JAC Volume 2, Issue 1/2

The study of composition processes describes what writers do. The study of the art of composition describes methods for giving writers better control over what they do. This essay makes a contribution to both research concerns. It contributes to the study of composition processes by describing what ironists do when they refute an opponent. It contributes to the study of the art of composition by offering methods for giving writers better control over the adaptive strategies they use when attempting refutations.

We need both sorts of research in composition, but do we need to study the composition of irony? This essay contends that the study of irony solves a common problem in the teaching of adaptive writing. Although rhetoric texts invariably mention that writers of persuasive discourse need to adapt to opposing viewpoints, they fail to explain how to do this. Granted, the texts encourage students to keep in mind the beliefs and goals of the opposing readership; yet this hardly insures that the writer will know how to make good use of this knowledge.

To solve this problem, we need to know more than what expert writers do. We need empirical research stated with enough specificity to provide guidelines for teaching novices how to perform as experts. In other words, we need research in the composition process that bears directly on the art of composition. Such research is often difficult to devise as there are many things we can learn about expert performance that have only marginal value in the training of novices.

It does not, for example, suffice to study the discourse of highly adaptive writers for criterial benchmarks and then tell novices to make sure their prose incorporates these same features. Empirical evidence¹ shows that highly adaptive writers make more frequent references to the concerns of potentially hostile readers than less adaptive writers. But simply telling novice writers to make more references to the readers' concerns will not necessarily generate more adaptive discourse.

An example shows why this is so. As a classroom exercise, one of the authors had students write a "defense" of abortion to a hypothetical group of "devout Catholics." The students were given detailed information about the concerns of Catholics and were explicitly instructed to "address these concerns in your defense of abortion." Many of the students made frequent reference to these concerns—but nonadaptive references (i.e., "your concern with the protection of fetal life is ignorant."). These writers followed the surface benchmarks of highly adaptive prose even though their own prose was not adaptive.

The lesson to be learned is that instructors, when teaching students to adapt persuasive discourse to opponents, must devise tasks that (1) force writers to practice the cognitive skills basic to adaptation—role-playing or pretending to share the hostile reader's point of view; (2) make it easy for the writer to monitor his or her own writing for its adaptiveness to the reader; and (3) allow instructors to measure adaptive skill in degrees rather than absolute terms. This last condition makes it possible to decompose adaptive skill into component skills. An instructor can then train (in natural sequence) a student in these component skills.

This paper contends that one type of ironic writing meets all three criteria. In the third and fourth section below, we develop an "irony game" with rules for marking adaptive skill and training exercises for improving writer's performance at the game. The first section examines the adaptive demands that the composition of irony makes. The second section discusses the more specific adaptive demands imposed by the type of (refutative) irony used in the irony game.

I. The General Adaptive Requirements of Ironic Composition

To motivate the adaptive requirements of irony, let us work toward a precise description of ironic discourse. As a first approximation, let us say that a discourse is ironic when its literal sense is perceived to be purposely "at odds with" the writer’s (or speaker’s) communicative intentions. This formulation accounts for the understanding of simple ironies. An ironist says "it's a beautiful day"' during a downpour. The understander perceives that the speaker’s communicative intentions undermines the utterance's literal sense.

Further reflection shows that this formulation includes too much. For there are many other discourse types which fit this description. Hidden requests, puns, some forms of metaphor, are all examples. Kant associated these subversions between form and intention as the basis of many jokes.² There is no humor in Thurber's joke: "Well, I'll say one thing for you," she remarked, "when you throw a party, it always hits somebody," if one does not recognize the literal sense collapsing with the unexpected shift of assumed communicative intents.

To distinguish irony from these other types of verbal subversions, we need a more restricted formulation. The restrictions we make are of two sorts: those on (1) the pragmatic function of the ironist's actions, and those on (2) the context by which readers make sense of these actions.

As for the pragmatic functions of the ironist's actions, we first need to examine what these actions are. Stated simply, the ironist composes pretended evaluations. False statements, even when deliberately composed, do not necessarily amount to ironies. A reader will likely recognize the statement "Columbus discovered America in 1900," as false rather than ironic because this statement lacks a clear evaluative point. Conversely, readers are more likely to recognize a sentence like "Einstein was dumb" as ironic than false because of its clear (but insincere) evaluative tone.

The question remains, what is the ironist's pragmatic aim when making such pretended evaluations? The ironist means to convey to the reader disavowal of the literal evaluation. The ironist who writes, "Einstein was dumb," wants to be understood as seriously renouncing this judgment. It is this serious intention to disavow their literal evaluations that distinguishes ironic writers from those jokesters who don't intend to convey any serious point through their words. As we will see in the following section, the irony game is played by having a writer disavow an opposing reader's position while pretending to advocate it.

As for the contexts in which readers make sense of ironic discourse, let us examine how these contexts are more restricted than the contexts in which literal evaluations are understood. Consider first the contexts in which a reader understands a literal evaluation. Suppose among the many statements a writer composes in a text is the statement "Capitalism is good.”

Some readers may have extensive background knowledge about the author (from surrounding text, previous works, reputation, etc.) and thus they will know the reasons behind the evaluation; others may have no such knowledge and will wonder aloud why the writer has come to make such judgments; still others may not be at all bothered by this lack of background. They will register the writer's opinion and not care to know what underlies it. Despite this wide variety of contexts (and readers), all such readers will understand the evaluative statement even though they will have understood it with varying degrees of specificity.

The understanding of an ironic evaluation, on the other hand, is limited to contexts (and readers) with special background knowledge. A reader cannot recover an irony without recognizing the contrast between the writer’s real and stated values. But this means that the reader of irony must know the ironist's real values. More to the point it means that the ironic writer cannot expect to communicate irony unless the writer adapts to the reader’s state of knowledge about these values. For this reason, writers must compose irony with special attention to the reader’s knowledge of, or at least capacity to infer, their real values. This is an adaptive constraint general to all writers who wish to communicate irony to their readers.

II. The Specific Adaptive Requirements of Ironic Refutations

The irony game bids a writer not only to address a reader ironically, but also to assume the reader an opponent. The writer is told that irony must be used to expose flaws in the opponent's position, flaws which a rational opponent would acknowledge. The purpose of this section is to show how this refutative use of irony imposes an additional set of adaptive constraints on the ironic writer.

Let us first examine the principles of argument by which irony can be used as a refutative strategy. These principles are connected with the principles underlying reductio arguments. Both the refutative ironist and the reductio arguer proceed from the opponent' s premises to draw conclusions they hope the opponent will disavow.

However, it would be a serious mistake to identify ironic refutation with reductio arguments. For unlike reductio arguments where the writer informs the reader of the pretended commitment (i.e., "let us assume for the sake of argument that P is true—although, of course, I really want to show P false"), ironic refutations are "dramatized" in the sense that the ironic writer does not make explicit mention of this pretense.

Ironic refutations achieve most of their persuasive force simply through their power to supply a dramatic replica of the opponent's position—while still showing the replica to be deficient or incomplete. In this way, the ironic refuter seeks to build the opponent's confidence before taking it away. All refutation, ironic or no, is essentially an "ironic" process in that a refuter can prove to be a convincing "outsider," only by proving to be a convincing (but subversive) "insider."

But in ironic refutation, the demand to write like an "insider" is especially pressing. This is because, whether adopting the voice of an insider or outsider, the literal refuter essentially makes arguments against the proponent of the to-be-refuted view. The ironic refuter, on the other hand, makes arguments as a plausible proponent of that view. The "insider's voice" is the ironist's chief weapon for attack. With the loss of that voice, the ironist's attempted refutation takes on the appearance of superficial attack. Because the irony game has writers undertake ironic refutations, it tests a writer's skill to write like an insider. It rewards writers when they retain the opponent's voice, and punishes them when they lose it.

In addition to the general adaptive constraints on ironic composition (constraints pertaining to the knowledge of the reader), the irony game also imposes constraints on the writer to attack the reader from within the reader's perspective.

III. The Irony Game

We are now ready to discuss the irony game in more detail. A writer plays this game by using irony refutatively, by attacking the views of an opposing reader through pretended praise. As a result of playing the irony game, the writer’s refutation can be evaluatively scored either as sarcastic, satiric, or Socratic. Sarcasm is a losing position because, in being sarcastic, a writer adapts to the opponent's claims, but not to the manner in which the opponent advances those claims nor to the premises the opponent actually uses when advancing those claims. Should a liberal democrat sarcastically "praise" Reagan's tax cuts as "a good way of ripping off the poor," the irony will not discomfit a conservative Republican who would never praise Reagan's tax cuts in those words and on the basis of those premises.

Satire is also a losing position but less so than sarcasm because the satiric ironist adapts not only to the opponent's claims but to the manner in which those claims are presented.3 Had our liberal Democrat said, "It will put a stop to welfare fraud," when ironically "praising" Reagan's welfare cuts, the writer would have been more satiric than sarcastic, mimicking not only the opponent's claims but also the "voice" of the opponent (who would make such claims). Note, however, that the writer would still not be faithfully adapting to the premises of the opponent, for conservative Republicans would not likely cite putting an end to welfare fraud as a major premise for advocating cuts in government spending.

To make these observations on satire even more concrete, let us draw from a writing protocol we took of a Swift expert ironically attacking the tenure system. The writer was told that his audience was a committee of tenured professors who were strong advocates of the tenure system. The writer was also instructed to do whatever possible to attack the tenure system but at the same time not offend his readers. He was told to make the irony sharp, but still subtle enough so that the readers might come to see real deficiencies with tenure.

We found that our Swift expert, in contrast to a sarcastic ironist, had the ability to "sublimate" his attack by replacing harsh words with more ambiguous evaluative terms that could reasonably have come from the opponent's mouth. When, for example, our expert originally "praised" tenure for contributing to the outpouring of papers, he softened this ridicule by revising outpouring to growth. When trying to describe what he believed to be the "sycophant assistant professors" who must curry favor with senior faculty, he revised his description to "the younger generation." When attacking senior professors for becoming deadwood after tenure, he made reference to our" colleagues who have settled into a more casual mode of research."

Through this strategy, our expert captured the opponent's tone of expression. It is less clear he captured the content. From the vantage of tenure opponents, his essay was a humorous and cogent indictment. It exposed the hypocrisy of tenure advocates who clothe unsound motives in dignified language. The rub is that a true proponent of tenure would not allow that tenure needs to be argued from such questionable motives. While a tenure proponent would agree with the literal claims made by our expert (i.e., that tenure is a sound and defensible system), and the dignified tone in which they were made, s/he would not agree with the premises from which our expert launched these claims.

Thus, if the satirical ironist provides a mirror of the opponent that casts a more " authentic" reflection than the mirror offered by the sarcastic ironist, it is still an imperfect mirror, it is a mirror that will fall short of giving the opponent full confidence that the writer has adapted completely; and so, when the satiric writer exposes flaws in the representation of the opponent's views, the opponent will be under no obligation to acknowledge these flaws as self-indicting. The Socratic stance represents a winning position in the irony game. Socratic irony presupposes more extensive adaptive skills than either sarcastic or satiric ironies.

The Socratic ironist achieves not only an accurate representation of the claims and manner of expression exhibited by the opponent, but also achieves a veridical representation of the major premises the opponent brings to bear when reasoning through to these claims. If the Satiric ironist adroitly mimics the opponent's mannerisms of argument, the Socratic ironist mimics in addition the chain of reasoning by which the opponent arrives at conclusions. Unlike the "less adaptive" ironies, the attacks leveled by the Socratic ironist are more likely to function as convincing refutations. For any attack on the opponent from the Socratic perspective will be an "internal" attack that uses the opponent's premises against him/herself.

We took the following example of a Socratic refutation from a writing protocol of a philosopher whose purpose was to attack affirmative action programs through ironic "praise."

A proponent of affirmative action would be hard pressed not to feel bothered by the irony in this passage. Many reasonable proponents make the distinction between hiring goals and quotas exactly as the ironist here. All proponents would have to acknowledge that administrators have been called in to explain why certain goals have not been met. Yet, if goals are really optional (as the affirmative action proponent contends), in what sense must administrators be held accountable for not meeting them? The writer's irony calls attention to a "gap" in the affirmative action position. Assuming that the affirmative action proponent feels confident that the writer has faithfully represented his position, the opponent must lose some of that confidence when exposed to this irony. We do not mean to suggest that the irony alone will force an opponent to recant a position. We do mean to suggest that it should force him/her to rethink the position more carefully-with the result that premises are added or deleted.⁴

As we mentioned in the first section, our major contention is that the irony game is a useful task for training students in adapting to an opponent. We can now see why this is so. We mentioned that any suitable training task would force (and not simply encourage) the student to adapt to the perspective of the opponent. The instruction to use irony forces the student writer to do just this. We also mentioned that the training task should be easy to self-monitor so that the writer knows whether the irony is Socratic, satiric, or sarcastic. As we will see in the next section, writers have an easy time monitoring their own adaptive performance when playing the irony game.

Finally, we suggested that the task should allow us to think of adaptation as a skill analyzable into several constituents rather than an "all or nothing" ability—and the irony game does allow this. Instead of having to say that the writer did/didn't adapt, the irony game allows us to say to what extent the writer adapted, whether to the opponent's claims, the opponent's manner of presenting those claims, or the premises on which those claims are advanced. This makes it easy to think of training tasks expressly designed to develop one or another of these component skills.

IV. Teaching a Student to Play the Irony Game

We conducted a series of three successive writing protocols on a single novice (i.e., a freshman) writer, asking him to play the irony game. Approximately three weeks elapsed between each protocol. For each protocol, we had our subject write about an issue in which he had expressed interest and about which he had expressed knowledge. For each issue, we prepared an "opposing position" which our subject was asked to read and to refute ironically. We gave our subject practice in the writing of irony and training in the differences between overt sarcasm, satire, and Socratic refutation.

Our interest was to see how much of the opponent's position (i.e., claims, manner, premises) our novice could faithfully incorporate in his attempt to refute that position. With respect to the study of composition processes, we were interested in acquiring a detailed record of how our novice adapted to his hypothetical opponent. With respect to the study of the art of composition, we were also interested in devising methods that would allow our novice to exercise more control over his adaptive strategies with each successive protocol.

In the first protocol, the novice writer, though explicitly warned against lapsing into sarcastic attack, reluctantly found that sarcasm was his principle refutative tactic. His first writing assignment was to address an opponent who thought the drinking age should be raised from eighteen to twenty-one. The opponent had conceded that such a law is difficult to enforce, but that it was nonetheless enforceable. Our subject sarcastically misrepresented the opponent by writing:

Our subject was aware of the misrepresentation as shown by the verbal transcript of his protocol:

Sure, enforcing a twenty-one year old drinking law is difficult—that’s not sufficient reason for there being no law at all. It is not—that is not—that is not a sufficient—is sufficient reasoning reason for not having a law at all. It is always best to have a law, it is always best to have—it is always better to have a law we can't enforce (laugh)—that's pretty sarcastic . . .

This example, one of dozens like it, shows how writers can monitor their own adaptive performance when playing the irony game.

Our subject also misrepresented the opponent's statement of “regret" that many innocent teenagers should be punished for the sins of a few. The subject represents the opponent as a careless reasoner.

After analyzing the first protocol, we hypothesized two explanations for the subject's inability to overcome sarcasm as a refutative strategy. The first had to do with his lack of verbal facility. We had written the opposition letters in a highly literate style. It is difficult to ask a person with a restricted vocabulary to mimic perfectly a more elaborated one. We thus could think of no expedient intervention to move our writer from sarcastic to satiric modes of attack—the latter depending upon a keen insight into the "insides" of words.

The second explanation for the subject's failure to move beyond sarcasm was his apparent inability to reconstruct the premises underlying the opponent's utterances. Instead, our novice seemed to cue on the opponent's utterances themselves. Part of the verbal protocol makes the subject's cueing strategy clear. When giving himself directions about how to undertake a subtle refutative strategy, he stated, "I just want to keep to their [the opponents] facts and turn them around and make them look really weird."

Although we could not train our novice in Swiftian satire, we could, it seemed to us, "extend" the range of his refutative perception from the opponent's utterances to the premises that supported them. To do this, we gave our subject a "press conference" before the second protocol. We specifically assigned him to play the part of the opponent and then fired questions at him, questions which forced him not only to become facile with the opponent's stated premises, but also to supply "plausible" premises when gaps were exposed in the opponent's position. This intervention helped, although not as much as we had hoped.

On the second protocol, our subject composed ironies that incorporated more of the opponent's premises than in the first. Nonetheless, despite long "build-ups" of faithfully represented premises, our novice still punctuated the refutative ending with sarcasm. The following is a typical passage taken from the second protocol. The opponent here is a military general who believes the U.S. should seek nuclear superiority over the Russians:

What started off as a promising line of refutation ends in a superficial equivocation on the implications of "initiative." Clearly, being a competent Socratic ironist requires more than just knowing the premises of the opponent. It also means being able to reason from those premises in an abstract way. We found from the second protocol that our novice had a comfortable understanding of the opponent's premises, but still could not reason from those premises so as to expose their limitations. What was our novice missing?

As mentioned above, the mechanisms of ironic refutation are similar to the mechanisms of reductio arguments. The reductio arguer not only knows the premises of his/her audience, but deduces general principles from those premises which can then be extended to contexts in which their application is absurd. For example, suppose an opponent claims that teenage drinking must be banned because it causes rowdiness or vandalism. The reductio arguer must take this premise and deduce a general principle like: anything which causes rowdiness or vandalism must be banned. Then the arguer must find a context where the application of this principle seems unacceptable: teenage parties also have been known to cause rowdiness and vandalism, so teenage parties must also be banned.

Prior to our third protocol, we trained our novice in deducing general 100 principles from the premises of an opponent, and finding contexts where the principle's application would be damning. More specifically, we gave our novice a list of twenty-eight argumentative paragraphs and asked him to extract from each paragraph an implied principle from which the conclusion seems to follow. Having identified the implied principle, our subject was then given practice in "searching for plausible contexts" where the application of the principle yields absurd or unacceptable consequences. Although the third protocol retains its share of sarcasms and stylistic infelicities, it does exhibit some strategies that pass for Socratic refutation. Consider the following passage excerpted from that protocol. The subject addressed an advocate of gun control legislation who had argued that people are likely to be injured when trying to protect themselves against a criminal intruder:

Although some of the paragraph's bite is lost because of its stylistic inelegance, the novice has moved in the direction of an honest refutation. He accepted the opponent's premises without distortion. He then elicited a plausible context (i.e., a life and death situation) in which the strength of the opponent's premises could rightfully be called into question. He had, more or less, "won" at the irony game.

V. The Irony Game in the Classroom

Thus far we have only talked of the irony game being played in conjunction with protocol studies, studies which are time-consuming, tedious, and expensive. True enough, protocol data is necessary in order to make empirical claims about the processes writers go through when playing the game. Yet there is no reason why the game can't be played in the classroom as the scorings (sarcastic, satiric, and Socratic) can easily be based on the product of the irony without tapping the process. Teachers can assign students to the task of attacking ironically an opponent and then judge the student product for its sarcastic, satiric, or Socratic features. Teachers can also use the intervention methods offered in this paper as means for improving student performances at the game.

While the irony game has uncontroversial classroom benefits as an exercise to sharpen adaptive skills, it remains to be seen whether the adaptive skills practiced by this game will directly generalize to the writing of nonironic persuasive essays. Could a person who has become an expert at the irony game automatically claim expertise in adapting to an opponent in a nonironic persuasive context? This question can be easily answered by giving one set of novices training in the irony game (to the Socratic level) and another set, no training. If there is a clear difference in these groups' abilities to adapt to the opponent in a literal persuasive environment, then this will be evidence that the adaptive skills developed by this game will generalize. The next phase of research on the irony game will be directed to assessing this question of generality.