Awakening the Topoi: sources of invention in the New Rhetoric's argument model.

• Author: Crosswhite, James
• Position: Report

Models are simplified representations of complex phenomena that--we hope--will allow us to understand and interact with those phenomena better than we ordinarily do. "Better" in this case means "in such a way that we can more consistently predict and control the outcomes of the processes in which we are interested." In the very event of modeling, then, we can already see the hermeneutical erosion of the reality of argumentation. Naturally, we want to control the outcomes of argumentative processes, and we want a model that will help us to do this effectively. However, one of the realities of argumentation is that we ourselves, our desires and goals, are implicated and at risk in the process. Argumentation presents the opening to follow an argument wherever it leads, to change one's mind about what outcomes are desirable. So just following the reductionist road of constructing a model already pares off some very interesting features of the phenomenon of argumentation.

So, more specifically, what is it that we want from such a model? It could be any of many things. Models are pursued mostly because they promise to aid us in our efforts to predict and control the outcomes of argumentative processes, but also because they aid us in our attempts simply to understand the causes and effects in and of argumentation processes, to discover the measurable empirical features of these processes, to be able to build machines (which are themselves also models) that can produce these processes, and for other reasons as well--many of which are probably linked to Aristotle's still pertinent observation that we imitate simply because it delights us to do so and because it is the primary way we learn (Poetics, 1448b). However, we should note--sharply--at the outset that it is one thing to imitate models of what other people are doing and another thing to imitate those people and those actions themselves. This natural distancing of invention from the primary phenomenon is a recurring feature of attempts to use theories and models.

One of the primary results we want from a model of argumentation is a reliable method of inventing and teaching the invention of arguments. Of all the parts of rhetoric, invention is the most valuable, the one that historically dignifies rhetoric and lifts it from the occasional declines to which it has been subject. However, invention has also proven itself resistant to modeling. Take the case of the most widely used argument model of our time, the Toulmin model (Toulmin, 1958). Its features are well known. Essentially, there is a claim, there are data supporting that claim, and there is a warrant that justifies moving from the data to the claim. There is backing for the warrant, ultimately in the form of a rational enterprise that helps to constitute a field or domain of argument. Arguments also have qualifiers and conditions of rebuttal. All of this can be graphically exhibited in boxes and lines that almost resemble a flow chart. It appears to be a model of how to generate arguments and set them into motion so that one can move reasonably from data to a conclusion.

However, it is almost anything but. Its creator, Stephen Toulmin, never intended to write a book on rhetoric or invention (Olson, 1993, p. 284). However, this fact is also evident on simple examination. There are no actions specified by the model. First of all, there is no context or motivation for argument in the model itself. To be fair, Toulmin explores argumentative contexts in detail in his writings (1972, 1979), but this is not evident in the model, in which the context begins slowly to emerge only when one questions the backing for warrants. There are no exigencies, no rhetorical situation. Further, though, there are no procedures for generating arguments. Just how one gets a claim and data to show up in the boxes is a kind of mystery. There is a mostly suppressed dialogue hovering in a ghostly way around the model. Consider the model's first appearance in The Uses of Argument. Someone asks whether Harry is a citizen. We do not know why, or whether the question is appropriate, but we do know that, for some reason, someone (else?) answers "Yes, presumably . . ." But we are not at all sure how this happens, or who knows where Harry was born or who knows about the citizenship laws and offers this knowledge to justify the answer. The truth is that it is only when we already have an argument, or something like an argument, that we can use the model to represent it, and then analyze it--and, some say, better evaluate it. But where do arguments come from? The Toulmin model may be a reasonably good way of diagramming an isolated argument for purposes of analysis and evaluation, but it does not provide an answer to that question.

The New Rhetoric describes argumentation in a way that provides more satisfactory answers, but it too leaves us with some problems when it comes to invention. The New Rhetoric does not intend to offer a model, and the complexity of what it does offer raises doubts about whether it can even be referred to as a model. The aim of the treatise is: "The justification of the possibility of a human community in the sphere of action when this justification cannot be based on a reality or objective truth" (Perelman & Olbrechts-Tyteca, 1969, p. 514). This is a philosophical aim. To determine the conditions for the possibility of something sounds like a Kantian project. And there is also a Kantian aura here around establishing what we may hope--whether we may hope for reasoned accord in the sphere of action. The project also involves developing an identity between freedom and reason, another Kantian theme. There is in fact a consistent philosophical project going on throughout The New Rhetoric and Perelman's other writings. However, it is not a project of transcendental reflection or transcendental philosophy. Within the philosophical clearing the treatise creates, the main argument is an argument by example, showing that that there are forms of reasoning in conditions of uncertainty that have been successfully used across a broad range of cultures and times and occasions. Yet these forms of reasoning have been neglected in modernity, a period during which certainty and self-evidence came to be taken as essential marks of the rational, and logic became essentially connected with necessity. In the course of this philosophical effort, the new rhetoric project had to say a great deal about what argument was and how it worked. Accidentally, but inevitably, the project expressed something like a model of argument.

However, it is nowhere near as simple as the Toulmin model, with its six elements and its resemblance to a flow chart in which reasonable warrantability runs from data through warrant and backing and back down, with qualifications, to claim, though a claim still subject to rebuttal. Toulmin's model has no agents and no social relations, aside from the ghosts who prompt with questions and activate the model. It has no account of what arguers must know or what deep skills or virtues they must have. There is no sense of a historical or ethical framework. Its marketability as a model is due to its simplicity. The new rhetoric project has never been as marketable.

If there is something like The New Rhetorids model of argumentation, it is set forth in the division of the treatise into its parts and perhaps some of their sub-subsections. The first part of such a model would delineate the framework of argumentation, les cadres de l'argumentation (pp. 11-62), which could also be translated, "the bounds of argumentation." Identifying the bounds helps to distinguish between what is argumentation and what is not, first and essentially by distinguishing between demonstration and argumentation (pp. 13-14), but then by setting forth the necessary conditions for argumentation to occur. The first condition is that there must be a meeting of minds (contact des esprits, pp.14-17), and in fact all of argumentation is a continuous and specific kind of contact des esprits. This meeting of minds is made possible and sustained by having a common language (hardly a simple matter); a reason to argue, and so a goal that has a plausible chance of being achieved by argument; a situation or conflict about which the parties are willing to change their minds--that is, conditions in which people are receptive to arguments; rules that govern the beginning, the conduct, and the ending of arguments, including rules for turn-taking, the length of the arguments allowed, and so on; interlocutors who are willing and able to argue with one another, who respect each other enough to change their minds because of what the other says; interlocutors with knowledge of the other party sufficient and accurate enough to permit appeals to what is held in common and the use of appropriate argumentative forms. Further, there must be no violence or bribery or any other form of coercion (pp. 54-59). The reasons for being convinced must be discursive.

There must also be the equivalent of a speaker and an audience, the esprits who will modify their ordinary non-argumentative ways, and both restrain themselves and also assert themselves in the ways that will allow them to fulfill their argumentative roles in conformity with the bounds of argumentation (pp. 17-23). Ultimately, it will be the role of the audience to judge the arguments that are offered. In this model, all evaluations of arguments are ultimately supported by an audience.

If one were beginning a diagram of this, one might imagine moving from a box marked "preargumentative conditions" along an arrow to another box marked "argumentative conditions." However, the border between these two sets of conditions is anything but simple. The accomplishment of any of the many conditions I have listed is a matter of degree, and the judges of whether the conditions actually achieved are good enough are, first...