ISSA Proceedings 2002 – Rhetorical Argumentation And The New Journalism: A Case Study
1. Critical thinking in journalism: the old model
One way to analyze argumentation in journalism is to use the tools of logic to evaluate the traditional medium of journalism – the printed word. Editorials and other forms of commentary are transformed into logical arguments that are checked for validity, fallacies and dubious premises. We argue that this approach is no longer adequate because journalism has changed. An increasing amount of journalism is on the Internet where journalists use many forms of media and new technology to inform and persuade. Internet articles debate public issues by combining text, audio and video, supplemented by hyperlinks and ‘chat’ forums. Articles, written with specific audiences in mind, are embedded on web sites surrounded by persuasive elements, from background information to dynamic images. Sites encourage readers to go back and forth between the various levels and components of the presentation.
Let’s begin with a summary of the standard approach to evaluating journalistic argumentation. The analysis begins with a text, usually a newspaper editorial, opinion column or letter-to-the-editor. The text is put into the logical form of an argument – a string of premises attempt to support a conclusion(s). In many cases, a relatively clear logical form is extracted from an opaque, rambling text. The reader, or the student in a classroom, is asked to consider whether the argument is deductive or inductive, whether there are missing premises, whether the argument commits a fallacy and whether the piece contains ambiguous or loaded language. Elements that are not directly relevant to how the premises logically support the conclusion are irrelevant to the evaluation, and treated as distracting rhetoric.
This analysis can be found in journalism classrooms, media textbooks and manuals on how to detect biased media messages. The approach depends on assumptions about the form of communication being used and the most appropriate logical tools for this style of journalism.
The journalism to be analyzed is assumed to have the following features:
1. The argument uses only one medium, typically print, and the nature of the medium does not play a substantial role in the evaluation process.
2. The journalism article uses a “transmission” form of informing and arguing:
(i) The journalist (author) of the article transmits the results of his research and analysis
The journalist is active in researching, verifying and constructing an argument. The reader is a relatively passive recipient of the transmission.
(ii) The journalist writes from the article from a position of ‘authority’- a firm, settled view about a topic or issue.
2. The act of communication is from one person to many. It is not interactive.
3. The communication is a completed text, a finished product, not an on-going process or dialogue.
Given this form of journalistic argumentation, a traditional logical approach to the text is presumed to be adequate for evaluation. The main features of the analysis are:
1. The analysis is linear and propositional: The text is reconstructed into a set of propositions leading to conclusions.
2. The analysis pays limited attention to a range of contextual elements:
(i) The arguer
(ii) The background of the debate
(iii)The role of the intended audience
(iv)The forms of media used
Making these features explicit explains why a new model of evaluation is needed. The assumptions about journalism no longer apply to a good deal of the new journalism. Journalism on the Internet employs a different model of communication. Instead of using a transmission model, the journalist uses a dialogic model that stimulates interaction between journalist and audience. On-line argumentation uses multiple forms of media and layers of information. The reader is invited to actively explore various aspects of the issue that fan out in many directions. The result is a non-linear argument that encourages the reader to think ‘horizontally’ and in depth. The idea of the journalist as an authoritative voice transmitting an argument in one direction to an awaiting audience is replaced by the idea that the writer is one voice among many, the initiator of a discussion. New media journalism, therefore, needs a more flexible model of argumentation that captures and evaluates these new features.
We argue that the rhetorical theory of argumentation (Tindale, 1999) provides this model. To explain our approach, we will analyze a multi-media article from the Salon.com site in the United States, a popular source for on-line information and argument. The article attacks George Bush’s opposition to a Federal bill of rights for patients. It argues that a similar bill in Texas did not lead to dire consequences, as Bush had claimed. We turn now to relevant details of the model and our analysis of the example.
2. Central features of a rhetorical model
Given what has been said so far, it is probably not too difficult to see why the traditional notion of argument has been deemed insufficient to capture what is at the heart of the process of journalistic argumentation. That traditional notion has encouraged the idea that an argument consists only in a collection of propositions, torn from the flow of speech and thought and frozen on the page; and that its proper evaluation lies in understanding the relationship between those isolated statements (premises and conclusions). Ignored is any interest in the source of the argument (arguer) or its target (audience) and the relationship between them, nor in the medium employed for communication. So, if our aim is to study journalistic argumentation, we must do so with a more expanded sense of ‘argument’, one that is rooted in a rhetorical approach to argumentation.
There are several important features to this approach. Perhaps chief among these is the idea of an invitational rhetoric (Foss & Griffin, 1995). This perspective views persuasion not as an activity imposed by one individual (or group) on another (or others), but as a joint venture of equal partners whereby people are presented with an opportunity to persuade themselves. The audience shares in the argumentative activity. When it is the understanding of an issue that is the arguer’s goal, rather than changing or controlling others, then an invitation may be extended to the audience to enter the arguer’s world and see it as he or she does. The audience listens and then presents its own position. Thus the examination of an issue is a co-operative venture. There is something of the pragmatic dialogue at the heart of this proposal. Rather than treat the parties in the dialogue as protagonist and antagonist, it recognizes them as equals, mutually respectful of each other. Beyond this, the stress on invitation is reminiscent of a similar sensibility in Aristotle’s Rhetoric, where the audience is invited to complete the argumentation. Thus, the all-important element of audience self-determination pervades both the Aristotelian model of rhetoric and more recent proposals. The audience, when persuaded, is persuaded by its own deliberations, after reflection on reasoning that it has understood in its own terms and may even have had a hand in completing.
Treating argumentation as invitational or dialogical in this way involves adopting other audience- or context-related features, which might briefly be described here. To begin with, there is the idea of a cognitive environment (Sperber & Wilson, 1986). An interest in the beliefs and tolerances of audiences leads us to consider the environments in which audiences assess arguments and make their judgements. When we think about collectivities of people who may be the intended audiences for pieces of argumentation, it is natural to think about what these groups have in common, what kinds of knowledge inform their joint association? This idea is often captured by talking about the common knowledge that a group has. In practice, though, it is notoriously difficult to know what other people know, or what kinds of mutual knowledge exist between members of a group. The idea of a cognitive environment circumvents this problem by encouraging us to think not about what people actually know, but what they might be expected to know given the kinds of (cognitive) environments in which they move. According to Sperber and Wilson, a ‘cognitive environment’ is the set of all the facts manifest to each of us that we can perceive or infer. “A fact is manifest to an individual at a given time if and only if he is capable at that time of representing it mentally and accepting its representation as true or probably true” (1986, 39). A mutual cognitive environment is a cognitive environment in which it is manifest (clear to all) which people share it. Game theorists, for example, share a mutual cognitive environment. As individuals they are essentially different, but there is an overlap in that certain facts and assumptions, and the meanings of terms in a shared language, are manifest to them. They may not make the same assumptions, but it is possible for them to do so. Mutual manifestness is weak in the right sense, since a claim that an assumption is mutually manifest will not be a claim about actual states or processes but about cognitive environments; not about what people know or assume, but what they might be expected to know or assume.
We need, however, more than the idea of a cognitive environment to fully explore the context of an argument. The first substantial model of rhetorical argumentation, provided by Aristotle, understood context in terms of the relations between logos, ethos, and pathos. Since then, other features like the “message” and the “common code” have been added. One such idea is that of locality, the time and the place in which the argument is located. Depending on the issue, such considerations can vary from being peripheral to playing a central role in the recognition and assessment of arguments. Where and when people live affects the nature of their thinking and therefore their arguments.
A second consideration is background, by which we understand those events that are instrumental in our understanding of an argument: the occasion of the exchange or discourse; prior argumentation on the issue, including argumentation between the arguer and audience; current social and political events which give clarity, urgency or irony to the argumentation; the consequences for the participants of the outcome of the argumentation. A third, and central feature, is that of the arguer, the intelligent initiator of the meanings inherent in the argumentation (although those meanings will be modified once the audience becomes active). While locale and background contribute to the source, the arguer is its principal constituent. In the case of Internet journalism, as we will see, the ‘arguer’ can be complex, comprising the author of a text, owners of a site, and advertisers. Another important consideration involves modes of expression: the utterance involved and the force of its expression; what is said and what is left unsaid; the mannerisms of the arguer (when present); and the medium used to convey the argument, along with the conventions of that medium.
Finally, there is the audience itself. This is another complex idea, and a fluid one. Audiences change, even in the course of argumentation. As Perelman reminds us: “We must not forget that the audience, to the degree that speech is effective, changes with its unfolding development” (1982, 149). Or, as Neil Mercer puts it: “Context is created anew in every interaction between a speaker and listener or writer and reader” (2000, 21). The rhetorical audience is not a passive consumer of arguments; it plays an active role in the argumentation. The nature of the audience sets the terms of the premises, which are formulated in light of theses accepted by those to be addressed. The audience also contributes assumptions to the reasoning. And the audience can interact with the argumentation in the mind of the arguer or in dialogue with the arguer and become a co-arguer.
These aspects of audience indicate that a principal way in which arguments are judged successful and evaluated is not directly in terms of their internal logical support, but in terms of their impact on the audience. The aim of argumentation is the adherence of audiences to its theses, which requires the development of an underlying understanding. Argumentation is judged strong or weak to the degree that this aim is accomplished. But this leads us quickly to two stumbling blocks. Since success is viewed only in terms of the audience accepting a thesis, then it would appear that anything goes in persuading an audience; and, connected with this, that the model of argumentation involved is thoroughly relativistic. We address this concern through the distinction between the particular audience and the universal audience.
In most everyday contexts, argumentation is directed at some particular audience. But lying within that audience is a sense of universality connected with what is understood as reasonable. That is, a standard for evaluating argumentation that rules out prejudice and illegitimate bias can be active in any audience. The universal audience is not a model of ideal competence introduced to a situation to moderate or judge what is to be reasonable; it is developed out of the particular audience for a situation, thus, in a strict sense, describing what is reasonable. The universal audience, as a representation of reasonableness in a specific context, cannot value effectiveness over reasonableness. This would be self-contradictory. On this model, the particular audience is brought to agreement on its own terms; on terms that are internal to it; on terms that it recognizes and supports. Producing and evaluating argumentation involves learning about what is reasonable, rethinking it, adding to it, and taking from it. The source for this is the particular audience. There is no other empirical ground. There is nowhere else to look for our standard of reasonableness other than to the reasoners themselves as they self-consciously engage in this activity.
3. Analysis of the example
The example in question is an article by journalist Jake Tapper entitled, ‘The healthcare disaster that wasn’t’ (July 17, 2001). It shows parallels between the arguments Bush used in 1997 when opposing the passage of the Texas Patient Protection Act and those he was using in 2001 to oppose a patients’ bill of rights that was making its way toward debate in the House. Bush presented both scenarios as having potential to set lawyers against doctors, to the detriment of the latter. He had let the Texas legislation become law without his signature, but predicted it would drive up health costs. Tapper canvases various opinions in arguing that Bush’s predictions have not come to pass, and would not be expected to transpire on the federal level either. As an example of web journalism, the article contains various links to other articles (usually by Tapper) providing further information, and is accompanied by an image of a man in a dark suit (presumably a lawyer) wrestling to the ground a man in a smock and wearing a stethoscope (undoubtedly a doctor). Various additional links surround the front page of the article, taking people to other pages on the Salon site, offering pieces of advice, or simply selling products like vacations and Cuban cigars.
The first thing we should draw on in analysing this example is its mode of expression. The argument is not a linear product but more like a three-dimensional ‘thing’, pointing to many things around it. It has a ‘depth’ that is missing from the usual journalistic argumentative text, but which, interestingly, is more characteristic of the argumentative exchanges that go on between live arguers, where people introduce other points. In the Salon.com example, the links can support the claims being made by providing warrants, but they also place an onus on readers to go beyond the principal text to ‘verify things for themselves’. These depth-providing links lend an important new sense to the notion of ‘background’. One of the points we are concerned about under this heading is prior argumentation on the issue in question. This is very important to how an audience will interpret an argument and respond to it. Those of us who teach argumentation or journalism in the classroom often have to fill in the background to texts so that students will understand what is going on. Tapper’s ‘text’, though, brings with it its own background in the form of links to previous argumentation.
However, it is important for us to critically appraise this component. While the links are an important aid and addition to argumentative discourse on the web, they are also a feature to be assessed. In Tapper’s article, for example, one of the key reasons given for why people should not oppose a federal patients’ bill of rights is that the Texas bill (passed in spite of Bush’s opposition to it) has had “none of Bush’s apocalyptic predictions” come true. But what were these predictions? This is where we would expect a link, especially since the reference to these predictions as ‘apocalyptic’ colours the tone of the piece in a way that favours support for the federal bill. The most relevant link that preceded this part of the text was where the phrase “fought the law tooth and nail in Texas” was highlighted. But that link is to an August, 2000 web article ‘Patient Politics’ which provides little insight into the 1995-97 debate in Texas. All we are told by Tapper is that Bush said of the Texas law: “I am concerned that this legislation has the potential to drive up healthcare costs and increase the number of lawsuits against doctors and other healthcare providers.” Hardly an apocalyptic prediction, and a concern actually borne out by some of the information provided elsewhere in Tapper’s article.
So, we must be alert to the absence of links just as we would challenge the absence of claim-support in a ‘regular’ text. But the links where they are provided give us a far more dynamic sense of claim-support.
There are two other important observations to make about the mode of expression. The first is the inclusion of visual images to accompany the Tapper text. In this case we have a picture of a snarling man in a business suit throwing a medical practitioner to the ground. This gives some visual insight into the article, which on one level is all about a struggle between the legal and medical professions. It’s a very active image calculated to grab the browser’s attention and draw her into the text. Although in this case it isn’t actually a part of the argument, it has the potential to be part of the persuasive package. Of course, visual images can accompany the static texts of other media, so this isn’t such an innovative feature. But its potential exceeds this, because we can imagine (or expect) the future inclusion of animation or video, thus enriching the dynamic presentation of this medium.
The other feature to note is the matter of iteration or repeatability. All texts are repeatable in the sense that we can come back to them again and again, and they may not be saying the same thing to us insofar as we have changed and we bring different things to the text. But in the case of the medium in question, it brings different things to us each time we return to it. The iteration that we are analyzing is that on the pages of Salon.com on July 17, 2001. But each time we return to the ‘url’ for Tapper’s article, we find it embedded in a different set of external links (as opposed to the internal links that act as warrants). This means the article itself is juxtaposed against different links with respect to health care issues and issues of American politics, as well as advertizing. This last point leads us to think about arguers and aims. The principal arguer is Jake Tapper, author of the article ‘The Healthcare Disaster that Wasn’t’ at the centre of the site. His text is organized in support of the claims that the Texas legislation has not been a disaster and that, therefore, Bush should not oppose the Federal Patients’ Bill of Rights. His aim is to bring his audience to accept his position. A quite traditional aim; it is only the method that is different here. But even he is also doing other things. Many of the supporting internal links take us to other pieces he has written. And at the end of his article, there is a link to the Amazon.com page where Tapper’s book, Down and Dirty: The Plot to Steal the Presidency, can be immediately purchased. So, he is doing more than presenting a claim; he is presenting a whole perspective, a complete persona, of which the current argument is only a sampler (the reason for buying the whole, as it were). Likewise, the people behind Salon.com are selling their site as an informative, useful site that we will come back to again and again to receive information. This differs markedly from other media sources for information. Even though our favourite newspapers hope to make subscribers of us, they don’t assault us with such a plethora of images and invitations, they don’t offer us the immediacy and the interactive experience that comes with the web-based journalism. Finally, there are, of course, the advertisers offering their wares. And not just any wares, but wares selected for the types of audience that will likely be perusing Tapper’s piece. Again, it is the immediacy of the ‘sell’ and the interactivity of the invitations that separates these advertisers from those in the regular media.
The audience is a central consideration in all of this, and it is time to turn to that consideration. The site, the article and its links suggest a number of things about the intended audience, its makeup and cognitive environment. The audience is relatively affluent, as is suggested by the advertisements for Cuban cigars and vacations. It is also an audience concerned with healthcare issues, since this is an article dealing with healthcare, with links to topics like sunburn. The article assumes an audience that is politically savvy and, if not already critical, prepared to be critical of George Bush. But it also anticipates a reactive audience (rather than a passive one): an audience that will become engaged in the issue and article, will follow links, investigate, and enter the debate. This last feature is specifically encouraged by the addition of a link at the end to ‘Sound Off’ and an invitation to ‘Send us a Letter to the Editor’. These responses can in turn be followed and responded to. Hence, the article and its author expect a relatively dynamic audience, and the argument is structured accordingly: posing questions, offering depth-links into the issue, presenting positions and countering those positions.
Given all these features, it seems the argument is not structured to be a ‘stand alone’ final statement on the issue, but part of an ongoing dialogue. We can expect other pieces to follow, just as pieces preceded it (and those that follow will link here). So an aim of the argument is to gain the adherence of the audience in the sense of bringing them into the debate (and hence back to the site), rather than simply gaining the adherence of the audience for the claims made. This makes it difficult to measure success on a number of levels. Insofar as rhetorical argumentation looks to audience adherence as one measure of success, it is difficult for an evaluator to assess this. But we can assess how successfully the argument works as a dynamic exchange of ideas and invitation to become involved. Let’s consider this. Ostensibly, Tapper is offering three principal claims, within an informative ‘reporting’ of the facts. He is proposing that the Texas bill has not been a disaster (the lead paragraph clearly asserts this) and marshals support for this claim through testimony and links. He also suggests that there is a need for a Federal Bill of Patients’ Rights. And so, Bush should not oppose the federal bill (as he is currently doing). Important here is the support for the first claim, because establishing this would show both support for a federal bill and that Bush has been wrong on this issue in the past. We have already seen an absence of evidence in support of the claim that Bush projected apocalyptic consequences from the Texas bill. Instead, what we are told is that Bush was concerned that the legislation would drive up healthcare costs and increase lawsuits against healthcare providers. To show that the Texas bill has not been a disaster, then, Tapper needs to demonstrate that both these outcomes have not materialized. But the article is ultimately ambiguous on these points. Tapper writes: “Far from becoming a bonanza for avaricious trial lawyers, the right to sue an HMO or insurance company in Texas has been exercised just 17 times.” Admittedly, Tapper’s principal source, George Parker Young, is biased on the issue or is a “partisan on the topic” since as a lawyer he prosecutes such suits and stands to gain from them. But Tapper also cites the assistant director of the public information office as confirming that there have not been “a rash of out-of-control- suits.”
On the other hand (and there is an attempt at balance here), a critic of the legislation, Lara Keel, points out that it is really too soon to tell if the 1997 law will create a flood of suits, since none has yet made it to jury trial. That is, things move slowly in such matters, and four years is not sufficient time to judge; and people may well be waiting to see how juries deal with suits such that there could yet be a flood of suits (as Bush had feared). Keel also points out differences between the Texas legislation and the proposed Federal bill (such that the internal reasoning of Tapper’s argument may not necessarily follow). Keel further indicates an adverse effect of the Texas bill in the statistic that 18 per cent of small business in Texas has dropped their health insurance: “The premiums have risen enough to put the cost of health insurance out of reach” (which was another of Bush’s concerns).
This position is again countered by other statistics and other perspectives, arguing that even where HMO premiums have increased, the bill has not been a factor. There is some confusion around the statistics Tapper provides, because matching a decline in the percentage of employees of small businesses to a decline in the percentage of such employees with insurance is not significant if the second statistic measures only the current employees (that is, even if the number of employees has decreased, there has still been a decline of health insurance among those still employed.) Even the further statistic from the Texas Association of Health Plans that the number of Texans enrolled in plans has actually increased since 1997, need not be significant because this figure represents an increase of the total number of Texans covered and we need to know how many of these are employees of small businesses. The ambiguity of the statistics presented, then, does not sufficiently countered Keel’s claim about the harm done to small businesses.
Given what we know about the audience profile for this argumentation with its political savvy and openness to the issue, we would expect the universal audience (the principle of reasonableness) within that audience to pick up on these problems and recognize that, ultimately, Tapper does not provide sufficient support for the claims (explicit or implicit) that he makes in his article. In particular, it has not been shown that the law has been a success in Texas and a reasonable audience, even one predisposed to supporting such a law, should set aside their interests in the issue enough to see this. Moreover, the argumentation itself does not warrant the force of the lead to the article (‘The healthcare disaster that wasn’t’), where Bush’s warning of problems is asserted to have been wrong. We say the force of this lead because much of the article shows that the full extent of predicted problems hasn’t materialized yet. But some, it seems, have, and more may yet emerge.
The invitational nature of this kind of text, assuming and encouraging a responsive audience through things like the links provided, allows for this disagreement with the claims. It allows the audience to make up its own mind, even though the text is slanted towards a certain reading of the information. And this, we might suggest, is an important feature of journalistic argumentation and one that is particularly supported by the type of medium employed here.
4. Conclusion: Importance of a rhetorical model
This example shows how the new journalism applies its technology to persuasion on-line, and how a rhetorical model can explore this form of communication. The strength of the model is that it is not restricted to statements in the text. It can look at the presentation as a whole, and the interplay between arguer and audience. It can take into account the layers of information contained in hyperlinks, and it is sensitive to the media in which the argument is constructed.
In conclusion, we would like to explain the importance of developing a new model of media criticism. One reason has already been noted: Multi-media journalism is ubiquitous, with the spread of the Internet. Even traditional media, such as newspapers and public broadcasters, have interactive web sites.
Another reason is that a rhetorical model is better equipped to understand the rapid changes taking place in journalism today. Multi-media journalism is redefining journalism, the forms of public deliberation and the relationship between the public and their media. In newsrooms, journalism is being transformed. Journalists must learn to report for a variety of media – print, TV, and the Web. Powerful new technology is making available new and global sources of information, and increasing the speed of newsgathering. The online journalist is more the initiator of a conversation than a transmitter of pre-digested facts or settled opinion. Journalistic arguments are only one part of web sites that seek to attract audiences by being interactive. Their audiences expect to be given the opportunity to engage in debate, to pursue their own research on the Net, and respond immediately by e-mail to what the journalist writes.
This new environment changes how journalists see themselves and how they regard public deliberation. Instead of claiming to write the last word, online journalists explicitly acknowledge that their articles are ‘spins’ on complex topics. A commentary is a fallible stab at a defensible point of view. Readers are expected to make up their own minds. They must triangulate to the most plausible position by using a plurality of perspectives. For example, Jon Katz, the American columnist, has described how writing commentary online contrasts with writing editorials for a newspaper. Within minutes of posting his articles on the Internet, a host of experts from around the world use e-mail to provide Katz with counter-evidence, counter-argument, original documents, graphics and new studies. Katz often finds himself writing not one article on a topic but a series of articles with updated facts and alternate perspectives. Katz says interactivity can be a humbling experience: “The only thing I can compare it to is being tied to the back of a car and dragged through the street” (Shapiro, 1999, 39).
In addition, a rhetorical model helps us assess whether journalistic arguments promote or inhibit rational deliberation in the public sphere. Does online argumentation, with its new approaches, really advance journalism’s democratic function of creating an informed citizenry? John Pavlik (2001), a new media expert at Columbia University, thinks that the Internet’s ability to provide layers of information should result in a more “contextualized journalism” for a pluralistic public sphere. Yet we know that Internet journalism also disseminates unreliable rumour, manipulated images, and uninformed ranting. The edgy “in-your-face” style of writing that is popular on the Internet can devalue logic and reasonable deliberation.
A rhetorical model can’t by itself answer these large questions about journalism’s impact on the public sphere. But it can help us to address these questions by re-shaping how we think about argumentation in journalism today. In an era of global media, we need sophisticated tools to critique how the media influences us and our fellow citizens. A rhetorical model is one of those tools.
Aristotle (1991). On rhetoric: A theory of civil discourse. George A. Kennedy (Trans.). Oxford: Oxford University Press.
Foss, S.K. & Griffin, C.L. (1995). Beyond persuasion: A proposal for an invitational rhetoric. Communication Monographs, 62, 2-18.
Mercer, N. (2000). Words and minds: How we use language to think together. London: Routledge.
Pavlik, J.V. (2001). Journalism and the new media. New York: Columbia University Press.
Perelman, Ch. (1982). The realm of rhetoric. William Kluback (Trans.). Notre Dame: University of Notre Dame Press.
Shapiro, A.L. (1999). The control revolution. New York: Century Foundation.
Sperber, D. & Wilson, D. (1986). Relevance: Communication and cognition. Cambridge: Harvard University Press.
Tindale, C.W. (1999). Acts of arguing: A rhetorical model of argument. New York: State University of New York Press.
Tapper, J. (2001). The healthcare disaster that wasn’t. http://www.salon.com/politics/feature/2001/07/17/patient/print.html