Specificity of comprehension of natural-language arguments in scholarly texts

Kaluga State Pedagogical University, Russia
Department of Foreign Languages, 26 Razin Street, Kaluga 248023, Russia

Summary

Argumentation theories are competitive as well as conceptions of compre-hension of arguments. An argumentation comprehension linguistic theory must concentrate on both strategic and tactical organization of natural language ar-gumentative texts. Non-rigidity of arts texts serves as a ground for applying a non-rigid principle to the text analysis. Being inapplicable to natural-language argumentation in arts texts, sentential logic can still be used if an informal sys-tem is employed. Comprehension of monologue text correlates with a commu-nicative method of argumentation as opposed to an interactional one. Felicity conditions in the communicative method are those of informal-logical correct-ness and validity of argument organization and of predictive procedures of ar-gument comprehension in the absence of feedback. Semiosis of a monological argument viewed as a sign is primary to a dialogical argument. Argumentation is viewed as a subtype of reasoning, the other subtype being explanation. Ex-planations are usually found in everyday discourse; they also occur in research texts but are given as evidence, as a rule.

Because of the semiotic nature of a text a system for its analysis must be semiologically-oriented. The system must be intensional because reasoning texts are also intensional. The system must be able to do with both the form and the meaning of the text. Two aspects of creating such a system are described.

The first has to do with a tactico-strategic organization of the text argu-mentation. It is a modified version of Toulmin's (1958) argumentative-functional system. The model is modified because of a number of drawbacks in treating validity, linguistic expression of Warrant, style-specificity, exces-siveness of the functions of Backing, Qualifier, and Rebuttal for elementary ar-guments (cf.:Eemeren, Grootendorst, Kruiger 1987: 199-207).

To modify the Toulmin model we differentiate between basic and non-basic argumentation units, the former having the structure [Claim + Data + Warrant]. The distinctive feature of our approach has to do with scholarly rea-soning and is this. If an argumentative paragraph consists of only one statement (expressed by a declarative sentence), the analyzer defines its argumentative function using the Principle of Charity. Thus it would be wrong to treat the sen-tence in question as Data (Datum), because the same Data can lead to very dif-ferent conclusions; instead, the recipient treats it as an enthymeme with the ex-plicit Claim where the author leaves the opportunity to find the Data to the re-cipient because it is easier to do. In case of a two-member argument one of its components is regarded by the recipient as the Claim, the other - as its Data. It is not a Warrant, because, following the Principle of Charity, the most important premise should be given first; the Warrant can be reconstructed because it con-nects the Data and Claim on the grounds of cause, correlation, identification and ethos. It is true that in everyday reasoning Data are often implicit because they reflect facts presumably known to the recipient, and the only premise is often a Warrant. In scholarly texts, however, Data do not at all always denote some-thing the recipient is or should be aware of. On the contrary, the author, em-ploying the Principle of Charity, presupposes that the recipient may not be aware of Data (some facts, details, and observations are regularly described in scholarly works) but can easily make a mental connection between the two ar-gument components thus reconstructing the Warrant; therefore, Data but not the Warrant should be given first.

The second aspect deals with the tactical argument analysis where we de-tect if the Data are coherent with the Claim, i.e. if they really support the latter but not something else instead. It is rational search for the correct Warrant. The method works even for full-member arguments, because if the recipient finds the correct Warrant, he/she can compare it with the explicit one thus estimating the quality of the author's argument.

For scholarly texts, we can apply a subject-predicate system because lin-guistic actant structures in scholarly texts are very limited in diversity. The sys-tem that can be used here is a modified version of scholastic syllogistic. Tradi-tional syllogistic is inapplicable for natural-language scholarly text analysis be-cause it is a tectonic model not dealing with the grammatical form of arguments, which often leads to a grammatical many-term syllogistic fallacy, cf.: All bodies (1) moving in elliptic orbits (2) are subject to the law of gravitation (3). Comets (4) move in elliptic orbits (5). Therefore comets are subject to the law of gravitation (Anderson, Belnap 1961: 713 - the numbers and bold type are mine and correspond to quantity and quality of the terms of the syllogism). To have a valid form, this formally valid syllogism must undergo certain language-specific grammatical transformations. Therefore, the extended syllogistic must:

• take into account the specificity of a concrete language and possibilities of different language-specific transformations;
• provide a non-ambiguous grammatical form of the subject of the judge-ment (p.ex., if the subject can be expressed by parts of speech other than nouns and pronouns); define the range of morphological variety of the predicate;
• detect the types of referential semantics of subjects and predicates (p.ex., class, description etc.)
• establish, how important is the standard (subject-predicate) form of the judgement for syllogisms and what are the necessary linguistic transformations of reducing/reconstructing natural-language utterances to that form;
• define the number of figures and modes for such syllogistic modification and what will be critical for the definition of the figure (p.ex., the position of the Middle term or something else); establish, if replacement of premises is possible in the new system.