Kintsch and van Dijk (1978): Proposition analysis

Kintsch and van Dijk (1978) approach text from the comprehension side. With the help of their analytical model, the Process Model, their aim is to “describe the system of mental operations that underlie the process occurring in text comprehension and in the production of recall and summarization protocols”

(p.363). They claim that the surface structure of discourse is interpreted as a set of propositions, ordered by various semantic relations. Some of these are explicitly expressed in the surface structure of the text, while others are inferred during the process of interpretation with the help of context-specific or general knowledge.

The semantic structure of texts can be characterized at two levels: a micro-structural and a macromicro-structural level. The microstructure is the local level of

93 Discourse analysis and language teaching

discourse, the structure of the individual propositions and their relations. The macrostructure, on the other hand, constitutes the global level of discourse and characterizes it as a whole. The two structural levels are connected via so called semantic mapping rules, or in short, macrorules. Kintsch and van Dijk (1978) assert that a text is coherent “only if its respective sentences and propositions are connected, and if these propositions are organized globally at the macrostructural level” (p.365).

Arriving at the macrostructural level (i.e., seeing the set of macroproposi-tions which constitute the text) is the result of different processes. Because the macrostructure must be implied by the explicit microstructure from which it is derived, first, the propositions in the text base need to be identified (this is also called microstructural information). They form proposition sequences, from which, with the help of semantic mapping rules, the semantic macrostructure of the discourse can be derived. This is what is called gist in every-day terms.

The analytical tool with the help of which the analyst can describe the processes which underlie the process of text comprehension and production is called by Kintsch and van Dijk as the Process Model. Since the meaning of a text can be described in terms of a structured list of propositions ordered according to the order of words in the text, the input of the analysis is a list of propositions. The notion of proposition is used by Kintsch and van Dijk in a special sense to denote concepts and not words. Table 4 presents the list of propositions for the extract below (taken from Kintsch and van Dijk’s sample analysis).

The horizontal lines in Table 4 indicate sentence boundaries, the numbers equal propositions, the numbers in brackets refer to the propositions with that number, the words in capital letters in brackets are the actual propositions, and the capital letters are to emphasize the fact that these are concepts and not words from the text. Kintsch and van Dijk admit in their study that this is a rather superficial presentation of the “meaning” of this text, but claim that, for their psychological processing model, they needed a non-elaborated semantic representation that is close to surface structure; something similar to what goes on in the reader’s mind while constructing potential interpretations. The goal of their analysis is to present these interpretative processes which change the initial semantic representation to a deeper, less surface-dependent structure.

After making a list of all the micropropositions, the first step in the processing model is to organize them into a coherent graph that visualizes the structure/hi-erarchy of the propositions. Kintsch and van Dijk claim that a text is processed in cycles, the first n propositions in cycle 1, then the second n propositions

94 Krisztina Károly

in cycle 2, in temporal order, in chunks of several propositions at a time (for a detailed description of this process, see p.379).

As regards the nature of coherence, Kintsch and van Dijk (1978) draw the following conclusion on the basis of their investigations:

Sentences are assigned meaning and reference not only on the basis of the meaning and reference of their constituent components but also relative to the interpreta-tion of other, mostly previous sentences. Thus, each sentence or clause is subject to contextual interpretation. The cognitive correlate of this observation is that a language user needs to relate new incoming information to the information he/she already has, either from the text, the context, or from the language user’s general knowledge system. We have modeled this process in terms of a coherent propo-sitional network: The referential identity of concepts was taken as the basis of the coherence relationships among the propositions of a text base. (p.389)

95 report of a study that I undertook to assess

(LOCATION: AT, CAL STATE COLL LOS ANGELES) (IS A, STUDENT, BLACK PANTHER)

(IS A, STUDENT , BLACK PANTHER) (BEGIN, 17) Table 4: Proposition list of the extract

96 Krisztina Károly 2.3.7. Kaplan (1966, 1987): Contrastive rhetoric

It may be stated with considerable confidence that the most outstanding figure in the study of contrastive rhetoric is Kaplan, whose “doodles article” (1966) is cited in almost every single study that has been born in the subject ever since.

His work has provoked lengthy debates and an infinite number of attempts to prove or undermine his beliefs.

Kaplan (1966) starts out from a predominantly pedagogical perspective, claiming that “a fallacy of some repute and some duration is the one which assumes that because a student can write an adequate essay in his native language, he can necessarily write an adequate essay in a foreign language”

(p.3). He claims that proficiency in a language, demonstrated by, for example, syntactic mastery, does not necessarily entail adequate composing. Comments such as “out of focus,” “lacks organization,” “lacks cohesion” on the part of the teacher are all due to unexpected thought-sequences, or in other words, strange rhetoric. The English thought pattern, he claims, is dominantly linear in its development. It starts with a topic statement in each paragraph, supported by examples or illustrations, and proceeds to develop the central idea and relate that idea to all the other ideas in the whole essay. To bridge this gap between the essays of foreign students and the expectations of the teachers, Kaplan set out to describe the rhetorical organization of texts, written by foreign students who speak different native languages, in order to be able to include the results of contrastive rhetoric in composition teaching.

In his experiment 700 compositions were analyzed, all written by foreign students. Out of the 700, 100 were disregarded for being too small samples.

The remaining 600 compositions were broken down to three main groups:

GROUP I. (129 compositions): Arabic, Hebrew

 GROUP II. (381 compositions): Chinese (Mandarin), Cambodian,

 Indochinese, Japanese, Korean, Laotian, Malaysian, Thai, Vietnamese GROUP III. (88 compositions): (Spanish-Portuguese) Brazilian, Central

 American, South American, Cuban, Spanish, French, African, (Italian) Swiss

He found that in the case of native speakers of Arabic and Semitic languages paragraph development is based upon a complex series of “parallel” construc-tions. He identified four main types of parallel structures: (1) synonymous parallelism (mostly realized by the use of the coordinating conjunction and), (2) synthetic parallelism (this means the completion of the idea/thought of the first part in the second, using conjunctive adverbs such as therefore), (3)

97 Discourse analysis and language teaching

antithetic parallelism (this means the presentation of a contrasting idea in the second part, using but in the majority of cases), and (4) climactic paral-lelism (when the idea of the passage is not completed until the very end of the passage).

In Oriental (mainly Chinese, Korean, but not Japanese) cultures, the de-velopment of paragraphs seem to follow an indirect approach: the ideas are

“turning and turning in a widening gyre” (p.10). The circles or gyres of ideas turn around the subject and show it from a variety of tangential views, but the subject is never looked at directly. As Kaplan puts it: “things are developed in terms of what they are not, rather than in terms of what they are” (p.10).

Native speakers of Neolatin languages (French, Latin American Spanish, Spanish) displayed digressions, illustrations in their compositions, which were interesting in many cases, but did not contribute significant structural material to the basic thought of the paragraph.

Finally, texts composed by Russian writers also show considerable rhetori-cal differences. Most of these differences could be explained in terms of the Russian sentence structure, which greatly differs syntactically from English. It allows very long, complex sentences, with only commas separating the various ideas presented. These texts also mention numerous irrelevant details, which constitute only parenthetic amplifications of structurally related subordinate elements.

Kaplan’s main conclusion is that second-language teaching inevitably faces the problem of cultural variation, at the sentence level (grammar, vocabulary, sentence structure) as well as at the discourse level. Logic, the basis of rhetoric, is evolved out of a culture, hence it is not universal: it varies from culture to culture, and from time to time. Kaplan does not say that any of these writing patterns are inferior, but what he claims is that they exist, and create difficulties for even advanced speakers of English. He sees the solution of this problem in conducting further research to acquire a more accurate description of these cultural differences so that the explicit teaching of rhetoric could be imple-mented in second-language instruction.

Kaplan was fiercely criticized for his 1966 article. This made him return to the ideas he put forward and revise some of his stronger statements as well as provide evidence for others which have been tested since then (Kaplan, 1987).

He still claims that there are important differences between languages in the way in which discourse topic is identified in a text and in the way discourse topic is developed in terms of, for instance, exemplification or definition.

He admits, however, that the case he made in his earlier article is too strong.

98 Krisztina Károly

The characteristics mentioned do not always occur, but languages have certain preferences towards certain types of development, so that “while all forms are possible, all forms do not occur with equal frequency or in parallel distribu-tion” (p.11).

The reason why Kaplan considered and still considers these issues important from a pedagogical point of view is because while native speakers are able to re-cognize the circumstances in which various forms are used and that the choice he/she makes constrains any text which may follow, the non-native speaker is not. In addition, in contrast to native speakers, the non-native speaker also lacks a complete inventory of possible alternatives that could be used. Thus it is the responsibility of the teacher to help the non-native speaker make up for these defects through explicit teaching and awareness raising.

3. Academic discourse: cross-cultural comparisons of the