WHAT MAKES AN ARGUMENT A GOOD ARGUMENT?
Dr. Jon Bruschke
Department of Speech Communication
California State University, Fullerton
PO Box 6868
Fullerton, CA 92832-6868
Paper presented to a "top three" panel at the 1998 National Communication Association Convention, New York, NY.
WHAT MAKES AN ARGUMENT A GOOD ARGUMENT?
Perhaps because of its unique subject content, the field of communication is awfully self-reflective (e.g., Powers, 1995; Wartella, 1994; Zarefsky, 1993). And while communication scholars of all stripes struggle to find their place and purpose in the academic mosaic, argument scholars face questions no less profound than those Zarefsky asked of the broader discipline: What holds the field together? What are its core concepts? Imposing questions indeed. This paper will tackle those questions in a way, it is hoped, that will not simply locate argument as a field in its proper place on the intellectual organizational chart. Instead, serious consideration of those questions ought to yield insightful ways of thinking about how arguments are studied, defined, and evaluated. Essentially, this paper will contend that current approaches to the study of argument tend, in general, to focus attention away from the logical components of argument. It is simply a question of emphasis, because there is nothing wrong with the other sorts of judgements that argument scholars make, but a profound one since assessments of what constitutes a "correct" or "rational" argument remain one of the most powerful tools of rhetors. What should be the vital core of the field is instead a rarity. The central issues, therefore, are why a focus on the rational elements of arguments is useful, why it is possible, and how it can be done. These issues will be fleshed out by concentrating on three central questions: (1) Do current conceptions of argument include rationalist approaches? (2) What theoretical justification can be made for a rationalist approach? (3) How can scholars make rationalist evaluations of arguments?
Do Current Conceptions of Argument Include Rationalist Approaches?
Defining an academic field can be accomplished either by looking at the scholarly efforts to define it or by looking at the current work underway in our journals. The former approach examines theoretical attempts to define what the field should be while the latter examines what scholars are actually doing. An interesting exercise is to compare the theoretical definitions with the actual work and look for overlap. Taking the theoretical view first, how has argument been defined in academic works? The answer, of course, is that it has been studied in a variety of ways. Argument has been subscripted (argument1, argument2, and even possibly argument0), it has been alliterated (process, product, and procedure, in Wenzels  terms), its verb has been altered (having rather than making an argument), it has been converted to acronyms (CRC approaches versus the MDLs in TOA [Willard, 1989]), it has been converted to its grammatical uses (argument as noun versus argument as a verb [Bruschke, 1995]), or its features have simply been listed (Brockreide, 1990). Without belaboring the intricacies of all of these definitions, it is safe to say that they range from the broad to the narrow. Those taking a more post-modern de-differentiating attitude can easily dismiss strict boundaries with ready examples that defy the borders and theoretical reasons that the perimeters cant exist; these approaches produce very broad definitions. Those taking a more narrow view, conversely, have a damnable time maintaining clear distinctions for their categories in an era where critical theories are increasingly rendering distinctions of any variety untenable (even simple things like "yellow" or "ball" are categories that are hard to defend, as Rouse  has explained Hesses examples). The most narrow views consider only reasoned suasory discourse to be "argument" (in its various forms referred to as making an argument, argument as a noun, argument as product) while the most broad definitions include virtually any human action that can have some symbolic value assigned to it (called argument as process, argument as verb, argument as "disagreement relevant expansion" [Willard, 1989]).
Another striking element of these current definitional efforts is that, with rare exception, few scholars find it necessary to choose between the broad and narrow interpretations. Thus, while a certain definition may distinguish between, say, the making of rather than the having of an argument, there is not necessarily a need to choose between the two. Either is acceptable, either leads to insights, and the question is one of preference rather than correctness. Even Charles Willard has protested that his broad theories arent meant to preclude the serious study of issues that fall under the more narrow definitions of argument. Even the narrow definitions seem to be aimed at defining a core of study rather than denying that there is insight to be gained at the periphery. Thus, there seems to be ample room within the current scholarly definition of argument to allow for the study of the rational elements of argument. Given the definitional allowances for the study of argument logic, it might seem reasonable to expect that the core of current argument research centers around logic-centered assessments or at least that a decent chunk of what the field does includes such analyses. As it turns out, this is not the case.
Put directly, argument theorists have flatly backed away from evaluations of arguments as being either correct or incorrect. The problem is not so much that we are unable to locate Platonic truths that we should be able to; it is that we have discovered that Platonic truth does not exist and have abandoned the search for truth entirely. The baby is gone with the bath water. As Young, Launer, and Austin (1990) comment: "Many contemporary theorists would have the critic abandon traditional forms of argument analysis and evaluation for an audience-centered perspective" (p. 89). A quick review of the articles in the past seven years' issues of Argument and Advocacy confirms this assessment. The discipline commonly engages in rhetorical criticism, conducts interpersonal studies of disagreement behavior, issues cross-cultural and cross-disciplinary studies of debates, offers persuasion and quasi-persuasion studies, and serves as a sounding board for intercollegiate debate professionals. Only two articles (Walton, 1991; Young, Launer, & Austin, 1990) have concerned themselves with the "correctness" of an argument in the traditional sense. Even then, Young and Launer opt for an approach that blends audience-centered and formal approaches. Introductory texts (e.g., Rieke & Sillars, 1993) generally have more discussion about the connection between argument and logic than contemporary argument writings.
Ed Schiappas collection of papers under the title Warranting Assent provides another poignant case in point. Schiappa (1995) is vociferous and eloquent in making his case for argument evaluation: "Argument evaluation has many uses, including improving argumentative competence, teaching critical thinking, and improving decision making. The benefits of argument evaluation are coterminous with the benefits of the study and practice of argumentation itself, for one cannot be successful at either without the ability to recognize the difference between good and bad arguments" (p. ix). Given the editors desire to separate good and bad arguments one might expect several papers devoted to analyzing the logical strength of various arguments. As it turns out, the book is divided into four sections, three of which do not concern themselves with logic at all. Two sections focus on ideological and axiological evaluations, which are expressly the sort of approaches that have been contrasted with logical assessments. The fourth division is called "epistemological" approaches and most nearly mirrors the logical perspective. Of the three essays contained in that section, one is a re-print of the Young and Launer article, one is a narrative critique (narrative, by the way, is the approach Young and Launer explicitly identified as the most obvious attempt to move away from logical evaluations), and the third is an essay by Dennis Gouran discussing logical failures in the Challenger accident. Even Gourans essay isnt so much dedicated to evaluating arguments as valid or not, but instead takes for granted that huge logical errors were made and focuses on proving that these errors were argumentative in nature and not procedural. Thus, even in a volume dedicated to argument evaluation with an editor concerned about critical thinking and sorting arguments out between the valid and faulty, less than one twelfth of the pages are dedicated to any concern with logic at all and even those do not constitute paradigm cases of logical criticism.
After a roughly two thousand year tradition of trying to discover a system that can speak to the validity of a claim, with the hope that systematic thinking can lead to a way to compare incompatible claims at a logical level, we have in large measure given up the fight. For all the critical worrying about the dominance of logical positivism, one is hard pressed to find any evaluation of logical correctness issued by an argument scholar this decade. To re-emphasize, the problem is one of focus. We have not failed in our inability to uncover unalterable truths (they probably dont exist), but we have failed to fully explore what sort of truth we may find instead.
Why are contemporary argument scholars so reluctant to engage the question of whether or not an argument is correct? Certainly one reason could be that present attempts at identifying when an argument is a good argument have not been satisfying. In general, there have been two strands of scholarship that attempt to define when an argument is correct. First, the work of Stephen Toulmin has sought a field-based definition of argumentative correctness. The notion of a field, however, has been roundly debated to the point that any definition of a field immediately sparks a debate and any discussion of when an argument advanced within a field is a good argument is immediately pre-empted by the inability to conclusively define what is meant by the term "field." Indeed, given the expansive definitions of fields that have proliferated of late it is difficult to imagine any situation where two communicators were not arguing from at least different interpersonal fields. Additionally, the work of Toulmin seems to have given way to new theories; Toulmin seems to have become more of an obligatory footnote or a text chapter rather than a driving force behind serious scholarship. Of course, the point here is not to thoroughly review in fine detail the many excellent points made about argument fields, it is simply to take a step back and ask what contribution the field debate has made. One would be hard pressed at this point to say that there is even common agreement on what criteria define a field, much less common understanding of what the extant fields are. (For a more extensive critique of fields, see Bruschke .) It is also fair to say that the field concept has not spawned a large body of insightful research. As Yogi Berra might have said of fields, everyone is talking about them but nobody is doing anything with them.
Second, some scholars have moved to an "informal logic" system (e.g., Walton, 1987; 1991). This model has been largely ignored, perhaps because of its inability to apply beyond a limited number of arguments. Most arguments that we seem to encounter in our lives seem infinitely more complex than a series of claims that can be tested with a list of fallacies; the fallacies can take on the flavor of "nit-picking errors of form" (Rieke & Sillars, 1993) rather than honest efforts to take the arguments seriously. There can be little doubt that, whatever other reasons scholars have for shying away from declarations about when an argument is a good argument, the failure of current systems to provide useful and satisfying conclusions is certainly a contributing factor.
In addition to these pragmatic limits against scholarly analyses based on logic there are theoretical precepts that may have discouraged scholars from issuing the more traditional analyses. The theoretical drive behind this relatively recent shift in emphasis is quite probably the question of indeterminacy. Indeterminacy is a broad and powerful concept that has been applied to many areas of inquiry; uniting all its applications is a belief in the inability to come to fixed and certain conclusions. Using indeterminacy, for example, epistemologists have concluded knowledge cant be fixed, legal scholars have concluded laws cant have fixed meanings or applications, and communication scholars have concluded that meanings cant be certain. In fact, the power of indeterminacy may also have served as the primary rationale behind the broad definitions of argument and the ease with which they destroy easy distinctions.
Given the starting point that neither meaning nor knowledge can be fixed, it is unsurprising that argument scholars have had difficulty pursuing questions of truth. The problem is that on the one hand, as Fritch and Leeper (1993) describe, argument scholars are almost universally interested in "discover[ing] truth as a product of the valid form" (p. 186), while on the other hand scholars in general have had great difficulty generating an epistemology identifying truth in any form (Roth, 1987). A recent twist is that any claim to truth may inadvertently serve dominant interests. Cox (1993) summarizes the dilemma well: "(1) we may fail to mediate difference or achieve a shared rationale for action; or (2) we may in fact succeed but at the cost of being complicitous with the dominant reason or common sense of a culture" (p. 11). Unable to discover an absolute truth and not wanting to take on all the critical claims at once, many scholars have sought a middle ground by seeking methods of evaluation that fuse the logical components of arguments with their ethical, emotional, and poetic features (see e.g., Bruschke, 1995; Fritch & Leeper, 1993; Klumpp, 1990, 1993; McKerrow, 1993; Nelson, 1991; Young & Launer, 1995). The general hope is that arguments, evaluated in this new light, will be more accurately evaluated and, more grandly, that the arguments most worthy of adherence will enjoy a logical and poetic advantage. Lost in the shuffle is a lucid definition of how an evaluation of logical correctness is even possible. Willard (1990) speaks for the majority: "Thought and emotion are inseparable Rationality comes easily in armchairs. But it may not mean much to say that one should be free of prejudice, bias, power, and politics. One never is" (pp. 224-225). The result has been that more energy has been directed towards analyses of pathos and mythos than logos. The driving theoretical force behind this distrust of rationality may bear the mark of indeterminacy.
It is not necessary, however, to slight the critical approaches to assert that the field of argument should have something to say about which arguments are correct. Even Willard (1983; 1989), who takes as his starting point Gadamer's indeterminacy claims and argues that criticism should be evaluation free (1983, p.245), understands that even if we cannot know the truth the rigors of life demand that we act as if we do (1983, p. 235) and act as lay "scientists" trying to make sense of the world. In the end, Willard concedes that there are situations where the use of argument standards may be "plausibly defended" (1989, p. 318) and believes that we should help our readers develop "good judgment." The difficulty is that the concept of "good judgment" is poorly defined (Bruschke, 1992) and Willard (and recent scholarship in general) has precious little to say about when argument standards may be "plausibly defended." The point is that even scholars who adopt the most indeterminate theories must concede that humans make real choices with material outcomes and argument scholars do well when they speak to how those choices can be made in a more informed fashion. Put simply, while it is true that from one philosophical perspective there is no necessary connection between any symbol and a referent, practical experience confirms that a surprising amount of the time people do seem to share meaning and live their lives in that space. Informing the choices that humans make is surely a worthwhile endeavor. Of course, it is one thing to identify how wonderful it would be if we could grapple with indeterminacy and it is quite another to provide an alternative theoretical basis for a more satisfying system. That will be the work of the next section.
What Theoretical Justification can be made for a Rationalist Approach?
If we are to take the more narrow definition of argument seriously we need a decent definition of it, a theoretical justification for it, and a working system to represent it. Given the proliferation of definitions the terminological questions will be left open for now; logic is defined here solely for the purposes of providing a general understanding of what is being discussed. Logic is defined as the rational, reasoned, and sound; the more logical an argument is the more likely it is to be correct. A more careful definition may ultimately be needed, but between here and there much work remains to justify why we should even be thinking about rationality, much less how we ought to precisely define it. It is my belief that there is enough shared understanding of what is meant by "rationality" to enable a useful discussion without becoming bogged down in definitional questions.
On, then, to the task of defending what value logic might have. It is my contention that the CRC model claims connected to reasons serves as a general model for all human thought (Billig, 1987). There are infinite manifestations of how claims are connected to reasons, what count as good reasons, and especially what the true claims are, but all reasoning begins by taking something that what we think that we know the relatively indisputable facts -- and trying to use them to prove less indisputable claims (Brockreide  has made this point most clearly and is clearly influenced by Toulmin in his exposition). Once again, Willard seems to accept this by adopting the human-as-scientist metaphor. I take the central question to be whether or not this process has value.
A defense of the process can be found in the pragmatic school which, oddly enough, arose from a group of critics of science. No article-length paper could possibly cite all the work encompassed in a thorough discussion of epistemology and scientific criticism; I will rely heavily on Rouses (1987) excellent summary of the various authors and positions for the sake of compactness. My point here is not to make a definitive epistemological statement, only to identify that there does exist a theoretical basis for accepting the CRC view of argument as valid. Whether that viewpoint is ultimately correct is the subject of another debate, but much as Willard has adopted but not proven his psychological perspective as the basis for his forays into argument, I simply identify that the pragmatist perspective is most persuasive to me and serves as the assumptive backdrop to that which follows. As Willard (1989) cleverly put it: "The purpose isnt to defend a theory from scratch but to consider its implications for a view of argument" (p. 318). Pragmatism is my starting point; if readers dont like it, they wont like the rest of this article.
Rouses account of pragmatism begins by contrasting "realists" who believe that science can accurately represent things with its theoretical terms with "non-realists" who assert that it cant. Obviously, indeterminacy is the weapon of choice for the non-realists. In Rouses view, the non-realist account simply fails to account for technological progress: "non-realist accounts of science (such as the post-empiricist model described above) seem at first glance to make the technical success of science a miracle" (p. 6). If no scientific theory ever has a referent it is awfully difficult to explain the development and existence of, say, airplanes. Partial representations can be accurate and useful if not Platonically true and the means by which these partial truths are revealed is dialogue. Drawing heavily on Rorty, the pragmatists have faith in dialogue and contend that "truth is what would emerge as the result of unconstrained inquiry pursued indefinitely" (p. 7). Although any claim, method, or standard of evaluation might ultimately be disproved, concerns about Sisyphian failure "can be dismissed as incoherent or vacuous unless it can be connected to the specific inadequacies of specific theories or disputes. But such specific criticisms bring us back into the conversation, require us to take specific stands, and thereby remove the general worry with which we began" (pp. 7-8).
A group of scholars who take the argument one step further is a camp Rouse identifies as the "new empiricists" (including Hesse, Hacking, Cartwright, and Laudan), a group Rouse sees as reacting against pragmatism but whose work could instead be viewed as an extension of it. The scholars begin by acknowledging that knowledge is historically situated and that observation is colored by pre-existing theoretical conceptions: "These new empiricists do not attempt to resurrect the distinction between theoretical and observational statements or languages, nor do they generally defend a naive realism about observation. The world as we encounter it is already interpreted, and theories play an important role in our received interpretations" (p. 9). Nonetheless, empirical phenomena exist and some theories allow for the successful manipulation of it; the only difficulty arises if we doggedly insist on equating our empirical prowess with truth: "The idea is that we can agree in identifying an empirical phenomenon despite even radical differences in our theoretical interpretations of it. We can also determine whether a theory accounts for such an identified phenomenon in the theorys own terms without necessarily accepting the theory of even acknowledging the legitimacy of its success" (p. 10). In the end, theories are evaluated in the very pragmatic terms of how well they solve the problems they attempt to solve. Rouse restates the point most succinctly by stating that "all things being equal, what makes for a better scientific theory is empirical growth (or expansion of problem-solving ability)" (p. 10).
Three things are striking about pragmatism as articulated by the new empiricists. First, the new empiricism can account for technological progress in a way that the more extremely subjective theories can not. Scientific progress ceases to be miraculous. Second, the new empiricism recognizes and accounts for the subjectivity that may shape observation, an issue that has concerned Willard and more recently (and more generally) Berger (1994). The new empiricism cannot be dismissed with simple appeals to the innate subjectivity of all things or the concern that all observations, when peered at deeply enough, embody some sort of theoretical suppositions. Of course they do, but some appear to solve problems better than others. Knowing which appear to work better seems like extraordinarily valuable knowledge for humans who act as scientists and must make difficult choices with profound material and emotional consequences. Third, this is all accomplished with the premise that agreement is possible on empirical facts if not the ultimate meaning for them. Theories may be indeterminate, but that doesnt mean that facts are.
This final point deserves more expansion because of its profound power. The implications of the point are reserved for the next section, but for now the point is important enough to dwell upon. Facts are knowable, and the case for it abounds. It has already been said that all human reasoning starts with that which is knowable and proceeds to that which is not; Toulmin has taught us that much (and Billig has elaborated on it). Even Willard, who starts with indeterminacy, makes no bones appealing to the "facts of communication." Even Foucault, no cheerleader for the cause of reason and the patron saint of deconstructionism, has been found to use a historical method and can be found reasoning from what he takes to be the facts of penal systems. Even Roth, a modern expositor of indeterminacy, concludes: "The parallel I am urging suggests that the process of checking the empirical adequacy of translations follows basically the same pattern found when examining just how well a scientific model is capable of providing solutions to the questions it generates" (p. 240). Even Carlson (1994), who argues well for the expansion of what rhetorical scholars should use as evidence, does so by appealing to what she takes to be the facts about "new rhetorics" that are contemporary in our culture. Our journals overflow with reports of data from social scientists, with the belief that the data can speak to a theory. Twice the Western Journal of Communication has found that broad methodological issues that separate scholars are reducible to questions of evidence. What counts as a valid fact in any particular case can be debatable, of course, and such concerns become part of Rortys conversation. Although articulated in obtuse ways and in a variety of formats, and often voiced implicitly rather than expressly, the premise that evidence, facts, or "data" can serve as the common ground from which arguments can proceed seems to be at least implicit with the writings of virtually everyone. It is the engine that powers the new empiricism and the force that animates and makes possible all scholarly discussion.
In sum, while indeterminacy provides a strong case for non-rationalist evaluations of argument, pragmatism and the new empiricism provide an effective theoretical base for engaging in rationalist evaluations of argument. The final question is what such an evaluation would entail.
How Can Scholars Make Rationalist Evaluations of Arguments?
Platonic truth is dead. It may not even have survived Platos student Aristotle. If it does exist it is in exceedingly rare circumstances. What is offered below is not a conclusive system that will uncover the ultimate truth of any given issue. What I hope it will provide is a scheme that makes it easier for scholars to lay out an argument and a vocabulary for assessing when a point is well proven and when it is less well proven. This does not represent revolutionary thinking so much as a refinement of the tools of analysis that are presently available. Argument scholars who forage into the analysis of disputes that necessitate the resolution of factual claims may not be able to come to decisive conclusions about the absolute veracity of the claims they study, but they should at least be better equipped to make systematic judgements.
Arguments may be evaluated on a number of different criteria. As noted above, scholars may issue ethical, moral, aesthetic, and rhetorical critiques of arguments. What is at present lacking is a solid way to evaluate the logical correctness of an argument. Figuring out how to represent an argument's logical components is the first necessary building block to issuing those evaluations.
At present there are two dominant ways that the logical components of arguments are represented. The first way is to identify the type of claim made by an argument. Arguments are thought of as being cause-effect, slippery slope, question-begging, or sign-based arguments; they are usually attached to some sort of fallacy or potential fallacy (e.g., Hamblin, 1970; Kahane, 1984; Walton, 1987; 1991). This first method of representation will not be considered here solely because it is not comprehensive. While such representations can be effective for evaluating certain types of arguments they are not general in nature. The second method of representing an argument's logical structure is the Toulmin model (Toulmin, 1958). While other scholars have worked on ways to represent arguments (e.g., van Emeren & Grootendorst, 1990) they tend to end up looking like Toulmin diagrams or perhaps are less detailed. The Toulmin model (as slightly modified by Rieke & Sillars ) serves as the starting point here because it is the best worked-out and most widely recognized representation of an argument's logical structure. In general, while the Toulmin model is a useful advance in contemporary thinking and unquestionably applies to areas not captured by the inflexible syllogism, it attempts to identify the logical value of individual claims and does not represent well arguments as they so often actually happen, with interlaced and overlapping dimensions. Several of the modifications offered below were originally offered by Manicas (1966) who advocated rejection of the model; I restate and add to them here to show how a modified Toulmin model can serve useful analytical functions.
Speaking broadly, Toulmin believed an argument may have as many as six parts (see Figure 1). Included are a claim, grounds for the claim, a warrant that connects the claim to the grounds, backing, rebuttal, and qualifiers. (For writing ease, I will use the terms "grounds," "data," and "evidence" interchangeably.) Readers unfamiliar with the model are recommended to Toulmin's original work or Rieke and Sillars' slight modification. Three particular points, however, are of present interest. First, the qualifier is only a particular type of claim. For example, if I were to assert "Bill Clinton will lose in the 1996 election" I have made one claim and if I were to say "Bill Clinton will almost certainly lose in the 1996 election" I have merely made a different type of claim. The phrase "almost certainly" does not function as a discrete part of the argument. Because the purpose here is to arrive at the probability that any claim is accurate (or how likely it is to be accurate), the qualifier is merely a move in the direction of reducing the probability that an argument is true or identifying its boundary conditions. As such it will not be considered separately from the claim. Rebuttals serve the same function and will be similarly eliminated.
Secondly, there is an undeniable similarity between a claim and the grounds for the claim (Manicas, 1966). In fact, in many actual arguments the grounds also serve as a claim of its own that may be refuted. For example, Anita Hill may claim that Clarence Thomas would make a bad Supreme Court Justice and use as her grounds the fact that he sexually harassed her. The grounds, however, may be called into question and thus the statement "Clarence Thomas sexually harassed Anita Hill" serves both as grounds for the broader claim and as a claim in its own right. Therefore, instead of statically defining a given statement as being either claim or grounds assertions may be thought of as grounds/claims (or "data/claims") that may be refutable and may be interconnected. They are tied together so that an argument does not consist of several discrete claim-warrant-data "arguments" but several connected data/claims. Figure 2 is the refined representation of the argument: "Clarence Thomas is unfit for the Supreme Court. He sexually harassed me and other women as well. I have the phone records to prove he called me at home and I am a credible witness."
Notice that the argument does not have to end there; arguers can (and did) call into question Ms. Hill's credibility and the phone records she offered. Notice also that the middle data/claim is supported by not one but two pieces of information; this is a situation that occurs frequently but is not contemplated in Toulmin's model. Finally, notice that the chaining together of multiple claims allows the representation of complex arguments. Toulmin's work was truly groundbreaking but his examples tend to be limited to simple statements with single sources of proof. Perhaps one reason that Toulmin has fallen out of contemporary favor has been the difficulty in applying his model to actual arguments. If so, the present modification may correct that shortcoming.
Third, there is an uncanny similarity between Toulmin's grounds and backing and his grounds and claim (see Manicas, 1966). If, as Toulmin recognizes, it is necessary to have a warrant (the "inferential leap") connect the grounds to the claim, what is it that connects the backing to the claim? Can the connection be made without an inferential leap? Should the leap not be subject to the same tests that other warrants are? If we can accept that grounds function more as data/claims and are usefully represented in serial connection to one another as they are in Figure 2, it is relatively unproblematic to conclude that backing fails as a useful concept. What Toulmin thought of as backing is represented here as its own data/claim with its own warrant connecting it to the higher-ordered data/claim.
These three changes, taken together, have simplified the number of components in Toulmin's model and allowed them to be represented in serial connection with one another. The serial connection has the advantage of representing arguments as they actually occur and recognizing more readily the interplay between the arguments.
In addition to these structural modifications there are some other points that bear emphasis that are not necessarily inconsistent with Toulmin's thinking but have not been emphasized. First, there may be a number of warrants that connect grounds to a claim and not just one. As Brockreide and Ehninger (1960) describe warrants, they are "that part of an argument which authorized the mental 'leap' involved in advancing from grounds to claim... Its function is to carry the accepted grounds to the ... proposition" (p. 45). There is no reason, however, why only a single statement authorized the connection between the grounds and a claim. Consider the example Brockreide and Ehninger use: "Russia would violate the proposed ban on nuclear weapons testing (claim); Russia has violated 50 of 52 international agreements (grounds); Past violations are symptomatic of probable future violations (warrant)" (p. 45). While the warrant is certainly accurate, there are other possible warrants as well. For example, "The CTBT is similar to past treaties," "There have been no changes in Russian leadership since the 50 violations," and "The US did not violate those agreements first" are all similar statements that facilitate the connection between the grounds and the claim. In short, a warrant, in addition to connecting the grounds to the claim, may be thought of as any assumption that must be true for the grounds to be relevant to the claim. There need not be only one such assumption.
Secondly, warrants themselves have factual components and thus also take on the character of data/claims. It is useful to retain them as a distinct component because they have a distinct function: A warrant is a data/claim that connects two other data/claims. However, it is often necessary that the warrant itself requires backing. This is a point made by Toulmin when he discusses "established" warrants (p. 175-177) and in the many discussions of Toulmin that have followed (e.g., Brockreide & Ehninger, 1960; Golden, Berquist, & Coleman, 1976, chpt 18; Manicas, 1966).
Finally, there is no reason that only a single piece of data need support a higher-ordered data/claim. Although this is also consistent with Toulmin and his explicators it is less readily apparent from their writing. But surely a claim like "We should abolish Affirmative Action" has more than a single bit of data to support (or refute) it. The more highly ordered the claim the more evidence there is that is likely to be advanced in support of it. To sustain a claim advocating the abolition of Affirmative Action will require, in all likelihood, many forms of support. On the other hand, the claim "the University of Utah's basketball team lost in the second round of the 1995 NCAA tournament" requires only the final score as its grounds.
In sum, the current proposal eliminates backing, qualifiers, and rebuttals from the representation of arguments and orders all the data/claims in an argument serially. It recognizes that there can be a number of warrants that connect any two data/claims and that any data/claim can be supported by more than one other such element. Finally, the current model recognizes warrants as a special class of claim that serves a discrete function but are limited by the same requirements as other claims, that is, they require well supported data of their own that must be connected by valid warrants. One final emphasis is necessary, and it is that the components of the model are functionally defined. Nothing makes a statement intrinsically the data, warrant, or claim. The claim of one argument may be the warrant of another and so forth.
If this model is an improvement over current representations it should be able to adequately represent an actual argument. An editorial offered by the Dallas Morning News, contained in the appendix, is diagrammed in Figure 3, and it is left for the readers to evaluate whether the representational structure represents an improvement over fallacy-like configurations or more traditional Toulmin conceptions. It should be noted, however, that even an argument this brief would be difficult to represent with a single Toulmin model; a series of such models could be constructed, but that representation would lose the interconnectedness and interplay between the various parts of the argument as it is advanced.
Now that a system for representing arguments is in place a much more daunting task lies ahead: How can we tell when an argument is a good argument? This paper will advance the following proposition to define a good argument: An argument is a good argument to the extent that the claim and only the claim is supported by the grounds and to the extent that all relevant grounds have been presented. One can test whether the grounds support the claim and only the claim by offering alternative explanations for the grounds. The remainder of this section will attempt to unpack these concepts.
The first requirement of a good argument is that the claim be supported by the grounds. Imagine a trial where a popular football star is accused of murdering his ex-wife. The prosecution advances as evidence the fact that blood found at the murder scene matches blood on the defendant's car. In this case the grounds, the blood on the car, is consistent with the claim advanced by the prosecution (that the defendant is guilty.) As a second example, consider the claim: "It must be winter; my cat is sleeping on the heater." The data is that the cat is sleeping on the heater; it is certainly consistent with the claim that it must be winter since the cat would ostensibly be seeking heat and winter is ostensibly cold. Notice that in neither case do the grounds necessitate the claim, but each element of evidence might make the claim more likely. It is, of course, possible to add more data in support of the claims. The prosecutor may show that the defendant's whereabouts during the time of the murder are unaccounted for, that the defendant hated his ex-wife, that genetic material found on the victim's fingernails matches the defendant's DNA, and so forth. In the alternate example, to further prove the claim that it must be winter one could point to the temperature or simply the calendar. Notice that the effect of the multiple strands of data may be cumulative (though not necessarily in a linear fashion); the more evidence that is offered that is consistent with the claim the more likely it appears that the claim is correct. This is because as more grounds accumulate it is more difficult to imagine alternative explanations for the grounds at hand, a point that will be discussed further below.
The second requirement is that the claim, and the claim alone, is supported by the data. This does not require, as was mandated by the traditional syllogism, that only claims that are supported by the evidence and only the evidence are worthwhile. It does suggest that the most sound argument is one where the claim and only the claim is supported by the grounds. To rephrase, a claim may become more sound the more that it appears that it is the only claim supported by the grounds. The means for testing whether the claim alone explains the grounds is by suggesting alternative claims that can explain the grounds. The more claims that can explain all of the grounds, the less sound the claim is believed to be. Consider the trial example and imagine that the prosecutor has only presented the blood as evidence. A defense attorney might suggest an alternative claim that explains the grounds: Perhaps the police planted the evidence in the defendant's Bronco. This alternative claim is equally consistent with the evidence and serves to undermine the original claim because of it. In this case the alternative claim did sufficiently "loosen" the connection between one piece of data and the claim. If the blood evidence were the only evidence rational arguers would not accept the claim of guilt because the grounds would be equally well supported by an alternative claim (assuming that some additional grounds supported the "evidence planting" claim). Of course, additional evidence to support the "police plant" claim (such as blood splatter patterns) aids the alternative explanation. Similarly, if one has claimed that it must be winter because a cat is sitting on the heater, a counter-claim that supports the grounds equally well might simply be that the cat is tired. A sleepy cat must lie somewhere, after all.
For an alternative claim to adequately defeat the initial claim it must meet all the requirements of the initial claim: (a) It must explain all the grounds offered in support of the claim, (b) it must explain the grounds equally well, and (c) it must be consistent will all other relevant grounds that come to light. As an illustration of requirement (a) consider that it is possible to accept that the police planted the blood in the Bronco but still accept the claim that the defendant is guilty if other grounds, such as the DNA tests, come to light. The (b) and (c) requirements recognize that as soon as the alternative claims are offered they must also be supported. Obviously, the more grounds that can be offered to show that the police may have planted the blood the more likely that claim is to be true. If one claims that the cat is simply tired the question of whether or not the cat is actually sleeping or warming herself becomes pertinent. In short, a counter-claim may equally well support the grounds in question but, if it lacks support of its own, may be implausible. I could claim, for example, that when Bill Buckner muffed the ball hit to him in the sixth game of the 1986 World Series he was distracted by a giant, invisible slug that oozed up his back and broke his concentration. This is an alternative claim that equally well explains the grounds (or in this case, the grounder) and counters the claim "Bill Buckner choked." However, unless I could offer additional evidence to support my claim there is little reason to accept it, especially if there is other evidence to support the "choke" claim.
There will be instances where some claims survive all attempts to offer alternative hypotheses. Consider the claim "It must be winter" that is supported by three bits of data, including the cat sleeping on the heater, the temperature, and the date. It would be relatively easy to offer alternative explanations for the grounds that the cat is sitting on the heater. It might even be possible to offer alternative explanations for the temperature; it might, for example, be an extremely cold (but not unprecedentedly cold) day in July. It would be nearly impossible, however, to offer an alternative explanation for the calendar date. Not all (or even many) disputes will be this easy to resolve and most will likely involve the suspension of judgment and greater gaps in knowledge. Nor will all disputes necessarily entail a claim that is clearly more correct than another. However, there will be cases where one claim emerges as clearly more correct and in all cases the method of reaching a decision will be more sound if the process suggested here is adhered to. If we must act with incomplete evidence, as we so often do, it makes sense to at least make the best guesses that we can.
The only additional requirement is that the claim account for all the relevant grounds. Imagine a situation where the estranged ex-wife of a physicist is killed with a letter bomb. While the bomb was not produced by the physicist, a check from the physicist to the bomber shows up in the bomber's bank account. At this point, two pieces of evidence, the fact of the estrangement and the check, point to the guilt of the physicist; clearly, he wished to kill his ex-wife whom he disliked and paid the bomber to do it. The claim that the physicist paid to have his wife killed is perfectly consistent with the grounds. However, there may be additional evidence that casts doubt on the claim. If it is discovered, for example, that the bomber is also a physicist, that the bomber had one of his ideas stolen by the physicist and had his life subsequently ruined because of it, and that the bomber was unaware of the divorce and believed that the physicist still lived at his ex-wife's residence, the claim is now much less strong. The bomber may have tried to kill the physicist but sent the bomb to the estranged wife by accident. The check could be explained as a payoff rather than blood money. New grounds have been introduced that the original claim cannot account for.
Arguments need not be thought of as true or untrue, but could be evaluated in a number of ways, including "sound," "compelling but not conclusive," "plausible but indecisive," or "unacceptable." In a given dispute, neither side may enjoy much of a logical advantage. A critic might conclude that one side has made a more compelling case, that neither side has, or that the evidence is not strong enough to point to a winner at the particular time. To re-state, there is no need for an argument to attain the level of Platonic truth before it is given serious consideration by argument scholars.
Of course, this scheme does not necessarily apply to all claims. It is very likely that several sorts of claims do not lend themselves to rational resolution. Ethical, aesthetic, and value claims come quickly to mind. This is not troubling. Not all arguments turn on rational points, and many decisions may incorporate value and rational elements. In the abortion debate, for example, there are moral as well as factual medical points. The point is not that the logical elements should always carry the day (as critics of science fear they will always do), but it is still the case that if the decision involves factual and moral elements it would be a better decision if the factual elements were reasoned soundly. In other words, while most contemporary theorizing recognizes a fusion of logical and non-logical elements, there is very little guidance about how to reach a conclusion about the logical elements before they enter into the broader debate. And there are an enormous number of vitally important arguments that doturn on logical issues; it is not wise to look past those incidents. If unimportant or uninteresting to argument scholars they are of vital necessity to the body politic.
The Importance of Data
A careful reader will notice at this point that little has been said about the content
of warrants and what makes them good or bad. This is unusual against the profound impact
that the concept of warrant has played in modern argument theory. A. Cheree Carlson
(1994), making a comment rather incidental to her overall point, wrote profoundly:
"When discussing criticism, the focus has been less on data than on warrant. The
evidence is almost a given. What matters her is how we justify making the leap from there
to our conclusion" (p. 21). Why the lack of concern about the warrant here?
The answer is that, in practical experience, most arguers seem to have a strong and largely correct intuitive sense about what makes grounds relevant to a claim. The instances where truly irrelevant grounds are offered get quickly dismissed, and what seems to count most is that good information (grounds, data, evidence) is available. For example, on the recent partial-birth abortion debate, one factual element that was relevant to the discussion was the number of such abortions actually being performed. Ron Fitzsimmons, executive director of the National Coalition of Abortion Providers, initially claimed that the procedure was fairly rare and was used only when there was some danger to the mother or the when fetus was extremely damaged. Closer examinations of the medical records indicated that these claims were essentially incorrect and Fitzsimmons later publicly declared that he had lied (Reed, 1997). The important point is that the warrant was never a question in the debate; both sides immediately understood the importance of the frequency of the abortions. If partial birth abortion is a rare procedure that is invoked mostly when there is a threat to the womans life then it is probably an issue best left unregulated. If it is a common procedure conducted mostly to terminate an otherwise normal pregnancy then the issue seems like simply another manifestation of the abortion debate and one for which the normal pro-life claims apply as well as they ever do, or perhaps more so since the fetus is so close to birth. It is interesting to note that some pro-choice advocates (e.g., Feldt, 1997), perhaps understanding the significance of the point, decided to hold to the claim that partial-birth abortions are rare and offered other sources (the Alan Guttmacher Institute) to confirm the point. At any rate, we can notice that (a) factual data is relevant to an issue that is frequently cast as a moral one, (b) the warrant is never in question, only the accuracy of the data, and (c) the point seems to turn appropriately on which source is providing a more accurate count of the number of partial birth abortions.
This is not to claim that the factual points in the debate ought to carry the day on the abortion debate necessarily. It may be that one side can demonstrate a moral imperative for protection of the fetus so compelling that it demands that abortions are never performed, even in cases of rape, incest, or health of the mother (as a surprising number of politicians in Texas advocate). Alternatively, it may turn out that the pro-choice side can sustain the moral claim that a womans right to control her body is so compelling that the state is never justified in attempting to regulate abortion choices, even for late-term abortions conducted solely for birth control reasons. If, however, neither side ultimately wins an absolute moral claim, which quite frankly appears more likely to be the case, there are going to be a number of difficult issues that arise where moral and factual claims intertwine. In those instances factual claims will be relevant, the quality of the data will be in question, and argument scholars ought to play a valuable role in assessing the quality of those arguments.
A final point about data is that, while it appears to be more powerful than concerns about warrants, there may be a number of cases where the available data is not sufficient to demonstrate a point conclusively. For instance, we could ask whether Lee Harvey Oswald acted alone when he shot JFK. A careful critic assessing the evidence might reasonably conclude two things. First, there is sufficient evidence the implausible path of the "magic bullet," the quick and mysterious death of Oswald and later Oswalds assassin, the amazing mortality rate of potential witnesses that could counter the Warren Commission conclusions to conclude that the grounds are not consistent with the single gunman theory. Second, one could conclude that there is not sufficient evidence to demonstrate who was acting with Oswald. Simply put, we dont have enough data to know the answer.
For the present theory it is enough to note that the rational conclusion is to suspend judgment, that the crucial part in resolving the question would be the uncovering of sufficient data to rule out possible explanations, and that no amount of correctness in the warrants will help us decide between competing conspiracy theories if the data is inconclusive. If we dont know, we dont know. There does seem to be a temptation, however, to stare at a important question that the data does not answer and take a number of dangerous routes. Some might move to moral or other non-rational judgements, since at least that discussion can (seemingly) proceed in the absence of data. Some might retreat into an ideological stance and conclude that, although there is not evidence in this case to prove the point either way, the assassination must be a government plot (as the government is an essentially corrupt institution that lies and plays under-handed tricks all the time) or perhaps that the conspiracy theorists are crazy (as there are always lunatics trying to undermine legitimate state authority). Some might start examining the warrants that extend from the inadequate data and see if any are faulty leaving some conclusions more plausible. Some might dismiss the entire rational project as unable to ever lead to a correct conclusion. Perhaps a better course than any of these is to bracket all these issues and get on with the task of finding the data that will answer the question.
The warrant is an indispensable theoretical instrument that explains why human reasoning proceeds as it does, allows us to contemplate questions for which there are not certain answers and consider content in addition to form, and introduces indeterminacy into our thinking. It is data, however, that provides humans-as-scientists with the answers that are necessary to navigate their way through the praxis of daily existence.
Limitations and Possible Objections
As with any perspective, the ideas offered here will contrast with other perspectives and no doubt objections will arise. This section will explore four possible objections to the proposed theory: (a) contextual factors doom any evaluations of correctness and incorrectness, (b) paradigm filters will limit the value of any evaluation, making the format proposed here futile, (c) the proposal eliminates the community as the basis of the warrant, and (d) the proposal eliminates power from the central consideration of argument.
The first objection would go something like this: There is no absolute or Platonic truth, merely local truths, and the attempt to define a system that identifies Platonic truth is thus doomed to failure. My answer to this objection has been alluded to during the course of the paper but it bears re-stating with more emphasis here. Arguments are good arguments when all the relevant grounds have been considered and fit with the claim. There are two conditions when the argument judged to be the better argument cannot have any claim to absolute truth; either the arguers don't have all the relevant grounds at their disposal, or the grounds themselves change and thus cease to be "relevant." As an example of the former situation, two arguers may be disputing whether the world is flat but lack any observation other than that which their eyes can see. Without pictures from a high vantage point or advanced mapping technologies the arguers would lack relevant data and could come to an erroneous conclusion, even if one person argued the case better. As an example of the second condition, consider a case where two marketers are trying to decide which color to make a new model car. One marketer might suggest powder blue and cite a marketing study from 1954 as evidence that the public is likely to respond favorably to the color; the other might suggest simple red but be able to cite no study. At first blush the powder blue advocate may seem to have the better argument; the marketer has grounds for the claim that the other marketer lacks. However, if the public's reaction to powder blue has changed since 1954 (imagine, in the name of good taste, that it has) the grounds are no longer relevant, and thus the claim is not well supported.
This system cannot overcome these limits. No intellectual system devised yet can. But the point behind the new empiricism isnt that we have arrived at a Platonic truth that will answer all questions, it is that as we act with incomplete knowledge we can at least select coherently among competing explanations for the empirical facts we think we do know. Knowledge may not be final, but it can at least advance. All of this discussion circles back to the central thesis of this article: Arguers must act as if they have complete knowledge. When they lack all the necessary information, when the information at their disposal is incomplete or not relevant, it is often the case that arguers must still act. The marketers must still decide what color to paint their cars. If the system proposed here makes arguers better able to act on the information they have, and if argument scholars can identify and critique public figures who advocate solutions to public problems based on irrelevant information or misreadings of information, or at the extreme cases where public figures advocate positions inconsistent with relevant information, argument will serve the body politic well.
A second possible objection is that the search for the correctness of arguments excludes critique; history and Thomas Kuhn (1970) have shown us repeatedly that paradigm systems lock us into certain ways of thinking and thus paradigms limit the questions that can be asked at any given time. If the system proposed here were adopted and left out cultural critiques humans could never find the relevant grounds or imagine alternative hypotheses to explain the relevant grounds. This criticism might be valid if the proposed system were intended to supplant the critiques of thought systems; it is intended to celebrate them. Paradigm critiques may provide the most profound ways of identifying new explanations for relevant grounds. In fact, one reading of Kuhn is that paradigms shift when grounds cease to be relevant or when new grounds come to light that are not explained by prior knowledge claims (viewed synonymously with "arguments" here). If a critique can show how grounds are not relevant, how relevant grounds have been excluded, so much the better. In fact, whether one believes that paradigms change via revolution (as Kuhn does) or evolution (as Toulmin does), even Kuhn thinks that the changes occur because of unexplained empirical observations, a thought that rest very comfortably with the new empiricists.
A third possible objection is that the force of the warrant has been moved out of the community. This is an objection of some concern as argument scholars have become intensely interested in the idea of community of late (e.g., Klumpp, 1990). There are two answers to this. The first is that, based on the objections noted earlier in this paper, the notion of a "field" or community has become so loose that it has failed almost entirely to provide relevant insight into the conduct of any particular argument. Stated differently, while theorists have had a "field" day demonstrating that any argument is field dependent, there has been precious little work that has been able to demonstrate how in any particular argument arguers can (and do) choose different warrants or how in any particular argument scholars can usefully analyze the dispute. Put in the most banal terms, scholarship of late has been content with the answer: "Yep, those two people disagree because they are from different communities" and leaving the analysis at that. While entirely true such conclusions are hardly illuminating. We are left to wonder how the arguers will resolve their dispute, if there are better ways to resolve it than others, and whether one fields warrant might have more to it than anothers. The second answer is that, when warrants are conceived of as claims of their own, it is entirely possible to recognize that their truth value will always be localized but still offer arguers and analysts some means of identifying which field has elaborated more sound support for their warrants.
A slightly different explanation might make both answers to the objection more clear. At present, in any disagreement where warrants differ argument analysts would try to construct the different communities with the varying positions. There would be inevitable impossibilities in defining the boundaries of the fields, disagreement within the fields about what the rules were that created the warrants, and at any rate the study of the fields would make argument scholars into second-rate sociologists. The present system would simply lay out the different warrants and identify the grounds each side offered to support them. Such an approach would build a third community -- those interested in the dispute without partisan interests -- and would at least lay out a grammar to be the first step in resolving the inconsistencies between the two communities. In short, the method of analysis enables discussion of the differences and builds a new community where so many present approaches simply identify differences.
The fourth objection might arise from post-modern circles, most prominently the deconstructionists. Inspired by Foucault and Derrida, one recent position in argument is that, because language is indeterminate, arguments never prove or disprove anything and that in the end what wins and loses arguments is power. (Although there are important differences between Foucault and Derrida, and probably any two such prominent critics for that matter, I believe that they are united at least in terms of their concerns about power and would object to the project presented here for essentially similar reasons.) Those in power set the rules for argument and usually in a way that serves their own interests and pre-determines the outcomes of many arguments. Time and space preclude a detailed examination and critique of these positions, but the position has become prominent enough to deserve at least a quick consideration (see Willard  for a brief explanation). And certainly, if the premises of the deconstructionists are accurate, they indeed doom the position taken here. Initially, however, I believe that the deconstructionists over-estimate the indeterminacy of language. Much as with the fields debate, while it is possible to prove theoretically that no shared meaning is possible it is also true that a surprising amount of the time people do seem to understand what language is intended to mean. Screaming "Kiss off!" at a fellow motorist who cut you off could have many possible meanings, none of necessity better than the other, but perceptions about the meaning of that message would likely be shared with amazing accuracy. Any theory that does not recognize and give equal weight to the meanings that are shared is incomplete.
Secondly, power may not be as sealed and unchangeable as some imagine it to be; as Willard (1990) notes, there are optimistic and pessimistic versions of power. The fact that oppressed groups can turn to rationality to empower their critiques is a profound historical phenomena. For example, The Bell Curve is an argument made by empowered people, playing by their rules, written in the language of social science (which has been roundly critiqued as serving dominant power interests.) However, and by utilizing the same system of rules, Gould (1981) and others have been able to cast serious doubt on the central truth of the claim being made. Offering new grounds that disprove dominant claims and demonstrating that dominant claims do not comport with accepted grounds is a crucial step in empowerment. Other historical examples abound, from Jewish intellectuals utilizing rationality to elude and critique cultural oppression to Martin Luthers 95 Theses nailed to the Wittenberg Cathedral. It simply does not seem to be the case that the only way to deal with rationalitys encroachment beyond its legitimate boundaries is to discard it; often, offering better reasons is sufficient.
In summary, let me state what I hope this paper has accomplished and what remains unfinished. It has provided a way to represent arguments as they occur; present models are of little utility in breaking down the logical structure of actual arguments. Both fallacies and the Toulmin model are only capable of representing fragments of whole arguments. However, this paper has suggested only in sweeping terms how the statements should be ordered; in that regard, the work of van Emeren and Grootendorst may greatly inform the method. As an additional contribution, this paper has offered a lexicon for evaluating what is already understood, implicitly, as rationality. As Michael Billig (1987) has contended, all thinking is arguing, and by extension what is called argument is the process of making public those internal rationalities. It is hoped that this paper has provided a vocabulary to better understand and then evaluate how those rationalities operate in practice. Perhaps with a better system to represent arguments more scholarship will evaluate the rationality of arguments.
Most importantly, this paper will be successful if it reinvigorates the discussion of logos, the embattled canon lost in contemporary thought. It is rather unlikely that the system offered here will prove itself to be the final word on what makes arguments sound. It may, however, spark an interest in exploring the question further, and at least begin work in that direction. It is ironic that so many contemporary scholars have attacked and discarded rationality by using rational argument; that fact alone suggests that it is a concept deserving further inquiry.
Figure 1. Traditional Toulmin Model (from Rieke and Sillars).
Clarence Thomas is Thomas sexually Phone records
unfit to serve on the harassed me and prove he called
Supreme Court others (grounds)
verifies my claims
Sexual harassers should
not be on the Supreme Court I am a credible
Credible witnesses do
Figure 2. Revised Toulmin diagram of a hypothetical Anita Hill argument.
Leads to a 2nd Dallas rape crisis center and
victimization CSAC confirm this
[Authorities in the area
[Must prevent prove the point]
Failure to cry out She was in a state of shock
Blaming passive doesn't equal (grounds)
victims is consent
She feared retribution
[If the victim (grounds)
she shouldn't be
blamed] [Other explanations disprove the consent notion]
Should include legal Must respond in the most Sue James and the RCC
arguments before the sensitive way say so (grounds)
incident, not later (grounds/claim)
(grounds/claim) [Authorities prove the point]
[Blaming is insensitive] (warrant) (warrant)
[Legal arguments Can't shirk responsibility 1 assault every few
are no excuse for blaming (grounds/claim) months (grounds)
(warrant) [Blaming shirks responsibility] [The problem is large]
Figure 3. Revised Toulmin diagram of the Dallas Morning News article (implied statements are in brackets)
Berger, C. R. (1994). Evidence? For what? Western Journal of Communication, 58, 11-19.
Billig, M. (1987). Arguing and thinking. Cambridge: Cambridge University Press.
Brockreide, W. (1990). Where is argument? In R. Trapp & J. Schuetz (Eds.), Perspectives on argumentation: Essays in honor of Wayne Brockriede (pp. 4-8).
Brockreide, W., & Ehninger, D. (1960). Toulmin on argument: An examination and application. Quarterly Journal of Speech, 46, 44-53.
Bruschke, J. C. (1992). Willard and the postmodern debate: What ground can he stand on? Paper presented at the 1992 Speech Communication Association Convention, Chicago, Illinois.
Bruschke, J. (1995). Deconstructive arguments in the legal sphere: An analysis of the Fischl/Massey debate about critical legal studies. Argumentation and Advocacy, 32, 16-29.
Carlson, A. C. (1994). How one uses evidence determines its value. Western Journal of Communication, 58, 20-24.
Cox, J. R. (1993). Postmodernity, crypotonomativism, and the Rhetorical: A defense of argument studies. In R. E. McKerrow et al. (Eds.) Argument and the postmodern challenge: Proceedings of the eighth SCA/AFA conference on argumentation (pp. 8-12). Annandale, VA: SCA.
Feldt, G. (1997, July). Partial-birth abortions should not be banned. Cosmopolitan, pp. 155, 224.
Fritch, J. E., & Leeper, K. K. (1993). Poetic logic: The metaphoric form as a foundation for a theory of tropological argument. Argumentation and Advocacy: The Journal of the American Forensics Association, 29, 186-194.
Golden, J., Berquist, G., & Coleman, W. (1976). The rhetoric of western thought (2nd ed.). Dubuque, IA: Kendall/Hunt.
Gould, S. J. (1981). The mismeasure of man. New York, NY: W.W. Norton and Company.
Hamblin, C. L. (1970). Fallacies. London: Methuen.
Kahane, H. (1984). Logic and contemporary rhetoric: The use of reason in everyday life (4th ed.). Belmont, CA: Wadsworth.
Klumpp, J. F. (1990). Taking social argument seriously. In R. Trapp & J. Schuetz (Eds.), Perspectives on argumentation: Essays in honor of Wayne Brockriede (pp. 110-120).
Kuhn, T. S. (1970). The structure of scientific revolutions. Chicago: University of Chicago Press.
Manicas, P. T. (1966). On Toulmin's contribution to logic and argumentation. Journal of the American Forensics Association, 3, 83-94.
McKerrow, R. E. (1993). Overcoming fatalism: Rhetoric/Argument in postmodernity. In R. E. McKerrow et al. (Eds.) Argument and the postmodern challenge: Proceedings of the eighth SCA/AFA conference on argumentation (pp. 119-121). Annandale, VA: SCA.
Nelson, J. (1991). Tropes of political argument in America. In D. Parson (Ed.), Argument in controversy: Proceedings of the Seventh SCA/AFA Conference on Argumentation (pp. 73-79).
Powers, J. H. (1995). On the intellectual structure of the human communication discipline. Communication Education, 44, 191-222.
Reed, R. (1997, July). Partial-birth abortions should be banned. Cosmopolitan, pp. 155, 224.
Rieke, R. D., & Sillars, M. O., (1993). Argumentation and critical decision making (3rd ed.). New York, NY: Harper-Collins College Publishers.
Roth, P. A. (1987). Meaning and method in the social sciences: A case for methodological pluarlism. Ithaca, NY: Cornell University Press.
Rouse, J. (1987). Knowledge and power: Toward a political philosophy of science. Ithica, NY: Cornell University Press.
Schiappa, E. (1995). Introduction. In E. Schiappa (Ed.), Warranting assent: Case studies in argument evaluation (pp. ix-xxviii). Albany, NY: State University of New York Press, Albany.
Toulmin, S. E. (1958). The uses of argument. Cambridge: Cambridge University Press.
vanEmeren, F. H., & Grootendorst, R. (1990). Analyzing argumentative discourse. In R. Trapp & J. Schuetz (Eds.), Perspectives on argumentation: Essays in honor of Wayne Brockriede (pp. 86-106).
Walton, D. N. (1987). Informal fallacies: Towards a theory of argument criticism. Amersterdam: John Benjamins.
Walton, D. N. (1991). Bias, critical doubt, and fallacies. Argumentation and Advocacy: The Journal of the American Forensic Association, 28, 1-22.
Wartella, E. (1994). Challenge to the profession. Communication Education, 43, 54-62.
Wenzel, J. W. (1990). Three perspectives on argument: Rhetoric, Dialectic, Logic. In R. Trapp & J. Schuetz (Eds.), Perspectives on argumentation: Essays in honor of Wayne Brockriede (pp. 9-26).
Willard, C. A. (1993). Argumentation and the social grounds of knowledge. University, AL: University of Alabama Pres..
Willard, C. A. (1989). A theory of argumentation. Tuscaloosa, AL: University of Alabama Press.
Willard, C. A. (1990). Argumentation and postmodern critique. In R. Trapp & J. Schuetz (Eds.), Perspectives on argumentation: Essays in honor of Wayne Brockriede (pp. 221-231).
Young, M. J., & Launer, M. K. (1995). Evaluative criteria for conspiracy arguments: The case of KAL 007. In E. Schiappa (Ed.), Warranting assent: Case studies in argument evaluation (pp. 3-32). Albany, NY: State University of New York Press, Albany.
Young, M. J., Launer, M. K, & Austin, C. C. (1990). The need for evaluative criteria: Conspiracy argument revisited. Argumentation and Advocacy: The Journal of the American Forensics Association, 26, 89-107.
Zarefsky, D. (1993). Does intellectual diversity always serve us well? Spectra, 29 (4), 2-3.
Copy of the Dallas Morning News editorial "School Assaults:
Don't blame the victim"
A 9-year-old girl is sexually assaulted in a Chicago school. A 12-year-old boy admits to the crime. Outraged, the parents sue the school district. Enter the school district's lawyers, who proceed to argue that it was her fault because she did not cry out.
Chilling? Revolting? You bet. By itself, the idea that a child can be sexually assaulted in a classroom without school authorities being able to prevent it strains credulity. Beyond that, however, the argument made by lawyers for the Chicago public schools last week that, in effect, "passive" victims should be blamed is outrageous.
According to the Dallas Rape Crisis and Child Sexual Abuse Center, old-fashioned notions about rape can and often do lead to a second victimization. Some victims of sexual assault may remain quiet during an attack because they are literally in a state of shock. Then again, others may fear life-threatening physical retribution, either during the assault or afterward. In any even, a failure to "cry out" should hardly be viewed as conclusive proof of consensuality.
Whatever happens in Chicago, public schools around the country must not be permitted to shirk their responsibility. Here in Dallas, police credit the Dallas Independent School District with making extraordinary efforts to monitor student activity in the schools. Unfortunately, bad things can and do happen nonetheless. At least one case of sexual assault comes to the attention of police liaisons in the Dallas schools every couple of months.
One such case is obviously one too many. But the point is, even though concerns about
legal liability and financial judgments are legitimate, such factors are best considered
before a sexual assault occurs, not after. As Sue James of the Rape Crisis Center rightly
notes, when sexual assault do happen, "schools need to respond to the problem in the
most effective and sensitive way."
This can and should include counseling and a review of security measures. But to blame the victim is neither effective nor sensitive; it is merely the cruel and gratuitous echo of the initial crime.