James S. Bielo, Why Some Theories Work and Others Do Not

Why Some Theories Work and Others Do Not

Analyzing Discourse in the First Bush-Kerry Debate

(or, Why George Bush is Incredibly Articulate…for real)

James S. Bielo

On September 30th, 2004, George W. Bush and John Kerry met at the University of Miami for the first of three presidential debates[1]. PBS news anchor Jim Lehrer moderated as Bush and Kerry answered questions about foreign policy and homeland security for 90 minutes. Both candidates consented to a 32-page agreement of rules and procedures for the debate[2]. In his opening remarks, Lehrer listed a few of these, as well as the expectation for those in attendance:

“For each question there can only be a two-minute response, a 90-second rebuttal and, at my discretion, a discussion extension of one minute.”

“Candidates may not direct a question to each other.”

“There will be two-minute closing statements, but no opening statements.”

“There is an audience here in the hall, but they will remain absolutely silent for the next 90 minutes.”

The highly scripted, highly regimented nature of this “debate” invited deserved criticism from the media and the general public[3]. Is it really a debate if the participants are not allowed to address each other directly? Does this format facilitate an educated electorate, or an attempt to preserve candidates’ self-images? The debate was, indeed, infuriating at times (for some, I imagine, the entire time). Yet, I warmed the seat in front of my television for the full 90 minutes, as did millions of others.

The morning following the debate I was in the mail/coffee room at the anthropology department of Michigan State University. I was filling up my first cup of the day when David Dwyer came in to heat up water for his own, flavored brand. David was a member of my dissertation committee, and a mentor of mine in social and linguistic theory. We also shared a common interest in political discourse – as polemic, rhetoric, ideology, and hegemony. Our conversation that morning proceeded directly to our thoughts about the Bush-Kerry debate. We talked for several minutes, and he eventually suggested we spend some time reading a transcript of the debate. I eagerly agreed.

The consensus in broadcast and alternative medias following the debate was that Kerry scored an impressive victory[4]. Yet, just over a month later, George W. Bush was re-elected to the U.S. Presidency. Following the election, I received an email from David with an attachment – a transcript of the first debate. He wanted to pursue his suggestion; in part, to glean some insight into the apparent dissonance between debate victory and election loss. We began meeting weekly – at the library, his office, and a local pub – to exchange observations and possibilities for analysis. This paper chronicles our attempts to apply theoretical frameworks in discourse analysis to the Bush-Kerry debate. After several (relatively) unsuccessful analyses we found one framework particularly insightful, both for the nature of this speech event and for the possibility that George Bush is incredibly articulate…for real.

Narrative Structure, Intertextuality, and Semantics

After our initial readings of the transcript, Bush and Kerry’s responses began to resemble narrative units. With each reading, the arrangement of their respective answers appeared increasingly similar. We decided to employ William Labov’s research on narrative structure to observe the logic of arguments and the organization of talking points.

Labov (1972) argued that spoken narratives have a beginning, middle, and end. The typical narrative structure includes six components: abstract, orientation, complicating action, evaluation, resolution, and coda. The abstract introduces the story by way of summarization. The orientation provides the necessary background information. The complicating action provides the general problem or situation being dealt with. The evaluation surmises the general point of the narrative, and the reason for its telling. The resolution explains what ultimately happened with the complication action. And, the coda signifies the end of the narrative. This is a progressive model, wherein the speaker moves through the components as stages[5].

Analysts have used this framework to better understand, news stories (Bell, 1991), narratives of religious conversion (Stromberg, 1993), and expressions of identity through storytelling (Lefkowitz, 2004). The application is rarely an exact replication of Labov’s components, but an attempt to identify an organizing structure that guides how particular genres unfold. In this wise, Dwyer and I were interested in what narrative structure was present in the debate responses of Bush and Kerry.

The first step in this process was to code the individual statements that comprise a particular response. Ultimately, we identified six types of statements: assertions, supporting statements, facts, positive evaluations, negative evaluations, and codas. The following examples from Kerry’s response to the first question demonstrate each type:

Assertion: “I believe America is safest and strongest when we are leading the world and we are leading strong alliances.”

Support (for the above assertion): “I’ll never give a veto to any country over our security.”

Fact: “We’re now 90 percent of the casualties in Iraq

and 90 percent of the costs.”

Positive Evaluation: “I have a better plan for

homeland security [than the president].”

Negative Evaluation: “This president has left

[alliances] in shatters across the globe.”

Coda: “[We can do a better job in] all of these, and

especially homeland security, which we’ll talk about a little bit later.”

We sampled the transcript to come up with this list, working inductively with the first four questions. Few responses contained all six, but every response was comprised of at least two of these statement types. Several questions could now be posed in regard to narrative structure. Are there patterns of organization within responses? Do certain statement types typically occur together, or separately? Do Bush or Kerry make particularly frequent use of, or avoid, certain types?

We quantified the coded statements by response and speaker, and then derived several possibilities for how the statements might be organized. We arrived at two conclusions. 1) Bush and Kerry made relatively equal use of the various statement types. 2) Bush and Kerry organized their answers in very similar ways, with a narrative structure of assertion -> supporting evidence (via facts and/or evaluations) -> coda to provide a neat conclusion. The second find is mildly interesting, but quite unremarkable. This is an expected organizational structure and reveals little about the larger discursive style of the debate. And, more important, it suggests nothing about why Kerry or Bush “won” the debate. In short, narrative structure was not a very productive framework in this case.

Not to be deterred, we returned to the transcript to consider other analytical possibilities. Over the past two decades, linguistic anthropologists have become increasingly interested in the theoretical insights from the Russian literary critic Mikhail Bahktin (1934). Namely, the concept of meaning as “dialogical” has been used to counter structuralist presuppositions about signification processes. A dialogical approach suggests that utterance meaning is inherently historical and predictive, riddled with the baggage of previous and expected uses. Julia Kristeva (1986) pursued this approach in her analysis of literary texts, identifying how a single text is the result of, carries the imprint of, and incorporates multiple texts - what she termed “intertextuality.”

Discourse analysts identify multiple forms of intertextual practice that social actors use in strategic, culturally significant ways. One form in particular has yielded insights about the dialogical quality of spoken discourse. “Reported speech” refers to the practice of blending another person’s speech into your own speech. This can be done “directly,” where the change in author is obvious, or “indirectly,” where the change is more subtle (Volosinov, 1929). Deborah Tannen observed this phenomenon in everyday interactions and argued that ‘reported speech’ is more like “constructed dialogue, that is, primarily the creation of the speaker rather than the quoted party” (1989: 99). As constructed dialogue, the reporting speech provides some meta-commentary on the reported speech (e.g., its veracity or authority). The goal is not so much to report another’s speech accurately or faithfully, but to use their speech in order to make a statement about one’s self or the person whose speech is being reported.

For the Bush-Kerry debate, the initial question was simple: how does each participant use the intertextual strategy of reported speech? In our readings of the transcript, both candidates clearly invoked other speakers’ words on numerous occasions. Our analysis would question what, if any, pattern or significance there was to the use of this strategy.

Bush and Kerry each reported speech in 31 instances. Kerry quoted a wider range of speakers, including: Bush prior to and during the debate, “the Bush administration,” Jim Baker, General Scowcroft, “war families,” “young returnees,” General Shinsheki, “the terrorism czar,” Prime Minister Allawi, John F. Kennedy, Charles DeGaulle, Colin Powell, and George Will. On the other hand, Bush reported the speech of Kerry prior to and during the debate, “the Kerry campaign,” Ambassador Negroponte, “the troops,” Missy Johnson, and Colin Powell.

By far, the most commonly reported speech by Bush and Kerry was from Bush and Kerry. Kerry, on 15 occasions, and Bush, on 21, invoked the speech of their opponent in an effort to criticize their policy and/or their character. They each had their favorites. For Bush, he used various iterations of “The troops are not going to follow somebody who says, ‘This is the wrong war at the wrong place at the wrong time.’” Kerry did the same with, “The president said he was going to build a true coalition, exhaust the remedies of the U.N., and go to war as a lost resort.” But, the use of a political opponents speech to argue against them is hardly a revealing finding. The question lingered, do Bush and Kerry use reported speech strategically, in ways that resonate with larger rhetorical goals?

The short answer to this question is yes. For example, Kerry framed the reported speech of Bush as a “promise” on three occasions, as in the following:

“The president, in fact, promised [the armed forces]. He went to Cincinnati and he gave a speech in which he said, ‘We will plan carefully. We will proceed cautiously. We will not make war inevitable. We will go with our allies.’”

A “promise” is an archetypal case of what J.L. Austin (1962) famously called a “speech act;” a class of utterance that does not just describe, but actually constitutes reality. In her own analysis of political discourse, Jane Hill (2000) suggests that promises are a significant linguistic category because they are equated with moral character. Fulfilling a promise reveals integrity, while breaking a promise reveals ethical negligence. Thus, to frame someone’s speech as an unfulfilled promise is to make an assertion about who they are as a person. In this sense, Kerry’s reported speech of Bush is carried out strategically because he delivers a critical remark without explicitly stating it.

Bush also used reported speech in creative ways. Consider his representation of Kerry in the following:

“I decided the right action was in Iraq. My opponent calls it ‘a mistake.’ It wasn't a mistake. He said I ‘misled on Iraq.’ I don't think he was misleading when he called Iraq ‘a grave threat’ in the fall of 2002. I don't think he was misleading when he said that ‘it was right to disarm Iraq’ in the spring of 2003. I don't think he misled you when he said that, you know, ‘anyone who doubted whether the world was better off without Saddam Hussein in power didn't have the judgment to be president.’”

In this example, Bush pieces together a series of quotations from Kerry. The final case of reported speech – “anyone who doubted…” – is the most strategic. Bush uses the deictic referent “anyone” to position Kerry as the subject of his own critique against people who “[don’t] have the judgment to be president.” This resonated with Bush’s campaign accusations that Kerry sends “mixed messages” and is a “flip-flopper.” Much like the use of “promise,” this calls character into question, not just politics.

Our analysis of intertextuality was somewhat more productive than the application of narrative structure. Still, as an analytical framework, reported speech is lacking in this case. Observations about promises and such are interesting, but only in sporadic fashion, providing no coherent statement about the debate discourse or the reasons for “victory” and “loss.” Encouraged, but tiring, Dwyer and I moved on to a third theoretical framework.

We hoped to identify the use of key words by Bush and Kerry. In particular, we were interested in words that resonated with and indexed larger discourses that the two candidates wanted to engage. Jay Lemke (1995) refers to this relationship between lexical items and societal discourses as “textual semantics.” He uses the following example to demonstrate:

“The freedom fighters are being kept in a concentration camp. [versus] The terrorists are being held in a prison” (ibid: 37).

These two statements can describe the same empirical situation, but they constitute the meaning of that situation very differently. “Freedom fighters” and “terrorists,” “concentration camp” and “prison,” each resonate with different audiences. This can be a very effective strategy for communicating with and appealing to particular segments of the voting public, and for subtly and efficiently conveying ideologies[6]. Our goal was to see if any such cases were operative in the Bush-Kerry debate.

We identified only a few instances. For example, Kerry used “outsourcing” to resonate with his critique of Bush’s economic policies, as in the following:

“Unfortunately, [Osama Bin Laden] escaped in the mountains of Tora Bora. We had him surrounded. But we didn't use American forces, the best trained in the world, to go kill him. The president relied on Afghan warlords and he outsourced that job too. That's wrong.”

Bush found several ways to connect with the conservative discourse of American patriotism, invoking “terrorist,” “freedom,” “America,” and the like throughout the debate. However, he did so with relatively the same frequency as Kerry. Evoking nationalistic sentiments was the business of both candidates, not something that distinguished Bush’s talk.