ISSA Proceedings 2006 – Question-Reply Argumentation Reconsidered: A Pragma-Dialectical Account Of How Questions And Answers Are Used In Critical Discussion
It is fairly well taken for granted that questions and answers can form the basis for an argumentative discussion [i] (see, e.g., van Eemeren, Grootendorst, Jackson, & Jacobs, 1992; Ilie, 1999; Walton, 1989). Yet, how questions and answers function in argumentative discussions and whether question-answer argumentative dialogues are subject to special soundness criteria are still fairly open questions. On the first point, as far as the existing extant literature is concerned, who exactly is allowed to make an argument – the questioner, the respondent, or both – is unclear. In his comprehensive analysis on question-reply argumentation, Walton (1989) suggests that questions are used to elicit premises for dialectical proofs from a respondent and that questions can be fallacious in that they trap the respondent into making certain concessions he or she should not make. In both cases, this appears to indicate that the actual arguer, our protagonist, must be the respondent. This leaves the role of the questioner, as far as his or her ability to actively participate in argumentation, at best unspecified and at worst, impossible. Determining what possible role the questioner can have in an argumentative exchange is the first issue taken up in this paper. It is argued here that the questioner should not, in principle, be only an elicitor of argumentation, but can also have an active role in making argument.
On the second point, Walton (1989) further suggests that question-reply argumentation is subject to special evaluation criteria. In his introduction, Walton (1989) explains the purpose of his research into question-reply argumentation as ” work[ing] toward establishing general guidelines that would enable a reasonable critic to approach a particular, given case of question-answer dialogue and to evaluate a question as reasonable or unreasonable in that given context relative to the given information” (p. 1). In this paper, I hope to demonstrate that despite Walton’s apparent claim, there is nothing particular special about question-reply argumentation that would necessitate unique standards for its evaluation. In contrast, this paper suggests that question-answer argumentation is simply an instantiation of critical discussion and can be reconstructed as such. Further, a justification is provided for why this is the case.
1. The roles of participants in question-reply argumentation
The characterization of question-reply argumentation suggested by Walton’s (1989) analysis, as presented above, has two central characteristics worthy of attention. The first is that questions are used to elicit dialectical proofs from the respondent. The second is that the person responsible for making the argument is the respondent. Example  demonstrates an exemplar case of what sort of argumentation this definition describes:
Polly: Does it rain? – (A?)
Annie: It rains. – (A)
Polly: If it rains, then does it pour? – (A à B?)
Annie: If it rains then it pours. – (A à B)
Polly: So, you must accept that it pours? – (B?)
Annie: Yes, it pours. – (B)
Example  shows what it would mean, using a very basic (and hypothetical) example, to use questions to get a dialectical proof. Polly leads Annie in justifying her conclusion B (“it pours”) though modus ponens reasoning by eliciting (or requesting) her commitments. This example is fairly formal and in such a way, only represents a narrow category of question-reply argumentation. Even Walton (1989) himself, while maintaining this formal definition of question-reply argumentation, considers cases of question-reply argumentation that are broader (or at least less formal) than what example  allows. In his analysis of Canadian parliamentary debates, Walton (1989) uses several examples of the questions that elicit argumentative answers that are not precisely formally dialectic in nature. An example of this sort of use of questions to elicit argumentation can be seen in the following example[ii]:
Between Paul Wolfowitz and a Newsweek reporter, Lally Weymouth, discussing Wolfowitz’s nomination (at the time) to head up the World Bank:
Q: Who is your biggest opponent among the Europeans? The French?
A: Well, I would say on the whole the reaction from the Europeans has been very constructive. They’re looking to make sure that if I’m approved that I have a good understanding of their concerns, one of which is the priority they attach to the bank’s work in Africa. I understand how important the bank is to Africa. (“The Last Word Paul Wolfowitz”, Newsweek 2005, p. 64)
Here, Wolfowitz gives an argument:
1. The reaction of the Europeans has been very constructive.
1.1a. They want to make sure I have a good understanding of their concerns, like the bank’s work in Africa.
1.1b. I understand how important the bank’s work is to Africa.
While the questioner suggests a track that Wolfowitz’s argument could take – “discuss why the French oppose you and what you think of that” – Wolfowitz is free, from a pragma-dialectical viewpoint, to put forward any sort of argument he wishes to (van Eemeren & Grootendorst, 1984; 2005). From a strictly informational perspective, the answer given by Wolfowitz may not be satisfactory as he does not provide the appropriate wh-content (who the biggest opponent is) and instead seeks to indirectly refute the presupposition that there is an opponent. The answer is evasive (Polcar, 2005), but it is an argument.
Thus, we see at this juncture at least two ways that questions and answers can be used argumentatively and both share one property in common: only the respondent makes an argument. In both cases, the questioner only exists to elicit argumentation, and while in a sense she can suggest certain lines of argument/certain standpoints that she wishes the respondent make, she does not actively argue. These two sorts then follow quite cleanly from Walton’s characterization of question-reply argumentation. Yet, the question still remains as to whether the questioner can take a more active role in the process of making an argument.
One well-researched area of argumentation that concerns questions and replies is in the use of questions and answers during legal proceedings, for example in American criminal trials. Anyone who has ever watched an episode of Matlock, Perry Mason, Law & Order, or the like, is familiar with how attorneys must get the arguments for their case from those who are testifying. A brief example should serve to illustrate how questions are used in the (criminal) legal context:
Excerpt taken from the victim’s testimony in People v. Jackson, the sexual molestation trial against celebrity Michael Jackson (March 10, 2005)
Q. Was there ever any occasion that you drank something other than wine?
Q. Tell us what it was.
A. I drank Bacardi, vodka and Jim Beam.
Q. Okay. And who provided you with the Bacardi?
Q. Who provided you with the Jim Beam?
Q. And who provided you with the vodka?
Michael Jackson was indicted on 10 counts, including performing lewd acts on a minor under 14, furnishing a minor with alcohol to facilitate lewd acts, and furnishing a minor with alcohol. Example  clearly relates to establishing that the victim not only drank alcohol as a minor, but that this alcohol was given to the victim by Jackson. In closing arguments, the prosecution referred to this excerpt, and others like it, as evidence for its argument that Jackson violated the law. What we see in this example, and in examples like it, is an indirect use of questions, and the resultant answers, to make argument. Here, the District Attorney asks questions to elicit arguments that he will use later in presenting his case before the trial jury. Thus, this example gives us yet another “type” of question-answer argumentation and is the first to allow the questioner the ability to take a direct role in a critical discussion.
Though Walton’s conception of question-answer argumentation was first discussed as, strictly speaking, allowing only the respondent in question-answer sequences to make argument, this is apparently only a surface reading of the possibilities a questioner has in arguing with questions and answers. As shown in example  a questioner can ask questions to elicit premises that he or she can use to later argue and as is apparent in example , a questioner can also use questions to directly argue with a respondent:
Between Terry Gross (TG), the National Public Radio host of the contemporary arts and issues weekday radio magazine Fresh Air, and Lynne Cheney (LC), wife of current American Vice-President Dick Cheney and the author of several children’s books on history (Fresh Air, March 9, 2005)[iii].
TG: In talking about the education of children which I know um is really important to you as an as the uh the former head of the NEH as someone who has written children’s books as a former educator yourself (breath) I want to ask you about something that the new head of the Department of Education, Margaret Spelling, did, (breath) she uh criticized public television for an episode of uh of um of um of a program um about an animated rabbit called Buster. And she asked pubic television to return the money that funded this episode and asked the network not to broadcast it.What happened on this episode is that this animated rabbit goes to Vermont to see how maple syrup is made and along the way he encounters a couple, two lesbian mothers, and that’s what Margaret Spelling found offensive. She said many parents would not want their young children exposed to the lifestyles portrayed in the episode. Um I I wonder what you thought of that move, as somebody interested in in how how information and history how how children are educated?
LC: Well I I certainly haven’t read uh Secretary Spelling’s letter but it does seem to me that when your um when you take up the question of sexuality it is something that when little kids are involved parents want to be in charge. And so I can I can understand that um this is not something the Department of Education would feel the government should be involved in and uh it seems to me that I read didn’t I though that some PBS stations are distributed the program so it is not a free speech issue because PBS has uh gone along those lines, but… [
TG: But this made me wonder about this question, if um if the Department of Education is uncomfortable about um uh uh a program that has two lesbian mothers in it um what about two lesbian mothers in real life who might go to an actual school or daycare center or kindergarten or sixth grade to pick up their their daughter or son. Should like what about the reality of them is that is that gonna make people uncomfortable? Do we – should they be escorted out because um it’s exposing our children to a lifestyle that some parents would be uncomfortable with?
LC: You know I think you have taken an issue of um public funding of public television to uh to a level that no one else has taken it to. Uh, the issue for in my mind is that uh government funding shouldn’t be teaching little kids about sexuality, it shouldn’t be uh teaching little kids about sex. It shouldn’t be uh um promulgating religious ideas there are um all sorts of things public funding shouldn’t be doing and that is one of them. And uh so, I think that uh uh the issue’s probably been resolved in the right way. he government has said we don’t do this and um at least some public television stations have said that this is a matter of free speech and we are going to distribute the programs. So insofar as these issues ever come to uh a resolution that um that seems to uh satisfy all sides I think this one did.
TG: But…t you could argue it is not about sex it’s just about you know … [
LC: It’s about sexuality.
You know little kids I think this is something that parents want to have want to be in charge of themselves, just as they want to be in charge of sex education – that’s why sex education in public schools is always uh a subject um that arouses a lot of passion on both sides.
TG: But its not it’s there’s nothing sexual going on there any more than uh if a husband and wife came in any more that would be sexual. I mean you aren’t describing their marriage in terms of sex, it’s just that, you know, it’s two people who who have a family together.
LC: I think that I think that parents probably want to approach the issue of homosexuality with their with their children themselves rather than having the government do it or rather than having public television do it. Though in this case if public television decides to do that um as an advocate of the First Amendment I think that’s fine.
TG: Lynne Cheney will be back in the second half of the show …..
Certainly, we see that Lynne Cheney makes an argument here, much like Wolfowitz did in example . Her argumentation is reconstructed in example :
1. You (Terry Gross) are making more of this situation than you should. (c8-9)
1.1. The situation (that some stations broadcast the program and some do not and no money should be returned to the government) has been resolved in the right way. (c13-14)
1.1.1. There is no violation of free speech (the First Amendment) here.
1.1.1a. Stations could make their own decisions about whether they were comfortable broadcasting this program or not without government interference.
1.1.1b. As an advocate of the First Amendment, that is fine (an acceptable decision).
1.2. It is explainable that [airing the “Postcards from Buster” program featuring lesbian parents] is something that the Department of Education would feel the government should not support/be involved in. (c4)
1.2.1. Public money cannot be used to promulgate religious ideas or to teach kids about sex/sexuality. (c10)
1.2.2. When you take up the question of sexuality it is something that when little kids are involved parents want to be in charge (not the government or media). (c2-c3)
Yet, Terry Gross seems to also participate in some argumentation – arguing whether the sub-standpoint of whether the Buster program was promulgating sex is acceptable (Cheney’s 1.2.1) – as is reconstructed in example :
1. This program is not about sex.
1.1. There is nothing sexual going on [in the program]. (t23).
1.1.1. It is the same as the program showed a (mixed-sex) husband and wife. (t24)
184.108.40.206. Generally, marriage is not described in terms of sex, it is just between two people who have a family together. (t25-t26)
This reconstruction only reveals a portion of what is occurring between Cheney and Gross, nonetheless, as is hopefully clear from the reconstruction, Gross, while still appearing to maintain her role as questioner, she also puts forward arguments against those given by Cheney. She is apparently in disagreement with Cheney on the issue of whether the lesbian couple featured in the Buster show actually advocates sexual choices (which, like the advocacy of religious ideas, does not warrant government funding) or is simply a demonstration of real-life (thus speech that can be governmentally supported).
Therefore, we see that when analyzing question-answer argumentation both the questioner and the respondent can have a role within the argumentative discussion. Identifying that the questioner can have a role in an argumentative discussion is not merely an exercise in over-precision or picking at definitions. There are some theoretical concerns that, on their face, might prevent us from considering at all the role of the questioner in argument. Arguments are thought of to consist of standpoints/claims that are defended by arguments/premises. One central component of this sort of idea is that argumentation is built up by statements. Using speech act terminology, standpoints should be assertives of some sort, which are supported by more assertives (see, e.g., Houtlosser, 1995; van Eemeren & Grootendorst, 1984). Questions are not assertive, they are instead directives – requests for information. Thus, there are reasons to ignore the role of questions in reconstructing argumentation since questions cannot “properly” be used as either standpoints or argumentation/premises. Yet, it has been fairly accepted, at least from the pragma-dialectical theoretical point of view, that there is absolutely no reason why standpoints and argumentation must be composed of direct assertives. For example, as Slot (1993) has demonstrated, rhetorical questions can be easily reconstructed as standpoints and argumentations (see also, Houtlosser, 1995, for reconstruction of standpoints in general in argumentative discourse). From this analysis then, we can be clear that when analyzing question-answer argumentation, both the questioner and the respondent can be parties in a dispute.
2. Question-reply argumentation as a “special” sort of argumentation?
That both the questioner and the respondent can make and respond to arguments during question-answer dialogues is also important to establishing whether it is necessary to propose special evaluation criteria for question-answer argumentation. To address this concern, it is necessary to think first about whether question-answer argumentation has a special goal that differs from ordinary argumentation. It is certainly the case that question and answers do not always have to be used argumentatively, and in principle, many activity types or discourse situations where questions and answers are the primary speech acts performed by participants do not have an overarching argumentative goal. For example, the “interview” discourse can be thought to contain information-seeking questions (from the questioner) and information-containing answers (from the respondent). While we can categorize the answers as argumentative, as was done in example , we can still say that the primary purpose of the discourse is not to resolve a disagreement or dispute, but to get information. Yet, this is no way precludes the use of questions and answers as argumentative. As shown by van Eemeren, Grootendorst, Jackson, and Jacobs (1993), many sorts of discourse activities that do not have the primary goal of resolving a difference of opinion can still be reconstructed argumentatively when argumentation occurs in these contexts. Van Eemeren, et al. (1993) use several sorts of examples to illustrate this point – ranging from witnessing on a college campus to child-mediation sessions – with each illustrating that while the primary goal of the discourse is not argumentative in nature, that argumentation can occur within these contexts.
Without fully recounting the arguments made by Van Eemeren, et al. (1993), how one is able to identify and reconstruct the argument has to do in part with applying pragmatic principles to the analysis of discourse. People sometimes do things with language when participating in a certain discourse that appear to, in principle, do absolutely nothing that relates the goal of the discourse. Thus, a divorcing couple who during a child-custody mediation session “digress” into personal attacks against one another, linguistic contributions that are irrelevant to reaching an agreement about who gets to see the child when, should and do appear on their face as totally irrelevant contributions that only frustrate the goal of mediation. Yet, if these digressions are analyzed as argumentation – that this divorcing couple is attempting to convince each other and/or the mediator that they are the only suitable parent to care for the child – sense can be made of these linguistic contributions. This is not at all to suggest that this argumentation is strong, but the reconstruction at least makes the utterances make sense and eventually allows a potential evaluation of the argumentation.
Walton (1989) appears to suggest that question-reply argumentation is a particular sort of discourse and in being so, requires a certain set of evaluative criteria. Yet, closer analysis reveals that Walton (1989) uncovers the sorts of questions and answers that are constructive to making an argument – or those that hamper open discussion – which certainly is a different concern altogether. Where he stands on whether we need to develop actual independent criteria for evaluating question-reply argumentation remains unclear, but from a pragma-dialectical perspective, there is absolutely no need for a special sort of criteria. Question-answer argumentation is just one way, of numerous ways, to engage in critical discussions and argumentation. Much as the “digressions” in mediations can be reconstructed, and thus evaluated, as argumentative, so can questions and answers when they are functioning argumentatively. The use of questions, and answers, can be argumentative, or informative, or what have you – what is important is that when they have argumentative function they can be reconstructed as argumentation and evaluated as such.
3. The reconstruction of question-reply argumentation
Therefore, the question remains, how do we know that questions and answers are argumentative, as opposed to information-seeking and giving, or whatever? Of course, this subject warrants an entire paper of its own. Nonetheless, I wish to provide a very rough answer here. The way we know is exactly the same way we know how “irrevelant” contributions in mediation function argumentatively, or how we know “irrelevant” contributions is other discourse/activity types function argumentatively. We refer, generally, to Grice’s (1969) Cooperation Principle (or something similar) and other pragmatic principles in order to determine the meaning of what people are communciating. In a nutshell, people say things that they mean their interlocutor to understand – when they are uncooperative (e.g., say something patently false), they are not being linguistic jerks – they intend their interlocutor to search for the implicated meaning that makes their contribution cooperative. This search for implicature can happen on a textual/global level or on an elementary/utterance level, yet in any case, the utterances are interpreted as saying more (or something different from) what was literally/directly communicated. While the sort of pragmatic theory adopted will change the justification of the reconstruction, using Grice (1969) and Searle (1979), we can see how this applies in the following example and reconstruction:
Don’t you think that rap music lyrics should be censored? I mean, all those violent lyrics cause all sorts of violence in society.
1. Rap music lyrics should be censored.
1.1. All those violent lyrics cause all sorts of violence in society.
While  should provide a perfectly reasonable and interpretable reconstruction of the argument presented, on its face  doesn’t literally communicate an argument. First, there is a question of some sort presented. As discussed above, this directly does not communicate a standpoint. To make things worse, when treating the question as a question, before there is a chance for an answer to be given, there is an answer provided. If standpoints must be assertive and questions must get answers (or in the case of rhetorical questions, that answers are unnecessary), this then violates all the rules of cooperative communication. So, how can we make sense of this? With reference first to speech act theory, we should understand that this question can’t be seriously intended as either information-seeking or rhetorical as it violates the felicity conditions for sincere information-seeking or rhetorical questions. Our next cognitive step would be to assume the person who uttered this example is acting uncooperatively, but as we apply Gricean theory, we seek a cooperative interpretation and therefore realize this example, as uttered, violates both the manner and the relevance maxims and so we seek an interpretation that makes the example cooperative (that makes the second sentence relevant and the first sentence clear). We can therefore realize that if we treat the question as an indirect standpoint and the second sentence as argumentation defending this standpoint, the entire example is absolutely understandable as argumentation.
How this process was explained for example  can be applied to discourses at large. When in an interview discourse type, the primary goal should be for the interviewer to elicit information from the respondent. When the utterances made by an interviewer diverge from this goal, there are two interpretations possible: the interviewer is being patently uncooperative and frustrating the goal of the interview just because they are irrational/mean/linguistically-jerky or the interviewer is being uncooperative because they are engaging in a different discourse activity and wish the respondent and audience to comprehend this. Looking back at example , Terry Gross received the information requested by the very first answer given by Lynne Cheney. Gross asked simply what Cheney thought about Spelling’s criticism and Cheney answered (to paraphrase) that it was probably reasonable and that, in any case, it was not of big importance. Yet, Gross continues to question Cheney about this very issue. If Gross’ overall discourse goal was truly to illicit information, there would be no subsequent turn by Gross on this issue. That she continues to question Cheney, even when Cheney has already presented her answer, indicates that either Gross is patently uncooperative or that she has some other reason to continue to “question” her respondent. One possible way to make Gross’ contribution continued questioning cooperative is to realize she is engaging in arguing, not simply information-seeking.
At this point an alternative explanation can be given that while Gross received some information, she did not yet receive complete information, thus the next questions she asked were directed at getting that information. This is unsatisfying however. First, Cheney’s answer was fairly complete – she gave her opinion as requested. Second, not only does Gross bother to introduce justification for her questions, which is odd (from a speech act perspective) if she simply wants information[iv], but she also stops questioning altogether in later turns in favor of asserting (“But you could argue it is not about sexuality” or “But it is not as if anything sexual is going on here”). Not only does the audience (and Cheney) understand that there are markers of disagreement present here, such as “but”, we should see that asserting propositions that contradict those made by the respondent are not in-line with the goal of getting (more) information, but in-line with the goal of resolving a difference of opinion.
Unfortunately, it is likely that there are no certain linguistic structures that exist that indicate a questioner is arguing instead of simply information-seeking or doing something else. Also problematic is that the analysis of the reconstruction of question-answer sequences as argumentative, as presented here, is incompletely argued. However, I believe that this cursory examination of how question-answer sequences can be reconstructed as argumentative at least provides some justification for understanding how questions and answers can be used argumentatively without having to specify certain criteria for the use of questions and answers in argumentation.
In many ways, what was presented in this paper should, hopefully, be uncontroversial. Ultimately, this paper was intended to simply detail what exactly question-answer argumentation entails. There were two major points presented here that I wish to draw attention to: first, in question-answer argumentation both the questioner and the respondent can have a role; and second, that there no special evaluation criteria needed to evaluate question-answer argumentation. Question and answers can be used argumentatively and reconstructed as argumentative when used in such ways. Finally, a preliminary way to approach justifying the reconstruction of questions and answers was presented. While certainly more explanation of how the principles of reconstructing questions and answers used in critical discussion is needed, I hope that at least a preliminary way of supporting how we can reconstruct question-answer argumentation was given.
i. In this respect, question-answer “dialogues”, such as interviews, press briefings, or ralk-showa, can be reconstructed as argumentative, when argumentative elements are present, just like other sorts of dialogues that exhibit argumentative properties. For examples of the reconstruction of argumentation present in non-exemplar cases of critical discussion, like mediation or witnessing, see van Eemeren, Grootendorst, Jackson, and Jacobs (1992).
ii. While this example does not come from Walton, it is representative of his examples. An example from Walton’s analysis of parliamentary debates was not included because of the length of his examples. However, see Walton (1989) for a fuller analysis. And in any case, this example provides is interesting given the presence of an evasive answer (Polcar, 2002).
iii. Transcript done by Leah Polcar. The original recording can be found in the Fresh Air archive linked from the National Public Radio website (http://www.npr.org).
iv. This is odd from, at least, a Searlean speech act perspective as a “true” information seeking question is uttered since the person asking does not know the answer. If any justification is to be given, it should a separate speech act of the type that excuses the person from not already knowing the answer (as in, “I am sorry, I should have understood this from the class readings, but what exactly does Dr. Smith mean by x on page 97?”).
Eemeren, F. H. van, & R. Grootendorst (2005). A systematic theory of argumentation. Cambridge University Press: Cambridge.
Eemeren, F. H. van, & R. Grootendorst (1984). Speech acts in argumentative discourse. Sic Sat: Amsterdam.
Eemeren, F. H. van, R. Grootendorst, S. Jackson & S. Jacobs (1992). Reconstructing argumentative discourse. University of Alabama Press: Tuscaloosa, AL.
Grice, H.P. (1975). Logic and conversation. In Cole, P. & Morgan, J. (Eds.), Syntax and Semantics, Vol. 3: Speech Acts (pp. 41-58). Academic Press: New York.
Houtlosser, P. (1995). Standpunten in een kritische discussie: Een pragma-dialectisch perspectief op de identificatie en reconstructie van standpunten. Amsterdam: IFOTT.
Ilie, C. (1999). Question-response argumentation in talk-shows. Journal of Pragmatics, 31, 975-999.
Polcar, L. E. (2005). The problem of evasion in question-answer argumentation: Are evasive responses fallacious or strategic communication? In C. A. Willard (Ed.), Critical Problems in Argumentation (pp. 466-473). Washington: National Communication Association
Polcar, L. E. (2002). Non-straightforward answers in question-answer dialogues. In F. H. van Eemeren (Ed.), Advances in Pragma-Dialectics (pp. 215-228). Amsterdam: Sic Sat.
Searle, J. R. (1979). Expression and meaning: Studies in the theory of speech acts. Cambridge: Cambridge University Press.
Slot, P. (1993). How can you say that? Rhetorical questions in argumentative texts. Dordrecht: Kluwer.
Walton, D. (1989). Question-reply argumentation. New York: Greenwood Press.