Questions, Presuppositions and Fallacies

Abstract

In this paper I focus on the fallacy known as Complex Question or Many Questions. After a brief introduction, in Sect. 2 I highlight its pragmatic dimension, and in Sect. 3 its dialectical dimension. In Sect. 4 I present two accounts of this fallacy developed in argumentation theory, Douglas Walton’s and the Pragma-Dialectics’, which have resources to capture both its pragmatic and its dialectical nature. However, these accounts are unsatisfactory for various reasons. In Sect. 5 I focus on the pragmatic dimension of the fallacy and I suggest amendments to the accounts mentioned drawing on the study of the phenomenon of presupposition in theoretical pragmatics. I argue that the central notion in the definition of the fallacy is that of an informative presupposition. In Sect. 6 I focus on the dialectical dimension of the fallacy. This dimension needs to be explicitly acknowledged in the definition of the fallacy in order to distinguish it from a different, non-dialectical, fallacious argumentative move involving presuppositions.

1 Introduction

In the literature devoted to argumentation theory discussions of the so-called “Fallacy of the Complex Question” (FCQ henceforth) usually start with a display of examples taken to be paradigmatic cases of this fallacy. Here are a few of them:

• 1. Do you regret having been unfair to your political opponents? (Greco 2003: 227)

• 2. When did you stop cheating on your income tax returns? (Walton 1999: 379)

• 3. Why does a live fish placed in a bowl already full of water not cause it to overflow, whereas a dead fish does? (Hamblin 1970: 38, slightly modified)

Hamblin tells the apocryphal story of Charles II, who is said to have asked question (3) to the members of the Royal Society, misleading them into taking up the challenge to find an explanation of an alleged fact that actually does not obtain. A distinction is usually made between a direct and a corrective answer to a question. A direct answer to the question carries the presuppositions of the question, a fact acknowledged in pragmatic literature (see, for instance, Kadmon 2001: 115). In contrast, a corrective answer (Cross and Roelofsen 2020: §1.1), such as a reply by which the addressee rejects the question and the terms in which it is formulated, does not carry the presuppositions of the question. Questions such as the above invite the audience to give a direct answer, and so to accept the controversial presuppositions. Such a question, Hamblin comments, “carries with it a presumption which may prejudice an attempt to give a straightforward answer” (1970: 39). This dialectical move, Hamblin (1970: 38) adds, “seems designed to force [the addressee] into admission of” a proposition she would not otherwise be willing to accept. Or, as Jacquette (1994: 286) puts it, the questioner “is unfairly eliciting an admission from the answerer”. Such questions have been considered fallacious and included in the list of fallacies since Aristotle.

While FCQ has been studied extensively in argumentation theory,¹ the phenomenon of presupposition and its various effects on discourse has been a classical topic in the literature devoted to the pragmatics of natural language. Apart from specific accounts of how presuppositions affect the dynamic of conversation, pragmatic studies have emphasized the role presuppositions might play in misleading the audience, as they are likely to be accepted on the fly, without much critical inquiry.² In this paper, I draw on both the existent literature in argumentation theory and that on pragmatics in order to shed light on the phenomenon. In Sects. 2 and 3 I reconsider the pragmatic and dialectical nature of the fallacy. In Sect. 4 I introduce two dialectical accounts of FCQ present in the literature, Douglas Walton’s and the Pragma-Dialectics’. In Sect. 5 I draw on pragmatic theories of presuppositions and presupposition accommodation in order to obtain a better understanding of the mechanism on which the fallacy relies and to distinguish it from non-fallacious uses of presuppositions. In Sect. 6 I focus on the dialectical dimension, and I distinguish FCQ from a different, non-dialectical, fallacious argumentative move involving questions that carry problematic presuppositions.

In what concerns the label, I will use in what follows “The Fallacy of Complex Question”, as it is one of the most used labels, and it is not inaccurate. It indicates that there is some sort of troublesome complexity with such questions. It is not uncommon to come across the label “Loaded Questions”, although not always as a name for this particular fallacy. In turn, “The Fallacy of Many Questions” seems misleading.³ Åqvist (1965) rightly observes that this label “misleadingly suggests that what is wrong about questions involving false presuppositions consists in their involving two or more independent questions” (Åqvist 1965: 75, quoted in Walton 1981). But the speaker in the examples given above, and other classical examples of the fallacy, is not asking several questions with one interrogative sentence (as is the case with questions of the form “Where and why did so-and-so happen?”). In fact, the arguer is asking only one question that carries one or more presuppositions, which are implicit statements, not questions.

2 The Pragmatic Nature of FCQ

Most accounts of FCQ in argumentation theory acknowledge its pragmatic nature, but there are exceptions. For instance, Åqvist writes in the same text mentioned above:

I do not think that anybody would really contend that the Fallacy of Many Questions is committed by every risky question, i.e., by every question having some possibly false presupposition. The alleged fallacy is rather taken to be committed only by such risky questions as indeed have a false presupposition. (Åqvist 1965: 74–75, italics in the original)

According to Åqvist, wh-questions and yes-no questions that carry presuppositions are fallacious when the presupposition is false. But this characterization of FCQ is inadequate. It is not the actual truth-value of the presupposition that makes the questions in the above examples fallacious, but the fact that the presupposition is controversial in the context of the dialogue. The presupposition might even turn out to be true, but if it has not already been accepted by all parts in the dialogue, or if the interlocutor is likely to reject is, the question is still potentially fallacious. As Frank Fair (1973) notes, “it is not so much false presuppositions as unwarranted or improper ones that need to be singled out in practice for condemnation” (Fair 1973: 217). In fact, the actual truth-value of the presupposition is irrelevant. Suppose you and I falsely believe (and believe the other believes) that France is currently a monarchy and has a king. Suppose I ask you the question in (4), which carries a false presupposition (that there is a unique king of France), introduced by the definite description.

• 4. How old is the king of France?

There is no point in calling such a question fallacious, as long as we both share the (albeit false) belief that the presupposition is true. Introducing such a presupposition in dialogue is not a problematic argumentative move.

Thus, in order to characterize FCQ we should not consider the truth-value of the proposition presupposed, but its acceptability to the audience. In examples (1), (2) and (3) the relevant feature is that the questioner presupposes a controversial proposition, that the respondent would reject upon careful consideration. This is a pragmatic fact, that is, it concerns the speakers’ attitudes and commitments.

This observation is not new: apart from Fair (1973), whom I already quoted, Douglas Walton also emphasizes the pragmatic nature of this fallacy. He writes that, “a question is fallacious when it is used as a sophistical tactic to interfere with the respondent’s ability to retract commitment to allegations made by the other party who is asking the question.” (1999: 382, my emphasis) As a result, in a particular context a question might be fallacious with respect to one presupposition that it carries, but not with respect to a different one, as Walton (2006: 193) observes. For instance, question (1) carries at least two presuppositions: that the addressee has political opponents, and that she has been unfair to them in the past. We can easily imagine a context in which the question is fallacious with respect to the latter presupposition, but not with respect to the former.

3 The Dialectical Nature of FCQ

It has been also pointed out that FCQ has a dialectical nature. An argument for this claim could be obtained if we proceed in a roundabout way, starting with the question whether FCQ’s are genuine fallacies or not. The discussion of this question will provide us with insight into the dialectical nature of the fallacy.

Hamblin, among others, observes that apparently the speaker that commits an FCQ is not offering an argument at all, neither a good one, nor a bad one:

it remains to ask what relevance these examples have in a list of fallacies. A fallacy, we must repeat, is an invalid argument; and a man who asks a misleading question can hardly be said to have argued, validly or invalidly, for anything at all. Where are his premisses and what is his conclusion? (1970: 39)

If FCQ is not an argument then it cannot be a fallacy, Hamblin reasons. The perspective on fallacies Hamblin takes is that these are arguments of a certain kind. In particular, according to what has come to be known as “the Standard Definition of Fallacy” (although Hamblin does not use this term himself) a fallacy is an invalid argument that appears to be valid (Hamblin 1970: 12; 194). But no argument is conveyed by the questions we have considered, so Hamblin comments that ‘fallacy’ “seems to be misdescription” (1970: 40).⁴

The Standard Definition of Fallacy has been criticized on many accounts, but many of the modified versions maintain what Hansen (2002) calls “the ontological component” of the Standard Definition, i.e. the idea that a fallacy is an argument of a certain kind.⁵ Even definitions of fallacy that slightly modify the Standard Definition with respect to the ontological component, taking a fallacy to be “a pattern of argumentation” (Johnson 1989: 416; Johnson and Blair 1994: 54), or “an argument or at least something that purports to be an argument” (Walton 1995: 255; 2013: 215) still rule out FCQ from the list of proper fallacies. On any such view FCQ is not a fallacy because it is not an argument. This conclusion is problematic, since the use of such questions in argumentation seems to be, in certain contexts at least, fallacious.

One strategy to avoid the undesirable conclusion is to reconstruct FCQ as an argument. This is what Jacquette (1994) proposes. He suggests that FCQ “can be reconstructed as an inference sequence” (1994: 286), as follows:⁶

• 5. (a) You can only answer ‘yes’ or ‘no’ to the question you are now asked: ‘Do you regret having been unfair to your political opponents?’

  [(a.i) You have political opponents.] (presupposed by (a)).

  [(a.ii) You have at least at some time in the past been unfair to your political opponents.] (presupposed by (a)).

  (b) If you regret having been unfair to your political opponents (if you answer ‘yes’), then you have at least at some time in the past been unfair to your political opponents.

  (c) If you do not regret having been unfair to your political opponents (if you answer ‘no’), then you have at least at some time in the past been unfair to your political opponents.

  (d) Therefore, you have at least at some time in the past been unfair to your political opponents.

On Jacquette’s account, the question mentioned in (5a) is not a premise of the argument, but only the “vehicle” by which the presuppositions (5a.i) and (5a.ii) are introduced in the argument. These premises are written between brackets because they are implicit, similar to the missing premises of an enthymeme. The conclusion Jacquette draws is that FCQ is an argument which instantiates the fallacy of begging the question: “it is obvious that the fallacy [in (5)] is just a particular example of the Petitio Principi, since the argument’s premise in [5a] presupposes the conclusion.” (1994: 286).

Although I agree that that the argument in (5) is circular, given that the conclusion is to be found among the premises—in particular premise (5a.ii)—I think Jacquette’s reconstruction is not tenable.⁷ His account suggests that the audience might miss the fact that premise (5a.ii) repeats the conclusion and might be tricked into accepting the conclusion on the basis of the valid deductive argument formed with the rest of the premises, the structure of which is the following:

• 6. $A\bigvee B;A\to C$; $B\to C;therefore, C$.

But there is a problem with this account: although this argument is valid, it is not sound, because the first premise (5a) is obviously false. It is not true that you can only answer “yes” or “no” to the question asked. Other options are possible, as the respondent might simply refuse to answer the question in the terms in which it is asked.⁸ As a result of having an obviously false premise, the argument is not at all appealing to the audience.⁹ We end up with an argument that is obviously bad. But a fallacy is not an obviously bad argument, but one that has the potential to seduce careless arguers. So, the account is not successful in saving a place for FCQ on the list of fallacies. Something must have gone wrong.

The problem with the reconstruction lies with the premise (5a), which is the assertion that there is no other choice but to answer with “yes” or “no”. This assertion is obviously false, and it calls attention precisely to the existence of the alternatives to a yes-or-no answer. In contrast, the questioner in standard examples of FCQ such as (1) is doing something very different from advancing an argument such as (5). The questioner invites the audience to answer with either “yes” or “no”, and thereby to accept the presupposition of the question. In contrast, the speaker in (5) is giving an argument, that is, she is providing justification for a conclusion. Offering justification for a claim is generally thought to be the main aim of argumentation.¹⁰ But a question such as (1) does not offer justification. Instead, it is an attempt to obtain justification, and it is successful only in case the respondent indeed gives a direct answer, which carries the controversial proposition. The answer is a key ingredient of the fallacy. Only if—and after—the respondent answers the question, one could infer from the answer that, for instance, the respondent has been unfair to her political opponents (as this is what the answer presupposes). But in the argument (5) the respondent’s answer plays no role at all. The conclusion is reached before the interlocutor’s reply and independently of whether this answers the question at all. That is, Jacquette’s reconstruction fails because it does not capture the dialectical nature of the fallacy. Although Jacquette acknowledges that in a case of FCQ the questioner is “unfairly eliciting an admission from the answerer” (1994: 286), this aspect is nowhere to be found in the reconstruction he offers.¹¹

Many fallacies do not require the context of a dialogue. Their success or failure as strategies to gain the adherence of the audience does not depend on whether they are performed in dialogue or in a written text. But questions such as (1), (2) and (3) do require the context of a dialogue. Their success or failure as argumentative strategies depends on whether the interlocutor answers the question as it was posed. A question that is not meant to be responded—for instance, a rhetorical question—might be fallacious in other ways—as I argue in Sect. 6—but it is not a case of FCQ.

4 Two Accounts of FCQ

We have reached the conclusion that genuine cases of FCQ are dialectical, a feature bluntly ignored by attempts to reconstruct the move as a single argument, such as Jacquette’s (1994) analysis. For this reason, such attempts fail, and so, Informal Logic approaches that take fallacies to be arguments of some kind (Johnson and Blair 1994, Walton 1995) are poorly prepared to include FCQ on the list of fallacies. However, Walton (2010) accommodates the fallacious character of FCQ by introducing a distinction between sophisms and paralogisms. Paralogisms are instances of identifiable argumentation schemes, but sophisms need not be arguments: “the sophism type of fallacy is a sophistical tactic used to try to unfairly get the best of a speech partner in an exchange of arguments” (2010, 171; see also 1995, 254). Walton places FCQ in the category of sophisms, acknowledging its fallacious nature without attempting to reduce it to an argument structure. He also characterizes FCQ as a “sophistical tactic” and “a sophistical trick” (Walton 1999: 382–3; 2006: 203), which hints at its dialectical nature. According to this account, “a question is fallacious when it is used as a sophistical tactic to interfere with the respondent’s ability to retract commitment to allegations made by the other party who is asking the question.” (1999: 382).

A different approach is offered by Pragma-Dialectics. As van Eemeren and Grootendorst (2004) point out, “Since many questions depends on the dialogue situation, this fallacy can only be analyzed adequately by means of a dialectical approach.” (2004: 160) Theorists that approach argumentation from a dialectical perspective take fallacies to be moves in the argumentative discussion that contravene the purpose of the exchange. For Pragma-Dialectics, “the reason for considering an argumentative move fallacious is always that it is in some way or other prejudicial or harmful to the realization of the general aim of resolving a difference of opinion on the merits” (van Eemeren 2013: 147). Fallacies are characterized as “derailments of strategic maneuvering in which a rule for critical discussion has been violated”, but which “seem to comply with the critical discussion rules although in fact they do not.” (2013: 148–9) An essential point here is the following: “Strategic maneuvering does not only manifest itself in the complex speech act of argumentation, but also in all other speech acts performed in argumentative discourse that are pertinent to resolving a difference of opinion on the merits.” (2013: 145) This way, the definition of fallacy allows for including in this category certain questions, which are speech acts of a different kind than assertions and arguments. Questions are illocutionary acts that belong to the class of directives, according to Searle (1975: 356), while arguing is a complex speech act composed of elementary speech acts that belong to the class of assertives (van Eemeren and Grootendorst 1984: 34–35).

For Pragma-Dialectics, FCQ is a fallacious maneuver because it violates one of the ten rules of a critical discussion, more exactly, Rule No. 6: “A party may not falsely present a premise as an accepted starting point nor deny a premise representing an accepted starting point” (van Eemeren and Grootendorst 1992: 209). According to van Eemeren and Grootendorst (1992: 214), FCQ consists in “Falsely presenting something as a common starting point by wrapping up a standpoint in the presupposition of a question.”

In the next two sections of this paper I take a closer look at these two accounts, identify problems and propose amendments, in order to get a better understanding of the fallacy.

5 Revising the Theory: the Pragmatic Dimension

One central concept to an account of FCQ is that of a presupposition. This brings us to the question: what is a presupposition? The place to look for an answer to this question is theoretical pragmatics, where presuppositions are usually characterized as having three features. Presuppositions—whether triggered by declarative sentences or interrogative ones—are implications of a certain kind. Three features are usually invoked in order to distinguish these implications from what is said (the content of the speech act that carries the presupposition), semantic and logical entailments, conversational implicatures and other kinds of implications. The first one is that presuppositions are not-at-issue contents, backgrounded contents, or contents that are peripheral to another speech act, such as an assertion (Abbott 2006). A second characteristic is that the speaker takes for granted the proposition which is presupposed (Chierchia and McConnell-Ginet 1990: 281, Kadmon 2001: 10). And a third feature is projection: if an utterance of a sentence carries a presupposition, so will the utterance of a new sentence that is created by embedding the original sentence in specific linguistic environments.¹²

The three features delineate a particular kind of linguistic data but do not yet provide us with an account of why presuppositions behave the way they do. The most successful attempt to account for the data is developed by Robert Stalnaker (1974, 2002). Stalnaker explains the phenomenon of presupposition in pragmatic terms, on the basis of “general assumptions about rational strategy in situations where people exchange information or conduct argument” (Stalnaker 1974: 205).¹³ In what follows I introduce briefly the main ideas of this approach. This will shed light on some of the limitations of the accounts introduced in the previous section, although the improvements I propose in the following sections do not require assuming all the details of the Stalnakerian picture.

One important notion on this account is that of common belief: a proposition p is common belief in a group if and only if (almost) everybody in that group believes that p, and believes that (almost) everybody believes that p, and so on. Another important notion is that of a common ground. A proposition p is in the common ground of a group if and only if it is a common belief in the group that everybody accepts p.¹⁴ The notion of common ground is important in order to model the dynamics of a conversation. When an assertion is made, the content of this assertion is added to the common ground, provided no one objects. Now, presupposing that p is, according to Stalnaker, simply believing that the proposition p is in the common ground. Presupposing is, therefore, a propositional attitude. This notion of a person presupposing a proposition is linguistically relevant because utterances of certain sentences trigger presupposition, that is, they require, for lexical or other reasons, that a certain proposition be part of the common ground for the utterance of the sentence to be appropriate. For instance, sentence (7) is used appropriately only if the presupposition that John used to smoke in the past is already part of the common ground.

• 7. John stopped smoking.

However, occasionally a sentence that triggers a presupposition p might be appropriate to utter even if p is not part of common ground. The utterance of the sentence is appropriate if the participants in the conversation are willing to treat p as if it were already part of the common ground when the sentence is uttered. This phenomenon is called presupposition accommodation. David Lewis (1979: 347) is the first to have called attention to it, writing that “presupposition evolves according to a rule of accommodation specifying that any presuppositions that are required by what is said straightway come into existence, provided that nobody objects.” Thus, an utterance of sentence (7) could be appropriate even if the proposition that John used to smoke in the past is not part of the common ground at the moment of the utterance, as long as everybody accepts the proposition and adds it to the common ground on the fly. Normally, speakers exploit the mechanism of accommodation when they have reasonable expectations that the audience is disposed to add the proposition to the common ground on the fly, although the proposition is new information to them. In those cases, the speaker treats a piece of new information as if it were already known to everybody in the conversation and expects the audience to do the same. That is why these uses of presuppositions are usually called informative. Informative presupposition is a common mechanism that effective communication relies upon constantly. It is, as Seuren (2000: 280) comments, “an extremely powerful device for saving time and energy in linguistic communication”.

There are various common reasons why hearers may accommodate an informative presupposition, and speakers may expect hearers to do so. Greco (2003: 223–224) notes that audiences accommodate informative presuppositions when they trust the speaker, as for instance, in case the speaker is an expert on that subject or is known to be better familiarized with the topic than they are. In the case of (7) accommodation might occur if the audience believes the speaker is better acquainted with John’s smoking habits. Audiences might also accommodate a presupposition when it is not very important to them whether the presupposition is true or false, and so the cost of disrupting the conversation in order to discuss the proposition presupposed is not worth it. This is a case in which the audience accepts a presupposition for the purpose of the conversation without believing it is true. Another type of context is that of a literary work of fiction, and there are more.

Let us now go back to the accounts of FCQ that we have briefly introduced above. In various places Walton (1989: 28; 1999: 380) defines the notion of presupposition as follows: “a presupposition of a question is defined as a proposition that is presumed to be acceptable to the respondent when the question is asked…”. We see now that this definition does not quite capture the notion of presupposition (be it of a question or of some other speech act). When presuppositions are already part of the common ground, they cannot be presumed to be acceptable, as they are already accepted by all participants to the conversation. Walton seems to be having in mind the phenomenon of informative presuppositions, but there is no reason to suppose that all presuppositions are informative.

Let us consider now the core of Walton’s account of FCQ, which is the following:

“… not every question containing a presupposition that the respondent is not committed to should be judged to be a fallacious question… Such a question is fallacious when it is used as a sophistical tactic to interfere with the respondent’s ability to retract commitment to allegations made by the other party who is asking the question.” (1999: 382).

Notice that in the first sentence quoted Walton acknowledges that informative presuppositions have non-fallacious uses. They are only fallacious when the speaker aims to interfere with the respondent’s ability to refuse to commit to the allegations made by the questioner.¹⁵ But when does this happen? After all, questions carrying informative presuppositions always interfere (to a certain extent) with the respondent’s ability to refuse to commit to the presupposition, given that the questioner invites the respondent to accept the proposition presupposed on the fly, without paying special attention to it. But this is not always a fallacious move. When is it? A speaker who invites the interlocutor to accommodate a presupposition expects the audience to do so on the basis of contextually available shared reasons to accept the presupposition (for instance, epistemic reasons resulting from the speaker’s expertise on the topic). In the paradigmatic examples of FCQ given, such as (1), (2) and (3), there are no such reasons. The informative presupposition is unacceptable to the respondent, and the questioner knows this. Still the questioner asks the question in an attempt to get the former to commit inadvertently to the presupposition. Simply put, FCQ involves a presupposition of a question which is both informative as well as controversial. As we have seen, this means that the presupposition carries new information, not already part of the common ground, but which is presented as uncontroversial, ready to be taken for granted. However, if this information is actually controversial, presenting it as ready to be taken for granted is a sophistic, fallacious, move.

Pragma-Dialectics’ account is closer to the one I have just scketched the fallacy consists in “Falsely presenting something as a common starting point by wrapping up a standpoint in the presupposition of a question” (van Eemeren and Grootendorst 1992: 214). All informative presuppositions involve treating a proposition that constitutes new information as if it were already part of the common ground (that is, as a common starting point), when in fact it is not. But if that proposition is also a standpoint, it should not be presented as uncontroversial and ready to be taken for granted.

6 Revising the Theory: the Dialectical Dimension

We concluded above that FCQ is committed when the presupposition of a question is both informative as well as controversial (or, in Pragma-Dialectics’ terms, it expresses a standpoint). But this characterization is not precise enough, as it does not capture the dialectical nature of FCQ. Pragma-Dialectics has the resources to account for the dialectical nature of the fallacy, but the characterization it offers does not do so. Consequently, it does not manage to differentiate genuine cases of FCQ from non-dialectical fallacious uses of informative presuppositions.

We have seen that the paradigmatic examples of FCQ are cases of questions that elicit an answer from the audience. But, of course, many questions are not of this kind. Questions that fulfill other functions than requesting information or explicit answers are sometimes called “non-standard” (Ilie 1999, 2015: 3). One class of non-standard questions are rhetorical questions, a term generally used for questions asked in order to make an indirect speech act, that of asserting the obvious answer to the question (Huddleston and Pullum 2002: 773, 835, 921).¹⁶ According to Ilie (1994, 2015) the aim of rhetorical questions is to “convey the addresser’s commitment to its implicit answer in order to induce the addressee’s mental recognition of its obviousness” (2015: 4). For instance, question (8) is normally used as a rhetorical question that conveys the assertion that nobody gives a damn anyway.

• 8. Frankly, who gives a damn anyway? (Huddleston and Pullum 2002: 773)

Although a rhetorical question is manifestly formulated in a way that suggests a particular answer, that does not mean that the audience finds the answer, with its associated presuppositions, acceptable. A speaker might formulate rhetorical questions with presuppositions that she knows are highly problematic. Here is one example, from Copi et al. (2013), quoted from an essay on the topic of climate change:

• 9. With all of the hysteria, all of the fear, all of the phony science, could it be that man-made global warming is the greatest hoax ever perpetrated on the American people? (2013: 139)

This is a rhetorical question which suggests an affirmative answer. The question presupposes that there is a phony science behind the claim of a man-made global warming (introduced by “the”), which is a highly problematic claim. A second example Copi et al. (2013) give is that of an imagined homeowner who asks, regarding a proposed increase in the property tax:

• 10. How can you expect the majority of the voters, who rent but don’t own property and don’t have to pay the tax, to care if the tax burden of others is made even more unfair?

This rhetorical question presupposes, among other things, that the tax burden on homeowners is already unfair (introduced by “even more”), a presupposition that is controversial in the context. The suggested answer, which also carries this presupposition, is therefore not something that the audience is disposed to accept without further consideration.

Copi et al. (2013) treat (9) and (10) as genuine cases of FCQ. In fact, the authors consider that most FCQs are of this kind: they are “likely to be rhetorical, with no answer actually being sought.” (2013: 138) However, these rhetorical questions are different in a significant way from the examples of FCQ we have discussed in the previous sections. Examples (1), (2) and (3) are standard questions, and therefore essentially dialectical, while (9) and (10) are not. Contrary to Copi et al. (2013), I maintain that the latter are not instances of FCQ. Instead, FCQ is a dialectical version of a more general phenomenon that involves the use of presuppositions that are both informative and controversial, which also has non-dialectical versions.

The non-dialectical version of the fallacy is instantiated not only by rhetorical questions, but also by declarative sentences. Consider (11), which results from transforming the interrogative sentence in (9) into a declarative sentence:

• 11. With all of the hysteria, all of the fear, all of the phony science, it could be that man-made global warming is the greatest hoax ever perpetrated on the American people.

An utterance of (11), same as (9), carries the controversial informative presupposition that the science studying global warming is a phony science. But (11) is not a question but an assertion, and there is nothing particularly dialectical about it.

Let us now take a closer look at the non-dialectical version of the fallacy in order to get a better understanding of the contrast between the two. Why would an arguer convey a controversial claim by way of a presupposition, as in all these examples? The answer has to do with the fact that informative presuppositions have a special persuasive power, a phenomenon often discussed in the literature devoted to pragmatics. For instance, Ruth Wodak (2007: 214) writes that, in contrast to assertions or obvious conversational implicatures, “Presupposed content is, under ordinary circumstances, and unless there is a cautious interpretive attitude on the part of the hearer, accepted without (much) critical attention”. Simon-Vandenbergen et al. (2007) contains the interesting observation that, when presupposing a controversial proposition,

… the speaker does not simply decline to offer any recognition that the proposition is in some way problematic or subject to contestation (as is the case with non-presupposing bare assertions). They go beyond this to present the proposition as simply not at issue, as a proposition which can be assumed and hence need not be asserted. (2007: 47)

Mazid (2007: 356), Kopytowska (2010: 255), Hoinărescu (2015: 109), Langton (2018), Lotze (2019) have all remarked the persuasive character that presuppositions tend to have and have emphasized the role they might play in misleading the audience, as they are likely to be accommodated on the fly, without much critical inquiry. Marina Sbisà (1999), one of the first to have identified this phenomenon, discusses the following example of a persuasive use of presuppositions:

• 12. The anti-cancer treatment invented by Luigi Di Bella, the professor from Modena, scores another amazing goal in its own favour. (Il Giornale 17/12/1997, quoted in Sbisà 1999)

Sentence (12) carries at least two presuppositions: one is triggered by the definite description that is the grammatical subject of the sentence, and it is that there is an anti-cancer treatment invented by Luigi Di Bella. A second presupposition is introduced by “another”, and it is that this treatment has already scored amazing goals in the past. The newspaper had not discussed the issue in the near past, so, the two presuppositions carry information that is most probably new to the audience. The readers are likely to be persuaded that the propositions presupposed are true, and to accommodate them when reading the sentence. The writer achieves this—at least in part—on the basis of conveying the controversial propositions as presuppositions, and so, implicitly, by presenting the information as uncontroversial. Given that the information is highly controversial, the argumentative move is fallacious.

To sum up, we have distinguished so far two related fallacies: FCQ and a non-dialectical use of informative and controversial presuppositions (call it FNP, abbreviating “fallacious non-dialectical use of presuppositions”). There are various features that distinguish the two fallacious moves. First, while both are pragmatic in nature (being a particular exploitation of the phenomenon of informative presupposition), only FCQ is dialectical. With FNP the audience is not invited to say anything, but only to make a certain inference as part of the process of interpretation of the speech act. By presupposing the proposition, a speaker invites the audience to conclude that the proposition is uncontroversial, a candidate for accommodation.

A second difference, strictly related to the previous, is that FCQ involves the use of standard questions, while the latter might involve non-standard questions or speech acts performed with declarative sentences.

A third difference, also related to the ones mentioned, concerns the aims pursued by the two strategies: FCQ is a tactic to get the respondent to express commitment to a controversial proposition she would reject under more careful consideration. The success of the “trick” depends on whether the respondent gives a direct answer or rejects the question. In turn, the aim of FNP is persuasion, and not merely getting the audience to express verbal commitment to a controversial claim. While FCQ cannot be reconstructed inferentially as an inference (as I have argued in Sect. 3 above), FNP does have an inferential structure.

Do the accounts of FCQ we discussed above distinguish the two fallacies? Neither Walton’s account nor the Pragma Dialectics’ do a good job in this sense. Walton (1999) comes closer to acknowledging the dialectical character of FCQ, but there is no mention of the non-dialectical version of the fallacy (FPN, that is). In turn, the Pragma-Dialectics’ account of FCQ is too encompassing to distinguish the two fallacies that have been described here. Both FCQ and FPN involve “falsely presenting something as an accepted starting point”. More recently, in van Eemeren (2015), a distinction is drawn between “the fallacy of making unfair use of presuppositions in making assertions,” and “the fallacy of making unfair use of presuppositions in asking questions (= fallacy of many questions)” (2015: 419). This gets closer to capturing the contrast between FCQ and FPN. However, the real distinction that needs to be drawn is not between questions and assertions with controversial presuppositions. It is that between a dialectical move involving standard questions (for FCQ), and a non-dialectical fallacious move involving non-standard questions, assertions, or other speech acts that carry controversial presuppositions (for FPN).

In conclusion, I propose to define the two fallacies as follows: both involve informative uses of presuppositions, both involve controversial claims conveyed as presuppositions, but only one has a dialectical nature, while the other does not. FCQ is committed when a standard question is asked that carries a controversial informative presupposition and is aimed at obtaining the respondent’s verbal commitment to the presupposition through giving a direct answer to the question.¹⁷ In turn, FPN is instantiated by sentences (interrogative or not) that carry controversial informative presuppositions aimed at persuading the audience of the content presupposed.

The persuasive character of presuppositions, which I have mentioned above, might play a role in misleading the audience in both cases. In cases of FCQ, such as King Charles’ question to the members of the Royal Society, it might contribute to persuade the interlocutor to hastily ignore the problematic nature of the presupposition and to give a direct answer to the question. Accommodation, although generally a powerful device for saving time and energy in linguistic communication, is in the case of FCQ and FPN a bad move from an argumentative point of view, when the proposition presupposed is a controversial claim and needs to be carefully examined.