The Problem with Presidential Databases

Abstract

There are many ways in which to study presidential communication from both humanistic and social scientific perspectives. From early study of presidential rhetoric (Aune and Medhurst 2008), to debates over rhetorical presidency (Ceaser et al. 1981; Tulis 1987; Medhurst 1996; Medhurst 2008), to generic studies of presidential oratory (Campbell and Jamieson 2008), to longitudinal studies of presidential style (Hart 1984, 1987), to content analysis of presidential discourse (Lim 2003, 2008), to examination of rhetoric's ability to affect public opinion (Edwards 2003), to constitutive function of rhetoric in formation of national identify (Beasley 2004; Stuckey 2004) scholarly attention to communicative dimensions of presidential leadership are ubiquitous. Into this plethora of scholarship step Coe and Neumann with a proposed database of what they label presidential speeches. Their work seems clearly to fall within a social scientific framework as they seek to standardize future analyses so that more meaningful generalizations can be made and more precise replications can be undertaken (Coe and Neumann 2011, 727). Insofar as standardization, generalization, and replication are hallmarks of standard social science, authors clearly seem to be on solid ground. But I wonder about rationale for a new database. only rationale that I see in their article is that proposed database will reduce absolute number of texts that scholars need to consult as they examine presidential rhetoric and that such a reduction and standardization will make generalizations easier to obtain. I do not doubt that such a claim is true. I do doubt its wisdom. First, there already exist at least two databases of presidential discourse--the Public Papers of Presidents of United States and American Presidency Project at University of California, Santa Barbara. Both of these databases have built-in search engines that allow for easy content-analytic searching. Coe and Neumann's database is, in fact, a subset of Public Papers. So only real advantage appears to be reduced amount of data (speeches) included in search domain. But is such a reduction really an advantage? If, for example, a scholar wanted to know what Jimmy Carter thought about Panama Canal Treaties, why would it be an advantage to search only nationally televised speech of February 1, 1978, when Carter gave several speeches on topic, at least one other of which was a address? It would not be an advantage. So if there is no advantage at level of individual president or particular issue, advantage must lie in institutional practice of rhetoric by presidency. But even at this level of analysis, kinds of payoffs envisioned--average word counts per speech, mean words per year, percentage of foreign policy versus percentage of domestic policy speaking--hardly seem worth candle. Worse, such analyses are almost always misleading. So I would like to see a stronger rationale, both in terms of need for a new database and in anticipated scholarly payoffs. Yet even if a stronger rationale could be provided, there are still several philosophical and methodological problems that must be addressed. Philosophical Concerns My concern starts with title of Coe and Neumann's article--The Major Addresses of Modern Presidents. Not or simply major addresses, or, better, selected addresses, but the addresses. One does not have to read very far to learn that the do not include some of most important in presidential history. So why insistence on the? Call me paranoid, but when rhetoricians see language deployed to define--to rope in or to rope out, as Kenneth Burke might say--we become suspicious. The addresses is a definition. Mutatis mutandis anything not included in database is not a presidential address. …