1 Fallacies as Violations of the Rules for Critical Discussion

In this article we revisit the pragma-dialectical approach to the fallacies in order to explain its design and development. Defining characteristics of the pragma-dialectical approach are the dialectical perspective on argumentative discourse as a critical exchange of argumentative moves between the parties in a difference of opinion and the pragmatic perspective on these argumentative moves as communicative acts with a specific function in the discourse. In order to resolve the difference, the critical discussion that is ideally conducted needs to go through four different stages: a confrontation stage in which the difference of opinion becomes apparent, an opening stage in which the starting points of the resolution process are established, an argumentation stage in which the standpoints at issue are defended against doubt or criticism, and a concluding stage in which it is determined to what extent the resolution has been resolved.

Unlike in the ‘logical standard approach’ to the fallacies discussed by Hamblin (1970), in the pragma-dialectical standard approach the fallacies occurring in argumentative discourse are not automatically viewed as violations of the logical validity standard, but can be violations of a variety of standards. These standards have in common that all of them pertain to the reasonableness of argumentative discourse, i.e., the appropriateness of the argumentative moves that are made for serving the purpose of resolving a difference of opinion on the basis of their merits. Logical validity is one of these standards, but this standard applies only to argumentative moves made in the argumentation stage of the argumentative process which are claimed to involve logical reasoning. To a great many of the argumentative moves made in the argumentation stage this claim usually does not apply, let alone to the argumentative moves made in the other stages.

The rules for critical discussion proposed in the pragma-dialectical theory of argumentation (van Eemeren & Grootendorst 1984; Eemeren & Grootendorst 2004) are incorporated in a less abstract practical ‘code of conduct’ for reasonable argumentative discourse, which contains standards representing the various kinds of soundness conditions that need to be fulfilled in making argumentative moves aimed at resolving a difference of opinion on the merits.Footnote 1 These standards define what argumentative discourse is like in a critical discussion in which reasonableness is fully maintained, i.e., when the discourse is optimally aimed at resolving a difference of opinion on the merits. Depending on what maintaining reasonableness in the various stages of the argumentative process involves, in addition to the set of communicative requirements that always apply, different standards apply to the argumentative moves made in the confrontation stage, the opening stage, the argumentation stage, and the concluding stage. Every argumentative move that is an infringement of any of the rules for critical discussion, whichever party performs it at whatever stage of the discussion, is in this perspective a fallacy (van Eemeren 2018: 62–69). The rationale for calling an argumentative move a fallacy is that it obstructs or hinders the resolution of a difference of opinion based on the merits of the argumentative moves that are made. The instrumentality of the rules for critical discussion in resolving differences of opinion on the merits and exposing counterproductive argumentative moves as fallacious, demonstrates their ‘problem-solving validity’.Footnote 2

A difference of opinion cannot be resolved by means of argumentative discourse if it has not become clear in the confrontation stage what the difference involves. The first rule of the code of conduct, the Freedom Rule (1), guarantees that in a critical discussion it is always possible to make clear which difference is to be discussed by requiring that standpoints and doubt regarding standpoints can be advanced freely. Even when the difference has been externalized, it cannot be resolved if the party that advanced the standpoint at issue is not prepared to take on the role of a protagonist defending this standpoint. The Obligation to Defend Rule (2) ensures that standpoints put forward and called into question will be defended against critical attacks. This defense cannot lead to a resolution of the difference if the antagonist who is to be convinced (or the protagonist who defends it) misrepresents the standpoint that is defended. The Standpoint Rule (3) ensures that this does not happen.

A difference of opinion is not resolved on the merits when the defense of the standpoint is based on argumentation that does not support this standpoint or merely on ethos or pathos. In such a case the defense does not comply with the Relevance Rule (4). The resolution of the difference is also obstructed when protagonists do not accept their responsibility for elements they have left implicit in their argumentation or antagonists do not stick to what can be assigned to the protagonist on the basis of a careful reconstruction. In such cases the Unexpressed Premise Rule (5) has been violated. A necessary requirement for an argumentative process that can lead to a resolution of the difference on the merits is also that the starting points are not used in an improper way. When something is treated as a starting point that is in fact not accepted or an accepted starting point is denied, the Starting Point Rule (6) is violated. The difference of opinion is, again, not resolved on the merits when explicit reasoning is advanced in the argumentation that is invalid in a logical sense: this goes against the Validity Rule (7). Neither is a difference of opinion resolved on the merits when the argumentation in defense of the standpoint at issue that is advanced in the argumentation stage does not rely on admissible argument schemes that have been used correctly, so that the defense is in agreement with the Argument Scheme Rule (8).

A difference is not resolved if the parties involved do not agree in the concluding stage whether or not the standpoint at issue has been conclusively defended. This means that the Concluding Rule (9) needs to be observed. Finally, a general prerequisite for resolving a difference on the merits in all stages of the argumentative process is that the participants do not create pseudo differences or pseudo solutions by not expressing or not interpreting their own and each other’s intentions as accurately as possible, without relying on not transparent, vague or equivocal formulations or inaccurate, sloppy or biased interpretations. This general prerequisite calls for observation of the Language Use Rule (10).

When taken together, the rules for critical discussion that constitute the code of conduct for maintaining reasonableness provide in principle all standards that need to be taken into account in resolving a difference of opinion on the merits. Every argumentative move made in the discourse that violates in the stage in which it is made a rule for critical discussion is a fallacy. In Argumentation, Communication, and Fallacies (van Eemeren & Grootendorst 1992: 93–127), a list of violations is provided that is, as a matter of course, not complete but gives a good indication of the different kinds of fallacies that can be committed in the various stages of the argumentative process to be passed through in resolving a difference of opinion. In this way, the various types of fallacies are in the pragma-dialectical approach systematically connected with the functional variety of standards that need to be observed in resolving a difference of opinion on the merits.

The fallacies distinguished in the literature can in this way be characterized more systematically and consistently. Certain fallacies that were in the traditional categories only lumped together are shown to have nothing in common and are clearly distinguished from each other while genuinely related fallacies that were earlier separated are brought together. The fallacy known as argumentum ad verecundiam, for example, has variants that are violations of different standards, which are in fact separate types of fallacies. In one variant a party makes an appeal to authority at the opening stage of the argumentative process by giving a personal guarantee of the correctness of a standpoint (“You can take it from me that every war leads to another war”). In this case, the fallacy is a violation of the Obligation to Defend Rule (2) that a party that advanced a standpoint is obliged to defend this standpoint if this is desired. Another variant occurs when a party is prepared to defend its standpoint in the argumentation stage but does so by only parading its own qualities. This fallacy constitutes a violation of the Relevance Rule (4), which outlaws non-argumentation and argumentation merely based on ethos or pathos. Yet another variant occurs when a party appeals in the argumentation stage to an authority that is in fact not an expert in the field the standpoint at issue relates to. The latter kind of fallacy constitutes a violation of the Argument Scheme Rule (8), which prescribes that the source referred to in argumentation from authority should indeed be an authority in the area concerned.

Other examples of variants of a fallacy which are not of the same kind when viewed from the perspective of resolving differences of opinion on the merits concern the fallacy traditionally viewed as an argumentum ad populum. In one variant, this fallacy constitutes a violation of the Relevance Rule (4), in another variant it is a fallacy that violates the Argument Scheme Rule (8). In contradistinction, the fallacy traditionally regarded as a variant of the argumentum ad verecundiam and the fallacy traditionally regarded as a variant of the argumentum ad populum which are both violations of the Argument Scheme Rule (8), are in fact variants of the same kind of fallacy when viewed from the perspective of resolving differences of opinion on the merits.

In addition, the pragma-dialectical approach to fallacies allows us to identify fallacies that earlier went unnoticed. So far, these obstacles to resolving a difference of opinion on the merits now distinguished as “new” fallacies were not named. They include declaring a standpoint sacrosanct (violation of the Freedom Rule, 1), evading the burden of proof by immunizing a standpoint against criticism (violation of the Obligation to Defend Rule, 2), denying an unexpressed premise (violation of the Unexpressed Premise Rule, 5), falsely presenting something as a common starting point (violation of the Starting Point Rule, 6), falsely presenting a premise as self-evident (violation of the Starting Point Rule, 6), denying an accepted starting point (violation of the Starting Point Rule, 6), and making an absolute of the success of the defence (violation of the Concluding Rule, 9).

2 Conventional Validity of the Theory

Next to ensuring problem-solving validity by being instrumental in resolving differences of opinion on the merits, the rules for critical discussion also need to be intersubjectively valid to the parties involved in the resolution process. Otherwise they cannot be helpful in doing their actual job of resolving the difference in actual argumentative discourse. As we made clear in Sect. 1, the problem-solving validity of the pragma-dialectical standards for reasonableness can be analytically determined on theoretical grounds. Their intersubjective validity, however, which lends them conventional validity in argumentative practices, can only be determined empirically.Footnote 3 In 1996 we therefore started the comprehensive experimental research project Conceptions of Reasonableness, which was completed in 2008 (see van Eemeren, Garssen & Meuffels 2009), to determine empirically whether the standards of reasonableness expressed in the rules for critical discussion of the pragma-dialectical code of conduct will be intersubjectively approved by the people involved in a difference of opinion. The research question of this project was: to what extent are the norms the arguers claim to use in evaluating argumentative discourse in agreement with the pragma-dialectical standards?

Instead of asking the research subjects directly for their views about the rules for critical discussion, in each of our tests the respondents were asked to judge the reasonableness of discussion contributions in which a specific discussion rule was violated. Otherwise they might be forced to make pronouncements about abstractly formulated matters that are too theoretical for them. In the tests, a variety of discussion fragments consisting of short dialogues between two interlocutors were presented to the participants. In order to create a baseline to make comparisons, the participants also had to judge discussion fragments which resembled the fallacious cases in appearance in which no rule for critical discussion was violated: non-fallacious direct personal attacks, for instance, were paired with abusive ad hominem fallacies.

For all our tests, we constructed clear-cut paradigmatic cases of the fallacies, always without loaded content. The argumentative dialogues concerned were put in a domestic, political or scientific context. The participants were invariably asked to judge the reasonableness of a particular contribution to the exchange that was offered to them. They had to indicate their judgment on a 7-point Likert scale, ranging from very unreasonable (= 1) to very reasonable (= 7), with neither unreasonable nor reasonable (= 4) in the middle.Footnote 4 In this way, we examined 24 different types of fallacies which are violations of rules for critical discussion spread over all four discussion stages. The rules violated are the Freedom Rule (Rule 1), the Obligation to Defend Rule (Rule 2), the Argument Scheme Rule (Rule 8), and the Concluding Rule (Rule 9). The general aim of the tests was to check to what extent the norms applied by ordinary arguers in judging the reasonableness of argumentative moves match the standards of reasonableness expressed in the rules for critical discussion.Footnote 5

Table 1 provides an overview of the reasonableness scores for all violations included in our research project.

Table 1 Overview of average reasonableness score for fallacious contributions and their reasonable counterparts (1 = very unreasonable; 4 = neither unreasonable, nor reasonable; 7 = very reasonable)
Full size table

In all cases the results were quite consistent. As can be seen in Table 1, except for the tu quoque variant of the argumentum ad hominem, all fallacious contributions to the discussion are rated below the neutral point (4) on the 7-point scale, and hence considered unreasonable. The non-fallacious counterparts of the fallacies are generally considered reasonable. The respondents prove to be consistent in making a clear, and statistically significant, distinction between the discussion moves that are unreasonable according to the pragma-dialectical standards because they involve a fallacy and the discussion moves that are not fallacious. Fallacious discussion moves are in general considered unreasonable by the respondents and non-fallacious moves are considered reasonable. These outcomes provide strong support for the hypothesis that the pragma-dialectical discussion rules are intersubjectively valid and are therefore entitled to be recognized as conventionally valid.

3 Fallacies as Derailments of Strategic Manoeuvring

The dialectical goals pursued in real-life argumentative discourse by the argumentative moves that are made always have a rhetorical analogue. When arguers are aiming to resolve a difference of opinion on the merits in actual argumentative practices, they are engaged in strategic manoeuvring in which they try to realize effectiveness through reasonableness (van Eemeren, Garssen & Meuffels 2012a). In principle, they thus combine, in every argumentative move they make in the discourse and every mode of strategic manoeuvring employed in making it, their dialectical and their rhetorical aims.

In order to cover the strategic manoeuvring taking place in argumentative discourse, in addition to the dialectical dimension of pursuing reasonableness, a rhetorical dimension of pursuing effectiveness is included in the extended pragma-dialectical theorizing (van Eemeren 2010). The strategic manoeuvring taking place in the argumentative moves made in the discourse consists of trying to keep a balance between these two dimensions of argumentative discourse in combining aiming for effectiveness with maintaining reasonableness (van Eemeren & Houtlosser 2002: 138–141). When any of the rules for critical discussion constituting the code of conduct for reasonable argumentative discourse is violated in the process, the strategic manoeuvring derails into fallaciousness (van Eemeren 2010: 187–212; van Eemeren 2018: 120–123).

In actual argumentative practices, fallacious derailments of strategic manoeuvring may occur in the empirical equivalents of all four stages of the argumentative process. In the initial situation, i.e., the empirical equivalent of the confrontation stage of a critical discussion in the argumentative discourse, the arguers will aim to define the difference of opinion at issue in the way most suitable to their purposes. In the equivalent of the opening stage in the discourse, they will try to establish the starting points that agree most with their own purposes – material as well as procedural starting points. In the equivalent of the argumentation stage, they will advance argumentation consisting of arguments selected for being optimally instrumental in realizing their purposes. In the equivalent of the concluding stage, finally, they will try to reach the conclusion that comes closest to the outcome they are out to achieve. In all stages, in the strategic manoeuvring something can go wrong that involves a violation of a rule of the code of conduct for reasonable argumentative discourse, and therefore amounts to committing a fallacy.

The extended pragma-dialectical approach enables us to do justice to the treacherous character of the fallacies and their potential persuasiveness in actual argumentative practices: it makes it easier to explain how the various kinds of fallacies “work” and why they may go unnoticed, so that they can be effective without being reasonable. Above all, the concept of strategic manoeuvring can help us understand why in certain cases sound and fallacious argumentative moves are hard to distinguish. One of the causes of this problem is that in argumentative discourse there is a presumption of reasonableness that is almost automatically attached to all elements in the discourse (Jackson 1995). The fact that in argumentative discourse reasons are offered in support of standpoints makes some people consider it likely that the treatment of these standpoints will be well-considered. By providing reasons, the arguers who advance argumentation are regarded as indicating that they respect the Principle of Reasonableness (van Eemeren & Houtlosser 2009 in van Eemeren 2015: 632; van Eemeren 2010: 32, 253).

Another reason why the difference between sound and fallacious argumentative moves is in some cases not immediately clear is that in their manifestations sound and fallacious argumentative moves are not distinguished from each other by their appearance. In each particular case, both of them are representatives of the same mode of strategic manoeuvring and in that sense they are one of a kind. Sound as well as fallacious uses of a personal attack, for instance, are manifestations of one and the same mode of strategic manoeuvring. In studying the role of argumentative discourse in resolving differences of opinion on the merits, however, it is necessary to make a clear separation between the two. Therefore, in the pragma-dialectical theorizing we make a terminological distinction between the mode of strategic manoeuvring that is called a personal attack and the fallacious use of a personal attack which is called an argumentum ad hominem. The same kind of terminological distinction is also made, for instance, between the neutral label argument from authority and the fallacious version of it, an argumentum ad verecundiam. The fallacious versions of the argumentative moves are generally given a more technical, often latinized name.

Just like in the case of a personal attack, if the required soundness criteria have been complied with, the use of an argument from authority can be a reasonable and effective mode of strategic manoeuvring. However, the strategic manoeuvring derails when, for instance, the authority appealed to does not pertain to the topic at issue, cannot be attributed to the source referred to or when this source is wrongly quoted or on a point where having it is not relevant (Woods & Walton 1989: 15–24; van Eemeren & Grootendorst 1992: 136–137). In these cases one or more of the critical questions associated with the use of an argument from authority cannot be answered satisfactorily, so that the Argument Scheme Rule (Rule 8) is violated and an argumentum ad verecundiam has been committed.

Another cause of problems in distinguishing between sound and fallacious argumentative moves is that certain modes of strategic manoeuvring have a continuum of possible realizations that goes from argumentative moves that are unquestionably sound to argumentative moves that are indisputably fallacious. There may also be manifestations of a mode of strategic manoeuvring that are situated in-between the two extremes, so that it is not always immediately clear to which category they belong. One can imagine, for instance, that there are personal attacks or arguments from authority that are neither clearly sound nor clear specimens of an argumentum ad hominem or an argumentum ad verecundiam – not because they are hard to interpret, but because they do not represent one of the outspoken extremes of the modes of strategic manoeuvring concerned that are for the sake of didactic clarity highlighted in textbooks.

If only for fear of loss of effectiveness of their discourse, arguers do not want their argumentative moves to be perceived as fallacious. In order to prevent this from happening, they will be inclined to try to stretch the scope of the soundness of the mode of strategic manoeuvring concerned to such an extent that their fallacious argumentative move is made to appear sound. An argumentum ad hominem, for instance, is then presented as if it were a personal attack that is fully justified in the circumstances in which it is made and relevant to resolving the difference of opinion at issue. Or an argumentum ad misericordiam is portrayed as an appeal to pity that is fully justified in the context concerned and decisive. In this sense, the inclination to strategically hide fallaciousness in argumentative discourse by moving the goal posts is another cause of the problem that sound and fallacious argumentative moves are sometimes hard to distinguish.

The deceptiveness that makes fallacies sometimes hard to detect is still increased by the fact that argumentative moves that are sound in the institutional context of the one communicative activity type (or cluster of communicative activity types) may be fallacious in the institutional context of another communicative activity type (or cluster of communicative activity types). This means that the criteria for deciding whether an argumentative move is to be regarded sound may vary to some extent depending on the institutional macro-context of the communicative domain in which the argumentative discourse takes place. As a consequence, the general standards for reasonable argumentative discourse incorporated in the rules for critical discussion that constitute the pragma-dialectical code of conduct need to be specified in accordance with the requirements of the communicative activity type, or cluster of communicative activity types, to which they are applied. This contextual differentiation may be another cause of the problems involved in distinguishing between sound and fallacious argumentative moves in actual argumentative practices.Footnote 6

4 Exploiting Hidden Fallaciousness

The results of our comprehensive empirical project ‘Conceptions of reasonableness’ described in Sect. 2 indicate that when confronted with clear cases of violations of rules for critical discussion people consistently judge these argumentative moves as unreasonable. The fact that most violations of the pragma-dialectical rules are emphatically rejected as unreasonable contributions to the discussion leads to the question how it can be explained that fallacies so often occur in oral and written argumentative discourse without being recognized as such by the listeners or readers.

In real life argumentative discourse, fallacies will generally not be committed straightforwardly and usually they will not appear in a clear way. This makes it much harder to recognize them straightaway as unreasonable discussion moves. Before they can be detected, it first needs to be determined to what extent the discourse concerned can be reconstructed as a critical discussion aimed at resolving a difference of opinion on the merits. For a better understanding of the problem of recognizing fallacies “in the wild”, we turn to the theory of strategic manoeuvring (van Eemeren 2010). Arguers themselves know that fallacious moves are unreasonable and may be detected by others. In order to prevent this from happening, they will sometimes try to manoeuvre strategically in such a way that their fallacious argumentative moves do not look unreasonable.

As we already explained, certain modes of strategic manoeuvring, such as a personal attack or an appeal to authority, are not automatically unreasonable by themselves, but only when in using them in the case concerned a rule for reasonable argumentative discourse has been violated. A personal attack, for instance, is in principle a legitimate argumentative move if it conveys the content of a standpoint that is fully appropriate in the institutional macro-context concerned: a politician may certainly call another politician unreliable without being considered unreasonable when making this personal attack does not involve an infringement of the Freedom Rule. In the same vein, it is also quite legitimate to criticize an arguer who wrongfully refers to himself or herself as an expert regarding a certain topic for misusing authority argumentation and who is thus committing an argumentum ad verecundiam.

In certain cases, strategic manoeuvring by using the fallacy of an abusive ad hominem can have a reasonable appearance because it mimics a legitimate reaction to such an abuse of authority argumentation (van Eemeren, Garssen & Meuffels 2012b). When arguers present themselves wrongfully as experts in a certain field or claim to be trustworthy when in fact they are not, it is perfectly reasonable to attack them personally about that. Due to the existence of such special cases, it is not always immediately clear whether a personal attack must be regarded as reasonable criticism or as a fallacious argumentative move which is an argumentum ad hominem. In two experiments we have tested empirically the hypothesis that abusive ad hominem attacks are considered substantially less unreasonable by people when they are presented as critical reactions to authority argumentation in which the person attacked parades as an authority (van Eemeren, Garssen & Meuffels 2012b). In both experiments this hypothesis was confirmed.

In a similar vein, the argumentum ad baculum can also be camouflaged. This fallacy counts as a violation of the Freedom Rule because by making a threat the arguer prevents the interlocutor from advancing a standpoint or doubt. While in its straightforward presentation this fallacy is recognized right away, the arguer can present it as reasonable by making the threat look like a well-intended advice to the interlocutor, so that its fallacious character is harder to recognize. This is made possible because the speech acts of advising and threatening share a couple of important characteristics that enable the arguer to present the threat in their strategic manoeuvring as a warning for something they are not really responsible for (van Eemeren, Garssen & Meuffels 2015: 316). The hypothesis we tested experimentally is that arguers will be more inclined to consider an argumentum ad baculum as reasonable that could also be seen as a piece of advice rather than a straightforward argumentation ad baculum. The results of our empirical research indicated that this was indeed the case (van Eemeren, Garssen & Meuffels 2015: 316–319).

In the argumentum ad consequentiam a non-legitimate step is made from a normative premise to a descriptive standpoint. By means of this argumentative move, the arguer tries to defend a descriptive standpoint by pointing at the negative consequences of the state of affairs mentioned in the standpoint. The argumentum ad consequentiam comes in two variants. The first one is based on a causal claim: the standpoint is true because it leads to a positive outcome. This variant resembles in its structure the pragmatic argument scheme. The difference is that in pragmatic argumentation the standpoint at issue is by definition prescriptive and in case of the argumentum ad consequentiam descriptive. In strategic manoeuvring to camouflage the unreasonableness of an argumentum ad consequentiam, the standpoint is presented ambiguously, so that it could be descriptive as well as prescriptive. This can be done by using phrasings such as x should (not) be seen as y or x should (not) be regarded as y.

The second variant resembles ad absurdum argumentation. In this fallacious variant, the reason that is given is: X is true, because if X is not true, then Y is true, and Y is not desirable, while in the reasonable ad absurdum argumentation the reasoning is as follows: X is true because if X is not true, then Y is true and Y is not true. The difference between the fallacious and the non-fallacious versions lies in the (bridging) premise: Y is not desirable vs. Y is not true. The descriptive proposition in the reasonable version becomes normative in the fallacious version. To avoid detection of the fallacy, in the strategic manoeuvring a phrasing is therefore sometimes chosen that allows for both interpretations (untrue and undesirable) (Garssen 2016: 251).

In an experiment we tested the claim that the camouflaged pragmatic variant of the argumentum ad consequentiam would be regarded less unreasonable than the straightforward argumentum ad consequentiam. Like in our other investigations of hidden fallaciousness, we found no empirical grounds for rejecting our hypothesis (van Eemeren & Garssen 2019: 331–332).

So far we focused in the hidden fallaciousness project on the principle of mimicry: by way of an ambiguous presentation, the fallacy is given a reasonable appearance. The arguer guilty of using a fallacy then counts on the presumption of reasonableness and hopes that the listener of reader will choose the reasonable interpretation. Other strategies that may have similar effects will have to be investigated. An example of such a strategy is making it hard for the interlocutor to ask relevant critical questions.

5 Contextualisation of the Study of the Fallacies

The standards for reasonable argumentative discourse that need to be observed to prevent fallacies from occurring are in the pragma-dialectical theorizing supposed to be agreed upon in the empirical counterpart of the opening stage of a critical discussion. This does not mean that in practice there is always an actual discussion between the participants about these procedural starting points: in argumentative reality, especially in strongly conventionalized communicative activity types such as a civil lawsuit, in a great many cases the crucial starting points, including certain evaluation procedures, are already given when people start taking part in a particular type of communicative activity.Footnote 7 Generally participation is in fact only possible on the condition that the participants know and respect these starting points. Nevertheless, in a great many communicative activity types there will still be some room left for procedural deliberations, particularly in informal communicative activity types where a formally approved conventionalization of the communicative activity type is lacking, such as a chat between friends. As a rule, such deliberations will start from the frameworks of starting points already familiar to the participants from their upbringing at home, in school or as members of a specific social or professional institution, i.e., from their primary or secondary socialization.Footnote 8

The evaluation procedure for argumentative discourse may in the various institutional contexts be implemented in slightly different ways. For the detection of fallacies it is therefore necessary to carefully examine all soundness criteria pertaining to a certain mode of strategic manoeuvring in order to determine whether they need to be specified for being applied to a particular argumentative practice – and if so, in exactly which way.Footnote 9 For the various communicative domains this may result in the articulation of (slightly) distinct sets of specific soundness criteria for particular modes of strategic manoeuvring in specific communicative activity types or clusters of communicative activity types. In the communicative activity type of a criminal trial in the legal domain, for example, the critical questions pertaining to appealing to authority which must be dealt with in the discourse will differ in some respects from the critical questions pertaining to appealing to authority in a scholarly paper or in other communicative activity types. In a criminal trial it is, for instance, appropriate to ask whether a witness whose testimony is used to support a juridical claim is indeed reliable, while in an academic debate and in certain other communicative activity types asking such a critical question would be inappropriate.

It always depends on the institutional point that is to be realised in a certain communicative activity type, its conventionalisation and the institutional preconditions going with it, which requirements are pertinent to that communicative activity type. In our future research concerning the fallacies, next to further studies examining the various kinds of fallacies and formulating their soundness conditions by making explicit the critical questions associated with carrying out the modes of strategic manoeuvring concerned, we need to situate these modes of strategic manoeuvring in the institutional macro-contexts of the various kinds of communicative activity types of argumentative reality in which they have been used. By reflecting upon the rationale of their conventionalisation captured in the institutional points of these communicative activity types and the practical circumstances of the institutional macro-context in whey they are carried out, we can determine the primary and secondary institutional preconditions constraining them and find out how the general reasonableness standards are to be implemented in the institutional macro-context concerned. In this way we can give a precization of the specific standards of reasonableness that need to be taken into account in that context in order to resolve a difference of opinion on the merits.