In Defense of Not Having a Data Set: A Call for Argument

Abstract

Kevin Coe and Rico Neumann are correct: there is no clear agreement on exactly which texts should be considered major presidential addresses of the modern era. Rather than viewing this state of affairs as creating a need for a standardized data set to be used by scholars interested in longitudinal content analysis, however, I would suggest that it is the very reason we should be careful about promoting one. I would rather we argue. I would prefer that scholars continue to be compelled to construct arguments and offer justifications for what they are examining when they say they are analyzing presidential discourse, regardless of methodology. There are two reasons for my stance. First, presidential communication is inherently difficult to define and/or delimit. It always has been, and I suspect that this difficulty will only increase over time, and thus the idea of fixed parameters may be more problematic with this type of discourse than with some others. Second, scholars of presidential discourse should be wary of the possible creation of a new canon, even if, as in this case, the intentions behind such invention are admirable. As debates in the humanities over the last 30 to 40 years have revealed, this potential is worrisome, especially for those of us interested in comparative work. In this response I will explain both of these reasons and then conclude on a final note that might seem like a contradiction: even if Coe and Neumann's argument is persuasive to me, and it is, I would ask us nevertheless to be very careful about promoting one data set over other possibilities. The institution of the presidency is more complex than it can appear, as readers of this journal know better than most. The public communication issued in its name is as well. On one hand, it is enduring, with deep roots in history, institutional practice, and generic constraints (i.e., an inaugural address can be expected to sound much different from a prepared statement delivered at a press conference). On the other hand, it is also perpetually changing due to the variety of communication skills and predilections of the individual office holders as well as the contingencies inherent in the presentation of unusual and/or unprecedented developments within its subject matter (e.g., national or international affairs), intended audiences, and potential media outlets. To my mind, the tension between those aspects of presidential communication that provide stability and continuity, and thus a basis for comparison between presidents, versus those that test individual presidents' capacities for invention and/or adaptation is fundamental to the object of study itself. That is, this tension is not just a descriptive feature of the discourse but is instead one of its basic characteristics; presidential communication is typically animated by a calculated negotiation between the systemic and the situational. This is the first reason behind my call for caution in response to Coe and Neumann's proposal. What seems institutionally stable at one time may be troubled, or at least complicated, by the unpredictable and/or the idiosyncratic at another, and focusing only on the most predictable forums for presidential discourse could mean that we lose sight of this complication and its entailments. Allow me to illustrate with two possible examples of how this unpredictability can play out in ways that would not be recognized by the proposed data set. Coe and Neumann correctly identify one of the recurring issues for scholars interested in the history of presidential discourse: what to make of the introduction of orality to the issuance of State of the Union messages. But this is not the only instance in which changing institutional practices can impact the study of presidential discourse. For students of the contemporary presidency, what are we to make of the increased use of social networking media as a means of enhancing the reach of presidential messages? …