In J. Verschueren (ed.), Language and Ideology. Selected Papers from the 6th International Pragmatics Conference, Vol. 1. Antwerp: Internatonal Pragmatics Association, 1999, pp. 492-509.

 

 

 

 

 

 

IDEOLOGY AND THE PERSUASIVE USE OF PRESUPPOSITION1

 

Marina Sbisā

 

 

1.   Foreword

 

Pragmatic presuppositions have been defined as assumptions shared by the interlocutors, which form the background of their ongoing discourse (Stalnaker 1973, 1974). Some if not all of these shared background assumptions have linguistic markers: thus Stalnaker (1973) (followed, among others, by Soames 1989) has spoken of presupposition requirements, while others have spoken of presupposition inducers or triggers (Levinson 1983, van der Sandt 1988), referring to those lexical items, morphological devices or syntactic constructions the use of which may require or activate a pragmatic presupposition.2 Inasmuch as this paper is an exercise in the pragmatic analysis of discourse and an attempt to reflect more deeply upon the theoretical concepts which make this possible, it will take into consideration only linguistically marked presuppositions. Therefore, I shall speak of presuppositions of utterances (Soames 1982, van der Sandt 1988) rather than of presuppositions of speakers. However, according to the received view of pragmatic presupposition, whenever an utterance has a presupposition its speaker may be said to make that presupposition.

It has long since been observed that presuppositions may have informative uses (Karttunnen 1974, Stalnaker 1974). In other words, when an utterance contains a presupposition inducer, the required presupposition still may not be included among the beliefs shared by the interlocutors and, in particular, it may be new information for the listener, who will "accommodate" the presupposition by adding it to the shared background beliefs (Lewis 1979). Such informative uses of presuppositions are frequent in the press, at least in Italy. A discussion of some examples drawn from the Italian daily papers will show how presuppositions are used to convey information. It will also become apparent how the received distinction between presupposition proper and informative presupposition does not reflect the reality of social communication processes, in which more subtle distinctions are involved.

This paper will focus on the fact that when what is presupposed has to do with values, social norms or ideals, or with perspectives on facts which are proper to a specific social agent, then informative presuppositions seem to serve persuasive aims. It will tackle the question of whether the received notions of pragmatic presupposition and of presupposition accommodation can account for the persuasive use of informative presupposition, and will argue that they cannot; a revised account of presuppositional phenomena will be put forward which can explain their persuasive function.

On the basis of the proposed description, it will be argued that presupposition is suitable for transmitting a kind of content which might be called ideological: assumptions, not necessarily conscious but liable to be brought to consciousness, about how our human world is and how it should be. By the same token, criticism of ideology may avail itself of a knowledge of linguistic presupposition inducers and of their communicative dynamics.

 

 

2. Informative and persuasive uses of presupposition in the Italian daily press

 

The headlines in Italian newpapers3 often contain linguistic presupposition inducers. It may be the case that the triggered presupposition is satisfied by the assumptions shared by the headline's author and the reader, but in many cases this is not at all certain. The reader may never have read a certain previous issue of the newspaper in which the information, now presupposed, was explicitly stated. Or he or she may have forgotten about that piece of information. In principle, it is also possible that the information concerned was never actually stated explicitly in any issue of that newspaper. It is therefore sensible to consider all presuppositions which convey a piece of information not explicitly stated in the same newspaper issue as at least partly informative.

Consider the following headline:

 

(1)        Anche Scalfaro dā l'ultimatum a Bossi

            'Scalfaro too delivers an ultimatum to Bossi'

            (La Stampa 8/9/1997, p.5)

 

The presupposition inducer anche 'too' triggers the following presupposition:

 

(1a)      Someone other than Scalfaro has delivered an ultimatum to Bossi

 

Scalfaro is the president of the Italian Republic and Bossi is the leader of the Lega Nord ('Northern League'), a party proposing federalism or even the independence of Northern Italy. If the reader knows that some days earlier, the Italian premier Romano Prodi had threatened to take Bossi to court if he insists in his claim that the electoral consultation his party is organizing in Northern Italy represents the election of a real Northern Italian autonomous Parliament, then the presupposition does belong to the set of assumptions shared by the author and the reader. But nothing is explicitly stated elsewhere in the same issue of the paper so as to ensure that this information is readily available to any reader before he or she reads the headline in (1). So the presupposition is likely to function as an informative one and, in the context of that issue, can be counted as such.

It is interesting to notice that the generic information conveyed by the headline's presupposition can be integrated by information conveyed by the article that follows it (which, however, is usually read after and not before the headline). In fact, in the article we find:

 

(2)   Scalfaro plaude a Prodi e al governo per il richiamo sulle elezioni padane.

'Scalfaro praises Prodi and the government for the reprimand about the Po Valley elections'

 

Here again, however, the information that Prodi and the government have reprimanded someone about the Po Valley elections is not explicitly stated but encoded in a definite description (il richiamo sulle elezioni padane 'the reprimand about the Po Valley elections'), the most classic of all presupposition inducers.

Our analysis of this example amounts to saying that a reader who is not quite sure who has given an ultimatum to whom about the Po Valley elections, is informed by (1) and (2) - thanks to the presuppositions they trigger - that the premier Prodi, prior to Scalfaro's ultimatum, has reprimanded Bossi about the Po Valley elections in a manner which can be considered equivalent to an ultimatum.

   Informative presuppositions play a prominent role in articles containing scientific information or involving reference to scientific knowledge. Consider:

 

(3)     L'Unione Europea era arrivata a Kyoto con la richiesta pių avanzata: una  riduzione dei gas nocivi che producono l'"effetto serra" (principalmente diossido di Ca) del 15% rispetto ai livelli del 1990.

'The European Union had come to Kyoto with the most radical request: a reduction in the harmful gases which produce the "greenhouse effect" (mainly carbon dioxide) by 15% with respect to the levels of 1990'

(La Stampa 10/12/1997, p.8: "Un giorno per salvare la Terra" 'A day to save the Earth')

 

Here, whether or not the reader has already been informed and remembers that the so-called "greenhouse effect" is produced by certain gases which are therefore harmful, he or she can gather this information from the presupposition triggered by the definite description i gas nocivi che producono l'"effetto serra" 'the harmful gases which produce the greenhouse effect'.

Sometimes, articles of political commentary use presupposition to activate assumptions they do not argue for:

 

(4)          Tutti i partiti ex cattolici sembrano assolutamente incapaci di dare una risposta a Giovanni Paolo II (...) sulle "emergenze" nazionali che egli ha individuato (...): politica a sostegno della famiglia, scuola libera e creazione di lavoro vero e stabile. Neanche si accorgono che D'Alema, su questi temi, tenta di scavalcarli (...)

'All the ex-Catholic parties seem quite incapable of giving an answer to John Paul II (...) about the national "emergencies" he has indicated (...): policies in support of the family, parity between state and private schools, and creation of jobs which are real and stable. They do not even realize that D'Alema, on these issues, is  trying to get there before they do'

(Il Giornale 9/9/1997, p.1: Antonio Socci, "La mosca cocchiera di D'Alema" 'The fly that pretends to drive D'Alema's coach')

 

Here the information that D'Alema, leader of the Democratic Party of the Left, is trying to supplant the traditional role of the Catholic parties with respect to some political issues singled out by the Pope, is conveyed as presupposed, since it is embedded by the factive verb accorgersi 'realize'.

Often, however, what is presupposed by a neswpaper headline or article has to do with values, social norms or ideals, or with perspectives on facts which are proper to a specific social agent. In September 1997, the daily press used different definite descriptions to refer to the electoral consultation organized by the Northern League for October 26: we find, for example,

 

(5)        le annunciate "elezioni" padane

                        'the announced Po Valley "elections"'

                        (Il Corriere della Sera 4/9/1997, p.1)

 

(6)        le elezioni del 26 ottobre

                        'the elections of 26 October'

                        (La Stampa 4/9/1997, p.5)

 

(7)        il voto padano

                        'the Po Valley vote'

                        (Il Corriere della Sera 8/9/1997, p.1)

 

(8)        le elezioni-sondaggio di ottobre

                        'the election-opinion poll of October'

                        (Il Giornale 12/9/1997, p.7)

 

The presuppositions triggered by these definite descriptions are somewhat different from each other and reflect the different attitudes of the newspapers or of the authors of the articles towards the political role of the alleged elections.

 

(5a) Something has been announced that someone is calling the PoValley "elections".

(6a) There will be elections on October 26.

(7a) There is or will be a vote in the Po Valley.

(8a) There will be an election-opinion poll in October.

 

(5a) and (8a) do not entail there being elections at all. (6a) and (7a) simply say that there will be elections, and although (6) and (7) are embedded in contexts which under one reading at least would filter their presuppositions, their most direct impact on the reader tends to activate them.

In these examples, it should be noted that the choice of an appropriate definite description is used in place of an assertion or even an argumentation to the effect that the Po Valley elections are, or are not, real elections. Since different political evaluations are associated with the two opposite assumptions, it is manifest that the use of a given definite description is not merely informative, but has persuasive aims: in particular, there is the aim to get the reader to believe that the Northern League really is going (or is not really going) to make the people of Northern Italy elect their own autonomous Parliament.

Another interesting case in which informative presuppositions shift towards being persuasive is the following:

 

(9)          La terapia anticancro ideata dal professore modenese Luigi Di Bella segna un altro clamoroso punto a favore. 

'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, p.1: "Il pretore ordina: usate l'anticancro Di Bella" 'Order of the judge: use the Di Bella anti-cancer treatment')

 

An inspection of the issue containing this article and of the preceding issues of the same newspaper authorizes us to consider the presuppositions triggered by the definite description la terapia anticancro ideata dal professore modenese Luigi Di Bella 'the anti-cancer treatment invented by Luigi Di Bella, the professor from Modena' and by the iterative un altro 'another' as informative ones. The Di Bella case had been absent from the press for months, and had in any case never received a great deal of attention. So readers were very likely to know nothing about it. Newspapers use various strategies to cope with this problem, ranging from inserting a brief summary of the previous stages in the story into the middle of the article (an explicitly assertive strategy), to drawing largely on informative presuppositions and on that feeling of familiarity with the presupposed news which they are able to produce.

A strategy of the first kind is used by La Stampa, where, however, the explicit report of the previous steps is introduced by a presuppositional expression (note the iterative riaprire 'reawaken'):

 

(10)        La (...) decisione č destinata ora a riaprire le polemiche sulla utilitā del metodo del professor Di Bella.

'The (...) decision is now destined to reawaken the controversy surrounding the usefulness of Professor Di Bella's method'

                        (La Stampa 17/12/1997, p.13)

 

A strategy of the second kind is the one used by Il Giornale in the article from which example (9) was drawn. Thus the following presuppositions are triggered:

 

(9a)        An anti-cancer treatment exists which was invented by Professor Luigi Di Bella from Modena.

(9b)        The Di Bella anti-cancer  treatment has already scored some amazing goals in its own favour.

 

which provide the reader with pieces of information he or she is unlikely to have and create the impression that there is an ongoing story which began some time ago, and which also try to get the reader to take it for granted that Di Bella has invented an anti-cancer treatment and that the goals scored in its favour are amazing. In fact, Il Giornale was about to launch a campaign in favour of the Di Bella therapy, and the persuasive use of informative presuppositions in this article was just a first step.

Another use of presuppositions which I take to be aimed at persuasion consists of conveying accusations or criticism without spelling them out explicitly. This use is specially frequent in Il Giornale, a right-wing newspaper clearly opposed to the government. Thus, in the same issue from which example (9) is drawn, we find:

 

(11)        La terapia del professor Di Bella, boicottata e osteggiata dal ministero della Sanitā, (...) finisce in Parlamento.

'Professor Di Bella's treatment, which has been boycotted and opposed by the Ministry for Health (....), ends up in Parliament'

(Il Giornale 17/12/1997, p.10: "Polo e Ulivo al parlamento: basta ostracismi al farmaco" 'Polo and Ulivo to the Parliament: stop ostracizing the drug')

 

Here an accusation levelled at the government (that it is boycotting Di Bella's treatment) in included in a presuppositional construction. Another example of the same strategy, drawing on the use of a factive verb, is the following:

 

(12)        (...) Eugenio Scalfari (...) vi ha avvisati: lasciate perdere le soluzioni forti (...) Ma so che non lo ascolterete e non ascolterete neppure me, perché un anno e mezzo al potere vi ha drogati.

'(....) Eugenio Scalfari has warned you: forget about strong solutions (....) But I know you won't listen to him or to me either, because a year and a half in power has intoxicated you'

(Il Giornale 4/9/1997, p.1: Vittorio Feltri, "Gli ercolini gonfiano i muscoli" 'The little Hercules show off their muscles')

 

In consideration of the fact that the author addresses himself to the politicians presently in power, the triggered presuppositions can be expressed as follows:

 

(12a)      The politicians presently in power will not listen to Eugenio Scalfari or to the author of the article.

(12b)    The politicians in power are intoxicated by a year and a half in power.

(12c)    The politicians in power have been in power for one and a half years.

(12d)    The fact in (12b) is a cause of the fact in (12a).

 

The author does not level an explicit accusation at the politicians in power, but takes the opportunity to presuppose that they are intoxicated by power by embedding his explanation of their behaviour as a complement clause of the factive verb sapere 'know'. (Of course, (12b-12d) count as presupposed if the causal subordinate clause perché... 'because...' is considered as subordinated to non lo ascolterete e non ascolterete neppure me 'you won't listen to him or to me either' and not directly to so 'I know'. But this is the most natural reading: the causal clause seems to explain why the politicians will not listen to the two journalists, not why it is that the author knows that they will not listen.) The author clearly feels hostility for the government coalition. It is likely but not at all certain that his readers already share his feelings. But, in a way, those readers who happen not to share them are the main target of the author's persuasive aims. Moreover, even in the case of readers who do share them, their hostile feelings for the government are reinforced by being provided with new input.

Finally, the persuasive use of presuppositions is manifest in political discourse, when a politician addresses the citizens directly in order to gain their approval, and can often be found in the quotations from such discourses reported by the daily press. There are clear examples in some statements attributed to Umberto Bossi (leader of the Northern League):

 

(13)        "Sono momenti in cui č in atto una riorganizzazione, una contrapposizione tra chi vuole mantenere Roma come baricentro dello sfruttamento della Padania e i patrioti padani."

'These are times when there is a reorganization going on, a contraposition between those who want to keep Rome as the centre of gravity for the exploitation of  Padania and Padania's patriots'

                        (Il Corriere della Sera 4/9/1997, p.11)

 

Some definite descriptions (as well as the relative clause chi... 'those who...'), the proper name Padania4, and the use of the verb mantenere 'keep' trigger the following presuppositions:

 

(13a)      Some people want to keep Rome as the centre of gravity for the exploitation of Padania.

(13b)    Rome is the centre of gravity for the exploitation of Padania.

(13c)    Padania is being exploited.

(13d)    Padania exists.

(13e)    Padania's patriots exist.

(13f)     Padania is a fatherland.

 

Belief in these assumptions (first of all, belief in the existence of Padania, presently not even a clearly defined region of Italy, and in its being a fatherland which has its patriots) is precisely what Bossi needs to elicit in order to gain wider approval for his policies.

Of course, reported statements are usually formulated so as to present the quoted politician's political line in such a way as to be functional to the political line of the newspaper. For example, Il Corriere della Sera seems to believe that Bossi is dangerous and that some of his attitudes are illegal and/or irresponsible, and the above quotation may be used to support this evaluation.

Both presuppositions and, as we shall see, the omission of presuppositions are prominent moves in the persuasive strategies of the press. I shall close this overview on informative and persuasive presuppositions in the Italian press with the following example of a declaration attributed to Premier Prodi:

 

(14)        "L'unitā morale e civile della nazione č radicata e resa salda anche nella vita religiosa e nell'appartenenza cattolica."

'The moral and civil unity of the nation is also rooted in, and held fast by, religious life and belonging to the Catholic faith'

(La Stampa 9/9/1997, p.1, "Prodi: la fede fa l'Italia unita" 'Prodi: the faith makes Italy united')

 

In this declaration, the definite description l'unitā morale e civile della nazione 'the moral and civil unity of the nation' and the adverb anche 'also' induce the following presuppositions:

 

(14a)    There is a moral and civil unity of the nation.

(14b)      The moral and civil unity of the nation is rooted in and held fast by something else besides religious life and  belonging to the Catholic faith.

 

In the context of the debates taking place at the time about the secessionist aims of the Northern League, presupposition (14a) - although not exactly "new information"! - has clearly persuasive functions: (14) takes one of the theses which is being debated for granted, by presupposing it. Presupposition (14b) should be shared knowledge for all Italians who have studied some history of their country; however, opinions differ about what exactly can be said to make for the unity of the Italian nation; so that (14b) is at least in part informative. Moreover, it is relevant that it is presupposed by the premier, Prodi, a Catholic himself, but leading a left-wing coalition. Thus one informative contribution conveyed by this presupposition is that Prodi believes that there are forces other than Catholicism which hold fast the unity of the nation. That is, by presupposing (14b) Prodi has implicitly recognized the role of lay society with respect to the unity of the nation. But it should also be noted that the speech concerned was made in a Catholic context, where its presupposition (14b) was likely to have the persuasive effect of getting Catholics to recognize - and moreover, recognize as obvious - that there are lay moral values in Italian society.

The newspaper omits anche 'also' from the headline ("Prodi: la fede fa l'Italia unita" 'Faith makes Italy united') and by so doing modifies the overall meaning conveyed by the statement quoted. This enables a political commentator to write on the same page:

 

(15)        Siamo uniti, ha detto sostanzialmente Romano Prodi a Loreto, perché siamo cattolici.

'Romano Prodi basically said in Loreto that we are united because we are Catholics'

(La Stampa 9/9/1997, p.1: Sergio Romano, "Se l'Ulivo diventa guelfo" 'If the Ulivo becomes a clerical party')

 

where the focusing adverb sostanzialmente 'basically' is hardly sufficient to make up for the omission of anche 'also' (with its associated presupposition (14b)) and where the elimination of this presupposition paves the way for an attack on the Ulivo coalition, at risk of being excessively conditioned by Catholicism.

 

3. Some doubts about the received account of pragmatic presupposition

 

Do the received notions of pragmatic presupposition and of presupposition accommodation allow a satisfactory account of the uses of presupposition illustrated above?

Firstly, I would argue that presupposition proper and presupposition accommodation should not be given two separate accounts, the one correcting and integrating the other. There is some continuity between the two phenomena, and when we are faced with actual samples of discourse containing presupposition inducers, it is hard to tell whether the triggered presuppositions should be counted  as satisfied (that is, as part of a set of shared assumptions) or not (and therefore, as informative presuppositions to be accommodated). In fact, the task of a receiver when faced with a text which induces a presupposition does not differ from one situation to the other: he or she is required to identify the presupposition and take it for granted, whether he or she was already acquainted with it or not.

As regards the supposedly shared status of presuppositions, we should distinguish at least between (a) cases in which the hearer holds the presupposed assumption and the speaker knows this to be so; (b) cases in which the hearer does not hold the presupposed assumption, but the speaker thinks he or she does; (c) cases in which the speaker does not know whether the hearer holds the presupposed assumption or not; (d) cases in which the speaker thinks that the hearer does not hold the presupposed assumption, while in fact he or she does; (e) cases in which the speaker thinks that the hearer does not hold the presupposed assumption, and he or she in fact does not. In the cases in (d) and (e) the speaker intends to make an informative use of the presupposition, but only in case (e) does he or she succeed. In the cases in (b), the presupposition turns out to be informative beyond the speaker's intentions. In the cases in (c), it is hard to tell whether the use the speaker makes of the presupposition should be counted as informative or not, since it may be either. This case is undoubtedly very frequent among our examples from the press: a journalist can never be sure which of his or her assumptions are already shared by the reader, and in addition has often to cope with an audience which is likely to be split, partly sharing and partly not sharing his or her assumptions. Therefore, it is apparent that we should not give separate accounts of presupposition proper and of informative or persuasive presupposition. A good account of presupposition should be extensible without modification to informative presupposition, as well as to the intermediate cases.

Secondly, I would like to recall some of the arguments against the received notions of pragmatic presupposition and of presupposition accommodation put forward by Gauker (1997, 1998). Gauker is concerned with a problem different from the one discussed here, namely with defining the "context" of the utterance, that is, that set of propositional elements on which the evaluation of the utterance of a sentence depends. He concludes that the context cannot be defined in terms of shared assumptions and that one should adopt two distinct concepts: that of objective context and that of the participants' take on the context. It is the objective context which must be taken into account in order to evaluate the appropriateness of utterances, while it is the participants' take on the context which is modified by what participants tell each other. On his way to establishing his thesis, he discusses informative presupposition and claims that no satisfactory account of it has been given so far (Gauker 1997:204-05, 1998). The "accommodation rule" proposed by Lewis (1979), according to which a presupposition comes into existence whenever that is required for the acceptability of what is said, is just the formulation of the problem, not a solution of it: one would like to understand how it happens in this way.  Stalnaker (1973, 1974) has claimed  that in informative presuppositions the speaker is pretending to presuppose something. Against him, Gauker objects that the point of an utterance like the well-known example by Karttunen (1974)

 

(16)        We regret that children cannot accompany their parents to the commencement exercises.

 

is just to inform that children cannot accompany their parents; if the speaker were acting as if everyone already knew that, he or she would not utter the sentence at all. Soames (1982), in his definition of utterance presupposition, has used the notion of "taking something as uncontroversial" and, in order to cope with informative presupposition, has analyzed this notion as equivalent either to thinking that the presupposition is a shared assumption, or to thinking that the audience is prepared to add the presupposition without objection to the context against which the utterance is evaluated. To him, Gauker objects that this would make informative presupposition indistinguishable from unobjectionable information, while, in fact, an utterance conveying an informative presupposition such as (16) may well raise objections;  it is therefore at least unclear in which other sense the audience should be prepared to add the presupposition to the context without objection.

By now, we may conclude that the received notions of pragmatic presupposition and of presupposition accommodation5 do not  account for the phenomenon of informative presupposition satisfactorily, and therefore cannot account for the persuasive uses of presupposition either.

 

 

4. Towards an account of the persuasive use of presupposition

 

I would now like to make a fresh start by focusing on the persuasive use of presupposition. The existence of such a use raises some questions which may emerge as being central to an account of presupposition: why should presuppositions ever be persuasive? Why should the audience not only identify the presuppositions triggered by an utterance, but also agree to take them for granted? Why is there a default tendency to do so (unless there are special reasons for rejecting the presuppositions)? I think that one answer to these questions may be couched in the following terms.

I propose considering presuppositions not as shared assumptions, but as assumptions which ought to be shared. Thus we allow them a normative feature. Such an admission is consistent with early insights about presupposition such as Strawson's (1950). There, presupposition (particularly, the existential presupposition associated with a definite description used as the subject of an assertion) is required in order for the assertion which presupposes it to have a truth value. One might say that when the context fails to satisfy the presupposition, the assertion violates one of its most fundamental norms - the one according to which it must have a truth value. Actually, Austin  seems to have read Strawson in this way, since he includes presupposition among the felicity conditions of assertive illocutionary acts (1962: 137). Ducrot too (1972) highlighted the normative or "deontic" factors in presupposition, although drawing the disputable conclusion (criticized in Sbisā 1979), that presupposition itself is an illocutionary act. In the recent contributions by Gauker mentioned above, normative factors emerge again. According to Gauker (1998), conversations are governed by objective contexts (that is, by sets of propositional elements, factually constrained and connected with the goals of the ongoing interaction, against which utterances are evaluated), since certain norms of discourse that interlocutors ought to adhere to are formulated in terms of these contexts. It seems to follow from this that an utterance should be regarded as in some way inappropriate or out of order if the speaker, in uttering it, does not take the objective context into due consideration.

In this vein, I would like to claim that whenever a hearer takes it that an utterance by a certain speaker presupposes something (because it contains a certain presupposition inducer), if the hearer takes the objective context not to contain the presupposed propositional element, he or she will be bound to consider the speaker not only as being wrong about the facts (as occurs when somebody says something factually false), but also as violating some norm of discourse. Violating norms of discourse may in turn be deemed a kind of uncooperative behaviour: it is in fact a kind of behaviour which makes it difficult to continue conversational cooperation. How can we reply to a conversational contribution whose presuppositions are not satisfied by the context? To use Strawson's words (1950:12), the question doesn't arise; we may find it difficult to continue the exchange at all. One solution which might preserve the communicative relationship is to initiate a discussion on those features of the objective context about which the participants disagree. This solution is laborious, because it involves a change of topic from what was explicitly at issue to what was merely presupposed, as well as being risky, because it amounts to openly challenging the entitlement of the speaker to issue the utterances he or she has issued, which may once again lead to a breakdown in the communicative relationship (why should anyone talk to someone who says things he or she is not entitled to say?). Since this breakdown in the communicative relationship may even damage the participant who has not violated any norm by leaving him or her deprived of a source of information (though dubious at times), and of a certain amount of cooperative resources (though not wholly reliable), it is obvious that, whenever possible, the hearer will avoid treating the speaker as someone violating norms of discourse. In all the cases in which there is some possibility left that the objective context does contain the presupposition, the default tendency6 will be to assume that it is so, namely, that the presupposition is in fact satisfied.

From all of this, we may conclude that the persuasive use of presupposition depends on certain normative or deontic features that presuppositional phenomena have. The utterance of a given sentence is really appropriate only if the objective context really has a certain content. It is among the speaker's responsibilities to make it so, that his or her utterance is appropriate. It is among the speaker's responsibilities to issue an utterance only if it is appropriate when the objective context is as it happens to be. Moreover, it is among the speaker's responsibilities to issue an utterance containing certain presupposition inducers only if the objective context really contains the presupposition they trigger. Thus, we are describing presuppositions as assumptions that the speaker ought to make, or, however, assumptions for which he or she is responsible. But in order to give an appropriate evaluation of the speaker's utterance, the hearer too ought to make the same assumptions, not because the speaker makes them, but because they are the assumptions which the speaker ought to make. Thus, we are describing presuppositions as assumptions which ought to be shared.

Although this description could be extended to those presuppositions of the speaker which are not linguistically marked, but can be inferred by the hearer from the objective context  on the basis of the assumption that the speaker has (as he ought to) a correct take on the objective context (Gauker 1998), in our present investigation we are applying it primarily to linguistically marked presupposition: those presuppositions of an utterance which are triggered by linguistic presupposition inducers. When faced with an utterance containing a presupposition inducer, and knowing that the speaker ought to issue it only if the context contains the triggered presupposition (provided he or she wants to keep on communicating with the speaker, and unless there are specific reasons not to do so), the hearer will hold that the presupposition of the speaker's utterance is contained in the objective context. This can be taken as a generally valid description of the hearer's position with regard to the (linguistically marked) presuppositions of the speaker's utterance. Thus, the phenomenon giving rise to the persuasive use of presupposition turns out to be quite general. We have perhaps reached a description of the dynamics of presuppositional phenomena which holds good irrespective of whether the presupposed propositional elements are old or new information for the hearer, and more generally, irrespective of whether they initially belong to the speaker's take on the context, to the hearer's, to both, or to neither.

 

 

5. Presupposition and ideology

 

The view of presupposition which emerges from the preceding discussion is that of a communicative device for constructing the participants' takes on the context, the functioning of which is guided by the underlying normative character of the objective context.

Presupposition, so intended, is clearly suitable for transmitting a kind of contents which may be called ideological: assumptions, not necessarily conscious but liable to be brought to consciousness, about how our human world is and how it should be. For various reasons, such assumptions are often difficult to be certain about, even though they are not in general unverifiable. Since assertion and argumentation commit the speaker to giving evidence or reasons for what is asserted or argued for, it may be difficult, or uncomfortable, to assert or discuss this kind of assumptions explicitly. However, they manage to play a prominent role in political or ethical choices or in the practical adhesion to ways of life, so that speakers may find themselves needing both to convey them and to win their audience's approval of them. From a consideration of some examples of the persuasive use of presuppositions, we can see that the presuppositions they convey display features of this kind.

The reader's attitude towards the Northern League and specifically his or her emotional attitude towards the likelihood of secession will change considerably depending on whether he or she assumes:

 

(7a)      There is or will be a vote in the Po Valley.

 

(presupposed by (7), il voto padano 'the Po Valley vote') or:

 

(8a)      There will be an election-opinion poll in October.

 

(presupposed by (8), le elezioni-sondaggio di ottobre 'the election-opinion poll of October'). It has to be noted that the contents of (7a) and (8a) are fully manifest in (7) and (8), so that their identification does not demand much processing. But the status of (7a) and (8a) is not that of explicit assertions; the speaker is required to take them for granted, unless he or she wants to challenge the author's reliability and thus the communicative relationship.

The evaluation of the behaviour of prof. Di Bella and of the Ministry of Health, and, in particular cases, even the reader's personal behavioural choices with respect to cancer, can certainly be influenced by

 

(17a)    There is a Di Bella anti-cancer treatment.

 

which is presupposed by  the headline

 

(17)      Il pretore ordina: usate l'anticancro Di Bella

            'Order from the judge: use the Di Bella anti-cancer treatment'

 

and by the other presuppositions (9a-b) triggered by example (9). It is not clear at all whether one and only one thing which can be described as the Di Bella anti-cancer treatment actually exists and whether Di Bella's prescriptions to patients suffering from cancer do anything at all to cure cancer. But presupposing (17a) cuts short of these doubts.

It is interesting to note that opposite evaluations of the policy of the Ministry of Health in the Di Bella controversy are often conveyed by constructions presupposing opposing conceptions of the rights of the patient and of the obligations of the State or of the doctors, which hardly ever come to an open confrontation. I would like here to add the following quotation to our sample taken from an article commenting upon the decision of an administrative court in favour of the free distribution to terminally ill patients of one of the drugs used by Di Bella.

 

(18)        Eppure, mi pare una cosa giusta. Perché tien conto che non si tratta di una sperimentazione che distribuisce incertezza, ma che distribuisce speranza.

'Still, I think it's the right thing. Because it takes into account the fact that this is not one of those experiments that distributes uncertainty, but one that distributes hope.'

(Il Piccolo, 12/3/1998: Fernando Camon, "Cittadini uguali solo davanti al Fisco" 'Citizens equal only before the tax-man')

 

The distribution of hope is here presupposed to be brought about by the experimental adoption of the Di Bella treatment (note the factive tiene conto che 'takes into accout the fact that') and taken as a criterion of positive evaluation.

Finally, consider presupposition

 

(13b)    Rome is the centre of gravity for the exploitation of Padania

 

triggered by mantenere 'keep' in the declaration by Bossi:

 

(13)        "Sono momenti in cui č in atto una riorganizzazione, una contrapposizione tra  chi vuole mantenere Roma come baricentro dello sfruttamento della Padania e i patrioti padani."

'These are times when there is a reorganization going on, a contraposition between those who want to keep Rome as the centre of gravity for the exploitation of  Padania and Padania's patriots.' 

 

Here, the content of (13b) is not fully manifest in (13). Speaking of someone who wants something to stay in a certain state, one is not spelling out that it has been in that state up to now, but one is nonetheless taking for granted and inviting the reader to take for granted that it has been in that state, and from linguistic material in (13) it is certainly not difficult to gather which. Moreover, the presupposition quite obviously suggests or fosters resentment against Rome. As to the other presuppositions of (13), (13 a-f) quoted above, it is clear that a follower of Bossi has to find a place in his or her universe for entities such as Padania (a fatherland, which has its patriots) and Padania's patriots, and that such a universe is different from the one, say, evoked by Prodi in example (14), where what is presupposed to exist is the moral and civil unit of the Italian nation (and therefore, at the same time, the Italian nation as well). Thus the presuppositions of (13) and those of (14) require the readers to have very different takes on the context. I shall not discuss under which conditions propositional elements such as (13a-f) or (14a-b) could be said to belong to an objective context. What is of interest here is that in the communicative situation in which (13) and (14) are produced, there is uncertainty and disagreement about the existence of things such as Padania, Padania's patriots, the Italian nation and its moral and civil unity and both (13) and (14) try to get beyond this state by their presuppositions.

What is implicitly conveyed in these examples has to do with ideology in one of the many senses which the word "ideology" may have. In the framework of the historical-terminological analysis proposed by Rossi-Landi (1978: 15-34), the sense in which we are using the word "ideology" here - assumptions  about how our human world is and should be, not necessarily conscious but liable to be brought to consciousness - can be collocated between ideology as common sense and ideology as false conscience (whether in the form of non-deliberate lie or in the form of conscious deception). On the one hand, common sense assumptions are regularly taken for granted by speakers and may act as criteria for value judgements or choices, but they do not need to be encoded by linguistic presupposition inducers. When this does occur, the discourse which triggers the presuppositions turns out to have hidden didactic aims. For example, in romantic fiction commonplace assumptions about male and female, the rich and the poor, love, sex, jealousy, marriage and the like have often become the content of a well-designed implicit sentimental education  (Pozzato 1982, Sbisā 1986). On the other hand, what marxism has called "false conscience" is characterized by being just false, although not always consciously recognized as such, and by concealing something non-ideological, thus serving the aims of some socio-economic or power group. Although its contents too can be conveyed by way of presuppositions, they are often explicitly theorized and even argued for.

The assumptions conveyed by our persuasive presuppositions are half-way between these two poles. They are not so clearly commonplace as not to be in need of some linguistic manifestation. In some cases (e.g. with iteratives or verbs of state change) the triggered presupposition can be reconstructed from the linguistic material present in the uttered sentence and the  presupposition inducer itself. In other cases, everything that has to be presupposed is in fact linguistically manifest, as in definite descriptions (il voto padano 'the Po Valley vote') or nominalized clauses introduced by the definite article (lo sfruttamento della Padania 'the exploitation of Padania') and in complement clauses of factive verbs (which formulate the presupposition which has to be triggered by the factive). But presuppositions are not fully explicit either: they do not present themselves as explicit claims, as assertions, evaluations or arguments. These would be verdictive illocutionary acts, therefore liable to be submitted to a truth/falsity, correctness/incorrectness judgement. Presuppositions avoid exposure to such a challenge by requiring the audience to take them for granted in order to keep on communicating. Thus they may act as a powerful tool for the organization of approval in the mass media: they combine common sense, elements of "false conscience" and other heterogeneous socio-cultural or evaluative assumptions into a mixture which, conveyed with the complicity of norms of discourse, seems to evade critical analysis.

However, it should be remembered at this point that presuppositions can be made explicit. All linguistic work on presupposition inducers has in fact used explicitation procedures. The classical discussions about presupposition projection, informative presuppositions, presupposition accommodation, and perhaps even more the recent works which parallel anaphor and presupposition (van der Sandt 1989, 1992; Beaver 1997), all remind us - quite beyond their differences and shortcomings - that presupposition explicitation is a possible procedure. Moreover, research on presupposition inducers and presupposition projection enables us not only to tell what presuppositions an utterance conveys (this could be done even in absence of such technical tools!), but also to do so on the basis of some specifiable reason or argument.

Explicitation of presuppositions and challenge to them are available moves in case of disagreement or misunderstanding, although usually dispreferred ones. If there is an interest, for example, in achieving conscious control over implicitly conveyed assumptions, explicitation may become a deliberate activity and the listener or reader practising it can consciously profit of the received knowledge about linguistic presupposition inducers. Exercises in presupposition explicitation with adults and even with children (Sbisā 1996) have convinced me that some acquaintance with the kinds of sentence components and constructions which are recognized as triggering presuppositions can be of help in overcoming resistance to focusing explicitly on a presupposition. Once it is made explicit, of course, a presupposition becomes just an assertion and, like all assertions, whilst the speaker is committed to defending it, the listener or reader is allowed to challenge it and even to ask the speaker for his or her grounds  for making that assertion. Thus by facilitating both presupposition explicitation and the monitoring of it, research on presuppositions can help improve reading and comprehension skills in general, and in particular promote and foster criticism of implicitly transmitted ideological contents.

 

 

NOTES

1 I would like to thank Paola Rodari for the many discussions we have had about the use of presuppositions in written texts since we wrote Sbisā and Rodari (1989) and Christopher Gauker for his insightful e-mail correspondence regarding context and assertibility.

2 It should be noted that the investigation of presuppositional phenomena had its starting point in reflections on some of these linguistic devices (Strawson 1950, Kiparsky and Kiparsky 1971, Fillmore 1971), thus highlighting, initially at least, certain semantic aspects of presuppositional phenomena rather than pragmatic ones.

3 The examples quoted in this paper are taken from La Stampa, Il Corriere della Sera and Il Giornale. The issues from 1-15 September, 1997 and from 15-31 December, 1997 have been consulted thoroughly. Material from other issues or from other daily newspapers (e.g. Il Piccolo) was also used, but apart from one or two exceptions, has not been quoted.

4 Padania is a proper name (geographical areas take the definite article in Italian). Therefore, its function is to point to an object which is taken to exist.

5 It should be noted that the notion of presupposition accommodation as used in Heim (1988) and even more in the anaphoric theory of presupposition (van der Sandt 1992; Beaver 1997) shifts from Lewis's original pragmatic notion to being a device for making sure that the presuppositions triggered by an uttered sentences are located in their due place in the representation of discourse. Thus, accommodation becomes a general tool for the logico-linguistic description of discourses containing presupposition inducers. There are no longer two separate explanations for presupposition proper and informative presupposition. But the pragmatic aspects of presupposition accommodation fade away, and the question of why discourse should function like this is not raised at all.

6 Default rules or defeasible reasoning strategies have often been included in accounts of presupposition. A default tendency to accommodation has been noted by Lewis (1979); in the context of the "projection problem", that is, the problem of calculating the presuppositions of compound sentences, Gazdar (1979)  has held that presuppositions are projected by default, unless there are reasons for cancelling them, while Soames (1982) has spoken of  "defeasible presumptions" which are created by speakers who utter compound sentences one of whose constituents contains a presupposition inducer; van der Sandt (1988, 1989) has dealt with presuppositions as default inferences; in more recent logico-linguistic theories of presupposition (van der Sandt 1992, Beaver 1997; but see already Heim 1988) there is the default preference for the so-called "global accommodation" of the presupposition, which collocates the presupposition in the global context of the discourse, over its "local accommodation", which guarantees the acceptability of the utterance without requiring the global context of the discourse to contain the presupposed propositional element. Gauker (1998) has claimed that there is a default tendency to assume that other participants presuppose whatever really does belong to the objective context; his claim is different from the one put forward here (which is, rather, that there is a default tendency to assume that what is presupposed by  the utterances of other participants really belongs to the objective context), but in the last resort the two claims are compatible, since both are rooted in the normative character of the relationship between utterance and context.

 

 

REFERENCES

Austin, John L. (1962) How to do things with words.  London: Oxford University Press. 2nd edition (1975).

Beaver, David I. (1997) Presupposition. In J. Van Benthem & A. ter Meulen (eds.), Handbook of logic and language. Amsterdam: Elsevier.

Davis, Steven (ed.) (1991) Pragmatics. A reader. Oxford: Oxford University Press.

Ducrot, Oswald (1972) Dire et ne pas dire. Paris: Hermann.

Fillmore, Charles J. (1971) Verbs of judging: an exercise in semantic description. In C.J. Fillmore & D.T. Langendoen (eds.), Studies in linguistic semantics. New York: Holt, Rinehart and Winston, 273-90.

Gauker, Christopher (1997) Universal instantiation: a study of the role of the context in logic. Erkenntnis 46: 185-214.

Gauker, Christopher (1998) What is a context of utterance? Philosophical Studies 91:  149-172.

Gazdar, Gerald (1979) Pragmatics. Implicature, presupposition and logical form. New York: Academic Press.

Heim, Irene (1988) On the projection problem for presuppositions. In D. Flickinger & al. (eds.), Proceedings of the Second West Coast Conference on Formal Linguistics. Stanford, Ca.: Stanford University Press, 114-25. Reprinted in S. Davis (ed.) (1991), 397-405.

Karttunnen, Lauri (1974) Presupposition and linguistic context. Theoretical Linguistics 1:181-194. Reprinted in S. Davis (ed.) (1991), 406-415.

Kiparsky, P., Kiparsky, C. (1971) Fact. In D. Steinberg & L. Jakobovits (eds.), Semantics. Cambridge: Cambridge University Press, 345-369.

Levinson, Stephen C. (1983) Pragmatics. Cambridge: Cambridge University Press.

Lewis, David (1979) Scorekeeping in a language game. Journal of Philosophical Logic 8: 339-359. Reprinted in S. Davis (ed.) (1991), 416-427.

Pozzato, Maria Pia (1982) Il romanzo rosa. Milano: Espresso Strumenti.

Rossi-Landi, Ferruccio (1978) Ideologia. Milano: ISEDI.

Sbisā, Marina (1979) Perlocuzione e presupposizioni. In F. Albano Leoni & M.R. Pigliasco (eds.), Retorica e scienze del linguaggio. Atti del X Congresso Internazionale di Studi della Societā di Linguistica Italiana. Roma: Bulzoni, 37-60.

Sbisā, Marina (1984) On illocutionary types. Journal of Pragmatics 8: 93-112.

Sbisā, Marina (1986) Atti linguistici in un romanzo popolare. In A. Arslan Veronese (ed.), Dame, droga e galline, Milano: Unicopli, 291-305. 

Sbisā, Marina (1996) La pragmatica nell'educazione linguistica e nella didattica. Parafrasi e acquisizione di informazioni. Paper read at the Seminario internazionale "Insegnare italiano nella scuola del 2000" (Trieste 7-9 novembre 1996).

Sbisā, M., Rodari, P. (1989) Il ruolo dell'implicito nella comprensione del testo. In A.M. Mitri & A. De Rossi (eds.), I luoghi della parola. L'educazione linguistica di base. Firenze: La Nuova Italia, 277-290.

Soames, Scott (1982) How presuppositions are inherited. A solution to the projection problem. Linguistic Inquiry 13: 483-545. Reprinted in S. Davis (ed.) (1991), 428-70.

Soames, Scott (1989) Presupposition. In D. Gabbay & F. Guenthner (eds.), Handbook of philosophical logic. Volume IV. Dordrecht: Reidel, 553-616.

Stalnaker, Robert (1973) Presuppositions. Journal of Philosophical Logic 2: 447-457.

Stalnaker, Robert (1974) Pragmatic presupposition. In M. Munitz & P. Unger (eds.), Semantics and philosophy. New York: New York University Press, 197-214. Reprinted in S. Davis (ed.) (1991), 471-81.

Strawson, Peter F. (1950) On referring. Mind 59: 320-44. Reprinted in P.F. Strawson (1971), 1-27.

Strawson, Peter F. (1971) Logico-linguistic papers. London: Methuen.

van der Sandt, Rob A. (1988) Context and presupposition. London: Croom Helm.

van der Sandt, Rob A. (1989) Presupposition and discourse structure. In R. Bartsch, J. van Benthem & P. van Emde Boas (eds.), Semantics and contextual expressions. Dordrecht: Foris, 267-94.

van der Sandt, Rob A. (1992) Presupposition projection as anaphora resolution. Journal of Semantics 9: 333-377.