Comparative Anthropology and Rhetorics in Cultures.

[Published in Norms in Argumentation (1989). Dordrecht: Foris, pp. 195–211.]


Perhaps it is best that we begin on a personal note. Neither of the two authors of this article is a ‘specialist’ in the domain of argumentation theory. Even though we have picked up smatterings of logic, we are not professionals working on either informal or formal logic. One of us is a practicing anthropologist, while the other is oriented towards comparative anthropology. And yet, we are invited to speak to an audience comprising of people specializing in dialogue logics, argumentation theory, psychology and such like. If the invitation to present a paper is not mistake or an oversight on the part of the organizers of this colloquium, then it could only mean that there is an expectation regarding our contributions: anthropologists have something to say about rhetorics, argumentations and their norms, which is of relevance to specialists like you.

Is this a reasonable expectation? Anthropologists, normally speaking, study and describe the habits, life-styles and beliefs of peoples and cultures. They may, if we restrict ourselves to the doubly normative theme of this gathering, describe the ways people argue in other cultures, outline some of the norms (if and where they exist) that guide discussions and so on. Specialists like you, on the other hand, cogitate about what an argumentation ought to be like, what its norms ought to be, whether the latter are defensible and many other similar questions.

How, then, are we expected to contribute to the colloquium? Perhaps, we could provide you with some reflections about the Tiv (which we shall do), or some facts about the way people resolve conflicts in Timbuctoo (which we shall not); or a colorful description of the curious habit of resolving disputes by bashing the opponent’s head, which the inhabitants of Ruritania seem to like so much. You may listen to all of this with mild curiosity or with scarcely disguised impatience. Either way, it does not mean much. Anthropologists appear to be at a disadvantage: we cannot hope to contribute substantially to the furthering of insights into the norms of argumentation.

At this point, one would expect a reasonable person to down his/her tools and take leave of this distinguished company mustering as much dignity as possible. As you will no doubt have noticed, we are not doing so.

Stubbornness apart, there is a deeper reason for not travelling that route. And that has to do with the fact that the theme of this conference reflects a set of highly culture-specific assumptions. As anthropologists we think that this set carries a mantle of universality which is rather illusory.

What we would like to do, in the course of what follows, is to share with you some of our reasons for thinking so. If we succeed in persuading you to reflect about our arguments seriously, then the expectations of both the organizers and ourselves will have been fulfilled: maybe, just maybe, anthropologists can contribute substantially in normative domains after all.

1. Norms of Argumentation: On Presuppositions

We would like to address ourselves to the task by first identifying some of the properties that are allegedly shared by argumentative interactions:

(a) An argumentation takes place between two parties, whenever their avowed opinions conflict;

(b) What makes something into an argument is the semantic relation between the utterances, and not the nature of either the speaker or the audience;

(c) Insofar as a speaker has a specific goal, it is to persuade or convince an audience to accept a position that s/he has put forward;

(d) This goal can be reached only if the norms and values held by both the speaker and the audience at least partially overlap.

While the above four are empirical properties of argumentative interactions, in almost all theories of argumentation there is also a normative assumption at a meta-level:

(e) It is rational to settle disputes argumentatively, and it is preferable to be rational rather than irrational. Consequently argumentative activity ought to be a norm-guided activity, where the norms are binding on all those who are participants in an argumentation.

While these assumptions are common to all argumentation theories that we know of, what divides them is the answer they give to the following two overlapping questions: what ought to epistemically count as a reasonable discussion? What ought to be the norms that guide a rational argumentation?

The first question that we should ask is about the status of the ‘properties’ of (a) to (d): Are they factual statements making claims about teh way people ‘argue’? Or are they ‘conceptual’ conditions that lay down criteria for considering only some verbal exchanges as argumentations? The second possibility is not very interesting because these properties get built into a very stipulative definition of argumentation: An argumentative situation is, by definition, one where there exists a conflict of avowed opinions; an argumentive process is, by definition, one of advancing, defending, retracting some opinions and so forth. While such stipulative definitions surely have their place in building formal systems (e.g. Barth and Krabbe, 1982), or even in circumscribing the domain of enquiry in a particular way (e.g. van Eemeren and Grootendorst, 1982), they are not cognitively productive. We need to be able to raise the question whether interactions involving the settling of disputes, without taking recourse to ‘violence’ or ‘fiat’, do exhibit the above-mentioned properties everywhere.

Therefore, let us not waste time fighting over the definitional ‘problem’. We will henceforth assume that theorists of argumentation are making the more interesting claim, viz., that wherever ‘rational’ procedures are used to settle disputes, the verbal interactions exhibit these empirical properties. In the subsequent sections, we will see whether the first two claims are true. Afterwards, we could assess the value of the normative assumptions.

1.1 The structure of the paper

It is not controversial to suggest that conversational activity is intimately connected to argumentative interactions. Therefore, it would be reasonable to accept the idea that theories of conversation and theories of argumentation are closely related too. Working on the assumption that there is a close relationship between the two, we will look at some cross-cultural scenarios in the second section. We show how our theories of conversations generate implausible conclusions if asked to account for these scenarios. These conclusions, we shall argue there, arise due to culture-specific assumptions made by theories of conversation and try to explicate some of them.

In the third section, we suggest that arguing is a specific way of conversing and that the assumptions of theories of conversations are also the premises of argumentation theories. Consequently, the implausible conclusions encountered in the previous section hold good here, too. But in the domain of argumentation, the implausibility is less obvious than it is with respect to conversation. We will look into the reasons for this appearance by analyzing two models that inspire theories of argumentation: legal disputes and scientific discussions. We shall how both models fail us–albeit in two different ways.

In the fourth section, we localize the reason for this failure in terms of a culture-specific approach to answering some particular issue. We shall identify the issue for what it is. This should enable us to answer the normative question regarding the rationality and superiority of argumentation by comparing it with the way other cultures solve them.

The next obvious task is to outline how other cultures solve them. But here, we bounce against the page-limit. Consequently, we are forced to issue a promissory note to return to this theme elsewhere. We end on an invitation to theorists of argumentation to work along with anthropologists before building their grand, ‘unifying’ theories. Though the paper is critical in tone, it is friendly in nature. Our hope, then, is that it will be so read and understood too.

2. Conversations and Conversational Theories

It is legitimate to claim that one does not need to know what a ‘conversation’ is in order to have a conversation. Therefore, whatever else we may want to say, it would be very difficult to maintain intuitively that there is no conversation in other cultures. Equally difficult would be to hold that some cultures are inferior to others regarding their practice of conversational activity. In other words, our intuitions do not allow us to assert either a factual or a normative claim, which would either deny the existence of conversation in other cultures or suggest that members from some cultures are incompetent conversationalists. Let us see how our theories of conversations help us relate to this intuition.

2.1 Some scenarios

1. As a white man and foreigner, you are a teacher in an English class in China. All of a sudden, one fine day, the Chinese director of the school drops in during the afternoon and requests to speak to you. After nearly a quarter of an hour, during which time the director has praised your work and qualities sky-high, and just when you thought that he was coming to the real point of his visit–he politely takes leave. You know that what he said was not ‘what he came to say’; nevertheless, you also realize that it has been said.

2. As a white man and foreigner, you are travelling somewhere in Asia (say, Thailand). Intending to take the public transport, you go to a bus stop and enquire a native passerby when the next bus is due. The native consults his wristwatch, assures you that you could expect it any minute and moves along. An hour and a half later, while still waiting for the same bus, you spy the same native returning from his errand. Furious, you collar him and ask him the same question. Though a bit surprised at your rage, when the native gives the same answer, you realize that you have been had: that guy had no more clue about the bus timings than you had!

3. As a white man and foreigner, you are travelling on a train in India. Seated next to you is a Brahmin, eager to strike up a conversation and impress you with his learning. Your sensibility is shocked by the poverty in India, and by the existence of ‘untouchability’ in that culture. Outraged and incensed by the indifference that Indians show to poverty and suffering around them, you quiz this Brahmin about why he is unmoved. The Brahmin assures you with great solemnity that most of the ‘beggars’ you have seen are really no beggars at all, but, to the contrary, rather wealthy. Many, in fact, are wealthier than either of you. In support of this fact, the Brahmin tells you a tale (an anecdote) about some beggars who turned about to be the greedy rich in disguise! Regarding the lot of those beggars who are not cheaters, he merely shrugs his shoulders and says that it is their karma. You are utterly shocked by both the flimsiness of his ‘explanation’ and by the total lack of humanity in that Brahmin. Both of you know that the ‘anecdote’ was a figment of his imagination.

Each of the three scenarios is a part of our folklore: almost each traveler and tourist who has been to Asia will have some such tale to tell. Before we reflect about any of the above three, allow us to shift our glance to the North American Indians. Here is an anecdote from the fieldwork days of one of the coauthors.

4. A student nurse came to the center of the Navajo reservation in a big and rather new American car. Her whole appearance radiated a “clean” upper middle class white background. On her arrival in Lukachukai, Arizona, she immediately went up to our interpreter (F.H.), urging him to address the monlingual medicine man (R.W.). She declared that she wanted to collect information from the latter about “diabetes” in the traditional medicinal lore. R.W’s reaction was as follows: he did not know anything about diabetes prior to the contact with the Whites and that he did not have time that day. “May be she should come back tomorrow”. The next day, upon her return, she was told that he did not have time for her because the “chickens have to be fed and watered”. The third day, she was told that there was no time because “the sheep have to be fed”. The Navajo, as it happens, do not feed either their sheep or chickens, let alone water them. As far as we know, she was given no information on those questions with which she approached the Navajo.

2.2 Conversational theories and their cultural assumptions

There will not be much of a controversy, we suppose, if we identify all three as conversations. That is to say, two or more people are involved; there was some kind of a question and answer, some exchange of information in some context or the other. We do not suppose that one needs to be knowledgeable about conversation theory in order to identify that a conversation has taken place. And yet, were we to look at some of the existing theories about what conversation is ( e.g. Grice 1975, Sperber and Wilson 1986), we are led to some nontrivial and startling conclusions: it would appear either that there has been no conversation at all or that in each of them some or other ‘maxim’ has been violated.

In and of itself, this piece of knowledge is not startling. But it does become so when we take into account that these types of conversation are not exceptions but standards, and that they all are examples from non-western cultures! (The ‘white man and foreigner’ bit is meant to help you relate to these scenarios better; perhaps, you have had some such experience yourself or you have heard it recounted by someone else. It has no other special significance in this context.)

In the first case, for example, the Chinese director was neither ‘brief’ nor relevant; in the second, the native was violating the ‘cooperation principle’ and was lying; in the third the same holds true too.

Our theories of conversation, then, generate the conclusions that other cultures either do not know how to have a conversation or that they always violate the norms of conversation or, even more crudely put, most of the time only westerners are competent conversationalists. This is a logical possibility, to be sure; but there is something utterly implausible about such conclusions. Our conversational ‘theories’ generate such conclusions because of some assumptions they make. Let me try and identify them.

The first assumption: Common to all (or nearly all) theories of conversation is the assumption that conversational relations (like, say, that of ‘relevance’) hold between beliefs. That is to say, the semantic relations between propositions (or whatever else you want to use in its stead) constitute the main area of inquiry. The not-so-recent ‘pragmatic’ turn has not seen it fit to challenge this basic assumption: the attention merely shifts to the changing set of background beliefs, which impose constraints on the conversational process.

There is, on the other hand, a more deeply-rooted feeling that conversational activity is something that only human beings indulge in. That is, we converse with you, or somebody tries to persuade someone else, and so forth. Contemporary cognitive scientists notwithstanding, we find it humorous when someone says “I am having a deep conversation with my computer” or “I have finally convinced the computer that it ought to accept my point of view” etc. We find such statements humorous, because we feel that some kind of ‘category mistake’ has been committed. And one of the ways of being humorous, as we know, is to commit some types of category mistakes.

The second assumption: That our theories of conversation do not refer to the fact that it is people who indulge in conversation, but that this fact is ineliminably present in our deep intuitions about conversational activities need not necessarily be opposed to each other. That is, the distance between the theories and our intuitions is bridged (often) by a metaphysical assumption, which is highly culture-specific: only human beings have beliefs or only they entertain propositional attitudes and such like, discussions in cognitive sciences notwithstanding.

Taken together, these two assumptions cloak our theories of conversation with a mantle of universality, which is highly illusory. How does this come about?

Because our theories of conversation do not refer to human beings (who are always spatio-temporal particulars) as they are involved in a conversational activity, the feeling is that they are general theories. Beliefs appear better behaved than their embodiments, viz., human beings, and even ‘pragmatics’ of communication becomes less shifty when conceived as the shifting set of background beliefs. Even though conversation occurs between some specific human beings, our theories suggest that it is an interaction between two species members, viz., between ‘partners’ of a dialogical process. Even where, as in some dialogical logics, words such as ‘opponent’ and ‘proponent’ are used, or as in theories of rhetorics concepts like ‘speaker’ and ‘audience’ are taken recourse to, they mean nothing more than the fact that the content of these beliefs constitutes the functional position of the generic individual. There is very little that distinguishes one such generic individual from another: they are all embodiments of beliefs and, as such, they are partners. With respect to the property of entertaining beliefs, it is not possible to distinguish between human beings. Because embodied beliefs are individuated the way beliefs are, the promissory note (one supposes this) that conversational theorists give us is that where one has to refer to those individuals who are having the conversation (a king talking to his subjects, the journalist interviewing the prime minister, the pope speaking to a priest, a Nobel laureate talking to a green doctoral candidate etc.), it could be done by referring to the belief sets. (A proponent is someone who has a pro-attitude to belief; a ‘Pope’ is an individuation of the belief set ‘Pope’, etc.)

The third assumption: Some activity is a conversational activity, or a reasonable discussion, if it instantiates or exhibits those properties required for it to be a conversational activity or a reasonable discussion. Insofar as a reasonable discussion is a norm guided activity, and norms qua norms are those which are universally obligatory or forbidding, to formulate a theory of argumentation (or conversation) is to build a general theory with ‘universal’ norms. How, after all, could one defend a norm of ‘rationality’ that is not universal?

In other words, even the kind of cursory and superficial glance as the one we have cast allows us to pick out at least three assumptions underlying theories of conversation:

(a) an assumption about the nature of human beings, which says that human beings embody beliefs;

(b) an assumption about the nature of persons, which says that humans are equal with respect to this property;

(c) an assumption about ethics and the nature of norms, which says that norms are obligatory in nature.

The first allows you to ‘discover’ relations between beliefs; the second enables you to provide a general structure of what it is to converse and the third helps you to propose and defend ‘maxims’ or ‘norms’ as the case may be.

If this caricature is even approximately true, we can raise the question now: What if there exist cultures and societies, where none of the above assumptions are true? That is, what if there are cultures and societies, which do not make this kind of a distinction between human beings and the rest of nature; to whom the ‘observational’ term, viz., human beings turns out to be a very ‘theoretical’ term; and who do act ethically without requiring or needing norms and values? Clearly, in such cases, our theories of conversation break down. By the same token, they cease being universal–in the sense of being theories of human communication.

We would like to propose is that this is indeed the case: in each of the scenarios sketched above, what we see is the breakdown of our theories of conversation. In these cultures, the kind of conversation held depends on exactly who the participants are; the ‘relevant’ answer to a question depends upon just who is asking the question and who is doing the answering; the conversational process is guided not by norms, but a learning process of executing appropriate actions.

We will not be arguing for any of the above points, much less provide you with overwhelming empirical data in support of them. Even though we are convinced that these assertions are true, candor requires us to warn you that our stance does not represent the consensus of the anthropological community.

3. Conversations, arguments and cultures

So far, we have spoken about conversation and conversational theories. What is their relation to argumentation and argumentation theories? We would like to suggest that the latter is extension of the former. That is to say, arguing is continuous with conversing. Because we cannot see what is objectionable about this point of view, we will not defend it. But to make this stance amenable to criticisms, we need to spell out some implications.

Argument, under this construal, is something like a ‘subset of conversation: properties of conversations are properties of arguments, too. An argument may possess more properties than a conversation; but no interaction can be an argument if it is not also a conversation at the same time. That is to say, not all conversations are arguments, but all arguments are conversations. By saying this, we believe that we are merely describing the consensus that exists in these two fields.

If this is true, the following is also true: all assumptions of theories of conversation are also the premises of theories of argumentation. From this it follows that if our theories of conversation lead us to say that people in other cultures do not converse ‘properly’, our theories of argumentation can do no better: people in other cultures do not argue ‘properly’ either.

It is here that we encounter the clouding of our intuitions: the conclusion that other cultures solve disputes ‘less properly’ (or more often a less rational way) appears less implausible. It looks as though it could be true: one is more inclined to retort with a ‘so what?’, and suggest that only an empirical enquiry could give lie to this answer. Even though we think that it would be wise to curb this inclination, it is more interesting for our purposes to pursue another question: Why is the implausibility less obvious in the case of argumentation than it  is in the case of conversation?

Our answer is that it is because of the ‘association’ that most make between rational argumentation on the one hand, and institutions that arose in the West on the other. Two such are important in this regard: legal institutions and scientific institutions. Given the historical fact that the West nurtured and developed the theories of natural sciences and the institutions of law, nothing appears to be seriously amiss if one believes that both the theories about and the practice of ‘rational discussions’ are more advanced in the western culture than elsewhere. That science appeared first in the West and that it underwent a long period of development there is a historical accident. It is equally a historical accident that legal theory and institutions based on them had their origin and growth in the West. Because of these accidents (convenient no doubt, but not providential because of that), there is a greater tendency in this culture, when compared to other cultures, to argue ‘rationally’. Even if one is normatively prepared to recognize that all cultures are equally worthy, it does not prevent one from suggesting that factually the West is more inclined to pursue argumentation in settling disputes than are other cultures.

What is the worth of this argument? We shall not take the easy road of pointing out to the records, which have chronicled the history of the West. We will travel a different route instead. To assess this claim, we will find out the specific problems confronted by argumentation theories that take legal disputes or scientific discussions as their models. That is the task of this section, beginning with legal disputes.

3.1 Rhetorics and others

The Berkeley group of anthropologists led by L. Nader (e.g. Nader and Todd, eds., 1978) has pioneered the attempts to understand the disputing processes and courts of diverse forms in other societies. In these cases one invariably gets a party which is wronged in some sense or other and a party which is litigated. Often, there is a third party which adjudicates the offence. Prima facie the relation appears to be the same as that we see in our courts. And yet the differences are of such a nature that the anthropologists are required to take recourse to generalities in order to make sense of the simple fact that there are disputing processes elsewhere too. Nader states, for example, that the growing insight of the anthropologists is that

(d)isputes are social processes embedded in social relations. The focus of attention shifts ‘from the dispute itself (and the techniques for handling it) to the social processes of which the dispute is a part’ (Starr & Yvgvesson, 1975 cited in Nader & Todd, 1977:16)

Now, this is hardly an earth-shaking statement. At first sight, it is even extremely trivial. This description is ‘true’ for any disputational process. For example, dispute in a court of law is a social process that is embedded in the social relations of a society, which ‘recognizes’ juridical relations between its members. Any person who studies a disputing process, say marital disputes, empirically focuses on the “social process of which the dispute is a part” and not just the “dispute and the techniques for handling it” after all! Why, then, are legal anthropologists forced to embrace this trivia as a great insight whenever they study disputing processes in other cultures? There are two obvious answers to this question. The first one is simply this: they are unable to capture the disputing processes in other cultures as disputing processes, if they merely extend (in a simple or even a complex way) our notions of disputations.

Why? What kind of problems do they confront? In legal anthropology, there are no clear answers provided. But the issues confronted have found some articulation. Consider what Bohannan (1975) has to say (see also Nader, 1975) about the problem of studying the process of discussion in so-called ‘tribal law’ (the Tiv). Approaching this from the point of view of characterizing the nature of legal discussion in the Tiv, he says: “But the two ‘languages’ per se–English jurisprudence, and the native language–are not equivalent entities” (1975: 404). The problem, as he sees it, is that British jurisprudence is a sort of formalized language. Consequently, he speaks primarily about the ‘vocabulary’ of the Tiv and the jurisprudence. But his problem is not simply one of presence or absence of some set of words in either of the two languages. Hence, on the same page, he says that we should think of the British jurisprudence as a whole ‘medium’, which is not at all similar to the Tiv tradition. A similar point is made by Nader too: what seems familiar in western and non-western traditions at first sight (disputing processes, argumentation, etc.) proves to be very different in practice. That is to say, the notion of ‘legal dispute’ makes it almost impossible to theoretically recognize as a legal dispute, even where it is perfectly obvious that the phenomenon is a legal dispute.

To understand this predicament better, we need to look at the second answer to the question why the anthropologists palm off trivia as deep insights. It involves a literal reading of their answer, but against the background of what we have identified as the assumptions of conversation and argumentation theories. If we do that, we can rewrite their general point thus:

(a) A dispute is not merely a confrontation of belief systems, but a relation between specific human beings as the socio-cultural entities that they are.

(b) The ‘parties’ of a disputing process are not mere partners, and a dispute cannot be reduced to a set of techniques used by them. What goes on in a dispute is a genuine change of social status, of ‘personhood’ etc., according to the tradition one lives in.

(c) The dispute is ‘relevant’ to or restricted to the tradition of the disputants. Indeed, in most disputes with members of another tradition (a foreigner, enemy, etc.) a different language and other procedures are used.

In other words, our theories of rhetorics that take legal disputations as models fail us if used to understand disputing processes in other cultures. They do not even allow us to identify the process which we need to study in order to settle the question of the existence or otherwise of rational procedures for solving conflicts. This, then, is the problem faced by one set of theories of argumentation.

What would happen if we were to take scientific discussions as our models? Could we at least make use of that to defend our normative assumption about what a reasonable discussion ‘ought’ to be like? Some have indeed travelled this path. Let us examine their attempt now.

3.2 On the irrelevance of ‘rhetorics’

All appellations such as ‘universal’ audience notwithstanding, the models that appear to inspire the rhetoricians are those involving an ‘ideal’ scientific community. There is, it appears, an entrance fee to be paid in order to be a partner in a rhetorical event: knowledgeability in the particular domain. This price is not only too steep for the ordinary citizen, but it also effectively excludes cogitations about almost all issues of social, political and moral significance. Because of this, theories of argumentation threaten to become irrelevant. Let us explain ourselves.

Consider, as an example, van Eemeren and Grootendorst’s book (or, for that matter, Barth and Krabbe’s dialogical systems). It is really difficult to take the rules that they propose for ‘reasonable discussion’ seriously: speaking in turn, discussing ‘critically’ (dialectically) some or other thesis put across by a proponent, etc. Surely, when Blackhole physicists come together in a conference and try to ‘evaluate’ the competing claims of theories they have a ‘reasonable discussion’ without ever having heard of rhetoricians. How, we wonder, would physicists react when one of their colleagues puts across a hypothesis and refuses to ‘test’ it? What would their response to a colleague be who tells them to accept some theorem as true, but refuses to prove it when asked to do so? (This is the ‘obligation to defend one’s position’–one of the norms in argumentation theory.) By and large, it appears to us, contemporary rhetorics is irrelevant to scientific discussions. Not only because of the fact that the latter functions as a model for the former, but also because the strictures of contemporary rhetorics are far too primitive to ‘regulate’ scientific discussions.

For the life of us, on the other hand, we cannot imagine what the demand for ‘critical testing’ really means when issues of social and moral concern are debated. How could one ‘critically test’ the claim that peace in Europe since the Second World War is due to the nuclear umbrella? Or, for that matter, its negation? How to ‘critically test’ the statement that eliminating the budgetary deficit ought to be the aim of a government which intends to guarantee the welfare of its people? How to ‘critically test’ the proposition that sanctions against South Africa are moral or that it is immoral? ‘Critical rationalists’, like van Eemeren and Grootendorst, do not give us any answers.

Such a theory of argumentation ends up becoming totally irrelevant: where we do argue ‘reasonably’, there we do not need a theory of argumentation to tell us that it ‘ought’ to be done. Where we do not know how to reasonably ‘argue’, there the theory of argumentation has little to tell us!

It must be pretty obvious what the problem is: when scientists come together to discuss and mutually evaluate the results of their colleagues, they come equipped with a whole arsenal. They have explicit theories, localizable problems, experimental techniques, accepted notions of ‘experimental evidence’, and many other things we do not know much about. The theories of argumentation that take scientific discussions as their model are not built after a thorough analysis of the structure of discussions in (m)any scientific domain(s). Instead, the theorists rush into the battle-field with pitiful slogans like ‘falsifiability’ and ‘anti-justificationism’ and such like. The problem with issues of social, moral and political concern is that none of the above conditions hold: most of our ‘theories’ are either implicit or fragmented; it is difficult to specify what the problem is or what counts as its solution, etc.

There is another way of putting this: theories that take ‘scientific discussions’ as their model suggest that a persuasive speech merely involves transmission of information. Where it does concern only transmission of information–such as reports of the results of experiments, elaboration of mini-theories and such like–it does make sense to focus upon the ‘efficacy’ of transmitting information, its brevity and relevance and such like. But as we have seen, in such situations the argumentation ‘theories’ are irrelevant. When it is about issues of social, political and moral concerns of communities, it is hardly clear what counts as information and what does not.

There is no great need, we take it, to summarize more than briefly the drift of what we have argued so far: our theories of argumentation threaten to become irrelevant. They become irrelevant to studying other cultures (in one case) because we cannot even identify the phenomenon that we are supposed to study; they become irrelevant (in the other case) because they have nothing to say!

What, then, are theorists of argumentation really doing when they speak of ‘norms in argumentation’? What are they studying? Why do they fail us, all tall claims notwithstanding? To get a grasp on this question, we need to identify the “problem situation” (as ‘critical rationalists’ put it) of theories of argumentation. To this task, we now turn.

4. Learning and rhetorics

Let us begin with the following two questions: when is someone a speaker? Under what conditions is someone the audience? In order to prevent possible misunderstanding, let us say straight away that we are interested in the nature of these ‘roles’. That one and the same person can be both a speaker and the audience within the same argumentative interaction does not, in any way, alter the answer that one gives to the two questions.

4.1 The speaker and his audience

Let us begin with the speaker: in an ideal case, the speaker wants to persuade his audience that the information he is about to convey to them is ‘true’ (or ‘acceptable’ or whatever). He also wants to win the adherence of the audience to his position because it is a rational position. Since we are talking about (and interested in) issues of social, political and moral concern, we can reformulate the above sentences thus: In an ideal case, a speaker has chosen for a course of action, his choice is a rational one, his goal is to convey reasons for his choice and persuade (where applicable) his public to be equally rational. The audience, if we assume that the speaker has been successful, has now learnt something from the rhetor: it has acquired a new piece of information. What competence is required on the part of the audience in order to be persuaded? The only condition is that it understands the speaker. The public requires linguistic competence, and nothing else. That is to say, learning (on the part of audience) is fundamentally a passive act of absorbing information! Consequently, if the audience has to learn, it has to be passive.

One might want to say that the audience is ‘really’ not passive, but active instead, because it needs to ‘think’, ‘analyze the argument’, ‘check for inconsistencies’ and so forth. Is this indeed the case? We do not think so: insofar as no axiomatized theory about issues of political, moral and social issues exist, the inconsistencies in arguments about these issues are found out by using the logical intuitions built into the language. To be a competent speaker-in-a-language (not an infallible one), to understand a language is to understand the logical relations between the sentences. No ‘extra’ activity apart from understanding a language is required. The same holds for ‘thinking’ and ‘analyzing the argument’ too. In other words, a successful speaker requires a passive audience; a passive audience is a pre-requisite for ‘rational’ argumentation.

This is a curious paradox: the alleged partnership (with all its connotations of ‘equality’ and such like) involves an undignified, asymmetrical relation. The speaker speaks, while the audience shuts up and listens. (In the hands of contemporary ‘critical rationalists’, it is even the norm!) If we look at a bit closely, the ‘paradox’ vanishes: if human beings are embodied beliefs, a learning process for such beings cannot be anything other than change in their belief systems. The culture-specific assumptions that we identified earlier are continuous with this state of affairs as well.

4.2 The field of argumentation

Let us shift our attention sidewards to bring the (problem) situation into better focus: Consider any situation where we intuitively would like a theory of argumentation to help us: abortion, divorce situations, discrimination on grounds of race and sex, nuclear umbrellas, unemployment, AIDS… Why exactly do we want to argue? Because it is not clear which action has to be executed next in that situation which called for an argument. That is to say, both the ‘speaker’ and the ‘audience’ are in a learning situation. The situation is new, the learnt repertoire of actions does not appear to be adequate; consequently, both the speaker and the audience have the same question to which they seek an answer together: what to do next? To this description of the situation, if you add such assumptions as:

(a) one decides first: actions are executions of decisions;

(b) an action is rational accordingly as whether the decision was rational or not;

(c) a human being is rational if he decides rationally; and such like,

then you will arrive at the more familiar picture of a person unable to decide which course of action to choose. To provide help in such contexts is to provide as much and as accurate information as possible. A ‘good’ speaker does precisely that.

In other words, the issue that argumentation tries to solve is one of providing a set of learning strategies. It is telling us how to go about solving a problem of learning to execute new actions. Its heuristic (even if the theorists call them ‘norms’) for learning is that we should try and acquire information by listening.

But, of course, this is not what argumentation theorists say: they say that the domain of rhetorics is about resolution of the conflict of avowed opinions. We would like to suggest that this ‘self-description’ of contemporary rhetorics is a false one, if it is taken to mean that both the domain and the problem exist quite independent of a highly culture-laden way of conceptualizing the domain itself. It is taken to mean this, because each writer informs his readers that “wherever there are people, there are differences of opinion” and so forth. Let us look at this point more closely.

Consider the following example: In order to say that Aristotelian physics is less preferable to Galilean physics, we need to identify the question which they were both trying to solve, viz ., Why do unsupported bodies fall downwards? It will not do to say that Aristotelian physics was solving only the problems of the existence of entelechy or energia. Rather, with these and other related notions, Aristotelian physics tried to solve, say, the problem of the motion of bodies. To be sure, it also formulated questions and answers that cannot even be conceptualized in Galilean physics: the relation between actuality and potentiality, for example. But when these two theories in physics emerged as rivals, and a choice was made between them, one had to (at the least) identify a set of common problems that both tried to solve. Analogously, the problem which argumentation theories answer is one of ‘How to learn what to do next?’ The problem ‘How to solve conflicts of opinions?’ is a problem within argumentation theory in exactly the same way the question ‘How does potentiality relate to actuality?’ was a problem in Aristotelian physics. In order to assess the worth of argumentation theories by comparing them with rival theories, their internal questions are of little use.

We could also reformulate this point differently. One could and should draw a distinction between (i) a field of study, and (ii) a particular approach or a collection of approaches typically used in that field of study. To say, as contemporary argumentation theorists do, that they study the field of argumentation is, we claim, wrong. The field of study is one of finding out how to execute an action when the next action is not obvious. That is, it is a field of action-heuristic. The characteristic property of contemporary argumentation theories is that they solve this problem by suggesting that we have differences of opinion and that we try to resolve it. This is but one answer, but one way of conceptualizing the solution. If this is the case, what happens to the normative position underlying the existing argumentation theories?

4.3 Why is it rational to argue?

One of the most simplistic answers to this question, unfortunately the only one that seems to exist in the literature, is the one that Popper has popularized. It appears as though Humankind knows of two things: either ‘argue’ peacefully or chop off the heads of those who disagree. As he put it once,

If the method of rational critical discussion should establish itself, then this should make use of violence obsolete: critical reason is the only alternative to violence so far discovered. (Popper 1972, p.292; emphasis by Popper)

Either people die, or people let ideas die in their stead. “In the face of argument of such quality” writes Gellner–a Popperian, though no neophyte–“one can only feel embarrassment”. (Gellner, 1985, p.43) Not all, apparently, share Gellner’s discomfiture. Barth and Krabbe more or less endorse the ‘reason or violence’ view of Popper (p. 25, 1982); the same motif appears to animate the work of van Eemeren and Grootendorst too.

We can appreciate why this alternative between ‘reason’ and ‘violence’ will not do. Because, to make this alternative stick, one will have to show that the way the argumentation theories look at their field of study is also the only way of ‘peacefully’ solving it. No ‘critical’ or ‘dogmatic’ rationalist has shown this; nor, for that matter, can they show it. The reason is as obvious as it is deplorable: they have not studied how other cultures have solved the same problem. Not only that. Only the arrogant or the foolish would claim that their pet ‘theory’ is the only way of solving some problem. History of sciences has taught us at least that much.

5. A Question

We would have liked to outline at least one other way of solving this problem by drawing upon one of the living cultures, viz., Asia. But the page-limit that is imposed upon us makes it impossible to do so. Consequently, all that we can do in the rest of the space allotted to us is to raise some questions.

The premises of argumentation theory, viz., the presentation of a thesis, adducing evidence for it, addressing a ‘universal’ audience, etc., are all culture-specific assumptions. They are not the result of considered reflections (even if they are the results of considerable efforts) on human situations. Stipulative fiats aside, we see no obvious ‘superiority’ to these notions as they are canvassed by theorists of arguments. Practically, these proposals are next to worthless. Empirically, with respect to other cultures, they are false anyway.

If you are with us so far, you have been with us almost to the end. Our theories of conversation, our theories of rhetorics suggest to us that others neither know how to talk nor how to argue. Factually these claims are false, normatively they are untenable. (Nor because, we would like to emphasize, of a dubious normative premise that all cultures ought to be considered equally worthy or equally rational or whatever else.)

The question that we should ask ourselves, at the end of an anthropological disquisition such as this one, is this: should we not attempt to develop our ‘general’ theories of communication after we have made some efforts to understand other cultures and other human beings on their terms? Should we not attempt to get rid of our illusions regarding the universality of what are really our provincial experiences before building theories that lady down ‘normative’ rules as to what counts and does not count as ‘reasonable’ discussions? As anthropologists (even if one of the coauthors does not think himself as an anthropologist!), our answers are in the positive. Whether we have able to convince you or persuade you is something that we shall know only later. It is our hope that we have at least made our claims appear plausible.


Barth, E.M. and E.C.W. Krabbe: 1982, From Axiom to Dialogue: A Philosophical Study of Logics and Argumentation, Walter de Gruyter, Berlin.

Bohannan, P.: 1975, ‘Ethnography and Comparison in Legal Anthropology’, in L. Nader (ed.) Law in Culture and Society, Aldine Publishing Company, Chicago, pp. 401-418.

van Eemeren, F.H. and Grootendorst, R.: 1982, Regels voor Redelijke Discussies, Foris Publications, Dordrecht.

Gellner, E.: 1985, Relativism and the Social Sciences, Cambridge University Press, Cambridge.

Grice, H.P.: 1975, “Logic and Conversation”, in P. Cole and J. Morgan (eds.), Syntax and Semantics, Vol. 3, Speech Acts, Academic Press, New York, pp 41-58.

Nader, L. and Todd, H. (eds.), 1977, The Disputing Process: Law in Ten Societies, Columbia University Press, New York.

Popper, K.: 1976, “Reason or Revolution”, in The Positivist Dispute in German Sociology, Heinemann Educational Books, London, pp. 288-300.

Sperber, D. and Wilson, D.: 1986, Relevance: Communication and Cognition, Harvard University Press, Massachusetts.