ISSA Proceedings 1998 – Emergent vs. Dogmatic Argumentation; Towards A Theory Of The Argumentative Process

From the mid-70s onwards, in line with the “pragmaticization” of research into argumentation, scholars have felt an increasing need to turn their attention to the argumentative process. Simplifying a bit, it may be said that they worked with Toulmin’s layout, or with the topical tradition into which Perelman & Olbrechts-Tyteca had put new life; but they began to be interested in how arguers actually sorted out what was claim and data and how they hung together by an inference warrant, or how exactly a topical inference was based on reality or actually reorganized the structure of reality.
In a text as early as Norwegian philosopher Arne Næss’s introduction to logic En del elementære logiske emner – English version Communication and Argument -, first published in Norwegian in 1941, a point is made in favor of taking into account, not only the argumentative product, i.e., the “completed” layout or topical inference, but also the process of “completing” it. For Næss has it that the bulk of an argumentative encounter is not about argumentative support proper, but about being clear what an utterer meant when he used a certain expression. Næss introduces the four procedures of ‘specification,’ ‘precization,’ ‘generalization,’ and ‘deprecization’ by which arguers can be clearer about what exactly they want an expression to say.
Few approaches to argumentation have taken up this process-orientedness of Næss’s account, among them Frans van Eemeren and Rob Grootendorst’s Pragma-Dialectics. Their meanwhile well-known and influential approach assumes that ideally a resolution-oriented discussion goes through four stages in each of which only certain resolution-furthering moves can be allowed. But furthermore, at every stage the discussants may perform speech acts specifying or precizating what they mean to say. However, these usage declaratives continue to be defined in the perspective of an argumentation that is successfully conducted to its fourth and concluding stage. That is to say, the argumentative process continues to be connected very closely to the product, i.e., the “completed” argumentation having successfully supported a standpoint which had been contested.

But, as van Eemeren & Grootendorst (1992 : chap. 1) themselves acknowledge, the connection of the process and the product of arguing in colloquial speech is not as systematic as the earlier version of their theory (1984) might suggest. What prima facie would seem to be irrelevant sidesteps or childish bickering may be revealed to have a determining influence on the outcome of the discussion (see Jacobs & Jackson 1992). A discussion about one contested standpoint may become more and more complex because clarification is needed as to some of the elements adduced in support of this standpoint (see Snoeck Henkemans 1992). That is to say, while the product of arguing is perhaps best analyzed as an inference complex that dialectically renders plausible a conclusion with the help of plausible premises, the communicative process of arguing deserves more attention as a particular kind of conversation and, therefore, is best analyzed, as are other kinds of conversation, as a step-by-step process extending in time and not necessarily being organized by a dialectical macrostructure.
This is possible with a joint dialectical and communicational reconstruction, prefigured by Normative Pragmatics as proposed by van Eemeren, Grootendorst, Jackson, & Jacobs (1993). In this framework, I shall give a different and more “communicational” interpretation to Næss’s four procedures. Thus, I will be able to reconstruct the argumentative process as a kind of communication organized, on the one hand, by a global dialectical goal and, on the other, step by step by local discursive moves. With Næss’s procedures of clarification in mind, I shall develop a tool for reconstruction starting from a model offered by Richard Hirsch in a different context. With this tool, it will be possible to show that the process of arguing is not always about the justification or refutation of a definable proposition on the background of presuppositions which are shared in principle, but very often about trying to match these presuppositions, these individual backgrounds, as best the arguers can, in order to overcome a problematic situation. In a sense, then, through the argumentative enterprise something individual becomes “inter-individual” or “intersubjective.” I shall show in this paper that this “intersubjectification” may work easily, may require considerable communicative co-operation, or may fail utterly – and this reflects whether or not at the outset the presuppositions of the arguers resembled each other closely. For obviously, an argumentation is more likely to succeed if the respective arguers’ unconstested starting points are quite similar and more likely to fail if they do not find enough common ground to start from (see, as to this, Willard’s (1983; 1989) theory of argumentative fields).

1. Discourse operations and their linguistic reflexes
Taking seriously Næss’s and van Eemeren, Grootendorst, Jackson, & Jacobs’s point that arguing has a justification-shaped dimension and a clarification-shaped dimension and implementing this point in a step-by-step analysis of the argumentative process requires that the reconstruction tool I will propose account indifferently for every step as a justifying step or as a clarifying step within an argumentative macro-structure. To do this, I shall elaborate on Richard Hirsch’s concept of ‘discourse operations’ (1989 : chap. 4). Hirsch conceives of arguing as an interactive problem solving activity carried out by collaborating interactors. When interactors feel that the information about a given subject which they have at their disposal is problematic, they start generating new information to handle the problem. Thus, the information state given at the outset is modified, and by evaluating all newly generated information as to whether it helps reach a less problematic information state, the arguers alter the general picture step by step and interactively in such a way as to arrive at an information state which is considered unproblematic. The interactive generation and immediate evaluation of information is called by Hirsch a ‘discourse operation,’ which has, accordingly, two phases and can be accounted for in terms of how an utterance reacts as an evaluation to a newly generated information state (1989 : 38-40). It may create a contrast or a complication, which conforms to doubting that the newly generated information state is promising as to arriving at an unproblematic picture (this would be the traditional opponent casting doubt on a proposition). It may consist of logic-like operations such as conjunction or conclusion; and it may be represented by semantic operations which help find a more adequate interpretation of information, such as precization or specification (this would be Næss’s clarifying procedures as part of the arguing) (1989 : 59-74).

All of these discourse operations, serving the purpose of processing information states in such a way as to come closer to an unproblematic state, have paradigm reflexes on the surface of a text; e.g., the connectors but for a contrast, therefore for a conclusion, or and for a conjunction, etc. And although I am not very at ease with Hirsch’s information theoretical background, which suggests that communication would rely on adequate and rather unproblematic mental representations of reality, I shall elaborate his concept of discourse operations which is worth closer examination. For it is likely to render what Normative Pragmatics assumes the process of arguing to be: the arguers’ co-operative step-by-step effort to sort out how they might overcome a communication problem (in the first place, a conflict of opinion). It is therefore necessary to give Hirsch’s concept a more “communicational” shape; and I shall, consequently, start from the assumption that discourse operations, whose surface reflexes are connectors like but or and, do not link information states but utterances. That is to say that by choosing a certain connector an interactor links his contribution in a specific – contrasting, complicating, etc. – way, to the communication as it has developed to the point where he chooses this connector.

Fig. 1 shows the last two sentences of the preceding paragraph as they are built up segment by segment with the help of connectors representing discourse operations which I felt were appropriate to develop my point about a “more communicational version” of Hirsch’s model being necessary for my purposes.

The proposed way of putting the concept of discourse operations might be called a “pragmaticization” of Jean-Claude Anscombre and Oswald Ducrot’s (1983; see Ducrot 1993, Anscombre, Ducrot, García Negroni, Palma, & Carel 1995 and the thematic issue Journal of Pragmatics 24 (1995)) structuralist Theory of argumentation in the langue. According to this theory, because of lexical and semantic properties of entities of the language system, Saussurean langue, sentences carry with them ‘implicit conclusions’ and hence have an ‘argumentative orientation.’ For instance,[i] a sentence like, ‘The movie is poorly directed,’ is more likely to argue for an implicit conclusion, ‘It is poorly acted,’ than for its opposite, ‘It is very well acted.’ Hence, the former conclusion has the same argumentative orientation as the sentence, and the latter has an opposite argumentative orientation. This is illustrated by the fact that, ‘The movie is poorly directed and poorly acted,’ sounds o.k. (same orientation), and that, ‘The movie is poorly directed and well acted,’ sounds somewhat odd (opposite orientation), whereas, ‘The movie is poorly directed but well acted,’ sounds o.k. The connector and, then, reflects the identical, the connector but the opposite, argumentative orientation of two connected sentences.

This is in line with my point that the discourse operation reflected on the surface by and creates a conjunction, and that that reflected by but creates a contrast. However, while Anscombre & Ducrot assume that this takes place at the level of the langue, the language system, and that the parole, the enactment of the language system, is sort of accessory, I shall argue that communication is more dynamical. When the addressee of an utterance connects to this utterance his own, following contribution by means of a connector that reflects a conjunctive, complicative, etc., operation, then this would in fact seem to suggest that the proposition conveyed by an utterance authorizes only certain pragmatically meaningful argumentative continuations – namely, the implicit conclusions it carries with it –, but others not. But it would seem, rather, that this is not in the first place a matter of langue but that it is up to the addressee/ respondent to choose one out of several possible meaningful continuations. Whether the continuation the addressee has chosen is in fact an appropriate one may be subject to closer scrutiny.[ii] For another interactor may go on with a contrastive or complicative discourse operation; and this complication, in turn, may involve precizating or usage declaring operations on lower hierarchical levels.[iii] Let us see how this works with a few examples.

2. Intersubjectification working without serious problems
I have said that by the discourse operations which interactors create by reacting in a specific way upon other interactors’ preceding contributions, something individual becomes intersubjective. This intersubjectification may work easily, as I will show now to illustrate how the concept of discourse operations “processing communication problems towards a solution” can account for the global dialectical and local step-by-step structure of argumentative encounters. The analysis to follow is displayed by Fig. 2.

Situation: In the French Enlightenment philosopher Fontenelle’s New Dialogues of the Dead (1686), Erasmus of Rotterdam reproaches Charles V. of Spain with the aristocratic privileges this latter would have, as son of a king, by mere chance without deserving them. Charles opposes to this that Erasmus must not appeal to his knowledge either; for this he has got from the wise men who preceded him, and learning everything that these knew, would not, says Charles, be more difficult than keeping the fortune an aristocrat inherits from his ancestors. To which Erasmus replies.

Erasmus: But let us not talk about knowledge, let us stay with intelligence; this quality in no way depends on chance.[iv]

Erasmus connects his utterance to what precedes as a contrast (but) designed to inhibit Charles’s equivalence of acquiring knowledge and keeping fortune. The contrast is, to look closer, a disjunction (let us not – let us) with its explication (for, which is unexpressed but can easily be reconstructed).

Charles: It does not depend on it? What! Doesn’t intelligence consist in a particular formation of the brain, and is there less chance in having been born with a well-formed brain than in having been born the son of a king? You were a great genius, but ask all the wise men the reason why you were not stupid and imbecilic: almost nothing at all, a slight change in the arrangement of fibers.[v]

By connecting rhetorical questions (recognizable above all by the negations) to the preceding utterance, Charles creates, on the dialectical level, a complication which, if successful, inhibits Erasmus’s contrast and hence strengthens his own equivalence ‘acquiring = keeping.’ This complication, in turn, conjoins (and) a precization of what intelligence is (consist in) and the claim that a well-formed brain comes about as much by chance as an aristocratic birth. The complication proper relies on a contrast (but) which elaborates on what has just been said.

Erasmus continues with a question: ‘Tout est donc hasard? //Everything, then, is by chance?’ That is, he fills in the ‘yes’ Charles’s rhetorical questions suggest, and by a conclusive discourse operation (then) he creates a slot in which Charles can fill in the henceforth intersubjective conclusion to be drawn from what precedes: ‘Oui, pourvu qu’on donne ce nom à un ordre que l’on ne connoît point. // Yes, providing this designation is given to an arrangement one is not capable of knowing.’ (French spelling normalized; my translation.)
The fact that Erasmus does not go on doubting or discussing but creates a slot for Charles’s conclusion reflects that the intersubjectification of Charles’s point of view has succeeded without major problems. Although Erasmus seems to learn something that fundamentally reorganizes his presuppositions about being proud of privileges, material or intellectual, once he has learned it, the agreement is unproblematic; the problem has been resolved.

3. Elaborate repair needed to process disagreement
It might have been that Erasmus had not created a slot for an intersubjectification of Charles’s position. He might have asked for further clarification about how the brain is formed, how intelligence depends on a particular formation of the brain, etc. In that case, intersubjectification might have been possible as well, but it would have required much more collaborative effort.
For reasons of space, I cannot fully discuss here an instance of arguing in which the position held by one arguer at the outset or a position emerging during the arguing becomes intersubjective because of elaborate interactive examination of the acceptability of the position. Let me just point to some characteristics of such instances of arguing by illustrating rather than analyzing a portion of the Nuclear Dialogues in which David Weinberger offers a critique of the Reagan administration’s policy of deterrence in 1980s. One dialogue is between two philosophers one of which, Emma, wears a pin reading ‘Ban the bomb.’ The other, Jennie, considers the slogan to be childish and simplistic, and disagrees that wearing it does any good opposing nuclear weapons.
Upon closer examination they discover that Emma is not even against all potential instances of use of nuclear weapons, which is why they shift to another, albeit related, topic, namely, what exactly Emma means when she says that she is against nukes. It turns out that Emma is against the policy of deploying nukes in Europe and threatening to use them. But this position, in turn, requires further examination; for now Emma’s “refined” position has it that, even though one should avoid using nuclear weapons as far as possible, there might be instances of legitimate use. This, however, is the position the “atomic hawks” have, which is why Emma and Jennie feel the need to turn to question where the differences are between the supporters of the policy of deterrence and their own position, which is that they are against this policy. It is only now, after one more topic shift, that they come to the position emerging from their discussion that ‘being against’ for them means in the first place that they are against producing and deploying more and more nukes although the number of nukes existing is largely sufficient to deter military action by anybody in their right mind. That is to say that in fact Jennie and Emma intersubjectify a position at the end of their discussion, but that without considerable topic shifts, precizations, specifications, etc. – in a word: without considerable interactive argumentative co-operation the intersubjectification probably would have been impossible.
To a certain extent, this discussion has the same characteristics as the one analyzed in the preceding section. However, here between the emergence and the succeeding intersubjectification of the relevant position, considerable topic shifts occur, and the collaborative effort will finally lead the discussants to intersubjectify a position which neither of them held a the beginning of the discussion. In Erasmus and Charles’s discussion the intersubjectification follows immediately the emergence of the position stemming from Charles’s precization of what intelligence is. In Emma and Jennie’s discussion, on the other hand, precizations and complications “lead the discussion astray.” That is, they cause considerable topic shifts, so that at the end the interactors are no longer really having the same discussion they had at the beginning. The preliminary steps, then, are in a sense “dialectically worthless” because they are not immediately connected to the position emerging from the discussion and finally being agreed upon. Nonetheless, they may not be eliminated from the discussion if it is analyzed in a communicational perspective. For it is obvious that without these preliminaries that gave rise to the precizations and complications leading to topic shifts, the discussants would never have gone on to that part of their discussion in which intersubjectification finally was successful and, accordingly, the problem was resolved.

4. Intersubjectification fails
The most important advantage of the processual reconstruction of arguing with the help of the step-by-step model I am proposing is that it can account not only for arguing that reaches its goal, i.e., arguing in which in the end the intersubjectification of a certain standpoint with respect to a contested position is possible. It can also account for arguing that does not reach this goal, i.e., arguing in which in the end no intersubjectification occurs. This is necessary to be able to model the argumentative process as an element of  its own, quite independent of the outcome this process may have.
In Louis Armand baron of Lahontan’s Conversations of a Native and the Baron of Lahontan, published in 1703, the author offers the Europeans a picture of a North American Native people whose chief, Adario, has been to France and tells Lahontan throughout the conversations about his people’s views on morals, politics, and ethics and about what the differences are of these views as compared to the European views.
Adario has just pointed to a gap that can be noticed between the religious imperatives Europeans use to preach and their own behavior which does more often than not deviate considerably from these imperatives. Lahontan concedes to what Adario has said:
I am unable to deny the contradiction you have noticed. But one has to take into account that humans sometimes commit sins despite the guidance of their conscience, and that there are learned people who lead a bad life. This may happen because of lack of attention or the power of their passions, because they have devoted themselves to worldly advantage: man, corrupted as he is, is driven towards evil in so many places and by an inclination so strong that, unless there is an absolute necessity, it is hard for him not to give in.

Lahontan tries to inhibit the destructive power Adario’s point would have for his attempts to bring him to a conversion to Christianity (see Fig. 3). After having acknowedged the inconsistency to which Adario has alluded, he goes on with a contrastive discourse operation (but) in which an explication is given (this may happen because of) for the apparent contradiction. Adario’s answer to this is a radical complication, which, in turn, inhibits Lahontan’s contrast, thereby giving his previous point all its destructive power:

When speaking of man, say: the Frenchmen; for you’re well aware of the fact that these passions, this striving for advantage and this corruption you are talking about, are unheard of amongst our people.[vi]
By specifying that about which they should be talking and by explicating this specification, Adario claims that Lahontan is right perhaps as far as Europeans are concerned. But since he takes what Lahontan says to be pointless as to the present discussion, he is not prepared to process any of Lahontan’s utterances. Therefore the intersubjectification of a standpoint with respect to a position, proposed by Lahontan through the discourse operations he has performed, is not possible. Accordingly, Lahontan’s attempt to bring Adario to a conversion will fail, and the discussion will not lead to any dialectical conflict resolution worthy of the name.

5. Conclusions to be drawn
The step-by-step analysis I have proposed for the process of arguing has yielded above all the following result: Categories and concepts of analysis which are applicable to the product of arguing, such as inferential connections or accepting or denying the justifiability of a position, are hardly adequate to an analysis of the process of arguing. For this process operates with more flexible communicative maneuvres. I have accounted for these maneuvres, on the basis of a reinterpretation of Richard Hirsch’s model, as discourse operations, i.e, a specific argumentative processing of a communication problem realized by the interactors through, e.g., connectors or entire phrases used to link their own utterance continuing the communication to the preceding communication in a specific way intended for collaborative problem solving.
The concept of discourse operations has the adavantage that it can account for at least two kinds of arguing. Until now I have drawn a distinction roughly between arguing that succeeds and arguing that doesn’t. It is more adequate, however, to speak of arguing in which positions that were not shared at first become intersubjective, and of arguing in which nothing becomes intersubjective. For if Charles V. succeeds in countering argumentatively Erasmus’s accusation, this is because something completely new emerges from the discussion for Erasmus: people are intelligent or not by (physiological) chance. On the basis of this newly emerged position, having become intersubjective, an argumentative agreement is possible. But it might well have been that this new position would have remained as controversial as its predecessor was, and then argumentative agreement would have been impossible. This kind of emergent arguing is therefore no warranty for an agreement being possible.
In the same way, if Adario and Lahontan do not agree on the merits of Christianity, this is because the position Lahontan proposes does not actually become intersubjective. For Adario’s and Lahontan’s presuppositions, the backgrounds that underly their communication are too different. Whereas Erasmus and Charles can match their communicative backgrounds to a certain extent to make agreement possible, this does not work for Lahontan and Adario. So it is not the absence of something emerging from the discussion for at least one of the participants that impedes agreement; it is, rather, that nothing emerges and that at the same time the backgrounds would have to be matched to a certain extent – which, in turn, is impossible as long as nothing new emerges. For if the communicative backgrounds of the arguers coincide sufficiently, then agreements are very possible without there emerging anything new from the discussion. This is the case, for instance, in forensic argumentation, proceeding from communicative backgrounds which are largely homologous for all the arguers.

The major conclusion to be drawn from my paper is the following: The analysis of the process of arguing is faced with different kinds of arguing which do not represent discriminate types of a strict classification but, rather, a continuum extending between two extreme cases. In one extreme case of arguing nothing at all becomes intersubjective and a position is justified or refuted on the basis of communicative backgrounds essentially identical for all the arguers. These backgrounds, then, in a sense acquire the status of an uncontested dogma. Therefore, I term this extreme case of arguing ‘dogmatic.’ Its characteristics are that rather few topic shifts occur and that the bulk of the discourse operations used are complications/contrasts and explications – which represent the “classical” product analysis categories of casting doubt on a position and justifying the doubted position.
The other extreme case is what I term ‘emergent arguing,’ for in this type of arguing arguers make a co-operative and collaborative problem-solving effort to match their communicative backgrounds. Because of this, something new emerges from the discussion, which is usually plain because topic shifts occur, because, while arguing, arguers notice that they have to submit a certain point to closer scrutiny, etc. Consequently, in emergent arguing discourse operations like precization, specification, exemplification, and conclusion are more frequent than in dogmatic arguing.
Most of the actual arguing in colloquial speech is somewhere in between the extreme cases, and hence this continuous scale from dogmatic to emergent arguing provides only for a possibility to classify a given piece of discourse as more clearly a form of emergent or of dogmatic arguing. Still, neither of the extremities of the scale guarantees that one or the other of them makes arguing more likely to succeed. Neither of them is “better” than the other. While scientific arguing usually aims at “intersubjectifying” positions and therefore is more emergent, forensic arguing aims at winning a case on the uncontested basis of the body of legislation and therefore is more dogmatic. Neither of them, however, is better than the other; for they obviously have different goals. Hence, as long as non-argumentative and extra-communicative features do not influence on the arguing to such an extent as to make it a pseudo-argumentation, the analysis of the ongoing argumentative process with the tool I have proposed allows for an account of how much the arguers’ communicative backgrounds coincided, or of how prepared they were to start from a shared point of view. If dogmatic arguing succeeds, two interpretations are possible: Either there were no noteworthy differences between the arguers’ respective communicative backgrounds, or those who accept an argumentative justification of a position accept at the same time all the presuppositions on which this rests. If emergent arguing succeeds, then the arguers felt that there were noteworthy differences between their respective communicative backgrounds, but they were prepared to examine more closely the point(s) at issue and to give up or modify part of their own communicative background in order to be able to arrive at a shared view of the position discussed.

NOTES
i. Example taken from Anscombre & Ducrot (1989 : 73), which is one of their rare English papers. (It is, in fact, a translation of Anscombre & Ducrot 1986). Rühl (1997b) gives a brief overview over the concept of implicit conclusions. Other sources in English as to their theory are the presentation in Fundamentals (1996 : chap. 11) and Snoeck Henkemans’s (1995) critique of their analysis of but as an argumentative connector.
ii.This is in line with Jackson & Jacobs’s (1980; 1982) point that ‘conversational argument’ comes into being because an addressee has not performed the conventionally expected second pair part of an adjacency pair, thereby creating a communication problem needing repair. The advantage of speaking of an addressee’s choosing one out of a variety of possible meaningful continuations is that no ‘structural preference for agreement’ (1980 : 261-262) of adjacency pairs has to be assumed a priori, which is in a way an idealization making the analysis depart from a strict descriptive account of the interaction.
iii. I have given a detailed account as well as defintions of discourse operations elsewhere (Rühl 1997a : 213-215).
iv. ERAS[ME]. Mais ne parlons point de la science, tenons-nous-en à l’esprit ; ce bien-là ne dépend aucunement du hasard. (p. 109) – French spelling normalized. My translation.
v. CHAR[LES]. Il n’en dépend point ? Quoi! l’esprit ne consiste-t-il pas dans une certaine conformation du cerveau, et le hasard est-il moindre, de naître avec un cerveau bien disposé, que de naître d’un père qui soit roi ? Vous étiez un grand génie : mais demandez à tousles philosophes à quoi il tenait que vous ne fussiez stupide et hébété; presque à rien, à une petite disposition de fibres (p. 109-110) – French spelling normalized. My translation.
vi. I have proposed such an analysis elsewhere (Rühl 1997a : 247-270). The example discussed there is a portion of the dialogue De grammatico, composed by Anselm of Canterbury around A.D. 1080 to deal with one of the favorite research topics of scholastic logic and semantics, namely, the logical status of the so-called paronyma, that is, simplifying considerably, of expressions which are adjectives but can be used as substantives, such as, e.g., grammaticus. Anselm’s actual problem, however, is not the morphological problem of derivation but the ontological implications this has in the perspective of the philosophy of early Scholasticism. For if there are expressions which can be adjectives and substantives as well, this would mean, in this perspective, that there are things which can be at the same time accidental (Aristotelian ‘kategoroúmena’) and substantial (Aristotelian ‘hypokeímena’), with which scholastic metaphysics is not very at ease. For more details about the problem, see the commented editions of De grammatico provided by Henry (1964) and Galonnier (1986).

REFERENCES
Anscombre, Jean-Claude & Oswald Ducrot (1983): L’argumentation dans la langue. Liège: Mardaga (collection Philosophie et Langage).
Anscombre, Jean-Claude & Oswald Ducrot (1986): «Argumentativité et informativité» in: Michel Meyer (éd): De la métaphysique à la rhétorique. Bruxelles: Éditions de l’Université de Bruxelles; pp. 70-94.
Anscombre, Jean-Claude & Oswald Ducrot (1989): «Argumentativity and informativity» in: Michel Meyer (ed.):
From Metaphysics to Rhetoric. Dordrecht: Kluwer; pp. 71-87.
Anscombre, Jean-Claude, Oswald Ducrot, María Marta García Negroni, Silvia Palma, & Marion Carel (1995): Théorie des topoï. Paris: Kimé.
Ducrot, Oswald (1993): «Les topoi dans la “Théorie de l’argumentation dans la langue”» in: Christian Plantin (éd.):
Lieux communs, topoi, stéréotypes, clichés. Paris: Kimé (coll. Argumentation – Sciences du langage); pp. 233-248.
van Eemeren, Frans H. & Rob Grootendorst (1984): Speech Acts in Argumentative Discussions. A Theoretical Model for the Analysis of Discussions Directed towards Solving Conflicts of Opinion. Dordrecht: Foris (Pragmatics and Discourse Analysis 1).
van Eemeren, Frans H. & Rob Grootendorst (1992a): Argumentation, Communication, and Fallacies. A Pragma-Dialectical Perspective. Hillsdale, NJ: Lawrence Erlbaum.
van Eemeren, Frans H., Rob Grootendorst, Sally Jackson, & Scott Jacobs (1993): Reconstructing Argumentative Discourse. Tuscaloosa: U of Alabama Press (Studies on Rhetoric and Communication) Fontenelle (Bernard le Bovier de Fontenelle): Nouveaux Dialogues.
des Morts. [1683] in: Fontenelle: Ouvres complètes. Texte revu par Alain Niderst. 8 tt. Paris: Fayard, 1990 sqq. (Corpus des ouvres de philosophie en langue française); t. I [1990]; pp. 47 – 263.
Fundamentals of Argumentation Theory (1996). A Handbook of Historical Backgrounds and Contemporary Developments, by Frans H. van Eemeren, Rob Grootendorst, Francisca Snoeck Henkemans, [et al.]. Mahwah, NJ: Lawrence Erlbaum Galonnier, Alain (1986): Anselmi Cantuariensis Quomodo grammaticus sit substantia et qualitas; i.e. Dialogus De grammatico _ (Au sujet) de grammairien [latin – français]. Introduction, traduction, et notes par A.G. in: L’Ouvre d’Anselme de Cantorbéry, sous la direction de Michel Corbin, S.J. Paris: Éditions du Cerf. Tome II, 1986; pp. 23-105.
Henry, Desmond Paul (1964): The De grammatico of St. Anselm: the Theory of Paronymy. Notre Dame, IN: University of Notre Dame Press (Publications in Medieval Studies).
Hirsch, Richard (1989): Argumentation, Information, and Interaction. Studies in Face-to-face Interactive Argumentation under Differing Turn-taking Conditions. Göteborg: Department of Linguistics, University of Göteborg (Gothenburg Monographs in Linguistics 7).
Jackson, Sally & Scott Jacobs (1980): «Structure of Conversational Argument: Pragmatic Bases for the Enthymeme» in: Quarterly Journal of Speech 66; pp. 251-265.
Jacobs, Scott & Sally Jackson (1982): «Conversational Argument: a discourse analytical approach» in: J. Robert Cox & Charles A.Willard (eds.): Advances in Argumentation Theory and Research. Carbondale, IL: Southern Illinois University Press; pp. 205-237.
Jacobs, Scott & Sally Jackson (1992): «Relevance and Digression in Argumentative Discussion. A Pragmatic Approach» in. Argumentation 6; pp. 161-176.
Lahontan (Louis Armand de Lom d’Arce, baron de Lahontan et seigneur d’Erleix): Suite du voyage de l’Amérique, ou Dialogues ou Entretiens d’un sauvage et du baron de Lahontan [1703]. in: Lahontan: Ouvres complètes. Édition critique par Réal Ouellet, avec la collaboration d’Alain Beaulieu. 2 tt. Montréal (Québec): Les Presses de l’Université de Montréal, 1990; t. II, pp. 789 – 885.
Næss, Arne (1966): Communication and Argument. Elements of applied semantics. Translated from the Norwegian by Alastair Hannay. London: Allen & Unwin [Norwegian Text: En del elementære logiske emner. Oslo: Universitetsforlaget, 11941 – multiple re-editions].
Rühl, Marco (1997a): Argumentieren und Autorität. Untersuchungen zu argumentativer Sprachverwendung in asymmetrischen Kommunikationssituationen. Ph.D. thesis, University of Freiburg i.Br., Germany.
Rühl, Marco (1997b): «Argument and Authority: On the pragmatic bases of accepting an appeal to authority as rational» Paper presented at the “Argumentation and Rhetoric” Conference, held by the Ontario Society for the Study of Argumentation at St Catharines, ON, May 1997.
Snoeck Henkemans, A. Francisca (1992): Analysing Complex Argumentation: The Reconstruction of Multiple and
Coordinatively Compound Argumentation in a Critical Discussion. Amsterdam: SicSat (Amsterdamse Studies over Taalgebruik 2).
Snoeck Henkemans, A. Francisca (1995): «’But’ as an indicator of counter-arguments and concessions» in: Leuvense Bijdragen 84; pp. 281-294.
Weinberger, David: Nuclear Dialogues. New York/Bern/Frankfurt am Main/Paris: Peter Lang, 1987 (American University Studies, Series V: Philosophy, vol. 35).
Willard, Charles A. (1983): Argumentation and the Social Grounds of Knowledge. Tuscaloosa: University of Alabama Press Willard, Charles A. (1989): A Theory of Argumentation. Tuscaloosa/London: University of Alabama Press (Studies on Rhetoric and Communication).