Practice," Journal of Communication 42 (winter 1992): 66‒81.
3. In his essay "Persian Gulf War: The Movie/' in Triumph of the Image, for example, George Gerbner provides a compelling precis of the instant "image history" beamed back from the Gulf. Daniel Hallin and Todd Gitlin, in "Agon and Ritual: the Gulf War as Popular Culture and as Television Drama,' Political Communication 10 (October‒December 1993): 411‒24, pro­vide an interpretive content analysis of a sample of national network, CNN, and local news in which they attend to images. And in Kellner's The Persian Gulf TV War, a meticulous, encyclopedic examination of the way information was controlled and manipulated in Gulf War reporting, he provides many discussions of the presentational "frames" and perceptual "images" or metaphors constructed to represent "good and evil" and characterize the "enemy" in media coverage.
4. In Kellner's Persian GulfTV War, he discusses the way that "technowar" images served U.S. military agendas, 157‒85.
5. Systematic analyses of TV news coverage during the Vietnam War have revealed that the popular perception of Vietnam as a "living room war," with nightly news coverage of events "in the field" and constant public exposure to the bloodshed of the war, is largely a myth. Early studies of Vietnam coverage including Lawrence W, Lichty, The War We Watched on Television,American Film Institute Report 4 (Washington, DC: American Film Institute, 1973), and Peter Braestrup, Big Story: How the American Press and Television Reported and Interpreted the Crisis of Tel 1968 in Vietnam and Washington (Boulder, CO: Westview, 1977), point up many of the inaccuracies of popular perceptions about the "living room war." These have been summarized and reconfirmed by Daniel Hallin in his comprehensive study The "Uncensored War": The Media and Vietnam (New York: Oxford, 1986). Similarly, despite the public fascination with "smart bombs" and "techno-imagery," we know that more of the Gulf War TV coverage consisted of such banal material as miliary briefings, commentaries by "experts," news studio dis­cussions, and interviews with politicians and government officials.
6. Dona Schwartz, "To Tell the Truth: Codes of Objectivity in Photojournalism," Communication 13 (1992): 95‒109; Kent Brecheen-Kirkton, "Visual Silences: How Photojournalism Covers Reality With the Facts," American Journalism (winter 1991): 27‒34; Jisuk Woo, "Journalism Objectivity in News Magazine Photography," Visual Communication Quarterly 1 (summer 1994): 9‒16.
7. Paul Messaris, Visual Literacy; Image, Mind, and Reality (Boulder, CO: Westview Press, 1994); James Flood, Shirley Brice Heath, and Diane Lapp, eds. A Handbook for Literacy Educators: Research on Teaching the Communicative and Visual Arts (NY: Macmillan, 1996).
8. A related distinction is made between "image-oriented" information and "more abstract and complex" information in Zhongdang Pan, Ronald E. Ostman, Patricia Moy, and Paula Reynolds, "News Media Exposure and Its Learning Effects during the Persian Gulf War,"Journalism Quarterly 71 (spring 1994): 7‒19.
9. After coding for the number of pictures falling into each coding category, we weighted the pictures according to their size to check whether frequency counts were adequately representing the prominence given to certain types of images. (It could be argued that a smaller number of pictures appearing in larger scale have greater prominence than a larger number of small pictures.) However, weighting the analysis for picture size made no appreciable difference; the same 18 categories as a percentage of picture weight still constituted 90% of the total, and frequency ranks remained virtually unchanged. Weighting for picture size significantly affected only one category: combat. The relatively small numbers of combat pictures appearing in the three magazines (about 3% of picture numbers) tend to be large in size (about 6% of picture space). On those relatively rare times when combat pictures were run they were reproduced on a prominent scale.
10. Low frequency coding categories were as follows:

MILITARY CASUALTIES (U.S.)	27 (2%)
DAMAGE AND DESTRUCTION (Saudi Arabia and Kuwait)	27 (2%)
PRISONERS OF WAR (Iraqi)	22 (2%)
POLITICAL LEADERS (Arab world, excluding Iraq)	19 (2%)
PUBLIC DEMONSTRATIONS (U.S.)	19 (2%)
WARTIME CIVILIAN LIFE (Kuwait)	19 (2%)
OIL AND ENERGY	15 (1%)
PUBLIC DEMONSTRATIONS (Arab)	14 (1%)
POLITICAL LEADERS (Other ‒ Soviet, Iranian, Chinese, United Nations)	13 (1%)
CIVILIAN CASUALTIES (Iraqi)	13 (1%)
ECOLOGICAL SUBJECTS	12 (1%)
PALESTINIAN LEADERS, ACTIVISTS, PROTESTERS	10 (1%)
WAR TIME CIVILIAN LIFE (Iraqi)	8 (1%)
WAR INFLICTED DAMAGE AND DESTRUCTION (Israel)	7 (.6%)
CIVILIAN CASUALTIES (Saudi, Kuwaiti, Israel)	6 (.5%)
MILITARY CASUALTIES (Iraqi)	6 (.5%)
TROOPS (Iraqi)	5 (<.5%)
WAR TIME CIVILIAN LIFE (Israel and the Occupied Territories)	5 (<.5%)
IRANIAN ACTIVISTS OR PROTESTERS	4 (<.5%)
WARTIME CIVILIAN LIFE (Saudi Arabia)	2 (<.5%)
PUBLIC DEMONSTRATIONS (Other Allied)	1 (<.5%)
MILITARY CASUALTIES (Saudi, Egyptian, British, French)	1 (<.5%)
MILITARY LEADERS (Iraqi)	0 (0%)
OTHER	7 (.6%)

1. Kellner, The Persian Gulf TV War, 157‒63.
2. George Gerbner, "Persian Gulf War: The Movie," in The Triumph of the Image, ed. Mowlana, Gerbner, and Schiller, 243‒65.
3. Katz, "The End of Journalism?" 8. In addition, our analysis shows that the pictures of weapons and technology that appeared were overwhelmingly American. Although Time's wall poster depicts "allied" aircraft, missiles, and armor manufactured and used by the United States, Britain, and France, depictions of weapons in the magazines invariably shows them in a U.S. military context. In the months during and after the war, it was reported that the success of U.S. weaponry was serving as a boost to foreign arms sales, particularly in the Mideast, prompting many commentators to suggest that Gulf War coverage had been, above all, a powerfully effective advertisement for the U.S. military and the American Defense Industry. See "High-tech's triumphs may bring a postwar arms race in Mideast," Boston Globe, 19 February 1991, Nat'l./Foreign sec., p. 1; "In war, nothing succeeds like success," Los Angeles Times editorial, 5 March 1991; Daniel Wood, "Post-Gulf weapons bazaar open," Christian Science Monitor, 18 March 1991, Economy sec., p. 8; "George Bush, in the arms bazaar," New York Timeseditorial, 6 June 1991; Adam Goodman, "Middle East arms bazaar draws fire," St. Louis Post- Dispatch, 16 June 1991, sec. E, p. 1; Andy Pasztor, "White House girds to promote huge arms sales to many nations," Wall Street Journal, 24 July 1991, sec. A, p. 7.
4. In fact, the British, French, and Saudi Arabian air forces participated in U.S.-led air strikes on Iraq from the beginning, and the same countries, plus Egypt, had significant numbers of combat troops under U.S. command in northern Saudi Arabia. Syria and Turkey officially contributed combat forces to the coalition but positioned their troops along their own borders with Iraq. Pakistan, Bangladesh, Morocco, Niger, Senegal, Czechoslovakia, and the Arab Emirates of the Gulf Cooperation Council contributed small numbers of troops or supporting personnel but played no major role in actual combat operations.
5. There are 33 such pictures in If. S. News and World Report (10% of all U.S. Nezvs and World Report pictures), 16 in Newsweek (4%), and 10 in Time (3%). Differences are statistically significant at chi-square=19.43, df=3, p<.00l.
6. CONTACT Press Images and Time magazine photographer Ken Jarecke, for instance, in a symposium at the University of Minnesota in April 1991, showed several photographs of incinerated Iraqi soldiers along the notorious "highway of death," which Time did not run and the Associated Press editorial offices in New York refused to put on the wire for their wire service clients to consider. Time published one of these photos more than nine months after the war in its 1991 year-in-review issue, 30 December 1991.
7. Sometimes such formal portraits were reproduced under headlines like, "The War Comes Home" (Time, 18 February 1991, cover and 2‒3) and "Heroes Remembered" (Newsweek, 11 March 1991,74-76). Since we wanted to note any and all pictorial references to human costs of the war, we coded all prior portraits and funeral photographs - even if they only showed the faces of grieving family members - as pictures of casualties.
8. These patterns of pictures parallel in an interesting way the categories Vincent found in his analysis of CNN stories. He found that press briefings and stated events by political and military leaders, tapes and pictures supplied by military or government sources, press pool stories of routine military life, interviews and commentaries by reporters, anchors, consultants and "experts," and stories focusing on the media itself, dominated CNN coverage and that more spontaneous or direct coverage of actually ongoing events was rare at best. Richard C. Vincent, "CNN: Elites Talking to Elites," in Triumph of the Image, ed. Mowlana, Gerbner, and Schiller, 181‒201.
9. David D. Perlmutter, "The Vision of War in High School Social Science Textbooks,"Communication 13 (1992): 143‒60.
10. This would be consistent with the reproduction of photojoumalistic codes outlined by Dona Schwartz in "To Tell the Truth: Codes of Objectivity in Photojournalism," Communication 13 (1992): 95‒109.

1 This paper continues the alliance between content analysis and the mass media.

2 The concern for 'systematic analysis' is typical of content analysis

3 This content analysis has been informed by these accounts, therefore.  Go to the footnotes and take a look at what they were.

4 This paper assumes that 'what the public see' are the images carried in the mass media.  This assumes a lot about how people engage with the mass media – above all, it assumes that people pay attention to what the hear on the radio or see on television or read in newspapers and magazines.  Is this always the case?

5 Here, the authors are asking a question about the typicality of the most famous images to have emerged from the media's coverage of the Gulf War.  They are interested, not in the most famous images, but in the most frequent.  This is precisely the strength of content analysis.

6 The concern for a 'concrete indicator' celebrates the quantitative quality of content analysis.

7 Note the method here does not involve sampling.  The large number of photographs analysed was the entire population of images.

8 Note the discussion of why these three magazines were chosen.  Most of the discussion focuses on the magazines themselves – the site of the image – with just some comment on how they are produced and their large circulation.  No attention is given to how they are read.

9 'Manifest content' is what the case study of National Geographic magazine in Chapter 5 worked with: that is, what the photos actually show.

10 This case study also codes other aspects of each photograph: 'visual style, pictorial genre, and event context'.  There is no effort to code the expressive content of the photographs, though.

11 This case study also coded for where the photographs came from: 'current on-location photographs versus file or archival photographs'.  This suggests that codes can focus on a range of aspects of an image, not just its content.

12 This is important.  Griffin and Lee are suggesting that the content of an image alone is not always sufficient to code it successfully, and that for some codes, examining the text that surrounds it may also be necessary.

13 Griffin and Lee explain here that they started by inventing a list of coding categories – forty in all – that they expected to find in the photos they were analysing.  This is not the same as developing an initial list of codes from a theoretical position.  Note their three sources for doing this.  Are they appropriate, do you think? Should they have looked at other sources? 

14 This is standard content analysis procedure: try out the initial list of coding categories, refine the list, then apply it to the whole set of images.  In this case, both researchers coded all the photographs, with a high level of agreement between them.

15 Forty coding categories is probably towards the maximum it is possible to work with – any more and you may forget some, which will make your coding inconsistent.  In fact, it seems that as they coded, four were removed.  (Table 1 on page 817 only lists 36 coding categories.)

16 This paragraph clearly states the first key research question that drove the formulation of the coding categories. 

17 Here is the second key research question.

18 And here is the third.  These research questions are informed by the debates about what the US media showed during the Gulf War.

19 Take a look at Table 1 on page 817.  Examine the 12 categories.  Are they enlightening, exhaustive and exclusive?  Look also at footnote 10 where the 'low frequency coding categories' are listed.  Is there a clear break in the frequency of the categories listed in Table 1 and those listed in footnote 10?  Why has the break been made where it has, do you think?

20 Read footnote 9.  Again, this raises an important point: the size of the photograph.  Assuming that large photos have more impact, should a content analysis take account of the size of each image? 

21 As it turned out, their initial analysis is based on 12 of their original 40 coding categories. 

22 The content analysis here is supplemented by a relevant description of a sequence of photo inserts that appeared in all three newsmagazines.  Content analysis does not easily examine how one image relates to other images.  This is one solution: taking notes of the surrounding material as you analyse every image.

23 Can you convert Table 1 into a graphic?  A bar chart, or a pie chart?  Which is better at conveying the main results of this content analysis?

24 Note that Grifin and Lee do not undertake extensive statistical analysis in this paper; in fact, this is their only use of a statistical test.  Their analysis depends on simple frequency counts.

25 The most numerous images are 'photographic records of military hardware' in non-combat situations.  Note how military hardware in combat is excluded from this category: coding categories must not overlap, remember.

26 Footnote 14 is interesting here.  It points out some of the 'facts' of the Gulf War which they assume would generate some coverage.

27 A key finding.

28 Another key finding. 

29 Here, Griffin and Lee suggest that the different content of the images of political and military leaders is representing different political systems: 'a solitary dictator'  versus 'a more pluralistic image of American democratic consensus'.  Where does the notion of 'consensus' come from here?  Is it grounded in the content analysis?

30 Note the clarity with which the results of the content analysis are presented, and the structure of that presentation.  While differences between the three magazines are noted, the authors have moved from an analysis of three major codes to a broader account of a 'pattern of visualization' which depends on the results of the content analysis.

31 Creating a list of coding categories that they expected to see allows Griffin and Lee to reflect on the categories whose counts were 'surprisingly low'.  Using this approach, it seems that content analysis can in fact acknowledge the significance of infrequency, as well as frequency.

32 This point is underlined by this section on 'pictures not seen'.  Griffin and Lees point out that the US media claimed to provide comprehensive coverage of the war; their logic is then to think of the sorts of pictures that truly comprehensive coverage would generate, and use these as their coding categories.  This method clearly shows the lack of comprehensiveness in the US media.  However, there may be other situations when imagining what images should exist is trickier, or less helpful. 

33 The discussion here of a study of the images in US school history textbooks is interesting.  Griffin and Lee did not consider such imagery as relevant at the beginning of their study, and they did not include it in their list of the three sources they used to generate their initial coding categories.  It clearly is a useful account, though, for their interpretation of their results.  What sorts of coding categories might they have generated if they had considered this analysis of history textbooks at the beginning of their study – or indeed if they had looked at computer games?

34 The knowledge that Griffin and Lee have of the US press coverage of other wars allows them to identify a new aspect of the coverage of this one.  This element of their analysis is similar to the contextual knowledge brought to bear in the analysis of National Geographic magazine discussed in Chapter 5, and emphasises the need to have a broad understanding of the genre of the images that are undergoing content analysis. 

35 The explanation of the pattern this content analysis uncovers is rather brief.

36 The paper ends with no reflexivity from the authors.  They are, though, very clear about the methods they used to reach their conclusions.