Analisi Linguistica e Letteraria 1/2008 Part 2 by EDUCatt

CoverALL200801Mount.qxp:Layout 1

24-07-2009

14:03

Pagina 1

ISSN 1122 - 1917

L’ANALISI

LINGUISTICA

LETTERARIA

2008

L’ANALISI LINGUISTICA E LETTERARIA FACOLTÀ DI LINGUE E LETTERATURE STRANIERE UNIVERSITÀ CATTOLICA DEL SACRO CUORE

ANNO XVI

FACOLTÀ DI LINGUE E LETTERATURE STRANIERE

L’ANALISI LINGUISTICA E LETTERARIA

2008

VOLUME 1

ANNO XVI - 1/2008

EDUCatt - Ente per il Diritto allo Studio Universitario dell’Università Cattolica Largo Gemelli 1, 20123 Milano - tel. 02.72342235 - fax 02.80.53.215 e-mail: editoriale.dsu@unicatt.it (produzione) librario.dsu@unicatt.it (distribuzione) redazione.all@unicatt.it (Redazione della Rivista) web: www.unicatt.it/librario

ISSN 1122 - 1917

EDUCATT - UNIVERSITÀ CATTOLICA DEL SACRO CUORE

VOLUME 1

L’ANALISI LINGUISTICA E LETTERARIA Facoltà di Scienze linguistiche e Letterature straniere Università Cattolica del Sacro Cuore Anno XVI - 1/2008 ISSN 1122-1917 Direzione GIUSEPPE BERNARDELLI LUISA CAMAIORA SERGIO CIGADA GIOVANNI GOBBER Comitato scientifico GIUSEPPE BERNARDELLI - LUISA CAMAIORA - BONA CAMBIAGHI - ARTURO CATTANEO SERGIO CIGADA - MARIA FRANCA FROLA - ENRICA GALAZZI - GIOVANNI GOBBER DANTE LIANO - MARGHERITA ULRYCH - MARISA VERNA - SERENA VITALE - MARIA TERESA ZANOLA Segreteria di redazione LAURA BALBIANI - GIULIANA BENDELLI - ANNA BONOLA - GUIDO MILANESE MARIACRISTINA PEDRAZZINI - VITTORIA PRENCIPE - MARISA VERNA

Pubblicazione realizzata con il contributo PRIN - anno 2006

© 2009 EDUCatt - Ente per il Diritto allo Studio Universitario dell’Università Cattolica Largo Gemelli 1, 20123 Milano - tel. 02.72342235 - fax 02.80.53.215 e-mail: editoriale.dsu@unicatt.it (produzione); librario.dsu@unicatt.it (distribuzione); web: www.unicatt.it/librario Redazione della Rivista: redazione.all@unicatt.it - web: www.unicatt.it/librario/all Questo volume è stato stampato nel mese di luglio 2009 presso la Litografia Solari - Peschiera Borromeo (Milano)

VOLUME I CONTENTS

Foreword

VII

Editors’ introduction

LINGUISTIC STRUCTURES AND PRAGMATIC FUNCTIONS IN ARGUMENTATION Identifying indicators of argumentative moves

FRANS H. VAN EEMEREN & PETER HOUTLOSSER & FRANCISCA SNOECK HENKEMANS

Dialogue–specific illocution indicators

FRANZ HUNDSNURSCHER

Presumption, burden of proof and lack of evidence

DOUGLAS WALTON

The argumentative power of words or how to move people’s minds with words

EDDA WEIGAND

CONSIDERING ARGUMENTATIVE DISCOURSE IN GERMAN

Argomentazione filosofica e traduzione: confronto e contrasto di idee e parole LAURA BALBIANI

Nominalprädikate und argumentative Funktionen: die Erklärung

107

SIBILLA CANTARINI

Zur deutschen Sprache der Kontroverse in der frühen Neuzeit

123

GERD FRITZ

Deutsche Entsprechungen zu magari? Eine Analyse anhand der multimedialen 139 Datenbank Forlixt 1 und anderer Corpora der gesprochenen Sprache CHRISTINE HEISS

Affektbeladene Ausdrucksformen in den Sitzungsdebatten des Europäischen Parlaments

151

LORENZA REGA & MARELLA MAGRIS

Marker impliziter Bedeutung in Werbetexten: eine kontrastive Analyse

163

FEDERICA RICCI GAROTTI

Sinngehalt im argumentativen Dialog: „Das ewige Leben“ im Evangelium nach Lukas (Lk.10, 25-37)

177

LUCIA SALVATO

Deutsche Entsprechungen zu mica: eine Analyse anhand Forlixt 1 und anderer Corpora der gesprochenen Sprache

191

MARCELLO SOFFRITTI

LINGUISTIC DEVICES IN ARGUMENTATIVE DIALOGUES

Words in context. Agreeing and disagreeing with allora JOHANNA MIECZNIKOWSKI & BARBARA GILI & CARLA BAZZANELLA

205

Russian “chasticy” argue in dialogues

219

ANNA BONOLA

Alcune considerazioni pragmatiche su “dovere” in italiano: usi deontici ed anankastici

229

MARCO CARMELLO

L’exagération dans le roumain quotidien. Bref regard sur les interactions verbales

237

MIHAELA VIORICA CONSTANTINESCU

Expressions used by young argentine spanish speakers: neutralization of discrepancies between interlocutors

253

MARINA GRASSO

Le rôle de la référence personnelle dans l’interaction: le cas du roumain

267

LILIANA HOINARESCU

Il dialogismo nelle postille dell’Ascoli

281

CELESTINA MILANI

“Today, we’re all Danes”. Argumentative meaning of the 1st and 2nd person pronouns in newspaper editorials on the Muhammad cartoons

289

MARTINA TEMMERMAN

Dialogical occurrences of bem in European Portuguese. Enunciative stability and deformability

305

HELENA VIRGÍNIA TOPA VALENTIM

Sens lexical et processus argumentatifs: le cas du connecteur français c’est-à-dire (que) HÉLÈNE VASSILIADOU

317

CROSS-LINGUISTIC PERSPECTIVES

Rinforzo argomentativo e attenuazione argomentativa in francese, italiano, 331 russo, spagnolo e tedesco: una proposta di analisi condotta sulla base di forum di discussione on-line VAHRAM ATAYAN

Introductory ‘it’ patterns in English and Italian academic writing: A cross-generic and cross-cultural analysis

343

GIULIANA DIANI

The verbal means in German and Japanese argumentation

357

MARION GREIN

A contrastive study of the introduction section genre of English and Persian 373 medical research articles AVISHAN MAHZARI

The role of nonverbal expressions as precursors to argumentative discourse in French, Japanese and American English conversation

385

CAROLINE ELISA NASH

THEORETICAL ISSUES IN ARGUMENTATION

Jeux et enjeux du discours rapporté pour le discours argumentatif

403

PRISQUE BARBIER

Approaching argumentative dialogue with formal models GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

415

Dire senza dire, argomentare senza argomenti

427

ADELINO CATTANI

Discursive perspective and argumentation in the Romanian parliamentary discourse. A case study

435

LILIANA IONESCU-RUXĂNDOIU

The dialectical-dialogical definition

443

FABRIZIO MACAGNO

The conjunction ‘and’: literal meaning and discourse values

463

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

L’engagement comme notion cognitive associée au destinataire

475

LOUIS DE SAUSSURE & STEVE OSWALD

Rassegna di linguistica generale

489

A CURA DI MARIO BAGGIO E MARIA CRISTINA GATTI

Rassegna di glottodidattica

497

A CURA DI BONA CAMBIAGHI

Rassegna di linguistica francese

501

A CURA DI ENRICA GALAZZI E CHIARA MOLINARI

Rassegna di linguistica inglese A CURA DI MARGHERITA ULRYCH

507

Rassegna di linguistica russa

513

A CURA DI ANNA BONOLA

Rassegna di linguistica tedesca A CURA DI GIOVANNI GOBBER E FEDERICA MISSAGLIA

517

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 253-266 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS: NEUTRALIZATION OF DISCREPANCIES BETWEEN INTERLOCUTORS MARINA GRASSO

1. Introduction We have grouped in this paper three expressions that, in our opinion, share important characteristics and are frequently used in the genre of informal conversations: the expressions obvio, más vale and ni hablar – that can be roughly translated as obvious, of course and needless to say. The basis for this analysis is a selection of argumentative fragments that form part of the corpus of the research project on “Genre in verbal interactions”, conducted at the University of La Plata, Argentina. This corpus is formed by twenty four informal conversations among university students between the ages of 18 and 26. We focused on the linguistic, discoursal and interactive behaviour of the expressions mentioned. A qualitative analysis, starting from the theoretical perspective of Systemic Functional Linguistics, was carried out, taking van Eemeren et al. (2000) as the reference for the analysis of the argumentative fragments. Work was done on the location of these expressions with respect to the syntactic organization of the utterances where they occur, and particularly on their function in the conversational discourse. The comparison and contrast of the expressions mentioned is the topic of this paper. 2. Analysis of the expressions Similarities

Table I Linguistic behaviour of the analysed expressions

MARINA GRASSO

254

As the table shows, the expressions under analysis share the same linguistic behaviour: the three of them can be found as part of a whole proposition, in isolation as the only component of the construction, and thematized with the conjunction ‘that’. They also share similarities as regards their interactive behaviour, since all of them are phoric constructions that “can only make sense in relation to a previous discourse” and “hence, can never appear inside the first turn of speech nor be the first part of an exchange, since they presuppose an anaphoric retroactive move” (Montolío 1996: 332-3). The three expressions in question could be included inside the class of markers that Martín Zorraquino and Portolés Lázaro (2000: 4146) call markers of epistemic modality that are used in declarative statements and that “build, themselves, an assertion, that reflexes how the speaker focuses the message introduced by the marker – or where the marker is immersed-, whether this message is considered for example, ‘evident’ or ‘known through someone else’”. The richest aspect to analyse is the discoursive behaviour where we find similarities among the expressions but also some interesting differences. The three items project a context in which there is only one possible option – in comparison with expressions such as ‘who cares?’, ‘what’s the use?’ where it is possible to consider other alternatives: ‘does it matter or not?’, ‘is it useful or not?’, respectively. Moreover, these discourse markers reflect an attempt to neutralize, up to a certain extent, the challenge of facing possible controversies among speakers. They are frequently used to orient the relationship between the participants towards a shared view of the world – although this common ground might not be taken as such – as the first case shows. Case 1. Topic of the conversation1 : Laura tells Romina about an evening out with friends to a pub called ‘Rektorado’ Argumentative chain: Assertion: Support:

I don’t like ‘Rektorado’ It’s stuffy – It’s expensive – The menu is not varied – the quality of the food is no good – Customers are not well treated

La: Después fuimos a … a Rektorado a comer. La: We then went to Rektorado to eat. Ro: Ah... las chicas fueron, con todo. Ro: Oh... the girls went, full of enthusiasm. La: Sí, yo re caliente, (risas) obvio. La: Yes, I was furious, (laughs) obvious. Ro: ¿Por? Ro: Why? La: Porque no quería ir al Rektorado, no me gusta el Rektorado. La: Because I didn’t want to go to Rektorado, I don’t like Rektorado. 1

Ungrammatical or odd versions in English might result from the fact that translation tries to respect the category of the word used in the original language.

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS

255

Ro: Yo este año voy a ir al comedor. A full. Ro: This year I’m going to the university canteen, for sure. La: Yo también. Bueno. Fuimos a Rektorado a comer. Obvio, que yo siempre “No, no”... La: Me too. Well, we went to Rektorado to eat. Obvious, I always “no, no…” Ro: Sí. Ro: Yeah. Case 1. Discoursive function of obvious shared with the other expressions: consensus

Rocío asks the reason of the affirmation ‘I was furious, obvious’ and it is necessary for her interlocutor to give more details so that she understands the context that Laura has projected. But the fact that Laura didn’t want to go to Rektorado is presented as the only possible alternative which, in fact was the expected thing to happen given the circumstances. The same kind of projection applies, in our view, to the other two expressions under analysis, as number 2 and 3 try to illustrate. Case 2. Topic of the conversation: Cecilia and Valeria talk about having a baby. Argumentative chain: Assertion: Having children now is inconvenient. Support: She hasn’t finished University yet – she has to find a job – she has to be independent. Ce: [No sé,] yo lo que pasa que por ahora no... Yo, Luis cada vez le están dando más ganas de [tener pero]… Ce: [I don’t know]. In my case, not now. I, Luis is more and more willing to have one [but] … Va: [¡Ah!, ¿sí?] Va: [Oh, really?] Ce: … igual yo por ahora no. No porque él, él: “¡Mirá qué lindo [bebé!”] Ce: All the same, not now, in my case. Not because he, he: “Look! What a cute [baby!”] Va: [Ceci], reci[bite y buscá…] Va: [Ceci], fin[ish your studies and then find] … Ce: [No, más vale] Ce: [No, of course] Va: … un trabajo primero. Va: … a job first. Ce: ¡Más vale! Ce: Of course! Va: No seas, no seas una madre que no... sabés tener un hijo y tener trabajo. Va: Don’t be, don’t be a mother who doesn’t… know how to have a child and a job. Ce: ¡No, me muero, más vale! Ce: No, I kill myself, of course! Case 2. Discoursive function of of course shared with the other expressions: consensus

Case 2 also shows that what is expressed in the stretch of discourse where of course occurs, is expected to be taken as the only option – or the best option – and could be paraphrased as ‘Surely, first I’m going to finish University and then look for a job.

MARINA GRASSO

256

Surely, I don’t want to be only a mother and not work’. Case 3 presents an instance of needless to say that goes on the same line as the previous fragments analysed. Case 3. Topic of the conversation: Romina and Laura talk about eating in the university canteen. Argumentative chain: Assertion: We must go to the university canteen. Support: Everyone goes there – it’s cheaper than other places – it’s really cheap. Ro: Yo este año voy a ir al comedor. A full. Ro: This year I’m going to the university canteen. Definitely. La: Yo también. La: Me too. Ro: Aparte, los chicos van todos los días. Ro: Besides, the guys go there every day. La: [¿Por qué?] La: [Why?] Ro: [Van todos] los días a la una. Ro: [They go every] day at 1. La: [¿Viste los horarios?] La: [Have you seen the timetables?] Ro: [Se juntan todos.] Ro: [Eveybody gathers together] La: Ah, ¿si? La: Oh, really? Ro: Ah, ¿son distintos? No. No. Ro: Oh, are they different? No, no. La: No, pero... los horarios son, salís a las doce, viste, más o menos, de la facultad, entonces... Pero bueno, hay que ir al comedor, ni hablar. Ro: No, but... the timetables are, you go out at twelve, you see?, approximately, from the university, so... Anyways, needless to say, we must go to the university canteen. Ro: Hay que ir al comedor. Un peso, es una ganga. Terrible. Ro: We must go to the university canteen. $1 is a bargain. Terrific. La: Lo que pasa que... bueno. La: The thing is... well. Ro: Lo que pasa que te.. te tienen que coincidir los horarios. La: The thing is that the canteen’s timetables have to fit yours. Case 3. Discoursive function of needless to say shared with the other expressions: consensus

Case 3 could be paraphrased as: ‘The fact that we must go to the university canteen is indisputable’. In movements where the studied expressions are part of a feedback move, obvious, of course and needless to say share the absence of an instance of negociation. A communion is created since the context of situation presented is accepted, and a high degree of intensification is shown. One participant manifests a certain way of seeing the world

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS

257

and the other one does not question the validity of what has been said. We associate this with the function of manipulation inside evaluation (Hunston & Thompson 2003), that can appear in the development of the interpersonal function, of building and maintaining the relationship between speakers. We could say that all the instances found used as part of a feedback move conform what Eggins & Slade call registering moves – in the sense of register of information– that is, “reactions that provide supportive encouragement for the other speaker to take another turn. They do not introduce any new material for negociation, and they carry the strong expectation that the immediately prior speaker will be the next speaker” (1997: 204). To illustrate this, we introduce the following cases from our corpus: Case 4. Topic of the conversation: Ana and Angelina talk about a friend’s new mobile phone Argumentative chain: Assertion: Clamshell phones are better Support: They are beautiful – they are more secure Ana: Está bueno pero no, viste cuando decís, yo hubiera elegido el otro, el Samsung, el Samsung, ¿no? Ana: It’s nice, but no, you know when you say `I would have chosen the other one, the Samsung’, the Samsung, don’t you think? Ang: Sí, pero XXX es caro. Ang: Yes, but XXX it’s expensive. Ana: Y bueno, me gusta. (Risas) Ana: Well, I like it (laughs) Ang: Qué raro, qué raro ella, gustándole todo lo que es caro. Ang: How strange, how strange, she liking everything that is expensive. Ana: Y obvio. Bueno. Ana: Well obvious. OK. Case 4. Obvious as part of a feedback move

There is not a change in the world view presented but a subscription and, sometimes, an enlargement of what has previously been said. This subscription to the interlocutor’s sphere, to the personal perception of the other speaker, allows for the development of strategies of positive politeness between the interlocutors, favouring mutual identification and solidarity (Martín Zorraquino & Portolés Lázaro 2000: 4156). In terms of the systemic functional linguistics, we would say that they contribute to the interpersonal function, creating affiliations. A fragment including another expression under analysis is number 5:

MARINA GRASSO

258

Case 5. Topic of the conversation: Belén, Mercedes, Paula and Romina talk about the requirements to pass a subject at the university Argumentative chain: Assertion: It’s better to pass the subject with a final exam than by working with other students Support: You don’t need to meet other people – you don’t have to coordinate timetables – the conditions of the place where the work should be done are no good – it’s far away and it implies a long journey to get there Bel: Es feo, yo prefiero que me tomen antes que hacer un trabajo, a mí no me gusta. Bel: It’s not nice. I prefer to give an oral presentation instead of writing a paper. I don’t like it. Mer: Y no. Mer: Right, no. Pau: O rendir un final… Pau: Or sit for a final exam… Bel: Sí, o rendir un final. Bel: Yes, or sit for a final exam. Pau: …porque aparte es de a tres, entonces entre que conocés a las otras dos XXX Pau: because, besides, the work should be done in a group of three, so you have to meet the other two Bel: Los horarios. Bel: The timetables. Pau: XXX y además lo estamos haciendo en el Observatorio, entonces entre que la gente del Observatorio pueda, que nosotras tres podamos, que lo podamos hacer… Pau: XXX and besides we are doing it in the Observatory, so we have to combine timetables with the people from the observatory, among us three... Bel: Claro. Bel: Sure. Pau: …que no se corte la luz, que no… Pau: and there there are the power cuts… Bel: Nosotros para hacer una charla de anemia en … y viajar hasta Berisso es un quilombo. Bel: In order to give a talk about anaemia we have to travel to Berisso. It’s total chaos. Rom: Sí, ni hablar. Rom: Needless to say, yes. Case 5. Needless to say as part of a feedback move

All these could be considered boosters, according to Holmes (1995), and upgraders, in House & Kasper’s words (1981) – that is, expressions that reinforce the meaning of the proposition where they appear (Watts 2004: 185). In Eggin’s (2004: 160) classification of modal adjunts, the expressions analysed would be mood adjunts which add interpersonal meaning to the clause, “that is they add meanings which are somehow connected to the creation and maintainance of the dialogue”. We can talk here of an emphatic use – in terms of Zorraquino & Montolío Durán (1998:33) – that “reinforces the assertion developing the informative value that the utterance has, in the sense of emphasizing the orientation with which it has been uttered”.

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS

259

As a partial summary of what has been said up to now, the following chart is presented, with the similarities found among the three expressions in question:

Table II Similarities among the expressions under analysis

We will now look at the differences in the discoursive behaviour of each expression. Peculiarities Obvious

To the idea that it is clear that facts could not have been otherwise, in the case of obvious a generalization is added, applicable to people or facts, depending on the case. Cases #1 and #4 previously presented show occurrences that contain generalizations related to the personal experience of the speaker. In Case # 1: When Laura says “Yes, I was furious, obvious” and “Well, we went to Rektorado to eat. Obvious, I always ‘no, no’ ...’’ the possible generalization related to the personal experience of the speaker could be: ‘I always get angry if we go to Rektorado. I always reject going there’. In # 4: When Ana replies “Well obvious. OK.” to Angelina’s “How strange, how strange of her liking everything that is expensive” we could say that the generalization with focus on an individual experience can be paraphrased as: ‘I always like expensive things’. In these cases, through the use of obvious the individual situation of the speaker is evaluated. “As we tell a story of personal experience, we remember how we were, how we believe we are, or how we wish to be. Evaluation in narrative exposes mirrors of multiple senses of the self ” (Cortazzi & Jin 2003:120) not only hypothetical but also dreamworld selves. Common ground with the hearer is projected. The speaker presents the situation under a structure that could be stated as: ‘When X happens, I always react in Y way’

MARINA GRASSO

260

There are other fragments – as the following – in which, we believe, the use of obvious is related to a less individual idea, that can be generalized, in one way or another, to any person and not to a single one. Case 6. Topic of the conversation: Paula talks with Belén about the changes in a subject at university as a consequence of a change in the curriculum Argumentative chain: Assertion: Some changes in the curriculum are negative Support: Students are taken as guinea pigs – the reduction in the number of teaching hours does not help anyone – the division of some subjects does not help either. Pau: … porque hay muchas materias, por ejemplo, que.. dicen.. en Contactología, se la pasaron diciendo “Hoy no llegamos, hoy no llegamos XXX que no llegamos”. Hay ocho horas en el.. en el plan de estudio y nos dan tres. Pau: … because there are lots of subjects, for example that... they say.. in Contactology, they kept saying “Today, we won’t be able to cover all the topics, today we won’t be able XXX we won’t be able”. The curriculum establishes 8 hours and they teach us only 3. Bel: Claro, obvio que no van a llegar. Bel: Obvious, they won’t be able to cover all the contents, for sure. Pau: Es que sí, si no llegás, entonces dame las que te corresponden… Pau: It’s that, if you can’t make it, teach me the corresponding number of hours… Bel: Claro. Bel: Sure. Pau: … no, no me digas “Ay, no llegamos” y te quedás de brazos cruzados. Hay mucha... no sé es todo un problema... Pau: … don’t, don’t tell me “Oh, we won’t be able to cover all the syllabus” and remain with your arms folded. There’s a lot… I don’t know. Everything is a problem… Case 6. Obvious. Generalization with a focus on people – collective experience

Unlike the previous case, here a collective scope is considered, that seems to manifest that what has been exposed is not only unquestionable but also applicable to any person. This generalization with focus on a collective experience can be paraphrased as ‘Nobody would find the given time sufficient to teach the complete subject’. Here the speaker presents the situation under a structure that could be stated as: We /all react/ in X way every time Y /happens/ /would all react /if Y happened’ In the rest of the occurrences found, we believe that the focus is not on the people but on the facts that are presented, in some way, as an inevitable consequence.

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS

261

Case 7. Topic of the conversation: Ana and Angelina talk about buying a new phone Argumentative chain: Assertion: Flip phones are better than the others Support: They are beautiful – they are more secure Ana: Me gustan los de tapita, [son los que vimos.] Ana: I like flip phones [those that we’ve seen.] Ang: [Sí…] Sí, son hermosos. Aparte son más seguros. Ang: [Yes...] Yes, they are beautiful. Besides, they’re more secure. Ana: ¿Por? Ana: Why? Ang: Y… no se te marcan solos si están en la mochila. Ang: Well... their keys are not pressed by themselves when you keep the phone in your rucksack. Ana: Ah, sí, eso… Ana: Oh, yes, that... Ang: Esas cosas. Ang: Those things… Ana: No, está bueno. Ana: No, it’s good. Ang: Vamos y compramos. Ana: Let’s go and buy. Ana: Por ahí son los más caros. Ana: May be they are the most expensive ones. Ang: Y, obvio. (Risas) Y, obvio. Como el de Sil… ése, de los más, digamos, [menos caros,] menos caros de los de tapita. Ang : Well, obvious. (laughs) Well, obvious. Like the one Sil has... that one, one of the most, let’s say, [least expensive ones,] least expensive with a flip top. Case 7. Obvious. Generalization with focus on the facts

The generalization with focus on the facts could be paraphrased in this case as ‘if the phone is more sophisticated, it will be necessarily more expensive’. Here the speaker presents the situation under a structure that could be stated as: ‘When the facts are X, Y always happens’

MARINA GRASSO

262

The following table shows the peculiarities of obvious in relation to the other two expressions analysed:

Table III Peculiarities of obvious in its discursive behaviour

These generalizations implied by obvious are exclusive of this expression. Of course As said before, a similarity between the three expressions in question, is the lack of negociation of the facts presented by one of the speakers. However, the meaning this marker suggests is different from the others. In case #2, Cecilia’s words can be paraphrased as ‘Surely I’m going to finish University and then I will look for a job before having a baby. Surely, I don’t want to be only a mother who doesn’t work’. The speaker presents the situation under a structure that could be stated as: ‘X said before, is evident’ Apart from this one, we find other cases in the corpus where the controversy does not consist of deciding whether to agree with the interlocutor or not, but to question the facts themselves. Case 8. Topic of the conversation: Enzo and Javier talk about passing a subject at university. Argumentative chain: Assertion: It’s better to sit for the final exam instead of re-attending the subject to avoid sitting for the written test. Support: It takes less time. You can attend other subjects. En: …si vas a final, hay muchos que van a final y… y… cómo te puedo decir, meten la cursada tienen que dar el final, y lo que hacen es cuando agarra, arranca otra vez la inscripción se vuelven a anotar otra vez en esa materia y la vuelven a hacer entera [para no dar el final.] En: … if you sit for the final exam and… and… how can I explain? There are people who pass the practical classes with a low mark and have to sit for the final exam, what they do is, when classes start again, they re – attend the subject to see if they can pass it with a higher mark so that they don’t have to sit for the final written part.

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS

263

Ja: [Ah, para salvar…] Ja: I see, in order not to… En: Sí, para no dar el final, eso es al pedo igual. Más vale prepararlo una vez y listo. En: Yes, in order not to sit for the written part; that’s useless anyway. Of course it’s better to study for the written part XXXXX Ja: Y prepararlo por ahí te lleva un mes [si hacés la cursada te lleva XXX.] Ja: And studying for the final exam takes a month [if you re-attend the subject it takes XXX] En: [Y pero… no perdés seis meses.] En: [But…you don’t waste six months.] Ja: Claro, por eso, no, no, por eso que por ahí es más conveniente hacer, eh, tirar el final que hacer toda la cursada de vuelta [el final…] Ja: Sure. That’s why, no, no, that’s why it’s more convenient to study, er, sit for the final exam than to attend the subject again [the final exam…] En: [Sí, ni hablar.] En: [Yes, needless to say.] Ja: …lo preparás en un mes o dos meses. [O menos, o menos.] Ja: … you study it in a month or two. [Or less, or less.] En: [Aparte con la cursada ya] te da, te da la posibilidad de seguir cursando la otra materia que son correlativas. En: [Besides, with the completion of the practical classes you have the possibility of attending other subjects.] Case 8. Of course. Consideration of more than one possibility

The context presented here could be paraphrased as ‘It’s better to prepare the exam than not to do so’. The other difference relies on the fact that in this case, more than one possibility is implicitly considered (the idea of sitting and not sitting for the exam), and that the option chosen by the speaker is shown as the most appropriate one according to him. Whereas in the previous case (#2) finishing university and getting a job before starting a family is presented as the only possibility. The following table shows the peculiarities of of course in relation to the other two expressions analysed.

Table IV Peculiarities of of course in its discursive behaviour

Needless to say Fragment # 3 (“needless to say, we must go to the university canteen”) exemplifies the similarity of needless to say with the other two expressions in question. The context

MARINA GRASSO

264

introduced here could be paraphrased as ‘It’s undoubtful that going to the university canteen is the best option’. In this case, the speaker presents the situation under a structure that could be stated as: X previously said is indisputable’ On the other hand, the following fragment introduces an instance with a function of needless to say different from the rest of the corpus. Case 9. Topic of the conversation: Enzo and Facundo talk about a football match with boys from another city called Pehuajó. Argumentative chain: Assertion: The boys from Pehuajó are stupid. Support: They are all gays – they are slow to play En: Al final tenemos que hacer el partido con los pelotudos estos de Pehuajó. En: In the end we have to play the match with the stupid guys from Pehuajó. Fa: ¿Sí? Fa: Really? En: Ni hablar. En: Don’t mention it. (1) Needless to say Fa: Se la comen. Fa: They are gays. En: Ni hablar que se la comen, toda, toda. En: (2) Needless to say that they are gays. Fa: Todos los de Pehuajó se la comen. Fa: All the boys from Pehuajó are gays. Case 9. Needless to say. Invitation to continue talking about a certain topic

We believe that the instance of needless to say thematized with ‘that’ (2) responds to the description previously mentioned: an ‘aligment’ of the speaker with his interlocutor (Carranza 1998:78). However, the first case in this fragment (1), not necessarily intends to present an option as apparently indisputable. At first sight, it seems to announce the preference of not dealing with certain topics; but, in our opinion, it functions as an invitation to talk, more than as an invitation to close the conversation. What we understand the speaker must have expressed in this case, could be paraphrased as ‘No comments’ (possible subtext: let me tell you why I say this).

Table V Peculiarities of needless to say in its discursive behaviour

EXPRESSIONS USED BY YOUNG ARGENTINE SPANISH SPEAKERS

265

3. Conclusion In the present work we have grouped three expressions that share the same linguistic and interactive behaviour: they appear inside a complete propositon, as the only element of the construction and thematized with the conjuntion ‘that’. Besides, they cannot be used at the beginning of a conversation because they are constructions dependent on previous moves. Regarding their discursive behaviour, we have seen that the three seem to cancel the option of considering other alternatives as possible in the same situation. They articulate what Ducrot calls cooriented members (Portolés 1998), that is, with the same argumentative orientation, and seek for concurrence – fostering the cooperation between interlocutors, politeness strategies in rapport management. It is in the discursive function where the most interesting differences are found. Obvious is the only expression that implies a generalization with different focuses, added up to the idea of lack of confrontation previously suggested. Of course and Needless to say also present a few cases that are different from the rest: one instance of use of of course that contemplates more than one implicit possibility and, in the second case, an invitation to continue speaking about a topic that has already been introduced into the conversation. The divergences in the discursive plane, are what make these expressions not necessarily interchangeable with one another. Bibliography Bosque, Ignacio & Violeta Demonte (ed.) (2000). Gramática descriptiva de la lengua española. Madrid: Espasa Calpe. Carranza, Isolda E. (1998). Conversación y deixis de discurso. Córdoba: Universidad Nacional de Córdoba. Cortazzi, Martin & Lixian Jin (2003). Evaluating evaluation in narrative. In: Hunston, S. & G. Thompson (ed.). Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford University Press, 102-120. Coulmas, Florian (ed.) (1981). Conversational Routine: Explorations in Standarized Communication Situations and Prepatterned Speech. The Hague: Mouton. Eemeren, Frans H. van, R. Grootendorst, S. Jackson & S. Jacobs (2000). Argumentación. In: Van Dijk, T.A. (ed.). El discurso como estructura y proceso: Estudios sobre el discurso I. Una introdución multidisciplinaria. Barcelona: Gedisa, 305-333. Eggins, Suzanne (2004). An Introduction to Systemic Functional Linguistics (2nd ed.). London: Continuum International Publishing Group. Eggins, Suzanne & Diana Slade (1997). Analysing Casual Conversation. London: Cassell. Holmes, Janet (1995). Women, men and politeness. London: Longman. House, Juliane & Gabriele Kasper (1981). Politeness markers in English and German. In: Coulmas, F. (ed.). Conversational Routine: Explorations in Standarized Communication Situations and Prepatterned Speech. The Hague: Mouton, 157-85. Hunston Susan & Geoff Thompson (ed.) (2003). Evaluation in Text. Oxford: Oxford University Press.

266

MARINA GRASSO

Hunston Susan & Geoff Thompson (2003). Evaluation: an introduction. In: Hunston, S. & G. Thompson (ed.) (2003). Evaluation in Text. Oxford: Oxford University Press, 1-27. Martín Zorraquino, Maria Antonia & Estrella Montolío Durán (ed.) (1998). Los marcadores del discurso. Teoría y análisis. Madrid: Arco libros. Martín Zorraquino, Maria Antonia & José Portolés Lázaro (2000). Los marcadores del discurso. Cap 63. In: Bosque, I. & V. Demonte (ed.) (2000). Gramática descriptiva de la lengua española. Madrid: Espasa Calpe, 4051-4213. Montolío Durán, Estrella (1996). Gramática e interacción, (ensayo metodológico para el análisis del español conversacional). In: Briz, A., M.-J. Martínez & Grupo Val. Ed. Co (ed.). Prágmática y gramática del español hablado. Valencia: Universidad de Valencia. Portolés, José (1998). La teoría de la argumentación en la lengua y los marcadores del discurso. In: Martín Zorraquino, M.A & E. Montolío Durán (ed.). Los marcadores del discurso. Teoría y análisis. Madrid: Arco libros, 71-92. Van Dijk, Teun A. (ed.) (2000). El discurso como estructura y proceso: Estudios sobre el discurso I. Una introdución multidisciplinaria. Barcelona: Gedisa. Watts, Richard J. [2003] (2004). Politeness. Cambridge: Cambridge University Press.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 267-279 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION: LE CAS DU ROUMAIN

LILIANA HOINARESCU

1. Dans le présent article, nous nous sommes proposé d’analyser les contextes communicatifs dans lesquels la fonction déictique de la référence personnelle (le pronom sujet) est doublée par des valeurs rhétoriques ou argumentatives. Nous aurons en vue exclusivement la référence personnelle (en particulier, les cas d’auto-désignation déictique) et nous utiliserons des extraits du corpus de roumain parlé contemporain, publié sous la coordination de Liliana Ionescu-Ruxăndoiu, L’interaction verbale dans la langue roumaine actuelle, ainsi qu’un autre exemple tiré d’un débat électoral télévisé (enregistré). Il est bien connu qu’en roumain (tout comme en italien ou en espagnol), l’expression du pronom sujet n’est pas obligatoire, les désinences verbales indiquant sans équivoque les rôles communicatifs assumés par les participants à une interaction. On a également souligné le fait que la présence du pronom sujet est liée à la nécessité d’exprimer l’emphase et/ou l’opposition (cf. Ionescu-Ruxăndoiu 1999: 87; GALR II 2005: 334335). À partir de ces remarques théoriques, à valeur axiomatique dans la linguistique roumaine, nous avons l’intention d’identifier les contextes où les locuteurs ressentent le besoin de renforcer leur engagement illocutionnaire, de les systématiser et, éventuellement, d’isoler certaines règles d’application discursive (évidemment, sans valeur absolue); si du point de vue grammatical l’utilisation déictique des pronoms est redondante et indifférente, du point de vue pragmatique, elle est significative et en conséquence implique une certaine récurrence. 2. On sait bien que, dans la perspective des théories de politesse, une attitude auto-centrique ou égocentrique (exhiber sa propre personne) reflète un comportement communicatif déficient et, finalement, antisocial. Conformément à la «loi de modestie», subsumée au principe plus général de politesse, être poli consiste à valoriser la face d’autrui au détriment de la sienne propre (cf. Kerbrat-Orecchioni 1996: 62). Dans cette étude, pour des raisons méthodologiques, nous désignerons comme égocentriques exclusivement ces situations susceptibles d’être sanctionnées du point de vue communicatif, et nous désignerons comme subjective ou réflexive toute forme discursive dans laquelle le locuteur se réfère avec insistance à soi-même, y compris par les moyens grammaticalisés, sans préjudicier par cela l’image publique ou la face de l’interlocuteur. Cette distinction nous semble nécessaire, étant donné que, dans un sens neu-

268

LILIANA HOINARESCU

tre, étymologique, tout énoncé est intrinsèquement égocentriques, c'est-à-dire lié d’une manière indissoluble, à son énonciateur (par l’acte-même de prendre la parole, le locuteur exprime un point de vue personnel et s’assume une responsabilité discursive. Même les situations de polyphonie énonciative – les énoncés ironiques citationnels ou intertextuels – ne font exception qu’en apparence à cette évidence discursive; en ce qui concerne le problème envisagé, c’est l’intention discursive de l’énonciateur, sa décision de transmettre un contenu informatif quelconque qui importe, et non les moyens d’expression utilisés – directs ou indirects, clairs ou ambigus). Comme nous essayerons de le démontrer par l’analyse des contextes concrets d’énonciation, l’accumulation des marques grammaticales (pronominales) pour désigner l’instance énonciatrice – «une syntaxe égocentrique» pour citer un syntagme des grammairiens roumains (cf. GALR II 2005: 335) – n’exprime pas de façon univoque une manifestation de l’égocentrisme pragmatique ou communicatif. 3.1 Une première illustration significative pour ce que l’on peut prendre superficiellement pour ‘narcissisme conversationnel’ est le récit personnel, la confession ou la confidence, qui représente le prototype des narrations subjectives et leur forme la plus radicale, parce que, dans ce cas, toute l’orientation discursive est centrée sur l’énonciateur. L’acte confidentiel correspond à un besoin essentiel de l’être humain, au sens psychologique et social, de chercher la compréhension et la sympathie de l’autre. Comme modèle discursif, la confession ou la confidence suppose un cadre donné (il faut remarquer le fait que ce contexte stable est prouvé par l’existence de certaines variantes institutionnelles: la variante traditionnelle – ecclésiastique ou la variante moderne – la psychanalyse): les rôles des inter-actants sont préétablis, le locuteur sollicitant non seulement le droit à la parole, mais aussi une certaine réaction affective et intellectuelle de la part de l’allocutaire1 . Le corpus nous a permis d’isoler un exemple représentatif pour ce type interactionnel. On peut observer, en confrontant les deux versions, roumaine et française, qu’en français il faut doubler par la forme disjointe le pronom sujet, afin d’obtenir l’effet rhétorique de l’utilisation du pronom sujet en roumain. Ex. 1: Confessions – interaction face à face, informelle; dominante coopérative; moi je (en roumain eu) – marque émotionnelle B: [e cazu clasic [al unei mame vitrege // B: [c’est le cas classique [d’une belle-mère A: [nu_ nu // A: [non_ non +B: copilu care află la UN moment dat că nu este copil↑ // +B: l’enfant qui apprend à UN moment donné qu’il n’est pas enfant↑ A: eu am ştiut↓ // A: moi je l’ai su↓ 1 Pour

une présentation théorique systématique et détaillée de la confidence comme type particulier d’activité discursive, voir Kerbrat-Orecchioni & Véronique Traverso 2007: 1-24.

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION

269

B: ai [ştiut dintodeauna? // B: tu l’as [depuis toujours su? A: [eu am ştiut dintodeauna↑ dar CE să vă spun↑ adică: starea asta a mea de copil adoptat adică:↑ nu mi-am pus F:OARte mari probleme cum: într-a-┴ cum este în filme. adică: fiind nepoată↑ N-AM considerat că <MARC ne desparte> adopțiunea asta. <R am considerat-o întodeauna ca pe mama.> dar DE la doisprezece ani cînd ea mi-a spus↓ deşi eu ştiam↓ eu ştiam (xx)↓ totdeauna că eu n-o iubesc pe ea foarte mult↑ ca şi cum: cum o iubesc eu pe mama↓ // A: [moi je l’ai depuis toujours su↑ mais qu’est-ce que je peux vous dire↑ c'est-à-dire: cet état à moi d’enfant adopté c'est-à-dire:↑ je ne me suis pas posé TROP de questions comme: vraim ┴ comme il se passe à la télé. c’est à dire: étant la nièce↑ JE N’AI pas considéré que cette adoption <MARC nous sépare>. <R j je l’ai toujours considérée comme ma mère.> mais À l’âge de douze ans quand elle me l’a dit ↓ bien que moi je le savais ↓ moi je le savais (xx)↓ toujours que moi je ne l’aime pas beaucoup↑ comme si: comme moi j’aime ma mère↓ (Ionescu-Ruxăndoiu 1999: 50-51)

Comme on peut le remarquer, la forte subjectivité de l’énoncé (induite aussi par le biais de la deixis personnelle) n’indique pas l’opposition entre les interlocuteurs ou encore l’emphase, mais au contraire la forte charge émotionnelle du récit, par laquelle le locuteur réclame la réaction affective de l’interlocuteur. C’est le contrat de confiance institué entre les deux partenaires qui est spécifique à ce type discursif: on dévoile son intimité devant quelqu’un que l’on considère capable de comprendre et d’approuver, capable d’accorder attention et sympathie et, non en dernier lieu, capable de discrétion. Par l’acte même de sélecter son confident, le locuteur le valorise, en lui attribuant implicitement ces qualités. Bien qu’il parle de soi-même, le locuteur ne transgresse pas la loi de modestie (en d’autres termes, il ne commet pas une infraction pragmatique), bien au contraire, il flatte son interlocuteur et se place soi-même dans une position d’infériorité, le dévoilement de soi étant un acte menaçant pour sa face négative. 3.2 Dans les contextes communicatifs à dominante compétitive, l’auto-désignation déictique a, de façon naturelle, un rôle important et systématique. Se référant à sa propre personne, le locuteur souligne, d’une part, l’opposition avec le point de vue avancé par son interlocuteur et, d’autre part, tente de soutenir et imposer sa propre thèse. Le degré de compétitivité de l’interaction, qui marque une progression à partir de la conversation → discussion → discussion conflictuelle (en contradictoire) → débat (expression de la divergence) → dispute, n’influe pas sur l’application de cette règle, qui reste relativement constante et générale. Ce sont les types de stratégies discursives utilisées qui diffèrent et qui entraînent consécutivement une nuance pragmatique sensiblement différente de l’expression du pronom sujet. Ainsi, dans une interaction face à face, informelle, où le statut social des participants est relativement symétrique et le dialogue a une finalité interne (les participants font connaissance dans le train et ont une discussion sur des

LILIANA HOINARESCU

270

questions générales), la principale stratégie argumentative est le récit autobiographique (dans un cadre institutionnel ou entre des personnes ayant un degré d’instruction plus haut les arguments sont plutôt d’ordre théorique). Ex. 2: Conversation dans le train – interaction face à face, informelle; dominante compétitive; moi je (en roumain eu/io) – rôle argumentatif persuasif a) A: eram în brăila↓ sînt brăileancă↓ şi ă# timpurile care le trăim acum leam trăit io atunci în brăIla. TOT aşa cu criza asta financiară↓ cu LIPsuri multe. […] şi tot aşa cu ┴ şi cu toate astea lumea ă totuşi îşi făcea: viața: aşa cum o putea↓ îşi făcea copii↓ // A: j’étais à Braila↓ je suis de Braila↓ et euh# les temps qu’on vit à présent moi je les ai vécus à ┴ cette époque-là à Braila. DE MEME il y avait cette crise financière↓ avec beaucoup de privations […]. et il en était de même avec ┴ et pourtant on euh toutefois on menait: sa vie: chacun comme il pouvait↓ on faisait des enfants↓ B: io nu vă contrazic↓ […] // B: moi je ne vous contredis pas↓ […]

b) B: io vă spun ALTĂ chestie. io sînt bugeTAR↓ şi TOT nu iau bani. adică: lucrez în mi:ni:sterul să:– sănătății. şi n-am luat banii dă trei luni. […] deci io n-am luat banii din# iu– iul–iulie […] eu nu mai am ZI de salariu de un an jumate eu iau ┴ // B: moi je vous dis une AUTRE chose. moi je suis fonctionnaire. et CEPENDANT je ne touche pas mon salaire. c’està-dire: je travaille au mi:ni:stère de la sa:–santé. et j’ai pas touché de sou depuis trois mois. […] donc moi j’ai pas touché de sou depuis #ju – juillet. […]. moi je n’ai plus le JOUR de paie depuis une année et demie moi je touche ┴ A: # [E nu↓ da’ da:’ vorbeam pentru copil şi pentru dumneavoastră în geneRAL // A: # [Euh mais non↓ mais mais : je disais cela de l’enfant et de vous en généRAL. B: [dacă eu n-am (securitatea unui) salariu↓ nu mai zic de alte chestii […] // B: [si moi j’ai pas (la certitude d’un) salaire↓ pour ne pas parler d’autres affaires […]

c) B: [doamnă↓ eu eu lucrez şi n:ici nu vreau să vă impresionez↓ ni:ci nu vreau să mă dau cum se spune# rotund. dar în sănătate la momentul actual e un HAOS toTAL. […] eu am avut ocazia să să să să dau un eXEMplu. deci am plecat într-o dimineață la serviciu↓ […] şi era un# bătrîn↓ căzut↓ de dimineață↓ i-a venit rău şi ă mă rog↓ l-a(u) luat cineva cu maşina şi i-am zis HAI domle (xxx) la policlinica […] ei↓ vreau să spun că l-a luat cetățeanu ăla-n maşină↓ am mers şi eu cu el↓ l-am dus la CINCI MEdici. […] <R aşa>↓ şi: l-am dus şi io la↑# MEdici […] m-a plimbat.#

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION

271

can final# la un ┴ la al cincilea mi-a venit dreacii. […] i-am zis↓ dom doctor↓ sînteți al CINcilea domne↓ […] al cincilea zice# <J du-l şi tu că eu nam timp↓ am treabă↓ am mu–mulți pacienți. […] <F,IM lasă-mă-n pace># <R du-l de-aici># <R n-am nevoie de el>. şi i-am zis. domle:↓ dacaveți nevoie: de PLAtă:↓ pentru a::– a:cordarea unui ajutor↓ cît de banal şi micuț ar fi↓ vă plătesc io. ştiți↑ fac io chestia asta↓ cît costă plătesc io↓ da’ n:u să poate._păi io trebuia(m) să fiu şi io la muncă._dumneata crezi că io mă plimb cu el pîn la doişpe ziua↑şi pîn la urmă l-a luat. // B: [madame↓ moi moi je travaille et j’ai pas non plus l’intention de vous toucher↓ ni de passer pour quelqu’un d’important. Mais dans le domaine de la santé à l’heure actuelle c’est un CHAOS toTAL […]. moi j’ai eu l’occasion voilà je je je vous donne un eXEMple. Donc je suis parti un beau matin au boulot ↓ […] et il y avait un# vieil homme↓ tombé↓ tôt↓ il a eu une malaise et euh bon↓ quelqu’un l’a emmené avec sa voiture et je lui ai dis ALLONS-Y monsieur à l’hôpital […] donc↓ ce citoyen l’a emmené avec sa voiture↓ je suis allé avec eux moi aussi↓ on l’a fait voir par CINQ MÉdecins. […] <R voilà>↓ et: moi aussi je l’ai accompagné chez les↑# MEdecins […]. on m’a promené.# pour qu’à la fin# chez le ┴ chez le cinquième j’ai pété les plombs. […] je lui ai dit↓ docteur↓ vous êtes le CINQuième↓ […] le cinquième m’a-t-il dit# <J emmène-le d’ici parce que j’ai pas de temps↓ je suis occupé↓ j’ai beaucoup de patients […] <F,IM fiche-moi la paix># <R emmène-le d’ici ># <R j’ai pas besoin de lui>. et je lui ai dit. écoutez docteur:↓ si c’est une question: d’ARGENT↓ pour lui do:: – donner votre aide↓ quelque banale et modique elle soit↓ moi je vous paye de ma poche. voilà↑ moi je le fais↓ combien ça coûte moi je paye↓ mais c’est pas possible. euh moi je devais être au boulot moi aussi. vous imaginez que je vais me promener avec lui jusqu’à midi↑ et finalement il l’a reçu. (Ionescu–Ruxandoiu 1999: 28-32).

Dans ce fragment, on constate une poly-fonctionnalité de l’auto-désignation déictique. Dans les exemples 2a et 2b au début du dialogue, le recours aux formes pronominales (des deux partenaires) reflète, en plus du désir de souligner le point de vue personnel, la tendance à restreindre le champ thématique, du général au particulier (notamment en ce qui concerne le locuteur B). Ce glissement est naturel, dans la mesure où toute conversation entre deux personnes qui ne se connaissent pas va des considérations générales jusqu’aux considérations personnelles; ainsi, les interlocuteurs s’offrent des informations contribuant à l’élargissement du territoire commun de discours. Orientant la conversation vers l’expérience personnelle, les deux partenaires montrent leur attitude coopérative, leur intérêt pour le sujet de la discussion et leur désir de continuer le dialogue. Dans les répliques de B, surtout dans l’exemple 2c ci-dessus, quand le récit est plus long, l’auto-désignation déictique soutient l’éloquence du locuteur, fonctionnant comme un marqueur de son implication émotionnelle. La narration acquiert un pathos

272

LILIANA HOINARESCU

particulier, par sa forte valeur personnalisée et continuellement authentifiée par le locuteur. Nous ne croyons pas qu’il s’agisse d’un auto-centrisme ou égocentrisme communicatif, étant donné que le locuteur oriente le discours vers son interlocuteur, cherchant à gagner sa bienveillance et sa compréhension (il s’agit d’une stratégie de la politesse positive). En d’autres termes, le locuteur garantit par l’insistance sur la référence personnelle la vérité de ce dont il parle, adhère à son propre discours et le certifie, obtenant de la crédibilité dans la construction de son message. En outre, le récit personnel permet au locuteur de se présenter soi-même, avec ses motivations et ses réactions, ce qui entraîne de manière implicite la coparticipation de l’allocutaire, sa réaction émotionnelle, créant ainsi les prémisses d’un accord tacite entre les partenaires. Comme stratégie argumentative, le récit personnel ou autobiographique n’a donc pas un effet de nature logique-rationnelle, mais empathique-affective ; d’ailleurs, on l’utilise avec la même efficacité tant dans le discours public (c'est-à-dire le discours des médias, le discours politique, le discours publicitaire) que dans le discours littéraire. Elle permet au locuteur de soutenir fermement son point de vue et de maintenir, en même temps, une relation de bienveillance et de coopération avec l’interlocuteur. De façon paradoxale, l’emphase n’exprime pas l’intention de dissociation, mais d’identification (affective) avec l’interlocuteur. Tout récit personnel a cet effet, mais, comme nous l’avons mentionné plus haut, on rencontre le plus haut degré d’implication affective dans les confessions, parce que, dans ce cas, le locuteur cherche exclusivement à se présenter soimême, et non à défendre en subsidiaire un argument. 3.3 Dans l’exemple donné sous 3, nous prendrons en considération la fonction différente acquise par la référence personnelle dans une interaction télévisée, de type conflictuel. Les deux interlocuteurs, le président et le vice-président de la Fédération Roumaine de Boxe, s’accusent réciproquement de mauvaise gestion de la fédération. La dispute a une finalité publique, externe, et la transmission télévisée augmente son caractère formel et sa charge symbolique. Ex. 3: Conflit verbal – interaction télévisée, formelle; dominante nettement compétitive; moi je – toi tu (en roumain eu/io – tu) – l’opposition; la négociation de l’identité sociale/de rôle B: ruDEle↓ da’ vreau să te-ntreb ceva↓ acu:ma:: NU ştiu:↓ <Z io văd aşa↓ <@ că toate> proble:me:le sî:nt bu:ne:↓> greşesc şi io:↓ greşeşti şi tu:↓ normAL deci eu nici n-am ZIS că sînt un_om perfect↓ şi nici nu pot să susțin↓ că nu este <R nimeni>↓ ((inspiră)) dar vreau să te-ntreb ceva↓((inspiră)) de CE ții aşa tu NEAPĂrat↓ deci tu eşti în stare să faci ORIce ca s-ajungi preşedintele acestei federații. // B: ruDEl↓ je veux te demander quelque chose↓ main:tenant:: je NE sais PAS↓ <Z moi je vois les choses ainsi↓ <@ que tous> les problè:mes son corrects↓> je peux me

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION

273

tromper moi aussi:↓ toi tu peux également te tromper:↓c’est bien normAL donc d’ailleurs moi je n’ai pas DIT que je suis parfait↓ et de même je ne peux pas soutenir↓ que <R personne> le soit↓ ((inspire)) mais je veux te demander quelque chose↓((inspire)) pourquoi EST-CE que toi tu tiens À TOUT prix↓ donc toi tu es capable de faire N’IMPORTE quoi pour devenir le président de la fédération. A: nu:. nu:. // A: non:. non:. B: ba DA↓ // B: mais SI↓ A: ştii bine. am discutat asta //A: tu le sais bien. nous en avons déjà discuté B: [NU-I adevărat. // B: [c’est PAS vrai. […] (Ionescu–Ruxandoiu 1999: 191) […] B: ruDEle↓ ruDEle↓ cred că tu nu-nțelegi un lucru↓ <Z tu NU-nțelegi un lucru>. tu nu TREbuie să mă INVI:ȚI pe mine↓ // B: ruDEl↓ ruDEl↓ je crois que tu ne comprends pas une chose toi↓ <Ztoi tu ne comprends pas une chose>. toi tu N’AS PAS À m’INVI:TER moi↓ A: nu te-am invitat la: [inaugurarea hotelului↑ // A: je t’ai invité à [l’inauguration de l’hôtel n’est pas↑ B: [deci tu nu nu trebuie să mă inviți pe MINE. tu tre:be să mă anu:nți pe mi:ne::↓ […] EU la RÎNdu me:u↓ trebe s-anunț TOT biROu federal↓ tu NU poți să faci de CApu tău # NIMIC. [absolut nimic. // B: [donc toi tu n’as pas à m’inviter MOI. Toi tu dois me l’annoncer à moi. […] MOI à mon TOur↓ je dois le dire à TOUT le buREAU fédéral↓ toi tu NE peux RIEN faire à ta TÊte #. [absolument rien. A: [eu nu fac de capu meu nimic↓ eu- // A: [moi je fais rien à ma tête↓ moi je (Ionescu–Ruxandoiu 1999: 193)

La dissociation nette entre les interlocuteurs est exprimée tant par le désaccord brutal, sans moyens d’atténuation – les études sur les corpus attestent le fait que l’expression du désaccord est, en principe, accompagnée d’une excuse (cf. Kerbrat-Orecchioni 1992: 235) –, que par la désignation pronominale des instances communicatives. De cette manière, les deux protagonistes précisent leur position discursive. On a souligné le fait que la dispute constitue un type interactionnel instable, qui débouche soit sur la violence, soit sur la rupture de l’interaction, soit vers la résolution par le retour à la discussion, voire à la conversation (Vion 1992: 139). Dans l’exemple analysé, le cadre discursif se maintient relativement constant, aucun des interlocuteurs ne se montrant prêt à céder. L’alternance des pronoms je/moi – tu/toi (en roumain eu/io – tu) est corrélative à la dynamique des échanges et également à une logique argumentative à double sens, par le biais de laquelle les locuteurs doivent, d’une part, protéger leur image, en réfutant les accusations reçues, et, d’autre part, discréditer l’image de l’adversaire. Cependant, la logique offensive prévaut, les adversaires répondant à une accusation par une autre accusation ou, ce qui est plus intéressant pour ce qui nous concerne, en démontant ironiquement les prétentions de l’adversaire.

274

LILIANA HOINARESCU

Ex. 4: Conflit verbal – interaction télévisée, formelle; dominante nettement compétitive; moi je – toi tu (en roumain eu/io – tu) – le sarcasme A: mi-ai sp- ┴ m-ai întrebat CE vei face. <R vei fi preşedintele-> şi <R io țiam spus> NU: nu-mi trebui AŞA ceva:: ↓ însă nu am încoTRO în momentu_ă:sta↑ federația NU POAte # să rămînă neconDU:să↑ ştii că SÎNT probleme la nivel internațional↓ ştii din auzite↓ eu ştiu [la modul conCRET // A: tu m’as di ┴ tu m’as demandé ce QUE tu vas faire. <R tu seras le président-> et <Rmoi je t’ai dit> JE N’EN ai PAS besoin::↓ mais je n’ai pas le CHOIX en ce_mo:ment↑ ON NE PEUT PAS # laisser la fédération sans direcTI:on↑ tu sais bien qu’IL Y A des problèmes au niveau international↓ tu le sais par ouï-dire↓ mais moi je le sais [concrèTEMENT B: şi tu te <F şi tu te sacri:fici> // B: et toi tu <F et tu te sacri:fies toi > A: exact. // A: c’est exact. B: şi tu te sacrifici // B: et tu te sacrifies toi A: exA:CT. // A: c’est exA:CT. B: [şi-o vei conduce tu. // B: [et tu vas la diriger toi. A: [exA:ct. exA:CT. […] // A: [c’est exA:CT. c’est exA:CT. (Ionescu–Ruxandoiu 1999: 192)

On remarque que la composante ironique est soutenue par la désignation déictique de l’interlocuteur et bien sûr par des moyens paraverbaux (l’intonation et l’intensité forte de la voix) et très probablement extraverbaux (mimo-gestuels). Les trois derniers échanges presque identiques marquent une augmentation de l’intensité ironique jusqu’au sarcasme, des deux côtés, parce que A repousse l’attaque de l’adversaire adoptant immédiatement le même registre expressif. Certes, ce contexte illustre l’égocentrisme communicatif, mais les conflits et, plus généralement, les interactions hautement compétitives (les débats électoraux, par exemple), par leur structure discursive, génèrent une telle attitude, les protagonistes étant obligés de défendre ou de négocier leur identité de rôle/sociale (voir aussi Ionescu-Ruxăndoiu 2006: 349-353). 3.4 L’analyse du corpus semble attester la récurrence de la structure pronom sujet à la première personne du singulier associé à un verbe d’opinion (voir aussi GALR II 2005: 335). La présence du pronom a la valeur d’un signal, à l’aide duquel les locuteurs attirent l’attention soit sur l’originalité de leur point de vue, soit qu’ils s’assument la responsabilité de l’énoncé, renforçant sa force persuasive. Ex. 5: moi je (en roumain eu) + verbe d’opinion; marque d’atténuation (politesse négative) D: [<P eu cred că eu cred că Ăla e cel mai bun pentru tine↓ ăla↓> // D: [<P moi je crois que moi je crois que CELUI-LÀ est le meilleur pour toi↓ celui-là↓> (L’appareil de massage, p. 55)

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION

275

B: eu aşa să zic că s-aranjează da’ eu cred că fiind subțiri aşa s-ar putea să: // B: moi je dirais qu’ils te vont bien mais je crois moi que étant si fins il peut arriver que: […] (Chez la couturière, p. 127) B: eu↓ eu cred că […] nimeni nu poate să pună o condiție↓# în respectarea ORDInii publice în capitală […] deci↓ cred că# cred că# atunci cînd v-ați asumat aceste efecte negative eu cred că ați studiat şi legea // B: moi je↓ moi je crois que […] personne ne peut imposer une condition↓# dans le respect de L’ORdre publique dans la capitale […] alors↓ je crois que# je crois que# quand vous avez adopté ces effets négatifs moi je crois que vous avez étudié aussi la loi (Proteste estudiantin, p. 166; 168)

II: <Î eu cred că e cel mai bun ambasador pe care la numit la numit românia după decembrie o mie_nouă_sute_optzeci_şi_nouă>. // II: <Î moi je crois qu’il est le meilleur ambassadeur que la roumanie a nommé a nommé après décembre mille_neuf _ cent_quatre_vingt_neuf >. (Des journalistes en entretien avec le Président, p. 152) (cf. p. 50; 56; 57; 83; 146; 209; 237) Ce qui nous semble intéressant c’est que l’on enregistre les constructions où les formes déictiques accompagnent le verbe d’opinion non seulement dans des contextes formels, mais aussi informels; dans le corpus, les contextes où le pronom est absent sont relativement égaux statistiquement à ceux où le pronom est présent. En ce qui nous concerne, nous considérons que cette modalité d’expression est justifiée du point de vue pragmatique par le respect de la loi de modestie. L’atténuation de la force illocutionnaire de l’énoncé par le recours aux modalisateurs du type je crois, je soupçonne, à mon avis constitue une stratégie de la politesse négative (dans la conception de Brown et Levinson). Exprimer une opinion sans adoucisseurs c’est l’apanage des instances d’autorité, de tels énoncés, qui ont un fort caractère prescriptif, étant spécifique en particulier au discours juridiques, où, en outre, l’instance énonciatrice est impersonnelle. Or, bien au contraire, dans la communication courante, mais aussi dans d’autres formes discursives, y compris dans le discours scientifique, les locuteurs ressentent le besoin de renforcer leur engagement illocutionnaire, par l’auto-désignation déictique, justement pour relativiser leur point de vue et diminuer son potentiel menaçant intrinsèque. 3.5 En examinant les exemples précédents, nous avons voulu insister sur le fait que, du point de vue pragmatique, la multiplication des marques de subjectivité de l’énoncé (par l’expression du pronom sujet) ne reflète pas nécessairement un comportement auto-valorisant, étant plutôt imposée par la logique argumentative et par la nature de l’interaction. En revanche, l’escamotage de la première personne nous semble extrêmement significatif comme marque d’une attitude présomptueuse et égocentrique. Cet

LILIANA HOINARESCU

276

effacement référentiel a la valeur d’un taxème2 , à l’aide duquel l’émetteur exhibe une supériorité réelle ou prétendue. Si en roumain, dans le langage infantile, se référer a soimême à la troisième personne est un moyen fréquent d’auto-désignation «Jeanne a faim»3 , dans le cas d’un sujet adulte (s’adressant à un autre adulte) ce glissage déictique indique justement le manque de modestie et l’affectation. C’est pourquoi dans la communication courante cette manière est doublement fâcheuse: elle génère des confusions et menace, par l’asymétrie, la face positive de l’interlocuteur. Ex. 6: Débat électoral (1996). Participants: Emil Constantinescu et Ion Iliescu. Interaction télévisée; dominante nettement compétitive; auto-désignation à la troisième personne – emphase, auto-éloge dissimulé; négociation de l’identité sociale/de rôle4 a) I.I. Eu cred că tocmai aceşti ani au arătat că preşedintele Iliescu a căpătat o mare credibilitate pe arena internațională, lucru exprimat de oamenii politici responsabili. // I.I. Moi je crois que ce sont justement ces années qui ont prouvé que le président Iliescu a acquis une grande crédibilité sur la scène internationale, ce qui est aussi exprimé par les hommes politiques responsables. b) I.I. Constituția României nu a fost redactată de Iliescu, de preşedintele Iliescu. // I.I. La constitution de la Roumanie n’a pas été rédigée par Iliescu, par le président Iliescu.

Voir Kerbrat-Orecchioni 1990: 74 «Nous appelons «taxème» […] tout comportement, verbal ou non verbal, susceptible de marquer une relation hiérarchique entre les interactants […]». 3 En roumain, la situation inverse est également possible, les adultes (femmes) s’adressant souvent à leurs enfants à la troisième personne (celle-ci peut substituer la Ière personne «Maman vient immédiatement » ou la IIème personne «Jeanne a faim?»). Ce type de mimétisme langagier connote fortement l’affectivité; ainsi on y recourt fréquemment pour atténuer les énoncés directives «Maman te prie de manger». 4 Cet effet antisocial provient, le plus probablement, du fait que l’auto-désignation du locuteur par d’autres formules que le pronom personnel de Ière personne singulier est, à l’origine, un signe de royauté (sur le plan politique, mais aussi symbolique), un moyen conventionnel pour souligner le pouvoir et l’autorité [absolue] du locuteur, reconnue officiellement par les autres. D’où la forte note auto-valorisante des formules d’auto-désignation atypiques et leur caractère foncièrement impoli (du point de vue pragmatique), qui rend leur utilisation très restrictive, voire interdite dans les relations égalitaires, démocratiques (bien sûr, nous avons en vue le registre sérieux, parce qu’une utilisation ironique, citationnelle, en réplique, est toujours possible et bien naturelle). Katie Wales (1996 : 70) relève le même effet de distanciation et d’autorité lorsque l’on substitue une autre personne à la première personne du singulier en anglais: “When non-first person singular forms are used for [+ego], the ‘distancing’ that results is a social one of ‘power’, confirming the speaker’s authority: […] Her Majesty is not amused”.

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION

277

Ex. 7: Interview – interaction complémentaire télévisée; dominante coopérative; autodésignation à la troisième personne; emphase/ironie; définition de l’image publique A: bun. DOMnule tudoRAN↓ # MArele public deci NU_ŞTIe despre dumneavoastră nimic↓ VREAU să le prezentăm personajul dorin [tudoran // A: bon. MONsieur tudoRAN↓ # alors le GRrand public NE_ CONNaît rien de vous↓ JE VEUX qu’on leur présente le personnage dorin [tudoran B: [să nu vă- // B: [ne vous A: CIne SÎNteți dumneavoastră. // A: QUI Êtes vous. B: să nu vă-nchipuiți că există un singur î: persoNAJ dorin tudoran↓ // B: n’ imaginez pas qu’il existe un seul euh persoNNAGE dorin tudoran↓ (p. 233-234)

Dans l’exemple donné sous 6, le locuteur recourt stratégiquement à la substitution pronominale pour souligner de façon impersonnelle sa propre valeur; la troisième personne crée l’impression d’objectivité et lui permet non seulement de dissimuler un auto-éloge, mais aussi de rejeter une critique. Dans l’exemple 7, tiré d’une interview à la télévision, la composante auto-valorisante est plus atténuée; une utilisation ironique est également possible, le locuteur citant (mentionnant)5 la formule de présentation à la troisième personne utilisée par le modérateur. L’égocentrisme dissimulé dans ces deux exemples est pourtant motivé du point de vue situationnel: dans l’exemple 6, le locuteur est obligé de négocier son identité sociale/de rôle, dans le cadre d’un débat électoral, pendant que, dans l’exemple 7 (l’interview), B est invité à se présenter soimême et à définir son image publique. 4. Certes, nos observations n’ont pas une valeur absolue; nous avons eu seulement l’intention d’isoler quelques contextes où l’usage de la référence personnelle s’explique notamment par des critères pragmatiques. Nous nous sommes penchés surtout sur les formes plus subtiles et ambigües où le pronom sujet «je» (en roumain «eu») acquiert un rôle argumentatif ou rhétorique: signaler l’implication émotionnelle, l’empathie, l’emphase, l’ironie ou le sarcasme. Nous avons également analysé les cas d’escamotage référentiel pour la première personne, fonctionnant soit comme indicateur taxémique, soit comme signal de l’(auto)ironie.

Pour la théorie de l’ironie comme mention, voir Sperber, D. & D. Wilson 1978: 399-412; Ducrot 1984: 171-233.

LILIANA HOINARESCU

278

Bibliographie Gramatica Limbii Române II [GALR II]. Bucureşti: Editura Academiei Române. Ducrot, Oswald (1984). Esquisse d’une théorie polyphonique de l’énonciation. In: Le dire et le dit. Paris: Minuit, 171-233. Goffman, Erving (1992). La mise en scène de la vie quotidienne. La présentation de soi. Paris: Minuit. Ionescu-Ruxăndoiu, Liliana (1999): Conversația: structuri şi strategii. Sugestii pentru o pragmatică a românei vorbite. Bucureşti: Editura ALL, ediția a II-a (revăzută). Ionescu-Ruxăndoiu, Liliana (coord.) (2002). Interacțiunea verbală în limba română actuală. Corpus (selectiv). Schiță de tipologie [IVLRA I]. Bucureşti: EUB. Ionescu-Ruxăndoiu, Liliana (2003). Limbaj şi comunicare. Elemente de pragmatică lingvistică. Bucureşti: Editura All Educațional. Ionescu-Ruxăndoiu, Liliana (2006). Prezentarea şi negocierea identității în dezbaterea electorală. In: Actele Celui de-al V-lea Colocviu al Catedrei de limba română. Limba română. Aspecte sincronice şi diacronice. Bucureşti: EUB, 349-353. Kerbrat-Orecchioni, Catherine (1990). Les interactions verbales I. Paris: Armand Colin. Kerbrat-Orecchioni, Catherine (1992). Les interactions verbales II. Paris: Armand Colin. Kerbrat-Orecchioni, Catherine (1996). La conversation. Paris: Seuil. Kerbrat-Orecchioni, Catherine & Véronique Traverso (2007). Introduction. In: Kerbrat-Orecchioni, C., V. Traverso, (éds.): Confidence/Confiding. Dévoilement de soi dans l’interaction/ Self-disclosure in Interaction. Tübingen: Niemeyer, 1-24. Sperber, Dan & Deirdre Wilson (1978). Les ironies comme mentions. Poétique 36: 399-412. Vion, Robert (1992). La communication verbale. Analyse des intéractions. Paris: Hachette. Wales, Katie (1996). Personal pronouns in present-day English. Cambridge: University Press.

Normes de transcription On utilise la minuscule pour l’initiale ou pour le début de la proposition/ phrase; la minuscule est employée aussi pour les noms propres. La majuscule marque l’emphase. Les mots dans les langues étrangères sont notés conformément à l’audition, entre les symboles «/ /» (ex. /niiorc/ pour New York). Accent á on le note seulement quand il diffère de l’accent courant TEXTE accent emphatique (d’un mot ou d’une syllabe) Intonation . contour mélodique descendant terminal

Hauteur Tempo

↓ ?

contour mélodique descendant non terminal contour mélodique montant terminal

↑ ! <Î > <J > <L > <R >

contour mélodique montant non terminal intonation rejective hauteur élevée de la voix hauteur basse de la voix tempo lent tempo rapide

LE RÔLE DE LA RÉFÉRENCE PERSONNELLE DANS L’INTERACTION

Intensité _ – u: / m: ’ ς # texte = = texte [texte +A: chée texte ┴ ,, “ <@ > <Z > <OF > s <MARC > forte <IM > <CIT > (( )) () (xxx) […] {}

279

<F > intensité forte de la voix <P > intensité faible de la voix (ligne basse, entre les mots) prononciation legato (ligne de dialogue) mot inachevé allongement d’un son; le nombre de symboles (:) indique, d’une manière approximative, la durée du phénomène l’apocope des sons; l’omission habituelle de «l» article définit est marquée systématiquement, sans utiliser l’apostrophe stop glottal pause; le nombre de signes indique, d’une manière approximative, la durée de la pause enchaînement immédiat entre deux tours (latching) le commencement du chevauchement des interventions successives continuation du premier tour sans tenir compte de l’intervention chevauconstruction abandonnée (false start) séquence de discours direct rapporté rire simultané à la prononciation sourire simultané à la prononciation équence prononcée par un mouvement expiatoire spécifique au soupir séquence mise en évidence par une prononciation rare (parfois syllabée) et imitation de la manière de parler d’une personne le locuteur lit un texte indications ‘scéniques’ transcription probable texte inaudible; le nombre de symboles (x) indique, approximativement, la longueur de la séquence inaudible séquence non transcrite intervention liée au contexte de la conversation (aside)

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 281-287 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

IL DIALOGISMO NELLE POSTILLE DELL’ASCOLI CELESTINA MILANI

1. In questa ricerca si presentano alcuni risultati dell’analisi condotta su varie postille scritte da G.I. Ascoli nella sua copia dell’Archivio Glottologico Italiano (= AGI), voll. 115 (1873-1900); i volumi sono stati donati alla Biblioteca Comunale di Milano, dove tuttora si trovano. Come già ricordato (Milani 1983: 219), A. Brambilla ha riconosciuto nella scrittura delle postille la grafia stessa dell’Ascoli, cfr. Rivista di Letteratura italiana 1 (1983): 187-192. Molte di tali postille costituiscono riflessioni o aggiunte dell’Ascoli ai propri lavori, ma si trovano anche annotazioni a lavori di altri studiosi. 2. Nelle postille dell’Ascoli si verifica una situazione di dialogicità col testo di AGI, che viene chiosato nel contesto della ricerca scientifica, producendo un’interpretazione nuova del problema e della realtà dei dati, non sempre riconducibile all’insieme dei punti di vista già espressi dall’Ascoli o da altri autori. In questo tipo di dialogo il proponente è Ascoli che porta delle correzioni o dei completamenti al testo e al tema trattato. Si tratta quindi di una rielaborazione del proprio o altrui punto di vista. Le note cooperano alla ricerca, approfondendo i tratti comuni e quelli differenziali che il metodo ermeneutico propone. Le note respirano nell’esperienza storica e comparativa dell’autore secondo il principio dialettico della compensazione e del contraddittorio (Galli 1983). Esse sono inserite nel contesto degli articoli glossati dall’Ascoli, che vengono così completati e aggiornati. L’oggetto del discorso viene ripreso e convalidato o modificato. Talora si riscontrano ripetizioni lessicali e variazioni, ma generalmente si tratta di elementi nuovi. Nell’insieme si trova un orientamento discorsivo più completo. Si instaura così un discorso dialogico con linee sobrie ed essenziali (Sitri 2003; Armossy 2005: 13-54; Perrin 2006: 5). Nelle note dell’Ascoli non si trovano mai connettori argomentativi; si tratta generalmente di enunciazioni riferite al testo accanto al quale si trovano. Il dialogo è col testo, con i lettori, con se stesso. Tramite le glosse, Ascoli crea un rapporto tra parlante e ascoltatore mediato dal testo. Gli elementi del dialogo sono quindi quattro: il parlante (Ascoli), l’ascoltatore (il destinatario delle note e del testo), le note e il testo che viene così completato e corretto (Spitzer 2007: 103-254).

282

CELESTINA MILANI

3. Passiamo quindi ad analizzare alcune postille presenti nel primo numero della rivista. Per queste note relative ad AGI 1 si rimanda a Milani (1984: 271-273). In AGI 1 (1873) si notano le seguenti postille all’introduzione dell’Ascoli. A p. x “Glottologia: la questione della lingua e gli studi storici in Italia del m.e. prof. G.I. Ascoli (squarci estratti dall’autore)”. Da questa annotazione appare chiara l’intenzione dell’Ascoli di pubblicare almeno le pagine introduttive di AGI 1 o parti di esse in una Memoria o saggio autonomo (Dardano 1974). A p. XII, a proposito delle differenze che caratterizzano i vari dialetti italiani e francesi, si trova il rimando “Mém. de Ling. I 364”. Si tratta dell’articolo di A. Brachet, Dictionnaire des doublets de la langue française, suppl., Mémoires de la Société de Linguistique 1 (1874): 358 ss.; a p. 363 ss. l’autore discute del trattamento delle vocali latine nelle lingue romanze in rapporto all’accento con particolare riguardo ai dialetti francesi. A p. XIII in fondo, a proposito della lingua tedesca che appare molto unitaria diversamente dall’italiano e dal francese frammentati in molti dialetti, l’Ascoli scrive “toccare nella seconda edizione dei rimpianti di Fernow e del de Bois Reymond”. Evidentemente egli vuole accennare a Carl Ludwig Fernow, autore dei Römische Studien, 3 voll., Zurich: Gessner 1806-1808, e a Emil Heinrich Du Bois-Reymond fisiologo e letterato tedesco. In fondo alla pagina si trova anche l’annotazione “V.v. Kohl, ap. Hasdeu, Cuv. d. Bătr. III 84”, per cui cfr. Bogdan Petriceicu Hasdeu, Cuvente den Bătrăni, III Istoria limbii române, Bucuresti: Editura Cultura Nationala 1881. A p. XIV riferita alla nota nella quale l’Ascoli accenna al problema della traduzione degli idiotismi latini nel tedesco del ’700, si legge questa postilla: “V.v. Saalfeld, Über die Fremdwörter, besonders in Deutschen (Sammlung gemeinnütziger Vorträge), Prag 1882”. A p. XV, accanto alla nona riga dall’alto, si legge: “Cfr. De volg. Eloq. I, VIII, v. f.”. Tale passo tratta dell’interscambio delle lingue, cfr. Dante, De vulgari eloquentia, ediz. a cura di Pier Vincenzo Mengaldo, Padova: Antenore 1968. Tale postilla illustra il passo di AGI 1: “Dice stupendamente il Vocabolario Novo che il laboratorio in cui la natura fa le lingue, le raffina e le perfeziona non può essere che una mera agglomerazione di uomini viventi in uno scambio continuo e obbligato di pensieri e di uffici”. A p. XVIII, riferito a “ad-mactatorio” si legge l’integrazione “t-i-rat-orio”, per cui si può richiamare tiratoio entrato nella lingua italiana nel 1332 e indicante “stanzone in cui si tirano i pannilani” dal latino medievale tiratorium (Piacenza, XIII secolo) (Battisti-Alessio 1975: s.v.). A p. XX, a proposito dei proverbi dialettali indicanti “trovare il pelo nell’uovo”, troviamo annotato “prov. côrso, Tomm. 375, va sempre circhendu u pelu in l’ovo”. Si rimanda a Nicolò Tommaseo & Bernardo Bellini, Dizionario della lingua italiana III/ 2, Torino & Napoli: Unione Tipografica Editrice 1971, 870; IV/2, Pisa & Roma & Napoli, Unione Tipografica Editrice 1879, 1681.

IL DIALOGISMO NELLE POSTILLE DELL’ASCOLI

283

A p. XXI, a metà, si legge “Ognuno direbbe e di certo anche il Manzoni avrebbe ditto e scritto, con priorità etimologica e letteraria, intenzione innovatrice e non intenzione innovatora, perché a Firenze pur si dice la lavoratora, ecc.”. L’aggiunta tende a chiarire le scelte lessicali convalidate dall’uso per cui “determinarsi un piccolo vano sia modo più naturale o consentaneo che non sia il venircisi a formare un bucolino”. A p. XXVIII, circa a metà, a proposito dell’apprendimento della lingua italiana, quasi idioma straniero per coloro che sono abitualmente dialettofoni, si trova la nota “Non si nega la sincerità, ma v’è grande esagerazione, ed è, come istintivamente, un discorso disforme dal pensiero. Pare il fervore di un padre, che raccomanda al figlio una cosa buona senza volergli dire dei pregi che più egli ne ammira”. A p. XXXI, riferito ad “infilare”, si trova “s’infilarono i loro cappotti, Grandi, L’abbandono, Firenze ’78, p. 214”. L’aggiunta chiarisce l’uso del termine. Nel testo infatti si ha “L’Arte … intende il linguaggio, non come una cute che sia il portato dell’intiero organismo della vita nazionale, ma come una nuova manica da infilare (modo veneto questo …)”. A p. L, accanto alla quinta riga dall’alto, si ha “cfr. Norcia: te ico ti dico ecc. ma chè ’n dicìa, Pap. 534. E Livorno Pap. 247”. La citazione è desunta da Giovanni Papanti, I parlari italiani in Certaldo alla festa del V centenario di Messer Giovanni Boccaccio, Livorno: Tipi Francesco Vigo 1875, pp. citate. L’aggiunta è relativa alle alterazioni transitorie del suono interno di un tema o nominale o verbale, come dice l’Ascoli a p. L, e in particolare riguarda l’influsso (o no) del suono finale di una parola sul suono iniziale della parola successiva, e nel caso della postilla riguarda la conservazione o meno di /d-/. Il passo di Papanti comprende una serie di note al testo livornese che le precede e che presenta parecchi casi di “alterazioni transitorie dei suoni”. Sotto si trova questa nota di Ascoli: “ cfr. la mia vart ecc. 100-101”. Lo studioso si riferisce alla sua analisi di vart-part nel dialetto ladino dei Grigioni di Sopraselva, cfr. Saggi Ladini, AGI 1: 100-101. A p. LIV, accanto alla quindicesima riga dal fondo, si legge “E il genov. e il francese (questo dinanzi a cons.) hanno au per allo; ma l’au genovese risale ad aru laddove nel franc. si tratterà di aul (aul-tien, aul-mien)”. Per le forme genovesi cfr. Ascoli, Del posto che spetta al ligure nel sistema dei dialetti italiani, AGI 2 (1873): 122 (/l/ > /r/) ed E.G. Parodi, Studi liguri, AGI 15 (1901): 6. Quanto ad aul si tratta di ipercorrettismo grafico sorto da basi etimologiche, cfr. aulx, chevaulx, consaulx, eulx, faulx, ecc. grafie del XV secolo (Brunot 1924: 495); la forma aul è puramente grafica e ipercorretta. 4. In AGI 2 (1873) si leggono le seguenti postille all’articolo di G.I. Ascoli, Del posto che spetta al ligure nel sistema dei dialetti italiani, pp. 111-160. A proposito della continuazione di /ă/ del latino nei dialetti italiani, a p. 113 tra le note 1 e 2, l’Ascoli aggiunge: “cfr. su’ anceddu de strania, Canti delle prov. mer. p. 290” ; a p. 114, vicino all’ultima

284

CELESTINA MILANI

riga del testo, a proposito del ligure repuaero (ripar-io? Cfr. d’accord-io d’accordo) si legge “s’accordio (Arch. di Lucca 1546 in Bonari, Francesco Burlamacchi, Napoli 1874, 56 e 58) e cfr. il Diz. it. ‘accordio’ e ‘discordia’, sardo merid. accordiu = ant. ital. accordio B. Bianchi, cfr. concordio Mon. Muss. 106”. L’opera di Bianco Bianchi qui richiamata è Storia della preposizione a e de’ suoi composti nella lingua […] Saggio di un dizionario etimologico e sintattico della lingua comune e dei dialetti toscani, Firenze: Tipografia della Gazzetta d’Italia, Acc. Crusca 1877. Quanto a concordio dal lat. concordium si trova nelle Rime genovesi v. 494.58 nella frase “Chi per no vorer concordio”; quanto a “Mon. Muss.” si tratta di Adolfo Mussafia, Monumenti antichi di dialetti italiani, Sitzungsberichte der Wiener Akademie der Wissensch. Phil.-hist. Klasse 46: 113-235, ma la citazione di concordio non si trova a p. 106, bensì a p. 211, G 441 nel poemetto in volgare intitolato De Jerusalem celesti et de pulchritudine eius et beatitudine et gaudio sanctorum. Per queste note ad AGI cfr. Milani (1983: 219ss). A p. 116, a proposito del dittongo /je/ esito in antico provenzale di /ĕ/ tonico del latino, accanto alla nota 3, si legge “tiers, cfr. cierge cereus”; la prima forma è provenzale, mentre cierge è francese, fatto derivare dal Bianchi da ceriō, cfr. AGI 13 (1893): 227; REW 1829 cēreum, cfr. sic. cilio, logud. kiriu, prov. cat. ciri, ecc. Alla p. 116 è stata unita la parte finale di una lettera di Mussafia con un’aggiunta dello stesso datata al 3 ottobre 1873; si trascrive l’aggiunta del Mussafia: “a p. 116 badate che c’è un errore di stampa: un’arbitraria variante di ie. Le ultime due linee della stessa pag. fanno supporre che prov. veills si pronunci veils e prov. vielhs si pronunci viełs. In verità la differenza di pronunzia consiste solo nel dittongo; la prima forma si pronuncia vełs. Il provenz. e per ł e per ñ vacilla fra il (ill) in e lh, nh. Date un’occhiata altresì al I volume delle Mém. de la Soc. de Linguist. e troverete un bello studio del Meyer sulle formule a cui in prov. si limitano ie ed uo (ue). Se non avete il vol., posso mandarvi un sunto della dissert.”. Sul carteggio Ascoli-Mussafia, cfr. Prosdocimi (1969). A p. 119 di mano dell’Ascoli si legge “tẹrbo, torbido (turbolo?), ma ghẹmo gomito come si spiega? V. St. cr. I 24”. Come si nota, terbo e ghemo sono voci del dialetto piemontese notate con la grafia tërbo e ghëmo1 in G.I. Ascoli, Studi critici, I, Milano 1861, estratto da Studi orientali e linguistici, 3 (1861), cfr. REW 8994 e 2354. A p. 128 si trova “piem. cadlé ecc. all. al prov. cadelá ecc. ‘mettere in bell’assetto’, *capitellare (Nigra, nov. 99)”. Si ha così un evidente termine post quem per datare la postilla che chiarisce con questo esempio il problema della palatalizzazione e della conservazione di lat. /k-/ in piemontese. Nella stessa pagina, alla nota 3, si trova una postilla che corregge il genovese noria da nutrica: “difatti è noriga p. 213” come si legge nel testo n. 40, 12 a p. 213 di AGI 2, presentato da N. Lagomaggiore, Rime genovesi della fine del secolo XIII e del principio del XIV, pp. 161 ss.

Col grafema ẹ l’Ascoli indica probabilmente il suono [œ] ed è probabile che tale suono sia designato anche dal grafema ë.

IL DIALOGISMO NELLE POSTILLE DELL’ASCOLI

285

A p. 132, all’altezza della quattordicesima riga, si legge “v. Della Barba, Sul parlare dei sardi, Reggio d’Emilia, 1880”; si tratta di un’aggiunta alla nota 1 che è una rapida presentazione dei dialetti sardi. A p. 134, par. 4-12 del sardo settentrionale “vocale tonica del latino” del testo appare corretto in “appaja continuaz. di”. A p. 141, riferito al logudorese ušai = usjai da ustulare2 , si legge “-are”. Sotto si trova “cfr. χ = rc ecc. Spano ort. 28”; l’esempio maχu per maschio si trova in Giovanni Spano, Ortografia sarda, Cagliari: Reale Stamperia 1840, 28. L’esempio chiarisce logud. chišu “cerchio”, chirğu, chiššu del testo che presentano un esito diverso del nesso /-rk-/ del latino. A p. 143, a proposito dei fenomeni del sardo centrale, si legge “cfr. campid. ghiáni e logud. morighessa (morus celsa), Arch. XV 487”. Il campid. ghiáni “morello” deriva dal lat. cyaneus prestito greco, il logud. –ghessa è opposto al campid. gessa; la citazione bibliografica è un rimando a C. Nigra, Postille lessicali sarde, AGI 15 (1900), pp. 481493. A p. 144, in relazione ai fenomeni fonetici del sardo centrale, dalla riga dodicesima della nota si legge “cfr. es. chenza = *cénsa (ristud.) = senza”; chenza è forma del logudorese. Sempre a p. 144 , dalla riga ventesima della nota in poi, a proposito del sardo centrale átta = merid. azza da acies, érittu da ericius3 ecc., si legge questa postilla dell’Ascoli: “putu piatta Caix 91”. Si tratta delle forme logudoresi putu dal lat. puteus (cfr. REW 6567) e piatta dal lat. platea (cfr. REW 6583); la citazione “Caix 91” si riferisce a N. Caix, Saggio sulla storia della lingua e dei dialetti d’Italia con un’introduzione sopra l’origine delle lingue neolatine, Parma: Tipografia Pietro Grazioli 1872, 91, cfr. Santamaria (1981: 123 ss.) e Renzi (1969: 596-597). Sotto a tale glossa si trova “litos licci fariniatu corriattu”: si tratta di esempi dell’esito in dentale sorda breve o lunga del nesso latino /cj/ in logudorese; tuttavia tale esito non è chiaro nel logud. corriáttu “flessibile”, campid. korriátsu, cfr. Wagner (1960: I, 33). A p. 146 si legge “mōns tria [tempora] sunt quando post vocalem natura longam duae consonantes sequuntur vel na dulex ut ‘mōns’ ‘rēx’ (Priscian. lib. V, de syll. in fin.)”; la frase di Prisciano illustra casi di /o:/ latino in cui l’allungamento è dovuto alla posizione. A p. 147, riferito a negghia del siciliano, si trova “sarà pari al nap. neglia?”. A p. 148, accanto al par. 22 della nota che tratta dei dialetti siciliani, a proposito del sicil. vénća da vendi[c]a, perca da perti[c]a, che l’Ascoli sospetta essere voci straniere, si legge “vencia Avolio 209, miliasti Avolio 172 cfr. ca percia qui perce Avolio 155”: si tratta di voci siciliane tratte da Corrado Avolio, Canti popolari di Noto, Noto: Tip. Zammit 1875.4 A p. 151 il genov. me[ŭ]n “mani” è corretto dall’Ascoli in m[ŭ]en5 . A p. 158 si trova un foglio inserito, sempre scritto dall’Ascoli: “Come asino (pronunc. aŝgino) si pronunciano da’ Massesi (di Massa ducale o di Carrara) le seguenti parole nelle quali trovasi la s proferita come la g de’ toscani e anzi tutto la parola dell’af2 Non si trova in Wagner (1960: II, 564) ma cfr. REW 9097. 3 Cfr. Wagner (1960: I, 151): logud.. átta, campid. áttsa e REW

107; cfr. inoltre Wagner (1960: 492): logud. árittu, érittu, irittu e REW 2897. 4 Cfr. ultima edizione a cura di Antonino Buttitta, Palermo, Ed. Regione Siciliana 1974. 5 Il termine appare con la grafia moen in Casaccia 1964: 488.

286

CELESTINA MILANI

fermazione: si pron. ŝgi; avverto però che la s di si somiglia un poco all’sc toscano; ma è assolutamente sg nelle voci spasimo pron. spaŝgimo, quasi pron. quaŝgi, biasimo pron. biaŝgimo.” A p. 158 l’Ascoli tratta della realizzazione del fonema /ž/ in Liguria, Sardegna, Corsica e nella pagina successiva accenna al fonema /ʃ/ in genovese, dialetti ladini e la nota riportata si inserisce appunto nel discorso della realizzazione delle sibilanti. Qui termina per ora il discorso di analisi delle postille ascoliane. Sebbene non manchino lavori dell’Ascoli (o paragrafi di lavori) dedicati alle lingue indoeuropee negli aspetti genetici e comparativi, l’attenzione più viva delle studioso è dedicata alle lingue e ai dialetti romanzi, come appare da AGI, rivista fondata dall’Ascoli nel 1873. Il Proemio, con cui si apre AGI 1, era “uno scritto di guerra con cui l’Ascoli, respingendo le esagerazioni del manzonismo fiorentineggiante (il manzonismo degli stenterelli, diceva Carducci), riproponeva con successo e con solide argomentazioni un ritorno alla lingua letteraria italiana quale si era venuta costituendo in secoli di una fiorente letteratura italiana facente capo a Dante” (Pisani 1972: 94). Con queste parole del grande maestro Vittore Pisani, chiudo questa breve ricerca sulle glosse dell’Ascoli ad AGI. Bibliografia Armossy, Ruth (2005). The argumentative dimension of discourse. In: van Eemeren, F. H. & P. Houtlosser (eds.). Argumentation in Practice. Amsterdam-Philadelphia: Benjamins, 87-98. Battisti, Carlo & Giovanni Alessio (1975). Dizionario Etimologico italiano. Firenze: Barbera. Brunot, Ferdinand (1924). Histoire de la langue francaise. I. Paris: Colin. Casaccia, Giovanni Dizionario genovese-italiano, II. Cosenza: Casa del libro G. Brenner 1964. Dardano, Maurizio (1974). G.I. Ascoli e la questione della lingua. Roma: Istituto dell’Enciclopedia Italiana. Galli, Giuseppe (1983). Introduzione: strutture interpersonali del dialogo fecondo. In: Galli, G. (ed.). Interpretazione e dialogo. Atti del IV Colloquio sull’interpretazione (Macerata 29-30 marzo 1982). Torino: Marietti, 12-17. Milani, Celestina (1983). Postille ascoliane inedite. Quaderni di Lingue e Letterature 8: 219-227. Milani, Celestina (1984). Note su Postille ascoliane inedite. Quaderni di Lingue e Letterature 9: 271-283. Perrin, Laurent (2006). Le sens et ses voix: dialogisme et polyphonie en langue et en discours. Metz: Université Paul Verlaine-Metz. Pisani, Vittore (1972). Graziadio Isaia Ascoli linguista. In: La filosofia friulana e giuliana nel contesto della cultura italiana. Udine: Arti grafiche friulane. Prosdocimi, Aldo L. (1969). Carteggio di G.I. Ascoli ad A. Mussafia. Archivio Glottologico Italiano 54: 1-48. Renzi, Lorenzo (1969). Napoleone Caix e Ugo Angelo Canello. In: Grana, G. (a cura di), I critici. Storia monografica della critica moderna in Italia. Milano: Marzorati, I, 595-618. REW (1935) Romanisches Etymologishes Wörterbuch. W. Meyer-Lübke ed. Heidelberg : Winter.

IL DIALOGISMO NELLE POSTILLE DELL’ASCOLI

287

Santamaria, Domenico (1981). Bernardino Biondelli e la linguistica preascoliana. Roma: Cadmo. Wagner, Max Leopold (1960). Dizionario etimologico sardo. Heidelberg: Winter. Sitri, Frédérique (2003). L’objet du débat. Paris: Presses Sorbonne Nouvelle. Spitzer, Leo (2007). Lingua italiana del dialogo. Caffi, C. & C. Segre, (eds.). Milano: Il Saggiatore.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 289-303 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

“TODAY, WE’RE ALL DANES”. ARGUMENTATIVE MEANING OF THE 1ST AND 2ND PERSON PRONOUNS IN NEWSPAPER EDITORIALS ON THE MUHAMMAD CARTOONS MARTINA TEMMERMAN*

1. The use of 1st and 2nd person pronouns in argumentative discourse This paper explores the use of 1st and 2nd person pronouns in newspapers’ argumentative discourse. The pragmatic meaning of these pronouns has already been studied in numerous articles. Some investigate the relation between text and context which is generally being established by the pronoun(s) (e.g. Bazzanella 2002), others focus on the meaning in specific kinds of discourse, like business communication (e.g. Dieltjens & Heynderickx 2003), political interviews (e.g. Bull & Fetzer 2006) or academic writing (e.g. Harwood 2005, 2007). Most of these articles limit their discussions to the 1st person pronouns. In this paper we want to investigate the argumentative meaning of the pronouns. We will show that not only 1st person pronouns, but also 2nd person pronouns play an important role in the argumentative meaning of a text. We have based our analysis on a corpus of newspaper opinion articles, and as a special case, we have chosen the articles in Flemish newspapers that were written in the aftermath of the publication of some satirical cartoons depicting the prophet Muhammad in the Danish newspaper Jyllands-Posten in 2005. This case has seemed to us specially relevant, as editorialists of other newspapers, writing about the publication in the Danish newspaper, must have been forced to consider the meaning of the personal pronouns they were using, thus taking a stance in whether they wanted to identify with their Danish colleagues or not and in how far they wanted their readers to identify with their viewpoints. The language the newspaper articles in our corpus are written in, is Dutch. According to Fowler (1991: 221), in newspaper editorials, [V]various techniques are deployed to make salient the illusion of utterance by an authoritative speaker, addressing a particular kind of reader embraced in an ‘us’ relationship and taking a particular marked stance in relation to the persons and topics referred to.

This us relationship Fowler mentions is very interesting. The author will try to convince the reader of a proposition, so that both will reach a common viewpoint, a we stance. Or, as Fowler (1991: 214) states: “Readers are implicated in the ideological posi* The

author wishes to thank Marit Trioen, who has carried out part of the research for this paper. The weform this paper is written in, is therefore not to be interpreted as an authorial we.

290

MARTINA TEMMERMAN

tion of the ‘we’ to the extent that they accept these propositions.” The use of us and we here already indicates the importance of personal pronouns in argumentation. As long as the reader does not have the same viewpoint as the author, there is no we, but only an I and a you, i.e. a first and a second person. That is why we will also include the second person pronouns in our analysis. 2. The traditional approach to the meaning of 1st and 2nd person pronouns: an overview The designation 1st and 2nd person pronouns covers three strongly related subclasses of pronouns: personal pronouns, possessive pronouns and reflexive pronouns. Their interrelatedness results from the fact that the grammatical categories of person, number and gender apply to the three of them: they all distinguish between 1st, 2nd and 3rd person1 ; between singular versus plural; and between personal versus non-personal gender, with a further distinction between masculine and feminine within the former. As a consequence, grammars often discuss these three subclasses under the same heading, presenting the personal pronouns as the prototypical class, which then has corresponding possessive and reflexive forms (see for instance Biber et al. 1999; Leech & Svartvik 2002). Table I offers an overview of Dutch 1st and 2nd person pronouns (the English forms are between brackets): Person

Personal pronoun Possessive Reflexive pronoun Subject Object Determiner Pronoun 1st Sg. ik (I) mij/me (me) mijn (my) mijne (mine) mijzelf/mezelf (myself) 1st Pl. wij/we (we) ons (us) ons/onze onze (ours) onszelf (ourselves) (our) 2nd Sg. jij/je/u (you) jou/je/u jouw/je/uw jouwe/uwe jezelf/uzelf (yourself) (you) (your) (yours) 2nd Pl. jullie (you) jullie/u je/jullie/uw uwe (yours) jullie zelf/uzelf(yourselves) (you) (your)

Table I: 1st and 2nd person pronouns in Dutch and English (after Biber et al. 1999: 328)2

As Wales (1996: 12) points out, tables “listing the different persons and cases” clearly dominate the overviews of personal pronouns, thus “reflecting the continuing preoccupation of grammarians with form”. Indeed, when it comes to a discussion of the meaning of 1st and 2nd person pronouns, grammars (even if they are called communicative grammars) remain rather silent. Haeseryn et al. (1997), which is a reference work for Dutch grammar, describes the primary meaning of the first person pronoun as referring to the speaker(s) and those s/he-they is-are related to in one way or another, and gives the same definition for the meaning of the second person pronoun but then referring to the hearer(s) (Haeseryn et al. 1997: 231ff – translation ours). 1

Hence the ambiguity of the term personal pronoun: both possessive and reflexive pronouns are also ‘personal’, taking their orientation from the speaker’s perspective (1st, 2nd and 3rd party). 2 Unlike English, Dutch distinguishes between a formal (u, uw) and an informal form (je, jou(w), jullie) of the 2nd person pronoun.

“TODAY WE’RE ALL DANES”

291

Nevertheless, various corpus-based linguistic studies have shown that 1st and 2nd person pronouns have strongly divergent uses which go way beyond their primary meanings of referring to speaker(s) and/or hearer(s). Particularly those studies analyzing the use of 1st and 2nd person pronouns in persuasive discourses have paid a considerable amount of attention to the variety of meaning of the 1st and 2nd person pronouns. For instance, different analyses of the use of we in academic discourse (Harwood 2005, 2007); in political discourse (Wilson 1990; De Fina 1995; Bull & Fetzer 2006); and in business discourse (Dieltjens & Heynderickx 2003; Van de Mieroop 2006) have pointed out that 1st and 2nd person pronouns are […] open symbolically to rhetorical manipulation and negotiation of meaning in different kinds of discourse, in the pragmatic interests of power, and subjectivity (speaker orientation), or modesty, empathy and politeness (addressee orientation). Just as politics is a question of pronouns (Middleton 1993: 118), so pronouns are a question of politics. (Wales 1996: 84)

In the following sections, we will give an overview of the secondary meanings of the pronouns. 2.1 Secondary uses of the 1st person pronoun – plural form In most grammars, attention is paid to the atypical uses of the 1st person plural we. It is then argued that, in comparison with the 2nd person singular, the use of we seems far more complicated, and that this extensive polysemy is to be explained by its referential ambiguity, since we embraces not only the three person categories (Wales 1996: 63), but also all possible combinations of person categories. Despite their general acknowledgment of the broad referential potential the pronoun we has, the different authors diverge strongly in their respective approaches to these secondary uses, particularly with respect to the amount and classification of the mentioned uses. Biber et al. (1999: 329-30), for instance, solely mention the use of we “in academic prose, where we may refer to a single author [often called authorial we (M.T.), ex. 1], a group of authors, to the author and the reader [often called inclusive we (M.T.), ex. 2], or to people in general [often called generic we (M.T.)]”. Leech & Svartvik (2002) discuss two secondary uses of we: the we in “formal writing”, exclusively referring to the writer (ex. 1); and a “playful, condescending use of we, referring to the hearer” (ex. 3): (1)We will explain this theory in the next chapter. (2)As we saw in Chapter 3, … (3)How are we feeling today? [doctor to patient]

Quirk et al. (1985: 350-51) overview finally lists five secondary uses. Their “inclusive authorial” (ex. 2) is to be found in “serious writing”, where it involves the reader “in a joint enterprise”. From their point of view, this inclusive authorial is not to be confused

292

MARTINA TEMMERMAN

with what they call an “editorial we” (ex. 1), the latter being used in scientific writing “in order to avoid an egoistical I”. The “rhetorical we” (ex. 4) comprises a “specific use of generic we” and is said to refer “to the nation”. Quirk et al. also mention “condescending we” in doctor-patient or teacher-student relations (ex. 3). The fifth and final use Quirk et al. distinguish is a we with reference to “a third party”(ex. 5): (4)We have to be proud of this country. (5)We don’t look happy today. [secretary about her boss]

Although this rudimentary overview contains excerpts from only three grammars, it gives an indicative impression of the traditional treatment of the secondary meanings of we. In this concern, particularly the two following observations deserve attention. First, the lack of systematicity and coherence is to be pointed out, both within one single approach and across the various authors. This problem is partly explained by the lack of a consistent criterion for classification. For instance, at least three different perspectives play a role in Quirk et al.’s overview of the secondary uses: the discursive context, the intended referent and the rhetorical function (which explains the difference between their inclusive authorial and editorial we). Secondly, these overviews strongly tend to be restricted to mere enumerations, in which the particular uses are approached as random, individual cases, “even oddities” (Wales 1996: 63). The actual meaning of the particular instances is generally discussed only in relation to the specific context in which they occur. As a consequence, no attention is paid to “general principles at work” behind the various atypical instances (Wales 1996: 63). 2.2 Secondary uses of the 2nd person pronoun – singular form An examination of the different grammars mentioned above leaves us with one secondary meaning of the 2nd person singular you3 . Besides its (standard) referential use, you may also be used generically. It then becomes an impersonal pronoun which refers to people in general, as in: (6) You never know what the future will bring.

In such instances where structural knowledge and general truth are expressed, the personal overtone of you disappears. 3. An answer to the traditional approach to the 1st and 2nd person pronouns: the pragmatic approach If we want to use the analysis of 1st and 2nd person pronouns in order to say something about argumentative techniques in newspaper editorials, a consistent and systematic approach to these pronouns is an absolute condition. Since various empirical studies

3 In Dutch, this use is restricted to the reduced form je.

“TODAY WE’RE ALL DANES”

293

that approach personal pronouns from a rhetoric point of view have shown that 1st and 2nd person pronouns, when not used in their prototypical ways, indeed do more than simply refer to the specific individuals for which they stand (Spiegelberg 1973: 132), and that, moreover, the particular secondary uses show recurring patterns of meaning which transcend the individual uses, this section is devoted to a systematic mapping of the pragmatic and argumentative functions of we and you. An analysis of personal pronouns from this point of view takes as its starting point an approach to personal pronouns as strategic devices rather than as purely “economy devices” (Biber et al. 1999: 327): a speaker does not only select a particular pronoun in order to refer to a specific (group of ) person(s), but his/her choice may also be motivated by certain communicative aims. On the basis of a thorough examination of the various secondary uses of the pronouns we and you as described both in traditional grammars and in empirical linguistic studies which pay attention to the occurrence of personal pronouns as rhetorical devices, four general rhetoric functions can be distinguished. In what follows, we will discuss these functions in more detail. 3.1 The establishment of solidarity A first communicative aim the speaker may have by using the 1st person plural is to express ingroup solidarity. This particular rhetorical function is the one most closely related to the primary function of we, i.e. referring to “a group of people, including the speaker” (Leech & Svartvik 2002: 57 – our emphasis). Also in its prototypical use, the 1st person plural characteristically implies a certain degree of relatedness (Haeseryn et al. 1997): in using we, the speaker inherently sets up a connection between himself and the person(s) he involves in the reference. Nevertheless, the rhetorical function of establishing solidarity goes further than the simple declaration of a link between two or more people. The use of we enables the speaker to define social groups: not only does the speaker state membership, and thus emotional and/or social connection to the group, but s/he also has the power to decide who else belongs to this ingroup, and, with the same token, who doesn’t (Helmbrecht 2002: 42). As Wilson (1990: 58, 76) argues, such ‘otherization’-strategies are inextricably related to the establishment of a communal identity: a definition of one’s supporters implies an identification of one’s enemies – us being “basically everyone but them”. Moreover, the use of the 1st person plural enables the speaker to designate group indexicality without the need for a confirmation of this “social bonding” (Mühlhäusler & Harré, 1990: 174) on the hearer’s side. Hence, Pennycook (1994: 176 – our emphasis) concludes that this particular rhetorical function is in essence related to a “covert assumption of shared communality”. 3.2 The mitigation of face threatening acts A second argumentative function of we is rooted in the theory of face threatening acts as developed by Brown & Levinson (1987). According to these authors, speakers con-

294

MARTINA TEMMERMAN

stantly negotiate face in linguistic interaction, with the maintenance of face being the main concern (Brown & Levinson 1987: 61). Their notion of face, derived from that of Goffman (1981), is to be interpreted as “the public self-image every member wants to claim for himself ” and is further developed in the notions of positive and negative face (Brown & Levinson 1987: 61-2). Every person wants to claim for him/herself two related things: first, a “positive consistent self-image”, and secondly, a “freedom of action and freedom for imposition” (Brown & Levinson 1987: 61). However, in everyday interaction there will always be “face threatening acts”: acts that intrinsically “run contrary to the face wants of the addressee and/or of the speaker” (Brown & Levinson 1987: 65). For instance, orders and requests typically threaten the freedom of imposition on the side of the hearer. In such face threatening situations, the speaker needs to employ linguistic strategies in order to avoid, or at least minimize, the potential face threat (Brown & Levinson 1987: 68). One of these linguistic strategies is the use of 1st and 2nd person pronouns. Let’s constructions are a typical example of such a strategy. Consider for instance the following request: (7) Let’s not digress. [chairperson in meeting] (Wales 1996: 67)

Here, the speaker includes him/herself in the request in order to minimize the threat to the hearer’s negative face which the request may cause. In this sentence, the speaker wants the hearer to perform an action that is mainly in the speaker’s own interest. But in using the 1st person plural instead of the 2nd person, a “surface meaning of modesty” or “joint activity” is generated, and the authoritative voice of the speaker is covered (Wales 1996: 66). This use of the 1st person plural in order to disguise a direct instigation makes the request more acceptable. A comparable use is the use of generic we or you in order to mitigate a face threatening act. In kindergarten, children might be taught: (8) We don’t eat meat with our fingers.

By using the we form, the distance between speaker and hearer is reduced and the utterance does not take the form of a direct order. Generic you can be used in the same way: (9) You don’t eat meat with your fingers.

This use seems to be more face threatening than the one in example (8), but less than a direct order. However, drawing up a scale of the degree of face threat would lead us too far for the scope of this paper. 3.3 The evasion of personal responsibility In persuasive discourses, both inclusive and generic we as well as generic you are often used in order to downgrade authorial responsibility (see for instance Harwood 2005; De Fina 1995; Wilson 1990). In such linguistic situations, the pronominal system is used in order to distribute, if not to project personal responsibility “from the I to the we” (Wilson 1990: 58), or from the I to humanity in the case of indefinite reference. A

“TODAY WE’RE ALL DANES”

295

prototypical example of this particular rhetoric function of we is the ‘academic we’, where “responsibility for what is said, suspected and so forth is diminished” (Mühlhäusler & Harré 1990: 175). 3.4 The enhancement of persuasion In this last strategic use of the pronominal system, the speaker claims the hearer as an associate in order to enhance the persuasive power of the argument. This effect can be elaborated both by we and by you. Following Spiegelberg (1973: 131), the intended effect of an ‘editorial we’ for instance is – amongst others – to “overwhelm one’s audience by persuasion, not giving it even a chance to test, accept, or reject one’s opinions”. The speaker thus involves the audience in the elaboration of a personal, subjective argument in order to enhance the persuasiveness of the assertion. A similar strategy, though built up in a different way, is to be found in the use of generic you or we. In using an indefinite agent, a maximum degree of objectivity and generalization is generated, again in order to conceal the subjective tone in the argument. In both substrategies, the speaker not only downgrades his/her own presence, but moreover elaborates the argument on behalf of the audience, and the rhetorical implication of this strategy is that the audience must share the speaker’s view as being the only correct and possible view (Wales 1996). In doing so, the weight of the argument is enhanced to a maximum degree – in the end even to a degree of universality. 4. Presentation of the data We have collected 25 newspaper editorials which were published between February 1 and February 15, 2006, in all Flemish newspapers. The editorials all deal with the publication of a number of satirical cartoons, some of them depicting the prophet Muhammad, in the Danish newspaper Jyllands-Posten on September 30, 2005 and the worldwide reactions to this publication in the beginning of 2006. Newspapers all over the world had to decide for or against (re)publication of the cartoons, and journalists were forced to take up an explicit standpoint in the debate. Due to the highly explosive context, these viewpoints had to be firmly substantiated and communicated to the readers. This makes the corpus we have collected very suitable for investigating argumentation. We would also expect argumentation on freedom of speech and freedom of press from an authoritative stance: journalists, being the executors of the rights of freedom of press, might be expected to defend this right in the case of the publication of satirical cartoons. Table II gives an overview of the number of pronouns in the corpus. All personal, possessive and reflexive pronouns have been included in the count.

296

MARTINA TEMMERMAN

Pronouns Newspaper (# articles)

1st person

2nd person

Total

Average

De Morgen (7)

De Standaard (5)

Het Laatste Nieuws (3) De Tijd (3)

56 4

9 1

65 5

21 1

Het Nieuwsblad (2)

Gazet van Antwerpen (2)

Belang van Limburg (2)

Het Volk (1)

Total

195

238

Table II: Number of 1st and 2nd person pronouns in the corpus

A first point to be made is that pronouns of the 1st person occur a lot more often than those of the 2nd person. In the category of the 1st person pronouns, the plural forms occur more often than the singular forms. The 2nd person pronoun, on the other hand, is always used in the singular form. 4.1 Pragmatic meaning of the pronouns in the corpus In order to find out the rhetorical functions of the 1st and 2nd person pronouns in the corpus, we always have to take the discursive context and the intended referent(s) into account. 4.1.1 Pragmatic meaning of the 1st person pronouns: singular form Out of the 195 occurrences of 1st person pronouns in the corpus, only 34 are in the singular form. Mostly, the 1st person pronoun is referring to the author of the article, as in the following examples: (10) Ik weiger om te geloven dat moslims anders zijn4. I refuse to believe that Muslims are different. (11) Ik sprak met imams, leden van de moskeeĂŤn en verenigingen. I talked to imams, members of mosques and associations.

The author explicitly formulates a personal conviction here (ex. 10), or describes a personal anecdote (ex. 11) (which is a well-known rhetorical technique). In one instance 4 All examples are quoted in their original Dutch form and translated into English as closely as possible.

“TODAY WE’RE ALL DANES”

297

(ex. 12), the author gives a free translation of a quote which is attributed to Voltaire, in order to make his own standpoint clear: (12) “Ik verafschuw uw mening, maar ik zou mijn leven geven opdat u ze zou kunnen blijven uiten.” “I detest your opinion, but I would give my life to make it possible for you to continue to express it”.

Citing somebody else’s words is a rhetorical technique, which we can consider as a case of shift in footing (Goffman 1981). Conversational footing is the degree of commitment the speaker expresses in connection with the message. By adopting the utterance of somebody else, the speaker makes a statement without taking the full responsibility for it. Still, in example (12) the I in the quote refers to Voltaire, but at the same time it stands for the journalist writing the article. 4.1.2 Pragmatic meaning of the 1st person pronouns: plural form For the pronouns in the plural form, we find the same meanings as in examples 10 and 11, i.e. the author expressing a personal conviction/hope (ex. 13) or describing a personal anecdote (ex. 14): (13) We hopen dat het bij woorden blijft. We hope that it won’t go further than words. (14) Het was een mopje waar we als communicantjes om giechelden tijdens de mis. It was a joke we giggled about during mass when making our first communion.

Here the pronoun only refers to the author him/herself and possibly to his or her personal sphere. We have found that the meaning of the 1st person plural form can also refer to a number of groups the author belongs to. The use of the 1st person plural here is the so-called exclusive use (the reader is not included in the reference). This use applies to the previous two examples and also to examples where the author refers to (the editors of ) the newspaper, as in: (15) Om duidelijk te maken hoe belangrijk de redactie van deze krant die waarden vindt, hebben we besloten om een aantal van de gewraakte cartoons vandaag opnieuw af te drukken. In order to make clear how important the editors of this newspaper find those values, we have decided to reprint a number of the contested cartoons today.

The author uses an argument of authority here. The newspaper has the capacity of reproducing the cartoons and spreading them to a larger audience, and it does so, in order to make a statement. In a few cases, we refers to a group the reader might or might not be included in. The first is the group of women in general. A female journalist writes mockingly:

298

MARTINA TEMMERMAN

(16) Zowel aan de top van de bedrijfswereld als in de academische wereld zijn wij ruim vertegenwoordigd […] We are amply represented at the top of the business world as well as in the academic world.

The author states ironically that a society which treats its women as equals, can boast of a certain superiority. In other cases, we refers to groups to which both author and reader belong. Here the 1st person pronoun is used inclusively (cf. section 2.1). Examples are: (17) Wij begrijpen de Franstaligen niet, zoals we de moslims niet begrijpen. We don’t understand the French-speaking, just like we don’t understand the Muslims. (18) Want eerst mag de profeet niet meer afgebeeld worden, daarna mogen we niet meer lachen met prins Filip. Because first the prophet cannot be depicted anymore, next thing, we’re not allowed to laugh with prince Filip anymore.

These examples are rather specific for the Belgian situation, Belgium being a country which is split up in three linguistic communities, and where the general feeling of alienation and bad understanding between the Dutch-speaking and the French-speaking communities increases. What still unites the country is the monarchy, but as example (18) shows, for a number of Belgians, the royal family is an object of derision. The author tries to convince the reader of his/her standpoint by dragging the reader in into a certain group, as in the following examples: (19) Als het aankomt op de verdediging van de vrijheid van mening en de vrijheid van drukpers, zijn wij vandaag allemaal Denen. If it comes to defending freedom of speech and freedom of press, today, we’re all Danes. (20) Gevoeligheden waar wij al lang niet meer mee zitten, zijn nog intact in de moslimwereld. Sensitivities we don’t know anymore for a long time, still exist in the Muslim world. (21) Vandaar ook dat wij de hele polemiek in het westen nogal over het paard getild vinden. That is why we in the West think the whole polemic goes over the top a little.

In these examples, we refers to the Western world, as opposed to the Muslim world. Here the authors revert to the us/them thinking which was also reported by Fowler (1991). Fowler has analyzed editorials in British newspapers after a bombardment in Libya and subsequent attacks and he found that authors often used the us/them contrast, where we (the British people) were represented as being completely different from and opposed to them (the terrorists).

“TODAY WE’RE ALL DANES”

299

Though there are at least 350 000 Muslim people in Belgium5 , the editorialists never seem to think of their readers as possibly being Muslims. There are only two instances in the corpus where a 1st person pronoun refers to Muslim referents: (22) “Het is alsof ik […] je vader en je moeder een klap zou verkopen […]. Zo diep zit het bij ons.” “It’s as if I […] would deliver your father and your mother a blow […]. That’s how deep it goes for us.” (23) In Vlaanderen […] stelt niemand de vrije meningsuiting of de persvrijheid in vraag. Ook onze moslims niet. In Flanders, nobody questions freedom of speech or freedom of press. Neither our Muslims.

In the first example, the author quotes a young Muslim explaining what impact the cartoons had on Muslim readers. In the second example, the Muslims living in Flanders are called our Muslims. They are being represented as belonging to the Flemish society, and thus being opposed to them, the extremist Muslims. Both uses imply that the reader cannot be Muslim. The Muslim world is being represented as completely separated from the group of readers of the newspaper. A last type of instances are those uses where we refers to mankind in general. This is the generic use of we, but still in these examples two camps are involved. The Western world and the Muslim world are summoned to work together, as two separate entities: (24) Dat we hier zonder dialoog niet uit zullen geraken, is duidelijk. It is clear that we won’t get out of this without a dialogue. (25) Een toekomst die we hoe dan ook samen zullen moeten opbouwen. A future we will have to build up together anyhow.

4.1.3 Pragmatic meaning of the 2nd person pronouns All 2nd person pronouns are used in the singular form. It is interesting to note that when the authors address their readers directly, they address them as individuals, not as groups. In most cases, the use of the 2nd person pronoun is rather straightforward, as in the following examples: (26) Uiteraard kent u als verstandige lezer het antwoord op deze vraag. Of course you, as an intelligent reader, know the answer to this question. (27) “Het is alsof ik in uw huis zou binnenstappen zonder te bellen.” “It is as if I would enter your home without ringing the doorbell.” (28) Er waren net die bloedige aanslagen in Londen geweest, je wist dus maar nooit. There had just been those gory attacks in London, so you never knew. 5 Interdisciplinary Centre for the Study of Religion and World View. “Islam in Belgium”,

http://www.kuleuven.be/icrid/religies/wr_nl_islam.htm#moslimsbelgie (accessed August 24, 2008).

300

MARTINA TEMMERMAN

In example (26), the author addresses the reader directly. Moreover, this is done in a flattering way, which is a frequently used rhetorical technique (the so-called captatio benevolentiae in ancient rhetorics). In example (27), the author is quoting a Muslim friend, who describes the impact the cartoons had on Muslim readers. This again can be considered as a case of shift in footing (Goffman 1981, see also section 4.1.1 above). Example (28) is an instance of the most frequent use of the 2nd person in the corpus, i.e. the generic or impersonal use (cf. section 2). A few cases of the generic use of the 2nd person pronoun deserve special interest. Examples are the following: (29) Je kunt ook in de tegenaanval gaan en tot je eigen stomme verbazing merken dat je de dreiging de baas kunt. You can also launch a counterattack and notice – to your own astonishment – that you can handle the threat. (30) Respect toon je niet door te vuur en te zwaard je eigen gelijk af te dwingen. You don’t show respect by exacting your own right by fire and sword.

In both examples, the 2nd person pronoun does not refer to anyone in general, but to a specific group. Example (29) describes a debate in the Flemish parliament on the publication of the cartoons and how the democratic parties were perceived to win the debate from the Flemish extreme right party. Je in the example refers to the Flemish democratic parties (a we-stance). Also in example (30), je refers to a very specific group, i.e. the extremist Muslims. We have found two instances of this use in the corpus, and both contain a negation. The generic phrase here defines what the others should not do. We can conclude that the generic form of the 2nd person is used here in order to mitigate the us/them contrast (cf. Fowler 1991). This is a use we haven’t found reported in the grammars and literature discussed in section 2. 5. Discussion of the data and conclusions The following general conclusions can be drawn from the use of the 1st and 2nd person pronouns in the newspapers’ editorials in our corpus. First of all, it should be noted that it is possible to write an argumentative editorial without any pronouns of the 1st or 2nd person. In our corpus we found 3 out of 25 articles like this. This may allow us to say that, when pronouns of the 1st and 2nd person are used in newspaper editorials, this is done in a purposeful or even deliberate way. We also have to remark that we always need the context in order to pin down the exact meaning of the pronoun. Pronoun and context together can form a persuasive unit. Apart from the cases where we found the pronouns in well-known rhetorical usage like recalling a personal anecdote, uttering a personal conviction or quoting another source, we also found a number of instances where we can apply the general rhetorical functions of the pronouns described in section 3. The evasion of personal responsibility (3.3) is a strategy which is less at stake here, but the other strategies do apply.

“TODAY WE’RE ALL DANES”

301

5.1 Establishment of solidarity Using pronouns for the establishment of solidarity is a technique which is applicable to the 1st person pronouns in the plural (“we, ons, onze”/ we, us, our) (cf. section 3.1). In our corpus, we have found a number of instances where we refers to groups the author and the reader belong to and which are being represented as being opposed to the Muslim world (Flanders, Belgium, the Western world). In the cases where we refers to the Western world as well as the Muslim world, the two entities are still represented as being separate, while having to cooperate (e.g. ex. 24). So a first strong argument which is being made by using we in the corpus, is that authors and readers belong to the same group, a group which must stand strong against an ‘otherized’ (cf. section 3.1) Muslim group. 5.2 Presenting the argument as a general truth The inclusive and or generic use of the 1st person plural and the generic use of the 2nd person singular allow the speaker to present an argument as a general truth. The techniques we discussed in sections 3.2 and 3.4 apply here. Inclusive we generates a maximum degree of objectivity. The rhetorical implication of this strategy is that the audience must share the speaker’s view as being the only correct and possible view (Wales 1996: 66). Whether the group of women, Flemish people, Belgian people or the Western World is referred to, the reader will identify him/herself as belonging to the group, and thus there is a chance that s/he will go along in the argumentation of the author. This also holds for the use of generic you, but here in a few instances we have seen that generic you was used to mitigate the us/them distinction (cf. section 4.1.3). The opposition of the Western world versus the Muslim world is again stated, but not in an explicit way (cf. ex. 29-30). 5.3 Authoritative argument The argument we would have expected to be made by the use of the 1st person plural, i.e. the authoritative stance of “we as newspapers and defenders of the free press” is an argument we did not encounter very often in our corpus. Ex. 15 is one of the few instances where the author makes a statement referring to the freedom of press. But on the whole, we must conclude that the Flemish newspapers in the period we have studied considered themselves to be the voice of the Western world and the defenders of Western values as opposed to the extremist Muslim world and values. Everywhere in the editorials, the us/them distinction is present, sometimes in an explicit, sometimes in a more blurred way. The main value that is being defended is freedom of speech and there is very little reference to freedom of press.

302

MARTINA TEMMERMAN

References Bazzanella, Carla (2002). The significance of context in comprehension. The ‘we case’. In: Edmonds, Bruce & Varol Akman (eds.). Context in context 7: 239-254. Biber, Douglas et al. (1999). Longman grammar of spoken and written English. London: Longman. Brown, Penelope & Stephen Levinson (1987). Politeness. Some universals in language usage. Cambridge: Cambridge University Press. Bull, Peter & Anita Fetzer (2006). Who are we and who are you? The strategic use of forms of address in political interviews. Text and Talk 26 (1): 1-35. De Fina, Anna (1995). Pronominal choice, identity, and solidarity in political discourse. Text 15 (3): 379-410. Dieltjens, Sylvain & Priscilla Heynderickx (2003). The indefinite “We” (het “wij”-gevoel/ le “nous” indéfini): sender and receiver references in top-down communication: a text type-based approach. Journal of technical writing and communication 33 (1): 3-27. Fowler, Roger (1991). Language in the news. Discourse and ideology in the press. London: Routledge. Goffman, Erving (1981). Forms of talk. Oxford: Blackwell. Haeseryn, Walter et al. (1997). Algemene Nederlandse Spraakkunst. Groningen: WoltersNoordhoff. Harwood, Nigel (2005). We do not seem to have a theory… The theory I present here attempts to fill this gap: inclusive and exclusive pronouns in academic writing. Applied Linguistics 26 (3): 343-375. Harwood, Nigel (2007). Political scientists on the functions of personal pronouns in their writing: an interview-based study of ‘I’ and ‘we’. Text and Talk 27 (1): 27-54. Helmbrecht, Johannes (2002). Grammar and function of we. In: Duszak, A. (ed.). Us and Others. Amsterdam: John Benjamins, 31-49. Huddleston, R. et al. (2002). The Cambridge grammar of the English language. Cambridge: Cambridge University Press. Kitagawa, Chisato & Adrienne Lehrer (1990). Impersonal use of personal pronouns. Journal of pragmatics 14: 739-759. Leech, Geoffrey & Jan Svartvik (2002). A communicative grammar of English. Third edition. Pennycook, Alastair (1994). The Cultural Politics of English as an International Language. London: Longman. Middleton, Peter (1993). Who am I to speak? The politics of subjectivity in recent British poetry. In: Barry, Peter & Robert Hampson (eds.). New British poetries. Manchester: Manchester University Press, 107-133. Mühlhäusler, Peter & Rom Harré (1990). Pronouns and people: the linguistic construction of social and personal identity. Oxford: Blackwell. Quirk, Randolph et al. (1985). A comprehensive grammar of the English language. 16th edition. London: Longman. Spiegelberg, Herbert (1973). On the right to say ‘we’: A linguistic and phenomenological analysis. In: Psathas, George (ed.). Phenomenological Sociology. New York: Wiley & Sons, 129-156.

“TODAY WE’RE ALL DANES”

303

Van de Mieroop, Dorien (2006). Identity construction in institutional speeches: the crucial role of pronouns. Lodz papers in pragmatics 2: 81-103. Wales, Katie (1996). Personal pronouns in present-day English. Cambridge: Cambridge University Press. Wilson, John (1990). Politically Speaking. The pragmatic analysis of political language. Oxford: Basil Blackwell.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 305-315 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

D IALOGICAL OCCURRENCES OF BEM IN E UROPEAN P ORTUGUESE. ENUNCIATIVE STABILITY AND DEFORMABILITY HELENA VIRGÍNIA TOPA VALENTIM

In this paper I intend to describe the plurality of values that the linguistic unit bem can assume in dialogical contexts in European Portuguese. Concerning the problem of the relation between linguistic signification and the construction of referential values, it is evident that there is a relevant dependence between the variability of sense affected to a unit and the linguistic and discourse context. The linguistic units change its meaning depending on the linguistic sequence in which they occur. Sometimes the modification of a single element of the linguistic context is enough to change, radically or imperceptibly, the semantic value assigned by a linguistic unit. For instance, in Portuguese, the order in which bem occurs in the utterances eu vi-te bem and eu bem te vi is a condition not only to its semantic values but also to each utterance’s meaning as a whole. In addition to this reciprocal dependence between any linguistic units that integrates the utterance and its linguistic context, each utterance acquires a meaning through its construction in reference to a situation, namely to a situational context. In other words, each utterance’s meaning is generated through those contextual features which are pertinent to its interpretation and which can be verbally explicit or not. Simultaneously, each utterance determines its occurrence’s situational context. We can consider, as in Franckel’s suggested formulation (see, for instance 2006), that the situational context isn’t defined only exteriorly in considering the utterance (as we notice in the pragmatic theoretical proposals); the situational context is the very condition itself for the interpretation of the utterance. Therefore, the situational context is also a construction. This fact seems evident given that each utterance is absolutely singular and strictly connected to a prosodic expression and to a certain contextualization. Besides, before the necessity to make explicit the difference between two sequences with an apparently close sense (A and B), it is common to resort to a contextual contrast (“we say A when… and B when…”). Sometimes this contextualization is even essential to give an account on the intrinsic difference of sense. Returning back to both sequences eu vi-te bem and eu bem te vi, we can say that the speaker produces an utterance as eu vi-te bem when he intends to signify the way in which he views the interlocutor, “clearly”, “well seen” (even if, with a particular prosodic modulation, it may include a rectifying value). The speaker produces an utterance as eu bem te vi to introduce an adversative, a contrary sense, or, with a strongly marked prosody, for intensifying the linguistic event, the “seeing you”.

306

HELENA VIRGÍNIA TOPA VALENTIM

In its most frequent occurrences, bem expresses a circumstance of mood. In the sequences 1 and 2 it functions exactly as a modal adverb1 . 1. A Sara expressou-se bem (Sara expressed herself well) 2. Eles fizeram bem os trabalhos de casa (They did their homework well)

Yet, the modal adverb’s syntactic behaviour in each of these sequences is different. In 1, bem corresponds to a phrasal constituent. It is included in the phrase’s structure. As a phrasal complement – thus, non-optional – it takes part in the expressed propositional content. The impossibility of the modal adverb suppression is a good test to prove that bem is subcategorized by the verb2 . In the sequence 2, the suppression of bem doesn’t bring about any un-grammaticalness. The adverb has an optional nature and, thus, it corresponds to a VP adjunction. In a semantic perspective, bem can be paraphrased in both utterances by “in an appropriate way”, “in a satisfactory way”, or even, “in an expected way”. These glosses allow us to identify a particular “enunciative” value in terms of modality, more specifically, an appreciative value3. The appreciative value is a conformity value: the specified value is considered as a speaker’s desirable value, and it can eventually correspond to an expected value. Thus, considering the theoretical presupposition that every relation is a complex relation, we can localize the linguistic occurrence as constructed in relation to another term – a pre-constructed one, a term which corresponds with a subjective origin. There is, therefore, a comparison of an occurrence in relation to the previously and subjectively constructed reference. The occurrence pi is not primary distinguishable from the other occurrences which set up P. But, because an occurrence is constructed in reference to an enunciative situation, pi is located in relation to the referential situation, namely, to a subject’s “previously fixed standard”. This relation is an identification. There is an identification or a conformity between the occurrence pi and the previously subjective construction as a reference point. So, the values affecting bem are the result of some operations, bem being its linguistic marker. When bem has an appreciative value, it marks two metalinguistic operations: the course through the different occurrences p, p’ and the identification of pi in relation to p (see Péroz, 1992: 40ss). In the next examples, bem is exchangeable with muito and hasn’t an appreciative value. 1 See Costa (2008), about the adverb in European Portuguese. 2 In Lopes (2004), some other syntactical tests are proposed,

confirming the functioning of bem in sentences as 1 and 2. 3 It was Antoine Culioli who proposed the classification of this kind of modality as an appreciative one (“modalité appréciative”, 1968: 112). It is characterized by the construction of a “[...] um juízo de valor, [...] uma apreciação sobre uma relação predicativa já constituída e validada (ou validável)” (Campos & Xavier 1991: 341).

DIALOGICAL OCCURRENCES OF BEM IN EUROPEAN PORTUGUESE

307

3. Ele é bem alto (He is very tall) 4. Eles são bem capazes de faltar (They are really capable of missing the lesson)

It functions as an intensifier, as the traditionally classified quantifier of degree (see Lopes, 2004: 439). Consequently, the sequences modified by bem with an intensive value are interpreted in terms of a scalar propriety. Therefore, bem only assumes an intensifying value when co-occurring with qualifying adjectives, and not with relational adjectives (*Este sistema é bem digital – “This system is very digital”). When bem modifies a deverbal adjective, it carries an appreciative value (Esta parede está bem pintada – “This wall is well painted”). There seems to be some restrictions in terms of the category of the processes which are compatible with the intensive value of bem: they correspond always to proprieties. The correspondence of an appreciative or an intensive value to bem depends on the way the notional domain to which the process is associated is structured. In the case of the appreciative value, the notional domain is organized in reference to a type or pattern occurrence, with the course through a class of discretive occurrences. In the case of the intensive value, the notional domain is organized in reference to the attractor centre, with the course through the different degrees of the propriety ordered through a gradient which begins in the attractor centre and finishes on the boundary of the notional domain (see Péroz, 1992: 71). But in certain occurrence contexts, in which bem doesn’t integrate the predicative syntactic domain, it displays a wide spectrum of semantic and pragmatic values. These linguistic contexts are, mainly, question / answer adjacent pairs, integrated in a colloquial conversation. But we can also have monologue examples. For the purposes of this paper – the identification of dialogical values of bem in European Portuguese –, a corpus of spoken and colloquial conversations, including a corpus of literary occurrences, is to be analysed. Most of the sample occurrences belong to the Davies, Ferreira Corpus of the European Portuguese4 , but we considered also a more restricted set of examples, collected in different communication situations. Let’s start by the next two examples: 5. – Então, posso concluir que só te deram o emprego porque falas alemão… – Bem, eu também tenho bons conhecimentos de gestão e de informática… ( – So, I can conclude that you got the job because you speak German… – well, I also have some management and computer knowledge) 4

The Davies, Ferreira Corpus of the European Portuguese has been funded by the US National Endowment for the Humanities, and is freely available online (http://www.corpusdoportugues.org). This corpus allows the search of more than 45 million words in more than 50,000 Portuguese texts from the 1300s to the 1900s.

308

HELENA VIRGÍNIA TOPA VALENTIM

6. Aquele foi um dia importante para mim. Bem, só no dia seguinte é que as coisas se definiram e se soube quem iria integrar a equipa. (That was an important day for me. Well, only in the day after things were defined that it was known who would integrate the team.)

We can recognize a rectifying function in these utterances and the consequent reformulation or change of an element of the previous intervention (5) or period (6). The linguistic forms também (“also”) (5) and só (“only”) (6) mark, through its semantics, this “revision” of the previous propositional content – in 5, through the enlargement of the extent of the reference; in 6, through its restriction. In a dialogical context, in which the interlocutor is the responsible for the previous intervention (5), bem introduces a retort that implies a certain disagreement with respect to the propositional content proposed by the co-enunciator. Bem marks a nonconformity value (in contrast to what we saw in relation to the examples 1 and 2 to whose appreciative value corresponds to a conformity value). This non-conformity value is simultaneously a concessive value. An evidence of this is the frequency of utterances with bem preceding the adversative conjunction mas (“but”). See the example 7: 7. – Eu nunca faria aquilo de que me acusam. Até tenho pena desse senhor… Acha que eu lhe faria mal, que eu ía mandar matá-lo? – Bem, mas há várias provas documentais de que ele trabalhou para si e de que terá participado em situações que o comprometem. (– I would never do what you are accusing me of. I even have pity for that man… Do you think I would issue orders to hurt him? – Well, but there are several documentary proofs that he worked for you and that he participated in situations that compromise you.)

Mas introduces an anti-orientation argument which expresses a speaker’s reticence or discordance, towards the forward intervention’s propositional content. But the cooccurrence of bem, as a linguistic resource that serves to attenuate such a disagreement (here as in the example 5), has an evident pragmatic effect that lessens the confrontation that comes upon that installed disagreement and the subsequent rectification. It is a mark of positive courtesy, a linguistic form that confirms the cooperation between the enunciator and the speaker and the interlocutor. It serves to reinforce the speaker’s positive image, protecting the interlocutor’s (negative) image. This kind of utterances starts with an agreement which is typically a dialogical feature, as pragmatic category, for expressing something non-compromising, i.e., a disagreement. As Borderia’s affirmation of similar linguistic contexts in Spanish, in these cases, “agreement is part of an argumentative move to express a hedged disagreement” (2003: 222). In this hedging operation, we have some modal determination operations. Bem hedges the previous propositional content by lowering the speaker’s epistemic compromise with it. Bem can occur with another associated value. Once again, this value is clearly constructed in reference to the speaker. This value is, therefore, a modal one. Take a look at the follow examples:

DIALOGICAL OCCURRENCES OF BEM IN EUROPEAN PORTUGUESE

309

8. – Professora, não fiz os trabalhos de casa. – Bem. Temos que conversar (– Teacher, I didn’t do my homework – Well, we must talk) 9. – Sabes quem vem connosco? O João. – Bem! Que alegria! (– Do you know who’s coming with us? John! – Well! What an unspeakable joy!)

This is a meta-discourse value, which signals the reception of the message, but can associate a negative appreciative value (adjacent pair 8) or a positive appreciative value (adjacent pair 9). The enunciator expresses himself relative to the bad or good, unfavourable or favourable character of the last sequence’s propositional content. The construction of these values – negative or positive appreciation – depends above all on the prosodic characteristics which are less easily formally tractable5 . The first value with which it is associated is a certain disenchantment (in the adjacent pair 8), which is marked by an intonation characterized by a voice tone that maintains the low levels. The second value (in the adjacent pair 9) is marked by an intonation characterized by an ascent of the voice tone in the production of bem, followed by a decrease that finishes in a suspension. In both cases, we have a very close usage of bem to that one when bem introduces a conclusive sequence (see after). We can also admit that a usage of bem through which the speaker introduces an intensifying value with the possibility of an ironical interpretation that entails a construction of the positive appreciative value. Here, once again, we have an ascendant intonation and a syllabic enlargement. See the example 10: 10. – O João é mesmo simpático – Bem! (– John is a really nice person. – Well!)

In another type and very frequent contexts, bem is used in association with question/ answer contexts, introducing the answer to a question, and, as a consequence, pragmatically assigning it. 11. – Está satisfeito com os serviços prestados pela impresa X? – Bem, não. (– Are you satisfied with the enterprise X services? – Well, no.)

5 The recognition of the prosodic features which characterize each utterance is essential to the description of its determination values. We subscribe the idea expressed in Campos (1998: 39): “[…] ao visar uma análise exaustiva da significação, a linguística não dará conta do fenómeno modal se não puder distinguir os esquemas de entoação específicos de cada valor, relacionando-os com os marcadores sintácticos com que esses esquemas se combinam”.

310

HELENA VIRGÍNIA TOPA VALENTIM

12. – Então, como é que é? Vens almoçar connosco? – Bem, sim. (– So, are you coming to lunch with us? – Well, yes.)

In interviews and inquiries (11), as well as in colloquial conversation, this using of bem has to do with the fact that the total interrogative pre-supposes a validation of the speaker’s pre-constructed propositional content. In other words, it anticipates a yes-orno answer. Because of the impossibility of corresponding in that way, the speaker doesn’t answer categorically, expressing hesitation in the first moment. Thus, in the examples 11 and 12 it comes to be again very evident the bem’s concessive value and the fact of bem being orientated towards a hedged agreement (11) or acceptance (12). Sequences such as bem, sim / não are usually uttered with a pause between bem and não / sim. And we have, in fact, two moments in terms of modal determination. In this sequence the speaker constructs what is called a “modal slide” (Campos, 1997: 155). Bem corresponds, in the first moment, to a hesitation, falling upon the propositional content. Sim / não, in the second moment, marks the assumption of the speaker of the previous propositional content validation, with an affirmative or a negative value. Thus, we have a modal phenomenon of “remodalization” (idem), through which two modal values result from two different modal operations: non-assumption and assumption of the propositional content validation. In this case the second operation always deconstructs the value constructed in the first operation. Through this use of bem – introducing an answer whose question the speaker, thus, denounces orientated in a different sense of what himself preconizes – the speaker can also express a greater complexity associated to the question’s topic. With this sense, in the following contexts, bem introduces a sort of a divergent answer, and its value results from the speaker’s attempt not to impose his opinion. 13. – Há muitos professores que, com metade dessa idade, se reformavam, se pudessem. Trata-se de um caso de coragem e de resistência às agruras da profissão? – Bem, por um lado, as pessoas são diferentes umas das outras. Por outro, como sabe, tenho estado a dar aulas com licença especial. (– There are several teachers who, being half as old as you, would retire, if they could. In your case, is it courage or a resistance? – Well, in one side, people are all different. On the other side, I have been teaching with a special licence.) 14. – Que geração era a sua? Como classifica a família ideológica que a envolvia e inspirava? – Bem, era confuso. Éramos uma geração muito anarquista. Cada um ia ao supermercado e abastecia-se daquilo de que gostava. (– Which generation is yours? How do you classify the family ideology that inspired you? – Well, it was confused. We were a very anarchist generation. Each one used to go to the supermarket and provided himself with what he liked.)

DIALOGICAL OCCURRENCES OF BEM IN EUROPEAN PORTUGUESE

311

15. – Para além da concepção havia alguma ideia sobre quem faria a montagem e o delinear do aspecto gráfico? – Bem, nós tínhamos a sensação de que a Revista com a missão que lhe tinha sido atribuída, foi fundada para não ser mais daquilo que ela é… (– Beyond the conceptualisation, was there some idea about who would do the setting and would outline the graphic image? – Well, we had the feeling that the Magazine, with the kind of mission it had, was founded not to be more than what it was…)

This way, bem occurs in utterances that constitute what pragmatically are reactive interventions and whose speaker refers the antecedent intervention’s propositional content and comment on it. See the next adjacent pair: 16. – O maior, o maciço apoio que a Renamo recebeu, de facto, foi o da própria população de Moçambique. Apoio interno! Em que moldes esse apoio se traduziu? – Bem, a Renamo expressava a revolta da própria população moçambicana. (– In fact, the major support that Renamo received was that of the Mozambican population. Domestic support! How was this support in concrete terms? – Well, Renamo expressed the Mozambican population’s indignation.)

In a great majority of the contexts in which bem marks a discourse value in an answer to a question, it precedes a subjective verb that marks the construction of an “uncertain” epistemic value – for instance crer (“to believe”), achar, pensar (“to think”), presumir (“to presume”), supor (“to suppose”). Less frequent is the co-occurrence of a verb like saber (to know), which marks the construction of a “certain” epistemic value. 17. – O que é bonito é ela ter alma, não é? – Bem, suponho que sim. (– The beauty about this is the fact that she has a soul, isn’t it? – Well, I suppose yes.)

In 17, the co-occurrence of two linguistic forms with a role in the modal determination of the utterance configures an “overmodalization” operation (see Campos, 1997: 151). Belonging to different categories, both linguistic forms (bem and suponho) cooccur marking the same modal value. Let’s examine another type of occurrence context and another kind of values that affect bem, namely, the meta-discourse values. The meta-discourse values very frequently affect bem. They are associated with the way in which the discourse is structured at an informational level and make part of a wide set of available procedure that the speaker uses in order to build up the subjective adjustment which is a cross dimension to all the linguistic expression, but much more evident in the conversation. The occurrence of bem corresponds in these cases to a linguistic marker of an intersubjective regulation effort. It is a determinative for the informational structure of the dialogical exchange.

312

HELENA VIRGÍNIA TOPA VALENTIM

Thus, as a discourse marker and, more specifically, a conversational marker, bem can also signal the opening of a conversation, as it is illustrated in the next examples: 18. – (the doctor to the patient) Bem, vamos lá a ver o que temos aqui (Well, let’s see what we have here) 19. – (an adult to children, even not knowing them) Bem, meninos, acabou a bricadeira (Well, children, the game is over)

Bem is always a reactive linguistic form which opens the conversation as seen in both examples. It signals the acceptance of the use of the word, introducing, in each of these sequences, cordial nuances. In the following examples, bem marks firstly the reception of the message and, secondly, the turn change associated with the theme change. Again, introducing cordial nuances, it signals a cooperation effort with the interlocutor. 20. – Cheguei também a escrever que a terra deveria ser nacionalizada e entregue – sem indemnização! – aos trabalhadores – Bem, então o que o separava da esquerda? (– It happened to me being writing that the land should be nationalized and given – without any indemnification! – to the workers. – Well, so, what separated you from the left?) 21. – Mas a sua embaixada aqui em Lisboa era favorável a uma situação revolucionária em Portugal. – Bem… para voltar à sua questão sobre a nossa derrota, devo dizer-lhe que os militares também ‘borregaram’ (– But your embassy here in Lisbon was in favour of a revolutionary situation in Portugal. – Well… coming back to your question about our defeat, I shall tell you that the military also renounced to a token decision)

Preceding a theme change, bem marks a sequential rupture, often introducing the “preconclusion” of the conversation. 22. Tive contactos directos com ele e percebi que não tinha capacidade para estar à frente de coisíssima nenhuma! Bem, para finalmente lhe responder, perante o impasse criado pelo recuo do PC e as reticências dos militares […], houve uma reunião no Palácio Foz (I had direct contact with him and I understood that he had no ability to lead whatever it was! Well, finally, in answering you, before the impasse created through the PC’s retreat and the military hesitation […], there was a meeting in the Palácio Foz) 23. […] para além dos ‘velhos do Restelo’ que diziam que nós não éramos capazes de fazer e manter uma revista, a ‘nossa revista’ afinal tinha sido e continua a ser um sucesso. Bem, em resumo, esta era a ideia inicial do Ministro… ([…] beyond the “Velhos do Restelo” saying that we weren’t able to make and to maintain the magazine, finally, ‘our magazine’ had been

DIALOGICAL OCCURRENCES OF BEM IN EUROPEAN PORTUGUESE

313

and continues to be a success. Well, in short, this was the Minister’s initial idea.)

Very often this pre-conclusion anticipates a farewell. 24. – Atão! Já anda para aí muita caloirada. – Bem, ainda ficas? (– So, there is already a lot of “caloirada” around – Well, do you still remain?) 25. – Bem, a ver se nos encontramos para conversar mais. (– Well, we’ll meet again to talk more) 26. – Bem, vou-me embora. (– Well, I’m going) 27. – Bem, adeus, até ao meu regresso. (Well, goodbye, till my return.)

As a meta-discourse marker, bem can establish a thematic continuity instead of a rupture. It serves the conversation progress when it processes the information, introducing a synthesis, a resume (28) or a systematization (29). 28. – “Last but not least”, o sector financeiro. – Sim, com o Banco Totta e a companhia de seguros Império. Bem… no essencial, em 74 era isso. (– Last but not least, the financial sector. – Yes, with the Totta Bank and the insurance enterprise Império. Well… essentially, in 74 it was like this.) 29. – …e tresleu? – Bem, há duas coisas: a primeira foi tudo o que lhe acabei de dizer sobre África […] (– … and did you read backwards? – Well, there are two things: the first was everything I just told you about Africa; the second […])

When bem establishes the thematic continuity through the information accumulation, it can introduce sequences that, in a discourse perspective, allow the recapture of the central theme (30) or, in contrast, that are in some respect a digression (31). 30. Ia com eles, atenta ao caminho e com pouca atenção ao que se passava ao nosso redor. Bem, o certo é que eles julgaram tê-lo visto… (– I was with them, attentive to the road and not paying attention to what was happening around. Well, the fact is that they thought they saw him…) 31. Bem, para nós isso não interessava nada, o importante era competir… (Well, for us it has no importance, the most important was to contest).

The introduction of a sequence which is, at a discourse level, digressive or lateral in relation to the preceding sequence can include a more precise expression. Therefore, we can consider that bem shares some functional properties with other “reformulators”.

314

HELENA VIRGÍNIA TOPA VALENTIM

Besides, this function and the one (already mentioned) that indicates the thematic continuity coincide because both introduce in the communicative context a sequence which the speaker considers more adequate. 32. Recordo-me que as palavras do meu tio me fazia sentir… bem, me fez sentir algumas vezes que seria capaz de superar aquele fracasso… (I remember that my uncle’s words used to make me feel… well, made me feel once that I should be able to excel beyond […])

In the example 33, bem marks the recapture of previously introduced theme, after some side remarks or digression: 33. Havia de tudo na vila: uma bela biblioteca, uma sala de espectáculos… Todas as semanas assistia a um concerto, uma performance… Como eu gostava. Bem. Tomar era o sítio ideal para se viver. (There was everything in that small town: a good library, an auditorium… I used to go to a concert or a performance on a weekly basis … How nice it was. Well, Tomar was the ideal place to live.)

Before the conclusion, I would like to make reference to the fact that the discourse marker bem can occur in duplication (bem, bem) and accepts a gradation (muito bem “very well”). In this case it implies a modal appreciation of the sequence it affects. This possibility comes from the appreciative value that the homonymous adverb marks (see sequences 1 and 2). The proposed description allows us to conclude that there is a strict relation between the discourse functions of bem and the variability of its semantic values, namely its modal values. Every time that bem, as a reactive form, expresses a cooperation strategy between the interlocutors and every time that it expresses the non-acceptance of a preceding sequence (contributing, thus, to the conversation progress and structure), the different values that bem assumes are constructed either in reference to the speaker either in reference to the discourse sequence, defining its degree of validation and its discourse orientation. In a theoretical level, these descriptive observations allow us to conceive a semantic approach about the meaning of the linguistic units: 1. The linguistic units meaning is not exclusively inherent; it is constructed in and through the linguistic context, at the same time, it determines the signification of the entire sequence. In other words, the meaning of a linguistic unit doesn’t exist by itself; but it is defined via the several ways in which it is related to the linguistic context. 2. The identity of a linguistic unit doesn’t correspond to any basic sense; it has to do with the specific role it plays in the interactions that constitutes the general meaning of the linguistic sequences. In other words, the sense of a linguistic unit is not apprehensible as the sense of the unit itself, but through the variation of the outcome of such interactions. 3. A linguistic unit does not possess proper sense and figurative sense. We have, consequently, the hypothesis of a multidimensional organization of sense.

DIALOGICAL OCCURRENCES OF BEM IN EUROPEAN PORTUGUESE

315

References Borderia, Salvador Pons (2003). From agreement to stressing and hedging: Spanish bueno and claro. In: Held, G. (ed.). Partikeln und Höfflichkeit. Bern: Peter Lang, 219-236. Campos, Maria Henriqueta Costa (1997). Tempo, aspecto e modalidade. Estudos de linguística portuguesa. Porto: Porto Editora. Campos, Maria Henriqueta Costa (1998). Dever e Poder. Um subsistema modal do português. Lisboa: JNICT/FCG. Campos, Maria Henriqueta Costa & Maria Francisca Xavier (1991). Sintaxe e Semântica do Português. Lisboa: Universidade Aberta. Costa, João (2008). O Advérbio em Português Europeu. Lisboa: Edições Colibri. Culioli, Antoine (1968). La formalisation en linguistique. Cahiers pour l’Analyse 9: 106-117. Duprey, Daniel (1981). Bien et le concept: existence et modalité. BULAG 8, Université de Besançon: 16-58. Franckel, Jean-Jacques (2006). Situation, contexte et valeur référentielle. Pratiques 129/130, juin: 51-70. Lopes, Ana Cristina Macário (2004). A Polifuncionalidade de bem no PE contemporâneo. In: Silva A. S., Torres, A. & Gonçalves, M. (eds.). Linguagem, Cultura e Cognição, Vol. II, Coimbra: Almedina, 433-458. Mosegaard Hansen & Maj-Britt (1998). La grammaticalisation de l'interaction. Pour une approche polysémique de l'adverbe bien. Revue de sémantique et pragmatique 4: 111-138. Péroz, Pierre (1992). Systématique des valeurs de bien en français contemporain. Genève-Paris: Librairie Droz.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 317-328 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

SENS LEXICAL ET PROCESSUS ARGUMENTATIFS: LE CAS DU CONNECTEUR FRANÇAIS C’EST-À-DIRE (QUE) HÉLÈNE VASSILIADOU

Introduction Alors qu’une structure du type X à cause de Y signale explicitement qu’il s’agit d’une relation de cause à effet (ou l’inverse), il n’en va pas de même avec X c’est-à-dire Y. La seule chose qu’un locuteur ordinaire puisse spontanément comprendre de cette relation est que Y est à dire à propos de X. La plupart du temps, il peut lui associer également une valeur explicative. Or, c’est-à-dire (que) (ci-après càd), même s’il semble avoir un sens descriptif neutre, peut prendre des valeurs qui varient selon le type de contexte dans lequel il apparaît et le type de constituants qu’il relie (Vassiliadou 2004 et à par.). Dans cet article, nous essaierons de faire ressortir les valeurs de càd qui s’éloignent de la simple reformulation des contenus propositionnels et qui se rapprochent des valeurs argumentatives. Il est ainsi difficile de parler d’orientation argumentative pour un énoncé comme (1), car, selon les études qui portent sur l’argumentation dans le discours, un tel exemple ne présente pas d’«échelles» (Ducrot 1980) qui sont mises en rapport l’une avec l’autre comme c’est le cas de (2): (1) C’est une coccinelle, c’est-à-dire une bête à bon Dieu (Authier-Revuz 1987: 58). (2) Pierre travaille jour et nuit pour un salaire misérable, c’est-à-dire qu’il se fatigue pour rien.

Par ailleurs, en (1), on peut parler à l’instar d’Anscombre (1995) de blocage de l’utilisation de X en tant qu’argument pour Y à cause de la relation de synonymie entre X et Y. En revanche, plus la ressemblance entre X et Y est modérée, voire faible, plus le glissement vers la dissemblance peut avoir lieu, comme c’est le cas dans l’énoncé (2) où càd Y ne fournit qu’une conclusion parmi le nombre virtuellement infini de conclusions («càd que c’est un brave type» ou «càd qu’il doit avoir besoin d’argent» ou encore «càd qu’il a vraiment besoin de ce travail», etc.). En outre, càd exploite les deux topoï présentés en X (Carel et Ducrot 1999 & Norén 1999: 52-53): a) Il travaille jour et nuit, donc il se fatigue. b) Il travaille pour un salaire misérable, donc il se fatigue pour rien.

Dans le cadre de ce travail, nous porterons notre intérêt sur l’examen de quelques types d’argumentation (Stati 1990) signalés par càd (que ce soit uniquement grâce à ses instructions sémantiques ou grâce à la combinaison de son sens codé et communiqué). Chemin faisant, nous tâcherons de répondre à des questions comme par exemple: est-ce

318

HÉLÈNE VASSILIADOU

que càd entre dans un schéma classique argumentatif du type «résolution d’un conflit»? ou encore dans quelle mesure ses traits sémantiques lui permettent (ou pas) d’intégrer des configurations causales directes telles que X → cause Y? (Bach 1996; Doury 2003; van Eemeren & Grootendorst 2004; Dascal et al. 2005). Notons enfin que notre analyse se situe à un niveau local ou micro-textuel et ne prend pas en considération un cadre de communication précis. Elle fait en revanche appel à un panel de théories sur l’argumentation et le discours en général et à leur impact sur l’emploi des marqueurs discursifs en particulier (cf., entre autres, Perelman et Olbrechts-Tyteca 1958; Plantin 1990; Bazzanella 1990; Rossari 2000; Stati 2002; Carston 2004; Walton 2004). 1. Quel type d’argumentation pour c’est-à-dire? Càd entre dans une relation argumentative qui correspond globalement à trois grands types de cas: (a)Inférences, conclusions, abductions (§ 2); (b)Introduction d’un argument pour une conclusion à venir: (3) (…) qu’ainsi la décision comportant déclaration d’utilité publique et modification du plan d’occupation des sols ne pouvait intervenir que par décret en Conseil d’Etat en application de l’article L. 11-5 du Code de l’expropriation; que le décret attaqué est intervenu le 22 juin 1980 c’est-à-dire dans le délai imparti en ce cas au gouvernement par les dispositions précitées da l’article L. 11-5 du même code; que, dans ces conditions, le comité de défense des expropriations des deux rives du Fémur n’est pas fondé à soutenir que le décret attaqué est intervenu après l’expiration du délai légal» (arrêt du Conseil d’Etat, Bourcier et Bruxelles 1995: 4). (c)Justification (ou càd au début d’une réponse): (4) A: Ça te va Christine la semaine prochaine? B: Càd que la semaine prochaine je me fais opérer à l’épaule. (càd oral)

Dans tous les cas, les orientations argumentatives marquées par càd indiquent que le segment Y est un argument plus fort pour la conclusion visée et que X et Y sont coorientés (Y ne constitue pas un contre-argument pour X). Il signale ainsi une implication (X entraîne l’acte Y) et indique qu’un acte est rendu possible (ou, pour reprendre les termes de Berrendonner 1983, X rend légitime, autorise l’énonciation de Y). Pour des raisons évidentes de place, nous n’examinerons ici que la première catégorie d’emplois argumentatifs de càd, celle qui va de la ‘reprise’ d’un implicite, d’une inférence jusqu’à l’établissement des liens abductifs entre une cause et un effet. 2. Inférences, conclusions, abductions Dans cette triple configuration, il s’agit, selon Hybertie (1996), soit de présenter l’état des choses décrit dans l’énoncé introduit comme une conséquence de l’état des choses

SENS LEXICAL ET PROCESSUS ARGUMENTATIFS

319

décrit dans l’énoncé antérieur (§ 2.1), soit d’inférer un fait, non donné dans l’expérience de l’interlocuteur, d’un fait donné dans son expérience (§ 2.2), soit enfin d’introduire une conclusion présentée comme une conséquence logique des différents moments d’un raisonnement exposé antérieurement (§ 2.3). Il y a toutefois un fil conducteur qui unit ces opérations: càd introduit une forme d’inférence qui va des faits constatés vers une hypothèse explicative (Dendale et De Mulder 1996: 308). Ainsi, l’énoncé càd Y doit être compris soit comme une explication plausible de X implicite, soit comme une conclusion qui nécessite une construction mentale préalable. 2.1 C’est-à-dire introduit une conséquence factuelle Les relations argumentatives instaurées peuvent être du type: (5) Il y a des nuages, c’est-à-dire qu’il va pleuvoir.1

où c’est la connaissance du monde qui permet au locuteur de conclure qu’il va pleuvoir, ou du type: (6) A: Ah, j’en ai marre. Tous les hommes sont des salauds. B: C’est-à-dire que ton mari est un salaud aussi? A: Lui le premier. (càd oral)

auquel cas argumenter peut aussi signifier démontrer la vérité d’une assertion via un syllogisme, un raisonnement logique: «le locuteur B est obligé de conclure que le mari de A est un salaud, puisque c’est un homme et que A a dit que tous les hommes sont des salauds». On observe que le discours (6) ci-dessus fait apparaître une prise en charge du locuteur, en ce qui concerne l’événement introduit par càd, plus forte que celle en (5). Cette différence dépend du type de raisonnement sur lequel s’appuie la relation exprimée ici par le connecteur càd. Par ailleurs, il n’est pas toujours aisé de «distinguer, au moins dans un premier temps, entre une identification faible, type récapitulation, et une différenciation, cas où [le connecteur] introduit une conclusion: est-ce qu’on résume en concluant, ou est-ce qu’on conclut en résumant?» (Hybertie 1996: 10). La comparaison entre (7) et (8), d’une part, et (9) d’autre part, va permettre de distinguer entre ces deux valeurs de la relation: (7) Tu parles des chômeurs, c’est-à-dire / donc des demandeurs d’emploi. (8) D’une part, vous avez beaucoup de diplômes, d’autre part votre travail est très intéressant. C’est-à-dire que / donc vous avez toutes vos chances d’être engagé. (càd oral)

Signalons qu’il s’agit ici d’un des rares exemples où nous pouvons enlever càd en gardant la relation argumentative intacte. Ce n’est pas càd qui crée la relation; il ne fait que l’expliciter.

HÉLÈNE VASSILIADOU

320

En (7), càd établit une relation métadiscursive de type conclusif à la manière de donc2 en faisant appel à une norme qui légitime le processus opéré. L’instruction donnée par les marqueurs est que les chômeurs sont potentiellement des demandeurs d’emploi3 ce qui légitime leur mise en équivalence. Il s’agit donc bien d’une relation d’identification. En (8), le locuteur récapitule les raisons pour lesquelles le candidat a ses chances d’être embauché et présente sa conclusion avec càd (ou donc) comme l’inférence légitime de X; la conclusion est présentée comme une synthèse des états de choses X (cf. aussi Roulet et al. 1985: 149). En revanche, en (9) ci-dessous, càd marque une relation de cause à conséquence. Il ne s’agit plus d’une équivalence entre les termes mis en relation, mais d’une différenciation (Hybertie 1996: 13): (9) Il n’est ni énarque, ni polytechnicien, càd qu’il apporte un peu d’air frais dans l’entreprise (Hybertie 1996: 14 sans càd)

Toutefois, l’assertion de Y est nécessairement légitimée, et cela permet, en outre, tous les coups de force argumentatifs, dont (9) constitue un bel exemple. En somme, dans ce premier cas de figure, l’instruction de càd est de présenter la structure de consécution comme se fondant sur une relation plus générale, valant pour toute occurrence des relations prédicatives de X et Y, et de l’imposer de ce fait comme prise en charge par l’énonciation. Il ne s’agit pas cependant d’une spécificité de càd: l’effet visé dans les exemples examinés est d’imposer une conclusion subjective et/ou objective comme une vérité qui se tire nécessairement de la proposition précédente. Càd est compatible avec ce genre de raisonnement, car son instruction ‘Y est à dire à propos de X’ lui permet de présenter Y comme une conclusion légitime de X. En outre, le contexte, dans son ensemble, conduit à cette conclusion, indépendamment de la présence du marqueur. 2.2 L’emploi inférentiel Ce deuxième emploi se distingue du précédent (même s’il n’est pas toujours aisé de distinguer entre l’un et l’autre de ces emplois) par le fait que càd intègre des opérations inférentielles qui consistent à remonter d’un fait donné dans l’expérience du locuteur à l’énonciation d’un autre, non donné. Ce type d’opération se fonde sur une connaissance du monde, en l’occurrence sur la connaissance des deux états de choses représentés dans les propositions mises en relation. Contrairement aux emplois où il marque l’énoncé d’une conséquence factuelle, il s’agit ici moins d’énoncer la conséquence d’un fait que d’énoncer la conclusion d’un raisonnement et de clore ainsi un mouvement discursif. Les exemples qui appartiennent alors à cette catégorie ne se laissent pas gloser par ‘un 2

Il est intéressant de noter que, dans ce type d’exemples, donc peut-être traité comme un marqueur de glose (Leroy 2005). Quelle est alors la frontière entre glose et argumentation? Il s’agit d’une question qui mériterait d’être creusée. 3 On perçoit, toutefois, que dans cet exemple donc établit une relation argumentative plus forte que càd. D’ailleurs, l’impossibilité de càd dans certains enchaînements argumentatifs dans lesquels donc apparaît tient au fait que sa valeur fondamentale n’est pas de signaler une argumentation.

SENS LEXICAL ET PROCESSUS ARGUMENTATIFS

321

fait X entraîne un fait Y’, mais plutôt par ‘les idées X, Y… permettent de conclure Z’ (Hybertie 1996). Dans ce cas, càd introduit une conclusion à partir de la reconstruction de deux prémisses implicites et la relation établie par ce marqueur prend une valeur d’identification. De ce fait, le marqueur conserve sa valeur métalinguistique, puisque c’est par l’identification construite par càd entre le terme X et le terme Y que Y peut être effectivement conçu comme explicitation ou commentaire de X. Ainsi, un énoncé comme (10) peut, en surface, prendre les allures d’une équivalence. Toutefois, après un examen attentif, nous nous rendons compte que le résultat final ne va pas de soi, mais qu’il découle d’un raisonnement implicite: (10) Au cours du week-end, la célèbre braderie de Lille a connu son succès habituel, càd immense (Charolles et Coltier 1986: 55)

Charolles et Coltier (1986: 55) considèrent que le sens de habituel est spécifié par immense, et «bien que la définition n’ait rien de canonique et qu’elle fasse l’objet d’une mise à distance ironique», elle pose l’identité entre X et Y. Les auteurs glosent l’exemple (10) comme suit: «À Lille, pour parler du succès de la braderie, habituel veut dire immense». Ils rajoutent également que càd crée une relation d’équivalence de signification qui tend à se présenter comme ayant «une valeur conjoncturellement conventionnelle». Il est vrai que ce qui est mis en relation, en surface, c’est «habituel càd immense». Or, ce n’est pas ‘habituel’ qui est ‘immense’ et ce n’est en aucun cas le ‘sens’ de ‘habituel’ qui est spécifié. Càd entre dans un syllogisme que nous pouvons présenter comme suit: - Nous avons une vérité générale qui est: «La braderie de Lille connaît habituellement le succès»; - Ce succès est un succès immense; - Cette année, la braderie de Lille a eu lieu et a connu son succès habituel; - Ce succès est donc immense.

Nous pouvons illustrer ce syllogisme avec le schéma suivant: Si X (braderie de Lille) → X (habituellement succès immense).

Càd nous oblige à reconstruire deux prémisses implicites et introduit la conclusion d’un raisonnement qui découle de ces deux prémisses reconstruites. En d’autres termes, càd construit une relation fondée sur une vérité, sur une connaissance qui peut être ou ne pas être partagée. Dans le cas contraire, on aurait une espèce de décrochement du type: ‘La braderie de Lille a connu son succès habituel, càd nul’. On observe ainsi que càd entre dans une relation argumentative et marque en même temps une équivalence entre X et Y, d’où sa qualification de métalinguistique. La présence du marqueur signifie que l’énoncé introduit est le second terme d’une relation. Il faut préconstruire le premier terme et considérer le deuxième comme équivalent au premier. À la valeur d’identification peut s’adjoindre éventuellement une nuance conclusive et, dans ce cas, nous ne serions pas loin d’une fonction résumante.

322

HÉLÈNE VASSILIADOU

Dans les exemples du type inférentiel, on observe également que càd (comme c’est le cas de donc) peut indiquer que le locuteur prend en charge l’énonciation de ‘l’acte directeur’. Toutefois, càd se distingue de donc dans les cas où le locuteur indique qu’il se distancie de l’acte qu’il reprend tout en prenant en charge l’acte directeur de l’énoncé: (11) Tu dis que Paul s’est marié? C’est-à-dire que ça m’étonne quand même qu’il ne nous ait pas invités. (càd oral)

Comme le notent Roulet et al. (1985: 149), donc n’est pas compatible avec un énoncé comme (11), car «ce sont des motifs indépendants du raisonnement de l’énonciateur qui légitiment le processus déductif opéré»: le locuteur considère que X n’est pas possible, car il aurait été informé, si Paul s’était marié. En revanche, càd peut prendre en charge le doute exprimé en X et signaler la raison pour laquelle le locuteur est étonné. Càd reprend, dans ce cas, la situation créée par l’énonciation de X, et légitime ainsi la production de Y et de l’acte illocutoire que réalise cette énonciation (Hybertie 1996: 30-32). Ainsi, càd peut introduire un segment Y qui renvoie à de l’implicite et gloser l’acte illocutoire latent en l’explicitant: (12) A: La poubelle est pleine. B: Oui, je sais je l’ai vu. A: C'est-à-dire que tu devais la vider. Mais enfin tu ne fais rien dans cette maison.

Dans cet exemple, càd explicite l’inférence que le locuteur B aurait dû tirer de l’énonciation de «la poubelle est pleine». En contexte monologique, on peut paraphraser par un marqueur davantage argumentatif, donc: (12’) La poubelle est pleine. Il faut donc la vider.

Toutefois, si on passe en contexte dialogique, donc paraît moins naturel: (12’’) A: La poubelle est pleine. B: Oui, je sais je l’ai vu. A:? Donc tu devais la vider. Mais, enfin, tu ne fais rien dans cette maison.

Il apparaît clairement que donc suit un cheminement déductif du type quand une poubelle est pleine, on la vide. Càd en revanche reprend l’inférence de façon directe en l’explicitant et enchaîne sur le «non-dit» (Ducrot et al. 1980: 10). Du coup, il peut introduire l’inférence réelle de la poubelle est pleine. On s’aperçoit que le locuteur B n’a pas compris l’implicite et que c’est càd qui déclenche l’inférence il faut la vider. Cette caractéristique, à savoir pouvoir légitimer l’énonciation de Y, différencie aussi càd d’un connecteur davantage argumentatif comme de ce fait (Roulet et al. 1985: 140-145): (13) Alex ne répond pas au téléphone. C’est-à-dire qu’il a déménagé. (13’)?Alex ne répond pas au téléphone. De ce fait, il a déménagé.

On peut gloser (13) par le fait qu’il ne réponde pas au téléphone veut dire qu’il a déménagé. Càd présuppose que le locuteur est au courant d’un éventuel déménagement d’Alex, ce qui permet au locuteur d’inférer il a déménagé et de prendre comme justification le

SENS LEXICAL ET PROCESSUS ARGUMENTATIFS

323

fait qu’il ne réponde pas au téléphone. La réalisation du fait Alex-ne pas répondre-téléphone entraîne qu’il devient possible de poser que le fait il-avoir déménagé est réalisé. Il est intéressant de signaler que la caractéristique de càd d’être compatible avec des énoncés où le locuteur prend en charge son énonciation est largement exploitée par les interviewers. Il n’y a par ailleurs rien d’étonnant à ce que càd soit fréquemment employé dans les interviews comme un récapitulatif (càd que vous pensez que…), car il permet d’indiquer que l’énoncé de l’interlocuteur est inféré de l’énoncé produit par le locuteur. En somme, ce type d’emplois de càd peut être caractérisé comme une «reformulation inférentielle synthétisante»4 : le raisonnement inférentiel permet d’introduire aisément en Y un récapitulatif de X (Zenone 1983 et Rossari et Jayez 1996). Càd, dans un contexte dialogique peut porter sur l’assertion de X et introduire en Y une conclusion à valeur généralisante et/ou restrictive. Le marqueur se greffe, le plus souvent, sur une interprétation personnelle du locuteur inférée à partir d’un implicite: (14) A: La belle voix, c’est un don divin, mais par la suite soit on la met en valeur, soit on la laisse inexploitée. Quelqu’un peut naître avec une voix magnifique, mais il peut finalement ne pas devenir un artiste. B: C’est-à-dire que ça ne suffit pas d’avoir une belle voix, il faut la cultiver aussi. (càd oral)

Comme cela se sait, dans le contexte spécifique des interviews, l’interviewer peut avoir recours à des énoncés paraphrastiques pour ‘obliger’ l’interviewé(e) à expliquer, à décrire, à rendre plus clair, à approfondir ses paroles ou encore à admettre explicitement quelque chose qui est resté sous-entendu. D’une manière générale, selon Labov et Fanshel (1977: 100), l’interviewer connaît déjà les réponses, mais il agit toujours en ayant en tête la présence d’un troisième ‘sujet’, c’est-à-dire du lecteur ou de l’auditeur qui est invisible, mais fort présent. Les paraphrases devraient être neutres en surface, sans la marque de l’interviewer, ce qui n’est pas toujours le cas. En (14), l’interviewer reste effectivement neutre et reprend l’énoncé du locuteur A pour en donner une synthèse conclusive dans le discours Y. La paraphrase n’est pas polémique puisqu’elle va dans le même sens que les propos du locuteur A. Càd introduit la synthèse de X et marque que le locuteur B décide, par sa propre initiative, de procéder à une reformulation de X. En revanche, en (15): (15) A: Quand par exemple certains se sont rassemblés autour de lui, euh, lui, il en a éloigné d’autres qui étaient des proches collaborateurs depuis longtemps. Il avait déclaré à l’époque que cela était nécessaire pour le soi-disant renouvellement du gouvernement. Enfin, il a essayé de faire ce qu’il croyait être bon pour la France, selon son jugement.

On peut voir, à l’aide d’un autre exemple, que càd introduit une conclusion explicite (marquée par j’arrive à ma conclusion) tout en indiquant que X est à interpréter ainsi: «Le taux de chômage n’arrêtera pas de baisser, Monsieur Poivre d’Arvor, càd, et j’arrive à ma conclusion, que toutes les Françaises et tous les Français doivent travailler dans le même sens que notre gouvernement» (càd oral).

324

HÉLÈNE VASSILIADOU

B: C’est-à-dire qu’il s’est lancé dans des actions précipitées et qu’il a fait le mauvais choix A: Ce n’est pas moi qui dis ça. C’est votre constat et celui de la presse. (càd oral)

la conclusion proposée par l’interviewer a un caractère polémique dans le sens où la valeur énonciative de càd est de mettre en cause l’opinion de celui auquel le discours est adressé en lui faisant endosser la validation de la proposition reformulée. La marque reste en surface plus neutre avec càd qu’avec un vous voulez dire par exemple. Toutefois, le locuteur A réagit en refusant de prendre en charge l’énoncé de l’interviewer: «c’est votre constat et celui de la presse». La spécificité de càd, par rapport aux autres marqueurs de raisonnement, tient au fait que sa valeur fondamentale n’est pas d’introduire une conséquence. Cette spécificité tient à la fois à sa valeur sémantique (elle indique la conformité à ce qui vient d’être énoncé ou à ce que l’on constate dans la situation d’énonciation) et à sa valeur anaphorique. Il assure par conséquent des emplois consécutifs d’une manière différente de celles des autres marqueurs de consécution. Dans le cas de càd, l’opération de l’anaphore qu’il marque induit qu’il ne met pas seulement en relation deux énoncés; il reprend X dans Y et offre à Y la légitimité de son assertion en reprenant le contexte antérieur auquel il confère un statut de preuve ou de «raison de comprendre Y» (Zenone 1983). Cette valeur de càd est encore plus claire dans les enchaînements abductifs dont il est question dans la section suivante. 2.3 C’est-à-dire et les enchaînements abductifs D’une portée voisine à celle des deux emplois que nous venons d’examiner, nous trouvons le cas de figure où càd annonce une conclusion qui découle d’un X (implicite et/ ou explicite). Ce troisième emploi se distingue des deux précédents par le fait que càd introduit un segment qui découle d’un raisonnement inférentiel de type abductif. Eggs (1994) explique que, dans le cas de l’abduction, l’argumentation passe par des indices tandis que pour la déduction, elle passe par le vraisemblable (Walton 2001). De même, le propre du type d’opérations marquées par càd est de présenter Y comme plausible et non pas comme un fait certain: Y est l’une des causes possibles qui fait que l’on peut énoncer X (Rossari 1996 et 2000). Càd qualifie l’état de choses précédent comme entretenant une relation sémantique de raison, de motif avec celui introduit. Autrement dit, càd, en n’acceptant que la relation effet-cause, est sensible à la direction sémantique intervenant entre les états de choses représentés. On peut vérifier que l’emploi du marqueur est plus naturel quand la direction de la relation va de l’effet à la cause que quand elle va de la cause à l’effet à l’aide des énoncés (16) à (18): (16) Son permis lui a été retiré. C'est-à-dire qu’il a commis plusieurs infractions routières. (16’)?? Il a commis plusieurs infractions routières. C'est-à-dire que son permis lui a été retiré.

SENS LEXICAL ET PROCESSUS ARGUMENTATIFS

325

(17) Il boit trop, c'est-à-dire qu’il est malheureux. (17’)?? Il est malheureux, c'est-à-dire qu’il boit trop (Rossari 1996 sans càd) (18) Vous avez voté pour Le Pen, c'est-à-dire que vous acceptez sa politique. (càd oral) (18’)??? Vous acceptez la politique de Le Pen, c'est-à-dire que vous avez voté pour lui.

On remarque que càd mobilise la récupération d’une inférence pour obtenir Y: «quand on boit trop, c’est en général qu’on est malheureux» (Rossari 1996: 274-275). Il établit ainsi une relation d’implication qui permet de passer de la vérité de X à celle de Y via le recours à une prémisse implicite. En (18’), par exemple, quelqu’un peut accepter la politique de Le Pen sans forcément voter pour lui. On peut envisager à la place de càd un connecteur typique des relations effet-cause, donc. À la différence de donc qui peut figurer dans des configurations causales directes du type X→cause Y, où X est la cause et Y la conséquence, càd apparaît beaucoup plus facilement dans des configurations causales inversées du type abductif, X→abduct. Y, où X est la conséquence et Y la cause: (18’’) Vous avez voté pour Le Pen, donc vous acceptez sa politique. (18’’’) Vous acceptez la politique de Le Pen, donc vous avez voté pour lui.

S’il est plus naturel de trouver càd dans des raisonnements abductifs, c’est parce qu’il est censé fournir une explication, une justification à propos du segment X: l’état de choses décrit par l’énoncé X est une «manifestation de la conséquence et l’état de choses décrit par l’énoncé Y est une manifestation de la cause» (Rossari 2000: 49). Par ailleurs, càd implique que l’opération à gauche du connecteur garantisse le succès de la seconde, (à droite du connecteur): pour que l’opération soit réussie, la mise en relation entre les deux opérations doit être fondée sur un cheminement causal abductif. Si on essaie d’utiliser une conjonction telle que parce que pour expliciter le rôle justificatif et la légitimité de càd, on s’aperçoit que cela n’est possible que dans le cas de l’abduction: (19) Vous avez voté pour Le Pen, parce que vous acceptez sa politique. (19’) ?? Vous acceptez la politique de Le Pen, parce que vous avez voté pour lui.

Il est important de noter que càd obéit à un principe d’informativité qui postule que càd Y doit être toujours plus informatif que X (Vassiliadou 2004: 314) et que, selon les observations d’Eggs (1994: 52), «accomplir un argument par l’indice» n’est pas «trivial», mais «apporte évidemment une connaissance nouvelle». Ainsi, un argument du type déductif paraît trivial, car la conclusion est déjà incluse dans l’énonciation de X. En revanche, dans le cas des opérations abductives, la conclusion est toujours présentée comme une connaissance nouvelle, ce qui est en accord avec le principe de l’informativité auquel obéit càd. Dans les exemples (16)-(18), càd ne porte pas sur le fait X, mais sur «l’acte directeur» de l’énoncé pour reprendre la terminologie de Roulet et al. (1985) où l’argument càd Y vient appuyer le fait énoncé et/ou inféré X. On aurait affaire, selon Roulet et al. (1985: 130-132), à une «fonction interactive d’explication». L’argument introduit par le marqueur est présenté comme à la fois nouveau et pris en charge par le locuteur. Toutefois, cela ne dépend pas de càd, mais du type de raisonnement sur lequel s’appuie la relation causale (ici de type abductif ). Rossari et al.

326

HÉLÈNE VASSILIADOU

(2004: 58) signalent aussi que les discours de type abductif font «apparaître une prise en charge du locuteur en ce qui concerne l’événement introduit par [le marqueur] plus forte que dans les discours déductifs». Enfin le locuteur, en employant càd, évalue l’acte d’intervention et le présente comme une justification de X. On observe alors que càd ne se contente pas de se greffer sur une inférence implicite afin d’annoncer une conclusion, mais il explicite, il justifie l’inférence en question. Par ailleurs, un cheminement de type abductif s’appuie souvent sur les croyances du locuteur. Il n’est ainsi pas étonnant de trouver càd avec des emplois illocutifs comme dans (20) où càd marque le lien entre un fait et un énoncé: le fait X «il n’arrête pas de dire qu’il veut changer de travail» justifie l’énonciation de Y qui est, en quelque sorte, une conclusion du locuteur concernant l’état de Pierre: (20) Pierre n’arrête pas de dire qu’il veut changer de travail. C’est-à-dire qu’il ne doit pas être très content de ce qu’il fait.

3. C’est-à-dire et l’argumentation: discussion Il va sans dire que notre marqueur sous-tend souvent une relation de discours plus spécifique qu’un simple enchaînement reformulatif. Il y a en effet une dépendance forte entre càd (et ce qu’il implique) et les constituants qu’il relie. Il peut ainsi, dans le discours argumentatif, scander les enchaînements discursifs et intervenir là où le locuteur prend appui sur lui-même (ou sur son interlocuteur) dans un mouvement à la fois explicatif 5 et justificatif. Or, comme le notent fort bien Murat et Cartier-Bresson (1987: 13), càd ne peut que donner une consigne; «encore faut-il pour l’appliquer, pouvoir reconstruire une relation plausible entre les termes». En effet, même si càd, en tant que mot du discours, est pourvu d’une fonction communicative minimale, il n’accepte pas pour autant n’importe quelle combinaison entre X et Y. En résumé, càd, comme tous les connecteurs, impose des contraintes non seulement sur le type des entités qu’il connecte (on a vu par exemple que, contrairement à de ce fait, il accepte que l’unité gauche soit pourvue d’une valeur illocutoire autre que l’assertion), mais aussi sur la nature de la relation d’implication (déduction vs abduction). Par ailleurs, càd comprend dans son sens descriptif l’instruction dynamique de à dire et cela lui permet de se greffer sur des inférences, sur des actes illocutoires ainsi que de présenter un discours à venir comme une explication, une justification ou encore une conclusion. Reste à voir plus en détail ses spécificités par rapport à d’autres marqueurs davantage argumentatifs.

5 Pour une discussion sur la différence entre argumentation et explication, voir Walton (2004).

SENS LEXICAL ET PROCESSUS ARGUMENTATIFS

327

Références bibliographiques Anscombre, Jean-Claude (1995). Topique ‘or not’ topique: formes topiques intrinsèques et formes topiques extrinsèques. Journal of Pragmatics 24: 115-141. Authier-Revuz, Jacqueline (1987). L’auto-représentation opacifiante de dire dans certaines formes de couplage. DRLAV 36-37: 55-103. Bach, Carme (1996). Reformular: una operaciò argumentativa aséptica? Estudio del connector de reformulaciòn parafràstica és a dir. Sendebar 7: 255-271. Bazzanella, Carla (1990). Phatic Connectives as Interactional cues in spoken Italian. Journal of Pragmatics 14: 629-647. Berrendonner, Alain (1983). ‘Connecteurs pragmatiques’ et anaphores. Cahiers de linguistique française 5: 215-246. Bourcier, Danielle & Sophie Bruxelles (1995). Une approche sémantique de l’argumentation juridique. Dire et c'est-à-dire. L’Année sociologique 45/1: 35-57. Carel, Marion & Oswald Ducrot (1999). Le problème du paradoxe dans une sémantique argumentative. Langue française 123: 6-26. Carston, Robyn (2004). Explicature and Semantics. In: Davis, S. & B. Gillon (eds.). Semantics: A Reader. Oxford: Oxford University Press, 817-845. Charolles, Michel, & Danielle Coltier (1986). Le contrôle de la compréhension dans une activité rédactionnelle: Eléments pour une analyse des reformulations paraphrastiques. Pratiques 49: 51-66. Dascal, Marcelo, Frans Van Eemeren, H., Eddo Rigotti, Sorin Stati, & Andrea Rocci (eds.) (2005). Dialogue, Argument, Controversy. SComS Special Issue: 1-20. Dendale, Patrick & Walter De Mulder (1996). Déduction ou abduction: le cas de devoir inférentiel. In: Guentchéva, Z. (éd.). L’énonciation médiatisée. Louvain / Paris: Peeters, 305-318. Doury, Marianne (2003). Argumentation et mise en voix; les discours quotidiens sur l’immigration. In: Bondi M. & S. Stati (eds.). Dialogue Analysis 2000. Selected papers from the 10th IADA Anniversary Conference, Bologna 2000. Tübingen: Niemeyer Verlag, 173-183. Ducrot, Oswald (1980), Les échelles argumentatives. Paris: Minuit. Eemeren, Frans H. van & Rob Grootendorst (2004). A systematic theory of Argumentation. Cambridge: CUP. Eggs, Ekkehard (1994). Grammaire du discours argumentatif. Paris: Kimé. Hybertie, Charlotte (1996). La conséquence en français. Paris: Ophrys. Labov, William & David Fanshel (1977). Therapeutic discourse: psychotherapy as conversation. New York: Academic Press. Leroy, Sarah (2005). Sécurité, donc sûreté. L’emploi de donc marqueur de glose. In: Steuckardt A. & A. Niclas-Salminen. Les marqueurs de glose. Aix-en Provence: PUP, 159-170. Murat, Michel & Bernard Cartier-Bresson (1987). C’est-à-dire ou la reprise interprétative, Langue française 73: 5-15. Norén, Coco (1999). Reformulation et Conversation. De la sémantique du topos aux fonctions interactionnelles. Uppsala: Studia Romanica, Upsaliensia 60. Perelman, Chaïm & Lucie Olbrechts-Tyteca (1958). La nouvelle rhétorique. Traité de l’argumentation. 2 vol. Paris: PUF.

328

HÉLÈNE VASSILIADOU

Plantin, Christian (1990). Essais sur l’argumentation. Introduction à l’étude linguistique de la parole argumentative. Paris: Kimé. Rossari, Corinne (1996). Les marques de la consécution: propriétés communes et distinctives à la lumière de donc, de ce fait, et il en résulte que. In: Muller, C. (éd.). Dépendance et intégration syntaxique: subordination, coordination, connexion. Tübingen: Niemeyer Verlag, 271-283. Rossari, Corinne (2000). Connecteurs et relations de discours: des liens entre cognition et signification. Nancy: Presses Universitaires de Nancy. Rossari, Corinne & Jacques Jayez (1996). Donc et les consécutifs. Des systèmes de contraintes différentiels. Linguisticae Investigationes XX/1: 13-28. Rossari, Corinne, Anne Beaulieu-Masson, Corina Cojocariu, Anna Razgouliaeva (2004). Autour des connecteurs. Réflexions sur l’énonciation et la portée. Berne: Peter Lang. Roulet, Eddy, Antoine Auchlin, Jacques Moeschler, Christian Rubattel & Marianne Schelling (1985). L’articulation du discours en français contemporain. Berne: Peter Lang. Stati, Sorin (1990). Le transphrastique. Paris: PUF. Stati, Sorin (2002). Principi di analisi argomentativa. Bologna: Patron. Vassiliadou, Hélène (2004). Les connecteurs c’est-à-dire (que) en français et ðilaði en grec. Analyse syntaxique et sémantico-pragmatique. Thèse de Doctorat. Strasbourg: Université Marc Bloch. Vassiliadou, Hélène (à par.). Quand les voies de la reformulation se croisent pour mieux se séparer: à savoir, autrement dit, c’est-à-dire, en d’autres termes. In: Schuwer M. & M.-C. Le Bot (éds.). Voies de la reformulation: contraintes, objectifs, stratégies. Rennes: PUR. Walton, Douglas (2001). Abductive, presumptive and plausible arguments. Informal Logic 21/2: 141-169. Walton, Douglas (2004). A new dialectical theory of explanation, Philosophical Explorations 7/1: 71-89. Zenone, Anna (1983). La consécution sans contradiction: donc, par conséquent, alors, ainsi, aussi (2ème partie). Cahiers de Linguistique Française 5: 189-214.

CROSS-LINGUISTIC PERSPECTIVES

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 331-342 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

RINFORZO ARGOMENTATIVO E ATTENUAZIONE ARGOMENTATIVA IN FRANCESE, ITALIANO, RUSSO, SPAGNOLO E TEDESCO: UNA PROPOSTA DI ANALISI CONDOTTA SULLA BASE DI FORUM DI DISCUSSIONE ON-LINE

VAHRAM ATAYAN

Introduzione L’obiettivo del presente contributo è la presentazione di alcuni procedimenti e mezzi linguistici che permettono il rinforzo o l’attenuazione del potenziale argomentativo di un enunciato, e in particolare di certi meccanismi frequentemente usati nei forum di discussione on-line. La scelta di questo tipo di testi è legata al loro carattere intermediario influenzato sia dal medium scritto che dalla relativa vicinanza comunicativa tipica della particolare situazione comunicativa. Tali fattori caratterizzano i forum come appartenenti alla classe di testi che manifestano il tratto dell’oralità concettuale nel senso inteso da Koch & Oesterreicher (1997)1 . Gli esempi discussi sono stati tratti dai forum di discussione del Nouvel Observateur, di Die Zeit, di El País, di La Repubblica e delle Известия, uno dei più grandi quotidiani nazionali russi. La scelta di tre lingue romanze, una germanica e una slava dovrebbe permettere una certa generalizzabilità dei risultati, che in tal modo potrebbero essere validi in una qualche misura per i fenomeni argomentativi discussi nel contributo al di là dei limiti di una lingua naturale specifica. Nel presente saggio si discuteranno quattro punti: - la necessità di introdurre nella teoria dell’argomentazione i concetti di forza/di rinforzo e di debolezza/di attenuazione del potenziale argomentativo di un enunciato, - i procedimenti del rinforzo e dell’attenuazione argomentativi, - i meccanismi e i marcatori linguistici del rinforzo, - i meccanismi e i marcatori linguistici dell’attenuazione, in particolare i mezzi linguistici che permettono il passaggio dall’attenuazione argomentativa all’inversione dell’orientamento argomentativo (cfr. Ducrot 1983). 1. I concetti di rinforzo e di attenuazione argomentativa Gli studi sull’argomentazione sembrano essere soggetti a una continua diversificazione terminologica e concettuale, dovuta in gran parte a fattori oggettivi, quali ad esempio la forte interdisciplinarità del campo legato a retorica, filosofia, linguistica, scienze del1 Cfr. Pirazzini 2006 per un’analisi di procedimenti retorico-argomentativi usati nei forum di discussione.

332

VAHRAM ATAYAN

la comunicazione, logica ecc. Dato che tali situazioni creano di solito difficoltà e ostacoli per lo sviluppo della ricerca, un’ulteriore estensione concettuale e terminologica non ci pare accettabile se non è dovuta a una precisa e motivata necessità. Ora, se optiamo qui per l’introduzione di tali concetti accanto a quelli abitualmente usati per la descrizione micro- o macrostrutturale dell’argomentazione (cioè argomento, conclusione, argomentazione a grappolo, argomentazione multipla2 , controargomentazione ecc.), è perché questa decisione sembra essere fondata su una serie di considerazioni che provengono da diversi campi: - L’intuizione metalinguistica dei parlanti pare permettere senza alcuna difficoltà valutazioni del genere “un argomento forte/debole”. Questo fatto non può di per sé essere considerato come argomento decisivo a favore della ripresa degli stessi termini come elementi del metalinguaggio descrittivo3 , però può corroborare le intuizioni del linguista riguardo al valore funzionale di certe sequenze linguistiche partecipanti all’argomentazione. - Relazioni di forza argomentativa vengono usate in un’ottica semasiologica nella descrizione semantica di numerosi connettori testuali, tali mais (Ducrot et al. 1980: 97) o même (Anscombre & Ducrot 1983: 58) e di operatori come per es. i modificateurs réalisants e déréalisants introdotti in Ducrot (1995). - L’idea di forza argomentativa viene utilizzata inoltre al livello più astratto delle funzioni argomentative generali, come nel caso della categoria rinforzo di Lo Cascio (1991: 191) o anche nelle considerazioni più intuitive di Perelman & OlbrechtsTyteca (1958: 661-662) sull’ordine tipico della presentazione di argomenti per la stessa tesi, il quale può dipendere dalla loro forza relativa. - Le considerazioni semasiologiche e funzionali sembrano essere confermate infine dalla frequenza di costruzioni linguistiche che, strutturalmente, sono sufficientemente indipendenti per poter verbalizzare componenti di un’argomentazione, ma non paiono ciononostante corrispondere agli elementi di strutture micro- o macroscopiche “classiche”, né costituire strumenti di presentazione di una funzione modale nel senso di Toulmin (1958: 105). Le sequenze linguistiche atte ad essere utilizzate nell’argomentazione costituiscono un insieme abbastanza eterogeneo (cfr. Stati 2002: 30) e non possono essere ridotte a singole strutture sintattiche. Sembra essere potenzialmente argomentativa infatti qualsiasi sequenza linguistica la cui apparizione nella linearità testuale possa essere messa in relazione a un’intenzione separata del parlante, che non risulta obbligatoriamente dal co- e contesto. Questa visione corrisponde anche alla definizione di Ducrot (1984: 174-175) dell’enunciato come unità risultante da una scelta relativamente autonoma del parlante. In ge2

Riprendiamo qui per alcuni concetti la terminologia italiana di Lo Cascio (1991: 128-129), che corrisponde in parte alle categorie della scuola pragma-linguistica (cfr. van Eemeren, Grootendorst & Snoeck Henkemans 1996). 3 Cfr. Ducrot (1984: 117) sulla necessità di evitare la confusione tra il livello linguistico e quello metalinguistico.

RINFORZO ARGOMENTATIVO E ATTENUAZIONE ARGOMENTATIVA

333

nerale, sequenze di questo tipo costituiscono strutture linguistiche in grado di realizzare un’azione comunicativa, cioè un’illocuzione nella terminologia della teoria degli atti linguistici, ma anche un’unità informativa (cfr. Ferrari 2003: 28-29; Brandt 1990: 79). Ovviamente, strutture del genere possono essere realizzate da un ampio spettro di costruzioni sintattiche che sono perciò in grado di fungere per esempio da argomento in un’argomentazione. Nella sequenza costruita 1 una frase relativa non-restrittiva (prendiamo in considerazione questa categoria a titolo d’esempio) realizza un argomento per la conclusione verbalizzata nella principale: 1. Paolo, che attualmente è troppo occupato, probabilmente non potrà venire con noi.

Ora, le stesse strutture sembrano spesso partecipare a un’argomentazione minimale, costituita da una tesi e un argomento, senza però assumere il ruolo dell’argomento o della conclusione: 2a. La situazione economica dell’UE sembra essere soddisfacente: persino la Banca Centrale Europea, che di solito è molto prudente, ha pubblicato una previsione positiva per questo anno. 2b. Certo, la BCE, che di solito è molto prudente, ha dato una previsione piuttosto negativa, ma io penso che lo sviluppo economico dell’UE sia soddisfacente, dato che abbiamo un’inflazione molto bassa.

Nell’esempio 2a la frase relativa sembra far apparire l’argomento – il parere della BCE – come particolarmente forte, il che è anche compatibile con le istruzioni del marcatore persino (questa analisi corrisponde anche alla categoria di rinforzo introdotta da Lo Cascio 1991: 191). In 2b invece la concessione – un argomento potenziale per la conclusione contraria – sembra essere resa più debole dalla relativa. Possiamo perciò constatare che strutture linguistiche potenzialmente in grado di verbalizzare argomenti o conclusioni possono anche avere una funzione che corrisponde all’intuizione della forza/debolezza argomentativa. E proprio per descrivere tali procedimenti nei testi argomentativi reali ci sembra utile introdurre a livello teorico i concetti di rinforzo e di attenuazione del potenziale argomentativo di una sequenza4 . 2. I procedimenti del rinforzo e dell’attenuazione argomentativi L’analisi degli esempi autentici tratti da diversi tipi di testo ci ha permesso di identificare tre procedimenti semantico-pragmatici utilizzati per rinforzare o attenuare un argomento (cfr. Atayan 2004, 2006 e in stampa): - Sembrano essere considerati come forti argomenti quelli nei quali l’elemento lessicale argomentativamente più importante (spesso nella posizione di focus dell’enun4

Parliamo qui di “potenziale argomentativo” nel senso ampio previsto nel quadro dell’Argumentation dans la langue (Anscombre & Ducrot 1983), cioè della capacità di un enunciato di favorire certe sequenze conclusive (enchaînements) e escluderne altre.

334

VAHRAM ATAYAN

ciato) è modificato nel senso di un’intensificazione, legata spesso al posizionamento nella parte superiore di una scala concreta o astratta. Elementi del genere corrispondono almeno in parte alla categoria lessicale dei modificateurs réalisants (MR) di Ducrot (1995). Nella sua analisi Ducrot ha constatato in particolare che tali modificatori permettono la costruzione di sequenze accettabili di tipo [elemento lessicale + et même + MR (+ eventualmente l’elemento lessicale)]5 . Tali casi sono legati spesso a un paradigma di gradi che una qualità può assumere, dando origine in tal modo a una relazione di forza argomentativa tra enunciati argomentativamente coorientati: “Paolo è molto intelligente/intelligentissimo” è, ceteris paribus, un argomento più adatto per giustificare la conclusione “Sarà in grado di risolvere questo problema” rispetto a “Paolo è intelligente”, e questo sembra valere – in assenza di condizioni contestuali molto particolari come “Abbiamo bisogno di qualcuno che sia intelligente, ma non troppo” – per tutte le conclusioni che possono essere corroborate da “Paolo è intelligente”. Ovviamente argomenti che fanno riferimento al grado inferiore di una qualità vengono considerati deboli, spesso anche insufficienti a giustificare le conclusioni potenziali della stessa sequenza priva del marcatore di posizionamento sulla scala (modificateurs déréalisants nella terminologia di Ducrot 1995): “Sono un po’ malato” o “Paolo è un parente lontano” sembrano essere meno adatti o addirittura non più idonei a giustificare le conclusioni possibili di “Sono malato” o “Paolo ci aiuterà”. - Il secondo gruppo di argomenti rinforzati è costituito da sequenze nel cui contesto vengono verbalizzate circostanze che normalmente dovrebbero rendere improbabile la realizzazione dello stato di cose citato nell’argomento. Tale è il caso in 2: la generale prudenza della BCE dovrebbe rendere improbabile, controimplicare, una previsione positiva. Nell’esempio 2b invece viene presentato un fattore che normalmente dovrebbe implicare il contenuto dell’argomento rendendolo più debole6. - Il terzo procedimento consiste nell’asserzione del carattere sorprendente – per un argomento forte – o normale – per un argomento debole – dello stato di cose verbalizzato nell’argomento. Tali procedimenti possono anche apparire insieme in una sequenza di tipo [A1 + normale (A1)] – marcatore di contrasto – [A2 + sorprendente (A2)] con possibili variazioni nell’ordine. L’esempio 3, tratto dal sito internet animafrica.net, illustra il procedimento in questione: 3. Qui a Isiro tutte le comunità sono in fermento per preparare degnamente la festa della Mamma di tutti. (A1) Già da una settimana ci si prepara con preghiere e Messa, e questo è normale, ma ciò che è sorprendente (A2) è l’orario.

5 Cfr. “Paolo è intelligente, e persino/anzi molto intelligente”, “È un parente, e persino molto vicino”. 6 Per il legame generale tra controimplicazione e forza argomentativa cfr. la categoria di rinforzo

in Lo Cascio (1991: 189) nonché Blumenthal (1990: 37-38), Carel (1995), Kalokerinos (1995: 82, 96), García Negroni (2001: 50, 2003: 61, 78), Atayan (2004, 2006: 317), per il legame tra implicazione e debolezza argomentativa Ducrot et al. (1980: 166), Lötscher (1988: 83), Anscombre (1996: 255), García Negroni (2003: 20), Atayan (2004 e 2006: 317).

RINFORZO ARGOMENTATIVO E ATTENUAZIONE ARGOMENTATIVA

335

Alle 5.30 del mattino, è buio la chiesa si riempie di fedeli, piccoli, grandi, donne, uomini […] come alla domenica. (Sequenza di ricerca: “sono in fermento per preparare degnamente”)7

Nell’esempio 3 la valutazione “tutte le comunità sono in fermento” viene illustrata da due sequenze: la prima che menziona le attività che si svolgono da una settimana – il cui effetto argomentativo viene attenuato dal riferimento alla normalità, e la seconda che riguarda gli orari insoliti dell’attività, presentati appunto come sorprendenti. Il marcatore contrastivo ma permette anche una gerarchizzazione argomentativa: il primo argomento attenuato dalla constatazione di normalità non vale più come tale per la conclusione, mentre la seconda sequenza realizza, secondo le istruzioni di ma, un argomento anti-orientato, cioè corroborante la conclusione e più forte. Nei paragrafi seguenti discuteremo, vista la loro frequenza nel corpus analizzato, due procedimenti di rinforzo, legati all’intensificazione e alla controimplicazione. Per quanto riguarda l’attenuazione analizzeremo qualche esempio di passaggio tra l’attenuazione argomentativa e l’inversione dell’orientamento argomentativo segnalato dai marcatori di focalizzazione retrospettiva. Per i procedimenti analizzati presenteremo i meccanismi semantico-pragmatici più specifici e i marcatori linguistici che possono introdurre sequenze rinforzanti o attenuanti. 3. Meccanismi e marcatori del rinforzo 3.1 Meccanismi di intensificazione Oltre ai casi di intensificazione scalare realizzata da modificateurs réalisants generali quali quantificatori di alto grado, nel corpus analizzato sono frequenti i casi in cui il rinforzo argomentativo viene realizzato tramite un meccanismo più specifico, legato ai tratti spazio-temporali dello stato di cose menzionato nell’argomento e ad altri fattori. Frequentemente tali informazioni vengono introdotte retrospettivamente, immediatamente dopo l’argomento, e segnalate da un marcatore costituito dalla congiunzione coordinante e un pronome dimostrativo generico (et ce/ceci/cela, e questo, и это, y eso, und das/dies)8 . Le indicazioni di tipo spazio-temporale sembrano in particolare essere di carattere sufficientemente generico per fungere da rinforzo per una classe abbastanza ampia di potenziali argomenti. La durata (es. 4a, 4b) o l’estensione spaziale (4c) di uno stato di cose sembra per esempio valorizzarlo dal punto di vista argomentativo: 4a. Je pense que le Kosovo mérite l’indépendance non seulement parce que les albanais du Kosovo sont majoritaires et opprimés et cela depuis des siècles mais aussi car ce sont des gens motivés à avancer, à créer des projets d’améliorations des conditions de vie non seulement des albanais mais également des serbes du Kosovo. (Sequenza di ricerca: “majoritaire et opprimés et cela depuis”) 7

Gli esempi dalle fonti on-line vengono ripresi senza correzioni. Invece del sito internet vengono indicate le sequenze di ricerca con le quali gli esempi possono essere ritrovati tramite il motore di ricerca www.google.de. 8 Per altri marcatori di rinforzo cfr. Lo Cascio (1991: 191) e Atayan (2006: 381-402).

336

VAHRAM ATAYAN

4b. Hi @ll Ehe ist prima, nur sollte nicht vergessen werden das sie sehr häufig in die Brüche geht. Tatsache in dem Zusammenhang ist, das der Mann bei einer Scheidung der komplette Depp ist. Er muss einer Frau Unterhalt zahlen, die von ihm nichts mehr wissen will, und das wenn es dumm läuft ein Leben lang. (Sequenza di ricerca: “und das wenn es dumm läuft ein Leben lang”)9 4c. Антон Захаров, МНЕ не надо лишний раз демонстрировать Вашу п ол н е й ш у ю п р а в о в у ю б е з г р а м о т н о с т ь . По з а к о н у ( и э т о ВСЕМИРНАЯ 10 практика) ПО приравнивается к песне, или, например, фильму. (Sequenza di ricerca: “Антон Захаров, МНЕ не надо лишний раз демонстрировать”)11.

Possono inoltre avere l’effetto di rinforzo, soprattutto nel quadro di una valutazione, anche i riferimenti a una “circostanza aggravante” interpretati come indicatori di un atteggiamento morale riprovevole: 4d. Come al solito “Il Giornale” dà prova di alta obiettività, e questo senza neanche arrossire. Prodi, i ministri e sottosegretari si diminuiscono il salario del 30% (Prodi =37.000 euro) e lo straccio di Berlusconi titola : “Ci aumentano le tasse e si aumentano i salari” (Sequenza di ricerca: “obiettività, e questo senza neanche arrossire”)

Qui il rimprovero, formulato ironicamente, viene intensificato dalla sequenza “e questo senza neanche arrossire” che sembra poter essere considerata come una specie di modificateur réalisant generale per azioni riprovevoli. Infatti, per qualsiasi azione A di questo tipo la sequenza “Ha fatto A, e questo senza neanche arrossire” sembra accettabile senza ipotesi contestuali particolari, il che corrisponde al criterio di Ducrot (1995: 147). 3.2 Meccanismi di controimplicazione I casi di controimplicazione rinforzante costituiscono, nel corpus, tre sottogruppi. Il primo comprende argomenti nei quali viene descritto uno stato di cose che esiste nonostante una causa o un motivo che normalmente dovrebbero impedirne l’esistenza. Essi vengono spesso introdotti dai marcatori menzionati sopra (congiunzione coordinante + pronome dimostrativo) e un connettore concessivo. Negli esempi 5a e 5b possiamo osservare che gli stati di cose usati in funzione di argomento – un cambiamento minimo della temperatura o il mancato successo dell’Eolo – si verificano nonostante i fattori controimplicanti, cioè rispettivamente una maggiore emissione dell’anidride carbonica o i finanziamenti dell’UE. 9

Bruxelles (2002) constata che anche un’indicazione sulla velocità o immediatezza di un’azione o un evento possono avere un effetto di rinforzo sul potenziale argomentativo della sequenza nella quale l’azione o l’evento vengono verbalizzati. 10 Maiuscoli nell’originale. 11 Anton Zakharov, IO non ho bisogno di dimostrare ancora una volta la Sua incompetenza totale in questioni legali. Stando alla legge (e questo è la prassi MONDIALE) il software viene trattato come le canzoni o per esempio i film. (Traduzioni dal russo dell’autore).

RINFORZO ARGOMENTATIVO E ATTENUAZIONE ARGOMENTATIVA

337

5a. Wie selbstverständlich wird hier behauptet, dass der Klimawandel sich beschleunigt. Ein Blick auf die globalen Temperaturkurven der Bodenstationen zeigt aber nun, dass sich die weltweiten Temperaturen seit 1998 nur geringfügig verändert haben und keinen positiven Trend mehr zeigen. Ebenso gilt dies für die Messungen durch Satelliten. Und dies trotz stark gestiegenen CO2-Werten. (Sequenza di ricerca: “Und dies trotz stark gestiegenen CO2-Werten”) 5b. La Eolo si è dimostrata una bufala per un motivo molto semplice: nessun prototipo ce l’ha fatta a superare i problemi tecnici [...]. E questo nonostante i fortissimi finanziamenti della UE, dati praticamente a fondo perduto. (Sequenza di ricerca: “E questo nonostante i fortissimi”)

Il secondo caso, complementare al primo, di controimplicazione rinforzante contiene azioni o stati di cose che si realizzano nonostante una causa o un motivo insufficiente. Nell’esempio 6a il carattere insufficiente del motivo (il potere) viene marcato dal tipico modificateur déréalisant nur, mentre in 6b la causa che normalmente dovrebbe impedire la raccolta di una somma particolarmente alta (il fatto che si tratti di una sola persona) è segnalata dal corrispondente russo только: 6a. Koste es was es wolle, auch um den Preis der Rehabilitierung des Herrn Oskar Lafontaine, gibt die Sozialdemokratische Partei Deutschlands alle gültigen Vorstellungen von Ehrlichkeit, von Respekt vor dem Souverän (Wähler) auf. Sie ist sich nicht zu schade, ihre eigenen Grundsätze aufzugeben und dies nur wegen des Preises der Macht. (Sequenza di ricerca: “und dies nur wegen des Preises der Macht”) 6b. Грызлов и его дружок Володин ну никак не могут найти денег для приличной пенсии старикам [...]. Если бы Абромович платил в год налог хотя-бы 30-40 млрд. руб., то никакого кризиса с лекарствами не было бы. И это только с одного олигарха. (Sequenza di ricerca: “И это только с одного олигарха”)12

Un uso specifico del rinforzo argomentativo di questo tipo è costituito infine dai casi in cui la controimplicazione risulta da una contraddizione tra le dichiarazioni e le azioni in particolare nel campo politico (7a) o tra l’importanza di un mezzo di informazione e le posizioni da esso sostenute (7b). Le due situazioni sono particolarmente frequenti nella critica dei politici o dei media, sul cui sito si svolgono le discussioni, il che corrisponde ovviamente anche a una funzione di base dei forum di discussione dei grandi quotidiani: dare la possibilità ai lettori di esprimere la loro opinione, anche critica, sul giornale. Generalmente la posizione o l’azione in sé riprovevole – cioè un argomento a favore di una valutazione negativa dell’attore – viene presentata come particolarmente grave a causa dell’esistenza di tali elementi controimplicanti:

12 Gryslov e il suo amico Volodin non riescono proprio a trovare i soldi per una pensione di vecchiaia degna di questo nome. Se Abramovich pagasse almeno 30-40 miliardi di rubli di tasse all’anno non ci sarebbe una crisi delle medicine. E questo per uno solo degli oligarchi.

338

VAHRAM ATAYAN

7a. Au lendemain du Grenelle de l’environnement, il est choquant de voir se tenir un conseil des ministres décentralisé, avec tous les déplacements et la pollution que cela engendre. Et ceci en prétendant défendre l’écologie en Corse! Ou Sarko et sa clique se fichent de nous et les Français apprécieront, ou ils n’ont rien compris et c’est encore pire. (Sequenza di ricerca: “Et ceci en prétendant défendre l’écologie en Corse!”) 7b. Ну, совсем уже докатились. И это центральное издание... Позор Извес тиям! Янине Соколовской тоже! (Sequenza di ricerca : “докатились. И это центральное издание”)13

Possiamo constatare nel complesso che i marcatori (congiunzione coordinante + pronome dimostrativo generico: et ce/ceci/cela, e questo, и это, y eso, und das/dies) sono in grado di introdurre un rinforzo retrospettivo del potenziale argomentativo di una sequenza, rinforzo che può basarsi su meccanismi semantici molto diversi. 4. Meccanismi e marcatori dell’attenuazione Nell’ultima parte della nostra analisi ci limiteremo alla presentazione di un fenomeno particolare legato all’attenuazione argomentativa. Si tratta dell’effetto di possibile ambiguità argomentativa delle sequenze che contengono elementi attenuanti il potenziale argomentativo del resto dell’enunciato. Ducrot (1995) ha osservato che le sequenze che contengono combinazioni (modificateur déréalisant + elemento lessicale) possono essere sia coorientate che anti-orientate rispetto alla stessa frase senza il modificatore. L’effetto di inversione dell’orientamento è legato alla posizione rematica/focale del modificatore: mentre “Possiamo constatare un leggero CALO dell’attività economica” sembra essere orientato verso una conclusione come “La situazione è difficile”, la stessa sequenza con “un LEGGERO calo”/”un calo LEGGERO” diventa un argomento piuttosto per le conclusioni come “La situazione non è (particolarmente) difficile”. Ovviamente, nel caso delle coppie aggettivo + sostantivo in certe lingue (per es. in quelle romanze) questo effetto può essere marcato nella superficie dall’ante- o posposizione dell’aggettivo, mentre nelle altre l’unico marcatore sembra essere l’intonazione (“ein leichter RÜCKGANG”/“легкое СНИЖЕНИЕ” vs. “ein LEICHTER Rückgang ”/ “ЛЕГКОЕ снижение”). Si intende che, sia in casi come quelli appena discussi, che in presenza di altri elementi attenuanti, per i quali il sistema linguistico non prevede tali libertà posizionali, soprattutto nei testi scritti – come i forum – può rivelarsi necessario disambiguare l’orientamento argomentativo. Nel corpus analizzato tale funzione viene realizzata spesso da sequenze metalinguistiche che forniscono una precisazione14 della focalizzazione (je dis bien/j’ai bien dit, (e) sottolineo, подчеркиваю, (y) subrayo, die Betonung liegt auf/ich betone). L’analisi dell’uso di queste sequenze permette anche di identificare i meccanismi specifici dell’attenuazione argomentativa. Il primo tipo di ele13 È proprio impossibile! E questo in un giornale nazionale... E vergognoso per le Izvestia! E per Yanina Sokolovskaia! 14 Tale precisazione può avvenire retrospettivamente o essere parenteticamente integrata nell’enunciato.

RINFORZO ARGOMENTATIVO E ATTENUAZIONE ARGOMENTATIVA

339

menti attenuanti è costituito, come nel caso del rinforzo, dalle indicazioni di quantità/ grado (8a), estensione (8b) o dimensione temporale (durata o frequenza – 8c) relative allo stato di cose descritto nell’enunciato precedente: 8a. Bitte wahren Sie das Urheberrecht. Kurzzitate sind zulässig – die Betonung liegt auf kurz (daher gekürzt). Die Redaktion (Sequenza di ricerca: “die Betonung liegt auf kurz daher”) 8b. J’ai mis moi-même des années, et je ne plaisante pas, à admettre que, parmi ceux pour lesquels je prenais fait et cause, certains professaient une idéologie parmi celles qui sont à mes yeux les plus odieuses. J’ai bien dit “certains”. Ou “un certain nombre”. (Sequenza di ricerca: “odieuses. J’ai bien dit certains”) 8c. Los asesinos son asesinos, pero a veces nos devuelven una imagen de la justicia muy poco justa. Léase a Nietzsche. A veces15 es de un clarividente que asusta. Subrayo a veces. (Sequenza di ricerca: “asusta. Subrayo a veces”)

Il secondo tipo di attenuazione che può, nel caso di focalizzazione retrospettiva, creare l’effetto di inversione è legato all’inattualità dello stato di cose citato nella sequenza-argomento. Tale inattualità può derivare dalle indicazioni sul carattere approssimativo (9a), dalle modalità, in particolare epistemica (9b) e aletica (9c), dalle informazioni strumentali (9d) oppure anche dall’indicazione della soggettività (9e). 9a. Millionen Moslems leben zusammen, auch mit anderen, ohne sich jeden Tag zu prügeln. Irak, Afganistan […] auch Gewalt in Deutschland als Islamproblem darzustellen ist nicht nur kurzsichtig sondern DUMM! Das ist fast (ich betone fast) so ein Geschätz wie von den “Slavischen “Untermenschen” von vor 33. (Sequenza di ricerca: “ich betone fast so ein Geschätz”) 9b. C’è stato un omicidio perpetrato da un poliziotto ai danni di uno che forse, e sottolineo forse, si è dato un paio di sberle con altri ragazzi (forse ultrà). Ciò è innammissibile [...]. (Sequenza di ricerca: “e sottolineo forse, si è dato un paio”) 9c. Die Kinder können ihm fast straflos entzogen werden, obwohl Umgangsvereitelung sehr wohl sanktioniert werden kann. Die Betonung liegt auf *kann* weil kaum ein Richter das Instrumentarium was er zur Verfügung hat, gegen Frauen auch anwendet. (Sequenza di ricerca: “Die Betonung liegt auf *kann* weil kaum ein Richter”) 9d. Да, дорогая, ЕБН косвенно подтолкнул Шеви к началу акции. Подчеркиваю – косвенно. (Sequenza di ricerca: “акции Подчеркиваю косвенно”)16

15 Sottolineato nell’originale. 16 Sì, mia cara, Eltsin ha indirettamente

mente.

spinto Shevardnadze a iniziare questa azione. Sottolineo indiretta-

VAHRAM ATAYAN

340

9e. Aunque, ya que lo dices, si tú las aprecias y valoras, si tú crees en esas cosas, pues seguro que tienen todas las papeletas para ser otro disparate pseudocientífico de los que gustas. Esa ya es para mí (subrayo lo de mi subjetividad) una hipótesis provisional bastante consistente. (Sequenza di ricerca: “subrayo lo de mi subjetividad”)

Gli esempi discussi ci permettono di constatare la molteplicità e la varietà dei fattori semantico-pragmatici utilizzabili nell’attenuazione argomentativa. Essi possono essere immediatamente legati allo stato di cose in questione ma anche avere un carattere più generico e rispecchiare diversi atteggiamenti del parlante nei confronti del proprio enunciato. 5. Conclusioni Nella nostra analisi di una scelta di esempi autentici abbiamo constatato che i procedimenti linguistici generali del rinforzo e dell’attenuazione argomentativi, legati a fattori di (de)intensificazione in senso ampio, alla (contro)implicazione cotestuale e al carattere normale o sorprendente dello stato di cose, si possono manifestare tramite diversi meccanismi particolari, che abbiamo studiato in dettaglio per i primi due procedimenti. Abbiamo potuto identificare anche certi marcatori linguistici di tipo ‘congiunzione coordinante + pronome dimostrativo generico’ o indicatori metalinguistici di focalizzazione retrospettiva che segnalano frequentemente nei testi analizzati l’applicazione del rinforzo argomentativo o la disambiguazione tra l’attenuazione e l’inversione argomentativa. A conclusione riassumiamo i procedimenti e i meccanismi identificati nella seguente tabella:

RINFORZO ARGOMENTATIVO E ATTENUAZIONE ARGOMENTATIVA

341

Fenomeno

Procedimento

Meccanismi identificati nel corpus

Rinforzo

Rinforzo scalare

- Indicazione dell’alto grado - Durata, estensione ecc. (es. 4a-4c) - Valutazione dell’atteggiamento (es. 4d)

Controimplicazione nel cotesto

- Realizzazione dello stato di cose nonostante una causa contraria attuale (es. 5) - Realizzazione dello stato di cose dovuta a una causa insufficiente (es. 6) - Contrasto tra la posizione/il ruolo e l’opinione espressa/l’azione (es. 7)

Constatazione del carattere sorprendente Attenuazione

Attenuazione scalare

- Indicazione del basso grado (8a) - Durata, estensione ecc. (8b, 8c) - Inattualità approssimativa (9a), epistemica (9b), aletica (9c), strumentale (9d), soggettiva (9e)

Implicazione nel cotesto Constatazione del carattere normale

Bibliografia Anscombre, Jean-Claude & Oswald Ducrot (1983). L’argumentation dans la langue. Paris: Kimé. Atayan, Vahram (2004). Structures macroscopiques de l’argumentation dans l’analyse du discours – dialogicité, polyphonie et modificateurs réalisants. In: Suso López, J. & R. López Carrillo (éd.). Le français face aux défis actuels. Histoire, langue et culture. Vol. I. Granada: Università di Granada, 531-543. Atayan, Vahram (2006). Makrostrukturen der Argumentation im Deutschen, Französischen und Italienischen. Frankfurt: Peter Lang. Atayan, Vahram (in stampa): Pourquoi une ‘équipe-surprise’ gagne (presque) toujours. In: Atayan, V. & D. Pirazzini (éd.). Actes de la section ‘Argumentation’ du XXX. Deutscher Romanistentag, Vienne, septembre 2007. Frankfurt: Peter Lang. Blumenthal, Peter (1990). Oppositive Sinnverknüpfung im Deutschen und Französischen. In: Gnutzmann, C. (Hrsg.). Kontrastive Linguistik. Frankfurt: Peter Lang, 33-68.

342

VAHRAM ATAYAN

Brandt, Margareta (1990). Weiterführende Nebensätze. Zu ihrer Syntax, Semantik und Pragmatik. Stockholm: Almqvist & Wiksell. Bruxelles, Sylvie (2002). Topoï lexicaux et analyse interactionnelle: une mise en perspective sur des données recueillies en situation institutionnelle. In: Eggs, E. (éd.): Topoï, discours, arguments. Stuttgart: Franz Steiner, 27-48. Carel, Marion (1995). Argumentation by exception. Journal of Pragmatics 24: 167-188. Ducrot, Oswald (1983). Opérateurs argumentatifs et visée argumentative. Cahiers de linguistique française 5: 7-36. Ducrot, Oswald (1984). Le dire et le dit. Paris: Ed. de Minuit. Ducrot, Oswald (1995). Les modificateurs déréalisants. Journal of Pragmatics 24: 145-165. Ducrot, Oswald et al. (1980). Les mots du discours. Paris: Ed. de Minuit. Eemeren, Frans H. van, Rob Grootendorst & Francisca Snoeck Henkemans (1996). Fundamentals of Argumentation Theory. Mahwah, N. J.: Lawrence Erlbaum. Ferrari, Angela (2003). Le ragioni del testo. Firenze: Accademia della Crusca. García Negroni & María Marta (2001). Les connecteurs espagnols ENCIMA / ADEMÁS: argumentation transgressive et argumentation normative. Langages 142: 41-56. García Negroni & María Marta (2003). Gradualité et réinterprétation. Paris: Harmattan. Kalokerinos, Alexios (1995). Even: How to make theories with a word. Journal of Pragmatics 24: 77-98. Koch, Peter & Wulf Oesterreicher (1994). Schriftlichkeit und Sprache. In: Günther, H. & O. Ludwig (Hrsgg.). Schrift und Schriftlichkeit. Berlin: de Gruyter, 587-604. Lo Cascio, Vincenzo (1991). Grammatica dell’argomentare. Strutture e strategie. Firenze: La Nuova Italia. Lötscher, Andreas (1988). Textgrammatik adversativer, restriktiver und konzessiver Satzverknüpfungspartikeln. Kopenhagener Beiträge zur germanistischen Linguistik 24: 65-97. Perelman, Chaim & Lucie Olbrechts-Tyteca (1958). Traité de l’argumentation – La Nouvelle rhétorique. Bruxelles: Université de Bruxelles. Pirazzini, Daniela (2006). Sobre la retórica de los Foros de Debate. In: Gil, A. & Ch. Schmitt (eds.). Retórica en las lenguas iberorrománicas. Bonn: Romanistischer Verlag, 65-86. Stati, Sorin (2002). Principi di analisi argomentativa. Retorica, Logica, Linguistica. Bologna: Pàtron. Toulmin, Stephen (1958). The Uses of Argument. Cambridge: Cambridge University Press.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 343-355 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING : A CROSS - GENERIC AND CROSS - CULTURAL ANALYSIS

GIULIANA DIANI

1. Introduction Biber and Finegan’s (1989) work in the area of interpersonal meaning suggests that academic discourse is characterized by the relative absence of markers of stance – “the lexical and grammatical expression of attitudes, feelings, judgments, or commitment concerning the propositional content of a message” (Biber & Finegan 1989: 93). But recent research has suggested a growing recognition that there is room for negotiation of personal stance within academic writing (e.g. Hyland 1999; Bondi 2002). As noted by Hyland (1999: 120), “the use of stance is an important aspect of professional academic discourse, conveying the field-specific expressive and interpersonal meanings which help readers to evaluate information and writers to gain acceptance for their work”. As Conrad and Biber (2000) note, studies into the ways that speakers and writers mark their personal stance have been carried out from many different perspectives, i.e. from descriptions of a single text type to investigations of large computer-based corpora. Much work has been devoted to the examination of the ways indicators of stance are employed in specific disciplines (e.g. Peck McDonald 1994; Bondi 1999; Hyland 2000; Charles 2004; Samson 2004a; Diani 2006). The expression of stance has also been investigated, with several studies specifically focusing on particular lexicogrammatical elements contributing to evaluation function within and across genres (e.g. Bondi 2002; Silver 2003, 2004). However, little attention has been paid to the relationship between disciplinary and cultural variation in the expression of stance in academic discourse. One of the extremely interesting effects of such a comparative research is to contribute to understanding how far the expression of stance is influenced by national “disciplinary culture” (Hyland 2000), or by national academic culture in general. It is from these latter considerations that the pres-ent study takes its lead. An analysis of personal stance may focus on different lexical or grammatical items (e.g. Hyland 1998; Hunston & Thompson 2000; Biber et al. 1999). First-person markers, for example, would be the most obvious subjective forms of authorial stance in both spoken and written academic discourse (e.g. Hyland 2001, 2002; FortanetGómez 2004; Samson 2004b; Fløttum 2005; Bondi 2007a). The focus of this paper is on a less obvious marker of personal stance. Following Francis, Manning and Hunston (1988), and Hunston and Sinclair (2000), we will

344

GIULIANA DIANI

examine the introductory ‘it’ patterns it v-link ADJ that-clause and it v-link ADJ toinf. clause in comparable corpora of research article openings, i.e. “the opening section up to and including the second paragraph of each article” (Silver & Bondi 2004: 121), and book review articles in English and Italian in the discipline of history1. The context of this analysis is provided by a number of previous studies (e.g. Biber, Conrad & Reppen 1998; Biber et al. 1999; Charles 2000; Hewings & Hewings 2002; Murphy 2004; Groom 2005), which looked at these phraseological patterns as particularly salient in academic English. These patterns are commonly used to express evaluations, and one aspect considered here is that although they may appear to be impersonal, “in a way that allows the writer to remain in the background”, as Biber et al. (1999: 976) put it, we understand quite well that the writer is the source of the comment. This is not a new finding: as Charles (2000: 48) points out, “although these patterns appear on the surface to be impersonal, the adjective choice opens up a space which the author can use to indicate the nature of his/her comment on what follows”. And as Murphy (2004: 213) notes, in this type of pattern the speaker does not ‘assume explicit responsibility’ for the attitude expressed towards the proposition that follows, so the evaluation may have an aura of objectivity about it rather than personal resonance. While the evaluation is not explicitly averred, there are nevertheless some reasons why on occasions there is no aura of objectivity around the pattern. This happens when the evaluator is modified by an adverbial […]. Subjectivity creeps back into the text through these adverbials.

The aim of this study is to extend these observations to cross-generic and cross-cultural analysis, with a view to illuminating generic and cultural variation in the use of these phraseological patterns. Given the now significant body of work demonstrating linguistic variation within and across disciplines and genres, and given the broad consensus that such variation is not arbitrary but always rhetorically motivated (Hyland 2000), it is plausible to suppose that different genres and cultures might make differential use of these phraseological resources. It is precisely this possibility that the present study aims to investigate. After a brief presentation of the materials and the procedures adopted for this study, we will provide a preliminary overview of the pattern/meaning associations for these patterns across genres and cultures. The overview will include a cross-cultural comparison of selected lexical elements.

These patterns are phraseological units in which the dummy subject pronoun it is followed by a link verb such as be, become or seem, and adjective or noun group, and a finite or non-finite that-clause, to-infinitive clause, wh-clause or -ing clause. (Francis, Manning & Hunston 1998; Hunston & Sinclair 2000). For reasons of length, only the first two patterns are discussed here.

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING

345

2. Materials and methods The analysis is based on four small specialized corpora of book review articles and research article openings, which have been designed to study academic writing in the discipline of history in different cultural contexts – English and Italian. We made use of the following corpora: a) a corpus of 76 HIstorical Book Review Articles in English (Engl. HIBRA) published in five British and American academic journals spanning the years 19992005 (consisting of 304,981 words)2 . b) a corpus of 41 HIstorical Book Review Articles in Italian (It. HIBRA) published in three Italian academic journals spanning the years 1999-2005 (consisting of 189,346 words)3. c) a corpus of 280 historical research article openings in English (Eng. Openings) published in ten British and American academic journals spanning the years 19992000 (consisting of 95,682 words)4 . d) a corpus of 310 historical research article openings in Italian (It. Openings) published in eight Italian academic journals spanning the years 1999-2001 (consisting of 97,513 words)5 . The corpora are of different sizes because they were originally compiled for other purposes. All frequency data reported in this paper will be presented as normalised figures, calculated per thousand words. The two genres selected for analysis were judged suitable for present purposes as their very specific status in the field of genre studies. Keeping in mind the basically dialogic and argumentative nature of academic discourse, both the book review article and the research article represent the most distinguished channel of knowledge dissemination within the specific scientific community. Within an academic context, they play a crucial role in the process of knowledge construction and discussion by providing a forum in which academics can set out their views in the form of arguments. More specifically, focusing on the status of the research article openings in terms of the main function of the research article, Bondi (2007b: 72) sees the first two paragraphs of the research article introduction as offering “material that showed the starting point of the article, but also the direction taken, the dynamics of the beginning section of the text”. Like a book review article, a research article opening identifies a ‘research space’ for the writer’s own views within a disciplinary debate, and the creation of a research space 2 The journals considered are: Labour History Review, Journal of Interdisciplinary History, American Historical Review, Gender & History, Journal of American History. 3 The journals considered are: Meridiana, Passato e Presente, Quaderni Medievali. 4 The journals considered are: Labour History Review, Historical Research, Gender & History, Journal of European Ideas, Journal of Medieval History, Journal of Interdisciplinary History, Journal of Social History, Studies in History, American Quarterly, American Historical Review. 5 The journals considered are: Passato e Presente, Quaderni Medievali, Il Pensiero Politico, Intersezioni, Meridiana, Società e Storia, Studi Medievali, Dimensioni e Problemi della Ricerca Storica.

346

GIULIANA DIANI

is realized by a variety of voices that indirectly help establish the writer’s voice and construct a “niche” for his/her claims (Swales 1990). The choice of the disciplinary area for this study – history – is linked to its position in the field of disciplinary discourses. As Bondi (2007b: 68-69) observes, history stands out for the obvious tensions between narrative and argument in the basic structure of discourse. [...] The writer does not only interpret events: he or she also argues for his or her own position in the context of a disciplinary debate, in a complex dialogic pattern with the reader. History is not just account and interpretation of events, but also dialogic argumentation of the interpretation put forward.

The methodology adopted for this study combines a discourse and a corpus perspective. Discourse analysis contributes to the definition of pragmatic functions of introductory ‘it’ phraseological patterns under investigation, whereas corpus linguistics offers ways of looking at lexical patterns: in particular, using Scott (1998), we studied wordlist frequencies and concordances. These were worked out by comparing corpora to each other. Large-scale corpus research has established that the adjectives that occur in the introductory ‘it’ patterns fall into broad evaluative semantic meaning groups described by Francis, Manning and Hunston (1998) as “true/untrue, likely/unlikely, obvious, good/desirable, bad/undesirable, important/necessary, interesting/surprising”, and relabelled by Groom (2005) in terms of “adequacy, validity, desirability, difficulty, expectation, and importance”. Our cross-cultural comparison of evaluative adjectives in the patterns was conducted according to the evaluative semantic meaning groups indicated with the patterns in Groom (2005), who followed Francis, Manning and Hunston’s (1998) categories, and are reproduced here as Table I. It should be underlined here that adjectives were not assigned to particular meaning groups on an a priori basis. Rather, each attested phraseological sequence (i.e. attested strings of words describable by such formulae as it v-link ADJ that/to-inf.) was examined by looking at the co-text. The semantic values attached to each adjective by Groom (2005) were found to be appropriate for our data as well. As space does not permit detailed analysis of all finite verbs occurring in these phraseological patterns, the discussion presented here focuses on the most frequent in the data – is/was – and compares it with its equivalent in Italian: è/era.

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING

347

Evaluative category

(Engl.) it v-link ADJ that/ (It.) verbo+AGG+che

(Engl.) it v-link ADJ to-inf./ (It.) verbo+AGG+infinito presente

ADEQUACY

–

(Engl.) It was broad enough to incorporate …; (It.) È opportuno precisare…

VALIDITY

(Engl.) It is possible that…; (It.) È probabile che…; È chiaro che…

(Engl.) It is quite possible to write…; (It.) È dunque possibile rivedere…

DESIRABILITY

–

DIFFICULTY

–

EXPECTATION

(Engl.) It is not surprising here that…

IMPORTANCE

(Engl.) It is significant that…; (It.) È significativo che…

(Engl.) It is fair to characterize…; (It.) È giusto continuare a riflettere…

(Engl.) It is hard to believe…; (It.) È difficile negare … (Engl.) It is not therefore surprising to find…; (It.) È interessante notare… (Engl.) It is important to look…; it is not necessary to have…; (It.) Ed è importante osservare…

Table I: A classification system of pattern/meaning associations for it + is/was + ADJ + that/toinf. and their Italian equivalents across the corpora (based on Groom 2005).

In Table I, we see that not all the meaning groups occur in both patterns across the corpora. In particular, ‘adequacy’, ‘desirability’ and ‘difficulty’ are restricted to the it vlink ADJ to-inf./verbo+AGG+infinito presente patterns in both cultures. The fact that the it v-link ADJ to-inf. pattern is associated with the evaluative meaning groups ‘desirability’ and ‘difficulty’ echoes the findings by Biber et al. (1999: 720) in the Longman Grammar, who note that the most common adjectival predicates controlling extraposed to-clauses are (im)possible, difficult, hard, good, better, best, nice 6 .

6 Biber et al. (1999) talk of “extraposed complement clauses” rather than introductory ‘it’ patterns, and although they analyse and label meaning groups somewhat differently, their categories are sufficiently analogous for the above observations made.

348

GIULIANA DIANI

3. Introductory ‘it’ patterns across genres and cultures: the case of it + is/was + ADJ + that/to-inf. and their Italian equivalents The present analysis found significant variation in the distribution of the patterns under investigation across all four corpora. Space does not permit a full treatment of findings, so the present discussion will be restricted to dominant pattern/meaning associations for these phraseological patterns for each corpus only, and are reproduced as Table II. Table II: Distribution of dominant pattern/meaning associations for it + is/was + ADJ + that/toinf. and their Italian equivalents across the corpora (per 1,000 words). Corpus

Dominant pattern

Dominant meaning

Occurrences per 1,000 words

HIBRA (English)

It v-link ADJ to-inf.

Difficulty

0.19

HIBRA (Italian)

verbo AGG che/infinito presente

Validity

0.15

Openings (English)

It v-link ADJ that

Validity

0.19

Openings (Italian)

verbo AGG che/infinito presente

Validity

0.14

As Table II shows, ‘validity’ is the dominant meaning for the patterns it is/was ADJ that/è-era+AGG+che in historical research article openings in both cultures. If this finding is broadly in line with expectations – in registers which construct knowledge, such as research articles, the expression of degree of certainty is important: the “certainty parameter” (Hunston & Thompson 2000: 23), defining how certain the writer is of what is to follow – the finding that ‘difficulty’ is the dominant meaning for the pattern it is/was ADJ to-inf. in English historical book review articles is somewhat more unexpected. Given the evaluative purpose of the genre, it would be reasonable to expect the ‘desirability’ meaning group to be prominent here, which covers the basic good/bad evaluative polarity. But the present data find ‘desirability’ to be relatively dispreferred in the corpus, only constituting a frequency of 0.07 per 1,000 words in the pattern it is/was ADJ to-inf. in the whole corpus. Also the ‘validity’ meaning is not prominent (only a frequency of 0.12 per 1,000 words in the patterns it is/was ADJ that/to-inf. in the whole corpus). But this is not the case of the Italian HIBRA corpus where ‘validity’ is the main meaning group in both phraseologies. 3.1 Focus on the patterns in English and Italian historical book review articles When considering evaluative adjectives in the dominant pattern it is/was ADJ to-inf. in the English HIBRA corpus, we find that the most frequent items are associated with the notion of ‘difficulty’ like difficult (it is attested 23/60 times – the percentage

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING

349

amounts to 38.33%), impossible (12 occurrences/60 – 20%), not possible (12 occurrences/60 – 20%), and hard (7 occurrences/60 – 11.66%). Closer analysis reveals that this phraseology is frequently used for both negative assessments of specific theories: (1) Urton attempts to solve these problems by arguing that any given decimal khipu supported two readings, one numeric and one binary. It is difficult to imagine how these two functionally different yet materially isomorphic, overlapping semiotic systems could operate simultaneously on the same khipu. (Engl. HIBRA)

and for negative assessments of the reviewed book author’s argument: (2) Without such examinations, however, it is difficult to know what to make of Lu’s broadest arguments about how Old Shanghai’s traditions helped to facilitate the city’s new global status. (Engl. HIBRA)

Our data support the study by Hyland (2000: 61), which finds that “unlike research articles, outright criticism is not avoided in book reviews, indeed it is an integral feature of the genre, substantiating its claim to be a scholarly form of writing”. Along the same lines, Giannoni (2002: 356) points out that “conflict is the common thread that holds this genre together”. In explaining this finding, we cite evidence from our analysis, which shows that these adjectives frequently occur within the vicinity of some marker of counter-claims, such as however, but, yet so as to create pattern of contrast/ conflict. This result provides a particularly clear illustration of the role of the reviewer as academic arguer. (3) Top-down planning on a large scale did, of course, occur during World Wars I and II. But the postwar reaction to the distended wartime state made it clear that Seeing Like a State Scott-style was an anomaly born of emergency conditions, and it found little nourishment in American cultural soil. Aside from wartime, the American state-led programs that come closest to Scott’s high modernist model are the Interstate Highway System and NASA. But, for all their scale and cost, it is difficult to see either as an oppressive instance of topdown state planning. (Engl. HIBRA)

If we move to cross-cultural comparison, Italian HIBRA shows a different trend. The data reveal that ‘validity’ rather than ‘difficulty’ (where only 2 occurrences of the sequence non è più possible +inf. occur) is the main meaning group in the corpus, with a frequency of 0.15 per 1,000 words in the patterns verbo+AGG+che/infinito presente in the whole corpus, followed by both ‘desirability’ and ‘importance’, showing a similar frequency (0.05 per 1,000 words). The analysis reveals that the most frequently occurring evaluative adjective in the verbo+AGG+che pattern is vero that is attested 13/22 times (59.09%). Interestingly, the data show that the ‘validity’ sequence verbo+vero+che often pairs a concessive clause with a counter-claim:

350

GIULIANA DIANI

(4) Ci sono però altri risvolti del panarabismo che vanno segnalati. AJ è causa ed effetto dell’individuazione di un bacino economico e politico di estrema rilevanza. Forse non è del tutto corretto, come fa Della Ratta, affermare che soltanto dalla seconda metà degli anni novanta nella regione si è iniziato a ragionare in termini di mercato. È, però vero che solo allora si sono create quelle condizioni sociali, politiche, economiche, che hanno permesso di fare del “popolo arabo” un mercato unico a tutti gli effetti, di grande interesse per i governi e per le imprese, per il suo potenziale politico e per la possibilità di sfruttare le economie di scala a livello distributivo. (It. HIBRA)

The examples indicate that, although there is a significant variation across the two comparable corpora in terms of parameters of evaluation (English HIBRA seems to privilege the notion of ‘difficulty’, whereas Italian HIBRA favours ‘validity’), both English and Italian historical reviewers display a greater concern with long sequences of argumentative dialogue between the reviewed book author and themselves. From the point of view of the argumentative development of discourse in the genre, counterclaiming or pointing out gaps in existing research (including, of course, the book under discussion) contributes to representing the debate reviewers build not only with reviewed authors but also within the disciplinary area. 3.2 Focus on the patterns in English and Italian historical research article openings As Table II shows, ‘validity’ is the dominant meaning for the patterns it v-link ADJ that/ verbo+AGG+che in the two comparable corpora of openings in both cultures. In the Italian corpus ‘validity’ is also the dominant meaning for the pattern verbo+AGG+infinito presente. In English openings, the ‘validity’ sequence it is clear that is the most frequent (is attested 8/19 times – 42.10%), followed by it is (un)true that (3/ 19 – 15.78%), confirming the findings by Biber et al. (1999: 672) in the Longman Grammar, who note that “clear, (un)likely, (im)possible, true are the four most common adjectival predicates controlling extraposed that-clauses in LSWE Corpus”. In Italian openings, on the other hand, the ‘validity’ sequences è vero che and è possible+infinito presente are the most frequent (è vero che is attested 4/7 times – 57.14%; è possibile+infinito presente 7/8 – 87.5%). These results point to the possibility of a common English and Italian pattern involving markers of epi-stemic stance (evaluators of truth, certainty or likelihood). Cross-cultural comparison provides a particularly clear illustration of how English and Italian use the it v-link ADJ that and verbo+AGG+infinito presente patterns, respectively, to give two different types of ‘validity’ assessment, confirming the distinction made in Francis, Manning and Hunston (1998) between ‘likely’ and ‘obvious’: a) one providing evaluations based on appeals to possibility/likelihood that characterize the Italian corpus of openings:

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING

351

(5) Utilizzando una serie di testi che coprono un arco di tempo che va dall’XI al XX secolo, è possibile osservare la costante presenza di questo santo nella credenza popolare e la sua fortuna nel corso dei secoli. (It. Openings)

b) the other based on evidence and appealing to notions of obviousness that characterize the English corpus of openings: (6) […] On Crimes and Punishments appealed alike to sovereigns, statesmen and philosophes. In this regard, the work held great importance for the English intellectual environment, where Beccaria’s arguments drew wide and sustained acknowledgement from many sections of society. This is not to say that criticism of English penal practice and theory was unheard of before Beccaria’s work appeared. There are many examples of criticism being levelled both at the practices of English punishment and at the principles which supported them before 1764. Nevertheless, it is clear that the work was eagerly adopted, most obviously by lawyers and the rising middle classes, as a declaration of the fundamental principles that ought to underpin the application of the penal sanction in an ‘improved’ civilisation. (Engl. Openings)

If we consider (6) above, where the use of the contrastive connector nevertheless signals a counter-claim, allowing the writer to show contrast and signalling his opinion, we can see that in the research article opening, as Bondi (2007b) puts it, the writer is not only involved in interpreting historical sequence events, but in dialogically arguing a claim, so as to place it in the context of a debate, signalling the importance of his/her interpretation, rather than that of the subject matter. 3.3 A lexical case study: the lemma surprise in English Further reflections on cross-cultural analysis can be offered by studying the concordances of selected lexical elements. The choice of the lemma to be analysed falls on surprise (i.e. it is not surprising that, it is not a surprise that, surprisingly). When we analyse the evaluative adjectives used in the patterns across the two cultures, we find that most of them are equivalent (i.e. true/vero, possible/possibile, clear/ chiaro etc.). But this is not the case of surprising that occurs only in the two English corpora, with a frequency of 0.12 per 1,000 words in HIBRA corpus and 0.11 in English Openings. Both genres show a preference for negative polarity (not surprising, no surprise, not surprisingly), constituting 73.68% of all occurrences of the lemma surprise in the whole corpora (28 occurrences/38 occurrences) against 26.31% showing positive polarity (10 occurrences/38). (7) The list of Butterfield’s accomplishments and services to the field is so long that it is perhaps not surprising that he is generally also believed to have founded the Cambridge history of science committee, the body in whose activities the institutionalization of the subject at Cambridge is ultimately rooted. (Engl. Openings)

GIULIANA DIANI

352

(8) It is no surprise that those of us who make it our business to study the distant past should dwell so obsessively upon the written word. Written texts, after all, are far and away our most abundant resource for understanding the long defunct people and societies that constitute the subject of our investigations. (Engl. Openings) (9) Not surprisingly, books in these two categories differ from each other markedly on such basis issues as the influence of imperialism on China’s modern history. What must be stressed, however, is that […] (Engl. HIBRA)

As can be seen, the main function of the various modalizations of the lemma surprise is that of highlighting the expectedness of the statement in itself as in (7); it also, however, predicts that the obvious nature of the statement will lead to further elaboration or explanation that may be derived from it as in (8) and also contrasts sequences as in (9). The adjective surprising is also used to indicate a gap in the literature, by emphasizing it. An illustration is provided in example (10). (10) It is rather surprising here that a chronology of urbanization is not clearly related to a similar chronology of economic history. For example, little is made of the past role and heritage of the oil industry, of the military, both U.S. Navy and armaments industry, and of L.A. (Engl. HIBRA)

On the whole, such results support Bondi and Mazzi’s (2007) view that markers of (un)expectedness are crucial in English academic discourse. In terms of writer’s positioning, they observe, “markers of (un)expectedness can be used meta-discursively to signal ‘engagement’, i.e. resources by which the author negotiates (engages with) the various convergent or conflicting positions activated in the text” (2007: 131). One interesting question still unanswered is why no Italian equivalent of surprising occurs in the Italian corpora. We would presume the persuasive intent of reviewers and researchers is the same in both cultures, yet obviously their strategies for convincing readers of the plausibility of their interpretations may differ. 4. Conclusions The brief overview of academic phraseology across genres and cultures has shown that the evaluative meaning associations with the two grammar patterns it is/was ADJ that and it is/was ADJ to-inf. and their Italian equivalents vary across the four corpora studied here. We have seen that in historical book review articles parameters of evaluation in the two patterns under examination are different across the two cultures. On the whole, the English corpus privileges the notion of ‘difficulty’, that is frequently used for negative assessments. English reviewers present themselves as struggling to be both appropriately critical and fair. The Italian corpus, on the other hand, favours ‘validity’. Italian reviewers predominantly use these patterns to assess the validity of the reviewed

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING

353

book author’s ideas of the historical record. Both English and Italian writers in historical research article openings, in contrast, show a greater concern with ‘validity’. Italian historians tend to provide evidence on notions of ‘likelihood’, while writers in English on notions of ‘obviousness’. We have attempted to show how the patterns under investigation contribute to the construction of the reviewer’s or researcher’s argumentative position, in a way that, though apparently impersonal, you can discern their voices. This seems to confirm Murphy’s (2004: 219) remark that “it is largely impossible for a writer to remain in the background […] and what appears to be impersonal is merely a mask which the writer soon sheds”. In both corpora of book review articles and openings, history is placed under discussion within a disciplinary debate: both reviewers and writers interpret historical events and argue their claims in a debate within the discourse community. References Biber, Douglas, Susan Conrad & Randi Reppen (1998). Corpus linguistics: investigating language structure and use. Cambridge: Cambridge University Press. Biber, Douglas & Edward Finegan (1988). Adverbial stance types in English. Discourse Processes 11: 1-34. Biber, Douglas & Edwars Finegan (1989). Styles of stance in English: Lexical and grammatical marking of evidentiality and affect. Text 1: 93-124. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad & Edward Finegan (1999). Longman Grammar of Spoken and Written English. London: Longman. Bondi, Marina (1999). English across genres: language variation in the discourse of economics. Modena: Il Fiorino. Bondi, Marina (2002). Attitude and episteme in academic discourse: Adverbials of stance across genres and moves. Textus 15/2: 249-264. Bondi, Marina (2007a). Historical research articles in English and in Italian: a cross-cultural analysis of self-reference in openings. In: Bertuccelli Papi, M., G. Cappelli & S. Masi (eds.). Lexical complexity: theoretical assessment and translational perspectives. Pisa: Plus, 65-83. Bondi, Marina (2007b). Authority and expert voices in the discourse of history. In: Fløttum, K. (ed.). Language and discipline perspectives on academic discourse. Newcastle: Cambridge Scholars Publishing, 66-88. Bondi, Marina & Davide Mazzi (2007). Adverbial marking of stance and disciplinary culture in academic research articles. In: Hornung, A. & C. Robustelli (a cura di), Vivere l’intercultura – Gelebte Interkulturalität. Festschrift für Hans Drumbl zum 60. Tübingen: Stauffenburg, 121-135. Charles, Maggie (2000). The role of an introductory ‘it’ pattern in constructing an appropriate person. In: Thompson, P. (ed.). Patterns and perspectives: insights into EAP writing practice. CALS: The University of Reading, 45-69. Charles, Maggie (2004). The construction of stance: a corpus-based investigation of two contrasting disciplines. Unpublished doctoral dissertation. UK: The University of Birmingham.

354

GIULIANA DIANI

Conrad, Susan & Douglas Biber (2000). Adverbial marking of stance in speech and writing. In: Hunston, S. & G. Thompson (eds.). Evaluation in text: authorial stance and the construction of discourse. Oxford: Oxford University Press, 57-73. Diani, Giuliana (2006). Reviewer stance in academic review articles: a cross-disciplinary comparison. In: Del Lungo Camiciotti, G., M. Dossena & B. Crawford Camiciotti (eds.). Variation in business and economics discourse: diachronic and genre perspectives. Roma: Officina Edizioni, 139-151. Fløttum, Kjersti (2005). The self and the others: polyphonic visibility in research articles. International Journal of Applied Linguistics 15/1: 29-44. Fortanet-Gómez, Immaculada (2004). The use of ‘we’ in university lectures: reference and function. English for Specific Purposes 23/1: 45-66. Francis, Gill, Elizabeth Manning & Susan Hunston (1988). Collins COBUILD grammar patterns 2: nouns and adjectives. London: HaperCollins. Giannoni, Davide S. (2002). Hard words, soft technology. Criticism and endorsement in the software review genre. In: Gotti, M., D. Heller & M. Dossena (eds.). Conflict and negotiation in specialised texts. Bern: Peter Lang, 335-363. Groom, Nicholas (2005). Pattern and meaning across genres and disciplines: an exploratory study. Journal of English for Academic Purposes 4: 257-277. Hewings, Martin & Ann Hewings (2002). It is interesting to note that…: a comparative study of anticipatory ‘it’ in student and published writing. English for Specific Purposes 21: 367-383. Hunston, Susan & John Sinclair (2000). A local grammar of evaluation. In Hunston, S. & G. Thompson (eds.). Evaluation in text: authorial stance and the construction of discourse. Oxford: Oxford University Press, 74-101. Hunston, Susan & Geoffrey Thompson (eds.) (2000). Evaluation in text: authorial stance and the construction of discourse. Oxford: Oxford University Press. Hyland, Ken (1998). Hedging in scientific research articles. Amsterdam: John Benjamins. Hyland, Ken (1999). Disciplinary discourses: writer stance in research articles. In: Candlin, C. & K. Hyland (eds.). Writing: texts, processes and practices. London: Longman, 99-121. Hyland, Ken (2000). Disciplinary discourses: social interactions in academic writing. Harlow: Longman. Hyland, Ken (2001). Humble servants of the discipline? Self-mention in research articles. English for Specific Purposes 20: 207-226. Hyland, K. (2002). Authority and visibility: authorial identity in academic writing. Journal of Pragmatics 14: 1091-1112. Murphy, Amanda C. (2004). A hidden or unobserved presence? Impersonal evaluative structures in English and Italian and their wake. In: Partington, A., J. Morley & L. Haarman (eds.). Corpora and Discourse. Bern: Peter Lang, 205-220. Peck McDonald, Susan (1994). Professional academic writing in the humanities and social sciences. Carbondale and Edwardsville: Southern Illinois University Press. Samson, Christina (2004a). Taking a stance: evaluative adjectives in a corpus of written economics lectures. In: Anderson, L. & J. Bamford (eds.). Evaluation in oral and written academic discourse. Roma: Officina Edizioni, 151-162.

INTRODUCTORY ‘IT’ PATTERNS IN ENGLISH AND ITALIAN ACADEMIC WRITING

355

Samson, Christina (2004b). Interaction in written economics lectures: the meta-discursive role of person markers. In: Aijmer, K. & A-B. Stenström (eds.). Discourse patterns in spoken and written corpora. Amsterdam: John Benjamins, 199-216. Scott Michael (1998). Wordsmith Tools. Oxford: Oxford University Press. Silver, Marc S. (2003). The stance of stance: a critical look at ways stance is expressed and modeled in academic discourse. Journal of English for Academic Purposes 2: 359-374. Silver, Marc S. (2004). The argumentational consequences of disciplinary constraint: the case of naturally in American discourse. In: Bondi, M., L. Gavioli & M. Silver (eds.). Academic discourse, genre and small corpora. Roma: Officina Edizioni, 139-157. Silver, Marc S. & Marina Bondi (2004). Weaving voices: a study of article openings in historical discourse. In: Del Lungo Camiciotti, G. & E. Tognini Bonelli (eds.). Academic discourse – New insights into evaluation. Bern: Peter Lang, 141-159. Swales, John (1990). Genre analysis. English in academic and research settings. Cambridge: Cambridge University Press.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 357-372 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION MARION GREIN

Introduction The article aims at a comparison of the linguistic means of argumentation in German and Japanese. The first chapter will give definitions of both, argument and argumentation. Furthermore, some basics of argumentation theory are summarized. The second chapter is concerned with the means of language used in argumentation. These can be differentiated into several linguistic devices or markers, i.e. the lexical selection, the topic-comment or thematic structure, illocutionary markers and various connectors. Here, I will confine myself to the analysis of various connectors or discourse markers used in argumentative action games. The goal of my study was to compare these devices within the languages of German and Japanese. Due to my ample data set consisting of German and Japanese refusals (Grein 2007a), the study will be limited to justificatory argumentation in which one person at a time seeks to justify his or her refusal to undertake a specific action. The given role-play situation consists of a directive in which the test persons are asked to do some extra non-paid work on a weekend. The argumentation being that extra non-paid work is acceptable under special circumstances. Altogether 200 German and 200 Japanese test persons refused to work voluntarily and almost 70% of them, however more Germans than Japanese, argued why they would not work during their leisure without (financial) compensation. Both, German and Japanese, possess equivalent causal conjunctions, yet, next to using conjunctions, Japanese makes frequent use of conjunctive converbal constructions and nominalizations. 1. Arguments, Argumentation and the Minimal Action Game Within the field of linguistics, there are numerous definitions for the concepts of argument and argumentation (van Eemeren & Grootendorst 1992: 60ff.). In the year 1990, Lumer had already outlined a number of 18 diverging definitions (Lumer 1990: 26ff ). The most prominent definition harkens back to van Eemeren, Grootendorst & Kruiger (1987: 7), who define argumentation as a social, intellectual, verbal activity serving to justify or refute an opinion, consisting of a constellation of statements and directed towards obtaining the approbation of an audience, a form of interaction. An argument is, thus, often defined as “any exchange of information centered on an avowed disagreement” (Gilbert 1997: 104).

358

MARION GREIN

Or as van Eemeren et al. (1996: 5) put it: Argumentation is a verbal and social activity of reason aimed at increasing (or decreasing) the acceptability of a controversial standpoint for the listener or reader, by putting forward a constellation of propositions intended to justify (or refute) the standpoint before a rational judge.

Taking this definition into account, most propositions, oral and written, can be interpreted as a kind of argumentation. Weigand (1999: 54) argues: From a functional point of view it seems absurd to deny that we would not always try to achieve our communicative purposes by more or less effective means. Thus language use in dialogic action games would always be inherently persuasive from the very beginning.

Typical situations are the request of a child to stay up late, the attempts of an employee to increase his salary, the efforts of a politician to convince the opposition of any innovations or the endeavour of a lawyer to find an accused not guilty. Thus, most dialogic conversations or dialogic action games are arguments justifying one’s actions. Just as we are confronted with numerous definitions of argument and argumentation, there are frequent theoretical approaches. Van Eemeren & Grootendorst (1992: 6ff.) outline ten theoretical approaches. Bücker (2004: 16) summarizes thirteen approaches. In Weigand’s (2003, 2006, 2008, forthc. this volume) holistic approach argumentation is not interpreted “as a game in the abstract, but starts from human beings’ mind” (Weigand, this volume). Weigand (this volume) distinguishes between games of argumentation and moves of arguing. She considers the classical game of argumentation as a representative game of negotiation about the world between thesis and antithesis: thesis

antithesis

claim to truth

↑arguments

Fig. 1: Representative game of argumentation

Weigand (this volume) conceives moves of arguing, i.e. of providing reasons, as representative subordinate moves which can appear in every game, not only in support of representative claims but also in support of claims to volition: representative game

directive game

claim to truth

claim to volition

↑arguments

Fig. 2: Arguing in representative and directive games

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION

359

Both, the speaker with his or her specific interest as well as the hearer who either accepts or refutes the position of the speaker, verbalize their positions with the means of arguments. In any argumentative dialogic action game, the speaker is aware that the hearer might have a divergent opinion concerning the content of his or her argumentation. Thus, the speaker tries to verbalize his or her statements, his argumentation, by such means that the hearer will consider the argumentation as comprehensible, acceptable or at least admissible (van Eemeren et al. 1996). He or she needs to be convincing. The minimal action game analyzed here could be considered as a directive action game, in which a superior requests an employee to work during the weekend, nil-paid. The analysis, however, concerns the following representative game: thesis argument of the superior: It is perfectly all right to ask an employee to work on weekends in case of a special occasion.

antithesis argument of employees: - not ethical - impossible without financial compensation - impossible when other obligations or engagements exist - indirect argument

Fig. 3: The game of argumentation analyzed

Argumentation is understood as a process in which speakers, with their very own cognition, emotions, cultural backgrounds, beliefs etc., depict a specific situation with verbal means. Weigand (this volume) states: “Argumentation starts from divergent views. Different claims to truth are expressed by representative speech acts and then negotiated in argumentative dialogues”. Argumentation is not only subject to semantic content; it also exhibits a particular linguistic structure, including particular verbal means. Both, semantic content and linguistic structure are primarily dependent on contextual and pragmatic factors, i.e. the specific situation and the social distance between the speakers. Thus, the argumentation of the child who wants to stay up late, the attempts of the employee and the endeavour of the lawyer mentioned above will surely have a different structure and different verbal means. Furthermore, nonverbal communication is accompanied with verbal communication in argumentation but will be disregarded in this article. Van Eemeren & Grootendorst (1992: 6ff) – like many other approaches – operate with argument schemes. They differentiate symptomatic, analogical and causal argumentation. Without going into the controversial debate between philosophical ideals and rhetoric, I will confine myself to four basic types of argumentation schemes: moral, plausible, rational and tactic argumentation which are often discussed as fallacies (van Eemeren & Houtlosser 2008).

360

MARION GREIN

moral

Basic types of argumentation

tactical

rational

plausible

Fig. 4: Types of argumentation

It is possible to appeal purely on moral or ethical grounds. Common values and principles of society are mentioned. Moreover, ethical argumentation predominantly mentions persons of high standing as prominent examples of public opinion (‘argument of authority’). Moral argumentation is often found in China when people refer to Confucius. In plausible argumentation, arguments can be based on the possibility of adverse publicity or potential damaging effects of a particular action. Plausible evidence is often based on published data, observed experiences of others, public opinion and common sense (‘sanity and reason’). Plausible argumentation is difficult to distinguish from rational argumentation. In rational argumentation, however, the persuasiveness of facts (statistics, research data, and verifiable hypotheses) is predominant. In due form, the speaker suggests several competing counter-positions and then proceeds to make a rational choice between those positions, based on factual evidence. The conclusion often is not offered as the ultimate truth but rather as the most likely one out of several choices. Rational argumentation appeals to the listener’s intellect and appears to be objective. Again, tactical and rational argumentation is difficult to distinguish. Tactical argumentation, in fact, seems to be close to some types of fallacies. Instead of arguing the speaker emphasizes his or her alleged supremacy. He or she pretends to consider counter-positions objectively, yet, dismisses all counter-arguments as incomparable exceptions to the rule. 2. Verbal Means of Argumentation The means of language used in argumentation can be differentiated into a several linguistic devices or markers, i.e. the lexical selection, the topic-comment or thematic structure, illocutionary markers and various connectors (van Eemeren & Grootendorst

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION

361

1992; Eggs 2001). Furthermore, the argumentative function can be implicit, and thus linguistically unmarked (van Eemeren & Grootendorst 1981). On top of that, nonand paraverbal devices can supersede the necessity for explicit marking of the argumentative function. Here, I will take a look at the various connectors used in argumentative discourse. These markers are markers of adverbial subordination, i.e. conjunctions, causal adverbs and prepositional noun phrases. They link two propositions, usually indicating a causal relation. The argumentative function can, however, be implicit and thus overt marking is unnecessary. Moreover, non- and paraverbal means can be sufficient when the relation can be implied. Some examples will illustrate the verbal means: Conjunctions: (1) Ich möchte abends eine Stunde länger aufbleiben, weil alle meine Freundinnen auch erst um 21 Uhr ins Bett müssen. I wanna go to bed late because all of my friends don’t have to be in bed before 9 o’clock!

Causal adverb: (2) Ich möchte abends eine Stunde später ins Bett. Meine Freundinnen müssen nämlich auch alle erst um 21 Uhr ins Bett. [further German adverbs: daher, deshalb, darum (‘that’s why’) trotzdem (‘in spite of it, nevertheless’)]. I wanna got to bed late. You see/you know, all my friends don’t have to go to bed before 9 o’clock.

Prepositional noun phrase: (3) Aufgrund meines 12 Geburtstages wäre es nur fair auch bis 21 Uhr aufbleiben zu dürfen Because of my 12th birthday it would only be fair if I could stay up till 9 o ’ c l o c k.

Juxtaposition (not overtly marked):

(4) Ich möchte abends eine Stunde später ins Bett. Meine Freundinnen müssen auch alle erst um 21 Uhr ins Bett. I wanna go to bed late. All my friends may stay up till 9 o’clock.

Simple construction: Due to the fact that all these utterances are part of a dialogic action game, it is perfectly all right just to give the justification in a simple construction. The justification is given in reference to an initially uttered proposition. (5) Meine Freundinnen müssen auch alle erst um 21 Uhr ins Bett. All my friends may stay up till 9 o’clock!

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION

363

3.1 Conjunctions

80%

56% German

36,20%

Japanese

40%

Fig. 5: Conjunctions

Although Japanese has a more elaborated set of markers than German (i.e. converbs and nominalization), the Japanese use more conjunctions than the Germans. It is furthermore notable that in 67,9% of all conjunctional constructions the conjunction node (‘because’) is used. Actually, node has been grammaticalized into a conjunction, historically being the nominalization marker no and the converbal form de of the copula desu. A second causal conjunction, next to node, is kara. Yet, constructions with node are considered to be more objective than those with kara, and thus are, obviously, more frequent. Constructions marked with node put their emphasis on the result whereas constructions with kara focus on the reason or cause (Grein 1998: 158f ). In German, the conjunctions weil and da (‘because’) are most frequent. 3.2 Causal Adverbs Neither the German nor the Japanese data displayed any causal adverbs. This might be due to the (hierarchical) constellation of employee and employer. Causal adverbs seem to lack the appropriate politeness. Within different interpersonal constellations causal adverbs were used in both languages. 3.3 Prepositional Noun Phrases

1,54%

1,53% German

1,52% 1,50%

Japanese

1,50% 1,48%

Fig. 6: Prepositional noun phrase

Prepositional – or rather postpositional in Japanese – noun-phrases were equally rare in both languages.

364

MARION GREIN

3.4 Juxtaposition (unmarked)

24%

18,37% German

16%

Japanese 3,50%

8% 0%

Fig. 7: Juxtaposition

Whereas juxtaposition is the third most frequent means in German with 18,4%, it is hardly used in Japanese at all. In Japanese, a conversation between equals is strongly dependent on context and thus arguments and connectors are omitted. Yet, when talking to a superior the arguments and thus the verbal markers are obligatory (Grein 2007a, b). This is also reflected in the usage or rather the absence of simple declarative sentences in Japanese that will be presented in the next chapter. 3.5 Simple declarative sentence

37,24% 40% German 12,5%

Japanese

Fig. 8: Simple declarative sentence

As mentioned in section 3.4, the Japanese are obliged to mark their argumentation with connectors in the given situation. Thus, the employment of simple sentences â&#x20AC;&#x201C; in which the causal context is implicit â&#x20AC;&#x201C; is frequent in German (37,2%) and scarce in Japanese (12,5%). 3.6 Nominalization ( Japanese)

10%

German Japanese 0% 0%

Fig. 9: Nominalization

366

MARION GREIN

3.7 Converb

16,5%

20%

German Japanese 0% 0%

Fig. 10: Converb

Converbs are verb forms which are neither finite nor nominal (Grein 1998: 84). They are used in coordinative sequences of clauses or subordinate adverbial clauses. Tense, aspect and modal categories are dependent on the final finite verb. Coordinative sequence (11)

Asa morning

ni TEMP

okite wake-up:CONV

kohi coffee

o ACC

nonde, drink:CONV

gohan rice

o AKK

tabete, eat:CONV

shibun newspaper

o ACC

yonde, read:CONV

kaisha ni ikimasu. company DIR go:HON:PRES. I wake up in the morning, have some coffee, eat breakfast, read the newspaper and go to the company. (12)

Asa morning

ni TEMP

okite wake-up:CONV

kohi coffee

o ACC

nonde , drink:C O N V

gohan rice

o AKK

tabe t e eat:CONV

shibun newspaper

o ACC

yonde , read:CONV

kaisha ni ikimashita . company DIR go:HON: P A S T I woke up in the morning, had some coffee, ate breakfast, read the newspaper and went to the company.

The change of tense refers to all prior given converbs. Adverbial subordination: (13)

Sushi o tabetekara hon Sushi ACC eat:CONV book After having eaten Sushi, I will read a book.

o ACC

yomu. read:PRES

367

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION

(14)

Sushi o tabetekara hon Sushi ACC eat:CONV book While I am eating Sushi, I will read a book.

o ACC

yomu. read:PRES

(15)

Sushi o tabetekara Sushi ACC eat:CONV If I eat Sushi, I will read a book.

o ACC

yomu. read:PRES

hon book

Grein (1998) differentiates altogether 13 converbs in Japanese. In fact, German provides two infinite verb forms that can be considered as converbs: present participle and present perfect participle. (16)

Von about

ihren her

Tanzfähigkeiten abilities to dance

singend sing:CONV

betrat enter

sie she

den Saal (present participle) ART hall. While singing about her dancing abilities she entered the hall. (17)

Von about

ihren her

Tanzfähigkeiten abilities to dance

überzeugt convince:CONV

betrat enter

sie she

den Saal (present participle) ART hall. Being convinced that she could dance she entered the hall. (Grein 1998: 84)

Within the data-set, there were no converbal constructions in German but 33 (16,5%) occurrences in Japanese. Semantically, most of these constructions do not offer an antithesis but rather the speaker puts up an argument why he or she will not be able to work on the weekend. Indirectly, the speaker thereby indicates that his or her argument is superior to the employer’s argument. Direct argumentation: thesis

antithesis

argument

claim to truth

REPRESENTATIVE

NON-ACCEPTANCE

It is perfectly all right to work honorary at special occasions

Extra work has to be paid

Fig. 11: Representative game of direct argumentation

368

MARION GREIN

Indirect argumentation: DIRECTIVE

NON-CONSENT

‘You have to work this weekend’

‘No’

thesis

argument

claim to truth

REPRESENTATIVE

REPRESENTATIVE (NUNTIATIVE)

It is perfectly all right to work honorary at special occasions

I have an important engagement My mother is sick

Fig. 12: Representative game of indirect argumentation

Indeed, one could argue that these cases are no argumentations. Yet, I conceive them as indirect argumentation. The employee, by uttering his or her argument, implies that the employer’s argument is of less importance or even non-discussable. (18)

haha my mother

no GEN

joutai health

ga NOM

warukute bad:ADV:CONV

sono hi wa ikesou ni arimasen that day TOP go:POT:HON:PRES:NEG Since my mother’s health is quite bad, that day won’t work. (19)

sono that

hi day

wa TOP

to ikenai COM P go:POT:NEG:PRES

doushite how youji plan

mo too ga NOM

ikanai go:NEG:PRES atteka r a have:CONV

dekin dekimasen work can:HON:NEG:PRES Since on that day I have plans that can by no chance be cancelled, I can’t work.

369

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION

3.8 Question

10% 7% German Japanese 1% 0%

Fig. 13: Question

Again, this type of argumentation is considered as indirect argumentation. thesis

questioning the thesis

argument

question

claim to truth

claim to knowledge

REPRESENTATIVE

EXPLORATIVE

It is perfectly all right to work honorary at special occasions

Considering my salary, just a joke, right?

Fig. 14: Explorative game of indirect argumentation

Within the German data 13 employees put forward their argumentation by means of a question. In Japanese only two occurrences were found. Questions are not sufficiently marked for politeness. Concerning the contents of utterances, the Japanese examples are diplomatic, not really giving any argument, while in the German examples the antithesis is easily recognizable. (20)

sore that

wa TOP

watashi I

ni DAT

shika alone

dekinai can:NEG:PRES

koto na node shimau ka. thing because finish:PRES QU Is that a job that can only be done by me? ( 2 1 ) Bei meiner Bezahlung wohl With my salary probably Considering my salary rather a joke, right?

eher rather

ein a

SpaĂ&#x;? joke

370

MARION GREIN

(22)

In meiner Position? In my position In my position? Are you joking?

Ist Is

das that

ein a

Witz? joke?

4. Comparison

2% 2%

1% 7%

question 0%

nominalisation

8% 17%

converb juxtaposition

Japanese German

18% 13%

declarative sentence

37%

conjunction

36%

57%

40%

Fig. 15: Comparison verbal means

The Japanese prefer complex constructions and employ overt verbal markers, i.e. conjunctions and converbs (83%). In German, however, simple constructions, i.e. simple declarative clauses, juxtapositions and questions, are preferred. Within the German data merely 37,8% of the sentences are complex and overtly marked with an adverbial subordinator. Whereas we find ‘real’ argumentation within the German data, the Japanese rather employ the indirect strategy given in Fig. 12: DIRECTIVE ‘You have to work this weekend’

NON-CONSENT

‘No’

thesis

argument

claim to truth

REPRESENTATIVE

REPRESENTATIVE (NUNTIATIVE)

It is perfectly all right to work honorary at special occasions

I have an important engagement My mother is sick

Fig. 12: Representative game of indirect argumentation

THE VERBAL MEANS IN GERMAN AND JAPANESE ARGUMENTATION

371

A first glance at the types of argumentation displays that Germans tend to argue plausible and tactical while the Japanese favour implicit moral argumentation. What exactly is implicit moral argumentation? When talking to superiors, the social distance has to be taken into account. Thus, in the given interpersonal constellation, many Japanese do not argue against the employer’s proposition that working on an honorary basis is permissible on special occasions but rather adhere to society’s moral conventions of non-acceptance by using set phrases in which further obligations or even liabilities are mentioned. The establishment and maintenance of harmony is the most important value of Japanese society (cf. Grein 2008a: 195), in order to preserve or maintain harmony, each individual has to adhere to his or her obligations (jap. giri). Obligations are of greater moral value than other arguments. As mentioned before a set phrase like ‘the circumstances are a bit bad, but I have another obligation’ is considered a stronger argument than the superior’s argument. Therefore, Japanese games of argumentation need further analysis within different interpersonal relations, especially in constellations where the arguers possess the same social status. 5. Relevance The study has shown that both, semantic contents (argumentation type) and linguistic devices differ within the analyzed languages German and Japanese. Further analysis and comparison is indispensable since the given interpersonal constellation had an impact on the argumentation types. These findings are of relevance for argumentation theory, interactional linguistics and cross-cultural pragmatics, where speech acts uttered in identical interaction settings are compared (Blum-Kulka et al. 1989; Grein 2008b: 21). Results of crosscultural pragmatics are of relevance for second language research. The findings of the present study could and should be considered in second language instruction and textbooks. Structure, contents and devices diverge. Not considering these differences in a cross-cultural argumentation would lead to miscommunication. References Blum-Kulka, Shoshana, Juliane House & Gabriele Kasper (1989). Cross-cultural pragmatics: requests and apologies. Norwood, N.J.: Ablex Publishing Corporation. Bücker, Jörg (2004). Argumentationstheorie und interaktionale Linguistik. Münster: SASI – Studentische Arbeitspapiere zu Sprache und Interaktion. Eemeren, Frans H. van & Rob Grootendorst (1981). Verzwegen argumenten. Tijdschrift voor taalbeheersing 3: 130-156. Eemeren, Frans H. van, Rob Grootendorst & T. Kruiger (1987). Argumentatietheorie. Utrecht & Antwerpen: Het Spectrum. Eemeren, Frans H. van & Rob Grootendorst (1992). Argumentation, communication, and fallacies. A pragma-dialectical perspective. Hillsdale: Erlbaum.

372

MARION GREIN

Eemeren, Frans H. van et al. (1996). Fundamentals of argumentation theory. A handbook of historical backgrounds and contemporary developments. Mahwah: Erlbaum. Eemeren, Frans H. van & Peter Houtlosser (2008). Rhetoric in a dialectical framework. Fallacies as derailments of strategic manoeuvring. In: Weigand, E. (ed.). Dialogue and rhetoric. Amsterdam: John Benjamins, 133-151. Eggs, Ekkehard (2001). Argumentative Konnektoren und Textkonstitution. Am Beispiel von deduktiven und adversativen Strukturen. In: Cambourian, Alain (ed.). Textkonnektoren und andere textstrukturierende Einheiten. Tübingen: Stauffenburg, 71-90. Gilbert, Michael A. (1997). Coalescent argumentation. New Jersey: Lawrence Erlbaum Associates. Grein, Marion (1998). Mittel der Satzverknüpfung im Deutschen und Japanischen. Eine typologisch-kontrastive Analyse. Wiesbaden: Deutscher Universitäts-Verlag. Grein, Marion (2007a). Kommunikative Grammatik im Sprachvergleich. Die Sprechaktsequenz Direktiv und Ablehnung im Deutschen und Japanischen. Tübingen: Niemeyer. Grein, Marion (2007b). The speech act of refusal within the minimal action game – A comparative study of German and Japanese. In: Grein, M. & E. Weigand (eds.). Dialogue and Culture. Amsterdam & Philadelphia: John Benjamins, 95-114. Grein, Marion (2008a). Sociological concept and their impact on rhetoric. Japanese language concepts. In: Weigand, E. (ed.). Dialogue and rhetoric. Amsterdam: John Benjamins, 195-207. Grein, Marion (2008b). Der Sprechakt des Kompliments im interkulturellen Vergleich. Dialogue in and between different cultures. IADA online Publications, 15-32. Lumer, Christoph (1990). Praktische Argumentationstheorie. Theoretische Grundlagen, praktische Begründung und Regeln wichtiger Argumentationsarten. Braunschweig: Vieweg. Weigand, Edda (1999). Rhetoric and Argumentation in a Dialogic Perspective. In: Rigotti, E. (ed.). Rhetoric and argumentation. Tübingen: Niemeyer (Beiträge zur Dialogforschung 19), 53-69. Weigand, Edda (2003). Sprache als Dialog. Sprechakttaxonomie und kommunikative Grammatik. Tübingen: Niemeyer. Weigand, Edda (2006). Argumentation: The mixed game. Argumentation 20/1: 59-87. Weigand, Edda (2008). Rhetoric in the mixed game. In: Weigand, E. (ed.). Dialogue and rhetoric. Amsterdam: John Benjamins, 3-21. Weigand, Edda (forthc.) The argumentative power of words or how to move people’s minds with words. In: L’analisi linguistica e letteraria 16.

Glossary ACC ADV ART CONV COP DAT DESI GEN

Accusative Adverb Article Converb Copula Dative Desiderative Genitive

HON NEG NOM PAST POT PRES QU TEMP TOP

Honorifics Negation Nomination Past tense Potential Present tense Question marker Temporal marker Topic

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 373-384 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

A CONTRASTIVE STUDY OF THE INTRODUCTION SECTION GENRE OF ENGLISH AND PERSIAN MEDICAL RESEARCH ARTICLES AVISHAN MAHZARI

Introduction It has been difficult to avoid the word genre in language education in recent years. Many studies have been carried out to investigate the effect of genre awareness on producing straightforward and sound types of research articles. University students need to both write and comprehend research articles. Within this text type, the introduction section of research article is a genre in itself (Swales 1990, Bhatia 1993) as distinct from other sections of research articles. The introduction sections of research articles have been widely studied because, as Swales (1990) puts it, they are the most difficult part for writers, because they are forced with numerous options and decisions in this section: the amount of background knowledge, the authoritative versus sincere tone, the winsomeness of the appeal to readers, and directness of the approach they should incorporate into their writing. This study is an attempt to investigate the introduction section genre of American-English and Persian medical research article. The study aims to find out whether there is any difference in the introduction sections of these two languages as far as the frequency of moves and steps are concerned. Method Swales’ (1990) CARS model was utilized for the analysis. The model could effectively explain the introduction section genre of English and Persian medical research article. Swales (1990:141) posits a three move pattern for article introduction, as shown below. MOVE 1:

ESTABLISHING A TERRITORY

Step 1

Claiming centrality and/or

Step 2

Making topic generalization and/ or

Step 3

Reviewing items of previous research

MOVE 2:

ESTABLISHING A NICHE

Step 1A

Counter-claiming or

Step 1B

Indicating a gap or

Step 1C

Question-raising or

374

AVISHAN MAHZARI

Step 1D

Continuing a tradition

MOVE 3:

OCCUPYING THE NICHE

Step 1A

Outlining purposes or

Step 1B

Announcing present research

Step 2

Announcing principal findings

Step 3

Indicating RA structure

The Corpus Two hundred introduction sections of research articles were randomly extracted from three English and three Persian medical journals, (100 per each language). The data obtained for analysis were drawn from the English journals namely, American Journal of Medicine, Pain, Medicine and the Persian Journals namely, Pazhoohande (Research Medical Journal of Shahid Beheshti University), Research Medical Journal of Ahvaz University, Research Medical Journal of Kerman University. Criteria for Article Selection To choose the articles, the following criteria were taken into account. First due to dynamic nature of genre, the corpus was restricted to a period of seven years (19972003). Second, all English journal articles were published in the US and all writers assumed to be English native speakers as far as it could be inferred from their names, affiliations, and occasionally biographical notes attached to the articles, and their native-like command of English. On the other hand, the Persian articles were written by Iranians, so we expect Persian to be the writersâ&#x20AC;&#x2122; native language. It should be noted that there was no instance of translation in the Persian articles. Third, experts of this field (medicine) were consulted to choose journals with high importance, reputation, and readership. Their suggestions were of great help in deciding on the number of articles to be included in the study. For example, 57 articles were selected from one journal, but only 8 from another. Fourth, to keep the naturalness of the sampling, the length of the introductions was not controlled. Fifth, the introduction sections were from articles which had the traditional Introduction, Method, Result, and Discussion (IMRD) sections. The Identification of Moves To find the communicative purpose of each sentence, the texts were meticulously read sentence by sentence, because sentences were considered to be the unit of analysis and communicative purpose was the identifying feature of each move (Bhatia, 1993). The process of identification of moves involved focusing on formal clues such as explicit

375

ENGLISH AND PERSIAN MEDICAL RESEARCH ARTICLES

lexemes and expressions, verb forms, discourse conjuncts, etc. and focusing on function of sentences. Of course in some cases there were some instances of discrepancy between the two formal and functional approaches. In such cases, priority was given to the function of the fragment. Results Move and step frequency across the corpus were statistically analyzed. Then inter-rater and intra-rater reliability of the judgments were estimated. Move and Step Frequency The frequency of moves and their constituent steps were analyzed in the corpus. Tables 1 and 2 show the move and step frequencies in the corpus. Moves

Move 1

Move 2

Move 3

English

155

109

104

Persian

153

102

Table 1: Move frequency of introduction section of medical research articles Steps

1.1

1.2

1.3

2.1A 2.1B 2.1C 2.1D 3.1A 3.1B

English

133

116

Persian

169

140

3.2

3.3

Table 2: Step frequency of moves in the introduction section of medical research articles

Chi-squares showed no significant difference between introduction sections of English and Persian medical research articles as far as their move frequency were concerned. But the results of chi-square tests for step frequency showed that there is a significant difference between introduction section genre of English and Persian medical research articles concerning the frequency of steps. Tables 3 and 4 show the results. Moves

Significance

D.F

.05

χ². Value .163

χ². Critical 5.99

Table 3: The Chi-square indicating the significant difference of move frequency across the corpus

Steps

Significance

D.F

.05

χ². Value 32.128

χ². Critical 18.3070

Table 4: The Chi-square indicating the significant difference of step frequency in the moves across the corpus

376

AVISHAN MAHZARI

The results showed that there is no significant difference between move frequency of introduction sections of English and Persian medical research articles, but there is a significant difference between their step frequencies. This means that both English and Persian writers utilized the three moves with similar frequencies, but the realization of these three moves was different in these two languages. In order to realize the three moves, both English and Persian writers utilized the Eleven steps, but with different frequencies. In order to gain a better understanding of the issue a review over the features of each move and its constituent steps along with examples from each language are presented here. Move 1 Every research to be reported needs to create the general territory in which it operates. The analysis revealed that 100 percent of English and 100 percent of Persian RAs included move 1 in order to establish a territory. Percentage of occurrence of move 1 and its steps are summarized in the following table. Move 1: Establishing a Territory

Percentage of occurrence

Steps

English

Persian

Establishing a Territory

100

1) claiming centrality and/or

2) making topic generalizations

3) reviewing items of previous research

Table 5: Percentage of occurrence of move 1 and its steps

To establish a territory of the research to be reported, both English and Persian authors tried to assure the readers of the importance, relevance, and recency of their research project. To achieve this, they made centrality claims, topic generalizations, and reviewed items of previous research. The ultimate goal was to re-establish the significance of the research field for the discourse community. Examples of steps of move 1 across the corpus Step 1.1: Claiming Centrality English RA.No.Â 85: Diagnostic and Interventional procedures using radiocontrast media are being performed with increasing frequency. 84:

377

ENGLISH AND PERSIAN MEDICAL RESEARCH ARTICLES

Step 1.2: Making topic generalization English RA.No.Â 2: Because of the efficacy of multiagent combination chemo therapy, a significant proportion of patients are cured by current approaches. 31: â&#x20AC;&#x201C; 15

Step 1.3: Reviewing items of previous research English RA. No. 76: Initial studies found that IL.1B produces hyperalgesia, following either peripheral or central administration. (Ferreira et al., 1988; Oka et al., 1993; Watkins et al., 1994)

15: 1372 1373 )

Move 2 In move 2, there is a transition from the generally established context by focusing on inadequacy in previous research that needs new investigation. Percentage of occurrence of move 2 and its steps are presented in table 6. Move 2: Establishing a niche

Percentage of occurrence

Steps

English

Persian

Establishing a niche

1A) Counter-claiming

or 1B) Indicating a gap

or 1C) Question raising

or 1D) Continuing a tradition

Table 6: Percentage of Occurrence of Move 2 and its steps

As shown in table 6, 79 percent of English and 72 percent of Persian authors established a niche for their study. To attain this objective, they resorted to one or to a combination of steps included in move 2. Considering both English and Persian RAs,

378

AVISHAN MAHZARI

indicating a gap (step 1B) was the most favorite step, while question raising (step 1C) for English RAs and continuing a tradition (step 1D) for Persian RAs were the least. A possible interpretation can be that the authors avoid direct, explicit, and strong challenges to previous research, and they try to state the necessity of current research indirectly. 21 percent of English and 28 percent of Persian RAs did not include move 2 at all. In other words, the authors of these RAs have not stated any negative attitude or challenge to the previous research because they did not feel any need. Examples of steps of move 2 across the corpus Step 2.1.A: Counter-claiming English RA.No. 37: However, a cross-sectional PCR-based study reported a high prevalence of common respiratory viruses, which was not clearly related to the fatal attack. 64:

Step 2.1.B: Indicating a Gap English RA.No. 19: Surprisingly, the effects of late life anemia have not received much research attention.

19:

DSM â&#x20AC;&#x201C; IV

Step 2.1.C: Question-raising 1:

Step 2.1.D: Continuing a Tradition English RA. No. 12: These analyses were necessary since psychological factors may be related to MFP and bruxism in the absence of disease.

78: NUD.

Move 3 After establishing a territory and niche which are achieved through move 1 and move 2, authors find themselves in a better position to occupy the niche. In fact, by using previous moves, authors pave the way for putting forward their own research. Now

379

ENGLISH AND PERSIAN MEDICAL RESEARCH ARTICLES

that the reader is convinced about the importance and necessity of the current research, he may wish to be informed briefly about the exact goals of the research, procedure and methodology of the research, principle outcomes, and the structure of the present research. The results showed that 97 percent of English and 91 percent of Persian writers included this move in the RAs. Table 7 shows the percentage of occurrence of move 3 and its steps. Move 3: Occupying the niche

Percentage of occurrence

Steps

English

Persian

Occupying the niche

1A) Outlining purposes

or 1B) Announcing present research

or 1C) Announcing principle findings

or 1D) Indicating RA structure

Table 7: Percentage of occurrence of move 3 and its steps

Move 3 is typically presented as the final move of the introduction section. In one English RA, this move appeared at the beginning of the introduction. In other words, the writer occupied the niche at the very beginning of the RA. To Swales (1990), this refers to the fact that the author of the RA does not worry about the appreciation of his/ her product by the reader. This confidence could be because of the popularity of the author and well establishment of the importance and necessity of the research. In other words, the writer may be a well-known expert member who is reporting an issue which is very significant for other members. Examples of steps of move 3 across the corpus Step 3.1.A: Outlining purposes English RA. No. 18: The aim of our study was to assess the long-term association of C-reactive protein levels with the risk of death, nonfatal myocardial infarction, and repeat revascularization after coronary angioplasty. 46:

380

AVISHAN MAHZARI

Step 3.1.B: Announcing present research English RA. No. 7: We report the results of a cross-sectional analysis of the association between glycemic control and prevalent cardiovascular disease in patients with NIDDM.

45: .

Step 3.2: Announcing principle findings English RA. No. 5: We found no case in which A.nidulans was a pathogen in patient without CGD, despite the fact that the patient population at NIH is heavily weighted towards immunocompromised groups. These observations show that A.nidulans has distinct mechanisms of pathogenesis that are clinically important in patients with CGD.

8: .

Step 3.3: Indicating RA structure English RA. No. 5: Characteristic morphologic features of A.nidulans are shown in figure 1.

Reliability The next step was taking care of the reliability in order to avoid subjectivity in the analysis. Intra-rater reliability To this end, a sample of 60 articles (30 from each language) was extracted out of the corpus and was analyzed by the researcher two weeks after the first rating. The results of spearman rank-order correlation showed that there was high correlation between the researcherâ&#x20AC;&#x2122;s two counting of moves and steps. The details of the correlational analysis are presented in Appendix A. Inter-rater reliability In order to avoid subjectivity in the analysis of the data, a sample of 60 articles (30 from each language) were randomly extracted and were analyzed by an MA holder of TEFL, who was familiar with genre analysis (her own thesis was on genre analysis). The results of her analysis were correlated with those of the researcher. The results of spearman rank-order correlation showed that there was high correlation between the frequency of moves and steps counted by the researcher and the second rater. The details of the correlational analysis are presented in Appendix B.

ENGLISH AND PERSIAN MEDICAL RESEARCH ARTICLES

381

Discussion The results of this study shed some light on the generic organization of the introduction section of medical research articles across English and Persian and showed that there is no significant difference between them as far as the move frequency is concerned. The results also showed that the frequency of steps in the introduction section of English and Persian medical research articles is radically different. The findings revealed that the introduction section of a journal article serves a set of communicative purposes which could be presented in a specific order: Establishing a Territory, Establishing a Niche, Occupying the Niche. Introductions are organized in such a structure to persuade the discourse community that the present research is something worthy of attention. In addition, the results obtained in the study show that the moves of the introduction section across the two languages are not of equal frequency. That is Move 1 appeared to be the most frequent moves across the corpus. Move 2 places second, while move 3 is the least frequent move to be presented in the corpus. The findings also focus on the linguistic features the writers utilize to serve their communicative purposes. For example, Move 2 usually includes some negative forms (negative or quasi-negative quantifiers, lexical negation, negation in the verb phrase) because the writers try to imply that the previous research suffer some limitations, that there is a problem that has not been dealt, and that a question remains unanswered. Results of this study can help students to write research article introduction in a correct genre. University students need to both comprehend and produce academic prose of various related disciplines. Genre analysis can be of great help to instructors to make learners well-aware of how genres differ from and within each other and how learners can go about discovering these differences. The findings of this study may also contribute to the field of syllabus design because to solve the problem of academic isolation, syllabus designers must also include genre awareness courses to sensitize the students to the features that make the text a good and standard product. Raising genreconsciousness is a key factor in solving the problems that are barriers for scholarly, academic communication. In addition, it may be of great value if medicine practitioners tend to send the report of their research to medical journals and tend to know the generic organization of introduction section of their field. References Bhatia, Vijay K. (1993). Analyzing genre. London: Longman. Bhatia, Vijay K. (1997). Genre-mixing in academic introductions. English for Specific Purposes 16/3: 181-185. Bley-Vroman, Robert & Larry Selinker (1984). Research design in rhetorical grammatical studies: A proposed optimal research strategy. English for Specific Purposes 1/4: 82-83.

382

AVISHAN MAHZARI

Brett, Paul (1994). A Genre analysis of the results section of sociology articles. English for Specific Purposes 13/1: 23-45. Connor, Ulla M. & Anna Mauranen (1999). Linguistic analysis of grant proposals: European Union research grants. English for Specific Purposes 18/1: 47-62. Crookes, Graham (1984). Towards a validated analysis of scientific text structure. Applied Linguistics 7/1: 57-70. Dudley-Evans, Tony (1987). Genre analysis and ESP. Birmingham: University of Birmingham. Dudley-Evans, Tony (1994). Genre analysis: An approach to text analysis for ESP. London: Routledge. Duszak, Anna (1994). Academic discourse and styles. Journal of Pragmatics 21/8: 291-313. Flowerdew, John (1993). An educational, or process, approach to the teaching of professional genres. ELT Journal 47/4: 305-316. Flowerdew, John (2000). Using a genre-based framework to teach organizational structure. ELT Journal 54/4: 369-378. Henry, Alex & Robert L. Roseberry (1998). An evaluation of a genre-based approach to the teaching of EAP/ESP writing. TESOL Quarterly 32/1: 147-156. Holmes, Richard (1997). Genre analysis, and the social sciences: an investigation of the structure of research article discussion sections in three disciplines. English for Specific Purposes 16/4: 321-337. Hopkins, Andy & Dudley-Evans, Tony (1988). A Genre based investigation of the discussion section of articles and dissertations. English for Specific Purposes 7/2: 113-121. Howe, P.M (1990). The Problem of the problem question in English for academic purposes. English for Specific Purposes 9/3: 215-236. Hutchinson, Tom & Alan Waters (1989). English for specific purposes. Cambridge: Cambridge University Press. Hyland, Ken (1994). Hedging in academic writing and EAP textbooks. English for Specific Purposes 13/3: 85-92. Mustafa, Zahra (1995). The effect of genre transfer on linguistic awareness. English for Specific Purposes 14/3: 247-256. Nwogu, Kevin N. (1997). The Medical research paper: structure and functions. genre analysis and the identification of textual boundaries. Applied Linguistics 15/3: 288-298. Nwogu, Kevin N. (1998). Page to screen: Taking literacy into the electronic era. New York: Routledge. Paltridge, Brian (1995). Working with genre: a pragmatic perspective. Journal of Pragmatics 24/14: 393-404. Swales, John (1981). Aspects of article introductions. The University of Aston: Language Studies Unit. Swales, John (1983). Vocabulary work in LSP: A case of neglect? CILA 14: 21-34. Swales, John (1985). Episodes in ESP. Hemel Hempstead, UK: Prentice Hall. Swales, John (1990). Genre analysis. Cambridge: Cambridge University Press.

383

ENGLISH AND PERSIAN MEDICAL RESEARCH ARTICLES

Appendices Appendix A Table 8: Intra-rater correlation of English moves English moves in first and second rating

Significance .01

D.F 1

ρ Value 1

ρ Critical .99

Table 9: Intra-rater correlation of English steps English steps in first and second rating

Significance .01

D.F 9

ρ Value .99

ρ Critical .73

Table.10: Intra-rater correlation of Persian moves Persian moves in first and second rating

Significance .01

D.F 1

ρ Value 1

ρ Critical .99

D.F 9

ρ Value .99

ρ Critical .73

D.F 1

ρ Value 1

ρ Critical .99

D.F 9

ρ Value .98

ρ Critical .73

Table 11: Intra-rater correlation of Persian steps English steps in first and second rating

Significance .01

Appendix B Table 12: Inter-rater correlation of English moves English moves by second rater and researcher

Significance .01

Table 13: Inter-rater correlation of English steps English steps by second rater and researcher

Significance .01

384

AVISHAN MAHZARI

Table 14: Inter-rater correlation of Persian moves Persian moves by second rater and researcher

Significance .01

D.F 1

ρ Value 1

ρ Critical .99

D.F 9

ρ Value .99

ρ Critical .73

Table 15: Inter-rater correlation of Persian steps Persian steps by second rater and researcher

Significance .01

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 385-399 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

THE ROLE OF NONVERBAL EXPRESSIONS AS PRECURSORS TO ARGUMENTATIVE DISCOURSE IN FRENCH, JAPANESE AND AMERICAN ENGLISH CONVERSATION CAROLINE ELISA NASH

1. Introduction Numerous studies have been done on nonverbal communication with attention focused primarily on gaze behavior patterns. Extensive work has been done on the role of gaze and “gaze shift” in the “turn” of turn-taking in English (Duncan & Fiske 1977, 1985; Goodwin 1981; Schegloff et al. 1984). It is now widely accepted that the division between language and gestures is not as distinct as previously believed (Kita 2003; Kita & Ide 2007; McNeill et al. 2005). Studies in conversation analysis that incorporate the nonverbal component reveal important facts about the relationship between language and gesture. The use of gestures in a natural and interactive conversation requires observable contextual phenomenon as well as assumptions or inferences about the speaker’s beliefs and intentions. The mutually-shared background information of the speaker and addressee depends to a great extent on their cultural background. Although studies on gaze behavior describe observed patterns of predominantly American subjects, studies on culture-specific gaze behavior have been conducted since the early 20th century that reveal distinct cross-cultural differences in certain patterns of gaze behavior between interlocutors engaged in interactive conversation. Most notably, Whiffen (1915), who conducted studies on gaze behavior of American Indians, attested that Indians do not look at each other while speaking – neither the speaker at the listener, nor the listener at the speaker (Whiffen 1915: 254). Yet, ethnocentric studies still dominate kinesic research and the constructed models and postulated rules are often generalized to apply to the social behavior and organizational structure across languages and cultures. Speakers and addressees across cultures do not use the same techniques in gaze behavior patterns and hand and head gestures that regulate the conversation as will be revealed by our study of argumentative discourse in French, Japanese and American English. We further seek to show that the role of the gesture as a conversation marker is significant in any model of talkinteraction, due to the fact that overt linguistic cues to regulate natural conversation and express interlocutors’ attitudes are not usually expressed during natural conversation.

386

CAROLINE ELISA NASH

This study addresses the following research questions: 1. How do interlocutors communicate intent to disagree or argue? 2. Since nonverbal expressions convey more than is actually said, are there certain gestures which signal that the forthcoming discourse is argumentative? (i.e. nonverbal regulators that function as a precursor to argumentative discourse?) 3. Are there cross-cultural differences or similarities (perhaps universals) in the nature of these preverbal regulators? In this paper I examine the means by which the addressee signals forthcoming argumentation as they effect their turn transition, and the means by which the speaker averts argumentation by the addressee. I discuss the role of gaze and assert that gaze avert is crucial in the argumentation process. 2. Regulators defined Schiffrin (1987) uses the term discourse markers to cover a wide spectrum of discourse functions including those in the areas of pragmatics and conversation analysis. Discourse markers have defined functions in both written text and verbal communication. In conversation, the “participation framework” (Schiffrin1987: 27) includes “the way in which speakers are related to their turns: claiming them, fighting for them, and relinquishing them”. Fraser (1996: 168) defines pragmatic markers as “linguisticallyencoded clues which signal the speaker’s potential communicative intentions”. According to Fraser, these pragmatic markers correspond to four different types of messages but he clearly makes a distinction between the first three: basic, commentary, and parallel markers, and the fourth: discourse markers. For Fraser, discourse markers are only those markers that specify how the message is related to foregoing discourse. Based on this refinement of the definition of discourse markers, regulators cannot be classified accordingly. Manoliu (1999) argued for the necessity of further refining the categories of pragmatic markers and proposed the term conversation markers as a sub-category of pragmatic markers whose functions are to organize the talk-interaction. These functions include their role in negotiating the turn, their role in controlling the addressee’s attention and understanding, their role in accepting or rejecting the speaker’s topic, and crucial to this study, their role in conveying speaker-addressee attitude. Since regulators function to control and maintain the flow of natural conversation, in my view, regulators fit under the rubric of conversation markers. 3. Data collection and method The data for this study were collected via video recordings of native French speakers residing in various regions throughout France, Japanese native speakers residing in Japan and American English native speakers residing in California in the United States.

THE ROLE OF NONVERBAL EXPRESSIONS

387

All participants were taped while engaged in natural interpersonal and interactive conversations and were not at any time aware of the nature of the study and the specific topic of research prior to or during the taping. The subjects are of five different sociolinguistic groups: 1. French speakers 30-65 years of age; 2. Japanese speakers 30-65 years of age; 3. Japanese speakers 20-25 years of age; 4. American English speakers 3065 years of age and 5. American English speakers 20-25 years of age. 4. Addressee signals of argumentative discourse In this section we examine some nonverbal expressions displayed by the addressee that serve as precursors to argumentative discourse. These signals communicate intent to disagree or argue on the part of the addressee. The abbreviations used in this section and their corresponding referents are as follows: S = Speaker; A = Addressee; A➝s = Addressee-turned-speaker (i.e. addressee took the speaker-turn); GD = Gaze Direct; GA = Gaze Avert; HT = Head Tilt. 4.1 The French addressee The French addressee displays the head tilt as a turn-taking strategy but crucially, the head tilt bears the meaning of a certain degree of disaccord. The head tilt gesture conveys the addressee’s intent to argue or disagree to the speaker, and precedes verbal expressions of disaccord. In example (1), a French addressee displays the head-tilt preceded by gaze avert prior to his argumentative discourse. 1.

(A averts his gaze, displays head-tilt, then interjects S) A➝s: (GA➝HT) “Oh…mais je me fous de ce que vous et les autres en pensent! Ça vous regarde pas! Il m’a dit ce matin même: (GD) “J’aime bien ma vie ici à Paris avec toi!”” A➝s: (GA➝HT) ‘Oh … but I don’t give a damn about what you and the others think! It’s none of your business! He said to me this very morning, (GD) “I like my life here in Paris with you!”’

The addressee-turned-speaker redirects his gaze at his interlocutor as he quotes an absent referent. In example (2) shown in Image I, a French addressee uses the head-tilt to signal forthcoming argumentative discourse in response to what the speaker has just said and prior to explaining or justifying his disagreement. The addressee averts the gaze just prior to uttering “mmm…”, then pauses, and begins to tilt his head prior to uttering “ah” and continues tilting his head as he utters “non!”. The addressee-turned-speaker then commences his argumentative discourse.

388

CAROLINE ELISA NASH

Image I.

A➝s: (GA) “mmm…”

(HT) “ah”

(➝ HT➝ HT) “non!” (‘no!’)

Other French verbal expressions that signal forthcoming argumentation preceded by the addressee head-tilt extracted from the data are “alors, là...” (‘now that there...’); “ben, écoute...” (‘oh now, listen....’); “pas forcément...” (‘not necessarily...’); “pas spécialement…” (‘not particularly’); “oui, mais...” (‘yes, but...’). These utterances respond to what a speaker has just said and prior to explaining or justifying his disagreement. Addressee gaze avert and head-tilt may also bear the additional meaning of negative attitude on the part of the addressee as shown in example (3). In this example there are two addressees who take the speaker-turn argumentatively and clearly express a negative attitude towards the speaker’s discourse: 3.

(GD) “En tout cas [unintelligible]… de venir récuperer tes affaires!” (GD) ‘In any case [unintelligible]… to come get your things!’

A1➝s: (GA➝HT) “C’est bon! Commencez pas!” (GA➝HT) ‘That’s enough! Don’t start!’

389

THE ROLE OF NONVERBAL EXPRESSIONS

A2➝s: (GA➝HT) “Ohhhh!” (GA➝HT) ‘Ohhhh!’

Other French expressions of negative attitude uttered by the addressee preceded by gaze avert and head-tilt extracted from the data are as follows: “n’importe quoi” (‘(you just say) whatever/anything!’); “tu te racontes!” (‘you daresay!/oh really?’); “je m’en fous!” (‘I don’t give a damn!’); “c’est pas beau”! (‘that’s not nice!’); “il comprend rien!” (‘he doesn’t understand anything!’); “j’avais rien compris de ce que tu as dit!” (‘I hadn’t understood anything you said!’); “moi, j’aime pas trop!” (‘I don’t like it much!’). Gaze avert followed by the head tilt performed by the French addressee are also displayed preceding expressions of sarcasm such as, “ah bon!?” (‘oh really!?’) and “ah… tu vas prendre ça là-bas!?” (‘oh… you’re going to take that over there!?’). French verbal expressions of disagreement, negative attitude and sarcasm on the part of the addressee are almost always uttered in conjunction with gaze avert followed by a head gesture, one of which is the head tilt as presented in this discussion. 4.2 The Japanese addressee A side head tilt by the Japanese addressee conveys disaccord to the speaker. The addressee is in disagreement with the speaker’s opinion, comment or topic. If the speaker yields the floor at the head tilt display, then the addressee has succeeded in taking his turn to counter the speaker. Otherwise, the head tilt is held longer and firmer until the addressee successfully takes his turn. The initial head tilts may be unaccompanied by verbal utterances, or pre-verbal followed by the utterances, “mmm...”, “saa...” (expression of doubt) or “demo...” (‘but...’), as shown in example (4) and Image II. The expressions “saa…” and “mnn…” are never uttered without first averting the gaze then tilting the head, whereas “demo…” is uttered either following gaze avert or the head tilt, or both. Image II

“mnn…”

“saa…”

390

CAROLINE ELISA NASH

In example (5) shown in Image III, the addressee who is in disagreement with the speaker, averts the gaze and displays the head tilt just prior to his discourse. Image III

(a) 5.

(b)

(a) S:

“Meiji ka nan ka.” ‘Like Meiji or something.’

(b) A➝s:

“Meiji ja naku te yo,…” ‘Not Meiji…’

The Japanese addressee head tilt and gaze avert are also displayed when expressing negative attitude and sarcasm. Some expressions extracted from our data with approximate glosses are as follows: “ii jyanai!” (‘so what!/you should be pleased!’); “jyoodan jyanai yo!” (‘you’ve got to be kidding !’/no way !’); “yoku yuu yo!” (‘you exaggerate!/you just say whatever!’); “kankei nai yo!” (‘that’s got nothing to do with it!’); “komaru yo!” (‘that’s problematic!/that’s an imposition!’); “mata yuu!” (‘there you go (saying that) again!’). 4.3 The American addressee A very common American gesture used by members of certain socio-linguistic groups is the “Talk to the Hand” open palm gesture displayed by the addressee preceding the utterance, “Whatever!” to convey to the speaker that she or he wants the speaker to stop talking, and is usually followed by another contentious utterance leading into mutual argumentative discourse. This gesture is performed by a twist of the wrist and circular hand motion, partially extending the arm towards the speaker’s face, positioning the tense open hand between the interlocutors as illustrated in Image IV. The addressee precedes the hand display with gaze avert.

THE ROLE OF NONVERBAL EXPRESSIONS

391

Image IV

Gaze avert is the primary means by which an addressee signals forthcoming argumentation in American English. In example (6), when the speaker offends the addressee by uttering “…show off your talents…”, the addressee immediately averts his gaze and then counters by denying that he was showing off, all the while averting his gaze. He redirects his gaze towards his interlocutor when he corrects the statement made by the speaker. 6.

S: “… show off your talents…” A➝s: (GA) “I didn’t want to show off, (GD) I just wanted to stay awake.” ← Gaze Avert →

American English verbal expressions of negative attitude, sarcasm and disaccord on the part of the addressee are often uttered following gaze avert. Some examples extracted from the data are as follows: “what is her problem?”, “I have a problem with what she’s wearing!”, “hell no!”, “who cares?”, “what the heck?” “you’re/she’s hella dumb!”, “you’re retarded!”, “you’re smoking dope!”, “in your dreams!”, “oh yeah, I forgot, you were dropped on your head when you were a baby!”, “yeah, right!”, “don’t think so!”. 5. Speaker means of averting the argument In the previous section, we examined certain nonverbal expressions that serve as precursors to argumentative discourse. These signals communicate intent to disagree or argue on the part of the addressee. In this section we examine some means by which the speaker averts the addressee’s forthcoming argument which oftentimes serves a dual purpose, i.e. not merely to avert an argument, but also to signal forthcoming argumentation on the part of the speaker. The abbreviations used in this section and their corresponding referents are as follows: S = Speaker; A = Addressee; GD = Gaze Direct; HD = Hand Display.

392

CAROLINE ELISA NASH

5.1 The French speaker A regulator used by the French speaker to signal that it is still his or her turn to speak and that he or she is not yet willing to yield the floor to another participant at a transition relevance place, comprises the index finger vertically placed in between the speaker and the addressee with the arm bent at a 45 angle, elbow forward and slightly raised. A variation is the open palm held vertically between the speaker and the addressee, the palm facing the addressee. In example (7) shown in Image V, prior to the first frame (a), the speaker utters, “…parce que si vous donnez de trop, vous avez un contrôle fiscal.” She continues to say, “On vous dira…” while at the same time the addressee interjects with “Tu exagères…” resulting in overlap. At this juncture in frame (a), the speaker, anticipating an argument on the part of the addressee, displays her index finger gesture and continues to speak, countering with “…on dira… on vous dira: ‘Mais vous avez servi deux repas. Vous n’en avez déclaré qu’un. Donc, vous volez l’État!’”, all the while maintaining the display of her hand gesture as well as gaze direct towards her addressee. Still in Frame (a), the addressee interjects again with the utterance “Tu exagères! Peser tous… chaque… chaque…” while the speaker is still holding up her index finger and overlaps with “Il y a eu un restaurant… Il y a eu…”. It isn’t until frame (b) after the speaker restarts her argument with the utterance “Il y a eu un restaurant…” when she no longer anticipates an argument from the addressee, i.e. quells the argument and succeeds in denying the addressee his speaker-turn, that she terminates the gestural hand display. Image V

(a)

(b)

394

CAROLINE ELISA NASH

Image VI

S1: “Si tu veux être décontractée, il faut fumer le bédo tous les deux et tu seras plus comme ça (gesturing hands trembling motion).” ‘If you want to destress, you have to smoke the bedo, the two of us together and you won’t be like this (shaking) anymore.’ S2: (HD/GD) “Après! Après! Si tu veux, après… après les examens, si tu veux… mais avant, non.” (HD/GD) ‘After! After! If you want, after… after the exams, if you want… but before, no.’

Speaker 2 then continues to defend her position by expressing the seriousness of the exam-taking process that Speaker 1 dismisses as an almost frivolous undertaking, thereby perpetuating further argument that ensues between these two participants. 5.2 The Japanese speaker A tense nod held 2 to 3 seconds by the Japanese speaker performed in sentence-final position, functions to avert a forthcoming argument by the addressee. This head gesture conveys to the addressee that the speaker wants to change the topic or terminate the conversation. Attempts on the part of the addressee to further contribute to the conversation are usually futile. This gesture conveys negative attitude in essence saying: “I am having the last word on this topic and now it’s the end of this conversation!” With regard to gaze patterns, Japanese speakers do not maintain gaze direct in the process of averting an argument as do the French. The Japanese speaker averts the gaze just prior to the head nod display. 5.3 The American speaker One nonverbal means by which the American speaker averts an argument is by averting the gaze. In the sequential organization of the turn-at-talk, gaze direct is a turnyielding regulator for Kendon (1990). According to his model, during the speaker-

THE ROLE OF NONVERBAL EXPRESSIONS

395

turn, the speaker averts the gaze; therefore in displaying gaze direct, the speaker is signaling a response from the addressee. Our data support Kendon’s turn-taking model for American English speakers. However, in the argumentation framework, I find that when the speaker anticipates argumentative discourse from the addressee, she or he redirects the gaze towards the addressee for a brief moment and then averts the gaze just prior to an attempt by the addressee to counter the speaker’s discourse. The hand gesture, “Talk to the Hand!”, discussed in Section 4.3. and shown in Image IV is also used by American speakers to avert an argument. This hand gesture differs from the addressee signal in that the purported meaning does not necessarily bear a negative attitude on the part of the speaker. Here again, the speaker first redirects the gaze to the addressee, then averts the gaze followed by the display of the hand gesture. 6. The significance of the role of gaze avert In English, the “nod” is a back-channeling regulator used by the addressee with utterances such as “uh-huh”, “I see”, etc., to signal to the speaker that she or he is listening, following and/or is in agreement with the speaker’s opinions, comments and/or topic. This gesture maintains the conversation flow and conveys “positive attitude”. The nod is always displayed while gazing at the speaker. We find the “nod” to be a backchanneling regulator also used by addressees in French and in Japanese. Regulators perform other functions, however, such as convey negative attitude, request or reject further information, control the addressee’s attention and understanding, accept or reject the speaker’s topic, confirm the speaker’s hypotheses about the addressee’s background knowledge via tag questions and negotiate turn-taking. In this study we found that the addressee back-channels utterances that 1. convey disagreement with the speaker’s discourse and possibly interjecting or taking the speaker turn to explain or justify the disagreement; 2. convey negative attitude; 3. express sarcasm and 4. express disaccord. Negative back-channeling, therefore, is expressed with gaze avert1 , while positive back-channeling is always uttered with gaze direct. The expression of disagreement and negative attitude can also be viewed as a departure or detachment from the mutual engagement and from the progression of the topic. As such, gaze avert functions in a similar manner to digress from the here-andnow with the intent of cooperating in the natural course of the conversational exchange in order to arrive at a resolution or conclusion of sorts. Thus, the detachment from the mutual engagement and the digression from the here-and-now – this discourse feature is manifested by the aversion of gaze.

1 This is not to say that one never displays gaze direct in negative back-channelling. Certainly in confrontational situations this would be the case; however, we are working within the context of an ordinary conversation which follows the Gricean maxims and the Cooperative Principle.

396

CAROLINE ELISA NASH

Crucial to this study is accounting for cultural differences which I find to be significant with respect to gaze behavior patterns and the form of the head or hand gesture. As the data have shown, French speakers maintain gaze direct prior to, during, and after the gestural signal; whereas Japanese and American speakers avert the gaze prior to and during the gestural display. Perhaps of greater significance is the pattern that is observed across all three of these languages and cultures. This study yields results that suggest that gaze avert functions across these three languages as a precursor to argumentative discourse but with the following condition: Only when it is the addressee who displays a nonverbal signal that functions as a precursor to argumentative discourse, is the precursor displayed in conjunction with gaze avert.

Hence, gaze avert is an addressee precursor to argumentative discourse. Assuming that the interlocutors are adhering to the principles and parameters of the turn-taking model for American English conversation which stipulates that the addressee is gazing at the speaker during the speaker turn, gaze avert does not indicate a lack of interest in or disapproval of the speaker’s topic on the part of the addressee as previously suggested by Argyle and Cook (1976) and others, but on the contrary, the addressee is expressing his or her vested interest in the topic by virtue of the disagreeing, disputing and disproving of the speaker’s discourse. Although it may be argued that the former is found to be true in expressing negative attitude and sarcasm, the lack of interest or disapproval is actually not in the speaker’s topic when expressing disagreement, rather, it is with the speaker’s comment, opinion or utterance, i.e. discourse, that the addressee is in disagreement. Further, according to the French turn-taking model proposed by Nash (2001), speaker-addressee mutual gaze (i.e. gaze direct by both interlocutors) is high during the speaker turn. Hence, gaze avert on the part of the addressee has a functional purpose, and one such function is that in conjunction with utterances that express disagreement, negative attitude, or sarcasm, gaze avert signals forthcoming argumentative discourse. Finally, the Japanese exhibit low mutual gaze, i.e. high gaze aversion during a conversation. The Japanese model in and of itself belies the notion that gaze avert indicates a lack of interest. 7. Preliminary models The data from this study suggest that as interlocutors engaged in argumentative discourse, we exhibit a predictable pattern, i.e. we adhere to specific rules in the interactive exchange from which we can construct a model of behavior for the addressee and the speaker. I propose here a 3-tiered model that accounts for the effecting of the addressee’s argumentative discourse. First and foremost we presuppose that there is no speaker ex-

THE ROLE OF NONVERBAL EXPRESSIONS

397

pectation of a forthcoming argument by the addressee. Secondly, the addressee performs an act that functions to signal to the speaker his or her intent to argue. Finally, the act is successful if the intended result is achieved, i.e., the addressee takes the speaker-turn and initiates his or her argumentative discourse. I. Premise: There is no speaker expectation of a forthcoming argument by the addressee. II. Performance: The addressee signals with gestural display. III. Result: The addressee takes the speaker-turn interrupting the speaker with a contentious comment. I propose another 3-tiered model that accounts for the effecting of the speaker’s averting the addressee’s argumentative discourse (and secondarily, effecting the speaker’s argumentative discourse). First and foremost we presuppose a forthcoming argument by the addressee. Secondly, the speaker performs an act that functions to block the addressee’s argument. Finally, the act is successful if the intended result is achieved, i.e. the addressee does not realize his or her argumentative discourse. I. Premise: The speaker has the expectation of a forthcoming argument from the addressee. II. Performance: Anticipating the argument, the speaker gestures thus denying the addressee the speaker-turn to counter. III. Result: Any attempt on the part of the addressee to counter are futile hence the argument on the part of the addressee is not realized. 8. Conclusion Nonverbal regulators as conversation markers perform various functions in the organization of the talk-interaction. The goal of Pragmatics is to account for those mechanisms that are employed in communicating more than is actually said or conveyed by a verbal utterance; hence the principles and parameters that account for pragmatic competence in native speakers are not delimited to linguistic expressions. On the contrary, a comprehensive account of native speaker pragmatic competence cannot preclude nonverbal behavior, particularly in the form of gestural patterns that are displayed in conjunction with verbal utterances, as these features of the language system are observed in the interaction of native speakers which only occur within socio-cultural contexts. This paper addressed this issue by identifying those nonverbal expressions that regulate conversation, focusing on signals that serve as precursors of argumentative discourse which encompass negative attitude and disagreement in French, Japanese and American English conversation. We sought to reveal cultural differences and in so doing we discovered a cross-linguistic pattern, i.e. gaze avert displayed by the addressee functions across these three languages as a precursor to argumentative discourse.

398

CAROLINE ELISA NASH

We proposed models that account for the speaker-addressee realization of argumentation in a conversational exchange that necessarily incorporates the gestural code due to the fact that the verbal code is not sufficient in describing the actual exchange in all of its intricacies. This is merely a first study and in view of the now widely-held belief that language and gesture function together as one unit â&#x20AC;&#x201C; inseparable and indispensable in communicating in any language, further research in the role of nonverbal expressions as they relate to argumentative discourse is surely warranted. References Argyle, Michael & Cook, Mark (1976). Gaze and Mutual Gaze. Cambridge: Cambridge University Press. Duncan Starkey, Jr. & Fiske, Donald W. (1977). Face-To-Face Interaction: Research, Methods and Theory. New Jersey: Lawrence Erlbaum Associates Publishers. Duncan Starkey, Jr. & Fiske, Donald W. (1985). Interaction structure and strategy. Cambridge: Cambridge University Press. Fraser, Bruce (1996). Pragmatic markers. Pragmatics 6/2: 167-190. Goodwin, Charles (1981). Conversational Organization. Interaction between Speakers and Hearers. New York/Toronto/Sydney/San Francisco: Academic Press. Kendon, Adam (1990). Conducting Interaction. Patterns of behavior in focused encounters. Studies in Interactional Sociolinguistics, 7. Cambridge: Cambridge University Press. Kita, Sotaro (2003). Pointing : where language, culture, and cognition meet. Mahwah, NJ: Lawrence Erlbaum. Kita, Sotaro & Ide, Sachiko (2007). Nodding, aizuchi, and final particles in Japanese conversation: How conversation reflects the ideology of communication and social relationships. Journal of Pragmatics, 39/7: 1242-1254. Manoliu, Maria M. (1999). Interrogative utterances as non-questions. Romanian markers of talk-interaction. Romanistik in Geschichte und Gegenwart. Trier/Hamburg: Helmut Buske Verlag. McNeill, David (2005). Gesture and Thought. Chicago: Chicago University Press. Nash, Caroline (2001). Language and Gestures in Conversation: A Cross-Cultural Study of the Usage and Functions of Regulators and Illustrators in French, Japanese, and American English. (Doctoral dissertation, University of California, Davis). Dissertation Abstracts International 62, 07: 2404. Sacks, Harvey, Schegloff, Emanuel A. & Jefferson, Gail (1974). A Simplest Systematics for the Organization of Turn-Taking for Conversation. Language, 50: 696-735. Scheflen, Albert E. (1964). The Significance of Posture in Communication Systems. Psychiatry, 27: 316-331. Schegloff, Emanuel A. (1984). On some gesturesâ&#x20AC;&#x2122; relation to talk. In: Atkinson J. M. & Heritage J. (eds.), Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press, 266-296.

THE ROLE OF NONVERBAL EXPRESSIONS

399

Schiffrin, Deborah (1987). Discourse Markers. Studies in Interactional Sociolinguistics, 5. Cambridge: Cambridge University Press. Whiffen, Thomas (1915). The North-West Amazons: Notes on Some Months Spent among Cannibal Tribes. London: Constable.

THEORETICAL ISSUES IN ARGUMENTATION

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 403-413 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

JEUX ET ENJEUX DU DISCOURS RAPPORTÉ POUR LE DISCOURS ARGUMENTATIF

PRISQUE BARBIER

Introduction Cette étude s’intéresse au rôle joué par les procédés polyphoniques dans les discours argumentatifs. Elle s’appuie sur notre travail de thèse qui interroge l’influence du contexte sociolinguistique de la Côte d’Ivoire sur les pratiques discursives de locuteurs et scripteurs ivoiriens, et notamment sur les procédés de prise en charge. Tout d’abord, afin d’exposer le cadre dans lequel se situe notre étude, nous commencerons par la description du contexte sociolinguistique de la Côte d’Ivoire. Ensuite, nous présenterons le cadre théorique sur lequel elle s’appuie, à savoir celui de la polyphonie. Enfin, nous consacrerons la troisième partie de cet exposé à la présentation des résultats de notre analyse des procédés polyphoniques observés dans notre corpus d’étude, à savoir des entretiens et des débats. 1. Situation sociolinguistique de la Côte d’Ivoire La Côte d’Ivoire est le creuset de plusieurs langues vernaculaires, mais comme l’indique la constitution adoptée en 1960 à l’indépendance, la langue officielle de la Côte d’Ivoire est le français. Cette langue est, comme beaucoup d’autres institutions en Côte d’Ivoire, un héritage de la colonisation française, introduit historiquement à travers le système scolaire. Duponchel (1979) explique que «dans cette optique le français paraît devoir jouer pour l’essentiel en Côte d’Ivoire, le rôle qu’il joue pour un Français de souche». Par là, il souligne toute l’ambiguïté du rôle assigné au français, non seulement canal de diffusion de la «Connaissance» et de la Culture, mais aussi langue «seconde» et non pas langue maternelle. En outre, il constate que malgré son importance, le français, langue officielle, n’est pas senti comme une réalité nationale. En effet, il observe que les Ivoiriens revendiquent «le droit de parler le français à leur manière». Ainsi, pour lui, «ces faits … semblent montrer que la langue seconde n’est plus tout à fait sentie comme un corps étranger et intouchable, mais comme une réalité vivante pouvant s’adapter à un contexte socio-culturel très spécifique». Lafage (1980) pense elle que pour comprendre la manière dont se construit le «français populaire», il faut d’abord faire une analyse sociolinguistique de la communication entre locuteurs d’une langue africaine. Ainsi, elle observe que la communication en français se calque sur les modèles communicatifs africains, notamment en ce qui concerne la stratification de la société. Simard (1994) considère lui que aujourd’hui «il ne fait aucun doute que le fran-

404

PRISQUE BARBIER

çais de Côte d’Ivoire se soit ivoirisé. C’est-à-dire qu’il y a une norme locale, endogène qui y régit maintenant les usages» au point que l’on puisse même parler d’un «français de Côte d’Ivoire». Pour lui, la langue française s’est fondue dans le moule de la société ivoirienne, pour en arriver à «une symbiose entre la langue et la société», et il ajoute qu’on peut même parler de «vernacularisation» du français, car «cette langue est le reflet et l’expression de la société ivoirienne, tant au plan de sa structure sociale qu’à celui de sa façon d’appréhender le monde et d’en rendre compte». Pour lui, ceci s’explique par le fait que la variété centrale de français pratiqué en Côte d’Ivoire, bien que fortement marqué par la norme académique a aussi pour origine le Français Populaire Ivoirien et la structure des vernaculaires ivoiriens, ainsi que «le mode de conceptualisation propre à une civilisation de l’oralité». Il remarque donc que lorsqu’un Ivoirien parle français, même s’il emploie les règles de la grammaire française, il utilise aussi le français «selon son propre mode de production de sens en fonction de son identité culturelle». D’après Kouadio (1998), le français tel qu’il est pratiqué en Côte d’Ivoire s’est même tellement particularisé qu’on peut dire aujourd’hui qu’il est devenu une variété autonome par rapport au français central. En effet, il souligne qu’en Côte d’Ivoire «plusieurs variétés de français coexistent, se concurrençant souvent, s’interpénétrant parfois correspondant toujours à des besoins et à des situations de communication spécifiques». Manessy (1990) pense également que cette norme endogène vient de «la culture traditionnelle» sur laquelle s’appuie la société africaine d’aujourd’hui, et notamment sur le statut particulier qui y est reconnu à la parole, dont les règles se reflètent dans une rhétorique et une stylistique spécifique. Il reconnaît lui aussi que ces spécificités sont à la fois d’ordre phonétique et prosodique, mais aussi lexical (dans les calques et les emprunts), ainsi qu’au niveau de la phraséologie, et constituent «un lieu d’exercice privilégié pour la fonction identitaire du langage» car «les traits y acquièrent aisément une signification sociale» et permettent de catégoriser l’expérience et d’organiser le message et l’énonciation. Ainsi, comme nous venons de le voir, le contexte sociolinguistique de la Côte d’Ivoire paraît jouer un rôle important dans le mode de construction des discours des Ivoiriens. Pour notre part, nous étudierons plus particulièrement ici les procédés énonciatifs par lesquels les Ivoiriens véhiculent dans leurs énoncés différents points de vue, et les finalités stratégiques de ces différents modes de prise en charge discursive. 2. Théorie de la polyphonie Notre analyse s’appuie sur les recherches en pragmatique qui s’intéressent à l’étude des phénomènes linguistiques en contexte, c’est-à-dire en tenant compte de l’usage qui en est fait par l’utilisateur, et des effets que celui-ci a l’intention de produire, et notamment aux travaux concernant l’énonciation qui s’attache à décrire l’emploi des procédés linguistiques (embrayeurs, déictiques, modalisateurs, procédés polyphoniques) par lesquels un émetteur imprime sa marque à son énoncé, et les relations qui se tissent entre l’énoncé et les différents éléments constitutifs du cadre énonciatif.

JEUX ET ENJEUX DU DISCOURS RAPPORTÉ POUR LE DISCOURS ARGUMENTATIF

405

Nous nous intéresserons ici plus particulièrement au phénomène de la polyphonie, qui désigne la présence de voix différentes à travers une même énonciation. Bakhtine (1929), et à sa suite Ducrot (1980) utilisent la notion de polyphonie pour décrire le caractère dialogique du discours. Pour Ducrot, il y a polyphonie quand on peut distinguer dans une énonciation deux types de personnages: le locuteur et les énonciateurs. Le locuteur est la personne responsable de l’énoncé (le producteur physique de l’acte), les énonciateurs sont les différentes voix présentes dans l’énoncé, et l’énonciation permet d’exprimer leur point de vue. Ducrot a parlé lui de polyphonie énonciative, et Authier-Revuz (1995) d’hétérogénéité montrée, pour parler des multiples traces dans l’énoncé de la présence d’une autre source énonciative. Nous ne présentons dans cette étude que les procédés polyphoniques observés lors de l’analyse de notre corpus: l’emploi du discours rapporté, de la citation, du proverbe, de la polyphonie, et de la présupposition. 2.1 Le discours rapporté Le discours rapporté est la manifestation la plus classique de l’hétérogénéité énonciative, par le biais de l’emploi du Discours Direct et du Discours Indirect qui sont deux modes de citation indépendants. Maingueneau (1991; 1999) décrit le fonctionnement de ces deux types de discours. 2.1.1 Le Discours Direct Dans le DD les deux actes d’énonciation se trouvent parfaitement disjoints, étant rapportés à leurs situations d’énonciation respectives car le Discours Citant (DCt) et le Discours Cité (DCé) possèdent chacun un repérage distinct pour leurs embrayeurs. A l’écrit cette disjonction est manifestée par des guillemets (ou le tiret, s’il s’agit d’une réponse) qui sont le plus souvent renforcés par un verbe introducteur (dire, parler…). A l’oral, ces verbes sont souvent indispensables, associés à une pause, une variation dans le débit, l’intonation… lors du passage du DCt au DCé. Puisque le DD rapporte les propos tenus, sa principale qualité doit être une très grande fidélité, car une loi du discours voudrait que le rapporteur soit sincère et ne trahisse pas l’énoncé originel, mais rien ne l’empêche de rapporter des propos différents de ceux émis, et le fragment de DCé peut même paraître conforme à l’original et être déformé par l’intonation, une mise en contexte tendancieuse, etc. Ainsi, le DD demeure du discours rapporté et ne saurait constituer du discours au sens strict puisque la plupart du temps les propos cités ne sont pas assumés par leur énonciateur effectif, et même s’il s’agit du même énonciateur la situation d’énonciation est distincte. En outre, cette mise à distance du DCé par le DCt peut avoir différentes significations: un souci de respecter à la lettre le DCé, une volonté de ne pas prendre à son compte un énoncé que l’on rejette, un désir de paraître objectif, ….

406

PRISQUE BARBIER

2.1.2 Le Discours Indirect Concernant le DI, l’invariance entre énoncé originel et énoncé rapporté est au niveau de l’interprétation de l’énoncé originel, car il s’agit de la version qu’en donne le rapporteur. Ainsi, il n’y a plus autonomie syntaxique du DCé à l’égard du DCt, puisque dans le DI le DCé a le statut d’une simple complétive complément d’objet du verbe du DCt, où il n’y a de place que pour le DCt, cela impliquant que toutes les traces de l’énonciation du DCé (embrayeurs, tournures expressives, interrogations, etc.) soient effacées au profit de celles du DCt. De plus, étant «distant» de l’énoncé originel, le DI permet des contractions, ou des extensions de celui-ci car rien n’interdit de rapporter avec une seule phrase de DI un énoncé de plusieurs heures. Donc, un énoncé peut être rapporté au DI d’un nombre infini de manières selon les désirs de l’énonciateur du DCt, et en prenant en charge les propos d’autrui le rapporteur donne une certaine image de luimême, renvoie à ses positions idéologiques, affectives, personnelles, … . Il n’y a pas de différence importante entre le DI oral et écrit. Il introduit le DCé par un verbe introducteur suivi d’une complétive objet et il n’a donc pas de structure syntaxique qui lui soit propre. Ces verbes ne sont pas forcément ceux du DD, même si la majorité d’entre eux sont employés pour les deux types de discours rapporté (dire, avouer, crier…), mais certains ne sont utilisables qu’au DD («Enfin», soupira Elise). Ils apportent également divers types d’informations sur l’acte linguistique effectué, qui conditionnent la manière dont le récepteur interprétera le DCé. Certains verbes présupposent la vérité ou la fausseté du Dcé qui les suit (prétendre), certains situent le discours rapporté dans la chronologie discursive (répondre, répéter, conclure), certains explicitent la force illocutoire (supplier, promettre, …). 2.2 La citation Concernant la citation, Maingueneau (1976; 1991) repère différents types de citations ayant chacune leur fonction. - Les citations-preuves qui interviennent au cours d’une argumentation, soit pour réfuter, soit pour défendre, soit pour étayer un argument. Elles sont utilisées en raison de leur contenu ou de leur auteur, dans ce dernier cas on parle de «citation-autorité». - Les citations-reliques, servent elles à incarner un fragment de «discours vrai», authentique, afin d’authentifier le discours citant, de lui conférer de la validité. - Les citations-épigraphes sont destinées à relier le discours nouveau à un ensemble textuel plus vaste, à l’intégrer dans un ensemble d’énoncés antérieurs (l’intertexte). - Les citations-culture fonctionnent comme signe de connivence, signe de «culture». Elles varient selon les communautés, et de par leur fonction phatique provoquent une adhésion presque automatique.

JEUX ET ENJEUX DU DISCOURS RAPPORTÉ POUR LE DISCOURS ARGUMENTATIF

407

2.3 Les proverbes Concernant les proverbes, Maingueneau (1999) explique que lorsque quelqu’un dit un proverbe, il accomplit un acte de discours singulier, car il pose une assertion qu’il donne pour validée par une entité aux contours indéfinis, la «sagesse des nations». Il présente son dire comme l’écho d’un nombre illimité d’énonciations antérieures. On, l’asserteur du proverbe ne peut pas être considéré comme une instance vraiment distincte de l’énonciateur, car ce dernier fait partie de la communauté linguistique. Or les proverbes appartiennent au stock de savoir partagé par l’ensemble de la communauté linguistique, et c’est justement ce qui confère sa force au proverbe: alors que les citations d’autorité habituelles sont rapportées à des sources particulières (le Christ, tel savant, tel philosophe, etc.), l’autorité qui valide le proverbe coïncide avec l’ensemble des locuteurs d’une langue. 2.4 La diaphonie Au sujet de la diaphonie, Roulet (1985) introduit ce terme afin de préciser le concept de polyphonie. La diaphonie est «un des cas particuliers de voix dans l’énoncé, la reprise et l’intégration du discours de l’interlocuteur dans le discours du locuteur» (p. 70). La parole diaphonique s’oppose à la simple citation des paroles de l’interlocuteur en ce sens qu’elle suppose leur interprétation, et en ce sens «La structure diaphonique est ainsi une des traces privilégiées de la négociation des points de vue qui caractérise toute interaction» (p. 71). 2.5 La présupposition Pour Kerbrat-Orecchioni (1986), la présupposition réoriente la «vérité» des énoncés. Ce qui se passe au premier plan, c’est l’instance qui valide l’énoncé et comme elle le dit «toute assertion est prise en charge explicitement ou implicitement par un sujet énonciateur et c’est pour ce sujet d’abord qu’elle est vraie» (p. 56). Par le présupposé l’énoncé ne semble pas pris en charge par le locuteur, mais est présenté seulement comme une croyance. 3. Présentation des procédés polyphoniques relevés Nous allons maintenant passer à la présentation de l’analyse des procédés polyphoniques observés dans notre corpus d’entretiens et de débats. Nous présenterons tout d’abord l’analyse de l’utilisation du discours rapporté à partir de la présentation des différentes citations relevées, et ensuite nous passerons à celle de l’analyse des emplois des présuppositions.

PRISQUE BARBIER

408

3.1 L’emploi du discours rapporté Concernant l’emploi du discours rapporté nous avons relevé dans les entretiens comme dans les débats des citations en Discours Direct et en Discours Indirect. 3.1.1 En Discours Direct 3.1.1.1 Dans les entretiens -

Les citations par reprise

Dans les citations par reprise, les locuteurs reprennent des énoncés d’autres énonciateurs fictifs ou réels. 1. lorsqu’ils viennent aborder une fille il dit […] je vais t’envoyer dans tel endroit; lorsque je l’ai il me dit […] je vais passer 2. j’ai rencontré un homme de Dieu qui […] m’a dit Dieu fera je connais la peur de ton cœur 3. ma petite sœur […] m’a dit tu as vu ce que ce monsieur t’a fait … je dis non c’est pas grave de toute façon c’est la vie c’est comme ça il me dit […] je regrette de t’avoir abandonné mais je ne sais pas comment faire je dis non parce que si tu m’aimais tu n’allais pas me réveiller à minuit trente … ma petite sœur dit ah les mariages d’Abidjan c’est comme ça 4. elle ira dire à sa copine mais écoutez mon gars ne travaille pas alors bon celui-là […] celui-là ne lui donne rien moi mon gars m’a acheté tant

Dans les énoncés de I6 et I11, on a affaire à des citations d’autorité, car l’énonciateur du discours cité est une autorité reconnue par le locuteur (un homme de Dieu, ma petite sœur). Ces citations ont pour objectif de conférer de la crédibilité aux énoncés rapportés et donc au locuteur. - La citation d’une maxime 5. comme on le dit à l’école on va pour lire écrire et aussi on doit nous éduquer

Citer une maxime sert à rapporter un discours provenant d’une source indéfinie, et reconnu comme vrai. Ce type de citation fonctionne comme signe de connivence (Maingueneau 1991: 137). La maxime est donc utilisée pour provoquer une adhésion presque automatique de l’auditoire, car elle semble s’adresser à tout le monde. 3.1.1.2 Dans les débats La diaphonie La majorité des énoncés repris par les locuteurs sont ceux d’autres énonciateurs participants au procès d’énonciation.

JEUX ET ENJEUX DU DISCOURS RAPPORTÉ POUR LE DISCOURS ARGUMENTATIF

Débat 1: Les jeunes et la politique 6. comme le camarade l’a dit il doit mériter le jeune doit mériter sa place dans la vie politique 7. y a l’un des amis qui est un intervenant qui a dit la question d’être Ivoirien ou pas c’est un faux débat pourvu que la personne ait l’amour du pays 8. comme tout le monde on l’a dit la paix et que tout soit transparent

Débat 2: Les relations filles-garçons 9. euh parler des relations entre filles garçons comme certains l’ont déjà dit au delà des relations fraternelles on a aussi la camaraderie […] 10. je voudrais ajouter à ce qu’elle a dit s’il vous plaît je voudrais ajouter à ce qu’elle a dit c’est que l’amitié entre une fille et un garçon ça existe vraiment […] 11. moi je voudrais intervenir s’il vous plaît je suis d’accord avec Marcel quand il dit non qu’il faut d’abord établir une confiance entre les deux partenaires parce que les relations

Les citations Les citations par reprise 12. c’est ces mêmes gens qui ont dit si ils veulent pas aller à l’école laissez-les dans la rue nous on s’en fout déjà nous on mange

Les citations-preuve 13. excuse-moi ce que je voulais dire c’est que nous deux ça n’a jamais été le cas ça n’a jamais été le cas d’ailleurs c’est lui qui m’a dit Zénab t’es trop seule […] il a dit Zénab t’es tout le temps seule tout le temps seule c’est pas bien et tu donc je voudrais te présenter à un ami pour que tu sortes avec lui et tout j’ai dit non ça va pas marcher 14. mais si il revient et on discute de tout et de rien il me parle de ses problèmes il me demande des conseils il me dit ah voilà je vais prendre telle décision quel est ton avis? je dis ah voilà un tel il veut être un ami il cherche une amitié […]

Les citations-autorité 15. y a un leader politique africain qui a dit dis-moi quelle jeunesse tu as je te dirais quel peuple tu seras 16. oui je suis chrétien donc je suis la bible – qui dit pour qu’il y ait une véritable entre une fille et puis un garçon il faudrait que ces deux-là s’inspirent aussi de la bible il faudrait que ils prennent exemple sur ce qui a été écrit dans la bible

409

PRISQUE BARBIER

410

17. une fois que les rapports sexuels entre deux individus tout de suite il y a échange ces deux individus deviennent un une même chair alors que la bible nous dit l’homme quittera son père et sa mère pour ne s’attacher que à sa femme et tous deux ne feront qu’un

Les citations-culture 18. on a besoin toujours du plus petit que soi comme on le dit on a besoin aussi également des idées des autres […] 19. il ne faut pas dire que parce que on est des amis on ne doit pas par la suite devenir des amoureux puisqu’on dit l’amour naît au fil des temps […] 20. comme on le dit souvent sa bouche ne porte pas caleçon donc tes petits secrets que tu vas venir lui dire c’est sûr qu’elle va dire ça à quelqu’un d’autre […]

Ainsi, comme nous venons de le voir, les locuteurs se servent du discours direct pour intégrer soit des citations internes au procès d’énonciation pour corroborer leur thèse ou réfuter celle d’un autre participant, soit des citations issues d’une autorité ou de la culture générale pour étayer un argument. Par l’emploi des citations-autorité, le locuteur se sert de la crédibilité de la source de l’énoncé rapporté (ici le leader politique ou la bible) pour entraîner l’adhésion de l’auditoire à son argumentation. 3.1.2 En Discours indirect 3.1.2.1 Dans les entretiens Les citations par reprise Dans la citation suivante, la locutrice reprend les énoncés d’énonciateurs réels. 21. cet homme m’a dit qu’il ne voulait plus de moi; (il) m’a dit que il était l’homme que Dieu a envoyé pour moi; on me dit qu’il n’y a pas d’enfant dans mon ventre

Les citations par paraphrase 22. je suis d’accord pour ce que Savan Alla dit dans sa chanson disant que l’homme est un loup pour la femme 23. comme l’a dit mon ami c’est dû au fait que / les femmes surtout sont de plus en plus sont beaucoup matérielles

Dans ce cas, les Ivoiriens se servent de la paraphrase pour reprendre le discours d’énonciateurs absents ou présents (diaphonie) au moment du procès d’énonciation pour étayer leur argumentation et ainsi la renforcer.

JEUX ET ENJEUX DU DISCOURS RAPPORTÉ POUR LE DISCOURS ARGUMENTATIF

411

3.1.2.2 Dans les débats Les citations Les locuteurs se servent également de ce procédé pour citer des discours d’énonciateurs extérieurs au procès d’énonciation: dans le Débat 1 ils emploient des citations-culture, et dans le Débat 2 des citations-autorité. Débat 1: Les jeunes et la politique 24. on nous a toujours dit depuis l’école primaire qu’on ne doit pas faire la politique à l’école 25. je pense que on dit souvent que la femme est fait partie de la tranche de population la plus marginalisée la plus pauvre

Débat 2: Les relations filles-garçons 26. la bible dit que c’est normal c’est intéressant qu’une fille et un garçon aient des échanges – la bible dit qu’il faut être sincère il faut être sincère avec autrui donc si dès le départ tu es sincère avec la fille et que cette même fille-là est sincère avec toi vous êtes sûrs de ne pas déborder et c’est ce que la bible dit la bible dit aussi que c’est c’est l’homme l’homme il est appelé forcément à changer à vivre avec une femme […]

La diaphonie Dans ces exemples, les énoncés repris par les locuteurs sont ceux d’autres énonciateurs participant au procès d’énonciation. Nous avons ce type de reprise seulement dans le débat sur la politique. 27. pour ceux qui n’avaient pas suivi l’émission si Marius tenait à réfléchir sa qualité c’est parce qu’il avait dit qu’il avait derrière lui tout le peuple 28. voilà en fait je pense que dire le fait qu’elle dit que le président ou je sais pas doit avoir le l’amour du pays c’est un fait qui doit être pris en compte

3.2 L’emploi de la présupposition 3.2.1 Dans les entretiens 29. faudrait pas seulement voir le côté négatif des choses (comme certains) 30. on pourrait s’interroger d’abord sur / heu le mobile réel même de // disons l’intention profonde de Savan Alla

PRISQUE BARBIER

412

3.2.2 Dans les débats Débat 1: Les jeunes et la politique 31. l’idéal voudrait que le bon politicien soit celui-là même qui soit en mesure / d’écouter le peuple / d’être à la merci du peuple (sous-entendu: mais il n’est pas comme ça) 32. il faudrait pas qu’on se mette en l’idée que il faut forcément avoir des millions des milliards pour pouvoir briguer un poste politique (sousentendu: mais certains le pensent)

Débat 2: Les relations filles-garçons 33. je pense que sur ce point faudrait pas avoir de position tranchée (sous-entendu: comme certains) 34. qui dit pour qu’il y ait une véritable entre une fille et puis un garçon il faudrait que ces deux là s’inspirent aussi de la bible il faudrait que ils prennent exemple sur ce qui a été écrit dans la bible (sous-entendu: ce n’est pas toujours le cas)

On voit donc que dans les entretiens comme dans les débats, les locuteurs utilisent la présupposition pour invalider le point de vue d’énonciateurs fictifs, et ainsi introduire implicitement leur point de vue, mais sans l’assumer, notamment par l’emploi du marqueur générique On et de la modalisation passive faudrait, qui servent à généraliser les propos du locuteur afin d’impliquer l’auditoire. Conclusion Grâce à cette étude nous avons ainsi pu rendre compte du fait que les procédés polyphoniques sont utilisés dans un discours argumentatif pour exprimer l’opinion du locuteur de façon détournée. En outre, pour exprimer leur éthos, les Ivoiriens ont tendance à préférer un certain type de discours rapporté, notamment la citation de sources faisant autorité. Cette observation nous laisse à penser que cette forme de recherche de l’adhésion aux valeurs proposées, résulte d’un éthos hiérarchique dominant, d’une soumission aux valeurs religieuses, et également d’une surpolitisation du vécu des locuteurs. En effet, ce style discursif, corrélé à l’effacement énonciatif du Je, n’est pas étranger au statut de l’individu dans une société aux fondements collectivisants. Par conséquent, cette étude nous permet de dire que tout discours, et à plus forte raison encore un discours argumentatif, est le reflet des normes culturelles qui régissent le fonctionnement de la société dans laquelle il est construit et mis en œuvre.

JEUX ET ENJEUX DU DISCOURS RAPPORTÉ POUR LE DISCOURS ARGUMENTATIF

413

Bibliographie Authier-Revuz, Jaqueline (1995). Ces mots qui ne vont pas de soi: boucles réflexives et noncoïncidences du dire. Paris: Larousse. Bakhtine, Mikhail 1929 (tr. fr. 1977). Le marxisme et la philosophie du langage. Paris: Ed. de Minuit. Ducrot, Oswald (1980). Les mots du discours. Paris: Ed. de Minuit. Duponchel, Laurent (1979). Le français en Côte d’Ivoire, au Dahomey et au Togo. ILA: Abidjan. Kerbrat-Orecchioni, Catherine (1986). L’implicite. Paris: Armand Colin. Kouadio N’Guessan, Jérémie (1998). Le français devant une variété autonome de français. Document inédit. Lafage, Suzanne (1980). Petite enquête sur la perception du français populaire ivoirien en milieu estudiantin. Bulletin du Centre d’Etudes sur le Plurilinguisme 6/80. Lafage, Suzanne (2002). Le lexique français de Côte d’Ivoire, appropriation et créativité, (tome 1 et tome 2). RROFCAN 16 et 17, Nice: Institut de Linguistique française, CNRS. Maingueneau, Dominique (Ed. 1999). L’énonciation en linguistique française. Paris: Hachette. Manessy, Gabriel (1990). Modes de structuration des parlers urbains, in Gouaini E. & Thiam N. (ed.) Des langues et des villes, Paris: Didier Erudition. Roulet, E., Auchlin, A., Moeschler, J., Rubattel, C. & Schelling, M. (1985). L’articulation du discours en français contemporain. Berne: Peter Lang. Simard, Yves (1994). Les français de Côte d’Ivoire. Langue Française 104.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 415-426 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

APPROACHING ARGUMENTATIVE DIALOGUE WITH FORMAL MODELS GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

1. Introduction From the linguistic point of view, dialogue can be understood as an exchange of speech acts with its rules and constraints, with the main objective being communication. From pragmatic theory, dialogue could also be defined as a game where the participants try to maximize the possibilites of success in their argumentation in order to achieve their own goal. Any contribution to the study of dialogue is in debt with the theory of speech acts, introduced by Austin (1962) and Searle (1969), that has become one of the central issues of pragmatics and human communication traditionally tackled by artificial intelligence. The theory of dialogue acts has made it possible to establish units in dialogue and has allowed the start of the dialogue grammars approach (Sinclair & Coulthard 1975) as well as establishing several taxonomies of units (Sinclair & Coulthard 1975; Traum & Hinkelman 1992). Computer science has been more interested in the structure and the interaction of agents in conversation (Litman & Allen 1990) rather than in single utterances. From that perspective, some authors (Sinclair & Coulthard 1975; Coulthard et al. 1981) have distinguished different levels inside the dialogue: act, move, exchange and transaction. Such classification has become classical, especially referring to moves (Kowto et al. 1993; Traum & Larsson 2003). Traum and Hinkelman (1992) give a new perspective in the topic gathering different types of acts in four groups: turn-taking, grounding, core speech acts and argumentation. One of the novelties of this taxonomy consists in the inclusion of turntaking into such categories. A similar idea can be found in Bunt (2005), who defends the existence of taskoriented acts and dialogue control acts. Both types modify the linguistic and cognitive context. However, task-oriented acts only change the semantic context and dialogue control acts mainly affect the social or physical context. He considers turn taking to be included in interaction management functions, which belong to dialogue control functions. By means of Grice’s maxims, participants are required to be cooperative and they have to follow a behavior that is related more to politeness than to linguistics. But many times, contra to Grice’s claims, participants have secret purposes that are not known by the other agents. In this sense, Reed and Long (1997) make an interesting

416

GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

distinction between cooperation and collaboration. For a dialogue to be brought about, cooperation is necessary, but collaboration not always exists. For us, a crucial and non-static element in dialogue is context, understood as the environmental and personal states and circumstances that can affect the development of the dialogue. This context is in constant evolution, not only because of external factors, but also because of the speech acts performed by the participants. Therefore, like Bunt (1995), we think that the configuration of the dialogue is directly related to the intentions of the speakers/hearers and to the context. In what refers to the types of dialogues according to human argumentation, Walton and Krabbe (1995) introduced a taxonomy that has become classical. They distinguish between information seeking, inquiry, persuasion, negotiation, deliberation and eristic dialogues. Our work is mainly focused in deliberation, a kind of dialogue in which participants have to reach an agreement and make a decision. We approach deliberation from the perspective of dialogue games following the research line started by Carlson (1983). Many authors have tackled dialogue as a game (Mcburney & Parsons 2002a, 2002b; Prakken 2006), although most of them have focused on persuasion (Hitchcock et al. 2001; Glazer & Rubinstein 2001, 2005). Glazer and Rubinstein introduce an interesting variant, which they name debate, based on game theory. In their model, two opponents have to convince a third person, which is the one deciding. This paper is an attempt to start an approach for deliberation dialogue games accounting for argumentation, using some tools provided by mathematics, theoretical computer science and game theory. The definition of deliberation is derived from the general definition of dialogue, with the important restriction of the number of agents A, that in this case is given by A=2. In what refers to game theory, we use the extensive form of games representation because we assume the participation of the speakers is sequential and they alternate in turn. In this simple model, then, turn taking is already established, given by the number of agents. In deliberation, agents interact to decide what action or course of actions should be adopted in some situation. Grice (1975) claims that cooperation is basic for the correct development of conversation from the pragmatic perspective. However, our work is based on the ignorance of agents’ real attitude. We assume they could not follow Grice’s principles. Therefore, participants in the system have one of the following configurations: a) participants have secret intentions, and b) participants are neutral. In this research, we are mainly interested in defining games where the participants in the deliberation have secret intentions. In the sequel, the term dialogue refers to “deliberation dialogue’’. Summing up, we are interested in the formalization of deliberation with two agents that have secret goals, and in the analysis and optimization of their moves, tak-

APPROACHING ARGUMENTATIVE DIALOGUE WITH FORMAL MODELS

417

ing some methods from game theory. In Section 2 we introduce the main definitions of dialogue and deliberation. Section 3 shows an example. In Section 4 some strategies for improving the possibilities of success are discussed. Finally, Section 5 provides some discussion and future lines of work within this framework. 2. Definitions 2.1 Dialogue First, we introduce a general definition for a formal dialogue, that must be adapted in order to account for deliberation. The main features of the definition below are: a) the number of agents A is, by default, A>2; b) the sets of stores are selected for approaching spontaneous, non task-oriented conversations; c) turn taking is not established; d) during the conversation, agents can become disconnected, participants can leave the scene, and new ones can enter. The formal framework for dealing with this type of conversations would be the following: Definition 1: Dialogue systems can be defined as a 4-tuple (A, Θ, Σ, G), where: • A = {A1, A2, . . . , An} is the multi-agent structure, where every one of the agents has a configuration Ai = (Ri, Ci), being: − Ri the set of rules. Every one has:  At the left side turn-taking for the rule to be applied.  At the right side: a)the rule of generation, and b)the agent whom the speech act is addressed to, if it exists. − Ci is the state of the communication channels. Every agent has bi-directional (input/output) channels with others, which can be set either as open or closed. For the communication to be possible between two given agents An and Ai, the channel has to be open in both directions. For the other combinations (open-closed, closedopen, closed-closed) communication is not possible. • Θ = {κ, α, γ, φ, π, ε, σ} is the set of stores of core speech acts, being their configuration as follows: κ = {k1, k2, ... , kn}, query; α = {a1, a2, ... , an}, answer; γ = {g1, g2, ... , gn}, agree; φ = {f1, f2, ... , fn}, reject; π = {p1, p2, ... , pn}, prescription; ε = {e1, e2, ... , en}, clarify; σ = {s1, s2, ... , sn}, assert. • Σ={K (Query), A (Answer), G (Agree), F (Reject), P (Prescription),E (Clarify), S (assert)} is the set of turn-taking dialogue acts. • G = {G1, G2, ... , Gn} is the set of registers of the system, the place where the dialogue games are stored. There is a different generation register for every conversation started by the system.

418

GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

2.2 Deliberation Bearing in mind the general formalization of dialogue, now a formal description of deliberation has to be given, by constraining and adapting some of the elements of the previous definition. First of all, in order to approach deliberation, we establish the existence of only two agents A1, A2 which are the only participants. This is a methodological option to simplify the description of the system, but the formalization can be extended to n participants, only by adding the configuration of A in the definition we will introduce in this section. The existence of only two agents implies that the turn taking is previously established as an alternance of A1 and A2, in a way that no other protocol is needed. Therefore, Σ and the transitions and rules it implies are not necessary for this simple case. When the number of agents is A>2, then Σ must be included again in the definition of deliberation. Each one of the agents has a set of dialogue acts Θ (A1), Θ (A2). Each one of these sets is part of an acts store Θ = {p, r, s, a, q, x}. Such a store is different from the one of the general case of dialogue, as the acts are especially selected to fit an argumentative exchange. In Θ, p and s are two different types of arguments, r is a counter-argument rejection, a is acceptance, q is a question and x indicates that an agent is quitting the dialogue. We also establish that r and a, cannot be initial productions of the dialogue because they are only valid as a counter-argument. R is a set of combinations of argumentation-counter argumentation that relates elements from the set Θ (A1) to acts belonging to the set Θ (A2), or vice versa. These rules are different for each one of the agents or participants, and they have the form p → q. Every agent has its own set of rules, R1 for A1, R2 for A2, Rn for An. If single elements are found in the sets of rules of the agents, they can be used only as a “starting production”. They are, then, the starting symbols of the system. If both agents have starting acts, only A1 will be able to use them, since this is the agent, which, by definition, starts the debate, if it has arguments to do that. The participant that starts the dialogue is A1, if it has some argument that is not generated by another argument, i.e. some single symbol in R1. We denote a production w of an agent An in a given state as An (w), and the set of possible productions for an agent An in a given state as θ (A1). The possible outcomes of the deliberation are represented with upper-case roman letters. They are part of the set O, such that O = {A, B, C...}. Some of the elements of Θ are associated to elements of O by an application F. Such elements are named terminal acts.

APPROACHING ARGUMENTATIVE DIALOGUE WITH FORMAL MODELS

419

Definition 2: Having two speakers A1 and A2, a deliberation game G between them is defined as a 4-tuple: G = (Θ, R, O, F) where: • Θ is an acts store; • R = R1 U R2, ... U ..., Rn is the set of argumentation rules for each agent; • O is the set of possible outcomes of the deliberation; • F is an application relating elements of Θ to elements of O. Such application is denoted by the symbol ‘→’. If there is not an O element for a sign belonging to Θ, then the result is Ind, which means that the outcome is undecidable and the deliberation has to go on. The graphic representation of the deliberation game can be seen by means of a treediagram. For this we follow the regular elements and terminology. Using the tree graphic and the concepts that have been defined, we introduce several statements for the description and study of the properties of a deliberation dialogue. Description 1: Tree-Diagram The tree-diagram will show all the possible productions of the game, where the nodes are the agents speaking and the edges denote dialogue acts. Every different branch in the tree determines a possible trajectory. Description 2: Terminal nodes Terminal refers to the nodes that cannot be developed any more at the end of the debate, which corresponds to the classical definition of “terminal”. Description 3: Final nodes Final nodes are terminal nodes after a given move, that is, the nodes that, after the application of F are not labelled wit Ind. Nodes Ind are final but non terminal nodes. Description 4: Trajectories A trajectory of dialogue is every lineal path of the tree starting at the initial node. A complete trajectory is every path from the starting utterance to a terminal symbol. There are as many trajectories in a dialogue as the final nodes it has. There are as many complete trajectories in a dialogue as terminal nodes it has. There are as many possible trajectories in a given move as the new nodes it produces. Description 5: Trajectories in deliberation Being G a deliberation game for agents A1 and A2, and Θ = {w} the acts store, we denote a trajectory of this game with the form G(w1, w2 , ..., wn ), being w1, w2, ... ,wn the utterances generated to reach the final agreement in order of generation. Since a dialogue has as many trajectories as final results, then we say that a G = {G1, G2, ... ,Gn}.

420

GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

Description 6: Width The width of a dialogue w(G) is the maximal number of trajectories it has. The trajectories are ordered starting with the leftmost and finishing by the rightmost. Description 7: Paired trajectories We call paired trajectories those that have an even number of edges and unpaired trajectories those that have an odd number of edges. Description 8: Move We define a move M as an adjacency pair that consists of argument and counterargument. A sequence is a set of moves Mm, Mn, ...,Mi. A debate can have one or more moves. As in real life, some debates stop after a number of productions that has been determined before, and others can be calculated after all possibilities have been explored. Description 9: Move productions The productions generated after a move Mn are θ(Mn). In θ(Mn) two types of acts can be distinguished: final f(Mn) and terminal t(Mn). The state of the dialogue after Mn, denoted Θ(Mn), includes θ(Mn) and all the terminal acts before Mn, denoted by T(Mn). Being Mm, Mn the first and second moves in a debate, it is clear that in Mm, Θ(Mm) = θ(Mm), while in Mn, Θ(Mn) = t(Mm) U θ(Mn). Being M = {Mm, Mn, ... , Mi}, Θ(Mi) = t(Mm) U t(Mn)... U ...θ(Mi). If in a given move Mn, θ(Mn) = t(Mn), then the debate is complete. The results of the productions in a move Mn are designed by g(Mn), and they are obtained by applying F(θ (Mn ) =>O). The results of the debate after the move Mn are denoted by G(Mn), and they are obtained by applying F(Θ (Mn) => O). Description 10: Final Generation A generation is a final generation if it is the last generation that has been stipulated for a debate. A generation is a terminal generation if it produces an agreement between agents. Terminal generations produce complete nodes because, after the agreement, the dialogue does not develop any more. Description 11: Finished and complete dialogues We say a dialogue is finished, when the number of moves that have been assigned to it, have been performed. We consider a dialogue to be complete when every one of the acts generated in a production has an outcome in O. Description 12: Finite and infinite deliberation games We name finite dialogues those, which can be completed in n number of moves. For example a finite 2-complete dialogue is the one that can be completed performing two moves. An infinite dialogue is the one that cannot be completed. For debates that last until they have found an agreement for every possibility, a finished debate is a complete debate. This is only possible for finite dialogues.

APPROACHING ARGUMENTATIVE DIALOGUE WITH FORMAL MODELS

421

Description 13: Degree of a deliberation game Degree of a dialogue deg(G) is the length of the shortest trajectory to reach an agreement. Description 14: Depth of a deliberation game Depth of a dialogue depth(G) is the length of the largest trajectory to reach an agreement. For infinite dialogues the depth of the dialogue is ∞. Description 15: Monotonous / non-monotonous / undecidable deliberation dialogues If there is a debate where the deg(G) = depth(G) then such debate is called monotonous. When deg(G) ≠ depth(G) then the debate is non-monotonous. A debate where deg(G) = depth(G) = ∞ is an undecidable debate. 3. Example In this section, we introduce an example of debate. We do not establish a number of moves to compute, in a way that the dialogue will continue if there is any undefined result after every move. We will study the basic features of the dialogue taking into account the concepts introduced above. Example 1: Let’s take G = (Θ, R, O, F), with: • Θ (= Θ(A1) = Θ(A2)) = {p, s, r, a, x}; • O = {A, B, C}; • •

F = {A1 (x)

A, A2 (x)

B, a

C};

R = R1 U R2 being: R(A1 ) = {p, s, r → s, r → x, p → x}, R(A2 ) = {p → x, p → r, p → a, s → a, s → p}. The deliberation game resulting from this formalization is shown in Figure 1.

422

GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

Figure 1: Tree-diagram It is easy to see, observing the rules, that for A1, the secret goal is B, and for A2, the secret goal is A, since when A1 quits the deliberation the result is A, and when A2 quits the game the outcome is B. Looking at the diagram we can say that this is a finite 3-complete dialogue. The debate consists of two paired moves and one unpaired move: deg(G) = 2 and depth(G) = 5. This is, then, a non-monotonous debate. The width of the dialogue, width(G), is 7, being the trajectories G1 (px), G2 (prsa), G3 (prspx), G4 (prx), G5 (pa), G6 (sa), G7 (spx). G1, G2, G5, G6 are paired trajectories, whereas G3, G4, G7 are unpaired trajectories. The productions generated after the first move are θ(M1) = {x, r, a, a, p}. Θ(M1) = θ(M1). Applying F to this outcome we obtain the result of the deliberation after the first move, which is g(M1) = {B, Ind, C, C, Ind}. By means of the second move we obtain θ(M2) = {a, p, x, x}. Θ(M2) = t(M1) U θ (M2), being g(M2) = {C, Ind, A, A} and G(M2) = {B, C, Ind, A, C, C, A}. The third move is an unpaired one, because there is only one Ind node, and it is complete after one generation. Therefore θ(M3) = {x}. Θ(M3) = t(M1) U t(M2) U θ (M3), being g(M3) = {A} and G(M3) = {B, C, A, A, C, C, A}.

APPROACHING ARGUMENTATIVE DIALOGUE WITH FORMAL MODELS

423

4. Optimization Strategies The analysis of the formal properties of a dialogue must help the agents to calculate their possibilities to reach a good agreement. The example introduced in the last section can be a tool for introducing simple calculus of the probabilities of success. As it has been already said, the secret goal of A1 is B and the secret goal of A2 is A. Even if both of the agents want to reach an agreement, they both have a clear order of preferences. For both agents, the first option is their preference, and the second the neutral agreement C. Finally, the last option is the preference of the opponent. Since the debate is complete, there is not the possibility of leaving the dialogue without a result. There are three possible results, O = {A, B, C}. Let’s give a score for each one of them: 3 for the best option, 2 for the second, 1 for the worst. Table I shows the scores for agents A1 and A2. A

Table I: Scores Formally, it can be said that O(A1)={B, C, A}, O(A2)={A, C, B} With these scores for every result, two techniques can be implemented: horizontal scoring and branching scoring. Description 16: Horizontal scoring Horizontal scoring is the calculation of the possibilities of success for every agent by uttering each one of the possible speech acts in a given move. Description 17: Branching scoring Branching scoring is the calculation of the possibilities of final success in a trajectory of a deliberation game for every agent. Let’s take the perspective of A1. After M1, g(M1) = {B, Ind, C, C, Ind}. Applying the scores of Table 1, we obtain g(M1) = {3, 2, 2}. After M2 g(M2) = {C, Ind, A, A}, therefore g(M2) = {2, 1, 1}. Finally, after M3 we obtain g(M3) = {A} and g(M3) = {1}. From these scores, and in a scale from 1 to 3, the average score for A1 if the dialogue finishes in the first move is 2,3. The average score if the dialogue does not finish in the first move and a second move is necessary, 1,3. Finally, if the debate does not finish with two moves and a third move is necessary, the score for A2 will be just 1, this is, the minimal. Let’s take the perspective of A2. If the debate finishes in the first move, then the average score for this agent will be g(M1) = {1, 2, 2}, which is 1.6. However, if the debate goes on, then we get g(M2) = {C, Ind, A, A}, which is g(M2) = {2, 3, 3}, this is an average of 2,6. Finally, if the game needs a third move, then the average for A2 is the

424

GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

best, because g(M3) = {A}, and the score for A is 3 in A2. The outcome obtained by scoring the results in each move is shown in Table II. M1

2,3

1,3

1,6

2,6

Table II: Scores by moves Table II shows how, for A1 the best option is to try to finish as soon as possible. However, the possibilities of obtaining a good agreement increase for A2 when the debate lasts for more than one move. It seems, then, that its first act should be s, because the shortest trajectories come from the right. However, this is not accurate. In the first move, A1 has the possibility of finishing with a good result 2/3 on the left and 1/2 on the right. On the other hand, the only possibility to reach its secret goal is playing on the left. In general A1 has a 50% possibility of getting a good result starting on the right and 60% if starting on the left. A1 has to try to start on the left and finish at the first move. A2 cannot choose the starting symbol, and it has to wait for A1 in order to give its counter-arguments. If A1 starts on the right, A2 has an 83% possibility of success, whereas the percentage is 73% if A1 starts with p. Broadly speaking, A1 prefers trajectories starting with p and A2 prefers trajectories starting with s. But in all cases, it seems A2 has more possibilities to “win” the debate. The best trajectory for an agent is the shortest one with score 3. For A1 the best trajectory is G1. For A2 the best ones are G4 and G7, but it has to prefer G7 because starting with s it always achieves a good score. But if we examine the trajectories and their possible accomplishments, we have the following: G1 = 1/6, G2 = 1/24, G3 = 1/24, G4 = 1/12, G5 = 1/6, G6 = 1/4, G7 = 1/4. 5. Discussion By means of the method introduced in this paper we intend to explore some mathematical properties of deliberation dialogues, and the possibilities that the agents have to achieve a good agreement. By introducing this approach, we try to tackle dialogue, and especially deliberation, from a perspective that can help to: • simulate argumentative processes in/with computers, • explore some mathematical properties of argumentation and, • introduce a method for the formalization of some problems in pragmatics of dialogue.

APPROACHING ARGUMENTATIVE DIALOGUE WITH FORMAL MODELS

425

The system we present in this paper allows us to relate the behavior of the participants to pragmatics, especially to the context, to the behavior of the other agents, to the content of the utterances and to the secret intentions of the speakers. We think this method can be very useful to achieve an optimization of the arguments and counterarguments produced in deliberation dialogues. Since this is a first approach to the topic from the formal-computational perspective, the method is not very realistic, mainly because of three reasons: a) agents know everything about the other agents, they see all the time the general acts store, b) agents can change the strategy, but they cannot add new arguments to their stores, c) the final result of the dialogue can be calculated from the first move; it is deterministic. Therefore, in order to research this area more, we suggest performing the following improvements: a) the number of agents can be A>2, b) the agents can change strategies, add new acts to the stores and make new rules, c) agents can be blind, this is, they do not see the other agentsâ&#x20AC;&#x2122; stores. With these new features, the method can make a small contribution to achieving real micro spaces of artificial intelligence, being more realistic and human inspired. References Austin, John L. (1962). How to Do Things With Words. Oxford: Oxford University Press. Bunt, Harry (1994). Context and dialogue control. Think 3: 19-30. Bunt, Harry (2005). A framework for dialogue act specification. Presented at the Fourth ISO/ SIGSEM Workshop on Multimodal Semantic Representation, Tilburg: The Netherlands. Available at: http://let.urt.nl/general/people/bunt/docs/fdas.ps Bunt, Harry & Yann Girard (2005). Designing an open, multidimensional dialogue act taxonomy. In: Gaiffe, C.G.B. (ed.). Proceedings of the Ninth International Workshop on the Semantics and Pragmatics of Dialogue. Nancy, June 9-11, 2005. Carlson, Lauri (1983). Dialogue Games. An Approach to Discourse Analysis. Dordrecht: Reidel. Coulthard, Malcolm, Martin Montgomery & David Brazil (1981). Developing a description of spoken discourse. In: Coulthard, M. & M. Montgomery (eds.). Studies in Discourse Analysis. London: Routledge & Kegan Publishers, 1-50. Glazer, Jacob & Ariel Rubinstein (2001). Debates and decisions: On a rationale of argumentation rules. Games and Economic Behavior 36: 158-173. Glazer, Jacob & Ariel Rubinstein (2005). A Game Theoretic Approach to the Pragmatics of Debate: An Expository Note. In: Benz, A. & G. Jaeger & R. van Rooij (eds.). Game Theory and Pragmatics. London: Palgrave Macmillan, 251-265. Grice, Paul (1975). Logic and Conversation. Cambridge, MA: Harvard University Press. Hitchcock, David & Peter McBurney & Simon Parsons (2001). A framework for deliberation dialogues. In: H.V. Hansen, C.W. Tindale, J.A. Blair, & R.H. Johnson (eds.). Proceedings of the Fourth Biennial Conference of the Ontario Society for the Study of Argumentation. Kowto, Jacqueline & Stephen Isard & Gwyneth Doherty (1993). Conversational games within dialogue. Technical Report Research Paper HCRC/RP-31, Human Communication Research Centre, University of Edinburgh.

426

GEMMA BEL-ENGUIX & M. DOLORES JIMENEZ-LOPEZ

Litman, Diane & James Allen (1990). Discourse processing and common sense plans. In: Cohen, P. & J. Morgan (eds.). Intentions in Communication. Cambridge (Mass): MIT Press. McBurney, Peter & Simon Parsons (2002a). Dialogue games in multi-agent systems. Informal Logic, 22(3): 257-274. McBurney, Peter & Simon Parsons (2002b). Games that agents play: A formal framework for dialogues between autonomous agents. Journal of Logic, Language and Information 13: 315-343. Prakken, Henry (2006). Formal systems for persuasion dialogue. The Knowledge Engineering Review, 21: 163-188. Reed, Chris & Derek Long (1997). Collaboration, cooperation and dialogue classification. In Jokinen, K. (ed.). Working Notes of the IJCAI97 Workshop on Collaboration, Cooperation and Conflict in Dialogue Systems: 73-78. Searle, John (1969). Speech Acts. An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Sinclair, John & Malcolm Coulthard (1975). Towards an Analysis of Discourse: The English Used by Teachers and Pupils. Oxford: Oxford University Press. Traum, David (1999). Speech acts for dialogue agents. In: Wooldride, M. & A. Rao (eds.). Foundations of Rational Agency. Dordrecht: Kluwer, 169-201. Traum, David & Elizabeth, Hinkelman (1992). Conversation acts in task-oriented spoken dialogue. Computational Intelligence 8 (3): 575-599. Traum, David & Staffan Larsson (2003). The information state approach to dialogue management. In: Kuppevelt, J.V. & R. Smith (eds). Current and New Directions in Discourse and Dialogue, Dordrecht: Kluwer. Walton, Douglas & Eric Krabbe (1995). Commitment in Dialogue: Basic Concepts of Interpersonal Reasoning. Albany, NY: SUNY Press.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 427-434 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

DIRE SENZA DIRE, ARGOMENTARE SENZA ARGOMENTI ADELINO CATTANI

L’intervento tratterà dei significati e dei discorsi implicitamente argomentativi e si incentrerà sul valore persuasivo dei discorsi narrativi in apparenza non argomentativi. Descrivere e argomentare sono due atti molto diversi; e costituiscono una cosiddetta coppia filosofica. Vorrei qui, in primo luogo, mostrare che non si tratta di una coppia di nozioni esclusive, né semplicemente di una coppia di azioni complementari; in secondo luogo, intendo esaminare la questione di quando e come un testo descrittivo diventa un testo argomentativo. In altri termini e più precisamente, mi propongo di vedere quando e come una descrizione non è puramente la componente di un argomento (in quanto premessa esplicitamente esibita) o un’asserzione a sostegno di una tesi (è naturale che una narrazione o una descrizione vivide conferiscano forza al discorso), ma quando è essa stessa un argomento. Il presente saggio concerne pertanto la valenza argomentativa e persuasiva delle nostre descrizioni e narrazioni, che diventano quindi delle pseudo-esposizioni fattuali o pseudo-rappresentazioni, degli pseudo-rendiconti o pseudo-classificazioni. Prendi due e paghi uno, potrebbe essere lo slogan. C’è chi non ha nulla da dire. C’è chi non ha argomenti di conversazione. Chi, in mancanza di argomenti, ricorre all’insulto. C’è persino un sito web intitolato “parlare senza argomenti”. C’è viceversa chi dice senza dirlo. C’è chi ragiona facendo finta di raccontarci fatti. Cioè: chi dice (implicitamente) senza dirlo (espressamente) e chi argomenta raccontando. L’informazione e la formazione si attuano di norma su testi che possono essere di natura narrativa o di natura argomentativa. Narrare e argomentare costituiscono anche le due funzioni principali del linguaggio. Tra il narrare e l’argomentare sussiste una naturale opposizione, tale da rendere la coppia una polarità esemplare. Le narrazioni sono esposizioni di dati acquisiti, di fatti, di emozioni; dovrebbero consistere in una trasmissione di dati già condivisi da tutti o dai più o dai “migliori”, secondo la nota gerarchia aristotelica, e che basta assimilare, fare propri. Ciò che conta è il dato. Le argomentazioni invece sono proposte di ragionamenti da valutare o di tesi controverse, sostenute da ragioni più o meno buone, che vanno giudicate. Ciò che conta è il nesso tra i dati. Si tratta di due modalità discorsive diverse, ma accomunate da una ricerca di efficacia comunicativa. Essendo diverse, la loro analisi e valutazione richiedono criteri e strumenti diversi. Avendo uno scopo comune, vi sono però delle sovrapposizioni di standard d’analisi e di valutazione.

428

ADELINO CATTANI

Un esempio di narrazione pura è certamente il romanzo. Uno spot pubblicitario è altrettanto chiaramente e banalmente un caso lampante di discorso che mira non a descrivere, bensì a far fare (indurre all’acquisto, modificare le opinioni o i comportamenti del destinatario del messaggio). Capita però che testi in apparenza narrativi ed espositivi abbiano una valenza surrettiziamente persuasiva. Anzi, molto spesso l’esposizione risulta il migliore argomento. Disporre di qualche strumento di analisi e di valutazione che consenta di individuare questa duplice natura è utile, se non indispensabile. La domanda è: come si può fare un discorso argomentativo-persuasivo esponendo o raccontando (o facendo finta di esporre e di raccontare)? Che cosa conferisce natura inferenziale ad un enunciato-resoconto? Teniamo presente che un argomento è un ragionamento e un ragionamento deve avere per forza almeno tre componenti, quelle classiche del sillogismo: una premessa maggiore, una premessa minore ed una conclusione. Al fine che un testo espositivo diventi argomentativo si richiede che: vi sia un’intenzione argomentativa che viene espressa implicitamente da chi parla e che viene colta espressamente da chi ascolta sulla base di regole invalse di interpretazione. È chiaro ed innegabile che alcune affermazioni hanno un’inerente dimensione argomentativa: ad esempio, una biografia può essere letta come un’autocelebrazione o un’autodifesa. Ma non è questo senso intrinseco che vogliamo attribuire all’espressione “valore argomentativo”. La teoria degli atti linguistici ci aiuta a chiarire questo punto: ciò che si fa nel dire qualcosa è diverso da ciò che si fa col dire qualcosa. Dicendo (nel dire) “Ti avverto che piove”, sicuramente metto in guardia, in maniera convenzionale e automatica. Dicendo (col dire) “Ti avverto che piove” ti induco a non uscire, a prendere l’ombrello, a rammaricarti o a rallegrarti se la pioggia picchia argentina sui tetti dopo tre mesi di siccità. A differenza del primo caso, l’effetto del nostro dire richiede un’elaborazione diversa e non è garantito, in quanto non è determinato in maniera automatica e convenzionale, ma deve essere inferito. La cosa notevole è che l’inferenza riesce al destinatario più facile di quanto si penserebbe. Argomentare descrivendo e narrando Esempio 1a: “Milioni di persone hanno già comprato il telefono Nokia.” (Campagna pubblicitaria Nokia del 1994.) Esempio 1b: “Oltre due milioni di viaggiatori hanno già prenotato il volo on line con Easy-Flight.” (Da un pagina web di una agenzia di prenotazioni on line.)

Paiono delle constatazioni, delle informazioni documentabili rese pubbliche, resoconti di fatti obiettivi, dati numerici offerti alla pubblica opinione. In realtà 1a e 1b sono

DIRE SENZA DIRE, ARGOMENTARE SENZA ARGOMENTI

429

evidentemente un’operazione, neanche troppo nascostamente persuasiva, degli argomenti forniti per promuovere l’acquisto d’un prodotto o per valorizzare un servizio. Esplicitata la “forma ragionamento” sottostante è qualcosa come: Se milioni di persone acquistano un prodotto x, il prodotto x merita d’essere acquistato. Perché? Perché vale. E perché vale? Perché in milioni l’hanno comprato. Una mirabile petizione di principio e una premessa alquanto discutibile. Vi troviamo l’intenzione espressa implicitamente da chi parla e colta da chi ascolta, al quale rischia però di sfuggire la (inaccettabile) premessa che la regge. Esempio 2: “Anche ieri quindici milioni di telespettatori hanno visto Striscia la notizia”, ripetuto ogni sera dai conduttori della trasmissione, non è evidentemente un dato Auditel fornito in quanto dato, ma è destinato a suscitare la standing ovation e con essa l’autopromozione quotidiana del programma. Esempio 3: “È scozzese”. Significa che viene dalla Scozia, notazione etnica, oppure che è tirchio, notazione comportamentale, che si comprende solo grazie ad una implicita inferenza?

Anche questo è un caso di discorso intermedio che rientra al tempo stesso nella categoria del narrativo e dell’argomentativo. Scritto su una lavagna, senza altre informazioni, un lettore non potrebbe disambiguare il significato inteso. A questo provvede solitamente il contesto, così com’è il provvidenziale contesto a decidere se “gabinetto” è un “Ministrorum consilium” o una “Vespasiana latrina”. Ma anche il benedetto contesto e la pragmatica spiegano solo in parte perché la comunicazione sia felice quando potrebbe facilmente non funzionare. Esempio 4. Un quarto esempio è dato da un articolo di giornale. Un articolo de Il Giorno o de La Notte sulla legge finanziaria può sembrare espositivo, ma essere in realtà volto a convincere il lettore dell’opportunità o dell’iniquità di una data manovra economica e quindi essere di natura prettamente argomentativa sotto le mentite spoglie di un resoconto. Ribadiamo che molto spesso l’esposizione risulta il migliore argomento. La natura persuasiva dei fatti si desume anche dal loro essere considerati il linguaggio più efficace, intesi cioè come esempi offerti da individui credibili e coerenti. Questo dal punto di vista pedagogico. Dal punto comunicativo, si ricorda la raccomandazione che un capopartito rivolgeva ai suoi giornalisti: “fate parlare i fatti” - con l’ovvia intesa che i fatti sono manipolabili (Parlato 2006: 6). Esempio 5. Infine ricordiamo il famigerato “Pampers, i pannolini più usati nei reparti di maternità”. Perché sono i migliori oppure perché, ai tempi della propaganda in oggetto, erano distribuiti gratuitamente in tutti i reparti di maternità? Questo è un bell’esempio di come dire il falso dicendo il vero e testimonia l’importanza del rendere esplicite le premesse inespresse: in questo caso l’affermazione era effettivamente vera, com’era vero che negli anni Ottanta l’IBM era il computer più diffuso nelle scuole, ma solo in quanto ceduto gratuitamente alle istituzioni dalla stessa casa produttrice. Affermazione veritiera, pubblicità ingannevole.

430

ADELINO CATTANI

Come si può chiedere od ordinare semplicemente asserendolo? O zittire una classe proferendo solo un “pss, pps”? O addirittura far scoppiare una guerra semplicemente dicendo “guerra”? Anche in questo caso, John Austin ci offre una parziale indicazione di risposta. Aveva ragione il teorico dei cosiddetti atti linguistici: enunciato constatativo ed enunciato performativo si compenetrano. Il nostro dire è un fare. E viceversa, il nostro fare è sempre un dire. Facendo domande uno si racconta; chiedendo, ordinando, pregando otteniamo forse qualcosa, ma diamo sicuramente anche informazioni precise, a volte non volute, su di noi. Non solo nel senso, ovvio e banale, che se ordino un triplo whisky mi rappresento come un buon bevitore, ma soprattutto nel senso che in quello che diciamo c’è sempre molto, molto di più di quello che diciamo letteralmente. La nostra comunicazione è un iceberg, in cui la parte più consistente e più importante, quella che vogliamo trasmettere e anche quella che non vogliamo trasmettere, sta sotto il pelo dell’acqua. Eppure, pur essendo sotto traccia, viene colta di norma senza eccessiva difficoltà dal destinatario. La cosa sorprendente è che noi ci capiamo pur dicendo quasi sempre meno di quanto serve, mentre, si sa, fu la parte sommersa ed invisibile di un iceberg a far colare a picco il Titanic. Un buon insegnante per accertarsi se ciò che ha detto è stato inteso dalla classe può porre la domanda almeno in due modi: “Mi sono spiegato?” oppure “Avete capito?”. Le due modalità di interpellare gli alunni non sono affatto equivalenti. Se ricorre all’intercalare “Mi sono spiegato?” il docente mette in gioco anche se stesso e lascia aperta la possibilità che l’eventuale mancata comprensione da parte dell’uditorio sia determinata da una non perfetta spiegazione. La formula “avete capito?” invece scarica l’intera responsabilità della mancata ricezione sui poveri studenti. Anche questo è un modo di dire senza dire. Per non scoprire troppo quale sia il proprio modo di porsi nei confronti dell’uditorio, un relatore ha diverse opzioni alternative e asettiche rispetto all’ammissione o meno di responsabilità, quali “Chiaro?”, “Ci siamo?” o “Possiamo andare avanti?” Un uomo senza argomenti è il titolo di un saggio di Bjørn Ramberg (2000) su Richard Rorty e gli strumenti della filosofia. Diremmo noi, con un aforisma alla NietzscheMusil: un “uomo senza argomenti” non è un “uomo senza qualità” argomentativa. Già ne La filosofia come genere di scrittura, Richard Rorty afferma che “filosofi non kantiani come Heidegger e Derrida sono figure emblematiche di pensatori che non risolvono problemi, ma non hanno neppure argomenti o tesi da proporre” (Rorty 1986a: 109). Ma non è un rimprovero, anche se, a partire da Platone, “Lógoi dè toútou málista órganon” (Repubblica, IX, 582d), ovvero, “la ragione, l’argomentare è per eccellenza lo strumento del filosofo”. Se si identifica la capacità filosofica con l’abilità argomentativa, non v’è nulla che si possa considerare un argomento filosofico nel fiume di parole di Heiddeger e di Foucault. I filosofi analitici li ritengono persone che hanno tentato di essere filosofi senza riuscirvi, filosofi falliti, filosofi incompetenti. Ma “questo è tanto stupido quanto affermare che Platone era un sofista incompetente o che un riccio è una volpe incompetente” (Rorty 1986b: 223). In realtà “lo strumento del filosofo, l’argo-

DIRE SENZA DIRE, ARGOMENTARE SENZA ARGOMENTI

431

mentare, è, come la filosofia, un’impresa dialogica” (Rorty 1986b: 55). “La concezione dialogica dell’argomentazione ci permette di consegnare al passato l’opposizione fuorviante tra argomentazione razionale e riflessione creativa, innovatrice di vocabolari” (Rorty 1986b: 69). Ci sono stili filosofici diversi. Lo stile logico, razionalista di Cartesio e di Kant, i quali propongono catene inferenziali di ragionamenti. E lo stile aforistico, predicatorio di Nietzsche e Derrida, i quali enunciano e proclamano. Logos e pathos, direbbero i retori: ragionamento e illuminazione, convinzione e persuasione. L’importante è che ci sia almeno un rivolo di ethos che scaturisce dalla terza delle fonti aristoteliche della persuasione. Ci sono quindi modelli alternativi di argomentazione filosofica: quello sistematico (che può essere di stampo geometrico-deduttivo o di stampo scientifico-sperimentale) e quello oratorio (che può essere di stampo giurisprudenziale o di stampo prettamente retorico). I due stili convergono in quello che potremmo chiamare modello della finesse. La finesse non è solo una categoria estetica, ma anche logico-argomentativa: combina ordine e genio, rigore disciplinare e creatività inventiva. L’arte pascaliana dell’“user de finesse” consiste sì nella sottigliezza di ingegno ma anche nel giocare d’astuzia. In un saggio dedicato al ruolo della retorica nell’argomentazione razionale, Nicholas Rescher distingue due modi di fare filosofia, che chiama discorsivo ed evocativo. La filosofia di tipo discorsivo è imperniata su ragionamenti e caratterizzata dalla presenza di connettivi inferenziali. La filosofia di tipo evocativo è imperniata sulla retorica della persuasione e caratterizzata dalla presenza di espressioni di approvazione e di biasimo. La prima cerca di garantirsi l’assenso mediante giudizi argomentati, l’altra facendo appello a giudizi di valore. La filosofia della finesse e dell’indirettezza fonde in uno i due moduli stilistici. Rescher mette a confronto due passi, uno di Nietzsche, tratto dalla sua Genealogia della morale (II, 6) ed uno di Hume, tratto dal suo Trattato sulla natura umana (III, I, 2) facendo rilevare come in Nietzsche vi sia totale assenza di espressioni inferenziali. “Noi siamo chiaramente indotti a trarre certe conclusioni inespresse su base essenzialmente valutativa. Ma l’inferenza… è lasciata del tutto implicita come esercizio per chi legge. Questa conclusione non dichiarata alla quale mira la discussione è sottintesa ma mai affermata, implicata ma mai sostenuta” (Rescher 1998: 19). In questo modo Nietzsche “si sottrae alla fatica dell’argomentare e al fastidio delle obiezioni. Non curandosi di formulare esplicitamente la sua posizione, egli non sente il bisogno di darle un fondamento; si limita alla semplice insinuazione” (Ibidem). Facciamo un’operazione analoga, mettendo a confronto un passo di Cartesio ed uno di Nietzsche. …siccome i nostri sensi qualche volta ci ingannano, volli supporre che non vi fosse cosa alcuna che fosse tale e quale ce la fanno immaginare. E poiché ci sono uomini che si ingannano ragionando, anche intorno alle più semplici materie di geometria vi fanno dei paralogismi, giudicando che io ero soggetto a fallire al pari di chiunque altro, rigettai come false tutte le ragioni, che prima avevo preso per dimostrazioni. E, finalmente, conside-

432

ADELINO CATTANI

rando che tutti i medesimi pensieri, che abbiamo da svegli, ci possono venire anche quando dormiamo, senza che ve ne sia, allora, alcuno che sia vero, risolvetti di fingere che tutte le cose che mi erano mai entrate nella mente non fossero più vere delle illusioni dei miei sogni. Ma, subito dopo, posi mente che, mentre in tal modo volevo pensare che tutto fosse falso, bisognava necessariamente che io, che lo pensavo, fossi qualche cosa. E notando che questa verità: io penso, dunque sono, era così ferma e così sicura che tutte le più stravaganti supposizioni degli scettici erano incapaci di scuoterla, giudicai che potevo riceverla senza scrupolo come il principio della filosofia che io cercavo. (Cartesio, Discorso sul metodo, parte IV, trad. di G. Bontadini) Che cosa è buono? – Tutto ciò che nell’uomo accresce il senso di potenza, la volontà di potenza, la potenza stessa. Che cos’è cattivo? – Tutto ciò che discende dalla debolezza. Che cos’è felicità? – La sensazione del fatto che la potenza cresce, che una resistenza viene vinta. Non appagamento, ma più potenza; non pace in assoluto, ma guerra; non virtù, ma valentia (virtù nello stile del Rinascimento, virtù scevra da ipocrisia morale). Primo principio del nostro amore per gli uomini: i deboli e i malriusciti debbono soccombere. E bisogna anche dar loro una mano in tal senso. Che cos’è più nocivo di qualunque vizio? La compassione attiva verso tutti i malriusciti e i deboli – il cristianesimo… (Nietzsche, L’anticristo, § 2)

In breve, diremmo che i due stili filosofici, come gran parte delle coppie chiamate filosofiche, non sono in opposizione, né si escludono, ma sono più che complementari: una componente chiama e vuole l’altra. In una prospettiva ‘sistematica’ diremmo che non si tratta di due insiemi esclusivi. Con delle immagini ‘evocative’ diremmo che non sono come l’olio e l’acqua, o come il sole e la luna: se c’è l’uno non c’è l’altra. Quale delle due formule esplicative è migliore? Quella logica o quella oratoria? Al dilemma “Me ne vado o me ne vo?” il venerando filologo ricordò ai suoi fedeli discepoli riuniti al suo capezzale che “Ambedue dir si può”. Non farà scandalo quindi un Cartesio nietzscheano o un Nietzsche cartesiano e un Kant heideggeriano. Del resto i filosofi cosiddetti sistematici, alla Spinoza o alla Hegel, privilegiano la costruzione teorica, seguendo un movimento di pensiero dall’alto al basso: offrono cioè ragioni teoriche a giustificazione di fatti e preferiscono la pulizia teorica rispetto all’adeguamento ai fatti. Viceversa i filosofi evocativi, come Wittgenstein, partono dai fatti per risalire ai principi ed utilizzano i dati per coglierne le ragioni filosofiche, seguendo un movimento inverso dal basso verso l’alto. Quest’ultimo tipo di filosofo ama più del primo il dibattito e l’argomentare (cfr. Tomasini Bassols 1999: 242 e Dascal 1998: 19). Se può essere naturale l’impiego della narrazione nell’ambito quotidiano (cfr. Lundquist 1991) o nel contesto giuridico-processuale, in cui la sovrapposizione di storia e argomento è una regola e in cui il ‘fatto’ raccontato diventa un eccellente strumen-

DIRE SENZA DIRE, ARGOMENTARE SENZA ARGOMENTI

433

to di convinzione (cfr. Bennett 1978 e 1979, Bennet & Feldman 1981); se inoltre in molti si stanno occupando di “argomenti visuali” (altro modo specialissimo di argomentare senza argomenti), merita segnalare che c’è un recente interesse per il ruolo che la narrazione svolge in contesti impensati, come quello medico e quello matematico1. In conclusione, si può argomentare senza argomenti, come si può filosofare senza filosofia (nel senso classico del termine). L’indirettezza e la “finesse” non solo sono di fatto utilizzate, ma sono utilizzabili (cfr. Mason 1989, Warner 1989). Il ragionare semplice e chiaro, franco e diretto è una condizione ideale, ma sicuramente non reale. Anzi, in molti casi l’esplicito e il diretto non sono nemmeno il tipo di discorso ideale né ai fini estetici né ai fini della persuasione e della convinzione. C’è pure una finalità educativa della auspicata implicitezza, che è connessa alla varietà, alla versatilità, a quella che Quintiliano ed Erasmo da Rotterdam chiamavano copia; è, diciamo così, una finalità maieutica: “Non si deve scrivere in modo che tutti capiscano tutto, ma in modo da indurre uno a ricercare e a imparare per proprio conto” avverte Erasmo alla fine del capitolo diciottesimo del suo De Copia (Erasmo 1512). La copia quindi non è una mera regola estetica, ma altresì un mezzo funzionale: sfruttare la plasticità e la flessibilità del linguaggio dovrebbe diventare un'arte e un'abitudine in un quadro formativo. Proferiamo tante parole per nulla, direbbe l’aedo Shakespeare. Se ne dicon di parole, tante che ne può buttare, direbbe il vate-cantore d’oggi, all’insegna del risparmioso motto mercantile e cognitivo “due operazioni al prezzo di una”. Mostra e dimostra: esponi argomentando e ragiona facendo mostra di esporre, perché l’esposizione è il migliore argomento. È questo l’esito teorico-pragmatico di queste riflessioni. Ma attenzione a cosa c’è sotto quando si profitta dei saldi e quando si spacciano presunti fatti per argomenti e si fanno passare sottobanco premesse tutte da esplicitare e da soppesare. C’è il rischio Titanic di impattare contro la parte nascosta sott’acqua dell’iceberg e di affondare.

Un Convegno internazionale su “Le illness narratives nella professione infermieristica” si è svolto a Padova il 20-21 settembre 2005, a cura della Società Italiana di Sociologia della salute e del Dipartimento di Sociologia dell’Università di Padova, all’insegna della “narrazione come paradigma di cura”. Un convegno sul tema “Mathematics and Narrative” si è svolto nel 2005 a Myconos, con interventi, tra gli altri, di Karine Chelma, David Corfield, Apostolos Doxiadis e Timothy Gowers (cfr. www.thalesandfriends.org). Sul rapporto tra prova matematica e narrazione si può vedere Doxiadis 2005 e Thomas 2007.

434

ADELINO CATTANI

Riferimenti bibliografici Bennett, W.L. & Feldman, M.S. (1981). Reconstructing reality in the courtroom: Justice and judgment in American culture. New Brunswick, N.J.: Rutgers University Press. Bennett, W.L. (1978). Storytelling in criminal trials: A model of social judgment. Quarterly Journal of Speech 64, 1-22. Bennett, W.L. (1979). Rhetorical transformation of evidence in criminal trials: Creating ground for legal judgment. Quarterly Journal of Speech 65, 311-323. Dascal, Marcelo (1998). Types of Polemics and Types of Polemical Moves. In: S. Čmejrková et al. (eds.) Dialogue Analysis, VI, Proceedings of the Sixth International Conference. Prague 1996. Tübingen: Max Niemeyer Verlag, 15-33. Doxiadis, Apostolos (2005). Euclid’s Poetics. An examination of the similarity between narrative and proof. In: Michele Emmer (ed.). Mathematics and Culture II. Visual Perfection: Mathematics and Creativity. Berlin and Heidelberg: Springer Verlag, 175-182. Erasmo da Rotterdam (1512). De Duplici copia verborum ac rerum. In: Opera Omnia, Amsterdam, North Holland, 1969. Trad. inglese e commento di Betty I. Knott in: Craig R. Thompson (ed.) Collected Works of Erasmus 1 and 2, vol. XXIV, Literary and Educational Writings, 279-659, Toronto: Toronto University Press, 1978. Heitsch, Dorothea B. (1999). Nietzsche and Montaigne: Concepts of Style. Rhetorica 17: 411-31. Lundquist, Lita (1991). How and When are Written Texts argumentative? In: Frans van Eemeren & al. (eds.). Proceedings of the Second International Conference on Argumentation. Amsterdam: SicSat, 639-646. Mason, Jeff (1989). Philosophical Rhetoric. The function of indirection in philosophical writing. London e New York: Routledge. Parlato, Valentino (2006). La lezione di Togliatti e il macellaio Piero, Il Padova, Il Punto 5 novembre: 6. Ramberg, Bjørn (2000). Un uomo senza argomenti. Iride XIII: 52-69. Rescher, Nicholas (1998). The Role of Rhetoric in Rational Argumentation. Argumentation 12: 315-323. Rorty, Richard (1986a). La filosofia come genere di scrittura. In: Conseguenze del pragmatismo, Milano: Feltrinelli. Ediz. or. : Consequences of Pragmatism. Essays 1972-1980. Brighton: Harvester Press 1982. Rorty, Richard (1986b). La filosofia oggi in America. In: Conseguenze del pragmatismo, Milano: Feltrinelli. Thomas, Robert (2007). The comparison of mathematics with narrative. In: J.P. van Bendegem & B. van Kerkhove (eds.), Perspectives on mathematical practices: Bringing together philosophy of mathematics, sociolog y of mathematics and mathematics education. Dordrecht: Springer-Verlag, 43-60. Tomasini Bassols, Alejandro (1999). Los sofistas, Wittgenstein y la argumentación en filosofia. Tópicos. Revista de filosofia 17: 241-259. Voss, James F. (1998). The narrative as an Argument Component. In: Frans van Eemeren & al. (eds.) Proceedings of the Fourth International International Conference of the International Society for the Study of Argumentation. Amsterdam: SicSat, 846-848. Warner, Martin (1989). Philosophical Finesse. Studies in the art of the rational persuasion. Oxford: Clarendon Press.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 435-441 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

DISCURSIVE PERSPECTIVE AND ARGUMENTATION IN THE ROMANIAN PARLIAMENTARY DISCOURSE. A CASE STUDY LILIANA IONESCU-RUXĂNDOIU

1. Preliminary remarks On the 19th of April 2007, the two Chambers of the Romanian Parliament gathered in a joint session with a unique point on its agenda: to debate the proposal of President Traian Băsescu’s suspension from office. This debate was the result of an ambiguous decision of the Constitutional Court (CC), which admitted the infringement of the Constitution by the President, but declared his unconstitutional deeds as minor with regard to both their content and their consequences. Accordingly, the decision was transferred to the Parliament. As the great majority of the MPs voted in favour of the suspension (322 vs. 108), the President was suspended from office for 30 days and the president of the Senate was appointed to carry out his charges. The referendum organized on the 19th of May brought the President back to office, in spite of a high rate of absenteeism (participation of less than 45%) and of the fact that 25% of the participants were against. This paper has in view the intervention of a prominent member of the social democratic parliamentary group, the president of the Social Democratic Party (SDP), Mircea Geoană, in the above mentioned debate. The social democrats, the most important opposition party, were the authors of the suspension proposal, signed by 200 MPs. The speech under consideration has a very complex and elaborated structure. At the macro-discursive level, one can identify three basic units. The first, a long introduction defining the main goal of the parliamentary session, involves a sub-textual criticism directed towards the hesitating attitude of the CC. This institution considered that the notoriety of the facts invoked in the proposal of suspension from office could not offer a solid basis for a peremptory verdict and asked the Parliament to prove these facts. The speaker’s starting point is the axiom “The infringement of the Constitution by Traian Băsescu is a political reality”. The gravity of the President’s deeds is not proved, but only asserted in an insistent manner and in very general and abstract terms. The conclusion is that the existing system has reached its limits and that Traian Băsescu is “a failed political project”. His suspension from office would close a tragic chapter in our history: the reign of a leader’s subjective will and the myth of the providential leader. Our analysis will focus on this unit. The second unit lists the President’s major abuses. To his cynical formula “it will do”, the rhetor answers “it won’t do any more”, and concludes that this is “a worn-out

436

LILIANA IONESCU-RUXĂNDOIU

way of making politics”. The third unit presents the consequence of President’s behaviour, who continuously claimed his intention to be a “President-player”: Romania of the “Presidentplayer” is a country deeply affected by a profound crisis. The normal conclusion is: “Traian Băsescu is a closed chapter”; by voting in favour of his suspension from office, the MPs would offer the people the chance of a fresh start. Our analysis will focus on the first unit, setting out several basic argumentation strategies and techniques identified in the text. We shall try to bring forward – in a pragma-semantic analytical framework – the effects of some grammatical and lexical choices made by the speaker upon the argumentation processes. 2. The source of argumentation. The source-audience relationship. The discursive perspective As a component of the communicative activity, argumentation involves a source, an object (the target the arguments are focused on) and an audience that receives and evaluates the arguments. Two classes of personal forms, whose distribution is strongly unbalanced, make reference to the source of argumentation: the 1st person singular (A) and the quantitatively prevalent 1st person plural forms (B) of the personal and possessive pronouns, as well as of the verbs. A(a) The 1st pers. sg. appears mostly in meta-communicative utterances, which have a parenthetic use. They motivate some particular features of the discourse organization: (1) “I felt the need of an introduction – which you might find too long – in order to ensure a better understanding of our role today”, express or evaluate speaker’s own attitude in a certain moment of his performance: (2) “I can tell you, in front of God and of my own conscience that…”; “I am saying that in a very responsible way”, or express predictions about the attitude of the audience: (3) “Dear colleagues and citizens who are watching – I am convinced – with great anxiety this important session…”.

In example (2), the presence of the 1st person sg. seems to mark a self-disclosing attitude, but in fact speaker’s sincerity is part of the performance he is staging (see Goffman 1959: 17-18). (b) There is also an I marking the opposition and delimiting two polar points of view: (4) “Traian Băsescu says that only thieves need peace in order to better steal. I tell him that Romanians need peace because only in peace can they raise their children […]. And I also tell Mister Băsescu that thieves hide better and steal undisturbed when the riot is loud, as loud as possible.”

Repetition and parallelism amplify the antithesis.

DISCURSIVE PERSPECTIVE AND ARGUMENTATION IN THE ROMANIAN PARLIAMENTARY DISCOURSE

437

(c) A stronger expression of the speaker’s goals appears when urging the audience to adhere to his evaluation of President’s actions, in a formula with an explicit performative verb: (5) “Today I am solemnly asking you to show President Băsescu that this is a worn out manner of making politics”.

(B) Except for the above mentioned situations, the source of argumentation is ambiguously defined by the use of the 1st person pl. The inclusive meaning of this person creates a “useful ambivalence” (Wales 1996: 58), currently exploited in the political discourse. “We is one of the most political of all pronouns”, as T.A. van Dijk (2004: 360) puts it; it offers the speaker the advantage of disguising his subjectivity. We “is a shifting signifier” (Wales 1996: 62), placing the ego in various combinations (for their inventory functions, Wales 1996: 58-68), which sometimes are explicitly described by the speaker. In the case under consideration, there are several meanings covered by the 1st pers. pl. forms. (a) The speaker and the group of insiders from the audience (the other MPs). This meaning results from the reference to different actions involving only this part of the audience: (6) “What we are going to decide within this supreme forum of the Romanian democracy today is how serious the deeds of the Romanian President are”.

Still, from the very beginning, one can notice a strategic ambiguity introduced by the double opening formula of address: “Honorable colleagues, Dear Romanians”, which brings together insiders (MPs) and outsiders (ordinary citizens), as equally ratified participants in the debate. What follows is exclusively directed towards the MPs. One may wonder why the speaker explains in detail the aims and tasks of the parliamentary session, fixing an agenda to the MPs, when usually this is the obligation of the chairman of the session. This happens because – as already mentioned – his parliamentary group initiated the proposal of President’s suspension from office. The rhetor does not speak for himself, but in the name of the political group whose mandatory he is. His status of spokesperson of a certain group is strategically extended, tacitly including all the groups of MPs; the existence of largely shared views is consequently insinuated. This is a perverse persuasive strategy. One of its technical instruments is a repeated parallel between the unfulfilled mission of the CC and the mission of the Parliament, considered as a block. The formula repeated unchanged four times is: (7) “Legally, the Court did not have the possibility to evaluate the prejudices caused to Romania by [x]. Today we need to say that [y] is an extremely serious deed / is extremely serious.”

(where x and y are various labels used to classify different actions of the president, often in a metonymic cause – effect relationship. For example: x = “continuous attacks directed towards the main institutions of the state”; y = “blocking these institutions,

438

LILIANA IONESCU-RUXĂNDOIU

dragging them into a pointless war”). (b) The speaker and the political class in general. If in the case (a) there is no explicit reference to the MPs, in this case political class is explicitly mentioned, appositive constructions defining the speaker as one of its members: (8) “It is our collective fault, the fault of the whole political class…”; “The time has come for us, the Romanian political class, […] to propose a great national project to Romania”.

One can notice the same strategic tendency towards leveling the differences between the political parties. In spite of the objective semantic differences, the names political class and politicians are contextually used as quasi-synonyms of MPs, as it results from the following example: (9) “It is our duty as politicians towards the citizens who have elected us and who still have some hope from the Romanian parliament…”.

In comparison with the groups designated by different forms of the 1st pers. pl., the citizens or the Romanians – even if invoked as equally involved in the debate in the beginning (see above) – appear in fact not as outsiders, but rather as aliens (“not like us” – Wales 1996: 60). See the preceding example, as well as the following one: (10) “We should honestly tell Romanians what to expect in the future, and especially what we can and are trying to do for them.”

These examples fully reflect an elitist attitude of the politicians who consider themselves as a different community, in a superior position, in charge with the destiny of the whole nation, conceived as a passive mass of people. Accordingly, in this text, the distinction we/they does not concern different political parties (for a discussion about “them” and “us” attitudes, Wales 1996: 58-63), but politicians vs. citizens (or MPs vs. voters). This distinction is preserved in a more analytic review of those who could express an opinion: (11) “The Social Democratic Party and even I and maybe all of us and maybe the citizens of Romania hoped that at least at the 13 th hour, Traian Băsescu would eventually adjust his behavior”.

(c) Beside the above described cases of specific reference of the 1st pers. pl. forms (a and b), there are cases of generalized reference (for the distinction between specific exophoric reference and generalized homophoric reference, Wales 1996: 58-59). Sometimes the 1st pers. pl. forms express impersonal meanings: (12) “… we cannot speak about prosperity, equity, solidarity, justice or democracy in Romania” (=one cannot speak)

or indefinite meanings, with generic connotations (the Romanians):

DISCURSIVE PERSPECTIVE AND ARGUMENTATION IN THE ROMANIAN PARLIAMENTARY DISCOURSE

439

(13) “Our start in Europe risks being a failure”; “Unfortunately for him and for us, Traian Băsescu is…” “It is better to build (literal translation: that we build) a new Romania”.

Only in these situations, is the polarization politicians/citizens suspended in favour of the common national identity. What is sometimes suggested is the isolation of the President, who appears as the only person acting contrary to the general aspirations. Reinforcing the team views and extending them to the whole nation is another variant of the persuasive strategy of ignoring the differences, which we have already discussed. Speaker’s strategy of overlooking the politically heterogeneous structure of the Parliament and presenting the MP status as a unifying factor is determined by the basic goal of this speech: to get a vote in favor of President’s suspension from office. It has a strong impact on the general organization of the discourse as well as on the choices at the argumentative level (the nature and disposition of arguments, the means of argumentation). The speaker does not engage in polemics with other political groups. It is only an implicit criticism directed towards the CC, whose ambiguous decision was supposedly a result of the intimidating actions of the President. 3. The target of argumentation. Argumentative strategies and techniques The main target of an explicit sharp criticism is Traian Băsescu, pictured as an embodiment of a principle of evil. Ad personam sophisms, in their abusive variant (see van Eemeren & Grootendorst 2004: 177), aiming at discrediting the President, represent an important component of the argumentation system displayed by the text. This is a basic strategy in the process of motivating the proposal of the SDP parliamentary group. Closely connected with it is another type of strategic sophistic argumentation characteristic of the text: petitio principii. In the beginning, where the speaker was supposed to prove that president’s deeds are “extremely serious”, as repeatedly asserted, the whole presentation creates the impression of reminding facts which are very well known by everybody. In other words, the “facts” selected by the SDP spokesperson appear as part of the doxa. The doxa delimits a space which strongly determines the discourse situation where argumentation takes place (for the strategic functioning of the doxa in argumentation, see Amossy 2006: 99-125). In the case under consideration, it represents another strategic means of inducing the idea of a consensus between the MPs: no demonstration is necessary and, at the same time, being taken for granted, these facts cannot be contested. At a closer examination, the so-called “facts” are, actually, either too abstract, or subjectively interpreted. They are presented in a patterned sentential form, as possible definitions of some general terms, like deed or fact: (14) “Today, we need to say that [x] is an extremely serious deed / fact.”

We can add that the facts are named using infinitives with a nominal value (which can also be translated in English by –ing forms), whose meaning is abstract by definition

440

LILIANA IONESCU-RUXĂNDOIU

(to block the main institutions of the state; to drag them into a pointless war; to denigrate the magistrates and the law courts; to denigrate the Parliament, a fundamental institution of any democracy). As a matter of fact, all these are classificatory terms, involving generalizations. But as no mention appears about particular deeds of the incriminated President in order to ground such classifications, these terms reflect a rather subjective analysis of the facts. The key word of the introductory part of the speech is the adjective serious, which carries an intrinsic axiological (negative) meaning. It is the label applied to President’s deeds, getting an important argumentative function by its rhetoric repetition. There is a long chain of negative terms which strengthen each other: abuses, excesses, blackmail, intimidation, prejudices, refuse, denigration, attacks; the above mentioned nominal infinitives can be added too. They form an isotopic discourse core, boosting the evaluation expressed by the adjective more efficiently than the arguments themselves and accordingly assigning a certain argumentative orientation to the discourse (Amossy 2006: 160-162). At the same time, the isotopic series provides strong reasons to consider President’s behavior as non-democratic and to associate it with a detested past. Traian Băsescu is finally declared “the last President of the old system”. This creates the possibility to introduce another dichotomy: old/new, and to offer an optimistic prospective view based on the use of the met-aphoric cliché of building a house: (15) “It is better to build a new Romania on a new foundation, on a sound foundation, than to build it on a shaking, rotten and dangerous one, and to find, after several years, that the whole edifice collapses with the Romanian citizens inside.”

The antithesis is another means of indirectly arguing for the President’s suspension from office. The whole utterance carries an exhortative implicature: the MPs are indirectly urged (notice the impersonal construction) to make the right choice. They play the core participative role of decision-making Agents, whereas the citizens are mere Patients (Ilie 2005: 191-198). 4. Final remarks Our case study attempted to bring forward the impact of the discursive perspective adopted by the speaker on the structure, functioning and efficiency of the parliamentary discourse. In a parliamentary debate, an MP can speak for himself – as a rational being –, as a representative of a political group whose mandatory he is, or by virtue of his institutional role. Each of these perspectives has some advantages, but also some drawbacks, depending – among other factors – on the object and the main goals of the debate. At the same time, each perspective defines a specific type of ethos (for a possible typology, see Charaudeau 2005: 85-128).

DISCURSIVE PERSPECTIVE AND ARGUMENTATION IN THE ROMANIAN PARLIAMENTARY DISCOURSE

441

In the case under consideration, the last variant has been chosen. It enables the speaker to resort to an integrating strategy, using an ambiguous we, which has the advantage of apparently maximizing the agreement and minimizing the possibility of disagreement. The inclusive we creates the impression of a merger between the audience and the speaker, who accordingly avoids being perceived as a representative of an opposition group. As an effect, polemic forms of expression directed towards other MPs are excluded. Emphasizing the common ground – especially, the common goals of protecting the citizens and securing their welfare – , the speaker can focus his attacks on a target presented, with sophistic arguments, as the unique obstacle in MPs’ fight to fulfill their goals. The ethos projected by the speaker is that of a reasonable, unbiased person. If the strategies of argumentation could be more efficient in disguising speaker’s true intentions, lexical selections are more transparent for his / her subjectivity, as usually they include words with axiological (positive or negative) meanings, linked in complex chains. To conclude, one can admit that the choices made in a political institutional speech are hierarchically structured and the discursive perspective adopted by the speaker is placed at the highest level, determining all the other choices. References Amossy, Ruth (2006). L’argumentation dans le discours (2e éd.). Paris: Armand Colin. Charaudeau, Patrick (2005). Le discours politique. Les masques du pouvoir. Paris: Vuibert. Dijk, Teun A. van (2004). Text and context of parliamentary debates. In: Bayley, P. (ed.). CrossCultural Perspectives on Parliamentary Discourse. Amsterdam & Philadelphia: John Benjamins, 339-372. Eemeren, Frans H. van & Rob Grootendorst (2004). A Systematic Theory of Argumentation. The pragma-dialectical approach. Cambridge: Cambridge University Press. Goffman, Erving (1959). The Presentation of Self in Everyday Life. Garden City & New York: Doubleday & Co. Ilie, Cornelia (2005). An integrated approach to the analysis of participant roles in totalitarian discourse. The case of Ceausescu’s agent roles. In: de Saussure, L. & P. Schultz (eds.). Manipulation and Ideologies in the Twentieth Century. Amsterdam & Philadelphia: John Benjamins, 191-211. Wales, Kattie (1996). Personal pronouns in present-day English. Cambridge: Cambridge University Press.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 443-461 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

THE DIALECTICAL-DIALOGICAL DEFINITION FABRIZIO MACAGNO

The notion of “dialogical definition” seems to be in sharp conflict with the traditional views on definition. Definition for a long time was regarded as the expression of the immutable essence of the things. This traditional view was that the quest to put forward an acceptable definition requires finding the essential characteristic of the thing to be defined. However, the determination of what is essential has always been highly controversial in philosophy and in science (Walton 2005: 169-171). The conflicts of opinion on such matters tended to persuade philosophers and the general public that there is not such thing as “essence”, or, if there is, it cannot be known (Sager 2000: 217; Walton 2005: 169-173). This position led to the shared view that the distinction between essential and unessential characteristics is invalid (Sager 2000: 216-217). For this reason, the most important recent studies on definition tended to regard it in relativistic terms (see for instance Schiappa 2003). In fact, if the essence cannot be known or does not exist, any definition advanced can be good, as it cannot be verified or falsified. Any discussion about the definition of a thing seemed for this reason to be trivial, or outside the boundaries of scientific interests. The advent of the new approach to argumentation has changed this by seeing definitions as arguments that can be put forward and supported by good reasons or not (see Walton 2005: 179-184; Kienpointner 1992: 259). In argumentation, definitions are regarded as instruments for classifying1 , or rather naming, a fragment of reality. The acceptance of the use of a particular name (or, rather, a predicate) to denote a fragment of reality requires accepting the object’s possession of certain properties presupposed or implied by the name itself. For instance, the same aspect of reality (X) can be labeled as “monopoly” or “strong company”. However, whereas the concept of “monopoly” commonly implies “absence of the free market” or “destruction of the free-market economy”, a “strong company” usually implies that “the company is solid and defeats the concurrence”. Therefore, accepting the attribution of a predicate means accepting other properties that can be used to warrant the acceptance of a conclusion. The uses of the predicates “monopoly” of “strong company” in the propositions “X is a monopoly” or “X is a strong company” elicit arguments leading to different conclusions, for example “X is contemptible” or “X is laudable”. Schiappa (2003: 131) and Zarefsky (2006: 404) call this strategy argument by definition (Zarefsky uses also the name “persuasive definition”) and represent it as a kind of inference having the following structure: x is P (therefore, x is good/bad). 1 For the notion of classification as the attribution of a predicate to a subject, see Mill 1959: 76-77.

444

FABRIZIO MACAGNO

This type of argument can be analyzed by distinguishing the two distinct aspects of the process of attribution of a predicate to a subject, and the reasoning stemming from the predication and leading to an evaluation. The first process can be examined using the ancient concept of locus a definitione, namely the maxims describing the reasoning from definition. Definition is the (often implicit) semantic link between a factual premise, like “X controls the manufacture of ping-pong balls” and a classificatory conclusion, like “X is a ping-pong ball monopoly” (Windes – Hastings 1965: 160). The whole classification is grounded on the implicit definition of “monopoly”. The same aspect of reality (X) can be labeled as “monopoly” or “strong company” according to the definition of “monopoly” the speaker uses. The second part of the reasoning stemming from naming reality can be described using the concept of values and the link between the decision-making and the values. For example, it is commonly shared that “what damages the free-market is bad” and “the winner is laudable”. In this type of reasoning, the argumentation based on shared values and leading to an evaluative conclusion is dependent on the adopted definitions. The choice of a definition can be a powerful argumentative instrument, aimed at altering the evaluation of a state of affairs. Definitions, therefore, can be chosen to support a viewpoint; however, the choice of a definition is not arbitrary in argumentation. By interpreting the ancient concept of dialectical definition developed in Aristotle’s Topics (Giuliani 1972: 130), definitions can be conceived as endoxa, namely commonly accepted opinions. On this view, definitions are matter of commitment, that is, they depend on what is shared between the interlocutors. Definitions can be therefore the standpoint of an argumentative discussion, in which the controversial definition is assessed. The purpose of this paper is to show the argumentative structure of the reasoning from definition, and the criteria for the evaluation of a definition. Different types of definition will be analyzed, showing how different types of definition (for instance essential definition, definition by parts, or definition by etymology) trigger different types of inference (see Victorini Liber de Definitionibus). 1. The argumentative structure of reasoning from definition The most common use of definition is naming a fragment of reality. A being can be called a stone, an animal, or a man on the ground of the definitions of “stone”, “animal” and “man”. However, in argumentation naming often is not a standpoint in itself, but a premise for a consequent conclusion. Classifying a fragment of reality as “a nonhuman being” or as “a man” can be used to support contrary positions, like “It can be killed” or “It must not be killed”. In law, the importance of definition was highlighted by Cicero. He showed how the same action, stealing a sacred vase from a private house, could be punished as theft or as sacrilege according to the possible definitions of the latter term. If “stealing something sacred from a private place” was the definition of “sacrilege”, the action would fall into that category; on the contrary, if “sacrilege” was defined as “stealing something

THE DIALECTICAL-DIALOGICAL DEFINITION

445

from a sacred place”, the defendant would be punished for theft2. The relation between definition and consequences of attributing the defined term to a fragment of reality is the ground for another definitional argumentative move, namely the redefinition. The first uses of this argumentative move are again attested in Cicero. Redefining the concept of “personal enemy” (inimicus), he was able to classify Verres as a public enemy (Ciceronis In Verrem, II, 3, § 6): ‘Fidem sanctissimam in vita qui putat, religionem qui colendam esse existimat, is sine dubio inimicus est ei qui fana spoliavit, omnia templa violavit, delubra polluit’: hic etiam iam Verris idem factum, id de quo specialiter quaeritur, ita continetur et includitur, ut simul cum approbatione et deductione definitionis in speciem clausum teneatur.

Here the concept of inimicus is redefined by Cicero as the enemy of any good citizen. Definitions, at last, can be used as strategies for altering the evaluation of the thing defined. For instance, wisdom can be defined as the “ability of making money”, whereas foolishness “excessive desire of glory” (Ciceronis De Inventione, I, 49): ‘Sapientia est pecuniae quaerendae intelligentia’; […] ‘Stultitia est inmensa gloriae cupiditas’.

These strategies can be analyzed distinguishing between two argumentative processes. The argumentative uses of definition are based on a reasoning aimed at attributing a property to a fragment of reality, and a process of assessment grounded on how the attributed properties are commonly evaluated. 1.1 Reasoning from endoxical definitions Reasoning by definition has been often regarded as an indefeasible type of reasoning. Mill describes it in the following fashion (Mill 1959: 539): Some particular properties of a thing are selected, more or less arbitrarily, to be termed its nature or essence; and when this has been done, these properties are supposed to be invested with a kind of indefeasibleness, to have become paramount to all the other properties of the thing, and incapable of being prevailed over or counteracted by them.

For example, we can consider the following piece of reasoning. A body was defined as “what can move up and down”; in the void a body cannot move up and down; therefore, by definition, in the void a body is not a body anymore (Mill 1959: 539). This argument was used to disprove the existence of void, and was wholly grounded on the “nature” of the bodies. Similarly, by the definition of “man” a Negro was classified as a chattel (Dred Scott V. Sandford, 60 U.S. 393, (1856) at 404-405). In these arguments, 2

“Quare in eiusmodi generibus definienda res erit verbis et breviter describenda, ut, si quis sacrum ex privato subripuerit, utrum fur an sacrilegus sit iudicandus; nam id cum quaeritur, necesse erit definire utrumque, quid sit fur, quid sacrilegus, et sua descriptione ostendere alio nomine illam rem, de qua agitur, appellare oportere atque adversarii dicunt.” (Ciceronis De Inventione: I, xii).

FABRIZIO MACAGNO

446

the conclusion seems to logically proceed from the premises; however, despite the soundness of the reasoning, the conclusion cannot be acceptable. The critical aspect of these arguments lies in what is labeled as “the nature”, or “the” definition of something. In order to understand why a piece of reasoning from definition can be strong, weak, or simply unacceptable, it is useful to inquire into the structure of this type of argument. In argumentation theory, the type of reasoning from a definition to the attribution of the definiendum to a fragment of reality has been described by Hastings (Hastings 1963: 36-52) as Argument from Criteria to Verbal Classification. This pattern of argument has been later elaborated by Walton (Walton 2006: 129), who maintained Hastings’ argument structure: Hastings’ Argument from Criteria to Verbal Classification

Walton’s Argument from Verbal Classification

Event or object X has characteristics A, B, C… If x has characteristics A, B, C… then x is Q

INDIVIDUAL PREMISE: CLASSIFICATION PREMISE:

Therefore, event or object X is Q.

CONCLUSION:

a has property F. For all x, if x has property F, then x can be classified as having property G. a has property G.

Table I: Arguments from classification

Argument schemes from verbal classification seem to broadly represent the nature of the conditional premise “If p then q”. However, the notion of classification is in these patterns unclear. If a predicate is attributed to a subject on the grounds of one or more predicates, it remains unclear why should a predication imply another predication. According to these schemes, the conclusion “a is a cat” can logically follow from premises like “a is a dog” or “a is black”. These argument schemes do not specify the nature of the relation between the predications, and do not explain the reason why a property, or a classification, implies another property or classification. Walton’s argument scheme was further developed in (Walton, Reed & Macagno 2008). In the new formulation of the scheme, this semantic relation was made explicit in the classification premise: CLASSIFICATION PREMISE: For all x, if a fits definition D, then x can be classified as having property G. (Walton, Reed & Macagno, 2008: 68).

In this latter argument scheme, however, the relation between the definition and the thing defined is not clear, as well as the role definition plays in an argument. A better understanding of the reasoning process grounded on definition comes from the an-

THE DIALECTICAL-DIALOGICAL DEFINITION

447

cient loci 3 . In the ancient tradition, the rule of consequence that we represent nowadays as “p, if p then q, therefore q” was applied to the possible semantic relations. The p’s and q’s, in other words, were in the topical system predicates, and the maxims represented the acceptable semantic links, like “what the definition is said of, the definitum is said of as well”. The maxims were an abstraction of the possible reasonable (or semantic) relations that can be used in argumentation. The mechanism of the topics highlights the strict connection between the abstract rule of inference (the maxim), and the common knowledge. We represent the topical inferential structure of the argument: X exclusively controls the manufacture of ping-pong balls. Therefore X is a monopoly.

as below (see Rigotti & Greco 2006; Rigotti 2006; Rigotti 2007): MAXIM

ENDOXON

What the definition is predicated of, also the definiendum is predicated of Exclusive control of the market is the definition of monopoly PRELIMINARY CONCLUSION What “exclusive control of the market” is predicated of, also “monopoly” is predicated of The definiendum is predicated of X X exclusively controls the market of the manufacture of ping-pong balls Therefore the definition is predicated of X X is a monopoly Table II: Reasoning from definition

From this diagram, the relation between the common knowledge, namely the endoxa, and the inferential rule called maxim is made clear. The definition of monopoly as “Exclusive control of the market” is a commonly accepted proposition, on which the acceptability of the conclusion depends. If the definition were different, the plausibility of the conclusion would be different, even though the inferential mechanism and the inferential rule applied were the same. For instance, we consider the following argument: X is the biggest industry in the manufacture of ping-pong balls. Therefore X is a monopoly.

Here the conclusion is grounded on the endoxical premise “Monopoly is excellence in an economic field”. Anyone who knows the laws of a free-market economy would hardly accept this definition of monopoly, and would judge the argument as unreasonable or fallacious. However, the strength of reasoning from definitions depends on 3

The most extensive account of arguments schemes from classification is given in Kienpointner (1992). In Alltagslogik four schemes from definition are identified and the classifications by means of genus, and whole and parts are analyzed (Kienpointner 1992: 250-252).

FABRIZIO MACAGNO

448

what is actually shared by the interlocutor. If the concepts of monopoly, free-market economy, and competition are not clear to the hearer, he can easily accommodate his implicit knowledge to the unshared definition. His lack of knowledge makes it possible for him to accept as shared a definition that would be commonly judged as wrong. The acceptability of a definitional argument depends on what is commonly shared by the interlocutor, and on his knowledge. When definitions are not shared, detecting unacceptable or fallacious definitions becomes much harder. For this reason, the failure in providing a unique definition of some crucial political concepts, such as “terrorism”, “torture”, or (as analyzed in Schiappa 2003) “wetland” can be considered a real manipulative strategy. 1.2 Definitions and values As mentioned above, definitions are instruments of naming reality. However, the naming of reality is often an argumentative move aimed at supporting a standpoint. For instance, as seen above, characterizing an action as “a theft” or as “a sacrilege” implies different consequences. From a legal point of view, the types of punishment provided for the two crime types are different; from a social point of view, a man who committed sacrilege is usually regarded as worse than a simple thief. Naming reality can be an argumentative strategy for influencing the process of decision-making or the evaluation of the denoted fragment of reality. For this reason, definitions are regarded as “political” strategies, namely moves aimed at altering the interlocutor’s or the audience’s choices and assessments (Schiappa 1998: 3; Schiappa 1993: 404). In particular, on Schiappa’s view, defining and naming always express an attitude, orientating the interlocutor towards a certain conclusion. The relation between words and evaluations has been pointed out in argumentation theory in particular by Stevenson, Hare, and Halldén. Stevenson (1937; 1944) inquired into the nature of the ethical terms, that is, words that change the interlocutor’s attitude. Words like “culture” or “blackguard”, on Stevenson’s view, enhance an emotive reaction in the interlocutor, like admiration or contempt. Stevenson analyzed these “ethical” words using the categories of descriptive and emotive meaning. Whereas the descriptive meaning was identified by Stevenson with a cognitive reaction, because it was aimed at altering the knowledge of the world of the interlocutor, the emotive meaning was identified with the change of attitude a word can provoke in the hearer. The emotive meaning was described as the component of a sign that was intended to alter the emotions and the behaviour of the interlocutor. Hare (1952: chap. 7), following Toulmin (1950), examined the nature of the emotive meaning of ethical words, and pointed out how they can lead to action on the basis of shared principles of behaviour. For instance, the argument “Do your homework; do your duty” is grounded on the shared principle that “You should do your duty”. Similarly, the following argument, grounded on the qualities “sweet and juicy”, presupposes an implicit relation between evaluation and action: This strawberry is sweet and juicy. You should eat it.

THE DIALECTICAL-DIALOGICAL DEFINITION

449

In this example we can identify two types of shared premises, namely the evaluation premise (“What is sweet and juicy is good”) and the behaviour premise (“You should eat what is good”). The directive effect of emotive meaning of words can be therefore analyzed in terms of reasoning grounded on implicit norms of conduct. Halldén (1960: 73) focussed his inquiry on the nature of the reasoning from evaluation. He distinguished the descriptive criteria, used to describe the meaning of a word, from the evaluative criteria. These latter criteria can be interpreted as the shared propositions at the basis of the reasoning grounding the evaluation of an aspect of reality. For instance, “love” is positively evaluated in our culture because it is a morally noble feeling, and what is morally noble is considered to be good. Stevenson, Hare, and Halldén highlight two crucial aspects of the argumentative use of a name. A name can be used to influence the interlocutor’s emotions and choices, and this reaction can be analyzed in relation to the shared system of values, or “evaluative criteria”. However, in these authors the relation between evaluation and action is not clear. The argumentative link between assessment and decisions can be found in the ancient tradition, and in particular in Aristotle’s action theory. In the Nicomachean Ethics, Aristotle showed how a decision is always directed to a goal, and the goal can be what is good, or what appears to be good (III, 4, 1113a15); in fact, “everything aims at the good” (Topics III, 1, 116a 18). The emotive meaning can be explained in terms of reasons to act: the agent aims at obtaining what is good for him, and therefore his choices will be determined by what appears to be desirable to him. The agent’s “emotive reaction”, which can be identified with a decision to act, or with a behaviour of praise or contempt, is in this perspective the conclusion of a reasoning based on values, for instance, “Money is good. Therefore I should get it”, or “Violence is bad. Therefore I should despise it”4 . If we apply the Aristotelian view on action and values to the argumentation theory, we can represent the reasoning grounded on emotive words using the argument scheme from values (see Bench-Capon 2003a; 2003b). In argumentation, value can be described as what makes something desirable, and can be conceived as the reason leading somebody to desire something. Values are, in other words, reasons to act (Miano 1952: 657). The process of reasoning from values leads from a characterization of a fragment of reality to a decision about how to act. We can explain the complex argumentation as follows: x (an entity or a state of affairs) is characterized as P (for instance, x is a monopoly); P can be positively or negatively judged 4 Values can be different according to the culture and the people. What is good to somebody can not be good for somebody else. Aristotle tackles this problem distinguishing in the Ethics the absolute good from the relative or apparent good. In his Topics, he develops criteria for the choice between two things that appear to be desirable, like in the topics below (Topics, III 1, 28-34): “That which is desired for itself is more desirable than that which is desired for something else; e.g. health is more desirable than gymnastics: for the former is desired for itself, the latter for something else. Also, that which is desirable in itself is more desirable than what is desirable per accidens; e.g. justice in our friends than justice in our enemies: for the former is desirable in itself, the latter per accidens: for we desire that our enemies should be just per accidens, in order that they may do us no harm”.

FABRIZIO MACAGNO

450

according to a value V (monopoly is bad); according to the desirability of x, x can become a goal (or object of praise or condemn) for the agent. Using Walton and Krabbe’s terminology (see Walton & Krabbe 1995), we can say that the agent commits himself to an action (which can be a simple behaviour of praise or contempt) on the grounds of the desirability of x. Values are therefore the reasons of an action’s desirability, and the grounds of the agent’s commitments. We can represent the abstract pattern of inference as follows (Walton, Reed & Macagno 2008: 321): Argument from Positive Value PREMISE 1:

Value V is positive as judged by agent A (judgment value)

PREMISE 2:

The fact that value V is positive affects the interpretation and therefore the evaluation of goal G of agent A (If value V is good, it supports commitment to goal G). V is a reason for retaining commitment to goal G

CONCLUSION:

Table III: Positive values

Argument from Negative Value PREMISE 1:

Value V is negative as judged by agent A (judgment value)

PREMISE 2:

The fact that value V is negative affects the interpretation and therefore the evaluation of goal G of agent A (If value V is bad, it goes against commitment to goal G). V is a reason for retracting commitment to goal G

CONCLUSION:

Table IV: Negative values

In order to show how reasoning from definitions works, we can apply these argument schemes to the following simple arguments: 1. Pop Cola is a monopoly. It controls the market of soft-drinks. 2. Pop Cola is a monopoly. It is the strongest company in the market of soft drinks.

THE DIALECTICAL-DIALOGICAL DEFINITION

451

We can represent the underlying reasoning as follows:

Figure 1: Structure of the argumentative use of definition

In this scheme the argumentative move (the value judgment expressed on the entity “Pop Cola”) is grounded on the definition of “monopoly”. By redefining a predicate it is possible to change the assessment of a fragment of reality. For instance, in case Pop Cola were the strongest company in the market of soft-drinks, a negative assessment of it could be drawn from redefining “monopoly” as “the strongest company”. Leaving implicit the definitional premise and the evaluative premise, the same conclusion would follow from a different factual premise. An argumentation from values can follow directly from a definition, which can be used to alter our evaluation of its denotation. For instance we can consider the following cases (Naess 1966: 92-93): 1) Democracy is the policy of government that tries to bring

morality and politics closer to one another until they coincide; 2) Democracy is the form of government which gives, or tries to

give, the people the illusion of their own sovereignty.

452

FABRIZIO MACAGNO

In both cases definitions are used to alter the evaluation of the concept of democracy. In the first case, the positive assessment of the definiens affects the evaluation of the definiendum, whereas in the second case the negative value associated to deception alters the evaluation of democracy. This kind of reasoning from definition can be represented in the following fashion:

Figure 2: Altering evaluations through definitions

Arguments from values and classification represent abstract patterns of reasoning showing the argumentative structure of the most important argumentative strategies grounded on definition. However, in order to inquire into how definitions are used to persuade the interlocutor, it is useful to analyze the generic concept of definition in its specific types, namely the different types of definition. 2. Types of definitions and inferences from definition As seen above, the choice of a definition can influence the whole assessment of a fragment of reality. The same concept can be defined in different ways. For example, the concept of monopoly can be defined as “Exclusive control of the market”, or “Exclusive sale” (from the Greek μόνος, alone or single, and πολεῖν, to sell)”, or “Companies like

THE DIALECTICAL-DIALOGICAL DEFINITION

453

Standard Oil, American Telephone and Telegraph, British East India Company…” The structures of these definitions are different: whereas in the first case the predicates of the definition are the genus and the difference of the definitum, in the second case the definition represents its etymology, and in the third case its parts. These definitions are, however, argumentatively different. The fact that the same thing can be defined in different ways does not imply that all definitions are equal from an argumentative point of view. In other words, even though there can be different definitions of the same concept, the reasonableness and the force of the arguments grounded on them is noticeably different. Definitions, in fact, are endoxa, and the strength of the reasoning based on them depends on their acceptability. Moreover, the structure of the definition determines which inferences are possible, and their strength. The first type of definition we will consider is the essential definition, or the definition by genus and difference, for instance, “man is a rational animal” or “monopoly is the exclusive control of the market”. In the ancient tradition, this type of definition was considered to be the most complete (Victorini Liber de Definitionibus, 7, 10-16), because it wholly represented the concept defined, namely the core endoxa relative to the characteristics of the definitum. The essential definition is characterized by the genus, that is, a predicate expressing what a thing is in a generic fashion. For instance, the genus of man is animal; the genus of house is building. The genus represents the generic fundamental (or most shared) properties of the definitum (Stebbing 1933: 429). For instance, it would be unreasonable to say “It is a man, but he is not an animate being”, or “This is a house, but it is not a building”. Moreover, the genus explains the possible attributes which can be predicated of the definitum. For instance, a man moves, swims, eats, breathes…because only animate beings can do it (the predicate “to eat” presupposes the fact that the subject is animate: it would be meaningless to say “This table eats a lot”)5 . The difference distinguishes the thing defined from all the other concepts belonging to the genus expressed, and must justify the attribution of some predicates that can be predicated only of the definitum. For instance, only the man can laugh, talk, or deny, because he is reasonable, and these predicates presuppose a reasonable being (it would be meaningless to say “This cat talks pretty well”). The argumentative strength of essential definition lies in the fact that it represents the deepest level of shared proposition. Essential definitions can be considered a form of semantic analysis, and semantics is the deepest level of endoxa (or shared commitments): if the most basic semantic characteristics are not accepted, a communicative failure can easily occur. The essential definition is argumentatively powerful also because of the inferences it can trigger. For instance, we can examine some of the possible inferences based on the definition of “monopoly” as “exclusive control of the whole market”: 5

It is interesting to notice (see Rigotti 1997) that a man can be sitting or standing, or he can be stretched out, but he must be in a position. Similarly, a stone can be green or grey, but cannot jump. Accident is related to the possibility of predication, to the semantic genera of the predicates, the ten categories.

FABRIZIO MACAGNO

454

What the definition is said of, the definitum is said of as well. What the definitum is said of, the definition is said of as well. What is denied of the genus is denied of the species. What the species is said of, the genus is said of as well. What is said of the species, it is said of the genus as well. What is said of the whole genus is said of one of its species.

Pop Cola exclusively controls the market of soft drinks. Therefore it is a monopoly Pop Cola is a monopoly. Therefore it exclusively controls the market. Free market is not a form of control of the market. Therefore it cannot be a monopoly. Pop Cola is monopolizing the market. Therefore it controls it. Monopolies are bad. Therefore some forms of control are bad. Any form of control is bad. Therefore monopolies are bad.

Table V: Topics of definition

These inferences are all grounded on an essential definition (that is, definition by genus and difference) of “monopoly”. The structure of the definition allows the convertibility of definiens and definiendum, and the relation between species and genus allows the application of the inference rules of the genus (see Aristotle, Topics: book IV). In the definition by parts the definitum is defined by its constituent or essential parts. For instance, a house can be defined as “walls, foundation, and roof ” (definition by constituents), or as “apartment, villa, country-house, terraced house, cottage…” (essential parts). The possible inferences that can be triggered by these definitions are the following: Definitions

Inferences

• There are walls, the foundation, and a roof. Therefore there is By constituent parts: A house is walls, the a house (there might not be). foundation, and a roof. • A house is destroyed. Therefore walls, the foundation, a roof are destroyed (they might not be destroyed). • The walls, the foundation, the roof are solid. Therefore the house is solid (it can be weak anyway). • This house is nice. Therefore its roof is nice (it can be ugly). By essential parts: A monopoly can be a legal monopoly, a natural monopoly, a vertically integrated monopoly…

• This company is a legal monopoly. Therefore it is a monopoly. • This company is not a legal monopoly, nor a natural monopoly, nor…Therefore it is not a monopoly • Pop Cola is a monopoly. Therefore it is either a legal monopoly, or a natural monopoly, or… • Natural monopolies are good. Therefore monopolies can be good

Table VI: Definitions and inferences

THE DIALECTICAL-DIALOGICAL DEFINITION

455

The definition by constituent parts is extremely weak from an argumentative point of view. In fact, it is not convertible with the definitum, because the integral parts and the whole are not equivalent. On the contrary, the definition by enumeration of the species (essential parts) is argumentatively powerful, because it is convertible and it allows the inferences from genus and species. However, from a communicative point of view, the definition by essential parts is useless. This type of definition does not show what the thing is, but the knowledge of the definitum is presupposed by the definition. Moreover, definition by essential parts must list all the possible species of the definitum in order to be acceptable and convertible. A concept can be also defined by looking at the etymology of its significant. For instance, counselor, namely a lawyer who defends cases in court, can be defined by etymology as “he who gives counsels (from consulere)”. Even though the etymology is correct, like in this example, the definition can be wrong or unacceptable. From an argumentative point of view, definitions by etymology are not necessarily convertible, but trigger inferences based on the coniugates, namely the processes of lexical derivation. For instance, we can consider the following cases: He is a counselor, and therefore he should give counsels, and not decide. He is a teacher. He should teach, and not be in politics.

The argument based on definition by etymology is quite used in politics, as Mill (1959: 531) reports: “Perhaps no example of this can be found that is more extensively and mischievously employed than in the case of the word representative: assuming that its right meaning must correspond exactly with the strict and original sense of the verb ‘represent’, the sophist persuades the multitude that a member of the House of Commons is bound to be guided in all points by the opinion of his constituents; and, in short, to be merely their spokesman”. At last, some definitions are not aimed at showing what a thing is, but just at describing the thing defined. For instance, we can define “man” as “the being subject to sin” or as “the being that can pity the Gods”6 . These definitions are used to describe the man through non-essential characteristics, which are useful to support a value judgment on the defined thing. For instance, we can use the first description to support the evaluation of man as an evil being, whereas the second to praise him. These descriptions, we can notice, are both convertible with the definitum, but they do not show what it is. The concepts of “vice” and “pity” presuppose a previous knowledge of what a rational being is. These predicates can only be attributed to a rational being (it would make no sense to say “this mouse is greedy” or “my cat prays a lot”). Definite descriptions are not necessarily convertible with the thing defined. For instance, if we define “man” as “the being that can laugh (or talk)”, we distinguish him from all the entities; however, we can describe him also as “the biped animal” to differentiate him from quadrupeds. 6

Victorinus defines man as “ubi rursus malitia versutia ceteraque vitia versantur” (Victorini Liber de Definitionibus, 18, 19-20) or as “ubi pietas est, ubi aequitas continentia” (ibidem).

FABRIZIO MACAGNO

456

The same concept, to conclude, can be defined in different ways. However, definitions are different in structure and acceptability. Some definitions are stronger than others because they represent the core endoxa relative to a concept, and trigger particular types of inference. In particular, the definition by genus and difference can be interpreted as a type of definition showing the fundamental semantic properties of the concept defined. It can be conceived as an instrument of semantic analysis. 3. The Dialogical Definition Definitions, from an argumentative point of view, are commonly known propositions, or implicit commitments of the interlocutors, that can be more or less shared. The strength of a definition depends on its acceptability and structure. The existence of several definitions does not lead to a definitional relativism, but to conflicts of definitions, that is, to dialogues grounded on different definitions about the same concept. Dialogues on definition can be divided into two broad categories: dialogues on descriptive definitions, and dialogues on normative definitions. Whereas in the first case the dialogue is aimed at establishing what the shared definition is, in the second case the interlocutors are engaged in a discussion on which definition should be adopted. 3.1 Conflicting definitions: Defining “Ambassador” Dialogues on definition show how deep the relation between definition and implicit commitments or shared knowledge (see Walton & Macagno 2007) is. In particular, conflicts of opinion are sometimes based on different categorizations of reality, which in their turn are grounded on conflicting definitions. For example, we can analyze the following case, drawn from Manzoni’s I Promessi Sposi (Ch. 4). The interlocutors, Count Attilio and the Podestà, are discussing about the rightfulness or wrongfulness of a cavalier’s offence to a messenger. A Spanish cavalier sent a challenge to a Milanese cavalier; however, the messenger, not finding him at home, delivered the summons to his brother, who, after reading it, beat him. On Count Attilio’s opinion, the cavalier’s deed was right, whereas on the Podestà’s view the action was mean. The conflict of opinions can be represented as follows (Manzoni 2001: 209-246): Podestà: […] an ambassador is, in his nature, inviolable by the law of nations, jure gentium. But, without seeking so far, the proverb says, Ambasciator non porta pena; and proverbs, you know, contain the wisdom of the human race. Besides, the messenger having uttered nothing in his own name, but only presented the challenge in writing… Count Attilio: But when will you understand that this messenger was an inconsiderate ass, who didn’t know the first?… Count Attilio: […] What puzzles me is why you think so much of the shoulders of a mean scoundrel. Podestà: Who said anything about his shoulders, Signor Count? You would make out I had talked nonsense such as never entered my mind. I

THE DIALECTICAL-DIALOGICAL DEFINITION

457

spoke of his office, not of his shoulders; and am now considering the laws of chivalry. Count Attilio: […] according to the laws of modern chivalry, which are the only right ones, I affirm and maintain that a messenger who dared to place a challenge in the hand of a knight without having asked his permission, is an incautious fool, who may be beaten, and who richly deserves it.

This controversy is based on the classification of the messenger as an ambassador. Whereas the Podestà maintains that he actually was an ambassador, and therefore he couldn’t be beaten, Count Attilio does not acknowledge the status of representative to the messenger, and refuses to admit that the latter could benefit of the diplomatic privileges. The Podestà’s position is grounded on the definition of “ambassador” as “a messenger acting as a representative”: as the messenger presented the challenge in writing, his role in that situation was that of an ambassador. Count Attilio, on the contrary, considers an ambassador a person who knows and follows the rules of chivalry, in addition to act as a representative. The definition of “ambassador”, according to Count Attilio, encompasses not only the role of “acting as a representative”, but also the messenger’s behaviour and personal qualities. We can represent how the difference of opinions is grounded on a conflict of definitions as follows7 :

Figure 3: Conflicting Definitions 7

In this diagram, the premises in white boxes are the explicit premises of the arguments, which are quoted from Manzoni (2001: 209-246). The grey boxes represent the implicit premises.

458

FABRIZIO MACAGNO

In Count Attilio’s view, an ambassador is not a role (to act as), but a person characterized by certain qualities and a particular behaviour. On the contrary, the Podestà considers the ambassador a role, comporting only the requirements of “to be a messenger” and “to be a representative”. If we compare the two definitions, and analyze them as instruments of semantic analysis, we notice that only the Potestà’s definition explains the semantic characteristics of the predicate “to be an ambassador”: TOPICS AMBASSADOR IS A ROLE What the definitum is said This man came as an of, the definition is said of ambassador. This man came as well. as a messenger acting on my behalf. What is said of the This is a good ambassador. definitum is said of the This messenger represents me definition as well. well.

AMBASSADOR IS A PERSON This man came as an ambassador. (?) This man came as an educated person entitled to act on my behalf. This is a good ambassador. This educated person is well entitled to act on my behalf (follows well the rules of chivalry…)

Table VII: Conflict of definitions

The application of the topics of definition to the two definitions highlights how the semantic features of the predicate are better represented in the Podestà’s definition. Defining the definition as an instrument of semantic analysis allows one to compare definitions and choose between them. 3.2 Modifying a definition: Redefining “Virus” Some dialogues on definition, as seen above, can be resolved analyzing the shared meaning of the definienda. These types of conflicts of opinion are based on the interlocutors’ presuppositions. The proponent and the opponent take for granted definitions that in fact are not shared by the other party. These types of conflicts of opinion can be resolved through clarification dialogues, in which topics of definition are used to establish which definition is the best one. A different type of definitional conflict of opinion emerges when a shared definition is countered with a proposed redefinition. In this case the interlocutor advances his proposal of how the concept should be defined. A clear example of a debate on how to define a concept can be found in the discussion on the redefinition of “virus” after the discovery of a new giant virus, the Mimivirus (New Scientist, 25 March 2006: 37-39). In biology, viruses are placed outside the “tree of life”, being microorganisms smaller than bacteria, which cannot grow or reproduce apart from a living cell (New Scientist, 3 September 2005: 26). However in 2002 a virus was discovered that was bacterium-sized, and presented an unusual number of genes, some of which are involved in the first process that makes proteins from genes. The complexity of this virus and its likeness to cellular life was the ground

THE DIALECTICAL-DIALOGICAL DEFINITION

459

for the proposal of redefining the concept of “virus”. On the shared view, Mimivirus would be classified as a virus, as it “is still dependent on its host for basic metabolic processes such as protein synthesis and energy production. And it lacks proteinbuilding ribosomes, one of the fundamental features of life. So it is still a virus. […] And virus is a virus” (New Scientist, 25 March 2006: 41). This definition was countered by some scientists, who maintained that “a term broad enough to include everything from the minuscule hepatitis D to the giant Mimivirus is useless”, and that the same concept of life should be redefined, encompassing the Mimivirus. This virus is commonly considered alive by virologists, and its DNA contains an impressive number of genes. The conflict of definitions can be represented as follows:

Figure 4: Dialogue on definition

The arrows between the argumentations supporting the two conflicting arguments represent the conflict of definitions. The interlocutor uses an argument from consequences, namely a pattern of reasoning leading from the premise “A’s consequences are bad” to the conclusion “A should be avoided (not be chosen…)”. The traditional definition of life is shared among the scientists, and the interlocutor cannot deny the shared commitments. However, he advances a pragmatic argument to support his redefinition of “life” and “virus” in virology. He argues that by redefining these concepts a better explanation of the evolution of life could be possible. This debate shows how deeply connected definitions are with theoretical issues such as the origin of life, which constitute the cultural background of a community. Redefinitions, as shown in the example of the definition of “ambassador”, are often implicit and are used to manipulate. If undetected, the redefinition can lead the interlocutor to fallacious conclusions. However, when detected, redefinitions can be

460

FABRIZIO MACAGNO

countered and proved to be wrong by confronting them with the implicit shared commitments. Redefinitions, however, are not always manipulative strategies. They can be supported by arguments aimed at attacking or countering the presuppositions the shared definitions are grounded on. In these cases, redefinitions, and dialogues on definition, become dialogues on the endoxa of a community. 4. Conclusions Definitions from an argumentative point of view can be analyzed as endoxa, namely commonly accepted propositions. Definitional endoxa represent the deepest commitments of a community, as they are the basis of the mutual understanding. Definitions become argumentative strategies when they are used as premises in patterns of inference that are commonly called “argument from classification” or “loci a definitione”. In these types of reasoning, the definition is used to name reality, that is, to classify it. Different classifications of the same fragment of reality can lead to noticeably different conclusions, or influence the interlocutor’s process of decision-making. For this reason, conflicts of classifications are often based on conflicting definitions. This type of conflict can be resolved only by establishing first what a definition is, and what the criteria are to assess a definition. In the ancient tradition, the best definition was the definition that could license the strongest inferences, and it coincided with a semantic analysis of the definiendum. The essential definition can be therefore considered from an argumentative point of view the criterion to assess dialogues on definitions. References Aristotle (1968). Topics. Translated by W. A. Pickard-Cambridge. Oxford: Oxford University Press. Aristotle (1985). Nicomachean Ethics. Translated by Terence Irwin. Indianapolis, Ind.: Hackett. Bench-Capon, Trevor (2003a). Persuasion in Practical Argument Using Value-based Argumentation Frameworks. Journal of Logic and Computation 13: 429-448. Bench-Capon, Trevor (2003b). Agreeing to Differ: Modelling Persuasive Dialogue between Parties without a Consensus about Values. Informal Logic 22: 231-245. Cicerone (1998). De Inventione. Introduzione, traduzione e note a cura di M. Greco. Galatina: Congedo. Cicerone (2004). Orationes Verrinae. Tradotto da L. Fiocchi & N. Marinane. Milano: Rizzoli. Giuliani, Alessandro (1972). Aristotelian Theory of Dialectical Definition. Philosophy and Rhetoric 5: 129-142. Halldén, Soren. (1960). True Love, True Humour and True Religion: A Semantic Study. Lund: Gleerlup. Hastings, Arthur (1963). A Reformulation of the Modes of Reasoning in Argumentation. Evanston, Illinois: Ph.D. Dissertation, Northwestern University.

THE DIALECTICAL-DIALOGICAL DEFINITION

461

Kienpointner, Manfred (1992). Alltagslogik: Struktur und Funktion von Argumentationsmustern. Stuttgart: Fromman-Holzboog Manzoni, Alessandro (2001). I Promessi Sposi. Vol. XXI. The Harvard Classics. New York: P.F. Collier & Son, 1909-14; Bartleby.com, 2001. Miano, Vincenzo (1952). Dizionario Filosofico. Torino: SocietĂ Editrice Internazionale. Mill, John Stuart (1959). A System of Logic. London: Longman. Naess, Arne (1966). Communication and Argument. London: Allen & Unwin LTD. Rigotti, Eddo & Sara Greco Morasso (2006). Topics: the Argument Generator. In: Rigotti, E. et al., Argumentation in Financial Communication, Argumentum eLearning module. http:// www.argumentum.ch/. Rigotti, Eddo (2007). Can Classical Topics be revived within the Contemporary Theory of Argumentation? In: van Eemeren et al. (eds.), Proceedings of the Sixth Conference of the International Society for the Study of Argumentation. Amsterdam: Sic Sat, 1155-1163. Rigotti, Eddo (2007). Relevance of Context Bound Loci to Topical Potential in the Argumentation Stage. Argumentation 22: 519-540. Sager, Juan (2000). Essays on Definition. Amsterdam: John Benjamins Publishing Company. Schiappa, Edward (1993). Arguing about Definitions. Argumentation 7: 403-417. Schiappa, Edward (1998). Constructing Reality Through Definitions: The Politics of Meaning. A lecture presented for the Center for Interdisciplinary Studies of Writing and the Composition, Literacy, and Rhetorical Studies Minor. Speakers Series 11. Schiappa, Edward (2003). Defining Reality. Definitions and the Politics of Meaning. Carbondale and Edwardsville: Southern Illinois University Press. Stebbing, Susan (1933). A Modern Introduction to Logic. New York: The Humanities Press. Stevenson, Charles (1937). The Emotive Meaning of Ethical Terms. Mind 46: 14-31. Stevenson, Charles (1944). Ethics and Language. New Haven: Yale University Press. Victorinus (1997). Liber de Definitionibus. Mit Einleitung, Ă&#x153;bersetzung und Kommentar von A. Pronay. Frankfurt: Peter Lang. Walton, Douglas & Fabrizio Macagno (2005). Common Knowledge in Argumentation. Studies in Communication Sciences 5/2: 1-22. Walton, Douglas, Chris Reed & Fabrizio Macagno (2008). Argumentation Schemes. New York: Cambridge University Press. Walton, Douglas (2006). Fundamentals of Critical Argumentation. New York: Cambridge University Press. Windes, Russel & Arthur Hastings (1965). Argumentation and Advocacy. New York: Random House. Zarefsky, David (2006). Strategic Maneuvering through Persuasive Definitions: Implications for Dialectic and Rhetoric. Argumentation 20: 399-416.

Legal cases cited Dred Scott V. Sandford, 60 U.S. 393 (1856)

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 463-474 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

THE CONJUNCTION ‘AND’: LITERAL MEANING AND DISCOURSE VALUES

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

The present paper faces two basic challenges and aims at overcoming them while showing how to deal with one word, the conjunction ‘and’. What are the challenges? On the one hand, the claim that natural languages are imperfect, ambiguous, vague tools, so that to avoid misunderstandings we must regiment them by translating them into formalized languages. On the other hand, the opposite claim that natural languages are so rich, nuancées, full of ad hoc values in discourse, that no pretense of objective and rigorous treatment of the meaning of their constituents can be put forward. Let us see what happens in the case of ‘and’. The conjunction ‘and’ is formalized in first order logic by the connective ∧. The interpretation of such a connective is a function from truth values to other truth values: ∧ takes as arguments the truth values of the sentences it joins and results in the truth value of the whole sentence. The customary truth tables describe the functions attached to the logical connectives: p

p∧q

The formalization of the natural language word ‘and’ with the logical connective ∧ raises at least two problems: a) From the syntactic point of view, ‘and’ joins not only sentences but almost every part of speech and every phrase: nouns (B03 781 ‘She has worked extensively at Ronnie Scotts in London and Birmingham’, B29 1584 ‘The club holds regular social functions including wine and cheese parties and receptions for visiting companies which give members an opportunity to meet the actors and actresses’1 ), NPs (B1H 2090 ‘It was a fortress town and a trading centre, as it had been for the Romans 800 years earlier’), VPs (A7A 1292 1

The examples are taken from the British National Corpus (http://www.natcorp.ox.ac.uk/), which gathers 100 million of tokens in written and spoken English from a wide range of sources. The first three characters of the code denote the text from which the sentence is taken, the following numerals indicate the sentence number within that text.

464

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

‘Bodo put out a large hand and pulled the door to’), adverbs (ECD 508 ‘While market forces may undoubtedly bring many positive benefits, both socially and politically, there are many occasions when it is legitimate to subject them to regulation in order to secure confidence and fairness in the operation of the financial markets’), adjectives (KIR 3789 ‘On the second day of the Macari trial at Winchester Crown Court, Swindon players past and present have been giving evidence’) and even prepositions (ABU 1764 ‘But a range of people inside and outside the House of Commons, deliver’) and conjunctions (AD0 1270 ‘If your weight has gone up and down over the last year (or ten years!) try to remember when and why it was up or down’). Our aim in this paper is to show that: (i) the occurrences of ‘and’ that do not conjunct sentences can be rendered by ∧ only at the cost of distorting the syntactic structures of the sentences; (ii) some occurrences of ‘and’ cannot be formalized by ∧. So, ∧ is not a good formalization of ‘and’. b) From the pragmatic point of view, ‘and’ seems to take other meanings in addition to the one expressed by ∧: e.g. temporal order (FSE 676 ‘He stepped away from the command console and bowed’), cause-effect (4KT 4859 ‘Lawrence was unhappy over the outspoken challenge to his authority and axed Slaven from his Tranmere squad’), opposition to and disappointment with an expectation (KC3 235 He is ‘unemployed and receiving state benefit as employed’). Besides, the conjunction of NPs seems sometimes to indicate that the predicate has to be separately assigned to both conjuncts (CH8 1428 ‘Now my mum and dad come to just about every concert’), sometimes to indicate that the it has to be collectively assigned to the conjuncts (CAR 1718 ‘Friends and well-wishers gather at Kirkwall Airport to await the return of the children to Orkney’), sometimes it is ambiguous between these two readings (A0N 1958 ‘Alex McLaggan and Mary have their own roof at Grandtully’, which can mean either that Alex McLaggan has a roof at Grandtully and Mary has a different roof in the same place or that Alex and Mary have a roof together at Grandtully). We’ll show that ‘and’ has not so many meanings (it is not semantically ambiguous), and that the further meanings it seems to have (the distinction distributive/collective included) are indeed pragmatic enrichments due to inferences drawn from the text. Finally we shall sketch a positive theory of the meaning of ‘and’. 1. is not a good formalization of every occurrence of ‘and’ The formalization of (1) in first order logic is (2): (1) KDM 5350 Paul and Cathy are going (2) Go(Paul) ∧ Go(Mary)

THE CONJUNCTION ‘AND’: LITERAL MEANING AND DISCOURSE VALUES

465

So a conjunction of proper names is turned into a conjunction of sentences. (1) is considered as equivalent to: (3)

Paul is going and Mary is going

Indeed some syntacticians claimed that (3) is the deep structure of (1) and that (1) is the outcome of a syntactic transformation (cf. for example Gleitman 1965). While (1) and (3) have the same truth conditions, a conjunction of VPs often cannot be transformed into a conjunction of sentences. This happens when the NP which is the argument of that conjunction is indefinite: (4) FT9 151 Former England opener Geoff Boycott chased and caught a thief (5) Former England opener Geoff Boycott chased a thief and Former England opener Geoff Boycott caught a thief

(4) means that Geoff Boycott chased a thief and caught him while (5) could mean that he chased a thief and caught a different one. We face the same problem when a conjunction of adjectives is involved: (6) AYM 1288 You’ll always receive a warm and friendly welcome (7) You’ll always receive a warm welcome and you’ll always receive a friendly welcome These facts seem to prove that there exists no deep structure in which two different sentences occur. However, first order logic possesses the formal resources to express (4) and (6), although a price must be paid: the syntactic structure of natural language sentences is completely distorted. Indefinite NPs are expressed by the existential quantifier, which can take large scope and bound every open variable in the whole formula. (4) can be formalized by (4’) and (6) by (6’): (4’) x (Thief(x) ∧ Chased(Geoff Boycott, x) ∧ Caught(Geoff Boycott, x)) (6’) Always ( x (Warm(x) ∧ Friendly(x) ∧ Welcome(x) ∧ Receive(you, x)))

From the syntactic point of view, in (4) firstly the verbs are conjoined and then this conjunction takes its arguments. We give a very rough and simplified representation of the syntactic structure of (4):

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

466

Geoff Boycott

a thief chased

caught

Such a structure is not mirrored in (4’), where the noun ‘thief ’ and the verbs ‘chased’ and ‘caught’ are on the same level and their variables are all bounded by the quantifier expressed by ‘a’. The same holds in (6): firstly the adjectives are conjoined, then they modify the noun ‘welcome’ and finally the verb takes as argument the resulting construction: S

you

always

receive

warm

friendly

welcome

THE CONJUNCTION ‘AND’: LITERAL MEANING AND DISCOURSE VALUES

467

This is lost in (6’) where the verb, the noun and the adjectives are on the same level and their variables are bounded by the quantifiers expressed by ‘a’. However, when conjunctions of nouns are considered instead of conjunctions of adjectives or verbs, not only do the first order logic formalizations not mirror the syntactic structure of the sentences, they also give incorrect results. For example: (8) There were ten friends and colleagues of Paul’s at the party (8’)

10x (Colleague-of-Paul(x) ∧ Friend-of-Paul(x) ∧ To-be-at-the-party(x))

(9) There were ten friends of Paul’s at the party and there were ten colleagues of Paul’s at the party

Clearly (8) is not equivalent to (9); furthermore (8’) not only distorts the syntactic structure of (8)3 , but its formalization (8’) is not suitable to express (8) either: (8’) is true when ten people who are both friends and colleagues of Paul attended the party. But (8) can be true also when five friends (who are not colleagues) and five colleagues (who are not friends) attended the party; or it can be true when there were two people who are both friends and colleagues of Paul, two friends who are not colleagues and six colleagues who are not friends at the party. These possibilities are excluded by (8’). Things are even worse with regard to (10): (10) CJN 35 116 young men and women, between the ages of 17 and 22, took part in the research between January and December 1992 (10’)

116 x (Young-man-between17-22(x) ∧ Young-woman-between17-22(x) ∧ Tookpart-in-the-research(x))

The formula (10’) is true when 116 people who are both men and women took part in the research. Since no people who are both men and women do exist, (10’) cannot be true in any context; but clearly this is not true with regard to (10). In the same way, cannot formalize conjunctions of NPs that are arguments of a collective verb: (11) JT4 152 The vocalist and the guitarist met in a chip shop (11’) Met-in-a-chip-shop(the vocalist) ∧ Met-in-a-chip-shop(the guitarist)

(11’) is true when the vocalist met in a chip shop and the same thing is true of the guitarist. Indeed it is hard even to figure out what a sentence like that could mean. Finally, as ∧ is a function from truth values to other truth values, it is thought of as a formal device to render the conjunction of constative sentences to which a truth value can be assigned: so ∧ cannot formalize performative sentences which have no truth value.

2 As usually, ‘ nx P(x)’ is the abbreviation of: x1… xn (x1≠x2 ∧… xn-1≠xn ∧ P(x1) ∧…P(xn)) 3 In the syntactic tree firstly the nouns are conjoined, then they are modified by the PP ‘of Paul’s’

and then determined by the determiner ‘ten’; finally the DP built in this way becomes an argument of the verb. All this structure is clearly lost in (8’).

468

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

To sum up, in this first section, we have shown that ∧ cannot formalize every use of ‘and’ in natural language. In addition to ∧, the symbol of the intersection between sets , has also been used in order to formalize the conjunction ‘and’, for example by Partee & Rooths (1983). We could demonstrate that this formalization runs into problems similar to those in the case of ∧, but we will do not so for lack of space. 2. Discursive meanings As mentioned above, ‘and’ can carry many different meanings according to the discursive contexts where it is used. There are at least two possible positions regarding this: (a) firstly, ‘and’ is an ambiguous word with many possible meanings, one of which is selected by the context of use; (b) secondly, ‘and’ carries only one, very poor, meaning, while all the other meanings which seem to be possible are, in fact, the results of pragmatic implicatures bound to specific discursive contexts. In what follows, we defend the second position given above, starting with the argument about the alleged collective and distributive meanings. 2.1 The distinction collective/distributive is not a semantic one There are many different reasons in favour of a pragmatic solution. First of all, according to Winter (1996), if the meaning of ‘and’ were ambiguous between a collective and a distributive one, it should not be hard to find out languages where these two meanings are expressed through two different words. For instance, the Italian word ‘riso’ is ambiguous, since its meaning can be “laugh” (noun) or “rice”: in other languages, the two meanings of ‘riso’ are expressed through two different words, such as ‘Lachen’ and ‘Reis’ in German, or ‘rire’ and ‘riz’ in French. The same should also hold true for ‘and’: but this is not the case. In fact, as Payne shows in his comparative study (Payne 1985), there is no language which expresses the distributive and the collective meanings of ‘and’ through two different words. The fact that even very different languages systematically convey these two meanings using always just one word is a strong argument in favour of the conclusion that these two meanings do not really exist and that ‘and’ has just one meaning. Secondly, reasons of economy are involved. Claiming that the conjunction ‘and’ is ambiguous implies claiming that the ambiguity of ‘and’ is not among a limited number of meanings, but among an open, indefinite, and hardly manageable, set of meanings. As a demonstration of this, we come back to the distinction distributive/collective: some scholars, like Partee & Roth (1983), or Hoeksema (1988), agree that ‘and’ has two different meanings, namely a distributive and a collective one. But, let us consider the following sentence: (12) Students and professors wrote a manifesto

THE CONJUNCTION ‘AND’: LITERAL MEANING AND DISCOURSE VALUES

469

This sentence has at least three possible readings: (a) a collective one: all the students and professors have taken part in the writing of one common manifesto; (b) a distributive one: each student and each professor has written one single manifesto; (c) an intermediate one: all the students together have written one manifesto, while all the professors together have written another one. Claiming that ‘and’ carries an ambiguous meaning implies that the ambiguity of ‘and’ is at least in three senses, since there are at least three possible readings of ‘and’ (collective, distributive, intermediate). But there is more to it than this. Let us consider the following sentence (taken from Gillon 1987): (13) Rodgers, Hammerstein and Hart wrote musicals

Gillon makes it clear that Rodgers collaborated both with Hammerstein and Hart in writing musicals, while Hammerstein and Hart never worked together. Among the many readings of (13), the right one cannot be the distributive, since Rodgers, Hammerstein and Hart never wrote a song on their own, but always through some kind of collaboration; but the same holds also for the collective reading, since all the three songwriters never worked together. The predicate “writing musicals” neither has to be individually (i.e. distributively) attributed to Rodgers, Hammerstein, or Hart, nor collectively to all of them, but it should be attributed to two overlapping subsets of the set formed by the three songwriters. Thus, a further reading must be added to the distributive, the collective and the intermediate ones: in this fourth reading, the predicate is neither attributed to the single elements nor to the full set of the referents, nor to some partition of it, but to some overlapping subsets of it (technically, to a cover of it). The process of recognizing further readings does not stop here. For instance, arguing about an intermediate reading between the collective and the distributive ones, we should consider more than just one intermediate reading. In fact, a set can be subdivided in many different ways and the predicate can be attributed to the cells resulting from these different subdivisions. This implies an explosion of the number of possible readings which becomes difficult to manage since ‘and’ would be a word with really many different possible meanings. A final reason against the ambiguity of ‘and’ is that the collective/distributive difference holds when using plurals without the presence of ‘and’: (14) The four hundred firemen have put out fifteen fires

Two of the possible meanings of (14) are that the four hundred firemen have together put out the fifteen fires, or that everyone of the four hundred firemen have put out fifteen fires (which is pragmatically an improbable reading, even though semantically still possible). This is the same difference in reading we have noted for NPs which are joined through ‘and’. Claiming that ‘and’ plays a role in making the distinction collective/distributive leaves no way to explain all the different readings of sentences such as (14). Or, maybe, a different explanation must be taken into account: but giving

470

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

two different explanations for what seems to be just one phenomenon is just uneconomic. 2.2 Temporal, causal and opposition meanings are not semantic ones Although we have argued that the meaning of ‘and’ is not ambiguous between a collective and a distributive sense, it could still remain true that ‘and’ has other different meanings, such as temporal, causal, etc. According to the theory of the ambiguity of ‘and’, this word shows many possible meanings within the linguistic system and each specific context of use selects one of them. There are many arguments against such a theory. For instance, given that one of the possible meanings of ‘and’ is temporal order, this meaning would be selected by contexts such as the following: (15) He woke up and brushed his teeth Now removing ‘and’ from (15), the result is: (16) He woke up. He brushed his teeth

The meaning of temporal order of the two events seems not to be lost. But, if in (15) the meaning of temporal order is attained through ‘and’, why does this meaning still remain, when ‘and’ is removed? The Maxim of Manner by Grice requires the speaker to be orderly, i.e. expressing the contents of his talk with regard to their temporal order. Thus, unless further clarifications by the speaker or the context make the inference false, the addressee’s inference from (15) and (16) is that the speaker’s purpose is that the agent has performed the actions in the same (temporal) order of the clauses in the sentence. Another argument in favour of considering the meaning of temporal order as a pragmatic inference is the possibility of removing the implicature, without resulting in a contradiction: (17) He woke up and brushed his teeth, but not in this order

(17) is a non contradictory sentence. It would be such, if the temporal order was included in the semantic meaning of ‘and’: that things happened in a specific temporal order would be, in fact, stated and, later on, denied. The above remarks about translation are still worthwhile: if ‘and’ had two meanings (one including and one not including the temporal order), there could be languages where these two meanings are expressed through two different words. But this is not the case. The same holds good also for the alleged causal meaning of ‘and’: (18) He turned the key and the car started

Again, the causal relation between the two events is still expressed even if ‘and’ is removed: (19) He turned the key. The car started

THE CONJUNCTION ‘AND’: LITERAL MEANING AND DISCOURSE VALUES

471

Our claim here is that there is nothing in (18) and (19) which semantically conveys that the key-turning has caused the starting of the car. A possible reading of (18) and (19) is that the car starting has not been caused by the key-turning, but by something else4 . But such a reading is easily rejected, since key-turning usually causes the cars to start and speakers know it thanks to their encyclopaedic knowledge. Now, (18) and (19) report that the key has been turned and the car started: there is no reason to argue that the key-turning has not been the cause of the car’s starting. The speaker in (18) and in (19) does not say anything to deny this natural hypothesis: he knows that the addressee will guess this and does not act in any way to prevent it. The addressee has the right to make this inference and the speaker knows that the addressee has got this right. So, if the speaker is cooperative, he has used (18) and (19) in order to make the addressee understand that the key-turning has been the cause of the car’s starting. But that this causal relation between the two events exists is just a pragmatic inference derived from the context and our encyclopaedic knowledge about how cars work; it is not semantically conveyed by ‘and’. Levinson (2000), reforming Grice’s thought, suggests the following conversational maxim (here reported in a shortened form): The I[nformativeness]-Principle. Speaker’s maxim: “say as little as necessary”, i.e. produce minimal linguistic information sufficient to achieve your communicational ends Recipient’s corollary: amplify the informational content of the speaker’s utterance, by finding the most specific information, up to what you judge to be the speaker’s intended point. Specifically: i. assume that stereotypical relations obtain between referents or events, unless: (1) that is inconsistent with what is taken for granted; (2) the speaker has broken the maxim by choosing a prolix expression; ii. assume the existence or actuality of what a sentence is “about”; iii. avoid interpretations that multiply the entities referred to (assume referential parsimony). If the [I] principle by Levinson is indeed a maxim followed by speakers in their productive and interpretative strategies, this explains why from (18) and (19) the implicature that there is a causal relation between the two involved events can be easily drawn. In particular, the maxim requires that the addressee assumes the stereotypical relations among the events referred to by the speaker, unless there are contrary specifications. The addressee has to enlarge the informational content of the sentence, assuming 4

There are, in fact, contexts where the implicature is removed: “He turned the key and the car started. But he had not started the car: the battery cables had been unplugged. The car has been started through a remote control just in the moment when he had turned the key”. All this is not contradictory, as it would be if a semantic meaning of ‘and’ were here involved.

472

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

that these relations are verified although the speaker has not explicitly said so. Assuming that cars stereotypically start through key-turning, this is the most natural interpretation of (18) and (19). Another supposed meaning of ‘and’ would be to express the opposition between two conjuncts: in this case, the meaning of ‘and’ would be very close to that of ‘but’: (20) It is a beautiful day. Everyone is enjoying it outdoors and Paul is studying indoors (21) It was a very interesting proposal and you have rejected it In (20) and (21), the first conjunct establishes an expectation which is denied by the second. Here we have to wonder if the opposition between the two conjuncts is embedded in the meaning of ‘and’, or if it is inferred from the fact that what is expressed by the two conjuncts is itself in opposition (although not contradictory). An argument in favour of this second hypothesis is that it is possible to use ‘and’ instead of ‘but’ only if a clear and explicit opposition between the two conjuncts actually exists. When the opposition between the two conjuncts can be unclear to the addressee and must be explicitly expressed by the speaker, the use of ‘but’ is mandatory: (22) She is beautiful and clever (23) She is beautiful but clever That an opposition really exists between beauty and cleverness is a disputable matter: if the speaker claims that the two properties are in some way in opposition to each other, he must be explicit on this, since the opposition cannot be inferred merely from the meaning of the two conjuncts. Thus, he has to use ‘but’, as shown in (23); while in (22) the use of ‘and’ makes the opposition disappear. The alleged opposition meaning of ‘and’ would exist only in those cases where the opposition between the two conjuncts is clear: so, in the light of this, it is more economic to claim that the role of ‘and’ is just to join the two clauses, while the opposition between them is due to their own meanings and not to the meaning of ‘and’. Again, the opposition between the two clauses still holds good even without ‘and’: this shows furthermore that the opposition is due not to ‘and’, but to the meanings of the clauses themselves: (24) It is a beautiful day. Everyone is enjoying it outdoors. Paul is studying indoors (25) It was a very interesting proposal. You have rejected it To highlight the disappointment of what we would have expected, the final clauses of (24) and (25) are often uttered in a descending intonation and a lower tone. 3. Towards a minimalist theory of the meaning of ‘and’ We have excluded that the conjunction ‘and’ might be formalised with ∧; we have also excluded that it might have a rich and complex meaning or that it might be ambiguous

THE CONJUNCTION ‘AND’: LITERAL MEANING AND DISCOURSE VALUES

473

among various meanings. What has to be done then is to find a positive meaning of ‘and’, which does not coincide with the usual formalizations, but is minimal, i.e. which can be the semantic basis of all its syntactic occurrences and of all its uses. Our claim is quite simple: the function of ‘and’ is that of conjoining the semantic value of the second conjunct to the semantic value of the first conjunct in order to produce a structure which has the semantic values of both. In other words, taking two elements X and Y of the same syntactic kind, and assuming that the semantic value of X is α and that of Y is β, the structure ‘X and Y’ will have as its semantic value the result of the conjunction of β and α. The claim has obviously to be specified according to the elements joined by ‘and’, that is to say, according to the syntactic kind of X and Y: if X and Y are two declarative sentences, then ‘X and Y’ will be a structure such that both X and Y are asserted. If on the contrary the two sentences are directives, then the conjunction ‘and’ will establish a structure where both the orders are given. Otherwise if X and Y are two directive sentences, the conjunction will produce a structure where both the orders are given. If X and Y are predicates, then ‘X and Y’ will predicate both X and Y of the subject of the sentence; if X and Y are NPs, and if therefore their function is to refer to objects, then ‘X and Y’ will have as its semantic value the result of the sum of the references of Y to those of X, so that such a structure will refer to the objects to which X refers as well as to those to which Y refers. If X and Y are adjectives and if therefore their function is to ascribe properties to the objects to which a noun refers, then ‘X and Y’ will ascribe both the properties expressed by X and Y to the objects to which the noun refers. If X and Y are nouns and if their function is to refer to objects5 , once more ‘X and Y’ will refer to the objects to which X refers as well as to those to which Y refers. Eventually, if X and Y are adverbs and therefore their task is to modify and to specify an event, the state or the action expressed by the verb, or the property expressed by an adjective, then ‘X and Y’ will attribute to the verb or to the adjective both the modifications. We don’t think that a further meaning has to be attributed to the conjunction ‘and’. Obviously what is here outlined is just the draft of a theory: technical details depend on the adopted formalism. In any case we affirm that, whatever the formal theory adopted, if it aims to express the meaning of ‘and’ in natural languages, it has to equip itself with resources apt to add the referents of a noun to those of another one, to add the attribution of a property to that of another one, to add the modification of a verb to that of another one, etc. ∧ and can play this role with at least some of the syntactic types which ‘and’ joins, but not with all of them: they are more apt when it is a matter of adding the attribution of an adjective to that of another adjective, or the predication of a verb to that of another verb, or the modification of an adverb to that of another adverb, but in many cases they fail, when NPs and nouns are taken into account and what becomes necessary is to add referents to refer5 If the function of both nouns and noun phrases is to refer to objects, what is their semantic difference? We think that nouns have only a potential reference which is actualized or filtered by (possible) determinants of NPs. Yet, nothing in our claim about the meaning of ‘and’ depends on this particular theory.

474

ALDO FRIGERIO & MARCO PASSAROTTI & SAVINA RAYNAUD

ents. Two choices are possible, then: either to express the conjunction with different formal resources, using for instance for predicates, adjectives or adverbs and another sign for NPs and nouns, or to look for an alternative and more univocal formalization. In this second case, it would be a question of radically reforming our present formalizations. We do not think there can be anything to add to the semantic meaning of ‘and’. Our thesis is therefore a very weak one. Its strength is in stating that any stronger thesis either loses in generality or it attributes to the semantics of ‘and’ what is just a pragmatic inference. References Gillon, Brendan (1987). The Readings of Plural Noun Phrases in English. Linguistics and Philosophy 10: 199-219. Gleitman, Lila (1965). Coordinating Conjunctions in English. Language 41: 260-93. Hoeksema, Jacob (1988). The Semantics of non-Boolean and. Journal of Semantics 6: 19-40. Levinson, Stephen (2000). Presumptive Meaning: The Theory of Generalized Conversational Implicature. Cambridge (Mass.): The MIT Press. Partee, Barbara & Mats Rooth (1983). Generalized Conjunction and Type Ambiguity. In: Buerle, R., C. Schwarze & A. von Stechow (eds.). Meaning, Use and Interpretation of Language. Berlin: De Gruyter, 361-383. Payne, John (1985). Complex Phrases and Complex Sentences. In: Shopen, T. (ed.). Language Typology and Syntactic Description. Vol. 2: Complex Constructions. Cambridge: Cambridge University Press, 3-41. Winter, Yoad (1996). A Unified Semantic Treatment of Singular NP Coordination. Linguistics and Philosophy 19: 337-391.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 475-488 SPECIAL ISSUE: WORD MEANING IN ARGUMENTATIVE DIALOGUE

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

LOUIS DE SAUSSURE & STEVE OSWALD

Introduction La notion d’engagement du locuteur et la question de son utilité pour l’étude de l’argumentation concerne différents niveaux d’analyse, notamment l’identification des forces illocutoires et celle des propositions qui peuvent légitimement être tenues pour assumées, ou vraies, par l’un des participants d’une conversation. En d’autres termes, la question de l’engagement – commitment – concerne à la fois la force illocutoire (sur quel type d’acte de langage le locuteur s’engage-t-il – et sur quoi d’autre doit-il s’engager pour pouvoir s’engager sur un acte de langage déterminé –?) et le contenu lui-même. Nous nous concentrerons ici sur le deuxième aspect, en abordant la question de savoir à quelles conditions un destinataire peut tenir un locuteur pour responsable de, ou considérer que ce dernier s’engage sur, l’assertion d’un ensemble de propositions déterminées, dans un contexte donné. L’objectif de cet article est de proposer une approche psychologique de la notion d’engagement du locuteur vis-à-vis de l’énoncé qu’il produit, en suggérant l’hypothèse que plus un contenu est pertinent dans les circonstances, plus il est plausible que le destinataire considère que le locuteur s’engage sur ce contenu. Généralement, la notion d’engagement du locuteur fait référence au système de croyances du locuteur: ainsi, proférant un contenu vériconditionnel (une ou plusieurs propositions), le locuteur signalerait qu’il adhère à, ou s’engage sur, la vérité des faits couverts par cette (ou ces) proposition(s). Par sa nature, une assertion associe au contenu vériconditionnel une valeur de vérité positive: les conditions de vérité sont réputées vraies, et, ainsi, les propositions sont comprises comme décrivant des faits entretenus comme vrais par le locuteur. Pourtant, traiter des «croyances» entretenues comme vraies par un locuteur ne peut reposer que sur l’identification des contenus que le locuteur s’engage à communiquer par son énoncé1 . Or pour le destinataire, déterminer les contenus effectivement intentionnés par le locuteur, sur la base des formes linguistiques en contexte, n’est pas une opération transparente. Elle mobilise au contraire des processus sophistiqués qui impliquent une marge d’incertitude ou de risque. L’un des procédés d’argumentation fallacieuse qui montre de la manière la plus évidente l’importance d’identifier des contenus dont la communication est effectivement voulue par le locuteur est celui de l’homme de paille (strawman fallacy) qui con1

D’autre part les croyances individuelles sont privées et inscrutables. Seuls les indices de ces croyances sont publics et scrutables. Voir également à ce sujet Papafragou (2006: 1698).

476

LOUIS DE SAUSSURE & STEVE OSWALD

siste à attribuer à autrui une position, généralement facilement réfutable, sur laquelle, en fait, il ne s’est pas engagé. Or si une telle fallacie peut éventuellement fonctionner et convaincre, c’est probablement, pensons-nous, parce que les croyances sur lesquelles le locuteur s’engage ne constituent pas des faits objectivement scrutables, et que la plupart des formes linguistiques sont intrinsèquement ambiguës. Les théories de l’argumentation convergent pour considérer que l’une des clés de l’analyse des mouvements argumentatifs réside dans l’observation des considérations ou propositions qui peuvent raisonnablement (notre soulignement) être considérées (par un observateur externe) comme assumées par le locuteur, comme le pose Hamblin (1970). Hamblin considère l’existence de «stocks d’engagements» (commitment stores) produits par un locuteur au cours d’une argumentation, sans la connaissance desquels l’interlocuteur – ou l’analyste – ne peut identifier un sophisme comme celui de l’homme de paille. Ce ne sont pas tant les croyances d’un individu locuteur qui comptent, mais, dirons-nous, ce qui est tenu pour avoir été dit. Or il faut bien que quelqu’un tienne ce qui est tenu, et ce quelqu’un ne peut être que le destinataire, réel ou idéal. Ce que le destinataire tient pour avoir été dit correspond donc au stock des prises de position que Hamblin nomme stocks d’engagements. La définition que propose Walton (1993: 93), selon laquelle l’engagement est «une inférence tirée sur la base de ce que vous dites et comment vous agissez lorsque vous interagissez avec un autre participant (notre traduction)» exprime après tout la même chose: les engagements sont des inférences. Toutefois, plutôt que de placer au niveau du locuteur (comme le fait encore Walton ici), nous insistons sur le fait que ces inférences sont tirées par un destinataire et sont destinées à être tirées par un destinataire. Ainsi, toute verbalisation poursuit l’objectif de donner lieu à des inférences prévues par le locuteur, qui connaît intuitivement les mécanismes pragmatiques qui président à l’inférence, sur la base de sa psychologie intuitive ou théorie de l’esprit. Pour que le destinataire émette l’hypothèse que le locuteur s’engage sur le fait de communiquer P (et par conséquent, éventuellement, de croire P), il faut qu’il ait à sa disposition des éléments qui le conduisent à considérer que le locuteur s’engage, par son énoncé, à rendre manifeste le fait qu’il entretient la pensée P. La question qui surgit donc, et qui nous amènera vers la distinction entre les contenus explicites et implicites, est celle de l’identification des engagements du locuteur par le destinataire. Quant à savoir quelles sont les bases sur lesquelles un contenu peut être attribué à un locuteur en tant qu’engagement, la proposition de Hamblin (1970) reste évasive: les engagements, explique-t-il en substance, sont des faits observables de l’extérieur (par l’analyste, par exemple), qui peuvent être normativement évalués par un regard attentif sur ce qui est dit (what is said, dans la tradition gricéenne) et ce qui est fait (what is done, dans la tradition des actes de langage) par le locuteur considéré. La proposition de Katriel & Dascal (1989) est plus sophistiquée: les engagements sont, pour eux, constitués par ce que le locuteur peut être tenu pour avoir pris pour acquis en produisant son énoncé. Mais, c’est la position que nous voulons défendre dans cet article,

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

477

l’établissement de ce qui peut être tenu comme engagement d’un locuteur par un interlocuteur est un processus complexe, tout simplement parce qu’évaluer ce qui est dit par un locuteur requiert un niveau de contextualisation des énoncés, contrairement à la simplification opérée par Grice dans sa dichotomie dit – implicité. Cela est requis non seulement pour la saturation référentielle, mais aussi pour la désambiguïsation et différentes sortes d’enrichissements pragmatiques qui concernent bel et bien le niveau explicite du sens, c’est-à-dire celui qu’évoque Hamblin quand il parle de ce qui est dit, et qui, nous le répétons, n’est pas transparent. Qui plus est, la recherche actuelle insiste pour considérer qu’un locuteur peut être lié non pas uniquement par ce qu’il a dit, mais également par ce qu’il a communiqué implicitement, au moins à certaines conditions (Burton-Roberts 2006, Capone à paraître, Morency, Oswald & Saussure à paraître). Quant à ce qui est fait, nous considérons qu’il s’agit-là d’un problème encore supplémentaire, qui ne peut être considéré qu’en ayant déjà résolu la question de savoir ce que le locuteur peut être tenu pour avoir dit. Ces opérations de découverte des engagements du locuteur requièrent donc un modèle de la compréhension; nous proposons ici quelques pistes, en exploitant l’approche de Sperber & Wilson (1995), pour les considérer. 1. Engagement et types de contenus Depuis Grice, identifier les contenus pragmatiques implicites repose sur deux critères essentiels (nous laisserons le troisième, à savoir la détachabilité, de côté dans cet article). Le premier critère est la non-redondance de la verbalisation des contenus implicites. En effet, verbaliser le contenu implicite produit par un énoncé ajoute du sens à cet énoncé (contrairement aux implications). Ainsi, on ne peut considérer que l’information implicite serait contenue dans l’énoncé lui-même; dès lors, on est tenté de considérer qu’un locuteur ne peut communiquer qu’il s’est engagé sur son adhésion à ce contenu. La même conclusion vaut également, et de manière encore plus frappante, avec l’autre critère, à savoir celui de la défaisabilité: un contenu implicite peut toujours être défait, logiquement parlant, typiquement par une structure introduite par mais (ou d’autres marques de révision), sans produire de contradiction. De ce fait, la question de l’engagement du locuteur est au centre de la question théorique de l’interface entre sémantique et pragmatique: si on ne peut se rétracter sans produire une contradiction, alors on est engagé par ses paroles. Parmi les problèmes soulevés par cette conception se trouve le suivant: beaucoup de contenus que Grice disait implicites, et qui sont toujours considérés comme tels par de nombreux chercheurs, relèvent en fait beaucoup plus probablement du contenu explicite, tant la rétractabilité sur de tels contenus semble pratiquement impossible. Ainsi, considérer que l’énoncé Anne a quatre enfants ne communique exactement quatre que par implicature (tirée par la maxime de quantité) est très difficile à tenir, si l’on admet une notion plus intuitive de la rétractabilité, c’est-à-dire de l’engagement du locuteur. On imagine très mal un locuteur ayant produit cet énoncé déclarer ultérieu-

478

LOUIS DE SAUSSURE & STEVE OSWALD

rement qu’il voulait communiquer qu’Anne a quatre enfants ou davantage (dans un contexte neutre). A cet égard, l’introduction, par Sperber & Wilson (1986 / 1995) notamment, d’autres critères, comme le fait qu’un contenu développe ou non la forme logique de l’énoncé, en vue de statuer sur son caractère implicite, a été salutaire. En effet, au seul critère de l’engagement du locuteur, Anne a quatre enfants communique bien explicitement qu’Anne a exactement quatre enfants, et il n’est pas plausible que le locuteur ni l’interlocuteur puissent faire une autre hypothèse, hormis dans des contextes très inhabituels. Ceci ouvre une piste d’investigation supplémentaire: même les contenus explicites, produits par un énoncé, et sur lesquels un locuteur s’engage, sont obtenus de manière inférentielle. De la sorte, les contenus explicites dépendent également du contexte (cf. la notion d’explicature dans la théorie de la pertinence de Sperber & Wilson, et les notions apparentées dans d’autres traditions qui traitent de ce contenu pragmatique primaire qui est formé par les contenus explicites, comme la notion d’impliciture chez Kent Bach (1994)). La pragmatique s’accorde aujourd’hui à les nommer contenus pragmatiques primaires (primary pragmatic meaning), bien qu’elle diverge quant à leur nature. L’hypothèse que nous défendons est que la meilleure manière de discriminer ces types de contenus est de considérer le point de vue du destinataire. Il y a donc maintenant deux problèmes interreliés à considérer. Premièrement, la fiabilité ‘objective’ de la communication – et donc de l’évaluation d’un engagement du locuteur – dépend d’une tripartition, qui sépare trois types d’information en fonction de son degré de publicité (overtness): le contenu effectivement verbalisé, le contenu pragmatique primaire et le contenu implicité. Par contenu pragmatique primaire, nous entendons la catégorie que la théorie de la pertinence de Sperber & Wilson (1995) nomme les explicatures: ce sont des développements de la forme logique sous-jacente à la forme effectivement verbalisée. Ces explicatures produisent différents contenus, mais surtout, c’est ce qui nous intéresse ici, la forme propositionnelle, que le destinataire identifie comme ce que le locuteur communique explicitement (et qui auparavant était traité comme le dit de Grice, ou le contenu vériconditionnel de la sémantique, qui ne s’envisageait que par deux opérations simples, la saturation et la désambiguïsation). Il serait tentant de bâtir sur l’approche classique et de l’étendre avec une notion de gradation: le locuteur sera considéré comme s’engageant fortement sur des contenus verbalisés, moins fortement sur des explicatures, et faiblement (ou pas du tout) sur les contenus implicités. Toutefois, cette hypothèse est trop forte. Tout d’abord, parce que le contenu verbalisé n’est pas nécessairement propositionnel, et la notion d’engagement n’a de raison d’être qu’au sujet de contenus d’allure propositionnelle. Ainsi, un énoncé comme La pizzeria n’est pas une proposition, mais une verbalisation elliptique à laquelle des composants doivent être rajoutés pour déterminer si la forme propositionnelle explicite ressemble plutôt à Regarde la pizzeria s’il te plaît ou à La pizzeria conviendra pour le déjeuner, par exemple. Ainsi, le locuteur ne peut pas s’engager à propos d’un contenu verbalisé s’il n’est pas propositionnel (sauf sur le fait qu’il l’a prononcé); s’il s’engage sur quelque chose, ce sera sur le fait que le reste de la communication est

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

479

inférable à partir de la forme effectivement verbalisée. On objectera peut-être que cet exemple devrait être éliminé des observables, car typique de l’oral, la proposition sousjacente étant réputée saillante dans le contexte. Mais il s’agit là du cas le plus standard des stimuli linguistiques réels, et la saillance des informations contextuelles n’est pas radicalement différente selon qu’il faille ajouter des constituants inarticulés (Perry 1986, Carston 2004, Récanati 2002) pour une forme explicite ou qu’il faille recourir à une prémisse contextuelle pour tirer une implicature, puisque certaines implicatures sont très évidentes et que certaines explicatures sont au contraire peu évidentes. La forme propositionnelle, qui constitue ce qui sera évalué comme forme développée explicite, repose non seulement sur la saturation et la désambiguïsation2 , mais aussi sur la récupération de constituants inarticulés et la détermination conceptuelle. Ces deux derniers mécanismes supposent pour nous des opérations cognitives orientées vers la recherche de pertinence dans le cadre de la reconstruction d’un contenu explicite propositionnel ‘de premier degré’ que la tradition récente nomme l’enrichissement pragmatique (Carston 2004, Récanati 2002 par exemple)3 . Restent donc comme contenus susceptibles de produire une hypothèse sur l’engagement du locuteur la forme propositionnelle contextuellement dérivée et qui tiendra comme contenu explicite, c’est-à-dire réputé ouvertement communiqué par le locuteur, et les implicatures, qui sont d’autres propositions (certaines implicatures au sens gricéen se trouvant ramenées à des explicatures quand elles développent la forme logique, comme dans Anne a quatre enfants). Tout le problème réside donc en ceci que contrairement à ce qu’on suppose dans la tradition gricéenne, la plupart des contenus réputés explicites sont construits contextuellement sur la base d’indices incertains, tout comme les implicatures; ils ne sont donc pas donnés a priori. Quant aux implicatures, elles peuvent être tellement saillantes qu’il en devient implausible qu’elles n’engagent pas le locuteur, dans un sens intuitif qu’il importe de traiter. Qui plus est encore, plusieurs chercheurs admettent, bien qu’avec timidité parfois, que la perspective de donner un sens à un énoncé par une implicature détermine la sélection d’un développement propositionnel explicite adéquat; en d’autres termes, il arrive que la perspective d’un contenu implicite très pertinent détermine le choix d’un enrichissement particulier pour fournir le bon matériau explicite à partir d’un matériau verbalisé informationnellement incomplet. Il en va ainsi dans les exemples de causalité inverse implicite traités dans Saussure (2003 et 2005), où la référence temporelle des propositions – explicite – dépend d’une implicature causale (voir encore Carston (2004) sur la nécessité d’un principe de «mutual adjustement» entre explicatures et implicatures qui pourrait devoir, d’ailleurs, être généralisé à tout traitement cognitif de l’information linguistique).

Cf. Carston (2004) pour une argumentation contre la thèse codique d’une indexicalité généralisée des expressions ambiguës et de la saturation. 3 Cet enrichissement, pour Carston (2004), est soit contraint par la sémantique (c’est le cas de l’ellipse déterminée par un constituant linguistiquement requis mais manquant) soit opéré sur des bases purement pragmatiques (c’est l’enrichissement libre).

480

LOUIS DE SAUSSURE & STEVE OSWALD

2. Engagement, sens littéral et sens explicite Pour ce qui concerne l’ajout de précisions ou de constituants non requis par le code, c'est-à-dire pour des raisons de pertinence, il est crucial de noter que le locuteur a proféré dans les cas évoqués ci-dessus une proposition qui pourtant ne sera pas retenue comme la proposition explicitement voulue par le locuteur. C’est là que la distinction entre le contenu littéral, qui correspond à la proposition encodée, et le contenu explicite, qui sera le fondement d’une hypothèse sur l’engagement du locuteur, se trouvent disjoints. Ainsi, ce sur quoi le locuteur peut être considéré comme engagé n’est pas la proposition qu’il a proférée, mais une proposition plus complexe ou plus précise. Considérons les exemples suivants, avec leurs commentaires. (1)

Ça prendra du temps pour guérir ces blessures.

Ici, l’interprétation produit, sauf contexte marqué, quelque chose comme (1’): (1’) Ça prendra un temps considérable pour guérir ces blessures.

Il semble évident pour l’intuition que le destinataire ne fera pas l’hypothèse que le locuteur s’engage sur (1), très sous-informatif (puisque tout prend «du temps»), mais bien sur (1’), informatif: la pertinence de mentionner le temps se trouve atteinte par une majoration de l’importance ou de la quantité de temps en jeu. Pour se convaincre qu’il ne s’agit pas là de processus purement conventionnels, comme on peut être tenté de le croire, il suffit de rappeler que les mêmes processus sont en jeu dans des situations dans lesquelles la désambiguïsation joue au sujet de spécifications possibles du contenu conceptuel, lesquels sont purement contextuels et non conventionnels. Ainsi de (2), interprété comme (2’) ou comme (2’’) selon le contexte: (2) La pizzeria est ouverte. (2’) La pizzeria est ouverte à cette heure-ci. (2’’) La pizzeria est ouverte désormais (p.ex. après une fermeture pour travaux).

En (2’), le locuteur s’engage sur le fait qu’au moment de la parole, la pizzeria est ouverte, ce qui autorise d’autres implicatures, par exemple concernant la possibilité d’aller y déjeuner. En (2’’), le locuteur s’engage sur un état permanent d’ouverture de la pizzeria, mais qui n’implique aucunement que la pizzeria soit ouverte aux clients au moment de la parole. Les cas de discours rapporté présentent une difficulté supplémentaire en ce qui concerne l’engagement du locuteur, à cause de la remontée des informations attribuées à un locuteur source vers ce qui peut être attribué au locuteur lui-même (cf. à ce sujet Morency, Oswald & Saussure à paraître). Proférant (3), le locuteur peut être considéré comme s’engageant ou non sur (3’) en fonction d’indications contextuelles multiples et indécidables a priori: (3) Paul dit qu’il est temps de partir. (3’) Il est temps de partir.

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

481

Enfin, les cas d’ironie et de métaphore demanderaient d’autres développements pour être traités, mais on peut déjà remarquer que le locuteur de (4) dans une version ironique s’engage sur à la fois (4’) et (4’’), de même pour respectivement (5), (5’) et (5’’), et, pour la métaphore, il reste mutuellement manifeste que (6) est littéralement faux et doit être compris comme engageant le locuteur sur un contenu différent, par exemple que Paul est organisé à l’extrême (exemples (4) et (5) adaptés de Sperber & Wilson). (4) Il pleut quelques gouttes. (4’) Il pleut très fort. (4’’) Celui qui pense qu’il ne pleut que quelques doutes est ridicule.

(5) Beau temps pour un pique-nique ! (5’) Le temps est mauvais et inapproprié pour un pique-nique. (5’) Celui qui pensait qu’il ferait beau est ridicule.

(6) Paul est un classeur. La même analyse, finalement, tiendrait pour des questions plus complexes: ce que Grice appelle le dit se ramène parfois au contenu littéral, et certains de ses implicites ne sont que des explicites. Ainsi, dans Anne a quatre enfants, on peut admettre, conformément à l’intuition, qu’en contexte neutre le locuteur s’engage sur le nombre exact d’enfants (contenu explicite) alors que le contenu littéral, non enrichi et généralement sous-informatif, est Anne a quatre enfants ou plus. D’autres cas mériteraient encore de retenir notre attention. Dans un célèbre article de 1971, Geis & Zwicky signalent que l’interprétation la plus naturelle de (7) est (7’) alors que l’interprétation littérale attendue devrait être (7’’), si l’expression si du langage naturel correspond, comme ils s’y attendent, à l’implication logique: (7) Si tu tonds la pelouse, je te donnerai dix dollars. (7’) Si et seulement si tu tonds la pelouse, je te donnerai dix dollars. (7’’) Si tu tonds la pelouse, ou dans une autre situation possible, je te donnerai dix dollars.

A tout le moins, l’intuition insiste pour que le destinataire attribue (7’) comme engagement du locuteur de (7), et non (7’’), malgré le caractère plus littéralement conforme de (7’’). En réalité, il est bien loin d’être clair que le si de la langue correspond sémantiquement à une implication plutôt qu’à une biconditionnelle. La spécification de si vers l’une ou l’autre de ces valeurs semble très dépendante de facteurs pragmatiques; (8) reçoit ainsi soit l’une soit l’autre lecture, selon que l’une ou l’autre est plus propice à satisfaire la demande du destinataire en informativité, puisque (8) n’implique pas nécessairement que le locuteur communique que Marie ne sera pas contente si on ne va pas à la mer (mais à la montagne par exemple): (8) Si nous allons en vacances à la mer, Marie sera contente.

482

LOUIS DE SAUSSURE & STEVE OSWALD

Dans tous ces cas, bien entendu, la difficulté consiste à établir ce qui tiendra pour communication explicite, ouvertement admise et prévisible par les deux parties comme engageant le locuteur, à partir d’une forme littérale soit elliptique soit sous-déterminée. Comme cet engagement ne s’attribue que i) par le recours au contexte et ii) comme le fruit d’un processus d’attribution de pertinence aux énoncés produits4 , l’attribution d’un engagement à un locuteur est de nature contextuelle, pragmatique, automatique, et intuitive. A ce titre, le fait qu’un locuteur s’engage sur un contenu n’est pas prévisible sur la seule base du code. Ainsi, l’engagement n’est pas un fait objectif, c’est un fait éminemment psychologique et intersubjectif. Ce qui n’implique pas, nous insistons làdessus, que l’analyste ne soit pas capable, par ses compétences de sujet parlant et la connaissance qu’il peut avoir au sujet du contexte, de poser une hypothèse fiable sur ces commitment stores qui assurent la discrimination entre des positions argumentatives ouvertes à la discussion, et d’autres qui y sont fermées puisque le locuteur ne s’engage pas à leur sujet. Pourtant, qu’elles soient ouvertes ou fermées dépend d’un degré de fiabilité de l’hypothèse au sujet de l’engagement du locuteur. Il faut donc un instrument de mesure de cette fiabilité. Notre hypothèse est que cet instrument de fiabilité s’applique non seulement aux contenus explicites mais aussi aux contenus implicites. 3. Engagement et implicatures L’une des manières traditionnelles de distinguer les contenus explicites et implicites est de considérer que les explicatures sont des contenus à propos desquels le locuteur ne peut pas se rétracter sans produire une contradiction. Cela correspond à l’idée que les contenus explicites ou explicatures engagent davantage le locuteur que les implicatures. Toutefois, il ne s’agit là que d’une intuition, une conséquence, qui provient d’un principe plus fondamental: pour qu’une implicature soit considérée comme plausible par le destinataire, il est en effet nécessaire qu’elle s’accorde avec au moins une explicature préexistante logiquement (mais pas nécessairement temporellement, cf. Saussure (2005)). En cela, une implicature est une conséquence dépendant du contexte (puisqu’elle requiert l’importation d’une prémisse non dite) d’une explicature. L’important est que si une implicature se perçoit comme engageant le locuteur, alors il existe une explicature, qui la précède logiquement, qui est en tant que telle perçue comme engageant le locuteur également, tandis que l’inverse n’est pas vrai. Toutefois, et c’est le deuxième point qui nous semble crucial, comme dans un tel cas l’implicature constitue l’objectif de la communication, et qu’elle est considérée par le destinataire comme prévisible par le locuteur (sans quoi le destinataire ne la dériverait pas), alors il est plus important pour le destinataire de considérer l’engagement du locuteur sur l’implicature, l’explicature n’étant que secondaire, en tant que simple instrument de 4

Rappelons que pour Sperber & Wilson, tout énoncé communique la présomption de sa pertinence optimale, laquelle consiste en un équilibre entre les effets informatifs dérivables lors de sa compréhension, et l’effort impliqué par la reconstruction de ces effets informatifs.

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

483

déclenchement du travail inférentiel. Cependant, et c’est là une nuance nécessaire, attribuer un engagement du locuteur à une implicature est une hypothèse plus risquée pour le destinataire que de l’attribuer à une explicature. Ce fait a donné lieu à la position classique, mais par erreur, car il n’implique pas que nous n’attribuions jamais d’engagement à un locuteur à propos d’une implicature. Il suffit pour s’en convaincre d’observer le simple fait qu’une rétractation sur certaines implicatures est interprétée comme relevant de la mauvaise foi la plus criante, et son auteur, s’il peut échapper à la preuve de son inconsistance, suscitera néanmoins la conviction de son insincérité. Il nous faut donc admettre que le critère de l’engagement reste celui du coût de la rétractabilité: inconsistance logique, ou inconsistance pragmatique, ainsi que nous souhaitons la nommer, celle-là même qui déclenche la conviction de la mauvaise foi. A nouveau, prenons un exemple. Prenons l’énoncé (9) dans le contexte où il est accessible au destinataire que le locuteur, son père, a des raisons de se plaindre des dépenses engendrées par son mariage: (9) Il y a beaucoup de monde à ton mariage. L’implicature (9’) est alors anticipable: (9’) Il y a trop de monde à ton mariage.

Il s’agit bien d’une implicature, puisque (9’) transforme et ne développe pas l’explicature (9) (qui se trouve en l’occurrence identique à la forme littérale). Or, dans le cas où le contexte favorisant une plainte au sujet des dépenses est très accessible, le destinataire ne peut que supposer que le locuteur s’attend à ce que (9’) soit tirée, et il est ainsi amené à considérer que son père s’engage sur la plainte. Que cette plainte soit rétractable formellement, puisque située en dehors d’un développement de la forme logique verbalisée, n’implique pas que le locuteur ne s’est pas engagé sur cette implicature. Ce sur quoi il ne s’est pas engagé, c’est sur sa volonté de confrontation directe: procéder avec un énoncé comme (9) pour entamer une discussion sur (9’) permet, comme avec les actes de langage indirects et de nombreuses implicatures, de sonder l’acceptation par le destinataire d’un topique conversationnel qui pourrait menacer sa face. Si le fils se fâche et répond à son père qu’il ne devrait pas se plaindre pour des questions d’argent, le père pourra s’autoriser à se rétracter, mais cette rétractation n’en sera pas moins perçue comme étant de mauvaise foi. Le fait qu’elle soit ainsi perçue provient, suggérons-nous, du fait que l’explicature sans l’implicature qu’elle a déclenchée se trouve très pauvre sur le plan de la pertinence, se réduisant à une exclamation sur un fait plutôt banal, à savoir qu’il y a beaucoup de monde à une fête de mariage. Au niveau des implicites, nous assistons ici à l’émergence d’une inconsistance pragmatique: le destinataire n’a pas de raisons de penser que le locuteur de (9) puisse entretenir (9) mais pas (9’) dans les circonstances mutuellement manifestes, et surtout, il n’a pas de raisons de penser que, proférant (9) dans ces circonstances, il n’allait pas rendre (9’) manifeste au destinataire. Ainsi, quelque chose comme [(9) et non-(9’)] est

484

LOUIS DE SAUSSURE & STEVE OSWALD

pragmatiquement, bien que non logiquement, inconsistant, car (9) sans (9’) est dépourvu de pertinence et l’hypothèse que le locuteur s’engage sur (9) sans s’engager sur (9’) est implausible. Le sentiment de mauvaise foi réside sur ce type d’inconsistances. Toutefois, l’hypothèse sur cet engagement est néanmoins pourvue d’un risque interprétatif, car il existe un risque d’erreur sur la sélection des prémisses, et (9) peut donner lieu à d’autres implicatures dans d’autres contextes, par exemple le contentement face à un grand nombre d’invités. Il semble inutile de multiplier les exemples. Une réponse comme j’ai mangé en réponse à une invitation à dîner, sans que le locuteur annule explicitement l’implicature du refus de l’invitation (basée sur l’explicature le locuteur a mangé récemment), assortie de l’hypothèse très fortement entretenue que le locuteur a conscience dans les circonstances de la dérivation de cette implicature, amène inévitablement le destinataire à considérer que le locuteur s’engage sur le refus de l’invitation, pourtant implicité. Bien sûr il pourrait s’agir d’un cas de communication non coopérative: le locuteur pourrait faire de l’humour, mentir, être ironique; mais ces cas de figure jouent précisément sur le fait que le destinataire attribuera de manière prévisible un engagement du locuteur sur l’implicature. Nous devons encore mentionner deux types de problèmes. Tout d’abord, il n’est pas toujours évident pour l’analyste de déterminer si les contenus qu’il observe sont du ressort de l’explicature (auquel cas leur dénégation entraînerait une inconsistance logique) ou de l’implicature (auquel cas leur dénégation entraînerait une inconsistance pragmatique, pour autant qu’elle ait un fort degré de saillance dans le contexte considéré). Les critères de développement de la forme logique ou d’indépendance propositionnelle de l’implicature ne sont pas systématiquement fiables, et c’est sans doute le plus gros problème de l’interface sémantique-pragmatique. Mais il existe des critères psychologiques expérimentaux qui montrent que certains contenus pragmatiques ne sont dérivables qu’à partir d’un certain seuil de développement; ce point est certainement le meilleur critère pour distinguer les contenus implicites des contenus explicites mais déjà pragmatiques (ceux dont la forme propositionnelle diffère du contenu verbalisé même lorsque celui-ci a une forme littérale). Nous pouvons les appeler implicatures de haut niveau: l’ironie ou la métaphore en font partie, mais aussi certaines formes de discours rapporté ou de pensée attribuée comme le style indirect libre (sans verbe introducteur). Les travaux de Noveck (2001) ou de Breheny, Katzos & Williams (2005), entre autres, le montrent (bien qu’ils produisent des explications sensiblement différentes des processus en jeu). On sait par exemple (Noveck 2001) qu’il faut qu’un enfant ait atteint l’âge de 6 à 8 ans pour dériver pas tous à partir du quantificateur quelques, alors qu’il semblerait a priori complètement automatique de dériver quelques étudiants (ou un étudiant) ne sont pas venus à la soirée à partir de quelques étudiants sont venus à la soirée. Mais une fois la compétence pragmatique acquise, et c’est cela qui nous importe, le contenu pas tous – qui sera donc une implicature – sera, en circonstances normales, une dérivation tellement attendue à partir de quelques qu’il est prédictible que le destinataire considérera que le locuteur de quelques s’engage sur pas tous.

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

485

4. Perspectives: l’analyste, le destinataire et le critère de pertinence Le critère de non-rétractabilité ne peut être réduit à un critère uniquement logique mais doit être considéré du point de vue psychologique également. Pour ce qui concerne la rétractation sur des contenus explicites, elle entraîne une inconsistance logique, mais même ceci devrait, idéalement, être nuancé, puisqu’il existe des explicatures très dépendantes du contexte. Même les cas où une rétractation est contradictoire de manière logiquement dure, à savoir ceux où le contenu littéral est équivalent à la forme explicite, ne peuvent être considérés qu’en faisant abstraction de la saturation référentielle et d’éventuelles ambiguïtés fortes. Pour ce qui est de la rétractabilité sur les implicatures, réputées depuis Grice défaisables, donc rétractables sans contradiction logique, l’inconsistance sera pragmatique, pour autant qu’elle porte sur une implicature saillante dans les circonstances. L’inconsistance pragmatique procède d’une évaluation: le locuteur opérant une telle rétractation sera considéré comme non-coopératif, ou de mauvaise foi, à moins qu’il ne reconnaisse explicitement qu’il a produit une erreur de formulation, de distraction, etc. Comme nous défendons un point de vue psychologique – cognitif – sur l’attribution des engagements du locuteur, nous rejetons le point de vue qui voudrait que l’analyste puisse statuer sur ces stocks d’engagement sans recours à une dimension psychologique. Ainsi, il nous semble qu’une machine purement sémantique ne serait pas même capable d’attribuer des engagements avec fiabilité aux contenus dont la forme littérale équivaut à la forme propositionnelle explicite: il est en effet impératif que le traitement de l’information soit capable d’évaluer si cette forme littérale est susceptible d’atteindre les objectifs d’informativité requis dans un échange donné, selon un contexte spécifiable (sinon, que faire de l’ironie, des métaphores, et des innombrables cas de tautologie enrichie?). Or l’analyste n’est pas une machine sémantique, c’est un organisme à compétence pragmatique. Faisant œuvre de cette compétence – qui consiste à opérer les contextualisations nécessaires à la dérivation d’un sens dont il est plausible qu’il corresponde à l’intention informative du locuteur, l’analyste agit comme un destinataire générique. La formation d’une hypothèse sur l’intention informative du locuteur est réglée, selon Sperber & Wilson (1995), par un principe d’équilibre entre le caractère informatif de l’information obtenue et l’effort mental requis pour l’obtenir (principe de pertinence), l’échelle d’accessibilité de telles hypothèses étant à son tour guidée par une aptitude mentale humaine métareprésentationnelle (Sperber 2000). Cette aptitude nous permet de nous représenter les hypothèses et processus mentaux accomplis par d’autres, et repose sur la psychologie intuitive (théorie de l’esprit, empathie ou mindreading module selon les traditions). Cette aptitude permet à l’analyste de se mettre à la place du destinataire pour dériver les différents engagements raisonnablement, ou naturellement, attribuables au locuteur. L’analyste est un destinataire qui peut être handicapé lorsque l’échange implique des contextes ad hoc, connus et partagés par une partie des seuls interlocuteurs. Mais à l’inverse, l’analyste dispose du temps nécessaire à une évaluation offline, et d’un savoir spécifique. Ainsi, son évaluation ne concernera pas des

486

LOUIS DE SAUSSURE & STEVE OSWALD

stocks d’engagement objectifs mais psychologiques. Nous voudrions maintenant formuler une hypothèse centrale en suggérant que plus un contenu est pertinent, plus il est raisonnable d’attribuer ce contenu comme engageant le locuteur. Un contenu optimalement pertinent a plus de chances d’être celui intentionné par le locuteur et donc est plus susceptible d’être considéré par le destinataire comme une position engageant le locuteur, qu’un contenu moins pertinent. Une telle attribution d’engagement repose également, bien entendu, sur le caractère plausible5 de l’engagement du locuteur sur le contenu considéré. Nous suggérons, ainsi: Plus une information P obtenue sur la base de l’énoncé U est optimalement pertinente, et plus le destinataire a de raisons de penser que le locuteur pouvait s’attendre à ce que P soit dérivée de U, et plus l’attribution par le destinataire d’un engagement sur P au locuteur est motivée.

Il reste à dire un mot des implicatures douteuses, celles pour lesquelles on ne dispose pas de suffisamment d’éléments pour décider si elles sont intentionnées ou non: elles sont inférables, mais elles ne satisfont pas aux exigences que nous avons évoquées, puisque le destinataire n’est pas en mesure d’assurer que le locuteur s’attend à ce qu’elles soient dérivées. Ce cas n’est pas, à notre avis, le cas naturel de l’interaction communicative, puisque le locuteur ne produit pas là de garanties formelles suffisantes pour assurer la communication de ces informations dans le contexte. Des menaces voilées, des incitations subtiles, peuvent ainsi se produire, mais elles ne peuvent que laisser place au doute et donc exclure l’engagement du locuteur. Ainsi, seules les implicatures claires, du type de celles qui impliquent une perte significative de pertinence si elles ne sont pas tirées, tout en étant plausibles, entreront parmi celles qui donnent lieu à l’attribution d’un engagement du locuteur. Il y a d’autres cas épineux, qui concernent i) l’enchâssement d’engagements dans le discours rapporté, et ii) les expressifs; nous avons abordé spécifiquement ces questions ailleurs (Morency, Oswald & Saussure à paraître). Nous laissons ici de côté le cas des implicatures dites «faiblement communiquées» (Sperber & Wilson 1995) car elles ne sont pas comptabilisées comme intentionnées par le locuteur. De cet article, il ressort aussi que l’identification d’arguments sophistiques comme celui de l’homme de paille, ainsi que l’explication des mécanismes qui président à son existence même, doit pouvoir reposer sur la prise en compte des procédures pragmatiques qui conduisent à la genèse, chez le destinataire, d’hypothèses au sujet des contenus sur lesquels un locuteur s’engage. L’étude des mécanismes de construction du sens doit, de manière générale, occuper une place significative dans l’étude de l’argumentation.

La plausibilité est une question que nous ne pouvons traiter ici, faute d’espace. Nous postulons toutefois qu’elle est une condition de la pertinence.

L’ENGAGEMENT COMME NOTION COGNITIVE ASSOCIÉE AU DESTINATAIRE

487

Bibliographie Bach, Kent (1994). Conversational impliciture. Mind and Language 9: 124-162 Breheny, Richard, Napoleon Katzos & John Williams (2005). Are generalised scalar implicatures generated by default? An on-line investigation into the role of context in generating pragmatic inferences. Cognition 100/3: 434-463. Burton-Roberts, Noel (2006). Cancellation and intention. Newcastle Working Papers in Linguistics 12/13: 1-12. Capone, Alessandro (à paraître). Are explicatures cancellable? Towards a theory of the speaker's intentionality. Intercultural Pragmatics. Dernier accès le 29.09.2008 sur internet: http:// semanticsarchive.net/Archive/zExOThiM/explicatures.rtf Carston, Robyn (2004). Relevance Theory and the saying/implicating distinction. In: Horn, L. & G. Ward (eds.). Handbook of Pragmatics. Oxford: Blackwell. Geis, Michael L. & Arnold M. Zwicky (1971). On invited inferences. Linguistic Inquiry 2: 561-566. Kratzer, Angelika (1999). Beyond ouch and oops: How descriptive and expressive meaning interact. http://semanticsarchive.net/Archive/WEwNGUyO/. Katriel, Tamar & Marcelo Dascal (1989). Speaker’s commitment and involvement in discourse. In: Tobin, Y. (ed.). From Sign to Text: a Semiotic View of Communication. Amsterdam and Philadelphia: John Benjamins, 275-295. Lascarides, Alex & Nicholas Asher (1993). Temporal Interpretation, Discourse Relations, and Commonsense Entailment. Linguistics and Philosophy 16: 437-493. Morency, Patrick, Steve Oswald & Louis de Saussure (à paraître). Explicitness, implicitness and commitment attribution: A cognitive pragmatic approach. Belgian Journal of Linguistics. Moeschler, Jacques & Louis de Saussure (2002). Pragmatique du discours et interprétation. In: Roulet, E. & M. Burger (éds). Les modèles du discours au défi d’un ‘dialogue romanesque’: l’incipit du roman de R. Pinget : Le Libera. Nancy: Presses Universitaires de Nancy, 379-402. Noveck, Ira (2001). When children are more logical than adults: Experimental investigations of scalar implicature. Cognition 78(2), 165-188. Papafragou. Anna (2006). Epistemic modality and truth conditions. Lingua 116: 1688-1702. Perry, John (1986). Thought without representation. Supplementary Proceedings of the Aristotelian Society 60: 263-283. Récanati F. (2002a). Unarticulated Constituents. Linguistics and Philosophy 25, 299-345. Saussure, Louis de (2003). Cause implicitée et temps explicité. Cahiers de linguistique française 25: 119-136. Saussure, Louis de (2005). Parallélisme et linéarité de l’interprétation: remarques sur un cas de causalité implicite. Intellectica 40: 43-62. Saussure, Louis de (2006.). Implicatures et métareprésentations en contexte de presse écrite. In: Béguelin M.-J., M. Bonhomme & G. Lugrin (éds). Intertextualité et interdiscours dans les médias, TRANEL 44: 57-75. Sperber, Dan (ed.) (2000). Metarepresentations: a multidisciplinary approach. New York: Oxford University Press. Sperber, Dan & Deirdre Wilson ([1986] 1995). Relevance: Communication and Cognition, 2nd edition. Oxford: Blackwell.

488

LOUIS DE SAUSSURE & STEVE OSWALD

Walton, Douglas (1993). Commitment, Types of Dialogues, and Fallacies. Informal Logic XIV (2&3): 93-103. Walton, Douglas (1996). The Straw Man Fallacy. In: Bentham, J. van, F. van Eemeren, R. Grootendorst & F. Veltman (eds). Logic and Argumentation. Amsterdam: Royal Netherlands Academy of Arts and Sciences, 115-128.

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 489-496

RASSEGNA DI LINGUISTICA GENERALE A CURA DI MARIO BAGGIO E MARIA CRISTINA GATTI

FRANS H. VAN EEMEREN ed., Važneišie koncepcii teorii argumentacii/Crucial Concepts in Argumentation Theory, Filologičeskij Fakul’tet Sankt-Peterburskogo Gosudarstvennogo Universiteta, Sankt-Peterburg 2006, pp. 296 Dopo la pubblicazione nel 1992 di Argumentacija, kommunikacija i ošibki, traduzione dell’opera di Frans van Eemeren e Rob Grootendorst Argumentation, Communication and Fallacies: a Pragma-dialectical Perspective, seguita nel 1994 da Rečevye akty v argumentativnych diskussjach e nel 2002 da Argumentacija: analiz, proverka, predstavlenie – rese traduttive rispettivamente dell’opera di van Eemeren e Grootendorst Speech Acts in Argumentative Discussions: a Theoretical Model for the Analysis of Discussions towards Solving Conflicts of Opinion nonché del manuale di van Eemeren, Grootendorst e Snoeck Henkemans Argumentation. Analysis, Evaluation, Presentation – è possibile accostarsi in lingua russa a un ulteriore segmento della produzione scientifica degli studiosi di argomentazione della Scuola di Amsterdam. La teoria dell’argomentazione in Russia è una scienza relativamente recente. Attraverso una serie di contributi a più voci, gli studiosi di argomentazione del Department of Speech Communication, Argumentation Theory and Rhetorics dell’Università di Amsterdam presentano al lettore russo le tematiche più rilevanti al centro della teoria dell’argomentazione e i principali approcci avanzati in proposito. Apre il volume il contributo di Frans van Eemeren Sovremennoe sostojanie teorii argumentacii (pp. 14-33) sullo stato attuale della teoria dell’argomentazione. Qui l’Autore prende le mosse da una sintetica ma esaustiva presentazione di alcuni dei principali orientamenti teorici che si sono susseguiti nella

storia di questa disciplina, dal modello di Toulmin, sviluppato negli anni ’50, ai movimenti dell’‘informal logic’ o dell’‘argumentativisme radical’ di Ducrot e Anscombre degli anni ’70, fino agli orientamenti dialettici e retorici contemporanei, come ad esempio l’approccio di van Eemeren e Grootendorst alla ‘critical discussion’, nel quale la dimensione pragma-dialettica trova una feconda integrazione con quella retorica, grazie al processo dello ‘strategic manouevring’. L’Autore procede poi alla presentazione di alcune nozioni chiave al centro del dibattito contemporaneo, prese in esame in ciascuno dei capitoli successivi. Dalla caratterizzazione del concetto di punto di vista, preso in esame da Peter Houtlosser nel secondo capitolo (Točki srenija, pp. 34-63), si passa a una analisi delle premesse inespresse, svolta da Susanne Gerritsen nel terzo capitolo (Nevyražennye posylki, pp. 64-98), seguita dalla caratterizzazione delle nozioni di schema argomentativo, da parte di Bart Garssen nel quarto capitolo (Schemy argumentacii, pp. 99-122) e di struttura argomentativa nel quinto capitolo (Strukury argumentacii, pp. 123-161), ad opera di Franciska Snoeck Henkemans. Un ricco status quaestionis della riflessione sulle fallacie dall’antichità ai giorni nostri viene proposto da van Eemeren nel sesto capitolo (Ošibki v argumentacii, pp. 162197). Alla attività di interpretazione e ricostruzione degli argomenti è dedicato il capitolo settimo (Interpretacija i rekonstrukcija argumentacii, pp. 198-238), redatto da Agnès van Rees. Segue, infine, nell’ottavo e ultimo capitolo (Argumentacija v oblasti prava, pp. 239-268) una accurata ricostruzione degli innumerevoli studi sull’argomentazione in ambito giuridico, ad opera di Eveline Feteris. Non si possono non menzionare i prezio-

490

RASSEGNA DI LINGUISTICA GENERALE

si apparati bibliografici a chiusura di ciascun capitolo, con l’indicazione delle più importanti fonti per un ulteriore approfondimento delle tematiche trattate. Chiude il volume un prezioso Ukazatel’ terminov (pp. 276-293), contenente la terminologia russa utilizzata nel volume, seguita da quella originale in inglese. Questo ricco apparato terminologico documenta l’intensa attività lessicologica svolta dai traduttori, che si sono trovati a dover creare gran parte della categorialità linguistica necessaria per la traduzione in russo del lessico dell’argomentazione. Maria Cristina Gatti K ONSTANZE J UNGBLUTH , Pragmatik der Demonstrativpronomina in den iberoromanischen Sprachen, Max Niemeyer, Tübingen 2005 (Beihefte zur Zeitschrift für Romanische Philologie, 329), pp. 256 La monografia, ampia per la documentazione e caratterizzata da saldo impianto metodologico, presenta una teoria generale della deissi, che viene applicata nella descrizione dei pronomi dimostrativi usati per riferirsi alla persona. Le lingue considerate sono lo spagnolo, il catalano e il portoghese brasiliano. Nella parte teorica, esposta nel capitolo primo, il fenomeno della deissi è considerato alla luce della tradizione di studi avviati da Karl Bühler e sviluppati da Konrad Ehlich. Peraltro, l’Autrice è attenta alle implicanze che la diversità del genere testuale può avere per il funzionamento dei deittici. A questo proposito, ella riprende la tematica delle Diskurstraditionen cara a Eugenio Coseriu. Per l’aspetto applicativo, si considerano i tre pronomi dimostrativi este/esta/esto, ese/esa/eso e aquel/aquella/aquello dello spagnolo (cap. 2), aquest/aquesta, aixó, aquell/aquella/ alló del catalano (cap. 3), este/esta/isto, esse/essa/isso, aquele/aquela/aquilo del portoghese brasiliano (capitolo 4). Dopo un’analisi formale, attenta alle variazioni diatopiche e diastratiche, si considera l’uso dei pronomi nella deissi spaziale e temporale sia nella comunicazione face-to-face sia in altre situazioni

comunicative, nei quali emerge il ruolo della deissi testuale e della Deixis am Phantasma. Un aspetto innovativo dell’analisi consiste nel rilievo dato alla Hörer-Deixis (“deixis de l’allocutaire” nella terminologia francese). Termina la monografia un saggio di analisi comparativa delle descrizioni svolte per le tre lingue ibero-romanze. Al termine, si tenta una prima, provvisoria comparazione con i sistemi del finnico e del giapponese, caratterizzati anch’essi da tre pronomi dimostrativi. Forse un confronto con l’italiano, lingua romanza dotata di sistema tripolare, sarebbe stato più pertinente. Per la rilevanza scientifica della ricerca sarebbe opportuno integrare la descrizione con i dati del portoghese lusitano e, per l’aspetto storico-linguistico, con i documenti ladini delle comunità sefardite. Giovanni Gobber CLAUDIA CAFFI, Mitigation, Elsevier, Amsterdam 2007, (Studies in Pragmatics 4), pp. 342 In questo bel volume Claudia Caffi riprende approfondendolo il suo ricco contributo sul tema della mitigazione pubblicato nel 2001 (La mitigazione. Un approccio pragmatico alla comunicazione nei contesti terapeutici, LIT, Münster). In particolare, l’Autrice si impegna a ricostruire un dialogo molto ampio con le tradizioni di studio pertinenti per lo studio dei fenomeni linguistici connessi alla mitigazione: a tale obiettivo sono dedicati i primi due capitoli dell’opera. Viene studiato pertanto il tema del “soggetto” e della “soggettività”, partendo da Benveniste, ma tornando poi a Bühler per allargare la prospettiva alla lettura psico-pedagogica di Piaget. A ritmo serrato, vengono passati in rassegna numerosi studiosi di diverso orientamento, da Lyons ad Auer passando per Gumperz, Sbisà e molti altri, per approdare a una prima definizione del fenomeno in oggetto: mitigation è, sulla scorta della tradizione retorica classica della mitigatio (deminutio o imminutio), “understatement”, “Abschwächung”, “downgrading”. Il secondo capitolo esamina la prospettiva

RASSEGNA DI LINGUISTICA GENERALE

pragmatica a partire da Austin fino ai suoi sviluppi più recenti, per individuare i contributi relativi ai fenomeni linguistici che manifestano la modulazione mitigata: l’Autrice circoscrive attraverso tale percorso la scoperta degli indicatori “di sistema” che gestiscono (anche ) la politeness. Emergono dunque le “siepi” con Brown e Levinson e poi con Lakoff, l’indirettezza (a partire da Fraser, ma poi Leech ecc.) e gli “schermi” (di nuovo Brown e Levinson). Caffi ricostruisce inoltre lo studio del fenomeno dell’intensità espressiva, anch’essa collegata con il tema della mitigazione. Conclude questo secondo capitolo una descrizione assai ampliata del fenomeno, che tiene conto della “naturalità” della modulazione e pertanto di una oggettiva difficoltà a cogliere, nella realtà dell’osservazione, ciò che riguarda l’illocuzione da ciò che dipende da effetti di mitigazione o ancora da scelte stilistiche o ancora dal contenuto stesso del messaggio. Una posizione dunque più sfumata rispetto a quella del 2001. A partire da questa base storico-critica la Caffi sviluppa, nei quattro capitoli successivi, le sue osservazioni sulle strategie linguistiche della mitigazione, mettendole in rapporto con gli studi relativi al coinvolgimento emotivo. L’applicazione delle osservazioni svolte in sede teorica a un considerevole corpus di interazioni medico-paziente permette all’autrice di analizzare una cospicua quantità di dati empirici, dando un contributo importante alla conoscenza delle specificità verbali che caratterizzano i contesti “mitigati”. Sara Cigada VALERIA DELLA VALLE – PIETRO TRIFONE ed., Studi linguistici per Luca Serianni, Salerno Editrice, Roma 2007, pp. 665 Valeria Della Valle e Pietro Trifone sono i curatori di questa miscellanea che raccoglie gli studi offerti a Luca Serianni dai suoi allievi per il suo sessantesimo compleanno. Il volume si apre con la bibliografia delle opere di Serianni dal 1972 al 2007 e con l’elenco degli studenti laureati con lui dal 1982 a oggi: sin dalle prime pagine risulta così evidente come l’imponente produzione scienti-

491

fica dello studioso – la cui rilevanza è universalmente nota – e il suo costante impegno didattico siano caratterizzati da una notevole ampiezza di interessi di ricerca. La vastità e la ricchezza degli interessi scientifici dello studioso si riflette poi nella varietà di temi e di approcci metodologici dei 45 saggi raccolti nel volume, che investono numerosi campi di ricerca, come la filologia italiana, la storia della lingua italiana, la linguistica italiana e la didattica dell’italiano. In questa sede dobbiamo limitarci a rendere conto soltanto degli interventi di carattere specificamente linguistico, che trattano argomenti di linguistica italiana o di storia della lingua, pur sapendo di tralasciare dalla menzione numerosi lavori altrettanto degni di interesse e affidando dunque al lettore il compito di colmare questa lacuna. Alcuni studi linguistici si collocano in una prospettiva diacronica. Ne diamo una sintetica presentazione seguendo un criterio di ordine alfabetico, diversamente dall’ordine cronologico, secondo la data di laurea di ogni allievo, adottato nella miscellanea: Giuseppe Patota analizza gli usi della preposizione per considerandone la funzione agentiva attestata nell’italiano antico; Giancarlo Schirru, esaminando un fenomeno di mutamento fonologico, quale il dileguo di consonante in contesto vocalico, discute la classificazione delle proprietà distintive di vocali e consonanti dell’italiano; Francesco Sestito ripercorre la storia delle forme verbali tollere e vollere come varianti di togliere e volgere; Stefano Telve indaga la possibile alternanza degli ausiliari essere e avere con i verbi modali potere e volere seguiti da infiniti di verbi inaccusativi, sottoponendo le osservazioni dei grammatici a un riscontro con le occorrenze d’uso rilevate in un corpus di testi che coprono cronologicamente la storia dell’italiano dal periodo due-trecentesco al Novecento. Segnaliamo poi una serie di contributi che affrontano tematiche linguistiche di varia natura: Gianluca Biasci analizza il trattamento dei problemi di pronuncia in manuali operativi per logopedisti, focalizzando l’attenzione su

492

RASSEGNA DI LINGUISTICA GENERALE

due opposizioni fonologiche “critiche” del sistema italiano, l’opposizione tra la costrittiva alveolare sorda e sonora (/s/ e /z/) e l’opposizione tra l’affricata alveolare corrispondente sorda e sonora (/ts/ e /dz/); Rosarita Digregorio esamina la terminologia della biblioteconomia, segnalando l’ingresso di numerosi neologismi in questo settore; Cristina Faloci propone alcune considerazioni sulla terminologia della psicoanalisi, riflettendo sui problemi traduttivi emersi nella recente revisione della traduzione italiana del Freud di Stefan Zweig ; Vincenzo Faraoni presenta un’analisi a livello linguistico e testuale di un corpus di messaggi tratti da un forum interattivo presente nel sito web del partito di Alleanza Nazionale; Riccardo Gualdo prende in esame il fenomeno dell’interferenza lessicale, considerando l’influsso angloamericano sull’uso dell’aggettivo ironico, appartenente al lessico comune, e sull’uso scientifico-giuridico degli aggettivi sensibile e ragionevole; Marco Lanzarone analizza la terminologia della pianificazione pubblicitaria, in cui rileva la forte presenza di prestiti non adattati e un generale atteggiamento di freno nei confronti del processo di adattamento, volto a evitare l’uso di possibili traducenti, che risulterebbero di uso troppo comune, e a mantenere così l’alto tasso di tecnicismo del settore; Lucilla Pizzoli delinea il percorso storico intrapreso dalla legislazione in materia linguistica per gli italiani all’estero dalla fine dell’Ottocento a oggi; Danilo Poggiogalli, a partire dall’analisi di un corpus di testi di critica letteraria, esamina le diverse soluzioni adottate nello scritto dell’italiano contemporaneo in relazione a due tipi di accordo dell’aggettivo, l’accordo con nomi coordinati e quello con uno o più nomi in presenza di composti aggettivali; Lucia Raffaelli propone alcune riflessioni sul linguaggio giornalistico contemporaneo, considerando l’uso dell’interpunzione e le caratteristiche sintattiche dei titoli dei quotidiani; Alessio Ricci dedica il suo contributo alla scrittura degli studenti universitari, mettendone in luce gli aspetti critici in relazione ai diversi livelli linguistici; Francesco Zardo

documenta la presenza consistente di marchionimi nella lingua italiana, offrendo un aggiornamento dello spoglio lessicografico da lui condotto nel ‘94 e ora messo a confronto con dizionari di recente pubblicazione e con dizionari di neologismi. Silvia Gilardoni ELISABETTA FAZZINI – ELEONORA CIANCI ed., I Germani e la scrittura. Atti del XXXIII Convegno dell’Associazione Italiana di Filologia germanica, Pescara, 7-9 giugno 2006, Edizioni Dell’Orso, Alessandria 2007, pp. 242. Il volume si apre con un’approfondita presentazione di Fabrizio Raschellà il quale nota che la scrittura rappresenta per vari aspetti un confine e un passaggio tra realtà diverse: essa è strumento primario di trasmissione e di conservazione della memoria; per mezzo della scrittura i popoli possono rendersi consapevoli e diventare protagonisti della propria storia e delle proprie tradizioni. I Germani cominciarono a usare la scrittura agli inizi dell’era cristiana con un alfabeto in origine desunto da popolazioni italiche, stanziate nelle zone alpine di frontiera, e rielaborato in modo originale. Tale alfabeto oggi è chiamato runico o futhark (fuþark nome desunto dalle prime sei lettere); era usato per brevi iscrizioni commemorative o dedicatorie, talora di carattere rituale. Queste iscrizioni spesso presentano solo il nome del committente o dello scrivano e si trovano su oggetti diversi, il cui materiale è osso, metallo, legno, pietra. Poi, in seguito alla conversione al cristianesimo di vari popoli germanici, l’alfabeto runico viene sostituito dagli alfabeti usati dai popoli cristiani; i Visigoti adottarono l’alfabeto greco, gli altri popoli usarono l’alfabeto latino. L’alfabeto runico tuttavia fu ancora usato nella penisola scandinava come scrittura epigrafica, raramente applicata alla pergamena. Nel volume sono compresi i tredici interventi presentati al Convegno. Nicoletta Francovich Onesti (pp. 1-12) presenta le interferenze latine nella scrittura gotica, fondata sull’alfabeto creato nel IV secolo dal vescovo

RASSEGNA DI LINGUISTICA GENERALE

visigoto Vulfila guida dei Gothi minores della Mesia, che si basò sui modelli greco, latino e runico. La Francovich Onesti cerca di dimostrare come nel VI secolo i contatti col latino scritto e parlato abbiano portato dei mutamenti nell’ortografia gotica; tuttavia secondo la studiosa alcune innovazioni grafiche sono dovute ad evoluzione spontanea della fonologia gotica. Alessandro Zironi (pp. 13-38) studia lo sviluppo della scrittura gotica in età carolingia, esaminando in particolare il caso dei Gotica Vindobonensia, soffermandosi sul ms. 795 della Österreichische Nationalbibliothek di Vienna, che contiene anche lettere di Alcuino inviate a diversi destinatari tra cui Arno arcivescovo di Salisburgo. I ff. 21r-191v presentano scritti esegetici anche sulle lettere di S. Paolo e su scritti di autori del periodo tardo antico (ff. 21r-148v); si trovano anche alcuni commenti di Alcuino (ff. 148v-150v.) e lettere di Alcuino a vari destinatari (ff. 150v-183v). I ff. 192r-205r furono uniti più tardi al ms. e presentano lettere di Alcuino ad Arno (ff. 192r-197v), lettere di Angiberto abate di Saint-Riquier ad Arno (ff. 197v199r), un frammento poetico di Alcuino diretto al discepolo Candidus (ff. 199r-199v), una lettera di Carlo Magno ad Alcuino databile al marzo 790 (ff. 200r-203v). Molto importanti dal punto di vista della storia della scrittura sono i fogli 1r-20v, ff. 5r-18v Alcuino, Ortographia, ff. 19r-19v alfabeto greco con pronuncia; f 20r alfabeto runico con valore fonetico, 16 lettere dell’alfabeto gotico, una lettera gotica isolata. Il folio 20v presenta due alfabeti gotici con i nomi delle lettere, brevi testi gotici, regole di pronuncia del gotico. Il ms. Vindobonensis 795 si inserisce in una poliedrica attività che dà origine alla rinascenza carolingia aperta a vari orizzonti: gusto alfabetico, studi grammaticali, istanze politiche. Secondo Zironi, l’interesse per il gotico nel periodo carolingio potrebbe essere dovuto all’opera del grammatico Smaragdo e al poeta teologo Teodolfo d’Orleans. Elvira Glaser (pp. 39-50) analizza la prassi glossografica altotedesca antica, presentando una sintesi delle cosiddette “glosse a sec-

493

co” (cioè glosse incise sulla pergamena con la punta dello stilo priva d’inchiostro) frequenti nella glossografia tedesca medievale, di cui sono il nucleo più antico (metà VIII secolo). L’Autrice mette anche in evidenza la connessione tra glosse a secco e glosse crittografiche, eseguite dando alle lettere valori diversi da quelli tradizionali. Eleonora Cianci (pp. 5167) si occupa delle modalità di annotazione degli incantesimi nei mss. medievali di area tedesca e cerca di spiegare la collocazione (spesso problematica) all’interno dei mss. in cui si trovano. In genere questi mss. sono miscellanei e si possono suddividere in raccolte di scritti religiosi e in raccolte di argomento medico. L’analisi della studiosa si basa sullo studio di 32 mss. esaminati con profonda competenza. Simona Leonardi (pp. 69-94) studia il verbo scriban-schriben dalle origini alla prima età moderna. Il tema studiato riguarda il fenomeno della scrittura in prospettiva pragmatica e contrastiva: latino volgare e tedesco in fase evolutiva a confronto. La studiosa rileva che scriban riguardava l’insieme globale attinente alla scrittura, nel periodo tardomedievale e protomoderno scriben riguarda l’azione materiale dello scrivere, mentre tichten indica l’azione intellettuale del comporre. Concetta Sipione (pp. 95-112) studia i problemi della lingua theodisca in rapporto a Otfrid di Weissenburg che condusse uno dei primi tentativi di normalizzazione ortografica della lingua tedesca. Sono riflessioni del monaco e poeta francone Otfrid, presenti nella lettera con cui Otfrid presenta la sua versione dei Vangeli in versi tedeschi a Liutberto arcivescovo di Magonza. L’Autrice studia in particolare i passi in cui alcuni suoni della lingua materna di Otfrid sono resi in alfabeto latino. Valeria Di Clemente (pp. 113-125) studia il Prüller Kräuterbuch, esaminando gli aspetti paleografici e grafematici nel ms. Clm 536 della Bayerische Staatsbibliothek di Monaco di Baviera, formato da vari fascicoli rilegati insieme del periodo compreso tra la fine dell’XI secolo e la seconda metà del XII secolo con inserti del XIII secolo. Il ms. era in origine conservato nel monastero di Prüll presso Ratisbona. I

494

RASSEGNA DI LINGUISTICA GENERALE

testi perlopiù latini ma anche tedeschi sono vari: teologici, naturalistici, sapienziali; alcuni presentano visiones e scongiuri. Il testo esaminato dalla studiosa è un erbario latinobavarese dell’XI-XII secolo. Sono messe in evidenza le due mani a cui si deve la stesura del testo, il cui sistema fonologico è reso da un’evidente instabilità grafica dovuta anche al periodo di transizione dalla fase antica alla fase media dell’altotedesco. Maria Giovanna Arcamone (pp. 127-139) presenta le problematiche delle iscrizioni runiche d’Italia, studiando le iscrizioni runiche trovate nel santuario di San Michele a Monte Sant’Angelo sul Gargano e in alcune catacombe romane: si tratta di otto testi, dal punto di vista linguistico di area anglosassone (VII-VIII secolo). La studiosa presenta qualche rettifica ai testi, esaminandone i problemi linguistici e onomastici. Seguono le riproduzioni delle iscrizioni (pp. 140-149). Giulio Garuti Simone (pp. 151-160) analizza l’alfabeto runico di Modena, approfondendo le equivalenze alfabetiche runico-latine che si trovano nel ms. O.1.11 dell’Archivio Capitolare di Modena (inizi IX secolo). L’alfabeto runico è preceduto dalla presentazione dell’alfabeto greco. Carla Del Zotto (pp. 161-186) studia le rune fra magia e scrittura nella letteratura medievale. L’Autrice percorre il cammino della scrittura runica dalle più antiche iscrizioni, alcune delle quali si riferiscono a pratiche magiche, ai poemetti runici del tardo medioevo nordico, poemetti nei quali sono usate anche le rune come grafemi integranti l’alfabeto latino. La studiosa evidenzia anche l’uso di alfabeti runici in mss. latini medievali. Marco Battaglia (pp. 187-208) studia la scrittura come mezzo di appartenenza a una comunità etnica, soffermandosi in particolare su iscrizioni votive contenenti nomi di divinità pagane, trovate in luoghi di insediamento romano-barbarico nei primi tre secoli del periodo cristiano. Antonietta Amati e Lucia Sinisi (pp. 209-227) si occupano dell’identità culturale latino-longobarda in Puglia e in particolare in Terra di Bari (VIIIXII secolo), identità evidenziata anche mediante la scrittura “langobardisca” o beneven-

tana nella variante nota come “Bari type”. Il volume si chiude con un contributo di Fabrizio Raschellà (pp. 229-240), il quale rileva che manca un approccio comparativo nell’analisi dei sistemi di scrittura usati nelle lingue germaniche. Il problema resta aperto. Nell’insieme dei lavori raccolti nel volume è evidenziata la funzione storica e culturale della scrittura nel complesso contesto delle manifestazioni della creatività umana che, come rileva F. Raschellà, “concorrono a definire l’identità dei popoli e il suo mutare nel tempo e nello spazio”. Celestina Milani SIMONA MESSINA, Between the real and the unreal. The mimesis of Italian speech in TV series, “Linguisticae Investigationes”, 30, 2007, 2, pp. 261-290 L’Autrice elabora uno studio su alcuni aspetti dell’italiano basandosi sul linguaggio “trasmesso” delle serie televisive. A tal fine, l’Autrice prende in considerazione la family fiction, ossia quei cicli di episodi che narrano la vita quotidiana di una famiglia o di un gruppo di famiglie alle prese con avvenimenti più o meno significativi. Questo tipo di prodotto attinge per lo più al repertorio dell’italiano colloquiale, nel tentativo di ricreare per imitazione un certo contesto linguistico. Il corpus SLFF (spoken language in family fiction), analizzato nella ricerca, è tratto da due serie prodotte dalla RAI rispettivamente nel 1968 e nel 1988: La famiglia Benvenuti e Un medico in famiglia. Oggetto dell’analisi è l’imitazione della lingua parlata nel processo globale di messa in scena di uno stralcio di vita quotidiana. Per dare corso alla ricerca, l’Autrice individua alcuni fenomeni linguistici particolarmente significativi nel discorso spontaneo in italiano, quali il registro, le espressioni idiomatiche, l’uso polivalente del che, i meccanismi di focalizzazione, e ne analizza le occorrenze nei copioni delle serie considerate. Risultati differenti quanto a incidenza, distribuzione e morfologia dei suddetti meccanismi linguistici paiono suggerire che la serie televisiva del 1968 mostri una maggiore at-

RASSEGNA DI LINGUISTICA GENERALE

tenzione all’appropriatezza del linguaggio, tanto da proporre non di rado nei dialoghi note metalinguistiche – talvolta anche a carattere prescrittivo –, a scapito dell’efficacia mimetica. La serie più recente mostra invece un’attitudine realista più spiccata; in essa l’uso di un registro informale medio-basso, che spesso indulge in espressioni volgari, produce un’imitazione più fedele del discorso colloquiale. Al di là degli esiti raggiunti la ricerca pare particolarmente interessante in quanto rappresenta un tentativo di legittimazione e promozione del mezzo televisivo come fonte di materiale linguistico da analizzare. L’Autrice considera i prodotti televisivi come una risorsa di primaria importanza per lo studio del linguaggio, in quanto essi forniscono esempi di lingua parlata in una cornice temporale definita e rendono possibile, attraverso l’archiviazione digitale e cartacea, l’accesso ricorsivo ai dati, consentendo allo studioso anche la possibilità di istituire confronti diacronici tra stati successivi di una lingua. In questa cornice di studio va tuttavia tenuto in considerazione lo status particolare dei copioni di serie televisive come quelle oggetto della ricerca descritta, che non sono esempi di lingua “viva”, bensì tentativi di imitazione del parlato reale, in forma scritta. Chiara Beccalossi S ILVIA D AL N EGRO – W ILCO L ENSINK – CHRISTIAN UPMEIER – PAOLO VOLONTÉ, Visual communication in a multilingual context, “Zeitschrift für Literaturwissenschaft und Linguistik”, 148, 2007, pp. 113-131 L’articolo presenta i primi risultati di un work in progress a carattere interdisciplinare condotto da un team dell’università di Bolzano. Argomento della ricerca è la comunicazione visuale, intesa come processo di trasmissione di messaggi mediante l’interazione di elementi iconici e simbolici, e il suo ruolo di modificatore dello spazio urbano. Attingendo alla cornice teorica dei modelli inferenziali della comunicazione, e, soprattutto, agli sviluppi apportati dalla teoria della pertinen-

495

za, gli Autori sottolineano come i supporti di messaggi a carattere visuale (quali segnali, manifesti, ecc.) non si traducano sempre e comunque in atti comunicativi. Comunicare, infatti, non significa trasmettere un messaggio, bensì trasformare la realtà per far sì che altri producano atti significativi conformi alle nostre aspettative. La ricerca sul campo ha luogo – in questo caso – all’autostazione di Bolzano, un contesto multilingue che presenta un elevato livello di complessità. I ricercatori prendono in particolare considerazione gli avvisi a carattere pubblico-istituzionale, redatti nella maggioranza dei casi, secondo le norme vigenti, parallelamente in italiano e in tedesco, ove, peraltro, l’impiego di elementi iconici (disegni esplicativi e ausili grafici in genere) pare scarso (un caso, sugli otto analizzati). A questo proposito, essi notano come la leggibilità dei messaggi bilingui venga significativamente compromessa dalla scarsa distinzione tra le lingue impiegate, quanto a impostazione fisica dei messaggi e agli accorgimenti tipografici in essi adottati. Gli Autori propongono esempi di come la differenziazione tipografica può favorire in misura considerevole la comunicazione, ma si spingono oltre, proponendo un maggiore ricorso a una seconda modalità di realizzazione di messaggi destinati a un pubblico multilingue, ottenibile con un maggiore e più consapevole impiego contestuale di strumenti iconici e tipografici. Chiara Beccalossi P AOLA P IETRANDREA , The grammatical nature of some epistemic-evidential adverbs in spoken Italian, “Italian Journal of Linguistics”, XIX, 2007,1, pp. 39-63. L’articolo è contenuto in un numero monografico della rivista, curato da M. Squartini, dedicato al concetto di evidenzialità, ovvero a quella categoria che permette al parlante di segnalare la fonte dell’informazione o il modo in cui è venuto a conoscenza del contenuto proposizionale. In termini più generali, la categoria dell’evidenzialità rientra tra gli strumenti linguistici che permettono al par-

496

RASSEGNA DI LINGUISTICA GENERALE

lante di indicare l’evidenza di cui dispone (inferenza, discorso altrui, percezione sensoriale) per asserire la verità di una proposizione. Il crescente interesse intorno a questa nozione e alle marche che la manifestano ha reso centrale il dibattito sulla natura grammaticale o lessicale di tali marche evidenziali, individuando a questo riguardo lingue a base grammaticale e lingue a base lessicale. L’interazione tra evidenziale ed epistemico e tra grammaticale e lessicale è analizzata da Pietrandrea che rivisita il sistema epistemico-evidenziale dell’italiano integrando la categoria grammaticale dei verbi epistemici (futuro epistemico, dovere e potere) con quella lessicale degli avverbi (forse, magari) e delle costruzioni avverbiali modalizzanti (secondo me, si vede, dice che contrapposta a dice).

Sulla base dell’analisi condotta l’Autrice conclude che in italiano gli evidenziali lessicali e quelli grammaticali non costituiscono due paradigmi in opposizione bensì in stretta relazione, tanto che gli avverbi epistemici analizzati potrebbero essere considerati come forme debolmente grammaticalizzate, piuttosto che come forme totalmente lessicali, poiché formano una classe chiusa di opzioni omogenee e complementari e mostrano una ridotta mobilità sintattica associata a particolari configurazioni discorsive. Maria Paola Tenchini

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 497-499

RASSEGNA DI GLOTTODIDATTICA A CURA DI BONA CAMBIAGHI

C LARA M ONTELLA – G IANCARLO M ARCHESINI ed., I saperi del tradurre, Franco Angeli, Milano 2007, pp. 281 Sulla “traduzione” luogo di riflessione per eccellenza negli studi di linguistica, quinta abilità in quelli di glottodidattica, è apparso questo volume che si deve all’École de Traduction et d’Interpétration dell’Université de Genève, e all’Università degli Studi di Napoli, “L’Orientale”. Diamo conto brevemente di tutti i contributi ivi contenuti. Clara Montella, Le nozioni di significato e di senso in traduttologia tra storia, teorie e applicazioni pp. 11-44. Le nozioni di “significato” e di “senso” si possono far risalire a Schleiermacher, che fonda, secondo l’A., una ermeneutica della traduzione. Il teorico più citato nel saggio è Coseriu, cui l’A. collega il pensiero di Hagège presentato nella “teoria dei tre punti di vista”, a) cioè la zona del “senso come rappresentazione-descrizione”, b) cioè la zona del “senso come effetto” e c) cioè la zona “delle significanze inconsce”. Il paragrafo 2, relativo a significato e senso come “termini” nella didattica della traduzione, mostra la complessità dell’argomento e quindi la difficoltà alla riduzione dei 1419 termini raccolti, relativi a 838 nozioni in un apparato utile ad una disciplina che ha nel suo statuto epistemologico il concetto di “non ridondanza” qual è appunto lo statuto della “glottodidattica”. Complicazioni che hanno un riscontro, se possibile ancora maggiore, nella presa in conto del concetto di “traduzione” del processo studiato da Muntadas, il quale “analizza nelle forme consone all’arte multimediale virtuale la problematica della comunicazione, che egli identifica nella traduzione, nella misura in cui il net mette in contatto parlanti di lingue e culture differenti” (p. 37).

Giancarlo Marchesini, Teorie della traduzione e strategie traduttive, pp. 45-69. L’A., richiamandosi al saggio di Jirì Levì, La traduzione come processo decisionale, affronta il problema della traduzione come interpretazione e insiste sulla necessità di compiere comunque sempre una scelta nel momento traduttivo. Il movimento target oriented vs. source oriented permette di parlare di traduzione “libera” come dotata unicamente di una “libertà vigilata” e di “interpretazione” come manipolazione testuale vs. “interpretazione” come manipolazione concettuale. Laura Salmon, Sulla gestione inconscia del processo traduttivo: cosa sappiamo fare senza sapere come, pp. 71-99. Partendo dalla distinzione tra traduzione come “processo” e come “prodotto”, la Salmon propone una didattica della traduzione che sappia sfruttare le neuroscienze, cioè il funzionamento del cervello che, grazie ai neuroni specchio, mette in azione una memoria procedurale, legata e dipendente da una componente motoria del linguaggio umano. L’educazione e quindi l’avvio al “flusso coerente della comunicazione” (p. 90) tra una LP ed una LA mediante esercizi di imitazione, ripetizione e rinforzo, ispirano dunque una didattica atta a preparare buoni professionisti interpreti e traduttori piuttosto che grammatologi. Tali professionisti devono essere avviati al “frequente passaggio dall’automatismo alla coscienza e dalla coscienza all’automatismo” (p. 93). Una grammatica delle opzioni, collegata con l’enciclopedia del traduttore a sua volta connessa a una rete ipertestuale enorme e molto ramificata, deve sostituire una grammatica delle regole, nella convinzione che –ed è questa la conclusione della Salmon– “ogni volta che ci limitiamo a insegnare una regola esplicita, stiamo al contempo inibendo una procedura implicita” (p. 96). Erika Salsnik, Dagli universali traduttivi

498

RASSEGNA DI GLOTTODIDATTICA

all’italiano delle traduzioni, pp. 101-131. Partendo dalla presentazione di alcuni concetti-chiave negli studi traduttologici (relativi ad esplicitazione, semplificazione e normalizzazione) nei testi tradotti rispetto ai testi originali, l’A. presenta la “fragilità” del concetto di universale traduttivo, soprattutto in riferimento all’“italiano delle traduzioni”, ne esamina alcuni casi e ne mette in luce incongruenze e possibili, facili interferenze. Patrizia Mazzotta, Riflessioni glottodidattiche sulla traduzione delle microlingue, pp. 133-145. Il processo traduttivo è visto come un processo duplice endo- ed eso-linguistico di comprensione del testo di partenza e di riformulazione nel testo d’arrivo. Il primo momento implica conoscenze tecniche, settoriali, culturali, implicite e di convenzionalità che la didattica ha sempre sottovalutato; il secondo implica intuito, interpretazione e capacità di mediazione rapida, cioè “ricerca di un punto di incontro che riduca la distanza tra sensibilità, interessi, mondi diversi” (p.136). Sempre da un punto di vista didattico, non si dimentichi poi che la traduzione “costituisce uno straordinario strumento di riflessione metalinguistica” (p.143), per cui il suo uso in classe è di altissimo valore durante tutto l’iter apprenditivo. Enrico Borello, La traduzione nella storia della glottodidattica, pp.147-172. Excursus storico sui problemi della traduzione nella proto-glottodidattica e nella glottodidattica vera e propria. La traduzione nell’Antichità, nel Medioevo e nella Scuola di Toledo precede alcune considerazioni sulla traduzione nel 1600 e nel 1700, secoli in cui essa comincia ad essere considerata in chiave apprenditiva. Nel XVIII secolo la traduzione interlineare diventa “metodo” per la didattica linguistica, sfocia nella sua messa al bando nei metodi diretti e nelle teorie dei fonetisti del primo Novecento. Più interessante la seconda parte del saggio sul punto di vista degli strutturalisti, e quindi sulle sette categorie di relazioni tra parole di lingue diverse (R. Lado), sui tre tipi di traduzione endolinguistica, interlinguistica e intersemiotica (R. Jakobson), cioè su linguistica e traduzione, fino

all’elencazione delle sette operazioni traduttologiche della “stilistica comparata” di Vinay e Darbelnet. Newmark è ricordato come colui che distingue una traduzione semantica (più adatta ai testi letterari) da una traduzione comunicativa (meglio adatta a cronaca, saggistica, avvisi). Il “saper tradurre” è valutato secondo la scala di competenza traduttiva adattata da Munby. Johanna Monti, Localizzazione: il ruolo e i saperi della traduzione, pp. 173-197. Dopo aver definito il concetto di “localizzazione”, l’A. ne analizza il processo e ne traccia i contorni in una visione industriale, che considera il localizzatore come un nuovo profilo professionale, che si distingue dal traduttore tecnico per “le ulteriori competenze informatiche e gestionali richieste dal mercato e dal processo stesso” (p. 194). Véronique Sauron, Les outils d’aide à la traduction: ennemis ou alliés du traducteur?, pp. 199-226. Testo molto problematico, che mette a confronto i traduttori di un tempo, veri mediatori fra due lingue-culture e i traduttori moderni, che, avvalendosi delle tecnologie continuamente in evoluzione, disintegrano il processo del tradurre e trasformano il traduttore in un assistente di sistemi informatici, cui sono delegati quasi tutti i compiti sottesi alla complessa operazione del tradurre. Olivier Demissy Cazeilles, Traduction littéraire et langues vernaculaires: l’Ecosse, pour une éthique de la différence, pp. 227-243. L’A ., dopo aver presentato il “caso” della Scozia , dove coesistono almeno tre lingue ufficiali e sei lingue regionali, mostra l’insufficienza della dicotomia, che circola in traduttologia, tra traduzione source-oriented e traduzione target oriented, e, attraverso numerosi esempi tratti da testi letterari e dalle rispettive traduzioni o dai tentativi di traduzione, ci dice come “la traduction est dans la culture, la traduction est culture” (p. 242). Antonella De Laurentiis, Il compito del traduttore secondo Borges, pp. 245-262. “[…] l’idea che una traduzione non sia per forza di cose inferiore all’originale” (p. 246) sembra essere la “teoria” della traduzione per Borges,

RASSEGNA DI GLOTTODIDATTICA

che l’A. di questo saggio indica come colui che traccia in negativo una teoria dell’enunciazione, fino all’affermazione che “tutto è transitorio e niente è per sempre” (p. 258). A tale giudizio arriva l’A . attraverso citazioni di Eco, Steiner, Derrida ed altri studiosi di traduzione-interpretazione. Mauro Ferraresi, Perduto nella traduzione, pp. 263-278. Testo di sintesi finale in partenza dalla disciplina semiotica, valuta come deve essere tradotto l’errore, che è di volta in volta di diversa natura.

499

Ai tre ormai classici tipi di traduzione teorizzati da Jakobson (1959): la traduzione endolinguistica o riformulazione, la traduzione interlinguistica o traduzione propriamente detta e la traduzione intersemiotica o trasmutazione, l’A. ne aggiunge un quarto, la traduzione transemiotica o “sinestesizzazione”, “che consiste nella interpretazione di sistemi di segni non linguistici con sistemi di segni non linguistici” (p. 273), in tal modo allargando ulteriormente il concetto di traduzione, e apparentandolo a quello di interpretazione, ma tenendolo dallo stesso ben distinto. “Traduzione partecipata” sarebbe la proposta nuova, in cui “negoziazione”, da intendersi come negoziazione esplicita con l’autore del testo di partenza, costituirebbe la parola-chiave. Bona Cambiaghi

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 501-506

RASSEGNA DI LINGUISTICA FRANCESE A CURA DI ENRICA GALAZZI E CHIARA MOLINARI A LAIN R EY – F RÉDÉRIC D UVAL – G ILLES SIOUFFI, Mille ans de langue française. Histoire d’une passion, Honoré Perrin, Paris 2007, pp. 1465. Ce gros volume, fruit de la collaboration de trois éminents spécialistes, se signale tant par l’effort de synthèse que par la prise en compte des acquis les plus récents sans que soient négligées les études “classiques” dans le domaine de l’histoire de la langue. Les parties qui le composent (Le Moyen Âge, au sens le plus large du mot, puisque cette section englobe ancien et moyen français, jusqu’en 1530, par F. Duval; De la Renaissance à la Révolution par G. Siouffi; Du Premier Empire au XXIe siècle, par A. Rey), au-delà de l’autonomie de traitement qu’imposent la diversité des époques et l’évolution linguistique, abordent des thématiques récurrentes: on signalera notamment la coprésence de l’histoire interne de la langue (où, très grossièrement, l’accent porte plus sur la graphie, le lexique, la syntaxe, que sur la phonétique et la morphologie) et de l’histoire externe (notamment la question ancienne des contacts entre le français et d’autres langues, classiques ou modernes, et les polémiques liées aux emprunts et à la suprématie du français). Si une lecture suivie peut être conseillée aux étudiants et aux amateurs, les spécialistes pourront déceler des parcours de lecture privilégiés: la question du rapport entre latin et français, actuel jusqu’à la Révolution au moins; les variétés linguistiques (au sein même de l’Hexagone et à plus forte raison dans la francophonie); le long parcours qui mène à l’“institutionnalisation” du français et à sa normalisation; la “passion” des Français pour leur langue, thème de grande actualité mis en exergue dans l’intitulé même de cet ouvrage, mais dont les racines remontent bien loin dans l’Histoire; et encore – et c’est

un aspect qui démarque ce livre de bien d’autres – la dimension littéraire, réintégrée ici de façon consciente et voulue à côté de bien d’autres domaines d’emploi (pour les périodes les plus récentes: la chanson, le téléphone, les nouveaux médias, et jusqu’aux téléphones portables, le courrier électronique…). Un regret: les Notes sont réunies aux pp. 1321-1400, ce qui impose un mouvement de va-et-vient pas toujours commode; et une question concernant la Bibliographie (pp. 1401-27): pourquoi avoir marqué d’un astérisque, dans la section consacrée aux XIXe et XXe siècles justement, les ouvrages “publiés au XIXe siècle et au début du XXe” (marque qui n’est d’ailleurs pas toujours présente: cf. le Dictionnaire de Littré)? Signalons enfin les Cartes (1429-37), et les Indices: l’Index sélectif des noms propres (1439-48), Noms de langues et dialectes (1449-51), Domaines du français (1453-55), Quelques thèmes (1457-58), index qui réunit tant des intitulés généraux (Phonétique ou Norme, par ex.) que des sujets très précis (Clavier d’ordinateur ou Médias en Afrique): admettons toutefois qu’un tel tri n’était pas facile. Maria Colombo Timelli JEAN-PHILIPPE DALBERA, Des dialectes au langage. Une archéologie du sens, Champion, Paris 2007, pp. 464 Il saggio rende noti i risultati di una ricerca etimologica centrata sulle lingue romanze, ivi compresi i dialettismi, e si compone di una serie di schede lessicali – la cui area semantica è relativa per lo più agli zoonimi – redatte secondo i principi dell’analisi linguistica di tipo motivazionale: anche se, come osserva l’A., non sappiamo più perché il gatto sia denominato gatto, siamo ancora in grado di comprendere che il martinet probabilmente trae il suo nome dal martello, per metonimia.

502

RASSEGNA DI LINGUISTICA FRANCESE

L’analisi etimologica considera in particolare le denominazioni di animali quali la belette, il martinet, il rouge-gorge, la chouette, la grenouille, il renard. Quest’ultimo termine merita una citazione. È noto che in francese renard ha sostituito la precedente denominazione goupil (< lt. vulpiculus) e questo fatto è abitualmente attribuito all’influsso di un antroponimo Renard, tramite il Roman de Renart. L’A. cerca al contrario di giustificare etimologicamente il termine in questione partendo dal femminile renarde, e ipotizzando una forma originaria *rag-na (‘regina, madre nutrice’). Si tratta dunque di un lavoro approfondito ed erudito, che non esita ad avanzare ipotesi che pongono in discussione alcuni dei luoghi comuni più diffusi. Anna Slerca FRANCK FLORICIC ed., La négation dans les langues romanes, John Benjamins, Amsterdam/Philadelphia 2007 (Linguisticæ investigationes. Supplementa, 26), pp. 229 Ce volume est à la fois une contribution au renouveau des études romanes, moins pratiquées en France que dans d’autres pays européens, et au riche domaine de la négation, qui se prête à l’analyse tant sur le plan de la morphologie que de la pragmatique, de la syntaxe que de la sémantique. Les études rassemblées ici témoignent de diverses approches méthodologiques, en diachronie ou en synchronie, sans privilégier aucune orientation théorique particulière. A côté de contributions centrées sur le latin et l’italien, le catalan et l’espagnol, l’occitan, le roumain, relevons, pour le français: Franck Floricic, Françoise Mignon, Négation et réduplication intensive en français et en italien (pp. 117136), discutent la relation entre propriétés syntaxiques et sémantiques; Tine Van Hecke, La négation de la modalité déontique. Divergences et convergences entre français, italien et roumain (pp. 161-176), traite du problème de la négation des modaux du type ‘falloir’, ‘devoir’; Hélène Huot, La préfixation négative en français moderne (pp. 177-203), étudie les unités lexicales contenant le préfixe –in,

en particulier les dérivés d’adjectifs; G. Dal et alii, Les adjectifs en in-X-able en français (pp. 205-223), avancent l’hypothèse que le préfixe –in est l’exposant d’une règle de construction des lexèmes qui est exposée ici. Un index des formes linguistiques et des termes techniques clôture le volume. Anne Schoysman OLIVIER BERTRAND – HILTRUD GERNER – BÉATRICE STUMPF ed., Lexiques scientifiques et techniques. Constitution et approche historique, Les Éditions de l’École Polytechnique, Palaiseau 2007, pp. 269 Comme le soulignent tant Olivier Bertrand dans son avant-propos que Joëlle Ducos dans les pages conclusives, l’émergence d’un lexique va nécessairement de pair avec la construction d’une science, et la constitution d’un lexique scientifique, dans n’importe quel domaine, dépasse largement la simple question de l’emprunt translinguistique ou du néologisme. Après une réflexion théorique et méthodologique due à Robert Martin (Le traitement lexicographique des mots scientifiques et techniques), rapportée aux bases de données DMF, DMF2 et DMF3, les contributions réunies dans ce beau volume se rapportent à des domaines précis voire à des cas d’espèces: Thomas Städtler revient sur les traductions de Nicole Oresme; trois articles concernent le lexique de la médecine en ancien et moyen français: Isabelle Vedrenne-Fajolles (tradition hippocratique), Annelies Bloem et Michèle Goyens (traduction des Problèmes d’Aristote), Sylvie Bazin-Tacchella (lexique de l’anatomie); trois autres se concentrent sur la botanique au XVIe siècle: Éva Buchi (lexique pharmacologique chez Baptiste Cavigioles, 1541), Philippe Selosse (expression de la diversité végétale), Françoise Henry (vocabulaire viti-vinicole); d’autres domaines sont abordés dans les études de Stephen Dörr (astronomie en ancien français), Silvia Toniato (mathématiques au Moyen Âge), An Smets (fauconnerie en a. et m. fr.), Trung Tran (architecture dans le Songe de Poliphile, 1546), Franz Rainer

RASSEGNA DI LINGUISTICA FRANCESE

503

(comptabilité aux XVIe-XVIIe ss.), Elisabetta Deriu (art équestre aux XVIe-XVIIe ss.); une deuxième contribution de Philippe Selosse concerne les dénominations des savoirs en français préclassique. On signalera aussi l’intérêt de la bibliographie et sitographie générale (pp. 255-258), et surtout de l’index des mots et des concepts, outil indispensable pour une consultation ciblée de cet ouvrage. Maria Colombo Timelli

phones et anglophones enregistrés en milieu naturel (pp. 31-37). Nous signalons également, en dehors du dossier, l’article de Béatrice Fracchiolla, qui s’interroge sur le sémantisme des pronoms français dans une perspective à la fois énonciative, interactionniste et pragmatique, afin de montrer comment leurs emplois sont liés à la relation d’interlocution (pp. 43-48). Cristina Brancaglion

G ILLES R OQUES , Les régionalismes dans le fabliau de ‘Barat et Haimet’, in SARAH BAUDELLE -M ICHELS – M ARIE -M ADELEINE C ASTELLANI – P HILIPPE L OGIÉ – E MMANUELLE P OULAIN -G AUTRET ed., Romans d’Antiquité et littérature du Nord. Mélanges offerts à Aimé Petit, Honoré Champion, Paris 2007, pp. 719-727

M URIEL B ARBAZAN , Le temps verbal. Dimensions linguistiques et psycholinguistiques, Presses Universitaires du Mirail, Toulouse 2006, pp. 470.

L’étude des picardismes dans ce fabliau attribué à Jehan Bodel (bauc, bracon, cenaille, enclignier, esclignie, escuire, estel, garraz, laiszon, en maison, mont, orteille, faire la siue, sole, tempre, toitel, tremelerres) confirme la qualité du ms. D (BnF fr. 19152), qui conserve la quasi-totalité de ces mots régionaux et qui semble ainsi refléter au mieux le lexique de l’auteur. Maria Colombo Timelli “L’Information grammaticale”, 110, 2006 Dans le cadre du dossier consacré à “Le personnel sujet”, Danielle Leeman rend compte du traitement des cas d’omission du sujet dans quelques grammaires de référence du français et dans la grammaire chomskyenne; ce phénomène est ensuite analysé dans deux corpus de français parlé établis par Claire Blanche-Benveniste en 1990 et en 2002, ce qui permet à l’auteur d’avancer l’hypothèse que l’effacement du sujet grammatical manifeste le non-engagement, ou le désengagement, du locuteur (pp. 23-30). Dans une perspective cognitive, Aliyah Morgenstern étudie l’acquisition du processus d’auto-désignation avec le pronom je chez les petits enfants, à travers des corpus d’enfants franco-

L’ouvrage de M. Barbazan propose une description sémantique du système verbal du français, et notamment des temps du passé, selon une approche multidisciplinaire qui se nourrit de l’apport de la linguistique, de la psycholinguistique et de la méthode comparative. L’objectif est de parvenir à une définition des signifiés rendant compte des différentes valeurs contextuelles des tiroirs verbaux, en vue d’une application en didactique du FLE. Après un aperçu théorique et méthodologique sur le temps et sur l’aspect (ch. 1 et 2), l’auteure s’attache à examiner les valeurs contextuelles du passé simple et du passé composé (ch. 3), de l’imparfait (ch. 4) et de l’imparfait articulé au présent (ch. 5); l’analyse linguistique de ces tiroirs verbaux lui permet ainsi de parvenir à une nouvelle définition convoquant différents traits définitoires: les traits [-allocutif ] et [autonome] pour le passé simple; les traits [+allocutif ] et [accompli] pour le passé composé; les traits [anaphorique] et [inactuel] pour l’imparfait. L’activation de ces traits fondamentaux est liée à la prise en compte de trois dimensions distinctes: la dimension référentielle, la dimension énonciative, envisagée dans un cadre dialogique et une dimension conceptuelle de traitement textuel qui attribue aux morphèmes verbaux une valeur de marquage de macro-unités textuelles. Elisa Ravazzolo

504

RASSEGNA DI LINGUISTICA FRANCESE

BJÖRG LARSSON, Le sens commun ou la sémantique comme science de l’intersubjectivité humaine, “Langages”, 170, 2008, pp. 28-40 Dans cet article sur le discours et le sens commun, Larsson s’interroge sur la signification du terme éponyme (sens), et affirme que les locuteurs font appel à des catégorisations sémantiques différentes en fonction des domaines de sens. Pour lui, le sens n’est donc pas une propriété du monde mais est une unité mentale intersubjective qui doit être codifiée en signes. Il en découle que le terme ‘sens’ désigne un phénomène ayant une identité et un mode d’existence propre: il repose sur une asymétrique, une dynamique et une mémorisation, qui se découvrent dans les corpus authentiques et les interactions, beaucoup plus que dans les spéculations sur le système linguistique. Jean-Paul Dufiet PIERRE CADIOT, Proverbes, sens commun et communauté de langage, “Langages”, 170, 2008, pp. 79-91 Selon l’auteur, le sens commun propre au proverbe passe par un ensemble de techniques de discours qui se comprennent d’abord à partir d’une conception de la signification. Celle-ci repose sur deux versants: le premier, perceptuel, va du sensible au figural; le second, doxal, se déploie entre l’opinion et la croyance. Cadiot explique ensuite que la ritualisation formulaire du proverbe ne s’assimile pas à celles du cliché et du stéréotype. Son approche génétique remet donc en cause une explication fondée sur le principe du figement, et défend une conception qui repose sur trois notions: la communauté (qui est à la fois affaire de perception, d’éthique et d’esthétique), la modalisation-normativité (qui est le produit d’une diversité d’habitus), les genres (le sens commun se manifeste par sa conformité aux lois des genres). Jean-Paul Dufiet

M ICHELE D E G IOIA , Carrément. Lexique bilingue français-italien des adverbes figés, Aracne, Roma 2008, pp. 466 Se ci si sveglia al canto del gallo, si può tradurre letteralmente in francese au chant du coq? Sì, perché le due lingue hanno lo stesso avverbio idiomatico (o fisso), come dimostra la puntuale indagine condotta da Michele De Gioia su oltre 6000 espressioni avverbiali francesi e italiane. Un avverbio fisso francese privo di corrispondente diretto italiano è proprio quello del titolo, carrément, i cui equivalenti semantici sono decisamente, senza esitare, ecc. I risultati dello studio di De Gioia costituiscono un valido argomento contro lo stereotipo dell’intraducibilità delle forme idiomatiche che non solo sono traducibili, ma lo sono anche con equivalenti semantici aventi stessa funzione grammaticale e stesso carattere di fissità; partendo dal francese, ciò si verifica per l’italiano in 3 casi su 5. Lo studio tassonomico oggetto del volume è un lessico-grammatica comparato. Sia la metodologia sia il corpus francese di partenza sono stati definiti da Maurice Gross. De Gioia riprende la propria classificazione degli avverbi fissi italiani, approfondendo l’aspetto comparatistico in questo lessico bilingue che può essere gestito per il trattamento automatico con i programmi del lessico-grammatica. Interessante ipotesi, l’applicazione glottodidattica per l’apprendimento di espressioni idiomatiche divertenti: essere alto come un soldo di cacio / être haut comme 3 pommes. Per ricerche su questa categoria grammaticale, De Gioia propone uno strumento utilizzabile tous azimuts, a 360 gradi. Mirella Conenna M ARIA C OLOMBO – M ONICA B ARSI ed., Lexicographie et lexicologie historiques du français. Bilan et perspectives, Polimetrica, Milano 2008, pp. 290 Les articles réunis dans le présent volume éclairent divers aspects du lexique français en privilégiant, notamment, les perspectives lexicographique et lexicologique. Deux axes

RASSEGNA DI LINGUISTICA FRANCESE

de recherche y sont privilégiés: l’axe diachronique et l’axe diatopique. Les deux premières contributions de Gilles Roques reflètent ce choix: la première aborde la problématique des régionalismes lexicaux du français, tandis que la deuxième explore les principaux outils lexicographiques pour l’étude du français médiéval. Ensuite, Jean Pruvost parcourt l’histoire personnelle de Littré et de Larousse et prouve à quel point l’atmosphère de l’époque a influencé leurs ouvrages. La perspective diachronique est aussi à la base de trois contributions qui focalisent notamment la lexicographie bilingue italien-français: Nadia Minerva étudie le premier repertoire lexicographique français-italien et italien-français (Lyon, Rigaud, 1583); Nicole Maroger réfléchit aux critiques apportées au dictionnaire bilingue de Veneroni; Jacqueline Lillo explore les multiples éditions du dictionnaire d’Alberti de Villeneuve. Claudio Vinti préfère une démarche lexicologique pour analyser la relation entre le théâtre forain et le patois rustique. En revanche, André Thibault et Paola Puccini privilégient l’axe diatopique: le premier illustre les outils lexicographiques élaborés pour chaque zone de l’espace francophone; la deuxième, à travers l’ouvrage Profession lexicographe de Marie-Eva de Villers, esquisse le portrait du lexicographe dans une perspective anthropologique. Un autre groupe d’articles focalise la relation de la lexicographie et de la lexicologie aux nouvelles technologies: Annick Farina étudie les CD de trois dictionnaires bilingues italienfrançais et souhaite une meilleure exploitation de cet outil; Antonino Velez présente les principaux sites internet où l’on peut consulter dictionnaires, glossaires, banques de données; Geneviève Henrot-Sostero étudie une unité lexicale dans les contextes relevés grâce à internet. De son côté, Marie-Pierre Escoubas-Benveniste décrit les emplois prédicatifs des verbes de mouvement orienté, ce qui pourrait aider à leur traitement lexicographique. Chiara Molinari

505

L’implicite et les écrits de l’entreprise, “Les Cahiers de l’ILCEA”, 9, Grenoble 2007, pp. 221 Questo numero dei “Cahiers de l’ILCEA” pubblicato sotto l’egida del GREMUTS (Groupe de recherche multilingue en traduction spécialisée) all’Università Stendhal di Grenoble, sintetizza i contributi di una giornata seminariale dedicata a “Langues et cultures de spécialité”. Le aree geografico-culturali interessate sono, oltre alla Francia, gli Stati Uniti, l’Italia, la Germania, la Russia. In un’epoca dominata dalle TIC, la comunicazione che emana dalle imprese si moltiplica e si diversifica. I generi professionali oggetto di analisi vanno da quelli formali classici (rapporto annuale di attività, comunicato stampa) a quelli promozionali (campagne promozionali, pubblicità) ad altri più nuovi (mission statement, code of commit-tment) o presi dalla letteratura e messi al servizio dell’impresa (publifiction, autobiografia). Benché i settori dell’attività economica osservati coprano ambiti molto diversi (l’automobile, la finanza, lo sport , le banche, l’agro-alimentare ecc.), la coerenza epistemologica si costruisce attorno allo studio dell’implicito nella varietà degli scritti analizzati tramite approcci riconducibili all’analisi del discorso, alla pragmatica, all’analisi dei generi.

Emerge l’interesse epistemologico e pedagogico degli autori, ricercatori-insegnanti in curricola professionalizzanti, come pure l’interesse metodologico di un approccio interdisciplinare e interculturale che si colloca all’intersezione tra due vasti ambiti di ricerca: le lingue e culture di specialità da un lato e, dall’altro, le scienze sociali. Enrica Galazzi ISABEL DESMET, Terminologie, culture et société. Éléments pour une théorie variationniste de la terminologie et des langues de spécialité, “Cahiers du Rifal”, 26, 2007, pp. 3-13 Dopo aver considerato i caratteri degli approcci tradizionali – monosemia, univocità, opposizione terminologia/lessicologia –, moderni (analisi del testo di specialità, del contesto e della fraseologia) e contemporanei

506

RASSEGNA DI LINGUISTICA FRANCESE

(aspetti sociolinguistici, culturali, pragmatici), l’A. propone per la descrizione della terminologia e delle lingue di specialità un quadro teorico che integri la variazione in tutte le sue dimensioni linguistiche ed extralinguistiche. Principio basilare è la presa in considerazione della varietà dei discorsi di specialità, dei generi testuali e del contesto. L’analisi terminologica deve fondarsi sui livelli dell’analisi linguistica, di cui la pragmatica costituisce la base in quanto il ruolo degli interlocutori, del contesto e dell’ambiente spazio-temporale sono determinanti per la variazione. Pierangelo Bulgari MICHÈLE LENOBLE-PINSON, Traits communs et traits particuliers de deux banques terminologiques, l’une à Paris, l’autre à Bruxelles, “Cahiers du Rifal”, 26, 2007, pp. 36-47 L’A. presenta i tratti comuni e le peculiarità degli organismi terminologici francesi e belgi. La politica linguistica francese si organizza su tre livelli: le commissioni ministeriali per i lessici specialistici, la Commissione generale di terminologia e neologia, l’Académie française. La Comunità francese belga si avvale di una commissione di terminologia e di reti di esperti. L’A. sottolinea alcune procedure di creazione lessicale rilevate dall’analisi del corpus dei termini ufficiali: neologia semantica (bloc-notes, avant-première), neologia morfologica (francesizzazioni: bogue, hétérolyse; composti: banque à domicile, banque de données; formazioni neologiche scientifiche), morfologia lessicale (suffissazione). Pierangelo Bulgari LOUIS-JEAN ROUSSEAU, Élaboration et mise en œuvre des politiques linguistiques, “Cahiers du Rifal”, 26, 2007, pp. 58-71 L’articolo definisce i concetti di “politica linguistica” e di “aménagement linguistico”, delinea le componenti relative allo statuto ed al codice delle lingue, traccia una tassonomia ed i fondamenti politici e giuridici, descrive le esigenze metodologiche, i limiti e gli ostacoli, ne considera la valutazione ed i necessari continui aggiustamenti, traendo esempi dal-

l’area francofona (in particolare riferendosi alla riforma dell’ortografia, a lavori terminologici e di modernizzazione del lessico). Il contributo solleva il problema della rilevanza del prestito terminologico (perdita di campi per una lingua causata dall’assenza di sviluppo terminologico, ricorso all’orientation de l’usage), la questione dell’implantation e dell’utilizzo dei nuovi termini (con la nascita della terminométrie). Pierangelo Bulgari BÉATRICE FRAENKEL, Comment tenir un registre?, “Langage et société”, 124, 2008, pp. 59-71. À travers l’analyse de déclarations de naissance, B. Fraenkel se demande en quoi de tels écrits contribuent à la mise en œuvre d’un genre discursif et s’il est possible de séparer l’élaboration d’un objet (le registre d’État civil) de son énonciation. L’auteure remarque que ces écrits extrêmement formalisés peuvent être caractérisés comme “genres professionnels”, mais qu’ils ne s’expliquent historiquement qu’en tant qu’enjeux politiques. Ainsi, la rédaction d’un registre correspond -elle autant à la création d’un objet qu’au maintien d’une institution de la République et des normes graphiques qui l’accompagnent. Au-delà de l’acte de naissance, l’auteure s’interroge sur le statut des objets écrits en général. Françoise Favart MARIE-NOËLLE GUILLOT, Oral et illusion d’oral: indices d’oralité dans les sous-titres de dialogues de film, “Meta”, LII, 2007, 2 pp. 239-259 Dans son article, Marie-Noëlle Guillot analyse les dialogues et sous-titres correspondants du film 37°2 le matin (Beineix 1986) par rapport à l’oral spontané et souligne à quel point les contraintes du sous-titrage stimulent le potentiel à produire l’expérience de l’oralité. En se basant sur la théorie du mode de Fowler, et sans nier les dangers de l’économie des sous-titres, elle démontre que la “perte” est relative et que, pour ce qui concerne la production de l’expérience de l’oralité en soi, “le moins ne paraît pas nécessairement l’ennemi du bien”. Yannick Preumont

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008)507-511

RASSEGNA DI LINGUISTICA INGLESE A CURA DI MARGHERITA ULRYCH

J UCKER A NDREAS H. – I RMA T AAVITSAINEN (ed.), Speech Acts in the History of English, John Benjamins, Amsterdam/Philadelphia 2008, pp. 318 The aim of this volume is to sketch a speech act history of the English language. The volume contains a series of innovative studies which are divided into three different sections and provides accounts of specific speech acts at particular stages in the development of English. Part One contains significant contributions on directives and commissives (speech acts such as requests, commands and promises). Recent research has shown that older stages of English used different strategies for polite directives. Directives that would appear impolite or even rude today seem to have been common in the past. Kohnen’s paper focuses on typical manifestations of directive speech acts in Old English (his data are taken from the Helsinki Corpus – the first electronic corpus on historical principles – and the Dictionary of Old English Corpus). Pakkala-Weckström explores the speech act of promising in Middle English and in particular the different kinds of promises in their varying contexts in Chaucer’s Canterbury Tales. Archer develops a scale of directness for “Early Modern English” requests, considering variation across a range of social parameters, and comparing the results with research on present-day English. Busse first compiles an inventory of linguistic forms which were available to Shakespeare and his contemporaries to carry out directive speech acts and then assesses their precise discourse function in terms of their coercive force and their politeness, while Del Lungo analyses requests and commitments, frequently performed by nineteenth century commercial letter writers.

Part Two deals with speech acts such as greetings, compliments and apologies. Grzega shows that in Old English attention getters and wishes for good health were particularly common while in Middle English they were replaced by inquiries about the health of the addressee and wishes for a good time. Taavitsainen and Jucker outline the history of compliments in English while the last paper deals with apologies. The last section, Part Three, focuses on research techniques that are used to extract speech acts from computer-readable corpora. This part has been added in the hope that other scholars may profit from the problems encountered by the authors of these papers in the process of carrying out further corpuslinguistic speech act research in historical corpora. Erica Doppiati VICTORINA GONZÁLEZ-DÍAZ, Recent Developments in English Intensifiers: The case of ‘very much’, “English Language and Linguistics”, 12, 2008, 2, pp. 221-243 L’articolo prende in esame la semantica degli avverbi complessi e composti come very much, heretofore, anyway. Partendo dalla constatazione che tali costrutti avverbiali non hanno ricevuto sufficiente attenzione nella letteratura più recente, l’Autrice si concentra in particolar modo sulla semantica di very much. Mentre alcuni studi recenti lo descrivono come un avverbio clause-internal che modifica costituenti del sintagma, la letteratura precedente attribuiva agli avverbi nella lingua inglese uno scope più ampio. Attraverso l’analisi diacronica di un corpus contenente testi dal 1500 ai giorni nostri, González-Díaz mostra alcuni contesti nei quali il comportamento di very much contraddice le descrizioni più recenti. Le conclusioni dello studio

508

RASSEGNA DI LINGUISTICA INGLESE

suggeriscono come, a partire dall’ottocento circa, questo avverbio abbia sviluppato funzioni di modificatore di enunciato, collocandosi così sul continuum descritto da Traugott (1995): Internal Adverb > Sentence Adverb > Discourse Particle. Sarah Bigi

pragmatic markers come well differiscono chiaramente nel significato dai logical markers come but. In altre parole, il loro significato è interamente pragmatico, poiché non si riferiscono a contenuti proposizionali ma a funzioni modali del discorso. Sarah Bigi

B ELÉN M ÉNDEZ -N AYA , On the History of ‘downright’, “English Language and Linguistics”, 2008, 12, 2008, 2 pp. 267-287

A NNE -M ARIE S IMON -V ANDENBERGEN , Almost certainly and most definitely: degree modifiers and epistemic stance, “Journal of Pragmatics”, 40, 2008, pp. 1521-1542

Basandosi su una varietà di corpora e sul database di citazioni dell’OED, lo studio presenta un’analisi diacronica sull’origine e l’evoluzione della funzione di modificazione dell’intensificatore downright, sia esso in funzione di avverbio (it’s downright rude) che di aggettivo (downright nonsense). L’evoluzione di downright diventa anche un chiaro esempio dell’interazione tra processi di lessicalizzazione e di grammaticalizzazione nell’evoluzione di un singolo elemento. L’articolo sottolinea inoltre il ruolo del contesto e dell’inferenza che si rivelano determinanti nel processo di cambiamento semantico. Infine vengono messe in luce due diverse linee lungo le quali si sviluppano gli intensificatori: da una parte si osserva il passaggio da adjuncts a degree modifiers, dall’altra quello da descriptive adjective, ad affective adjective a intensifier. Sarah Bigi MARIA-JOSEP CUENCA, Pragmatic markers in contrast: The case of well, “Journal of Pragmatics”, 40, 2008, pp. 1373-1391. L’articolo prende in oggetto il discourse marker well, abbondantemente studiato, ma il cui significato a volte rimane ancora vago. La domanda posta dalla presente ricerca è quanto l’analisi contrastiva di un pragmatic marker può contribuire a metterne in luce il significato e le funzioni. Per rispondere a questa domanda, l’Autrice analizza le occorrenze di well nel film Four weddings and a funeral paragonate con i loro corrispondenti nelle versioni doppiate in spagnolo e in catalano. I risultati dell’analisi mostrano che i

La ricerca presentata in questo articolo prende in esame gli avverbi certainly e definitely. La loro somiglianza semantico-pragmatica, già descritta da dizionari, dati provenienti dai corpora e da questionari, costituisce il punto di partenza dal quale l’Autrice muove per sviluppare alcune considerazioni volte a precisare ulteriormente la descrizione di questi avverbi. In particolare, è posta attenzione alla preferenzialità degli avverbi considerati rispetto ai loro intensificatori di grado. Basandosi sul British National Corpus, l’Autrice mostra che mentre almost e most sono utilizzati per modificare entrambi gli avverbi, esiste una netta prevalenza delle combinazioni almost certainly e most definitely. L’articolo sostiene che tale preferenzialità è spiegabile a partire dalle caratteristiche semantico-pragmatiche dei due avverbi. Sarah Bigi ULLA CONNOR, – ED NAGELHOUT – WILLIAM R OZYCKI ed., Contrastive Rhetoric Reaching to intercultural rhetoric, John Benjamins, Amsterdam 2008, pp. 324 The volume deals with variations in the style and rhetorical structure of texts by non-native speakers of English and is divided into four sections comprising 14 chapters. In section I, “Current state of contrastive rhetoric”, Xiaoming Li (Chapter 1) focuses on the insights offered by the discipline to the teaching of L2 writing, while Ana I. Moreno (Chapter 2) highlights the importance of fully comparable corpora, which should

RASSEGNA DI LINGUISTICA INGLESE

be similar in all contextual factors and serve the purpose of carrying out reliable cross-cultural research. Section II, “Contrastive corpus studies in specific genres” (Chapters 3-8), with contributions by Annelie Ädel, Haiyeng Feng , Maria Loukianenko Wolfe, Chin-Sook Pak and Rebecca Acevedo, Lorena Suárez and Ana Moreno, Wei Wang, deals with student essays, grant proposals, sales/product promotion letters, newspaper editorials, academic book reviews of literature and newspaper commentaries. Diverse features such as metadiscourse, rhetorical moves, linguistic devices linked to Hofstede’s cultural dimensions, sentence and paragraph length and complexity, attribution to sources and personal voice are investigated in a variety of languages (American, British, Australian and nonnative English, Chinese, Russian, Spanish) using both quantitative and qualitative methods. Section III, “Contrastive rhetoric and the teaching of ESL/EFL writing”, comprising Chapters 9 to 12, written respectively by Virginia Lo Castro, Kara McBride, Xiaoye You and Joel Bloch, investigates Mexican students' writing practices in Spanish and in English, problems by Chilean university students when reading English websites, theme treatment in Chinese school essay writing and conceptions of plagiarism. The studies also show how methods such as questionnaires, participant observation, surveys, interviews and think-aloud protocols may be used in research in contrastive rhetoric. Section IV, “Future Directions”, consists of Chapter 13, by Paul K. Matsuda and Dwight Atkinson, and Chapter 14, by Ulla Connor. Chapter 13 reports a conversation in which the authors discuss the future of intercultural rhetoric, while Chapter 14 outlines the history of the discipline and argues for the need to encompass the study of the writing contexts and the social practices connected to writing. To respond to these broader aims the term “intercultural rhetoric” is suggested. Costanza Cucchi

509

ALISON IRVINE, Contrast and convergence in Standard Jamaican English: the phonological architecture of the standard in an ideologically bidialectal community, “World Englishes”, 27, 2008, 1, pp. 9-25 L’articolo indaga l’intersezione tra Creolo e Standard Jamaican English (SJE) ed in particolare esamina gli aspetti fonologici che caratterizzano e limitano i confini tra le due varietà. Nel contesto sociale giamaicano, il bilinguismo è visto come segnale di prestigio sociale. L’analisi, condotta intervistando soggetti che utilizzano regolarmente sul luogo di lavoro SJE, dimostra come fenomeni quali h-dropping, voiceless th-stopping ed uso di /ɔ/, siano percepiti come tipicamente creoli e come tali stigmatizzati dai parlanti; allo stesso tempo si dimostra come esistano elementi creoli in SJE che sono stati assimilati e resi accettabili anche in un ambito formale; è il caso di /kja:/ in posizione iniziale al posto di /ka/, della dittongazione di /o/ ed /e/, della variabilità nella resa di <-nt> con valore morfologico (per esempio in can’t, don’t). I dati raccolti evidenziano come nel ‘continuum’ linguistico jamaicano una definizione di standard dipenda dall’atteggiamento psicologico dei singoli parlanti e dal condizionamento sociale che ad esso si accompagna: i dati sembrano non permettere una distinzione netta e definita almeno nell’ambito del campione esaminato. Silvia Pireddu XIAOYE YOU, Rhetorical strategies, electronic media, and China English, “ World Englishes”, 27, 2008, 2, pp. 233-249 To date an inference model has been adopted to carry out research on rhetorical strategies employed in China English. Yet such a model assumes that China English presents linguistic and rhetorical features which exclusively originate from Chinese strategies and does not shed light on the growing semiotic potential of this new emerging variety of English. Therefore the author of the present article suggests an alternative model, a con-

510

RASSEGNA DI LINGUISTICA INGLESE

text-centred model (based on Halliday’s systemic linguistics) and starts to examine China English as it is used in particular contexts and for particular audiences. In his study on how Chinese youths use English in a bulletin board forum to foster an online community, You divides the discussion threads into four sets of contexts and analyses the rhetorical strategies that the forum members use in each set of contexts. The author maintains that Chinese youths develop different rhetorical strategies according to different contexts of situation and these strategies are somewhat different from those identified in other contexts by previous scholars. Context should therefore be the main variable in verifying the rhetorical potential of China English. Erica Doppiati AMPARO ALCINA, Translation technologies. Scope, tools and resources, “Target”, 20, 2008, 1, pp. 79-102 Alcina’s paper discusses translation technologies as a new field of interdisciplinary study that deals with the investigation of “the design and adaptation of strategies, tools and technological resources” (Alcina 2008: 90) that facilitate the translator’s job. Drawing on her review of existing literature as well as on her own experience in researching and teaching translation technologies, she illustrates her own proposal for a coherent internal structuring of the subject. She envisages this as consisting of five blocs: the translator’s computer equipment; communication and documentation tools; text edition and desktop publishing ; language tools and resources; translation tools. Although a number of questions remain to be defined, Alcina’s analysis is a welcome contribution to the “disciplinary, academic and professional consolidation and development” (Alcina 2008: 99) of the emerging discipline of translation technologies. Monica Pedrola

JEREMY MUNDAY ed., Translation as Intervention, Continuum International Publishing Group, London/New York 2007, pp. 165 This book is a collection of papers exploring a number of aspects related to the notion of translation as intervention. Translation is envisaged as a politically and socially active phenomenon which impacts in a dynamic way on a variety of communicative events. A range of situations and case studies are taken into consideration, including the European Union, translated literature in India and China, Arabic historical texts and court interpreting in multilingual South Africa. The result is a comprehensive analysis of the “intervenience” power of translators at different levels: their identity as mediating beings; their interpretation of the text; their intervention at the level of text, discourse and language functions; the role of technolog y and the management of related work practices and finally, the intervention in the translation process of participants other than the translators themselves. The contributions included in the book come from a range of cultures and disciplines, which reflects the way in which the whole field of Translation Studies has recently expanded, interfacing with other disciplines and broadening its horizons to embrace nonWestern traditions. This and the quality of the contributions included make “Translation as Intervention” an engaging collection of great interest to students and academics researching in this area. Monica Pedrola J ULIA H ÜTTNER , The genre(s) of student writing: developing writing models, “International Journal of Applied Linguistics”, 18, 2008, 2, pp. 146-165 L’articolo discute dell’utilità didattica dei modelli di analisi e di valutazione dell’EAP (English for Academic Purposes), evidenziando l’opportunità di utilizzare modelli diversi da quelli attualmente considerati. In altre parole le esigenze degli studenti chiama-

RASSEGNA DI LINGUISTICA INGLESE

ti a produrre testi scritti in ambito accademico non coincidono esattamente con i dati e le tipologie indagate dalla comunità scientifica a partire dagli studi di Swales. Occorre piuttosto distinguere tra l’inglese accademico inteso come lingua franca a livello scientifico e nell’ambito delle varie comunità di ricerca e gli obiettivi linguistici, i bisogni e gli scopi degli studenti. L’autrice, pensando alla situazione europea e svolgendo la propria indagine con un gruppo di studenti dell’Università di Vienna, considera i seguenti elementi: l’inglese può essere un mezzo didattico e venire utilizzato per prendere appunti o scrivere saggi ma la tipologia e la forma di questi è diversa da quella in uso nelle università anglosassoni e non coincide con le strategie retoriche dell’academic paper. La funzione didattica del saggio o meglio della dissertazione pone problemi di eterogeneità stilistica, linguistica e testuale che necessitano di un controllo da parte dei docenti. Siamo infatti in contesto L2 dove i modelli testuali della L1 interagiscono con gli usi linguistici EAP. Partendo dunque dai saggi prodotti dagli studenti si propone un modello d’analisi degli stessi che possa evidenziarne le caratteristiche comuni fornendo una base per lo sviluppo di materiale didattico che aiuti gli studenti ad esprimere i propri obiettivi comunicativi. L’articolo evidenzia inoltre i punti di contatto e le principali differenze tra academic papers e gli student’s essays sia a livello di intenzione comunicativa sia a livello lessico-grammaticale e delle strutture testuali. Silvia Pireddu “Studi Italiani di Linguistica Teorica e Applicata”, XXXVII, 1, 2008 Il numero monografico della rivista prende in esame i più recenti sviluppi della Linguistica Inglese in ogni suo ramo. Secondo le parole usate dal curatore, Giovanni Iamartino, nell’Introduzione, questo numero si propone “al lettore come una bussola per orientarsi nel mare magnum della recente produzione bibliografica in linguistica inglese” (p. 3). Il volume si apre con il contributo di

511

Cesare Gagliardi su fonetica e fonologia; prosegue poi con le voci di Antonio Bertacca sulla morfologia, Cristiano Moccias sulla sintassi e Stefania Nuccorini che insieme a Laura Pinnavaia affrontano la produzione bibliografica relativa al lessico. Marcella Bertuccelli Papi e Annalisa Baicchi collaborano nella presentazione degli sviluppi nel campo della semantica e della pragmatica, mentre Silvia Bruti e Lavinia Merlini Barbaresi danno il loro contributo sulla linguistica del testo e l’analisi del discorso. Maria Luisa Maggioni e Paola Tornaghi, a partire dalle radici storiche dei diversi Englishes offrono una panoramica sulle varietà dell’inglese nel mondo, con uno sguardo verso il suo sviluppo futuro. Giuseppina Cortese e Martin Solly pongono l’attenzione sull’inglese per scopi professionali e accademici mentre Marina Dossena e Richard Dury forniscono un’ampia descrizione di studi di linguistica diacronica. La corpus lingustics con i suoi metodi innovativi è trattata da Elena Tognini-Bonelli, Silvia Cacchiani e Giuliana Diani, mentre i translation studies, visti anche nella loro interazione con la corpus linguistics, sono presentati e descritti da Margherita Ulrych e Simona Anselmi. L’ultimo contributo è quello di Luciana Pedrazzini, che riprende gli sviluppi della didattica della lingua inglese. Tutti i capitoli presentano la stessa organizzazione: una introduzione contenente i riferimenti concettuali e bibliografici imprescindibili per ciascun ambito di ricerca; segue un’analisi critico-bibliografica della produzione scientifica anglosassone, che però non trascura i contributi dal mondo italiano. La sezione conclusiva si articola in una utile rassegna delle principali riviste e conferenze del settore, seguita da un paragrafo finale dedicato alle problematiche aperte e a possibili linee di ricerca futura. Sarah Bigi

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 513-515

RASSEGNA DI LINGUISTICA RUSSA A CURA DI ANNA BONOLA

AA.VV., Sovremennyj russkij jazyk. Aktivnye processy na rubeže XX-XXI vekov [Il russo contemporaneo. Processi attivi a cavallo tra il XX e il XXI secolo], L.P. KRYSIN ed., Jazyki slavjanskich kul’tur, Moskva 2008, pp. 709 O.P. Ermakova, R.I. Rozina, M.Ja Glovinskaja, L.L. Kasatkin, R .F Kasatkina, S.M. Kuz’mina, E.I. Golanova, E.V. Kakorina, E.A. Zemskaja: sono i nomi degli autori di questo volume, insigni studiosi che negli ultimi ’20-’30 anni si sono distinti per aver monitorato i numerosi cambiamenti avvenuti nella lingua russa e che, a distanza di poco più di un decennio dal noto Russkij jazyk konca XX stoletija (1985-1995), curato da E.A. Zemskaja, ripropongono un corposo aggiornamento sullo stato del russo contemporaneo a partire dalla metà degli anni Novanta. Le prime quattro parti del volume indagano i processi semantico-lessicali, grammaticali, fonetico-ortografici e comunicativi attualmente attivi nella lingua russa; la quinta parte si concentra invece sul russo della diaspora. Si rilevano i seguenti fenomeni: ingresso dei gerghi nel russo standard e allentamento della norma, diffusione di elementi linguistici marcati dal punto di vista sociale e professionale (i plurali in –à, l’ampio uso della preposizione po). Se negli anni ’50-’60 tra le lingue professionali dominava quella dei fisici, soprattutto nucleari, ora prevale l’informatica, oggetto della creatività linguistica dei parlanti russi anche al di fuori del settore (Emelja invece di emajl, Trupopaskal’ per Turbo Pascal…). Attraverso processi di metaforizzazione si realizza un intenso passaggio di termini informatici nella lingua standard e viceversa (komp’juter zavisaet, otkazyvaet…). Nelle varietà territoriali e regionali della lingua russa si registra la diminuzione dei

dialettismi e l’ingresso di elementi della lingua standard, conseguenza della diversificazione sociale che si sta accentuando anche nelle campagne in seguito ai maggiori contatti con il mondo urbano. Nelle parlate urbane, accanto al ‘vecchio’ prostoreč’e - ricco di dialettismi, parlato dalle persone anziane con basso livello di istruzione - si sta diffondendo una forma “giovane”, con parlanti di età inferiore ai sessant’anni, di istruzione medio-bassa e senza padronanza delle norme della lingua standard; a differenza della vecchia forma, è privo di dialettismi e ricco, invece, di termini gergali. Infine, si sta formando un gergo di uso comune basato sui gerghi dei lager e delle prigioni, già sviluppatisi in epoca sovietica. Nella morfologia si nota la tendenza alla semplificazione del sistema dei casi e all’analitismo. Numerosi anche i cambiamenti nella semantica, soprattutto nei rapporti lessicali paradigmatici (creazione di nuove coppie di antonimi e sinonimi). In questo clima di rinnovamento cresce infine sempre più la coscienza linguistica dei parlanti e la loro partecipazione al dibattito pubblico sui cambiamenti in corso nella lingua russa contemporanea. Anna Bonola AA.VV. Nacional’nyj corpus russkogo jazyka i problemy gumanitarnogo obrazovanija [Corpus Nazionale delle lingua russa e problemi della formazione umanistica], N.R. DOBRUŠINA ed., Gosudarstvennyj universitet – Vysšaja škola ekonomiki, Moskva 2007, pp. 237. A due anni dall’inaugurazione del libero accesso al Corpus Nazionale della lingua russa, escono gli atti della conferenza in cui, nell’aprile del 2007, si è voluto riflettere sulle applicazioni didattiche di questo nuovo strumento. La prima parte illustra l’utilità

514

RASSEGNA DI LINGUISTICA RUSSA

metodologica del corpus, che permette di modellizzare i processi di ricerca e rende l’apprendimento un cammino euristico più personale. Nello studio del russo come lingua straniera il corpus facilita la creazione di esercizi orientati ai testi, permette di chiarire le condizioni d’uso di forme linguistiche concorrenti (bit’ v – bit’ po; začem – počemu) e, più in generale, di avere contatto con l’uso linguistico contemporaneo. Tre gli interventi di studiosi dall’Italia: F. Fici tratta l’uso del corpus per la didattica della traduzione del gerundio russo in italiano; S. Slavkova scrive sull’apprendimento dei verbi di supporto appoggiandosi al concetto di funzione lessicale; F. Biagini fornisce un apparato di esercizi costruito con il corpus nazionale e riguardante la concorrenza fra začem e počemu. Nell’insegnamento del russo come prima lingua il corpus è stato utilizzato per la didattica di diversi aspetti del sistema linguistico: la derivazione lessicale, le categorie del lessico, i linguaggi di specialità, e l’analisi testuale. Anna Bonola E. FORTEJN, Polisemia imperativa v russkom jazyke [La polisemia dell’imperativo nella lingua russa], in “Voprosy jazykoznanija”, I, 2008, pp. 23-45 Il saggio analizza la semantica dell’imperativo nella lingua russa per mostrare la complessa polisemia che si cela dietro alla medesima veste morfologica. Accanto al valore proprio dell’imperativo (direktivnyj imperativ), usato per esortare l’interlocutore a compiere l’azione espressa dal verbo, E. Fortejn individua: l’imperativo necessitativo (dolženstvovatel’nyj), narrativo (povestvovatel’nyj), ottativo (želatel’nyj), condizionale (uslovnyj) e concessivo (ustupitel’nyj imperativ). Si evidenziano anche le caratteristiche morfosintattiche di ciascun tipo di imperativo (la disposizione del soggetto e del predicato nella frase, la persona, l’aspetto verbale e il tempo, la presenza dei suffissi – ka o – te) e le particolarità stilistiche (il povestvovatel’nyj imperativ è tipico del folclore e della lingua parlata, mentre l’optativnyj imperativ, produt-

tivo fino al XIX secolo, è vivo oggi solo nelle frasi idiomatiche). L’invariante semantica che accomuna i sei tipi di imperativo è il tratto dell’‘impulso’ a compiere l’azione o a rappresentarla, esortando l’interlocutore a parteciparvi. Nell’imperativo direttivo tale impulso è diretto ed esplicito; negli usi indiretti, invece, dominano altri tratti semantici specifici (le sfumature ‘soggettivo-modali’). Gli usi indiretti dell’imperativo vengono illustrati con numerosi esempi tratti da opere letterarie dell’Ottocento come dalla stampa degli ultimi anni. Maria Versace T.M. N IKOLAEVA , Neparadigmatičeskaja lingvistika (istoria “bluždajuščich častic”) [Linguistica non paradigmatica (storia delle “particelle erranti”)], Jazyki slavjanskich kul’tur, Moskva 2008, pp. 375. Le particelle con valore deittico-comunicativo sono patrimonio comune di tutte le lingue slave e fenomeno imponente, ma spesso ignorato dai linguisti. Si tratta di una strana asimmetria fra lingua e linguistica i cui motivi sono spiegati da T.M. Nikolaeva nel primo capitolo della sua monografia (le premesse teoriche). Le particelle - negli studi diacronici considerate resti di una classe antica e poco nota - nelle tassonomie vengono accomunate a congiunzioni, pronomi o alle parole discorsive ecc. T.M. Nikolaeva per esse sceglie il termine partikuly, evitando così una troppo stretta identificazione con le časticy. Nel secondo capitolo le particelle vengono confrontate con diverse parti del discorso e con i clitici; l’autrice si concentra sugli elementi *-s e *-t, centrali per la formazione dei paradigmi nelle lingue indoeuropee. La semantica delle particelle slave varia notevolmente a seconda che siano usate come elementi singoli (primarie: a, i, no, že, li,) o combinate fra di loro (li, per esempio, indica l’interrogativa indiretta, ma dà luogo a composti come libo, ali, toli, uželi ecc. che hanno tutt’altro significato). Delle particelle primarie e dei cluster che esse formano, dando luogo al processo di

RASSEGNA DI LINGUISTICA RUSSA

grammaticalizzazione, si occupa il terzo capitolo. Il patrimonio iniziale delle particelle slave, ristretto e con semantica indefinita, ma sempre riferita all’opposizione determinatoindeterminato, creerebbe, grazie a un metodo caleidoscopico (prima metafora di Nikolaeva), una elevata quantità di combinazioni dalla semantica più precisa. Le particelle originarie sono dunque un meccano primordiale (seconda metafora) delle lingue slave, attivo all’incirca fino al XVI secolo. L’agglutinazione (una sorta di ‘minisintassi’) sarebbe così il principio costruttivo dei paradigmi antico slavi, sorti dalla combinazione delle particelle con gli elementi più vari (è questo per l’autrice il senso dell’aggettivo “errante” riferito alle particelle). La monografia si conclude con utili appendici: l’elenco delle particelle primarie comuni a tutte le lingue slave e di quelle proprie di ciascuna lingua slava; l’elenco dei pronomi slavi composti da particelle, dei cluster di 10 particelle russe; infine un glossario delle particelle slave. Anna Bonola A.A. Z ALIZNJAK , Drevnerusskie enklitiki [Enclitici antico russi], Jazyki slavjanskich kul’tur, Moskva 2008, pp. 276 Prestigiosa premessa e insieme complemento al volume di T.M. Nikolaeva sopra presentato è questa monografia del grande filologo russo che lamenta la scarsità degli studi sugli enclitici, forse dovuta al fatto che del ricco apparato di enclitici antico slavi non è quasi rimasta traccia nelle lingue slave orientali, a differenza di quelle occidentali e meridionali (in particolare il serbo e il croato). Ma la scoperta delle gramoty su corteccia di Novgorod ha dato nuovo impulso alla ricerca su questo tema e dalla loro analisi, come di altri numerosi documenti antico russi, Zaliznjak giunge a una serie di importanti conclusioni che confermano le intuizioni contenute nel pionieristico intervento di R. Jakobson alla conferenza dei linguisti del 1933 Les enclitiques slaves (1935).

515

Gli enclitici antico russi comprendono le particelle že, li, bo, ti, by, i pronomi al caso dativo mi, ti, si, mja, tja, sja, ny, vy (successivamente anche le rispettive forme accusative), e le forme verbali Esm’, esi (che assumeranno poi la funzione di copula). La loro disposizione avviene secondo la seguente gerarchia di ranghi: 1. že 2. li 3. bo 4. ti 5. by, bych’, bychom’, byša 6. pronomi in caso dativo 7. pronomi in caso accusativo 8. Esm’, esi. La ricostruzione del preciso meccanismo degli enclitici antico russi, la cui posizione rispetta la legge di Wackernagel, si oppone alla tesi finora accettata che l’ordine delle parole in russo sarebbe sempre stato libero, e permette di correggere numerosi errori interpretativi dei documenti antico russi. Rilevante è il concetto di “barriera ritmico-sintattica”, un confine che distingue l’inizio della frase dalla parte rimanente e che permette di riconfermare la validità della legge di Wakernagel anche là dove sembrava essere contraddetta. L’autore segue l’evoluzione degli enclitici dal XI al XVI secolo, durante la quale essi sono lentamente scomparsi ad eccezione di -sja, diventato parola autonoma e suffisso verbale. Nel russo moderno restano solo že, li e by, che per posizione nella frase rispettano in parte ancora le antiche leggi. Anna Bonola

L’ANALISI LINGUISTICA E LETTERARIA XVI (2008) 517-521

RASSEGNA DI LINGUISTICA TEDESCA A CURA DI GIOVANNI GOBBER E FEDERICA MISSAGLIA

HILTRAUD CASPER-HEHNE, Deutsch-amerikanische Alltagskommunikation. Zur Beziehungsarbeit in interkulturellen Gesprächen, Niemeyer, Tübingen 2006 (Reihe Germanistische Linguistik, 265), pp. XIV+542 In dieser Monographie wird das interkulturelle Gespräch als ein kommunikatives Ereignis verstanden, das neues Verhalten außerhalb der Ausgangskulturen der Ge-sprächsteilnehmer schafft. Interkulturelle Unterschiede werden vor allem dort erkannt, wo Missverständnisse entstehen, die auf unterschiedliche Kulturen der Gesprächspartner zurückgeführt werden können. Auf der Basis eines reichen empirischen Materials untersucht die Autorin die Beziehungsarbeit in interkulturellen Gesprächen zwischen deutschen Muttersprachlern und US/amerikanischen Fremdsprachenlernern. Im Brennpunkt der Analyse stehen drei Gesprächssorten: Erstkontaktgespräche, Gespräche mit Anweisungscharakter und schließlich die sog. interkulturellen diskursiven Gespräche, in denen Geltungen und Normen thematisiert und problematisiert werden. Bei Erstkontaktgesprächen geht es um Herstellung , Aufrechterhaltung und Stabilisierung bzw. Veränderung von sozialen Beziehungen in Form von Kontakten. In solchen Gesprächen erscheint oft die narrative Dimension, die für sich keine Gesichtsbedrohung impliziert. Gespräche mit Anweisungscharakter und interkulturelle diskursive Gespräche sind durch direktive Handlungen gekennzeichnet, die gesichtsbedrohend wirken. Erstere bedrohen das positive face, da sie den Freiraum einer Person einschränken. Das negative face wird durch letztere insofern bedroht, als die Wertschätzung einer Person betroffen wird. Aus der Untersuchung geht hervor, dass US-amerikanische Fremdsprachenlerner

sowohl das positive als auch das negantive face ihrer Gesprächspartner unterstützen und schützen. Deutsche Muttersprachler dagegen zeigen eine Tendenz zur Verletzung sowohl des positive als auch des negative face. Die Kommunikationskultur der anglophonen US-amerikanischen Studierenden wird somit als eine “Kultur der Wertschätzung” bezeichnet. ‘Kultur’ wird hier von einem soziologischen Standpunkt betrachtet, wobei Goffmans Erklärungsmodell eine bedeutende Rolle zugeordnet wird. Darüber hinaus wird die Ansicht vertreten, wonach Werte an Kulturen gebunden sind und als Ergebnis vom Diskurs über die Erfahrung aufgefasst werden. Erst im Diskurs, d.h. in der Kultur würden Werte entstehen. Die Grundfragen menschlicher Existenz, die in den Philosophien thematisiert werden, wären folglich als kulturelle Artefakte anzusehen (wären Menschenrechte selbst kulturell bedingt?). Damit ist keine anthropologische Fundierung des Kulturbegriffs möglich, die zum Verständnis und zur Beurteilung einer echten interkulturellen Kommunikation beitragen könnte. Giovanni Gobber CARSTEN REINEMANN – JANA HUISMANN, Beziehen sich Medien immer mehr auf Medien? – Dimensionen, Belege, Erklärungen, „Publizistik“, LII, 2007/4, pp. 465-484 Die These der wachsenden Selbstbezüglichkeit der Medien gilt in der Kommunikationswissenschaft weitgehend als akzeptiert. Reinemann und Huismann gehen der Frage nach, ob sie sich wissenschaftlich belegen lässt. Dabei stellen die Autoren ein Defizit an empirischen Längsschnittstudien fest sowie eine unzureichende theoretische Durchdringung des Konzepts der Selbst- bzw. Medienreferentialität. Sie skizzieren die verschiede-

518

RASSEGNA DI LINGUISTICA TEDESCA

nen Dimensionen des Phänomens und grenzen aneinander orientierte Handlungen der Medienakteure von aufeinander bezogenen Medieninhalten und explizite (z.B. Medienzitate) von impliziten Bezügen (z.B. AgendaSetting) ab. Nach dem strukturell-individualistischen Ansatz begründet sich eine stärkere Medienreferentialität in der aktuellen Berichterstattung mit dem Wandel der Rahmenbedingungen journalistischen Handelns (Expansion des Medienangebots, Beschleunigung der Nachrichtenzyklen etc.). Zeit-, Kosten- und Rechtfertigungsdruck bei Publikationsentscheidungen führten zur Reduzierung eigener Recherche und zur Orientierung an anderen Medien. Der Handlungsalltag der Journalisten sollte daher empirisch untersucht und anschließend die Folgen der zunehmenden journalistischen Koorientierung (z.B. Mainstreaming) analysiert werden. Jan Henschel P HILOMEN S CHÖNHAGEN , Ko-Evolution von Public Relations und Journalismus: Ein erster Beitrag zu ihrer systematischen Aufarbeitung, „Publizistik“, LIII, 2008/1, pp. 9-24 Das Verhältnis von Public Relations (PR) und Journalismus wird in der Kommunikationswissenschaft meist als Zusammenspiel zweier interdependenter Systeme dargestellt. Unter historischer Perspektive beleuchtet der Beitrag, welche Wechselwirkungen sich dabei im deutschsprachigen Raum bei der Entstehung von PR skizzieren lassen. Schönhagen sieht deren Genese in engem Kontext mit der Geschichte des Journalismus und weist an historischen Beispielen nach, dass PR als Gegenreaktion auf die Entwicklung der „autonomen Nachrichtenproduktion“ der Massenmedien verstanden werden kann. Auf die Blütezeit der Persönlichkeits- bzw. Meinungspresse im 19. Jh. antworteten demnach Staat, Wirtschaftsunternehmen etc. mit PR, um der fehlenden oder aus ihrer Sicht verfälschten Berichterstattung entgegenzuwirken. Unter Beachtung der sozialen Entwicklung und des zunehmenden Legiti-

mationsbedarfs der eigenen Handlungen erkannten die Organisationen zudem, dass große Zielgruppen bzw. die Gesamtöffentlichkeit nur noch über Massenmedien erreichbar waren. Dies hatte wiederum Folgen für den Journalismus, der z.B. aufgrund vermehrter Stoffzulieferung mit einer Ausdifferenzierung der redaktionellen Strukturen reagierte und sich – bedingt auch durch technische Neuerungen – zu einer modernen Massenpresse entwickelte. Jan Henschel BERTRAM SCHEUFELE, Die These der Negativitätsumkehrung – Ein Experiment zur Wirkung von Negativität in den Politikerund Problemdarstellungen von Printmedien, „Publizistik“, LIII, 2008/1, pp. 48-64 Welche Wirkung hat die immer negativere Politikberichterstattung auf die Rezipienten? Dies untersucht Scheufele auf der Basis kognitiver, emotionaler und konativer Einstellungseffekte und entwirft seine „These der Negativitätsumkehrung“. Die These besagt, dass Negativität in den Politikberichten nicht zwingend zu ablehnenden Meinungen oder gar zu Politikverdrossenheit führen muss, sondern in konkreten Urteilssituationen auch positive Effekte in Form von Zustimmung mit sich bringen kann. Dies wird an Hand eines Zwei-FaktorenExperiments belegt, bei dem Probanden zuerst ein negativer (Artikel A1) oder positiver (A2) Zeitungsartikel über die Problemlösungsfähigkeit deutscher Politiker vorgelegt wurde, bevor sie dann einen Text erhielten, der sich mit einem konkreten landespolitischen Problem auseinandersetzte, was dramatisiert (Artikel B1) oder sachlich (B2) dargestellt wurde. Ergebnis war, dass die A1Leser durch die neuerliche Medienkritik einen permanent verfügbaren Bezugsrahmen („ungelöste Probleme“) aktivierten, der zu einem größeren Problembewusstsein führte und eine Kontrastfolie bildete, von der sich die konkrete landespolitische (Problemlösungs-) Entscheidung abhob. Folge war ein überwiegend positives Urteil, das sich im Falle der

RASSEGNA DI LINGUISTICA TEDESCA

Lektüre des dramatisierten Textes B1 noch verstärkte. Jan Henschel INGA HUCK – HANS-BERND BROSIUS, Der Third-Person-Effekt – Über den vermuteten Einfluss der Massenmedien, „Publizistik“, LII, 2007/3, pp. 355-374 Warum neigen Menschen dazu, den Einfluss der Massenmedien auf andere zu überschätzen, während sie eine Wirkung auf sich selbst weitgehend zurückweisen? Das von W. Phillips Davison unter dem Namen „Third-Person-Effekt“ (TPE) beschriebene Phänomen wird in dem Beitrag von Huck und Brosius auf der Basis knapp 100 veröffentlichter Studien weitreichend analysiert, indem ein allgemeiner Forschungsüberblick der empirischen Befunde, der verstärkenden und abschwächenden Faktoren sowie der theoretischen Erklärungsmodelle gegeben wird. Dabei entwerfen die Autoren ihr integratives Modell unterschiedlicher Wahrnehmungsphänomene. Für den TPE sehen sie die Ursachen besonders in der Angst, dass Medien manipulieren sowie der generellen Kontrollillusion der Menschen. Diese glauben, ihre Umwelt zu kontrollieren, auch wenn sie es – objektiv betrachtet – nicht tun, gestehen anderen Personen aber nicht die gleiche Fähigkeit zu, wenn sie sich einer sozialen, psychologischen oder politischen Distanz zu ihnen befinden. Zur weiteren Forschung über mögliche Verhaltenskonsequenzen der Menschen auf Grund des TPE lädt die ausgiebige Bibliographie im Anhang ein. Jan Henschel R OSMARIE Z ELLER , Bemerkungen zur Geschichte der Gattungsbezeichnung „Schelmenr o m a n “, “ S i mp l i c i a na”, X X I X , 2 0 0 7 , pp. 393-396 Der Beitrag befasst sich mit der Geschichte der Gattungsbezeichnung „Schelmenroman“ und zwar mit der Frage, wie das negativ besetzte Wort „Schelm“ Bestandteil einer Gat-

519

tung sbezeichnung werden konnte. Ursprünglich bedeutete es nämlich „Kadaver“, „Leichnam“, „totes Vieh“ und im Wörterbuch von Adelung treten auch die Bedeutungen „tödliche Seuche“ und „ehrloser Mensch“ auf. Die zeitgenössische Übersetzung für „pikaro“ war „Landstörtzer“ und auf keinen Fall „Schelm“. Die Abschwächung dieser negativen Bedeutung ist die Voraussetzung dafür, dass das Wort als Gattungsbezeichnung fungieren konnte. Am Anfang des XIX. Jahrhunderts findet man den frühesten Beleg dafür, dass der Pikaro-Roman als „Schelmenroman“ bezeichnet wird, obwohl die Erklärung in Klammern – d.h. „Schelmenromane (del gusto picaresco)“ – darauf hindeutet, dass der Ausdruck noch unüblich war. In den achtziger Jahren desselben Jahrhunderts gibt es Belege von der generalisierten Verwendung des Begriffs und sie beweisen, dass die ursprüngliche negative Bedeutung des Wortes „Schelm“ für den Pikaro abgeschwächt worden war. Alessandra Monti HEIDEMARIA ABFALTERER, Der Südtiroler Sonderwortschatz aus plurizentrischer Sicht. Lexikalisch-semantische Besonderheiten im Standarddeutsch Südtirols, Innsbruck University Press, Innsbruck 2007, 286 pp. 2004 erschien das erste plurizentrische Wörterbuch auf der Welt. Herausgegeben wurde es unter dem Titel Variantenwörterbuch des Deutschen. Die Standardsprache in Österreich, der Schweiz und Deutschland sowie in Liechtenstein, Luxemburg, Ostbelgien und Südtirol (Berlin et al.: de Gruyter) von dem Duisburger Germanisten und Initiator des Projekts Ulrich Ammon in Zusammenarbeit mit österreichischen und schweizerischen Kollegen. Heidemaria Abfalterer hatte im Rahmen dieses Projekts die Datenbank für den Südtiroler Raum betreut. Jetzt hat sie das Quellenkorpus des Südtiroler Sonderwortschatzes erweitert und ein eigenes Wörterbuch herausgegeben, das die Südtiroler Varianten der deutschen Standardsprache aus plurizentrischer Sicht systematischer und

520

RASSEGNA DI LINGUISTICA TEDESCA

umfassender darstellt. Damit hat sie ein Nachschlagewerk vorgelegt, das sich von der editorischen Anlage her sehr gut benutzen, in der alphabetischen Anordnung der Stichwörter leicht überblicken und dank der selbstständigen, inhaltlich fundierten, detaillierten und schematisch ausgeführten Wörterbuchartikel gewinnbringend lesen lässt. Mehrere dem Wörterbuch vorangestellte kurze und prägnante Abhandlungen führen in das Thema ein und eignen sich als Einstieg für eine breite, sprachlich interessierte Leserschaft. Neben der Einführung in die plurizentrische Auffassung von der deutschen Sprache wird auch die Sprachsituation in Südtirol ab 1920 behandelt, als Südtirol Teil des italienischen Staatsgebietes wurde. In einem dritten Kapitel gibt sie einen Einblick in ihre methodische Vorgehensweise, in der auch der Aufbau der Datenbank zum Südtiroler Deutsch und ihre statistische Auswertung zur Sprache kommen. Schließlich hat sie in einem vierten Kapitel von A (Abänderungsantrag) bis Z (Zweitsprachunterricht) insgesamt 621 Lemmata zusammengetragen, die als „eigenständige, in sich geschlossene lexikographische Arbeit zum Sonderwortschatz in Südtirol“ (71) gelten kann. Sandro M. Moraldo MARC RATHMANN, Anglicisms in the German Business Press. A Corpus-Based Study, VDM, Saarbrücken 2007, 135 pp. Rathmann untersucht anhand einer Korpusanalyse der Wirtschaftswoche (1973-2003) den Gebrauch von Anglizismen in der deutschen Wirtschaftssprache. Nach einem Überblick über den Stand der Anglizismenforschung und der Klärung terminologischer Aspekte geht Rathmann den Fragen nach, ob die Anzahl der Anglizismen in der Wirtschaftssprache steigt und von der Textsorte abhängt. In der Wirtschaftswoche stammt der größte Anteil von Anglizismen aus dem Allgemeinwortschatz (67,35%) gefolgt von spezifischen Wirtschaftsanglizismen (25,97%), während aus anderen fachsprachlichen Domänen nur wenige Anglizismen vertreten

sind. Die Zahl der Anglizismen hängt von der jeweiligen Textsorte ab. In Titelgeschichten ist im Vergleich zu Leitartikeln und Leserbriefen die Anzahl von Anglizismen am höchsten und steigt am schnellsten an. Eine mögliche Erklärung ist, dass Journalisten bei der Berichterstattung aus dem Ausland englische Wörter übernehmen oder aber auch aus Prestigegründen verwenden. Da Leitartikel zur Beeinflussung der Lesermeinung beitragen und damit möglichst verständlich sein sollen, treten dort weniger Anglizismen auf. Aus den Leserbriefen ergibt sich, dass Leser deutlich seltener Anglizismen benutzen als Journalisten. Stefanie Vogler CORDULA MEISSNER, Eine gebrauchsorientierte Beschreibung des Sprachsystems mit Hilfe der Korpuslinguistik – das Beispiel der Synonyme ewig und unendlich, “Deutsch als Fremdsprache” 2008/1, pp. 8-13 Auf der Grundlage der generativen Grammatik Chomskyscher Prägung und der davon berührten angewandten Linguistik zeigt die Autorin den relevanten Blick der Korpuslinguistik auf das Sprachsystem. Im Mittelpunkt steht die reale Sprachverwendung, die auch im Fremdsprachenunterricht eine besondere Relevanz hat. Ziel des Artikels ist es daher zu zeigen, wie die traditionelle Sprachbeschreibung mit korpuslinguistischen Mitteln präzisiert und ergänzt werden kann. Dies wird durch ein Beispiel auf lexikalischer Ebene durchgeführt, nämlich auf der Ebene der Synonymiebeziehung der Adjektive ewig und unendlich, wofür zwei Korpora konsultiert werden. Mit der Untersuchung wird gezeigt, dass die Bedeutungsbeschreibungen beider Adjektive in traditionell erstellten Wörterbüchern des Deutschen nicht immer ihre reale Verwendung in der Gegenwartssprache zeigen: Die am häufigsten verwendeten Bedeutungen seien nicht immer adäquat dargestellt und die Reihenfolge der Bedeutungsvarianten entspreche nicht immer der Häufigkeit

RASSEGNA DI LINGUISTICA TEDESCA

ihres tatsächlichen Auftretens in der Sprache. Und da sich die Darstellung als unvollständig erweise, wünscht sich die Autorin eine empirisch fundierte Überarbeitung der Wörterbuchdarstellungen. Als Hilfsmittel brauche man dazu Korpusanalysen, Korpusbelege und Korpora, deren Einsatz im Fremdsprachenunterricht eine neue Möglichkeit darstelle, Lerner die reale Verwendung der Zielsprache entdecken zu lassen. Lucia Salvato MAGDALENA SZULC-BRZOZOWSKA, Über die Bedeutung von aus und von – kontrastiv, “Deutsch als Fremdsprache”, 2008/1, pp. 46-50. Gegenstand der Untersuchung ist die Semantik des Präpositionspaares aus und von, dessen naheliegende Bedeutung oft von (polnischen) Deutschlernenden verwechselt wird. Problematisch hat sich in diesem Sinne die Frage nach den kontextuellen Bedingungen zu der richtigen Verwendung beider Präpositionen erwiesen. Ziel des Beitrags ist es darum, eine abstrakte, möglichst einheitliche bzw. nichtvariable Bedeutung beider Präpositionen zu präsentieren. Die Analyse konzentriert sich auf den Gebrauch von aus und von in der Präpositionalphrase mit dem syntaktischen Wert einer Adverbialbestimmung. Mittels einer kontrastiven Analyse werden die semantischen Merkmale beider Präpositionen in den Vordergrund gestellt. Da der Kontext dafür ein Unterscheidungsfaktor ist, wird die Untersuchung innerhalb des lokalen, temporalen, modalen und kausalen Kontextes, sowie bei der Präpositionalrektion von Verben durchgeführt. Die Ergebnisse zeigen, dass die semantischen Überschneidungen bei polnischen Lernenden auch wegen der irreführenden Äquivalenz in ihrer Muttersprache noch problematischer sind: Sie können kontextuell nur dann erfasst werden, wenn klare Unterscheidungsmerkmale vorliegen. Bei dem Fremdsprachenerwerb wäre darum das Ziel zu verfolgen, eine möglichst nichtvariable Bedeutung beider Präpositionen zu er-

521

reichen, um eine möglichst unterschiedliche Verwendung zu rechtfertigen und zu begründen. Lucia Salvato ULRIKE A. KAUNZNER, Grenzen überschreiten – Grenzen überwinden. Neue Herausforderungen an den DaF-Unterricht, in Ulrike A. Kaunzner (Hrsg.): Der Fall der Kulturmauer. Wie kann Sprachunterricht interkulturell sein?, Waxmann Verlag, Münster/New York/München/Berlin, 2008, pp. 1-26 In diesem Beitrag geht Kaunzner auf unterschiedliche Aspekte des Faches „Interkulturelle Kommunikation“ ein. Neben der Darstellung von Lehrinhalten und der Umsetzung des Faches im DaF-Unterricht ist vor allem ein Überblick über die Vermittlungsformen aufschlussreich. Die Lehrenden sollen so dazu angeregt werden, sich mit der Förderung von Selbstreflexion und Fremdverstehenstraining eingehend zu beschäftigen, um dann später die Schüler bzw. Studierenden zur Überwindung von Sprach- und Kulturgrenzen zu befähigen. Christine Arendt

CoverALL200801Mount.qxp:Layout 1

24-07-2009

14:03

Pagina 1

ISSN 1122 - 1917

L’ANALISI

LINGUISTICA

LETTERARIA

2008

L’ANALISI LINGUISTICA E LETTERARIA FACOLTÀ DI LINGUE E LETTERATURE STRANIERE UNIVERSITÀ CATTOLICA DEL SACRO CUORE

ANNO XVI

FACOLTÀ DI LINGUE E LETTERATURE STRANIERE

L’ANALISI LINGUISTICA E LETTERARIA

2008

VOLUME 1

ANNO XVI - 1/2008

ISSN 1122 - 1917

EDUCATT - UNIVERSITÀ CATTOLICA DEL SACRO CUORE