|
|
|
|
|
|
|
ARGUMENTATIVE-LINGUISTIC PRINCIPLES
OF THEORETICAL TEXT COMPREHENSION
Vasilyev Lev G.
Kaluga, Russia
Published: Collected research articles, Bulletin of Russian Communication Association "THEORY OF COMMUNICATION AND APPLIED COMMUNICATION", Issue 1 / Edited by I.N. Rozina, Rostov-on-Don: Institute of Management, Business and Law Publishing, 2002. - 168 p. P. 142-149
Abstract
The proposed method of linguistically-oriented reasoning comprehension is based on semiological principles of text comprehension where both content and form are essential. A text recipient is viewed as a rational judge trying to detect all the components of the argument he/she considers and thus to see if the argument is consistent. Elementary and higher level argumentative units of the text are discovered by applying a modified S. Toulmin's model of argumentative functions. Validity and correctness of arguments are established by means of a linguistic interpretation of traditional syllogistic.
The goal of this paper is to sketch a linguistic-argumentative method of analytical comprehension of theoretical texts in humanitarian sciences. The proposed method gives a program for a normative / prescriptive reconstruction of hidden necessary premises. In an argument (a standpoint supported by premises), there can be a number of premises, some of them inalienable (semantic), others - suitable (pragmatic). The problem at stake is how to define which premises are necessary and how they must be expressed linguistically. The latter is important if the linguistic aspect of argument is considered. I state that the idea that only propositional content, but not the form of the judgments within an argument is necessary for correct comprehension of the latter is wrong. Specifically, it is wrong when an enthymematic argument is reconstructed so that the necessary premises be found on the basis of an explicit standpoint. Necessary premises can be found by means of a syllogistic. Traditional syllogistic, however, allows relative linguistic freedom in formulating argument premises. Criteria for creating a system of a linguistic modification of syllogistic must be proposed. Depending on differences of the linguistic form of a standpoint its premises are reconstructed differently and different logico-linguistic syllogisms are deduced.
Studying the problem of comprehension depends on the method accepted, on a researcher's background, and on the field of research. Thus, approaches in psychology or psycholinguistics can differ from those in hermeneutics, literary criticism or philosophy. Scientific method is not the only one to be applied in solving the problem of the essence and mechanisms of comprehension; it can be supplemented by other methods. All that means that both the topic and the object of research matter in studying comprehension. By the topic I mean a particular kind of message for understanding. By the object I mean a chosen method and particular aspects of the message to be studied.
The topic of my study is a research text in humanitarian sciences. Such texts differ from scientific texts in a considerably less rigidity. The latter can be defined as the degree to which a text can be modelled. The higher the rigidity of a text, the easier it is to reconstruct: such texts are usually built according to certain schemes. Introduction, Main Part and Conclusion are the simplest parts, arguments are more difficult. In scientific texts, formal logical schemes can be observed. In humanitarian texts formal logic is hardly applicable, and we need a different system to analyze their organization. The subject of my study is a problem of comprehending a research monologue text. Since a research text is organically argumentative, i.e. constructed on the basis of certain principles of reasoning, its comprehension is essentially comprehension of the text argumentation. By argumentation I mean reasoning in its informal-logical aspects. Rhetoric is not viewed here as a component of argumentation, and the latter is treated from the informal-logical point of view, rather than from the pragma-dialectical one. Since theoretical texts are oriented at reasoning, their perlocutionary aspect lies in theoretical ways of convincing, not in rhetorical means of persuasion. Therefore, for our analysis, rhetorical factors are irrelevant.
Argumentation is viewed as a social symbolic sub-system, with the system being a language - natural or artificial, depending on which version of argumentation is chosen for consideration. Like any human knowledge, argumentation as a symbolic sub-system is generated by the power of human mind. According to Kant (1929), human knowledge is based on human mind-initiated operations of structuring that transform sensations into perceptions. Knowledge is thus a product of human mind and is formed through correlation of the 'knowable' and the cognitive potential of a recipient. Constructive sign-forming abilities of cogitant individuals are unitary. This, however, does not mean that all cogitant individuals create identical cognitive structures: variety of constructs at an abstract level reflects specific categories managing the process; these categories can be logical or argumentative.
An important factor in production or changing symbolic systems (a symbolic system is viewed as a logical or semiotic system encoding a knowledge structure) is acceptance or refutation of a knowledge structure, respectively. If an old system of knowledge is refuted or is found inapplicable for describing or explaining an object, it is substituted by a new or a modified one. Being social (inter-personal), such competitive cognitive systems are applicable for describing and explaining phenomena. Therefore it is possible to postulate coexistence of competitive cognitive structures / systems, none of which, as a product of human mind and interaction, can be absolutely true.
Coexistence means that there can be a number of ways to comprehend what the author means and to estimate if his / her reasoning is valid: different logical systems can be used for such estimation. For example, Aristotelian syllogistic has undergone substantial changes through the centuries (cf. scholastic, Leibnitzian, Peirce's, and Lukasevits's syllogistic systems). Thus, Ch. Peirce's syllogistic differs radically from traditional one in these aspects. Though there are 19 syllogistic modes in both systems, Peirce's model is linguistically positive because in the subject-predicate relations there are no negative operators, and the description want / is wanted is used instead (Peirce 1932: 2.521). Two universal quantifiers (all and every) are used instead of one traditional (all). In the first and the seventh types of relations we find an extended treatment of the term distribution: the predicate is now distributed in positive universal judgments, cf.: (1) Every S has every p; (2) Every S wants every p. Some standard rules of syllogisms are not valid in Peirce's system: the traditional rule "particular conclusion if a premise is particular" is not mandatory (cf. Mode 2: Every P has some m's as its only qualities. Some S has all m's. Consequently, every S has every quality of any P). Also, particular conclusions become possible under two particular premises (cf. Mode 10: Some m is absent from all P's. Some S has all m's. Consequently, some S has a quality that no P has). All that means that from the point of view of a scholar using a traditional syllogistic system for estimating validity of an argument, Peirce-oriented estimation may seem fallacious. Consequently, argumentation theories can be object-oriented and object-specific; they can also be competitive and differently plausible/valid for a specific object (some of them can be better, others - worse).
K.Popper's (1974) rationalism is taken as a basis of method here. It states that though there is truth, it is practically unattainable. Competing theories designed to find the truth (or, in weaker terms, solution of problems, p.ex. what techniques to apply for efficient and unambiguous comprehension of messages) can and must be discussed and refuted since any of them is only a further step to attaining the truth. Falsifiability of theories leads to falsifiability of particular claims and judgments. Taking into account the unique character of personal experience, we can state the uniqueness of scholars' theories.
Therefore truth of judgments is viewed here as always relative to a particular cognitive system. The common ground for comprehension here is conventions about the principal axioms and the meaning of terms (such as Argument, Premise etc.). The conventional character of terms can be stronger or weaker: cf. Informal Logic, Pragma-dialectics, Deduction, Induction as examples of the latter. Thus, the two former terms are clearer to the ISSA community than to outsiders (even those working in the field of philosophy or logic). Also, there is no unanimity in treatment of the two latter terms (cf.: Fohr 1980; Freeman 1983; Fumerton 1980; Govier 1980; Weddle 1979). No doubt, conventional force can depend on linguistic clarity and skill to formulate one's ideas: the better you encode your ideas, the easier it is for a recipient to see what you mean.
A recipient of an argumentative text is viewed here as a "rational judge", or as an analyzer of reasoning in the text. He / she uses a certain model of analysis to comprehend the author's reasoning. The model is stored in the recipient's memory and is based on some logic. Criteria of logical correctness (relative truth of premises + validity of reasoning) must correspond to the standards of rationality that are used by both the author and the recipient of the text. Supposedly, such criteria exist. The standards are manifested in a specific argumentative model because a theoretical text is based on logic of reasoning. Depending on the genre of the monologue, we may find different reasoning systems: reasoning for persuasion purposes in an advertisement evidently differs from reasoning for convincing in a theoretical text.
There are different approaches to argumentation. Still, to have even minimal explanatory force any approach must be based on principles of construction and analysis of reasoning. Rational attitude helps us to choose out of many logical systems a basic one maximally corresponding to the goal and the object of our research.
For example, traditional syllogistic can work well enough for formalization of premises and conclusions in a formal system using logical symbols. It is hardly applicable, however, to natural language syllogistic arguments. For example, the argument
All bodies moving in elliptic orbits are subject to the law of gravitation
Comets move in elliptic orbits
Comets are subject to the law of gravitation
is regarded as the mode Barbara (Anderson, Belnap 1961:713). Viewed purely logically, it may be valid. But from the linguistic point of view, its validity is doubtful. For logic, the grammatical form of the predicate (moving - move) might not matter. Not so it is for language. Language users easily mark this difference. It is important because when the difference is recognized we can speak about violation of the principle of identity of the Middle Term of this syllogism. The grammatical form of the term (bodies moving in elliptic orbits vs. move in elliptic orbits) is clearly different. To observe the identity principle we must reformulate the minor premise into comets are bodies moving in elliptic orbits. We make this reformulation by using certain morphological-syntactic transformations. The latter can be different for different languages. If we apply this extension to syllogistic as a universal system we must either consider possible transformations of the kind and compile a list of them for different (at least Indo-European) languages or hypothesize their existence in linguistic competence of native speakers. In the former case, syllogistic becomes inoperable, in the second we cannot guarantee the logical correctness, because people do not always follow logical rules when performing transformations.
To solve this problem, not only content, but also linguistic form of sentences must be taken into consideration. It means that we must strictly observe the rule of term identity. Grammatical meaning of the terms must remain the same, i.e. a noun phrase must not be replaced by a verb phrase, and vice versa. Semiologically, the NP describes to what type or class the object belongs, i.e. expresses referential (extensional) features of the term while the VP expresses the properties, i.e. reflects its designative (intensional) characteristics. It must also be noted that in theoretical texts predicate terms do not usually denote action, process or state. Instead, they denote descriptions, or characteristics of objects. Respectively, linguistic functions of arguments (Agent, Patient, Instrument, etc.) are singled out depending on the type of predicate. For example, Agent is used with the predicate of action, but not of process, cf.: Bill killed the rat. VS. Bill saw the rat (in the second case the semantic function of the subject is that of an Experiencer, not of an Agent). Therefore, a subject-predicate model, but not a predicate-argument functional system (cf. Fillmore's 1970 case grammar) works better for texts of reasoning. The most natural means of expressing the subject of the sentence seems to be an NP.
Since an argumentative text is regarded here as a theoretical text based on reasoning, it must correspond to the principle of strictness which can be deductive validity. Taking into consideration the semiotic nature of a text, we should choose a logical system oriented (at least partly) on semiological processes. Such a system must be intensional because theoretical texts are themselves intensional. If we have a suitable logical system applicable in all respects but the intensional one, the system must be extended thus having an opportunity to describe both form and content. Since a theoretical text is a natural language discourse, it is necessary to pay attention to linguistic categories proper, i.e. meaning, exponential and contentive parts of the sign. These factors can be covered by a modified version of traditional syllogistic.
Taking into account the specificity of the type of a theoretical text taken as the object, namely, a text in humanities that does not have a strict formal organization, it is necessary to apply an informal logical system to text analysis. Such a system could demonstrate that being non-rigid, the text is still logically organized, i.e. constructed in accordance with a scheme of reasoning representing a tactico-strategic aspect of argumentation. For those purpose an argumentative-functional model as a version of sentential logic is used. It is a modification of Toulmin's (1958) model.
The modification says that if only one statement in a paragraph is present, it is probably a Claim, because it does not support any other statement. Having the Principle of Charity in mind, we cannot say that it is Data, because Data can be interpreted differently and a number of different conclusions can be derived from one and the same Data. Instead, the Principle presupposes that an isolated statement express the author's standpoint (Claim), grounds to which can be derived by readers. In other words, the Principle of Charity is a method of identifying an argument. If there is only one premise to the Claim, it is viewed as Data because according to the Principle, the most important premise should be expressed first. In theoretical texts, Data do not necessarily denote facts; and in this respect they differ from texts in everyday situations: in the latter Data can be easily and unambiguously reconstructed, therefore explicit premises often perform the function of a Warrant. So for theoretical texts the Principle of Charity performs a prescriptive function making the analyzer think that the premise of the Claim is Data.
Comprehension is understanding another person through a discourse; it is thus not only subject-oriented, but also object-oriented. The object-oriented principle of understanding presupposes specific treatment of happiness conditions of reasoning and comprehension of argumentation in monological texts. The happiness conditions are divided into general argumentative and specific argumentative conditions. This differentiation is based on the dichotomy between pan-systemic and mono-systemic levels in argumentative analysis.
General argumentative conditions for adequate comprehension comprise besides the Principles of Charity, Principles of Argumentativity, and of Symbiosis of Systems of Reasoning. The Principle of Argumentativity presupposes co-direction of premises of an argument so that their use does not contradict to a claim being proved, and the combination of the premises makes the argument stronger. It is important for reconstruction of enthymemes: no premises weakening the argument (like Toulmin's Rebuttal) must be restored as necessary. The latter are those which strengthen the standpoint without reserve and which do not require additional mental operations of differentiating supportive from not-so-supportive or non-supportive premises, the later having to do with counter-arguments. This principle does not apply to syllogistic because premises in a syllogism always "work in the same direction"; it is also important that the notion of strength of the syllogism is inapplicable to syllogistic as a deductive system.
The Principle of Symbiosis of Systems of Reasoning presupposes division of application of systems of logical analysis in accordance with a strategic and a tactical approach to the text. There are two levels of argumentation in the text. The strategic level is responsible for description of the principal (general) organization of the text. For strategic analysis argumentative-functional model is used; this model resembles T.A. van Dijk's (Dijk, Kintsch 1983) superstructure describing real procedures of thinking. The tactical level in the proposed theory is the level of the argumentative elementary unit; this intra-argument level is used here for analysis of logical correctness of the unit of argumentation. Since the recipient has nothing but text as objective data for analysis, he can establish its logical correctness basing on the degree of its optimality of encoding. In other words, not only the contentive, but also the exponential part of the text matters for establishing its logical correctness as viewed by the recipient. For this level a new version of syllogistic is applied; its syllogisms are sensitive both to the form and to the content. The syllogistic is operating on the structures resulting from argumentative-functional analysis of the text. These structures are argumentative units.
Specific argumentative conditions are Principles of Maximalism and of Discretion. These principles are differently oriented. According to the Principle of Maximalism, if there is no explicit quantifier (which is most often the case with humanitarian text statements) in the Claim judgment of an enthymeme and, consequently, the scope of the Claim can be different (with different modes of syllogisms taken for restoration) the recipient should choose the universal option out of the alternative "universal VS. particular". It is thus presupposed that the author of the text has made the stronger statement. For example, in Argumentation theories deal with dialogues the recipient can use the Principle and thus attribute to the author the universal scope of argumentation theories, i.e. all (argumentation theories) instead of the possible some (in this, he might be right if argumentation is equalized with pragma-dialectics as opposed to formal logic). A straw-man fallacy does not take place here because the attribution is not done with the purpose of further defeating the argument - refutations are beyond the scope of our study and the recipient is viewed as a sympathetic rational being attributing sincere arguments to the author. The analyzer's role is to comprehend the arguments for his / her own benefit, and the more convincing (i.e. the stronger) they are, the better.
The Principle of Discretion is quite the opposite and is oriented at choosing a particular statement. Maximalism works in accordance with the Principle of Charity: it is oriented on a greater scope (and, hence, greater force) of the author's argument. Discretion is oriented at "saving face" of the author: sympathizing with the author, the recipient takes measures: if the argument claim only turns out to be a particular (as opposed to a supposedly intended universal) statement as a less committant one, i.e. having less force than it could have had, the author's reputation does not suffer because the author is not thought to have made a stronger statement. Discretion is also oriented at the recipient - it insures her / him from possible blame of making a quantitatively too strong conclusion. For example, the subject in the sentence Argumentation theories deal with dialogues seen from the angle of the Discretion Principle will probably be attributed a particular, not a universal quantifier: some argumentation theories (if the recipient regards argumentation as the general field for both pragma-dialectics and formal logic; if the recipient is not sure of his own attitude he / she has the more reason to use some to say to himself / herself "You never know").
The proposed model is a specific mechanism for dentifying composition of argumentative units. Validity and logical correctness of elementary units as well as detecting missing premises need a different logical mechanism accounting for the inner structure of sentences. This logical mechanism is a linguistic version of traditional syllogistic and its degree of precision as to restoring correct modes of syllogisms is higher than in traditional or Aristotelian syllogistic. Specifications of categorial-semantic meanings for finding out Middle Terms and restrictions on the linguistic formulation of the subjects of judgments can provide the basis for a new system of syllogism reconstruction. The system is necessary for the recipient who seeks for defining what are the necessary premises the isolated judgement is based on.
Bibliography
- Anderson, A., Belnap, D., Jr. (1961). Enthymemes // The Journal of Philosophy. Vol. 58. No. 23. P. 713-723.
- Dijk, T.A. van, Kintsch, W. (1983) Strategies of discourse comprehension. New York: Academic Press.
- Fillmore, Ch. (1970) Subjects, speakers, and roles // Synthese. Vol. 21. Nos. 3-4. P. 251-274.
- Fohr, S. (1980) The deductive/inductive distinction // Informal Logic Newsletter. Vol. 2. No. 7. P. 5-8.
- Freeman, J. (1983) Logical form, probability interpretations, and the deductive/inductive distinction // Informal Logic Newsletter. Vol.5. No. 2. P. 2-10.
- Fumerton, R. (1980). Induction and reasoning to the best explanation // Philosophy of Science. Vol. 47. No. 4. P. 589-600.
- Govier, T. (1980) More on deductive and inductive arguments // Informal Logic Newsletter. Vol. 2. No. 3. P. 7-8.
- Kant, I. (1929) Critique of pure reason. / Translated by N.K.Smith. London: MacMillan,.
- Popper, K. (1974) Conjectures and refutations: The growth of scientific knowledge. London.
- Toulmin, S. (1958) The uses of argument. London: Cambridge University Press.
- Weddle, P. (1979) Inductive, deductive // Informal Logic Newsletter. Vol. 2 No. 1. P. 1-5.
 Text in Word
Back to LIBRARY
Back to HOME PAGE
| |
