FLEXIBLE FACTS
A pragma-dialectical analysis of a burden of proof manipulation
Frans van Eemeren & Peter Houtlosser
University of Amsterdam
1Odd communications
Advertisements can be surprising.[1] About five years ago the following advertisement was during a series of months published in American and British magazines: “Carlsberg, probably the best beer in the world”. In the same period, the advertisers of the ‘fully rejuvenated’ Dutch morning paper Trouw [Trust] treated Dutch radio listeners to the following announcement – which has been repeated ad nauseam ever since –: “Trouw, perhaps the best newspaper in the Netherlands”.[2] It is not difficult to see that there is something awkward about these two slogans. But what is it? Is it the firm and at the same time surprisingly cautious wording that makes one wonder whether Carlsberg is indeed the best beer in the world and whether Trouw is indeed the best newspaper in the Netherlands? Or is it the fact that a qualifier has been added to the appraisal that is wholly out of tune with the convention of advertising that products are recommended only in the boldest terms (“Durex, the best there is”)? Or is it yet something else?
We shall attempt to shed some more light on this problem. For this purpose, we first analyse the Carlsberg and Trouw slogans dialectically as standpoints in a critical discussion. Next we give a pragma-linguistic interpretation of the way in which these standpoints are phrased. Finally, we explain what the advantageous effects of this particular way of phrasing a standpoint may be. In offering this explanation we apply our recently developed views concerning the ways in which parties engaged in a critical discussion may manoeuvre strategically to realise their rhetorical aim of winning the discussion.
2Standpoints in a critical discussion
In a dialectical analysis advertisements are, just like any other form of argumentative discourse, viewed as contributions to a critical discussion aimed at resolving a difference of opinion.[3] In an advertisement, the difference of opinion is whether or not the appraised product should be purchased. The advertiser is the protagonist who makes an attempt to convince the antagonist – the reader or listener – of the acceptability of the standpoint that the product should be purchased. In the advertiser’s message, however, this inciting standpoint usually remains implicit. As a rule, advertisers restrict themselves to advancing – and justifying – an evaluative standpoint in which the product they want to sell is positively assessed. It is merely implied that this positive assessment is at the same time the justification of the implicit standpoint that this product should be purchased.
In the Carlsberg and Trouw slogans evaluative standpoints are expressed.[4] In a dialectical analysis these standpoints must be situated in the confrontation stage of a critical discussion in which, ideally, the standpoint that is at issue in the discussion is presented. At the same time, the nature and the force of the standpoint are specified. Is the standpoint descriptive, evaluative or inciting? Is it upheld firmly or cautiously? In the confrontation stage, the presented standpoint meets with doubt on the part of the presenter’s interlocutor, so that a difference of opinion develops.
In the cases of Carlsberg and Trouw, there is, of course, no question of an explicit expression of doubt. There is no antagonist who responds and the discussion remains implicit – the standpoint is presented and that is all. The way in which Carlsberg’s and Trouw’s standpoints are phrased, however, causes a problem: it is not exactly clear what the dialectical positions are that Carlsberg and Trouw have assumed. On the one hand, the force of their standpoints is rather weak (‘probably’, ‘perhaps’), on the other hand the evaluation expressed in these standpoints is quite strong (‘the best beer’, ‘the best newspaper’). This makes it hard to determine what the burden of proof is that Carlsberg and Trouw would have to discharge in order to justify their standpoints in the argumentation stage of the discussion. The question, however, is now whether it is really the combination of a weak force and a strong evaluation in one and the same standpoint that leads to these problems of interpretation. Are these problems not caused by something else? A more detailed pragma-linguistic analysis may provide an answer.
3Standpoints and modalities
Our pragma-linguistic analysis of the Carlsberg and Trouw standpoints starts from the assumption that a standpoint is to be viewed as the verbal outcome of the performance of the speech act of advancing a standpoint. Just as any other speech act, this speech act has certain felicity – identity and correctness – conditions (van Eemeren and Grootendorst 1984). The most significant of the correctness conditions of the speech act of advancing a standpoint is the preparatory condition that says that a speaker or writer who advances a standpoint must assume that the listener or reader does not already accept the proposition expressed in the standpoint at face value (Houtlosser, 1995, 2002). This condition provides a useful clue for a more precise analysis of the standpoints advanced by Carlsberg and Trouw: it leads straight to the question why the speaker or writer expects the listener or reader not to accept the proposition expressed in the standpoint at face value.
What kind of answer can be given to this question generally depends on the nature of the proposition that is at issue. If this proposition is descriptive, the speaker or writer will expect the listener or reader not to consider the state of affairs represented in the proposition to be the case at face value. If the proposition is evaluative, it will be assumed that the listener or reader will not regard the value judgement represented in the proposition as necessarily shared by both parties. If the proposition is inciting, the speaker or writer will expect the listener or reader not to be immediately prepared to perform the action represented in the proposition. Therefore, the nature of the proposition expressed in a standpoint has consequences for the type of justification that is to be given in defence of that standpoint. In the case of a descriptive proposition, argumentation is called for that makes it credible (to the extent claimed in the standpoint) that the state of affairs represented in the proposition is indeed correct. In the case of an evaluative proposition, argumentation is required that shows that the value judgement expressed in the proposition is rooted in some normative common ground. In the case of an inciting proposition, argumentation should be advanced that underlines the urgency for the listener or reader to perform the action represented in the proposition.
The propositions expressed in Carlsberg’s and Trouw’s standpoints is in both cases evaluative. More precisely, in both standpoints a proposition is expressed that represents a subjective value judgement that must be assessed with the help of commonly shared assessment criteria and on the basis of some normative common ground for its intersubjective validity, not for its truth. Going by the nature of the propositions expressed in their standpoints, Carlsberg and Trouw have taken on the obligation to show, with the help of a set of intersubjectively shared assessment criteria and on the basis of a normative common ground, that their value judgements on the beer and the newspaper may claim intersubjective validity – in other words, that these judgements can – and should – be shared by their listeners and readers.
If advanced in its complete form, however, a standpoint expresses not just a proposition but also some specifications of the nature and the force of the speaker’s or writer’s commitment (van Eemeren and Grootendorst 1984; Houtlosser 1995; Snoeck Henkemans 1992). Some of these specifications are not related to the nature of the proposition expressed in the standpoint. This applies for example to the specification of the speaker’s or writer’s commitment to the expressed proposition as being positive or negative: “It is imperative to buy new shoes first and then a new pair of trousers, instead of the other way around” and “It is not imperative to buy new shoes first and then a new pair of trousers, instead of the other way around”. Other specifications, however, are closely related to the nature of the expressed proposition. Examples are the speaker’s or writer’s indication of his propositional attitude vis-à-vis the expressed proposition and the modal qualifications of his commitments to the expressed proposition. Both indications of propositional attitudes and modal qualifications of commitments have to be congruous with the nature of the proposition. Some combinations are ‘allowed’, other combinations are not. A descriptive proposition, for example, may not be combined with an indication of an ‘evaluative’ propositional attitude such as ‘in my view’: normally, the statement ‘In my view it is four o’clock’ is unacceptable. This is so because by definition ‘in my view’ introduces a personal opinion and ‘It is four o’clock’ is not a personal opinion.[5]
Conversely, an evaluative proposition expressing a value judgement may not be introduced by a descriptive modal qualification that specifies the extent to which the factualness of the state of affairs represented in the proposition is conceivable, such as ‘it is possible that’ (alethic modality), or knowable, such as ‘it is certain that’ (epistemic modality). Even if such combinations of value judgements and descriptive modal qualifications may merely sound a little bit odd in colloquial speech, conceptually they are absurd. Value judgements simply cannot be ‘the case’ or ‘not the case’. Whether something is the case or not is in principle an objective issue that is determined by the extent to which it can (conceivably or according to our knowledge) be considered a fact. Value judgements are merely ‘valid’ or not, i.e., valid for a particular person or a particular group of persons. Whether a value judgement is valid or not is not an objective but a subjective – and at best inter-subjective – matter.
In the Carlsberg and Trouw standpoints happens precisely that which we have just diagnosed as conceptually absurd: an evaluative proposition is combined with a descriptive modal qualification. In Carlsberg’s standpoint the evaluative proposition that Carlsberg is the best beer in the world is introduced by the epistemic modality ‘probably’; in Trouw’s standpoint the evaluative proposition that Trouw is the best newspaper in the Netherlands is introduced by the alethic modality ‘perhaps’. The answer to the question what makes Carlsberg’s and Trouw’s standpoints so awkward is therefore not that these standpoints are at the same time firm and cautious, but that they are at the same time evaluative and descriptive. They sound odd because they are absurd.
Should we now conclude that the advertisers of Carlsberg and Trouw are not in command of their own language? Or should we assume that they know very well what they are doing and that the reason why they are advancing an absurdity is that they try to gain a certain advantage that outweighs the ‘disadvantage’ of advancing an absurdity? Our pragma-linguistic analysis points already to the direction in which this advantage could be sought. After all, just as expressing an ‘evaluative’ propositional attitude such as ‘in my view’ seems to turn a descriptive proposition into an evaluative proposition (see also note 5), expressing a descriptive modality such as ‘probably’ or ‘perhaps’ seems to make an evaluative proposition descriptive. It is quite plausible that this is the very reason why these modal words are added. At the same time, it is obvious that neither Carlsberg nor Trouw is interested in presenting a claim that is merely descriptive. They want the public to agree that this beer is the best beer and this newspaper the best newspaper – and that this consent will incite them to dash to the shop where these products are for sale. We think that it can be explained why the advertisers nevertheless use a descriptive modality if our pragma-linguistic analysis is situated in a framework of analysis that combines the dialectical perspective of a critical discussion with a rhetorical perspective on the effects that are pursued in the presentation of a standpoint. In the next section we shall make clear how this combination can be realised.
4Strategic manoeuvring
In practice, standpoints are not usually put forward just to subject them to a critical testing procedure, but to have them accepted by one’s audience as well. Obviously, in most advertisements the latter aim will prevail. There is actually nothing against pursuing one’s ‘rhetorical’ intentions in addition to trying to live up to one’s ‘dialectical’ ambitions. In principle, there is no reason why the two should necessarily stand in each other’s way: the aims of resolving a difference of opinion and resolving the difference in one’s own favour are after all not incompatible. But the two aims may also clash. Therefore the arguers will manoeuvre strategically to prevent such a clash – or at least the appearance of such a clash – from occurring. They will do all they can to see their own position through while meeting their critical obligations at the same time. Such manoeuvring, however, can derail and merely serve to disguise that certain dialectical obligations have been ignored. To be able to determine whether a particular instance of strategic manoeuvring is sound or ‘derailed’ in this sense, an analytic framework is needed that makes clear what sorts of opportunities the dialectical discussion situation provides to maximize the chances of reaching a rhetorical effect by strategic manoeuvring, and how they can be exploited in an acceptable way.
A fruitful point of departure for developing the desired framework of analysis is the pragma-dialectical model of critical discussion. As we have argued earlier each of the dialectical goals of each of the four stages of a critical discussion can be rhetorically exploited by each of the parties to complete the stage concerned to their own advantage (van Eemeren and Houtlosser, 1998, 2002). In the confrontation stage, for example, the dialectical goal of externalising the difference of opinion can be rhetorically realised by the parties to come to a definition of the difference that suits their own purpose best. And in the opening stage, the dialectical goal of agreeing on procedural and material starting points for the discussion can be rhetorically realised by the parties to establish the point of departure that is most expedient for their own case.
In all stages, strategic manoeuvring will be aimed at maintaining a balance between realising dialectical goals and pursuing their rhetorical analogues. In our perspective, a party’s strategic manoeuvring is manifested in the selection that is made from the topical potential of the discussion stage concerned, in the assimilation to the other party’s beliefs and preferences, and in the use of specific presentational devices.
If we incorporate our pragma-linguistic analysis of Carlsberg’s and Trouw’s standpoints in the integrated framework of analysis just outlined, we can advance the hypothesis that the conceptual absurdities noted in these standpoints are the result of a deliberate attempt at strategic manoeuvring, in this case a form of strategic manoeuvring that utilises a certain presentational device. The next question is in what sense the presentational device used in presenting these standpoints is precisely strategic.
5Strategic absurdity
A strategic presentation of a standpoint in a critical discussion consists in phrasing the standpoint in such a way that it can be justified convincingly by the protagonist. Obviously, it cannot be predicted what will be a convincing justification in practice. In a concrete case, this ultimately depends on the concessions the other party is prepared to make in the opening stage of the discussion. The nature of the standpoint that is to be justified, however, plays an important role on a more abstract level. Standpoints of the descriptive type can in the best case be ‘proven’ whereas evaluative and inciting standpoints cannot. In principle, it can after all be decided objectively whether or not something is the case and it cannot be decided in the same sense whether something should be valued in a particular way or whether a certain action should be performed.
If it is not clear beforehand which concessions the other party will make, the protagonist of a standpoint cannot know for certain whether he will be able to give a convincing justification of the standpoint. In such cases it is difficult to create a solid point of departure for the discussion in the opening stage. One thing the protagonist can do is try to attribute the concessions to the antagonist that can be used effectively in the justification of the standpoint. Another possibility is to suggest that such concessions are not needed because the standpoint at issue can be justified on the basis of conclusive arguments, i.e., this standpoint is an objectively justifiable standpoint. The last strategy is precisely the strategy that Carlsberg and Trouw employ. By using the epistemic modality ‘probably’ and the alethic modality ‘perhaps’, they imply that their standpoints can be justified objectively. By using ‘probably’, Carlsberg suggests that it can be determined objectively whether Carlsberg beer is the best beer in the world and that chances are high that this is the case. By using ‘perhaps’, Trouw gives the impression that it could be an objective fact that Trouw is the best newspaper in the Netherlands and that it is not so silly to think that this may indeed be the case.
Instead of advancing one ambiguous statement, Carlsberg and Trouw make in fact two strategically motivated discussion moves. The first move consists in advancing the evaluative standpoint that Carlsberg is the best beer in the world and Trouw the best newspaper in the Netherlands. The second move consists in suggesting by means of ‘probably’ and ‘perhaps’ that these standpoints can be objectively justified. While the first move is really confrontational, the second move belongs to the opening stage of the discussion. It serves to suggest that there is agreement about the burden of proof that has to be discharged for the standpoints at issue to be defended satisfactorily.[6]
It will be clear that this way of strategic manoeuvring amounts to a manipulation of the burden of proof that is dialectically unacceptable. What, to be sure, would an objective justification of an evaluative standpoint amount to? Carlsberg and Trouw give the impression that it will be sufficient to make an inventory of the domain of entities that are to be compared or the criteria that are to be applied in the evaluation. They act as if such an inventory automatically yields the result that they suggest: “We have not yet tested all the existing brands of beer, but as soon as that has been done Carlsberg will (probably) turn out to be the best beer in the world”, “We do not yet know exactly what makes a newspaper a good newspaper, but once we have found out it will appear that Trouw is (perhaps) the best”. Yet, there is ample reason left for doubt. If it can indeed be established by means of a test that Carlsberg is the best beer in the world, why then has such a test not yet been run? And should we really assume that the editors of Trouw have not yet found out what a good newspaper is, but will do so soon?[7]