We execute our strategy of comparing explanations drawn from our theory to those drawn by generalized families of competing explanations, as well as rival explanations driven by idiosyncratic features of the specific cases, in an analytically similar but empirically different manner. In both cases, we test our theory’s explanatory power against rival explanations using a framework drawn from recent advances in the qualitative literature; we refer to this strategy as a “folk Bayes” approach (detailed more in the Supplementary Information) and argue that it allows us to isolate observations in which our theory better predicts observance (or lack of observance) of clues associated with the case compared to rivals given intuitively plausible prior beliefs.
The differing levels of documentation available in each case (even before considering language difficulties) led to different strategies for collecting evidence to allow for this testing. In the Ming case, we drew on secondary sources. This reflects the fact that neither author reads Mandarin nor any other language implicated in the Ming treasure-fleet voyages. It also derives from the fact that we understand that even much of the “primary sources” available in Mandarin are themselves secondary sources (e.g. the Ming Shilu 明实录 or “Veritable Records of the Ming Dynasty”, an official record of the Ming dynasty compiled by scholar-officials after the death of each emperor). As discussed in the case, many of the “primary sources” may have been lost not merely to time and the dislocations associated with dynastic successions in e.g. 1644, 1911, and 1949 but to specific bureaucratic sabotage during the later Ming dynasty.
The absence of such primary documentation and access to original-language literature meant that we were reliant on English-language sources. Fortunately, these include a vast array of specialist tracts. Instead of relying on standard popularizations (e.g. Levathes 2014, When China Ruled the Seas; 436 Google Scholar citations), we used works by scholars of China and Chinese history (e.g. Dreyer 2007; Needham 1971; Cham 1988, 2007; Finlay 1991). We believe that this allowed us to better survey disputes over interpretations of the voyages’ meaning and impact; furthermore, our more capacious selection mitigates the problems mentioned by Lustick (1996). We relied most heavily on those sources that themselves seemed to be closest to archaeological, documentary, and other more-primary records.
In the Kennedy case, we benefitted from greater availability of documentary records. We consulted the secondary literature (the work of John M. Logsdon and Walter A. Macdougall, as well as a dissertation by Teasel Muir-Harmony, was particularly helpful). We also consulted contemporaneous media and other sources. However, we viewed these works more as a guide to initial surveys in archival and other primary-source research. We conducted searches for relevant records in compilations of government records such as the Foreign Relations of the United States series (for both the Eisenhower and Kennedy administrations) and The American Presidency Project as well as searches among the CIA’s online reading room of declassified documents. Furthermore, we employed records held by the John F. Kennedy Library. Many of the Kennedy Library’s holdings have been digitized, but some have not. To collect those records that were unavailable, I visited the library in November 2016 to find and photograph relevant documents. (This required only a day of research, as opposed to the much longer periods that would have otherwise been necessary, because of the volume of records available online already, because of the narrow nature of our research interests, and because most of the processing of those records occurred offsite after my visit.) Documents were saved for later reference as computer files and prepared for sharing where necessary by converting them to PDFs. Finally, the processing of the Kennedy Library’s tapes allowed us to consult an unusually rich vein into the president’s thinking; the volumes compiled by the Presidential Recordings Project at the Miller Center of the University of Virginia helped us understand how the Moon project fit into the president’s thinking.
Our exposition of our research on the Kennedy case was limited in the text by the demands of the journal publication process. Consequently, as with the Ming case, much of our argumentation and fuller expositions of our findings are presented in a Supplementary Information.
References in this Data Overview
Levathes, Louise. 2014. When China Ruled the Seas: The Treasure Fleet of the Dragon Throne. Open Road Media.
Lustick, Ian S. 1996. History, Historiography, and Political Science: Multiple Historical Records and the Problem of Selection Bias. American Political Science Review 90(3):605-618.
Supplementary Information
This article’s online appendix is available at
International Organization, on
Harvard Dataverse, and http://www.paulmusgrave.info/