Castaños, Fernando. 1981. “Elements for a coding system of argumentative acts”. Papers on Work in Progress, no.6. Bologna. Cooperativa Libraria Universitaria Editrice. 6- 13.
Elements for a coding system of argumentative acts
Fernando Castaños
Introduction
Since the early 1970s considerable efforts have been made in applied linguistics to describe language in terms of categories that go above and beyond the sentence, specially in connection with English for specific purposes (ESP). The aim has been to focus the attention of foreign language teaching on the use of language to communicate scientific and technological information, and thus to show the limitations of viewing language merely as the manifestation of grammatical rules.
There has been much interest in units larger than the sentence, such as the paragraph and the text, the exchange and the dialogue. There has also been much interest in units that describe the contents of a sentence, and in units that reflect the role of sentences in cognitive and communicative patterns. We have, thus, talked of propositions and notions, of rhetorical functions and communicative purposes, among others.
The following is a good example of the kind of statements that we have aimed to make about the language we want to teach:
Thus we find commonly the rhetorical functions of description, definition and classification, and the rhetorical techniques of time order, space order and causality. In addition, manuals have two rhetorical features found less commonly in scientific and technical writing, the interpretation of illustrations and the rhetoric of instructions. (Todd Trimble and Trimble 1977, quoted by Robinson 1970: 22).
Unfortunately, however, words such as “description”, “definition”, and “classification” are not as precise as they seem to be, and coding systems based on them are very difficult to use by other researchers than their constructors. Indeed, definitions of such categories are always vague, if they are provided at all. Let us consider, for example, the following:
SPECIFIC: Gives more specific information about something that was stated in a general way.
EXPLANATION: Previously stated information is explained in a more abstract manner (for example: relating the information to a general principle) or more concrete manner. (Meyer 1975, quoted by Widdowson and Urquhart 1976:25).
What is the difference between more concrete and more specific? Undoubtedly there was one in Meyer’s head when the system was devised, but what it is is not clear from the definition. It is my own experience that when working with systems of the sort in question, a researcher can achieve a high degree of consistency, say between two codings of the same text, provided thay are done within a short period of time. But if the system is used again after a few months the basis for the agreement is forgotten.
The problem is documented in some important publications. For example, Candlin et al. (1975) attempted to use a system devised by Williams (1973):
Identification involving defining
questioning
Classifying taxonomy
matching
differentiating
Analysing evaluating
generalising
measuring
simplifying
concluding
testing
predicting
Process interaction Describing evidence
causality
change of state
inference
hypothesis
states
processes
quantification
explanation
instruction
Commenting on this system, Candlin et al. say:
We found… that in fact many of our categories were either too vague or were pseudo-categories or that categories overlapped considerably. (Candlin et al. 1975, 17).
Among the problems that one would find if one attempted to use Williams’s system is the following: does Process in the left hand column refer to the same thing as processes in the right hand column? If so, why does the category occur twice, and in different columns? If not, why do the two categories have the same name? Another question is: why is Analysing a higher order category and “generalising” a lower order one, and why is it not the other way round?
It is not my purpose here to do a detailed criticism of developments which have at certain stages been useful in applied linguistics. I simply want to briefly indicate that there is a need to go a step further and try to provide unambiguous definitions of discourse analysis categories, and to show the kinds of problems we have to avoid or tackle.
My aim is to present some explorations in a certain direction, the direction of semantico-pragmatic universals. It seems to me that by focusing on the basic acts of signification – reference and predication – and on the systems of meaning used and created by discourse, we will arrive at definitions of categories for coding systems which are less vague and ambiguous than the ones we have at present.
Initial Discussion
Let us begin by locating the source of the problems I have referred to in the above Introduction. In Castaños 1980, I discuss some apparent contradiction between accounts of the act of definition by Selinker and the Trimbles (1976), on the one hand, and by Widdowson (1978: 41), on the other. What emerges from that discussion is that definitions are acts which may fulfil various functions.
A definition may be used by a speaker (or writer) to explain the meaning a word has, to a hearer (or reader) who ignores such meaning. But a definition may also be used to establish the terms in which discussions about a given matter can take place, as for example, in mathematical theories or in the rules of ball games. Furthermore, a definition can be used to signal that something is important, as occurs in many textbooks. In other words, definitions may be facilitative, argumentative, and/or valorative in function.
The discrepancies between the accounts by Selinker et al. and by Widdowson can ben traced to their being accounts of (different) functions of definition. And it seems clear that what definitions have in common irrespective of their function is that they are equatives, but not equatives of reference, as those studied by Halliday (1970): rather, they are equatives of denotation – using “reference” and “denotation in the sense of Lyons’s Semantics (Lyons 1977: 177, 206).
This brief sketch of the analysis of definitions indicated that the problem with the categories I mentioned in the introduction is that they are supposed to correspond to what might be called everyday academic metalanguage” terms. And those terms are very much like ordinary language terms; they stand for amorphous and variable configurations of elements. Different speakers use them on different occasions to refer to different sub-configurations of elements.
The problems that arise when one starts with names and then provides definitions of categories, in order to devise an inventory of acts, is also noted in Mitchell 1980, a systematization of requests and offers, Mitchell says there is no reason to suppose that a set of lexemes from English, or any other language, has divided the semantic field of speech acts in the most convenient way for applied linguistics. To start with, sometimes the performative and the reporting verbs do not even come in corresponding pairs.
It would then seem that, if we are interested in an account of speech acts, rather than in producing a glossary of English words related to speech acts, then we should proceed not from name to analysis of features, but from the analysis to the name – if a name is required at all. This is what Mitchell has done in his case, and this is what I intend to do for the kind of acts I mentioned in the introduction. In the next section, I shall present and discuss a proposal to consider those acts in terms of the internal structure of the propositions that are used to realise them.
Proposal
The sketch of my analysis of definitions in the previous section not only shows the kinds of problems we face, but it a1so indicates a possible solution. There I said that what definitions had in common was their being equative, and implied that not all equatives were definitions, but only equatives of denotation. I am here considering two things: the kind of predication involved (eg. equative/non equative) and what the subject or subjects are being used to refer to (e.g. reference/denotation)- and this points to a need to distinguish reference as an act from reference as a semantic relation, but let us leave this aside for the present.
Now, it has long been recognized that many acts are not realized with single propositions, but with combinations of propositions. I think that these combinations have to be dealt with in terms of sense relations, if we are interested in scientific theories as systems of meaning, and in scientific discourse as the use and creation of those systems.
In sum, I am proposing that:
a) Some argumentative acts are combinations of types of reference and types of predication
b) Some argumentative acts are combinations of sense relations of elements in pairs or groups of propositions.
Let us consider some examples:
(1) John is the doctor
(2) John is a doctor.
What distinguishes (1) from (2) is that the first is equating two singular definite references, and the second is predicating membership in a class of one particular referent. These are the kinds of elements referred to in a).
(3) The north wall is blue. The north wall is big.
(4) The north wall is blue. The south wall is green.
What distinguishes (3) from (4) is that in the former we have two propositions about the same referent and which ascribe unconnected properties to it, while in the latter the propositions are about different, but semantically connected, referents; further, in these two propositions the predicates are also connected. The sense relation involved in both cases, the referents and the predicates, is co-hyponymy. These are the kinds of elements referred to in b).
At present I think the two dimensions identified in a) and b) suffice to distinguish argumentative acts, and though it will probably be necessary to subdivide them, e.g. to separate quantification or logical connectives, it will not be necessary to add any more dimensions. It is important to note that the elements I am focusing on also enter into the configuration of facilitative and valorative acts, arid indeed some of the examples I will provide are of facilitative, rather than argumentative acts. But a full account of facilitation and valorization does require other elements, such as word order and sentence length, with which I shall not deal here.
Before exploring in some detail the proposal I am making, I wish to discuss some implications of it. The first one has to do with the methodological issue of going from name to analysis or analysis to name. Categories such as “singular definite reference” are, after all, names. Then, the source of the problems mentioned in the introduction is not much different from the source of their solution. But there is some difference. I do not intend to do an extensive analysis of reference types; I intend to use the reference types to do the analysis of the argumentative acts.
The aim is to specify complex configurations from the level of rhetoric in terms of simple elements from the semantico-pragmatic level. At this lower level, agreement between researchers, and therefore replication of studies, can be more readily achievable. Indeed, it is relatively easy to say unambiguously what one means by “singular definite” or “plural indefinite” reference – though the field of types of referring expressions is not totally free from discussion in philosophical quarters.
The second implication of my proposal is that different argumentative acts are not necessarily different illocutionary acts, as they are often thought to be. Indeed, if the differences lie in the internal structure of propositions and not in the “external” conditions for their use, the question of illocutionary force does not arise, and matters of intention and belief would seem to remain constant. To see this more clearly, let us consider negation first with respect to the act of promise and then with respect to the act of definition.
As Searle tells us, “Propositional negations leave the character of the illocutionary act unchanged because they result in another proposition with the same illocutionary force” (Searle 1969: 32).
Thus, both (5) and (6) below are promises:
(5) I will come.
(6) I will not come.
But (7) is a definition and (8) is not:
(7) Horn—silver is native chloride of silver.
(8) Horn,—silver is not native chloride of silver.
That is, the ‘force’ of definition changes when we negate the proposition. But there is something which remains unchanged, the real illocutionary force, which I suggest is simply that of assertion. We then need to distinguish illocutionary acts from argumentative acts (all of which are perhaps assertions). The need for another kind of entity, which could be called the propositional act, will become clearer when we see that equivalent propositions (from a truth—conditional point of view) can be expressed in the realization of different acts, that is, by different combination of referring and predicating expressions.
The final implication I wish to consider focuses more specifically on part b) of my proposal. If an act is a combination of two (or more) propositions, then one proposition can enter into the realization of various acts, by entering into various combinations with other propositions. For example in:
(9) The east wall is blue. (10) The north wall is blue too. (11) But the south wall is green.
we can identify (at least) two pairs of proposition, (9) and (10) on the one hand, and (10) and (11) on the other. Two combinatorial acts involving (10) are being performed, one explicitly marked with “too”, and the other explicitly marked with “but”. And if our little paragraph was part of a longer text, nothing would stop us from being able to identify other acts involving (10), e.g. if there were other sentences that referred to “floor” or “ceiling”, as opposed to wall. And ah these combinatorial acts of (10) would obtain as well as the “purer” act defined by the specific referring and predicating expressions of (10). We thus have a systematic way of accounting for the fact that a sentence can be used to perform a multiplicity of acts, a problem that has troubled applied linguists for long.
In the next two sections, parts a) and b) of the proposal will be considered in some more detail.
(The second part of this article will be published in PWP 7)
REFERENCES
Candlin, C.N., Kirkwood, J.M., and Moore, H.M. (1975). Developing study skills in English. In English for academic study with special reference to science and technology: problems and perspectives. London: ETIC.
Castaños, F. (1980). Consideraciones sobre el Discurso científico y la Definición. Estudios de Lingüística Aplicada. CELE/UNAM.
Halliday, M.A.K. (1970). Language structure and language function. In J. Lyons (ed), New horizons in linguistics. Harmondsworth: Penguin.
Lyons, J. (1977). Semantics, Vol. l. Cambridge: Cambridge University Press.
Meyer, B.J.F. (1975). The organisation of prose and its effects on memory. New York: North-Holland.
Robinson, P. (1980). ESP (English for specific purposes).
Oxford: Pergamon.
Searle, J.R. (1969). Speech acts. Cambridge: Cambridge University Press.
Selinker, L., Todd Trimble, M., and Trimble, L. (1976). Presuppositional rhetorical information in EST discourse. TESOL quarterly, 10, 3.
Todd Trimble, M., and Trimble, L. (1977). Literary training and the teaching of scientific and technical English. English teaching forum 15, 2.
Widdowson, II .G. (1978). Teaching language as communication. Oxford: Oxford University Press.
Widdowson, H.G., and Urquhart, A.H. (1976). K.A.A.U. English for academic purposes project. Edinburgh: University of Edinburgh. Mimeo.
Williams, P. (1973). A function—based course in English as a foreign language in science and technology. Lancaster: University of Lancaster. Mimeo.
Comentarios recientes