Marcelo Dascal

1. Motivation

My first objective in this paper is to persuade dialogue analysts of all persuasions of the importance of polemical exchanges, and thereby to prompt them to apply to this kind of "dialogues" their analytical tools and skills. My second objective is to pre sent a number of distinctions I have developed in the course of my research on polemical exchanges, in the hope that they provide a useful framework for further research on this topic. My third objective -- in fact, my "hidden agenda" -- is to propose ne w grounds and new objectives upon which a fruitful interdisciplinary dialogue between philosophical and empirical approaches to the study of dialogue can be reinstated.

1.1 Is it really necessary to persuade anyone that polemical discourse occupies a central place in our public and private discursive lives? Hardly so. Whichever way we look, we are entangled in endless polemicizing: from daily domestic quarrels, through d isputes over parking places or office space, to political dissension; from talk-show discussions, through labor conflicts, to policy decisions; from mild disagreement, through bitter bickering, to schism; from critical book reviews, through congress round -tables, to scientific controversies; in matters of literary taste, in the courtrooms, in the workplace, in parliaments, and at home.

No doubt social and cultural (as well as individual) variation obtains here as elsewhere in communicative practices, and one may distinguish, as suggested by Kerbrat-Orecchioni (1994, 82ff.), between those societies driven by an ethos confronta tionnel (the example she gives is Israeli society) and those whose regulatve ethos is rather consensuel (like the Japanese society). Nevertheless, even in the latter, the elaborate politeness devices used to prevent explicit polemical confrontations that might harm one's face suggest that they do not overlook the importance of polemics but rather conduct them in a covert rather than overt way. It would seem, then, that in one way or another people everywhere are constantly engaed in either defending them selves, attacking somebody else, or avoiding open confrontation. So much so that it makes sense to argue, with McEvoy (1995), that "defensive invention" is a basic and universal communicative skill. No wonder that several cultural traditions assigned so m uch importance to developing, imparting and employing this skill: recall the importance of rhetoric in ancient and medieval education, the reliance on disputationes up to the seventeenth century and on their equivalents (oposiciones in Spain, defense de t hese in France) in order to grant university degrees, the Talmudic detailed record of the sages' discussions in establishing the halakha, the immense popularity of the Chan-kuo Tse in ancient China, even though the stratagems it teaches were considered un ethical, etc.

And yet, despite the obvious importance of polemical exchanges, dialogue analysts have not so far devoted to them the special attention they deserve. So much so that as recently as 1989 Roulet rightly pointed out that controversy was a "little studied" kind of "agonal exchange". Of course there are welcome exceptions. Prompted largely by Perelman's rehabilitation of rhetoric (cf. Perelman and Olbrechts-Tyteca, 1970), "argumentation" studies have flourished (see, e.g., Lempereur, 1991; and many articles in the journal Argumentation). The philosophy, sociology and history of science have recently turned their attention to scientific controversies (e.g., Engelhardt Jr. and Caplan (eds.), 1987; Gil (ed.), 1990; Gil, 1985; Granger, 1985), and it is hoped that the newly created "rhetoric of science" (e.g., McCloskey, 1984, 1995; Gross 1990; Pera 1991) will soon follow suit. But these studies have not, for the most part, been concerned with the dialogical characteristics of polemical exchanges, and co nsequently have not employed the conceptual tools of dialogue analysis and pragmatics. Noteworthy exceptions are attempts to chart the dialogical field to which controversy belongs (e.g., Jacques 1991), and the use of speech act theory (e.g., van Eemeren and Grootendorst, 19..), pragmatics (e.g., Dascal 1989, 1990a), rhetoric (Pera 1991), and related conceptual tools in the study of controversies. By and large, however, the theme of polemical exchanges is still awaiting for a concerted research effort by the dialogue research community.

1.2 Polemical exchanges are especially important in epistemology -- to wit the centrality of dialectics in Western thought, from Plato and Aristotle, through Kant and Hegel, to Popper and Kuhn. For these and many other thinkers, knowledge is achieved thro ugh the exercise of critical reason. No doubt many philosophers and scientists prefer to exercise criticism in the inner fore of their thoughts, or in monological discussions of others' texts, or else in fictive dialogues they write so as to be in full co ntrol of the opponents' replies. These preferences notwithstanding, criticism is primarily a form of dialogical activity, which manifests itself most naturally in polemical exchanges of various sorts. Furthermore, such a critical activity constitutes the most immediately relevant context both for understanding the meaning of a theory and for accounting for conceptual change in any domain.

With their professional bias towards abstraction, philosophers and historians of ideas have not showed particular interest for the empirical study of critical activity as it deploys itself in dialogical praxis. It is up to pragmatics, conversat ional and dialogue analysis, rhetoric, and related disciplines concerned with the study of language use, to take up this challenge, and to help epistemologists to flesh out their abstract models of criticism with fresh and insightful empirical evidence.

Belonging as I do to these two communities, it is only too natural that, lest I relished in a split self, I should try to harmonize them, by somehow making the toils of each of them useful for the other. In fact, there is also a bit of nostalg y in this endeavor. Once upon a time philosophy and rhetoric, logic and dialectics, and pragmatics and the philosophy of language were happily married. Wittgenstein, Grice, and Austin, each in his own way, raised the problems that opened a theoretical spa ce for pragmatics, and provided some of this discipline's foundational concepts, by way of pursuing typical philosophical inquiries (cf. Dascal, 1994).

But there is more to it than sheer saudades of the happy time of the pioneers, which may not have been, after all, so harmonious. Current debates about the "foundations" of speech act theory (e.g., Tsohatzidis (ed.), 1994 and its critique by W eigand, 1996) and of pragmatics in general (e.g., the special issue (in volume 17, 1992) of the Journal of Pragmatics devoted to this topic), about how to develop an action-based dynamic and dialogical grounding of the study of language use (e.g., Vernan t 1996), about the universality or culture-specific character of communicative competence and practice (e.g., Wierzbicka, 1993; Blum-Kulka, 1992), about the "correct" number of the conversational maxims (e.g., Rolf, 1989) and the (in)sufficiency of the pr inciple of cooperation (e.g., Attardo, 1997) -- all this shows that the field of research created by the pioneers is far from having secured sound philosophical foundations. Progress towards this aim requires further dialogue between dialogue researchers and philosophers.

But philosophy too -- in particular epistemology -- has much to gain from such a dialogue. It is no big news that, for quite a while, the "modern" notion of rationality has come under heavy "post-modern" fire (cf. Cohen and Dascal (eds.), 1989 ). Nor is it big news that the split between normativist and descriptivist approaches within epistemology in general and the philosophy of science in partiular -- a split that, among other things, touches upon the question whether one can speak of the "gr owth of knowledge" in terms of rational criteria or not -- rather than subsiding has been increasing, leading the field to a true impasse (cf. Dascal, 1995c). And it is also well-known that philosophers, who for some time have discussed the issue of "cult ural relativism" as if it did not affect the very nature of philosophy itself, are now aware that this is not the case (cf. Dascal (ed.), 1991).What is at stake, in all these crucial issues of present-day philosophy, is the very issue of the possibility o f communication -- between radically opposed "paradigms", between deeply different "conceptual schemes", between apparently irreconcilable conceptions of rationality and of its role in our knowing, acting, and communicating.

Under these circumstances, it is no longer possible to base one's theory of communication -- as both Grice and Searle did, and as action-based theories of dialogue do -- upon an unexamined principle of (instrumental) rationality and upon the pr esumption that successful communication and understanding are the rule and misunderstanding and miscommunication the exception. Nor is it possible anymore to sit back and wait until philosophy disentangles itself by its own means, so as to finally -- once more -- clear the conceptual ground for empirical research to proceed. The "hidden agenda" I suggest consists in emphasizing the importance of the opposite direction. I am convinced that, by studying empirically those cases -- such as polemical exchanges -- where communication seems to be at odds with the old comfortable notions of cooperation and rationality, and yet still to be possible, we may contribute to the solution of the philosophical impasses alluded to above. We may find out, for example, that controversies display a special kind of "rationality" and "normativity" which cannot be reduced neither to the strictures of idealized logic nor to the sheer exercise of power in what amounts to no more than conflicts of interests (cf. Dascal, 1996). Wha tever we find out, it is sure to be exciting and important: exciting because the dialogue with empiria, in order to be worth its while, must be open to surprises, to imprevisibility, just like the dispute with a real opponent always is; important because it will be relevant to some of the major intellectual (and practical) issues of our time, as well as to the conceptual foundations of the empirical study of dialogue.

2. Methodological remarks

I have been arguing for an empirical study of polemical exchanges as a means to solve certain acute philosophical problems -- in particular, as a way to address the problem of "rationality" by examining critical activity as it is displayed in polemical ex changes. But -- the reader might justly complain -- my argumentation here, and perhaps my methodology in general, is all but "empirical". In fact, it seems to be a classical example of a "top-down" strategy: I begin with providing a "philosophical" motiva tion for studying a certain kind of phenomena; I go on by anticipating the kinds of conclusions I expect from such a study; and, as will be clear in what follows, I propose a typology for the phenomena to be studied consisting in quite abstract "ideal typ es". I even dismiss in advance possible counter-examples to this typology (e.g., cases that do not fit any of the ideal types proposed) by claiming that "real cases" are always "impure", for they involve a mixture of the three ideal types -- a claim typic al of all those who favor theoretical elegance over a close fit to the data.[1]

So, I have to explain my apparently self-contradictory methodological stance. The explanation is that it is neither strictly top-down nor strictly bottom-up. In fact, when I began to study actual controversies, my main methodological assumption was that the conceptual tools of pragmatics could be applied to them -- a rather harmless top-down assumption (see Dascal, 1989, 1990a). I was surprised to discover a correlation between the contenders' preferred use of certain argumentative patterns and their epistemological conceptions about the nature of knowledge and of its production. It was only in the light of further studies of actual controversies that the typology presented below gradually emerged, and its correlation with epistemological issue s was further explored.[2] At present, I view my proposal as a general hypothesis, which I want to be put to test by further empirical studies of a broad range of polemical exhanges.

This sounds like a perfectly run-of-the-mill application of the "hypothetico-deductive method". I mention it because it highlights the fact that we never begin either at the "top" or at the "bottom", but rather somewhere in the middle, for t he simple reason that we don't know what is the top and what is the bottom -- both being relative to the current state of research. Progress is made by moving both "upwards" and "downwards", by treating "analysis" and "synthesis" (in Leibniz's sense) as a lways going hand in hand, and thereby by actually treating knowledge as both "bottomless" and "topless". This is, anyhow, my methodological stance.

A further methodological clarification is in order. There are many "levels of organization of dialogue", many ways of distinguishing between these levels, and many ways of analyzing their structure.[3] Polemical exchanges sh ould be studied at all of this levels, of course. Here, I will focus on two "macro" levels of organization, which might be called, respectively, "strategical" and "tactical". The former, which is akin to what Jacques (1991) calls "discursive strategies", has to do with the global pattern of a polemical exchange -- its overall aims, its general thematic and hierarchical structure, and the corresponding assumptions about its "rules" (if any) and its mode of resolution. The latter has to do with the nature o f the moves and countermoves employed at specific points in the exchange, in the light of the varying contingencies or "demands" (cf. Dascal, 1977) of the polemics as it unfolds; this "tactical" level is part and parcel of a polemical dialogue's "pragmati c structure" (cf. Dascal,1992) or sequential organization. In both, its strategical and its tactical level, polemical exchanges share much with other forms of dialogue, but I will naturally focus on their peculiarities, which I will present in terms of tw o sets of ideal types. The study of these "macro" levels does not preclude the need for a detailed analysis of of their "micro" components. Quite on the contrary, it should be complemented and supported by such an analysis. Once more, I am here beginning in the "middle" (with a slight tendency towards the "top"), relying on others to continue upwards and downwards.

3. Three types of polemical exchanges

A polemical exchange involves at least two persons who employ language to address each other, in a confrontation of attitudes, opinions, arguments, theories, and so forth. The important expressions in this definition are address each other and confrontati on. The former stresses the interactive aspect ("exchange", "dialogue") and the latter, the content of the interaction, as perceived by the participants. Both expressions require clarification, which is best achieved by pointing out how they determine the extension of the field of "polemical exchanges".

3.1 The address-each-other requirement excludes from this field all sorts of polemical discourse in which one of the "participants" is unable to actually participate, i.e. all those polemics where there is no real dialogue. For example, the "critical rece ption" of a (deceased) author's work -- e.g., Popper's diatribes against Marx, Hegel, and Plato, in his The Open Society and its Enemies. Similarly, it excludes the literary and philosophical genre "dialogue", where one of the parties is not the actual pr oducer of his interventions -- e.g., Plato's, Leibniz's, Hume's "dialogues". Although such works may echo polemical exchanges that really took place (and may thus help to reconstruct them), they cannot be taken as reliable records of such exchanges. The r eason for excluding these cases has to do with my insistence on viewing polemics as an activity, and -- more importantly -- an activity that always involves an element of uncertainty regarding the opponent's reactions. The amount of uncertainty varies wit h the type of polemical exchange (ritualized forms, such as the obligatio, allow for little inventiveness). Nevertheless, a live, real, and active (i.e., neither dead, nor imaginary, nor silent) opponent is unpredictable in his/her reactions. Although we may anticipate to some extent our opponent's reactions, and even undertake to manoeuver her to react in a certain way, polemics is essentially a game where our capacity for predicting the adversary's move is limited. For this to be possible, each contende r must be able to exercise the right to contest not only the opponent's views but also the latter's renderings (quotes, summaries, interpretations) of his [the former's] positions. Since this right can be put to use either privately or publicly, either or ally or in written form, either directly or indirectly (e.g., through intermediaries), all of these forms of confrontational interaction should be considered "polemical exchanges".

The second requirement amounts to the acknowledgment of the fact that every polemical exchange involves "opposed views" regarding some "content". Both of these notions shoud be conceived broadly, for both the nature of the content and of the op position leading to a polemical exchange may vary considerably. The latter may range from logical contradiction or weaker forms of logical contrariness, through differences in semantic or pragmatic presuppositions, up to pragmatically construed contrasts (such as those implicit in most uses of but (cf. Dascal and Katriel, 1977), etc. The opposed "contents", in turn, refer not only to specific propositional contents on a variety of matters (factual, methodological, evaluative, etc.), but also to illocutio nary force, attitudes, preferences, emphasis, judgments of appropriateness and relevance, etc. All these may give rise to polemical exchanges, regardless of whether they are "truly" opposed. What matters is that the contenders perceive them as being oppos ed and consequently engage in a debate having such contents as its explicit or implicit object.

As an object of study, polemical exchange thus consists primarily in those texts or utterances directly addressed by each disputant to the other (or others), privately or publicly. In addition to this "primary text", there is in general a va st "secondary text" which, at least partially, belongs to the exchange. It includes, for instance, works or other exchanges by the disputants where the polemics is reflected directly or indirectly, as well as letters to third parties where allusion is mad e to it. A broader circle of texts that are pertinent form its "co-text" which includes, for example, works or exchanges by prior or contemporary authors quoted and relied upon by both disputants. Finally, every polemics unfolds within a nondiscursive "co ntext', whose various aspects and levels have always a more or less important role in their content and development.[4]

3.2 The family of polemical exchanges thus described includes, among its many members, verbal quarrels between couples, political debates, round tables in scientific congresses, critical reviews of books and replies to them, medieval disputationes, etc. W ithin this family, I propose to characterize three ideal types, which I will call -- for the sake of having a handy terminology -- discussion, dispute, and controversy. The main criteria for this typology are: the scope of the disagreement, the kind of co ntent involved in it, the presumed means for solving the disagreement, and the ends pursued by the contenders. These differences belong to the "strategical" macro level in so far as they refer primarily to the overall structure of the exchange, the assump tion being that such a structure reflects (at least to some extent) the contenders' planning and performing its "larger ... movements and operations".[5] Although each of these types allows for the occasional use of the three types of "ta ctical" moves to be described in the next section, each has an inherent affinity with one of the types of moves.

• A discussion is a polemical exchange whose object is a well-circumscribed topic or problem. As the discussion develops, the contenders tend to acknowledge that the root of the problem is a mistake relating to some important concept or procedure within a well-defined field (even though they disagree regarding the nature of the mistake in question and about who commits it). Discussions allow for solutions, which consist in correcting the mistake thanks to the application of procedures accepted in the fi eld (e.g., proof, computation, repetition of experiments, etc.).
• A dispute is a polemical exchange which also seems to have as its object a well-defined divergence. But at no point do the contenders accept its definition as grounded in some mistake. Rather, it is rooted in differences of attitude, feelings, or pref erences. There are no mutually accepted procedures for deciding the dispute, that is, a dispute has no solution; at most it can dissolve or be dissolved.[6] Since "dissolution" is a form of closure that, ultimately, remains "external" bot h to the topic under dispute and to the participants' beliefs and attitudes, the underlying divergences tend to recur either in disputes over other versions of the same topic or in disputes over other topics. Some contenders see in the position held by th eir opponents and in their "stubborn imperviousness to rational argument" symptoms of an illness against which the only reasonable action to take is punishment or therapy, or disregard.[7]
• A controversy is a polemical exchange that occupies an intermediate position between discussion and dispute. It can begin with a specific problem, but it spreads quickly to other problems and reveals profound divergences. These involve both opposed at titudes and preferences and disagreements about the extant methods for problem solving. For this reason, the oppositions in question are not perceived simply as a matter of mistakes to be corrected, nor are there accepted procedures for deciding them -- w hich causes the continuation of controversies and sometimes their recurrence. However, they do not reduce to mere unsolvable conflicts of preferences. The contenders pile up arguments they believe increase the weight of their positions vis a vis the adver saries' objections, thereby leading, if not to deciding the matter in question, at least to tilting the "balance of reason" in their favor. Controversies are neither solved nor dissolved; they are, at best, resolved. Their resolution may consist in the a cknowledgment (by the contenders or by their community of reference) that enough weight has been accumulated in favor of one of the conflicting positions, or in the emergence (thanks to the controversy) of modified positions acceptable to the contenders, or simply in the mutual clarification of the nature of the differences at stake.

3.3 As mentioned above, actual polemical exchanges are rarely "pure" examples of one of these three types. One of the reasons for that is that the contenders' ways of perceiving and conducting a given exchange need not be identical. For example, in the co ntroversy about Newton's account of whiteness as a mixture of the other colors,[8] Newton's attitude was to consider it as a discussion, by asking his opponents to perform cautiously and without mistakes the experiment upon which he had b ased his account, as a way of solving the problem and finding the truth: "For this is to be decided not by discourse, but by new tryal of the Experiment" (Cohen, 1978, 153); "But this, I conceive, is enough to enforce it, and so to decide the controversy" (ibid., p. 131); "There are yet other Circumstances [i.e., other experiments, M.D.], by which the Truth might have been decided" (ibid., p. 130). However, his opponents -- especially Hooke and Huyghens -- refuse to see in Newton's experiment an experimen tum crucis, and consider the debate more as a controversy, for they reject the assumptions underlying Newton's hypothesis, and propose alternative hypotheses of their own, grounded on a different theoretical framework and supported by other experiments, a s capable to explain Newton's experimental results. Here is what Hooke says: "But, how certain soever I think myself of my hypothesis (which I did not take up without first trying some hundreds of experiments) yet I should be very glad to meet with one ex perimentum crucis from Mr. Newton, that should divorce me from it. But it is not that, which he so calls, will do the turn; for the same phaenmenon will be solved by my hypothesis, as well as by his, without any manner of difficulty or straining: nay, I w ill undertake to shew another hypothesis, differing from both his and mine, that shall do the same thing" (Cohen, 1978, 111).

Another example is the different attitudes of Searle and Derrida in their well-known debate. In this case, the contenders' attitudes seem to evolve in the course of the debate. As pointed out by Potte-Bonneville (1991, 231), at first both seem to be engaged in a controversy about the best way to interpret and pursue the philosophical project of Austin. From the outset it is clear that this topic should engender a controversy rather than a discussion, since the contenders are well aware that they belong to radically different philosophical paradigms, which question each other's most basic assumptions and methods. Searle, however, opts to treat the debate as a discussion, by attempting to show that Derrida's reading of Austin is simply mistaken, an d therefore refutable. He assumes thereby that the interpretation of a (philosophical) text is a decidable matter, and that his own decision-procedure is the one that must be applied in order to settle the issue. Derrida's response, in turn, consists in q uestioning the assumptions underlying Searle's decision-procedure, which -- according to him --need to be set aside in order to allow for a deeper understanding of Austin. At this point Derrida, although ostensively relying upon the "obvious" controversia l character of the opposition at stake, may be simply trying to impose upon his opponent the "rules" of his own (Derrida's) "language game". In other words, he too may be treating the debate, at this stage, as a discussion. When, finally, both contenders' attempts to transform the debate into a discussion abiding by each one's rules fail, both seem to shift to its perception as rather a hopeless dispute.

Just as participants in polemical exchanges often tend to perceive them as either dicussions or disputes, so too this dichotomy seems to exercise a powerful attraction upon theoreticians, leaving no room for the middle term -- controversy -- i n their accounts. Thus, in Kuhn's (1962) well-known schema, in periods of "normal science" disagreements between scientists are intra-paradigmatic, i.e., they arise against a background of shared decision procedures that regulate their "problem solving" a ctivity -- i.e., such disagreements instantiate our "discussion" category; the inter-paradigmatic conflicts, characteristic of periods of "extraordinary science", on the other hand, are often depicted as typical "disputes", in so far as their conduct and resolution depends, ultimately, upon preferences, public relations, interests, and power, rather than upon rational persuasion. At the other extreme (as far as positions in the philosophy of science are concerned), a similar dichotomic tendency to exclude the possibility of controversies can be found in Popper's (1991) attempt to sort out, in the polemics conducted by scientists involved in "scientific revolutions", between an "ideological" and a properly "scientific" component: the former clearly belongs to the category of "dispute", while the latter instantiates "discussion". Needless to say that, for Popper only the latter is of any value in an account of the "growth of knowledge".

I am persuaded that the neglect of the category "controversy" as a third alternative, between the strict rule-based notion of rationality that characterizes "discussion" and the conception of "dispute" as governed by extra-rational factors, ha s been a major setback for the history of ideas and for epistemology, depriving these (and other) disciplines from the possibility of identifying and developing an alternative model of rationality. This is why I recommend this category for special attenti on and careful empirical study.[9]

4. Three types of moves

Turning now to the "tactical" level, I will distinguish between three ideal types of moves employed in polemical exchanges. If "tactics" is related to "strategy" as means to ends, the question of the (in)dependence of the former vis-a-vis the latter arise s: Can any of the types of tactical moves be freely used in each of the three ideal types of polemical exchanges? Philosophy -- especially, but not only, ethics -- has devoted much effort to the analysis of the means-ends relationship. Personally, I am no t in favor of either strict dependence or strict independence, but rather of "relative (in)dependence", which might be described in terms of the existence of a "natural affinity" between certain types of means and certain types of ends. This is the kind o f conceptual relation I think obtains between the types of moves and of polemical exchanges discussed in this paper. Whether or not this is empirically the case is a task left for further empirical research

4.1 Like the typology presented in the preceding section, the present one does not purport to be neither exhaustive nor exclusive. Moves can been classified, for example, according to their "functional" roles, e.g., as "initiative" and "reactive" interven tions, and, within such broad classes, as "elaborations", "repairs", "digressions", "replicas", "counter-replicas", etc.[10] Or, in the medieval theory of disputations, a respondent's permissible moves are classified in terms of their se mantic relationship with the opponent's claims as "conceding", "denying", or "distinguishing". I do not dispute the usefulness of such levels of analysis. The typology here proposed, however, addresses another level of conceptualization, which captures p roperties of moves that apply across their "functional" and "semantic" roles. The main criteria employed in the present typology have to do with the immediate goal of the move, the nature of the means it employs to achieve its goal, the kind of "mechanism " it relies upon, the "force" with which it is supposed to achieve its goal, and its relationship with the "current state" of the (polemical) exchange.

• A proof is a move that purports to establish the truth of a proposition beyond reasonable doubt. For this purpose, it employs some inference rule that explicitly and recognizably leads from other propositions (i.e., the evidence -- which includes Toul min's (1962) "data" and "warrants") to the proposition to be proved. Both the validity of the inference rule and the truth of the evidence are assumed to be established, and therefore to be accepted by the addressee. The addressee is compelled (in so far as he behaves rationally) to accept also the conclusion. The possibility of proving a proposition is taken to show that it has successfully withstood a decisive test that guarantees its truth (or its high degree of probability).
• A stratagem is a move that purports to cause a relevant audience to (re)act in a certain way, by inducing it to believe that a proposition is true. It may make use of inference, but need not do so. If it does, neither the inference pattern is assumed to be valid nor the evidence true, but only "effective" vis-a-vis the intended addresse and audience. It may involve deception and dissimulation -- e.g., by manipulating the "current state" and "current demands" of the exchange. The causation involved nee d not be explicit and recognizable by the audience, provided it achieves its intended effect, namely to let its user "win the day" (at least momentarily) in the eyes of the relevant audience (which may or may not include the interlocutor). Hence the curr ent meaning of this word as "any artifice or trick; a device or scheme for obtaining an advantage" (OED). The particular kind of "force" of this move lies not in compelling the addressee to hold the intended belief or to perform the desired action, but ra ther in rendering him "speechless", i.e., unable to react with a satisfactory counter-move.
• An argument is a move that purports to persuade the addressee to believe that a proposition is true. Like stratagems and unlike proofs, arguments are not directly concerned with truth, but with belief. Unlike stratagems, arguments seek to achieve thei r effect by providing recognizable reasons for inducing in the addressee the desired belief. Unlike proofs, however, these reasons need not be based on valid inference patterns and truthful evidence, which are presumed to be accepted by the addressee; the y must take into account which propositions the addressee actually accepts (or is likely to accept) as evidence, and which inference patterns are likely to persuade her. Arguments, although not compelling the addressee to accept their conclusion, put her under some sort of obligation to do so -- an obligation that presumably stems from social norms, e.g., those of communicative cooperation.

4.2 Obviously, clarification and exemplification of the above descriptions are needed, especially in the light of possible terminological confusion and of the difficulties in identifying "pure" examples of these types of moves in actual polemical exchange s.

4.2.1 The term proof, as employed here, does not refer only to formal deductive demonstrations, as in logic and (parts of) mathematics. It applies also to the use of other forms of inference (e.g., inductive, non-monotonic, presumptive) that are supposed to establish the truth (or the high degree of probability) of a statement. Nor does a proof -- in the present sense -- necessarily rely on evidence that has been itself proved: the appeal to experiment, observation, testimony, common sense, etc., wheneve r these are presented as directly relevant to establishing the truth of a statement, counts as a move pertaining to the category "proof". What is important in this kind of move is the ostensive reliance upon a procedural process of justification whose "ob jectivity" resides in its being procedural, i.e., "neutral" vis-a-vis the beliefs and interests of the contenders. This is why proof is deemed able to bypass such beliefs and address truth, so to speak, "directly". Furthermore, it acquires additional pole mical weight thanks to the presumption that truth must be the decisive factor in determining belief. The most efficient countermoves to proofs are "counter-proofs" that question either the reliability of the evidence presented (e.g., by pointing out incon sistencies in a testimony) or of the inferential procedure employed (e.g., the method of counter-examples in logic). The use of both kinds of moves in polemical exchanges is widespread. Nevertheless, they are only decisive -- as their users expect them to be -- in the context of "discussions", where a decision procedure which is assumed not to be questionable gives them the necessary "backing" (to use another of Toulmin's concepts). Hence the special affinity between proofs and discussions.

4.2.2 I have borrowed the term stratagem (Ger. Kunstgriff) from Schopenhauer (1942),[11] who has a thoroughly negative view of this kind of move. He compares stratagems to feints in fencing, and describes them as the dishonest "tricks, dogdges, and chicanery" to which contenders resort with the sole purpose of "being right", regardless of whether their thesis is true or false. When arguing with an opponent that makes use of such tricks, he says, one "no longer has to deal with his intel lect, but with the radical part of the man, his will, to which the only thing that matters is that he ultimately triumphs either per fas or per nefas [by hook or by crook]" (1974, 25). In providing a formal anatomy of these moves, which should serve as th e backbone of "the science of [eristic] Dialectic", and in exemplifying 38 stratagems -- many of which are described in Aristotle's Topica (especially in Book VIII) -- along with appropriate "counter-stratagems", Schopenhauer's aim is to grant the (honest ) debater a tool for easily recognising and defeating such tricks (1942, 10-11). Among his 38 stratagems, we find, along with well-known fallacies, moves that are supposed to have, say, a direct and explicit causal effect upon the opponent's beliefs, such as:

Extension. "This consists in carrying your opponent's proposition beyond its natural limits; in giving it as general a signification ... as possible, so as to exaggerate it; and, on the other hand, in giving your own proposition as restricted a sense ... as you can, because the more general a statement becomes, the more numerous are the objections to which it is open." (Schopenhauer, 1942, 13);

Diversion. "If you find that you are being worsted ... you can suddenly begin to talk of something else, as though it had a bearing on the matter in dispute, and afforded an argument against your opponent" (Schopenhauer, 1942, 29-30);[12]

as well as moves that are intended to provoke certain reactions, which will then make the opponent's position (and beliefs) vulnerable, e.g.:

[Irritation]. "Contradiction and contention irritate a man into exaggerating his statement. By contradicting your opponent you may drive him into extending beyond its proper limits a statement which, at all events within those limits and in itself, is true" (Schopenhauer, 1942, 26).

To the latter kind, we might add Aristotle's

[Confidence building]. "You should also, yourself sometimes, bring an objection against yourself; for answerers are unsuspicious when dealing with those who appear to them to be arguing fairly" (Aristotle, 1976, 156b 17),

as well as Leibniz's

[Compensation]. "Il est quelquefois utile que nous souffrions qu'on nous fasse quelque tort dans une matiere de peu de consequence, car si quelque grand y a trempe, cela luy donnera quelque penchant (s'il est d'un bon naturel) a nous f aire du bien dans quelque autre rencontre, et on peut menager la chose en sorte, que la seconde soit plus importante a nous, que la premiere" [GRUA, 701-702].[13]

Stratagems of the latter type are mainly "offensive" moves, "traps" of sorts, quite similar to those described in game theory as "strategic moves", namely "moves that induce the other player to choose in one's favor" (Schelling 1960, 2 2). But they can also be used, like those of the former type, defensively. Both of these, as well as other types of stratagems abound in polemical exchanges (see Dascal and Cremaschi, forthcoming, for a sample). Empirical research should find out whether they are more frequent in the exchanges belonging to the category "dispute", with which they seem to have a natural affinity, in so far as they share the goal of winning per fas or per nefas.

4.2.3 The term argument, finally, is here employed in the sense it has in Perelman's nouvelle rhetorique, namely as a kind of move intended to modify beliefs by means of reasons which are neither logically compelling nor impersonal.[14]Arguments, in this sense, differ from proofs in that they may be logically invalid (e.g., the slippery slope, the ad verecundiam) or else may consist in showing the insufficiency of logical validity (e.g., the petitio principii). A "slippery slope" argu ment consists in pointing out that A would lead to B, and then to C, D, ... N, through a causal chain, and to claim that one should prevent A, because N is an undesireble consequence. In politics, this argument is known as the "domino effect". Logically i t is invalid, because the causal chain can be interrupted anywhere, not just at its initial point, as the argument presupposes. The Vietnam war is a counter-example to this argument. Nevertheless, it is a rationally persuasive argument, which is regularly used in deliberations, and whose persuasive weight depends upon the addressee's estimate of the cost of interrupting the causal chain at different points. A petitio principii charge, on the other hand, does not question the logical validity of the oppone nt's move (what could be more valid than "p, therore p). It simply points out the uselessness of such a move in order to establish the truth of a proposition.[15] A petitio charge is, in a sense, an example of an ad hominem argument, bel onging to the subset of tu quoque arguments. Though generally (though by no means universally) considered fallacious on the grounds that the (circumstances of the) person making a claim are not relevant to the truth of that claim, ad hominem arguments can be (rationally) persuasive: if you wouldn't trust a man to buy a car from him, it is reasonable for you to see in this a reason not to vote for him for president, if you think honesty is a quality a president should have.

The affinity between the "argument" type of move and the "controversy" type of polemical exchange lies in the fact that the former fits the latter's most typical features. First, the controversy's openness, namely, the fact that in a controvers y everything is up for grabs, no "sacred" assumptions or methods being preserved from unlimited mutual questioning. Arguments are both good tools for that purpose (since they go beyond purely logical considerations, and thus allow to question what the for mer take for granted) and also excellent targets (in so far as, when used by the adversary to ground her position, their "quasi-validity" makes them easy prey to orthodox logical hunting practices). Second, the fact that, even though in a controversy all is up for grabs, not "anything goes", i.e., some norms are respected and the ways of acting upon the opponent's beliefs are constrained.

4.3 The identification of a move performed in a polemical exchange as belonging to a given ideal type is no easy matter, especially due to co-text dependency. One might think that a proof is a proof, an argument an argument, and a stratagem a stratagem re gardless of the "polemical demand" to which they respond and of their influence upon the ensuing interventions. But this is not the case. In the case of moves, as in the case of the interpretation of other linguistic elements, we may have to distinguish b etween, say, the "literal" and the "actual". Just as an utterance which has the syntactic form of a question and, thus, whose sentence-meaning is that of a question, may serve to perform other speech acts, so too a move that is "literally" a proof may be actually used as a diversion stratagem or as a non-compelling argument.[16] Consequently, the empirical study of these moves requires both a syntactic-semantic and a pragmatic component. The problem of identification is compounded by the fact that, whereas proofs and arguments "advertise" themselves as such by the use of explicit linguistic markers, stratagems rather "disguise" themselves as proofs or arguments.

5. Concluding remarks

I have presented my two trichotomies on the assumption that it is possible to distinguish clearly between their respective levels of analysis. I do believe that such a distinction is necessary and fruitful, although certainly not easy to pinpoint. Aristot le, for one, was aware of the need to make the distinction and proposed trichotomies quite similar to those developed here. On the one hand, he distinguished three types of polemical exchanges,[17] "demonstrative", "dialectical", and "co ntentious" (Aristotle, 1976, 100a 25 - 101a 5). On the other, he coined special terms for characterizing the moves typical of each of these exchanges, namely "philosopheme", "epichireme" and "sophism", respectively (ibid., 162a 15). But he did not make cl ear the relationship between the two levels. For he apparently didn't take into account the fact that none of the types of exchanges neither consists nor can consist exclusively of the types of moves to which they bear the closest "affinity". No wonder th at many of the moves that he lists under "dialectical" exchanges turn out to be stratagems. Unfortunately, I cannot pretend to have been more successful than Aristotle, in this respect, and instead of attempting to solve this problem here, I will add a fu rther source of difficulty for its solution.

We have already noticed that contenders often tend to conduct and interpret the polemical exchanges in which they are involved as belonging to one of the ideal types (usually limiting their choice to two of them, discussions and disputes). Such interpretations -- as many other meta-discursive claims and attitudes -- may reflect "tactical" choices of a "macro" arena: a contender may choose, say, the "discussion" arena in order to take (local or global) advantage of his mastery of this special kind of battle and weaponry. On the other hand, "micro" can also become "macro", and "tactics" can become "strategy". For example, a repeatedly used stratagem may become definitory of a contender's goal in an exchange, thereby acquiring strategic import. Thus, the counterpart of the "extension" stratagem, namely, restricting the scope of one's claim so as to make it invulnerable to a given objection, may become what I have called the "insulation strategy" (cf. Dascal 1990b), when used, e.g., to protect on e's claims to knowledge from every possible sceptical objection.

This last example permits me to conclude by alluding once more to my hidden agenda. One of the strongest manifestations of scepticism against the existence of controversies as here defined, is Schopenhauer's. For him there cannot really be a "g ood" dialectic, for as soon as the best intentioned person engages in a debate, his will takes over his intellect, and all he cares for is to win at any cost. But even Schopenhauer acknowledges that this "natural baseness of human nature", this "innate va nity ... which will not suffer us to allow that our first position was wrong and our adversary's right" (Schopenhauer, 1942, 4), has a positive contribution vis-a-vis the discovery and preservation of truth: "Should we abandon our position at once, we may discover later on that we were right after all" (ibid., p. 5). He also admits that we cannot, in studying debates, assume that we are able to "separate actual from apparent truth, since even the disputants are not certain about it beforehand" (ibid., p. 13). Together, these two remarks suggest that, in spite of his stern scepticism, he acknowledges that there are polemics which can contribute to the "growth of knowledge", at least as their unintended effect. Contra Schopenhauer, and along with Aristotle and Perelman, I believe this is not an unintended effect, but the result of the special kind of rationality inherent to "controversy", which makes it different from both "discussion" and "dispute". Unlike them, however, I think controversy and its typical move, argument, have not been sufficiently studied as an empirical dialogical phenomenon, so as to reveal the precise nature of the rationality embedded in their use. Without such empirical support, the debate between optimists like Perelman and pessimi sts like Schopenhauer will remain a mere dispute; with empirical support, it stands a chance of becoming a rewarding controversy.

Notes

1. The opposition between these two kinds of methodological preferences often underlies historically important controversies (see Cremaschi and Dascal, 1996, 1998; Dascal and Cremaschi, forthcoming; Gil, 1985). In dialogue studies, this o pposition is most evident in the ongoing dispute between ethnomedologists and speech act theorists (see, for instance, Schegloff's article and Searle's rejoinder in Searle et al., 1992).

2. Mainly in interaction with the members of the research group "Leibniz the Polemicist", in a very fruitful year (1994/5)at the Institute for Advanced Studies at the Hebrew University of Jerusalem. Sergio Cremaschi, Gideon Freude nthal, Fernando Gil, Alan Gross, Massimo Mugnai, Carl Posy, Quintin Racionero, Elhanan Yakira, and all those who attended our seminars have contributed graciously to both the insights and mistakes you may find in what follows . For some initial results of our work, see Dascal, 1995b, 1995c; Cremaschi and Dascal, 1996, 1998; Dascal and Cremaschi, forthcoming; Dascal and Gross, forthcoming.

3. A conference on this topic was held in Geneva in 1995. For how some of these levels bear on the topic of controversies, see Dascal, 1995b. For other relevant material, see Fritz, 1994, 1995; Jacques, 1991; Mann and Thompson, 19 88; Dascal, 1992; Roulet, 1995.

4. For the types and uses of co-text and context, see Dascal and Weizman, 1987. For descriptions of such types and uses in controversies, see Dascal, 1990a; Cremaschi and Dascal, 1998; Dascal and Cremaschi, forthcoming.

5. "Strategy - The art of the commander-in-chief; the art of projecting and directing the larger military movements and operations of a campain. Generally distinguished from tactics, which is the art of handling forces in battle o n in the immediate presence of the enemy" (Oxford English Dictionary). "Strategie: umfassende [vorbereitende] Plannung eines Krieges unter Einbeziehung aller wesentlichen Faktoren" (Der Grosse Duden, Bd. 10). "Stratégie: partie de l'art militaire qui trai te de la direction d'ensemble ... dans la conduite de la guerre" (Nouveau Petit Larousse).

6. A dispute can be referred to some instituted authority, such as a tribunal, which may decide in favor of one of the parties. But in these cases, the conflict of opinions or attitudes is not resolved but merely subdued, since, a s pointed out by Leibniz, no one has the power to force the other to forget or to attend -- both indispensable conditions for having the power to make the other change her opinion (VOR, p. 19).

7. "It is not necessary to examine every problem and every thesis but only one about which doubt might be felt by the kind of person who requires to be argued with and does not need castigation or lack perception. For those who fe el doubt whether or not the gods ought to be honoured and parents loved, need castigation, while those who doubt whether snow is white or not, lack perception" (Aristotle, 1976, 105a 3). It is possible to read this passage as suggesting that debates about both kinds of questions are useless -- the former because it can only be a "dispute" with an obvious "external" solution, and the latter because it is a "discussion" with an obvious "internal" decision procedure.

8. I am grateful to Alan Gross for bringing this controversy to my attention.

9. In Dascal (1995), in addition to having argued at length this point, I have tried to single out some of the most important epistemologically relevant characteristics of controversies. An English version of the paper in question can be read in my website: http://www.tau.ac.il/humanities/hci/vip/index.htmll.

10. Functional characterizations of this type are shared by a wide range of approaches to dialogue analysis: ethnometodology, speech acts theory, action theory, rhetorical structure theory, the Geneva approach, dialogical logic, etc.

11. Around 1830, Schopenhauer wrote a little treatise, without title, which he never published. This treatise was posthumously published in 1864 and, more recently, in 1983, with the title Eristische Dialektik oder Die Kunst, Rec ht zu behalten, in 38 Kunstgriffen dargestellt (Zurich). My references are to the 1942 English translation.The introductory passages and some of the stratagems are reproduced in the chapter "Logic and dialectics" of his Parerga and Paralipomena. I am grat eful to Massimo Mugnai for having called my attention to Schopenhauer's little treatise and to have presented me with a copy of the Italian translation, which contains an excellent essay by Franco Volpi (Milano, 1991).

12. Diversion is, in a sense, a form of digression. But in order to be effective, it should not contain explicit "digression markers", which would enable the opponent to dismiss easily their relevance (on different types of digre ssions, see Dascal and Katriel, 1979).

13. Leibniz himself employed this stratagem in his correspondence avec Arnauld (cf. Dascal 1995a).

14. Perelman extends the scope of his theory of argumentation beyond the field of dialogical exchanges, as defined here, for he says that it will study even arguments one addresses oneself, in silent deliberation (Perelman 1977, 19). I do agree with him that there is room and need for research on, say, "polemical soliloquy", which would, among other things, take into account the polemical nature of the "polyphony" not only of our speech, but also of our thought. It would be inter esting to find out, among other things, whether the three ideal types of moves here distinguished could occur in inner polemics with oneself. Can one, for example, apply stratagems to oneself? In so far as there is such a thing as self-deception, apparent ly the answer must be yes (see, in this connection, the interesting collection of essays compiled by Elster (ed.), 1985). With all its interest, that part of the study of inner argumentation that involves language use belongs to the domain of what I have proposed to call "psychopragmatics", rather than to "sociopragmatics", which deals with the outer uses of language (cf. Dascal 1983), to which our present topic belongs.

15. Passmore (1961) argues that most philosophical arguments are of this sort, namely, that they are not strictly formal, but at most "quasi-formal".

16. It is more difficult to imagine the converse. If a stratagem is successfully "used" in a given context as a proof, it does not thereby become a proof, but remains a (successful) stratagem. There seems to be a presumption abou t the scalar ordering of the "forces" of the three types, which constrains their interchangeability in use.

17. He calls them "reasonings", but in the context of this passage, this term refers to ways of conducting a debate, where "questions" and "answers" are exchanged.

References

Aristotle (1976), Topica (transl. E.S. Forster). London, Cambridge,MA.

Attardo, S. (1997), Competition and Cooperation: Beyond Gricean Pragmatics. In Pragmatics & Cognition 5, 19-48.

Blum-Kulka, S. (1992). The Metapragmatics of Politeness in Israeli Society. In: Watts, J., Ide, S., Ehlich, K. (eds.), Politeness in Language: Studies in its History, Theory and Practice. Berlin, 255-281.

Cohen, A. and Dascal, M. (eds.) (1989). The Institution of Philosophy: A Discipline in Crisis?. La Salle, Illinois.

Cohen, B. (ed.) (1978), Isaac Newton's Papers & Letters on Natural Philosophy, 2nd. ed. Cambridge, MA.

Cremaschi, S. and Dascal, M. (1996), Malthus and Ricardo on Economic Method. In: History of Political Economy 28, 475-511.

Cremaschi, S. and Dascal, M. (1998). Persuasion and Argument in the Malthus-Ricardo Correspondence. In: Research in the History of Economic Theory and Method 16.

Dascal, M. (1977), Conversational Relevance. In: Journal of Pragmatics 1, 309-327.

Dascal, M. (1983), Pragmatics and the Philosophy of Mind, vol. 1. Amsterdam.

Dascal, M. (1989), Controversies as Quasi-Dialogues. In: Weigand, E. and Hundsnurcher, F. (eds.), Dialoganalyse II, Band 1. Tubingen, 147-159.

Dascal, M. (1990a), The Controversy about Ideas and the Ideas about Controversy. In: Gil (ed.), 61-100.

Dascal, M. (1990b), La Arrogancia de la Razón. In: Isegoría 2, 75-103.

Dascal, M. (1992), On the Pragmatic Structure of Conversation. In: Searle, J. R. et al., 35-56.

Dascal, M. (1994), Speech Act Theory and Gricean Pragmatics. In: Tsohatzidis (ed.), 323-334.

Dascal, M. (1995a), Strategies of Dispute and Ethics: Du Tort and La Place d'Autruy. In: Proceedings of the VI. Internationaler Leibniz-Kongress, vol. 2. Hannover, 108-116.

Dascal, M. (1995b), Observations sur la Dynamique des Controverses. In: Cahiers de Linguistique Francaise 17 (2ème Partie), 99-121.

Dascal, M. (1995c), Epistemología, Controversias y Pragmática. In: Isegoría 12, 8-43..

Dascal, M. (1996), La Balanza de la Razón. In: Nudler, O. (ed.), La Racionalidad: Su Poder y sus Limites. Buenos Aires, Barcelona, Mexico, 363-381.

Dascal, M. (ed.) (1985), Dialogue: An Inter-Disciplinary Approach. Amsterdam.

Dascal, M. (ed.) (1991), Cultural Relativism and Philosophy: North-American and Latin-American Perspectives. Leiden

Dascal, M. and Cremaschi, S. (forthcoming), The Malthus-Ricardo Correspondence: Sequential Structure, Argumentative Strategies, and Rationality.

Dascal, M. and Gross, A. (forthcoming), The Marriage between Pragmatics and Rhetoric.

Dascal, M. and Katriel, T. (1977), Between Semantics and Pragmatics: The Two Types of `but' - Hebrew `aval' and `ela'. In: Theoretical Linguistics 4, 143-172.

Dascal, M. and Katriel, T. (1979), Digressions: A Study in Conversational Coherence". In: Poetics and Theory of Literature 4, 203-232.

Dascal, M. and Weizman, E. (1987), Contextual Exploitation of Interpretation Clues in Text Understanding: An Integrated Model. In: Verschueren, J./Bertucelli-Papi, M. (eds.), The Pragmatic Perspective. Amsterdam, 31-46.

van Eemeren, F. and Grootendorst, R. (19..1984), Speech Acts in Argumentative Discussions. Dordrecht

Dominicy, M. (1994), How Scientists Argue. Two Case Studies. In: Parret, H. (ed.), Pretending to Communicate. Berlin, New York, 89-103.

Elster, J. (ed.), (1985), The Multiple Self. Cambridge.

Engelhardt Jr., H. T. and Caplan, A. L. (eds.) (1987), Scientific Controversies: Case Studies in the Resolution and Closure of Disputes in Science and Technology. Cambridge.

Fritz, G. (1994), Grundlagen der Dialogorganisation. In: Fritz, G./Hundsnurcher, F. (eds.), Handbuch der Dialoganalyse. Tubingen, 177-202.

Fritz, G. (1995), Topics in the History of Dialogue Forms. In: Jucker, A. H. (ed.), Historical Pragmatics. Amsterdam, 469-498.

Gil, F. (1985), Science and Controversy. In: Dascal (ed.), 353-366.

Gil, F. (ed.), (1990), Scientific and Philosophical Controversies. Lisbon.

Granger, G-G. (1985), Discussing or Convincing: An Approach towards a Pragmatical Study of the Languages of Science. In: Dascal (ed.), 339-352.

Gross, A. G. (1990), The Rhetoric of Science. Cambridge, MA.

Jacques, F. (1985), Du Dialogisme a la Forme Dialoguée. Sur les Fondements de l'Approche Pragmatique. In: Dascal (ed.), 27-56.

Jacques, F. (1991), Argumentation et Stratégies Discursives. In: Lempereur, A. (ed.), L'Argumentation Liege, 153-171.

Kerbrat-Orecchioni, C. (1994), Les Interactions Verbales, tome III, Paris.

Kuhn, T. S. (1962). The Structure of Scientific Revolutions. Chicago.

Leibniz, G. W. [= GRUA], Textes inédits d'après la bibliothèque provinciale de Hanovre, ed. G. Grua (1948). Paris.

Leibniz, G. W. [= VOR], Vorausedition zur Reihe VI - Philosophische Schriften. Munster (1982-1990).

Mann, W. C. and Thompson, S. A. (1988), Rhetorical Structure Theory: Toward a Functional Theory of Text Organization. In: Text 8, 243-281.

McCloskey, I. (1985), The Rhetoric of Economics. Madison, WI.

McCloskey, I. (1994), How Economists Persuade. In: Journal of Economic Methodology 1, 15-32.

McEvoy, S. (1995), L'Invention Defensive, Paris.

Passmore, J. (1961). Philosophical Reasoning. London.

Pera, M. (1991), Scienza e Retorica. Roma, Bari.

Perelman, C. (1977), L'Empire Rhétorique: Rhétorique et Argumentation. Paris.

Perelman, C. and Olbrechts-Tyteca, L. (1970), Traité de l'Argumentation: La Nouvelle Rhétorique, Bruxelles.

Potte-Bonneville, M. (1992), La Violence dans le Débat Philosophique: La Controverse Derrida/Searle. In: D. Vernant (ed.), Du Dialogue (= Recherches sur la Philosophie et le Langage 14), Paris, 219-248.

Rolf, E. (1989), How to Generalize Grice's Theory of Conversation. In: Manuscrito 12(1), 55-69.

Roulet, E. (1989). Une Forme peu Etudiée d'Echange Agonal: La Controverse. In: Cahiers de Praxématique 13, 7-18.

Roulet, E. (1995). L'Analyse du Dialogue dans une Approche Modulaire des Structures du Discours: l'Exemple du Dialogue Romanesque. In: Hundsnurscher, F. and Weigand, E. (eds.), Beitrage zur Dialogforschung, vol. 8. Tubingen, 1-34.

Schelling, T. C. (1960), The Strategy of Conflict. Cambridge, MA.

Schopenhauer, A. (1942), The Art of Controversy. In: Complete Essays of Schopenhauer (transl. Saunders, T. C. New York.

Schopenhauer, A. (1974), Parerga and Paralipomena, vol. 2 (transl. E.F.J. Payne), Oxford.

Searle, J. R. et al. (1992), (On) Searle on Conversation. Amsterdam.

Tsohatzidis, S.L. (ed.) (1994), Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives. London.

Toulmin, S. E. (1969), The Uses of Argument. Cambridge.

Vernant, D. (1996), Dialectique, Forme Dialogale et Dialogique. In Actes du Colloque "Le Dialogique".

Weigand, E. (1996), The State of the Art in Speech Act Theory. In: Pragmatics & Cognition 4: 367-406.

Wierzbicka, A. (1993), 'Cultural Scripts' as a Key to Cross-Cultural Pragmatics. Paper delivered at the 4th International Pragmatics Conference. Kobe.