The Nixon Tapes


As part of the UCLA Fall 2017 Digital Humanities course taught by Dr. Miriam Posner, our team was given the Nixon Tapes database to analyze and visualize. The two main objectives of this project have been 1) put into practice digital humanities methodologies of handling data and 2) answer humanist questions regarding the Nixon Tapes.


President Richard Milhous Nixon has been one of the most polarizing figures in modern American history. From re-establishing relationships with China to his ultimate downfall during the Watergate scandal, President Nixon will always elicit a strong reaction from people. Our team wanted to examine his foreign policy decisions with regards to the Vietnam and Cambodia bombings during the peace negotiations in 1972.


This website is organized into 9 sections, each section highlights a visualization of the dataset. These visualizations are meant to be interactive, so we invite you to explore the dataset yourself. Was President Nixon a power-hungry aggressor or a peace seeking diplomat caught in a complex situation? Were his actions a result of his environment or his ambition? Click around to develop a conclusion yourself.

Conversation Analysis

Topic 1 Topic 2 Topic 3 Topic 4 Topic 5 Topic 6 Topic 7 Topic 8 Topic 9 Topic 10

After months of negotiations between the North and South Vietnamese, by December 14th, the Nixon Administration thought all hopes of negotiations have disappeared and a bombing campaign was necessary. From December 18 - 19th, the United States Air Force, Navy, and Marine Corps flew 3,420 sorties over North Vietnam including up to 120 B-52 sorties per day. During this time, the public became frustrated at Nixon for failing to deliver on his peace talk rhetoric from his presidential campaign.


Following his presidency, President Nixon attempted to defend this decision by stating it was his secretary of state, Henry Kissinger, who advocated for the bombing while he, President Nixon, wanted to continue to pursue peace talks.


To evaluate this claim, the Nixon Tape team analyzed the transcripts of 17 conversations between October 8th to December 28th. The following word cloud was build using analysis produced by topic modeling (described in the section to the right) of the conversations. Topic modeling analysis generated ten topics, grouping of text, and each word cloud represents one of the groupings weighted by relevance to the topic. The labels above each word cloud were determined by the Nixon Tape team based on the words, their relative importance to the topic, and the context of the conversations.

Topic Modeling

This conversation in particular is interesting because this took place just following Secretary of State Kissinger's negotiations with the North Vietnamese in Paris. This conversation was also held just after the Washington Post had reported that an FBI investigation had found evidence of a campaign of political sabotage by the Nixon forces that was "unprecedented in scope and intensity." Thus this visualization represents the context as well as each speaker's focus in the conversation.


The stacked bar chart represents how similar each speaker's speech in this conversation resembled one of the 10 topics that Mallet had identified. For instance, it is clear that Henry Kissinger and President Nixon's speech topics most closely aligned with "Appeasing on Agreements with Thieu."


This visualization was made by first parsing out the conversation by speaker into speaker text files using Python. Then each speaker text files were inputted into Mallet's topic modeling analysis to get the percentage likelihood of the speech matching up with each topic. Finally these percentages were visualized in Google Sheets chart.


US and North Vietnam begin formal talks in Paris

Earle Wheeler suggests authorization of bombings in Cambodia

Creighton W Abrams seconds Cambodia bombing authorization.

Map to the right displays modern day Cambodia where the bombing took place.

Final decision to carpet bomb Cambodia. Written in Haldeman's diary about "Operation Breakfast"

US Bombings in Cambodia

"Our dealings...will require reliability, precision, finesse. If we can master this process, we will have made a revolution." - Richard Nixon

Nixon plans to visit China

"I am simply saying that we expand the definition of protective reaction to mean preventive reaction." - Richard Nixon

Nixon arrives in China

"We'll blockade those sons of bitches and starve them out.” - Richard Nixon

||

"The deal we got, Mr. President, is so far better than anything we dreamt of.” - Henry Kissinger (10.12.1972)

"We know that the enemy's hurting, or they wouldn't be talking." - Richard Nixon (10.22.1972)

||

"If .. the United States...were not able to deal with the entity of Vietnam...with whom can the United States ever deal successfully?” - Fritz Kraemer (10.24.1972)

"We'll make our own deal and you'll have to paddle your own canoe." - Richard Nixon (11.18.1972)

"They're tough on the points that are almost insoluble." - Richard Nixon (11.29.1972)

"We've been playing with fire." - Henry Kissinger (11.30.1972)

||

"We've got to continue to crack it up there, so that they know we can still come back." - Richard Nixon (12.20.1972)

||

Conversation between Nixon and Kissinger, decrying the South Vietnamese leader's lack of cooperation, calling him "demented"

"I am certain, now he's coming along." - Henry Kissinger (1.21.1973)

"The settlement will last only as long as our two governments go forward together." - Richard Nixon (1.30.1972)

Nixon's Network

Vietnam Network
Nixon's Circle

In this Nixon-centric network analysis, records of conversations used from October 12th, 1972 to December 15th, 1972 are arranged into a one-mode network. These key dates were chosen because they mark the establishment of the first October Agreement, leading up to the final decision to begin operations for the December Bombings. The network centers Nixon as the focal point and measures the weight, or the amount of conversations held, between him and another person. In this way, each node is a singular person, with each edge representing Nixon and his or her contact through conversations. Due to an extensive list of names recorded in the dataset, parameters were set for a minimum amount of conversations with Nixon, to ensure a manageable and legible network. For this reason, each edge has a minimum weight of five conversations with Nixon.


Through the creation of this network, we were able to see that Stephen B. Bull, H.R. Haldeman, Alexander Butterfield, and Henry Kissinger were the persons with the most weight during this period. In borrowing the term from network analysis, we see that these conversations do actually hold "weight” in the realization of foreign policy decisions, which ultimately manifested in the resumption of mass bombings in North Vietnam that December. We approach this network in the context of these conversations' actual gravity in determining the United States further entrenchment into Vietnam, and the furthering of imperial domination in the Global South. The network functions as a simple tool for gauging who took up the most space in the President's inner circle at the time, perhaps translating to taking up some influence on the administration's decisions.

This bimodal network analysis uses edges to denote connections between key figures, with each node having been present in conversation(s) together. Because of the sheer volume of names in the data, conversations were selected according to our list of key dates and decisions leading up to the December Bombings. With these parameters we only included edges that had a weight above five, meaning that each actor had to at least participate in five conversations together. Using the two-mode method of analysis, we were able to visualize the interconnectedness and density of powerful members in Nixon's cabinet, as well as those who are more on the periphery.


The network again shows Stephen B. Bull, Henry Kissinger, and H.R. Haldeman as nodes holding the most weight. However, amidst those with broad-reaching authority in the administration are important members of staff who facilitate these conversations and connections, such as the White House operator, the White House photographer, and Nixon's personal valet, Manolo Sanchez, who has a significant between-centrality. The role these staff members play is essential in allowing the network to function.
Each network was made using OpenRefine, RStudio, D3.js, and Cytoscape.

Conversation Influences

Topic 1 Topic 2 Topic 3 Topic 4 Topic 5 Topic 6 Topic 7 Topic 8 Topic 9 Topic 10

Data Critique

Our team was first given an Excel sheet that shows the start and end dates and times, participants in the conversation, a brief description of said conversation, the location of the conversation, and the release date of the recording. There are other columns as well, but those are for identifying the physical tape itself as opposed to information about the content. The online additions to the dataset provide full transcripts of some conversations and the audio recordings too. The dataset can illuminate the context in which the recordings took place in addition to the geographical and temporal data it provides.


Additionally, the data was first recorded by way of a audio taping system set up by White House officials. The system recorded many important conversations concerning the Nixon Administration- with recording devices set up in the Oval office, the Cabinet Room, and the Aspen Lodge at Camp David-dating from February 16, 1971 to July 18, 1973. The National Archives has made the recordings readily available to the public through the creation of a standardized and open format. Such formating includes comma-separated values (CSV) files and zipped folders containing seven JavaScript Object Notation (JSON) files. Moreover, the staff at the National Archives used varying analysis techniques to structure the metadata according time, date, and location of each conversation (which total over 4,000 unique data rows). Interesting enough, the staff used audio translating software to generate writing script of select conversations in order to preserve accuracy of such accounts.


Furthermore, the original sources of the data are from the national archives of the US government. The White House Tapes contain sound recordings from President Richard Nixon's telephone calls recorded at the White House in Washington D.C. and the Presidential Retreat at Camp David, MD, recordings from meetings in the White House's Oval Office, Cabinet Room, President's office in the Old Office Executive Building, and Aspen Lodge at Camp David. The Richard Nixon Presidential Library and Museum is part of the National Archives and Records Administration, and is trying to digitize all of the tapes for public use.


Moreover, the organization of the dataset is arranged using the given parameters of time, sequence, location, participant in conversation, and collection. As a dataset generated by the National Archives and Records Administration, the presented ontology is intertwined with state power on the federal level. The dataset's relationship with the government informs its ideological framework, using parameters that are in line with the predominating, Western conceptions of time and location. Being a catalogue of conversations held by the president, the situated data is inherently ego-centric. Each entry point revolves around the president, as every record represents a conversation held by officials surrounding him during his administration. There are dense power dynamics contained in the relationships expressed in the data, as these conversations were responsible for policies, and decisions with immense impact.


Finally, if our dataset (the initial CSV file) was the only source, some information that would be left out would include the audio recordings, transcripts, and political context of the dataset. While the existence of these tapes say a lot about President Nixon's own insecurities and paranoia within his own administration, the CSV alone does not fully convey these emotions and the user truly gets a better sense by listening to his voice and mannerisms as well as reading through transcripts. Who President Nixon talks to and how he talks with each individual often reflect the stress that he was under that the original CSV file did not since it was primarily made for record-keeping purposes for the Richard Nixon Presidential Library.

Credits

Digital Humanities 101 Team

Special Thanks

  • Cary McStay - Tapes Supervisory Archivist | Richard Nixon Presidential Library and Museum
  • Dave Shepard, Ph.D - Lead Academic Developer | UCLA Digital Humanities Department
  • Miriam Posner, Ph.D - Assistant Professor | Information Studies and Digital Humanities

Sources