Editorializing the Cold War: Cartoons and Commentary on Nuclear Fear and Anxiety, 1945-1989



Norman Cousins once wrote, “The beginning of the Atomic Age has brought less hope than fear. It is a primitive fear, the fear of the unknown—the fear of forces man can neither channel nor comprehend.” 1 In the United States, this fear and anxiety has manifested itself around various events and issues since the bomb was first dropped on Japan in August 1945: fears of fallout and radiation, concern over the dangers posed by nuclear power, anxiety caused by the possibility of more countries gaining access to the technology needed to create these weapons, consternation over the thought of those weapons being used against Americans, and disquiet about their ability to protect themselves from such an assault.

Some authors have examined these issues from the vantage point of the political leaders involved in the Cold War battle for the hearts and minds of the world’s population. Other scholars examined the cultural production of those populations as they struggled with the knowledge that man, not just Mother Nature, could end all life on Earth. In the epilogue of his seminal work, By the Bomb’s Early Light: American Thought and Culture at the Dawn of the Atomic Age (1985), historian Paul Boyer claimed that since 1945 the American public experienced cycles of nuclear awareness and amnesia. This theory is presented with a smattering of cultural evidence drawn from film, television, literature, and newspapers to argue that 1945-50, 1955-63 and 1979-85 were periods of intense nuclear awareness while 1950-55, and 1964-78 were years of nuclear apathy. 2 Nuclear historians, both cultural and political, often utilize the same periodization Boyer outlined in his theory, with 1964 and the passing of the Partial Test-Ban Treaty as a demarcation line between the “early” Cold War and the “late” Cold War. In addition, these historians often gloss over the years of 1963 to the late 1970s, what Boyer originally termed, “The Big Sleep.” Citing the passing of the Test-Ban Treaty when nuclear tests literally went underground and the growing preoccupation with the situation in Vietnam, the Civil Rights movement, and a misplaced faith in deterrence theory, Boyer argued that these years were ones of nuclear apathy and cultural neglect. 3 This project is aiming to add nuance and complexity to Boyer’s rough periodization by focusing, not on whether there were periods of apathy and neglect, but by taking a broad view of the entire Cold War and focusing on the context and content of what issues and events were discussed.

In “Editorializing the Cold War: Cartoons and Commentary on Nuclear Fear and Anxiety, 1945-1989,” I will analyze the work of three nationally syndicated cartoonists: Herbert Block aka Herblock, Paul Conrad, and Frank Miller, and the newspapers in which their work appeared (the Washington Post, the Denver Post and Los Angeles Times, and the Des Moines Register, respectively). Using a combination of traditional close reading and new digital methodologies, my aim is provide a comprehensive analysis of Cold War nuclear anxieties by examining the correlations between the discourses imbedded in the editorial cartoons, the letters to the editor and opinion pieces in the editorial page, and the articles published in the newspaper itself.

One of the advantages of this long view of the Cold War is a new sense of the reliability of editorial cartoons 4 as representative of a societal discourse. Editorial cartoons of the 19th century and earlier have a strong tradition of being utilized by historians as sources for that period’s political thought and culture. This critical reading of cartoons becomes nonexistent after World War II with the advent of new visual culture mediums: film and television. For most 20th century historians, editorial cartoons have been misused; introduced, as historians Richard Scully and Marian Quartly note, “as a kind of decoration, to break up the text and give an impression of historicity, rather than adding to the historian’s argument.” 5 Treating cartoons as texts, as cultural artifacts, means that historians will benefit by opening up a hitherto neglected source of historical inquiry. By analyzing the editorial page — with particular emphasis on the cartoon as the focal point — against the articles published in the newspaper itself, I hope to provide a broad understanding of Cold War nuclear fear and anxiety and to trace the content of these disparate discourses as cartoonists, editors, and the newspapers’ readers negotiated their fears and aspirations.

Research Questions and Methodology

This project aims to examine a broad view of the Cold War. In doing so, it will address three primary questions: What can an examination of the entire Cold War tell us about the discourses surrounding nuclear fear and anxiety? Is there any regional disparity to this larger view? How are editorial cartoons as evidence indicative of a societal discourse? For this project, methodology plays a key role in driving the historical inquiry. As such, I am combining my research questions section and methodology section. For clarity I will discuss each methodology in relation to the research question or corpus to which they are primarily integral, but that does not mean that these techniques will only be utilized in that sole instance. All methodologies, whether digital or analog, will be used in some capacity for each portion of this project.

My first step in this project is to gather all the articles, editorials, and letters to the editor from my cartoonists’ newspapers that reference nuclear events and issues. Two of the newspapers, the Los Angeles Times and the Washington Post, have already been digitized and are available in PDF format from ProQuest. The other two newspapers, the Denver Post and the Des Moines Register, are only available on microfilm. A crucial and fundamental step for all the digital methodologies I utilize is converting these documents into a plain text format. While technology is available to assist in this process, there is no industry-accepted procedure to accomplish this goal. To get these texts manageable I will be experimenting with different techniques to create an acceptable methodology that can be followed for future scholars interested in pursuing this type of research.

These texts, once gathered, provide the core of my primary research for investigating nuclear fear and anxiety during the Cold War. A key methodology I will use for this analysis is topic modeling. 6 As historian Ted Underwood explains, “topic modeling is a way of extrapolating backward from a collection of documents to infer the discourses (‘topics’) that could have generated them.” 7 Each document can contain many topics, and a generated topic model is a list of keywords that the historian can analyze and label. I can go further with this technique and use topic modeling in relation to networks, which allows a researcher to see, as Scott Weingart writes, “how documents relate to one another, how they relate to topics, how topics are related to each other, and how all of those are related to words.” 8

By parsing the articles for each newspaper by decade and running each set of texts through topic modeling programs, I can see topic variation over time. In contrast to methods that merely present word frequency counts, topic modeling provides complexity by grouping words together and can produce other topics — groups of keywords — that may not be otherwise apparent. This presents a unique opportunity to also explore the traditional themes utilized by nuclear culture scholars. These themes can be broadly categorized as radiation and fallout, defense, proliferation, the anti-nuclear movement, international talks and treaties, and nuclear energy. Topic modeling provides the ability not only to gain a new perspective on Boyer’s periodization, but also to add complexity to the categorical structure utilized in this scholarship.

A distant reading of the Cold War can also highlight any regional disparity. While newspaper articles can be drawn from numerous sources across the country, local staff and readers generate the letters to the editor and the editorials. By running these texts through topic modeling programs independently from the newspapers’ articles I hope to highlight any regional differences in the content and context of discussions pertaining to nuclear fear and anxiety. To better asses the contours of nuclear themes and discourses as shown through the editorial page, we can also consider the relationships and changing pattern between keywords. This can be done by examining the editorial cartoons as a body of texts alone and in juxtaposition to the letters to the editor and editorials.

First, all the nuclear themed cartoons for each artist will be entered into a database of my own design. 9 Each entry will consist of the artist’s name, newspaper in which the cartoon was published, date of publication, caption of the cartoon, a brief two-three sentence description of the cartoon, a thumbnail image, and associated metadata relevant to this project. 10 This metadata consists of three parts: any characters depicted in the cartoon, e.g. Richard Nixon or the Democratic Donkey, the specific historical event the cartoon is referencing, if applicable, and keywords associated with the theme of the cartoon. Once completed, the database 11 will be run through a program called Gephi, which is a free, open-source software for visualizing and analyzing large network graphs, and exploratory data analysis. Gephi is a useful tool for charting the cartoons’ keyword frequency over time, and will assist in visualizing if and how many times certain keywords appeared together in the same cartoon for each artist during each decade. For example, in Herblock’s cartoons from 1946-1949 the keywords “atomic bomb” and “United Nations” appeared together in eleven cartoons. 12 This is a visual representation of the importance of the political movement during this time period calling for a One World Government and international control over the atomic bomb, essentially removing America’s monopoly on the technology.

These digital methodologies are invaluable to a researcher examining a large corpus of texts over an extended period of time. Their function lies in providing a unique view of the data, offering new ways to examine, read, and discover clues hidden in the texts. They also highlight features that merit closer examination. These features can be an easily explained result like the correlation between “atomic bomb” and “United Nations” discussed above, or they can be anomalies, unexpected relationships that appear during a time period where no obvious historical event would indicate its presence. In either case, the features merit a close reading of the texts to determine the context and content of the discourse present in the visualizations. Drawing upon the works of Peter Burke, David Freedburg, and Marita Sturken and Lisa Cartwright, visual culture theory is integral to my analysis of the cartoons and their imagery. 13 While communications studies like Greg Barnhisel and Catherine Turner, Pamela Shoemaker and Stephen Reese, and Rodger Streitmatter, will deeply influence my examination of the newspaper articles. 14

When topic modeling, keyword visualization, and close reading is applied to the texts of the editorial page — the cartoons, letters to the editor, and editorials — I can trace whether certain events and issues were more dominant in one region, the East, Midwest, or West, at one time versus another. For example, were Herblock and his readers more attuned to issues surrounding the passing of various international treaties and the unfolding events at summits like Reykjavik since they were located in Washington, DC, the heart of political power? Does the topic modeling reveal a greater concern with defense, both civil and military, in Iowa given that the Midwest was the location of the majority of the United States’ missile silos and this made that region a high priority target for Russian missiles? Were West-coast readers disproportionately preoccupied with the effects of nuclear fallout and radiation due to their proximity to the Nevada Test Site and the Marshall Islands? And, finally, what exactly were the editors, cartoonists, and readers concerned about within these larger issues and how did they express their anxiety?

The process outlined above can also illuminate the relationship between the cartoonists and the editorial page. Noted cartoon historian Charles Press argues that there is a loyalty that drives all cartoonists’ vision. That loyalty is “less to a set of ideas than to a particular community of individuals that the cartoonists thinks need nurture.” 15 In my project I argue that each artist had a specific community of individuals to whom they felt responsible, and this influenced their political vantage point and subsequently the message in their cartoons about nuclear issues. By utilizing text mining techniques like n-gram analysis I can determine whether a particular keyword set from a cartoon appears in the editorials the next day, or vice versa, or both. Text mining is particularly useful for examining this relationship but will be used in conjunction with the other techniques and methods discussed above. N-grams are generated by filtering the plain text files of the editorials and letters to the editor through a programming script that return snippets of text. These snippets consist of a specified set of words from either side of a designated keyword. For instance, I can return five words in either direction of the keyword “atomic” and examine them to see if the term “radiation” or “fallout” occurs with more frequency next to “atomic” during a certain period of time in a specific newspaper than another. N-gram analysis provides a closer reading of the context of keywords than topic modeling and when used together, I can begin to postulate if there was any influence on the readers’ discourse by the cartoons or vice versa.

The length of the time period this project examines can also provide an opportunity to discuss whether the topical/thematic correlation between texts and images can assist in determining whether editorial cartoons are representative (or not) of other editorial work. While any art historian or visual culture theorist will agree with cultural historian Jacob Burkhardt’s assessment that images are “witnesses of past stages of the development of human spirit, [objects] through which it is possible to read the structures of thought and representation of a given time,” 16 a study of 20th century editorial cartoons on this scale has not been done before. It is much more common to find 19th century cartoons being examined en masse, although usually in respect to a single artist, theme, or topic. With the advent of radio, film, and television, the editorial cartoon has ceased to be recognized by historians as potentially representative of a societal discourse. The scope and methodological processes of this project enables me to form a preliminary conclusion as how 20th century editorial cartoons should be treated as sources of historical inquiry: are they cultural artifacts that can stand in to represent a broad section of American thought and culture? Or, are they the viewpoints of a single artist, and should be examined in conjunction with other sources in any effort to understand a society’s thought and culture?

This project is seeking to reexamine the Cold War through a wide-angle lens by analyzing the correlation between texts and images. The digital techniques discussed above offer the opportunity to represent and analyze this relationship on an entirely new scale, which in combination with close readings of the texts and images, will provide a new perspective for examining the context and content of various discourses surrounding Cold War nuclear fear and anxiety.



  1. Robert A. Jacobs, The Dragon’s Tail: Americans Face the Atomic Age (Amherst: University of Massachusetts Press, 2010), 11.
  2. Boyer revisits this theory in an article entitled, “Nuclear Themes in American Culture, 1945 to the Present,” published in The Atomic Bomb and American Society. Despite the almost twenty-five year gap between the two publications, the theory changed little. Instead of the cycles of nuclear awareness and amnesia he wrote about previously, Boyer describes “three cycles, each marked by a surge of political activism and cultural expression, followed by an interval of comparative quiescence and diminished attention.” These new cycles date from August 6, 1945, to the mid 1950s, the mid 1950s to the late 1970s, and finally from the late 1970s to the end of the Cold War.
  3. Paul Boyer, By the Bomb’s Early Light: American Thought and Culture at the Dawn of the Atomic Age (New York: Pantheon, 1985), 357.
  4. The terms “editorial cartoon” and “political cartoon” are often used interchangeably. They commonly refer to single-panel cartoons located on the editorial or op-ed pages of a newspaper. Other forms of cartoon can have “political” messages, like the comic strip Doonesbury, but are not considered editorial cartoons per se. To prevent confusion I will henceforth use “editorial cartoons” unless replicating a direct quote.
  5. Richard Scully and Marian Quartly, “Using Cartoons as Historical Evidence,” Drawing the Line: Using Cartoons as Historical Evidence, Richard Scully and Marian Quartly, eds (Clayton, Victoria: Monash University ePress, 2009), http://www.epress.monash.edu/dl/.
  6. The primary topic modeling program I will use is called MALLET. There are other programs available that I will experiment with and the technology changes quickly so other programs may be created while I am working on this project that I can use.
  7. Ted Underwood, “Topic Modeling Made Just Simple Enough,” The Stone and the Shell, accessed March 17, 2013.
  8. Scott Weingart, “Topic Modeling and Network Analysis,” the scottbot irregular, November 15, 2011.
  9. My initial round of cartoon entries will consist of cartoons that explicitly refer to or represent nuclear issues. I include in this any cartoon that has the worlds “nuclear” or “atomic” in the caption or labeled in the cartoon, any cartoon that has the atom symbol in it, and any cartoon that depicts a specific nuclear technology. I do not include any cartoons that just contain an image of a bomb that is not explicitly noted as being nuclear in nature.
  10. Due to copyright restrictions, the thumbnail of each cartoon will only be visible to myself for my own research until I can gain the appropriate permissions to make it public.
  11. While I am building this database for my own research, I am doing so with the eventual goal of making it available to the public to use. Advanced searching capabilities will not only assist me in quickly finding a cartoon I need to examine, but will be indispensable for other researchers and interested users to explore the cartoons.
  12. For a detailed description of my methodology for generating this statistic and a copy of the visualization presenting this data, see http://nuclearcartoons.sashahoffman.org/1940s.php
  13. David Freedberg, The Power of Images: Studies in the History and Theory of Response (Chicago: University of Chicago Press, 1991); Peter Burke, Eyewitnessing: The Uses of Images as Historical Evidence (Ithaca, NY: Cornell University Press, 2001); Marita Sturken and Lisa Cartwright, Practices of Looking: An Introduction to Visual Culture, 2nd ed (New York: Oxford University Press, 2009).
  14. Rodger Streitmatter, Mightier Than the Sword: How the News Media Have Shaped American History (Boulder: Westview Press, 1997); Michael Parenti, Inventing Reality: The Politics of News Media, 2nd ed (New York: St. Martin’s Press, 1993); Greg Barnhisel and Catherine Turner, eds., Pressing the Fight: Print, Propaganda, and the Cold War (Amherst: University of Massachusetts Press, 2010).
  15. Charles Press, The Political Cartoon (London: Associated University Press, Inc., 1981), 63.
  16. Burke, Eyewitnessing, 11.