January 2006

 

The Effect of Text Cohesion on Reading Comprehension

 

by Mohammad Hossein Parvaz Orumiyeh University

and Dr Muhammad Ali Salmani-Nodoushan, University of Zanjan, Iran  

 

INTRODUCTION

The present study is an attempt at examining the effect of cohesive ties on language comprehension.  Language comprehension is an interactive process consisting of  background knowledge, cognitive tasks and conceptual abilities.  These three factors contribute most to an individual's comprehension.  Cohesion (lexical or referential), being a text feature, is decisive with regard to an individual's comprehension of a passage, particularly to non-natives.  In the following sections the relationship between this feature and of the text and the cognitive processes involved will be discussed at large.

Over centuries language analysis has been approached analytically. The most important characteristic of these approaches is that they consider language to be a self-contained system which is independent of the pragmatic environment. Moreover, language was considered to be made up of parts and the study of language meant the study of its parts. This view is well illustrated in the traditional grammars so far written.

On the other hand, newer approaches to language have viewed language as a synthetic phenomenon. In other words, in the study of language, one should take into account a good number of social, cultural, and situational factors that are assumed to affect language use and its features. In such a view, not only the linguistic code but also a knowledge of the communicative value of the linguistic code in relation to its linguistic and situational context is considered.


TEXT AND DISCOURSE

A distinction is usually made between the words text and discourse. Nourmuhammadi (1988) defines text as "the formal properties of a piece of language. A text is regarded as an exemplification of the operation of the linguistic code at an intra-sentential level." So, a text is a combination of sentences as formal linguistic objects. On the other hand, the use of such a sentence combination is referred to as discourse.

Texts are better to be studied in terms of their own features. For one may think that because it is a combination of sentences, it should carry the characteristics of a sentence. Texture refers to a text with its related features. Different types of features have been distinguished and defined by researchers with three different viewpoints. These include: the procedural approach to text, the functional approach, and the schema-theoretical approach. Of these three approaches, we are interested in the third one i.e. the schema-theoretical approach. In this approach, the text itself does not carry any meaning; it is the text user who is responsible for the interpretation of the text on the basis of the clues that exist in the text. In this section, we will review the literature on this approach.


REVIEW OF LITERATURE

Cohesion has been defined in a number of ways. Widdowson defines it in terms of the distinction that is made between the illocutionary act and the proposition. In his view (P.52), propositions, when linked together, form a "text" whereas illocutionary acts, when related to each other, create different kinds of "discourse."

According to Halliday and Hasan (1976), cohesion and register enable us to create a text. Register is concerned with what a text means. It is defined by Halliday and Hasan as the "set of semantic configuration that is typically associated with a particular class of context of situation, and defines the substance of the text."

Cohesion, as contrasted with register, is not concerned with what a text means. Rather, it refers to a set of  meaning relations that exist within the text. These relations are not of the kind that link the components of a sentence and they differ from sentential structure. The discovery of these meaning relations is crucial to its interpretation. For instance, in the following text:

Mary bought a new pencil. She put it in her drawer.

The interpretation of the elements she and it is dependent on the lexical items Mary and Pencil. So, cohesion is in the semantic relation that is setup between these elements.

According to Halliday and Hasan, the function of cohesion is to relate one part of a text to another part of the same text. Consequently, it lends continuity to the text. By providing this kind of text continuity, cohesion enables the reader or listener to supply all the components of the picture to its interpretation. Halliday and Hasan hold that cohesion in its normal form, is the presupposition of something that has gone before in the discourse, whether in the immediately preceding sentence or not. This form of presupposition is referred to as anaphoric. The presupposing item may point forward to something following it. This type of presupposition is called cataphoric. On the other hand, exophoric and endophoric presuppositions refer to an item of information outside and inside the text, respectively.

Halliday and Hasan recognize five types of cohesive devices in English and in the lexicogrammatical system of the language. They are reference, substitution, ellipsis, conjunction, and lexical cohesion. Reference, substitution, and ellipsis are grammatical; lexical cohesion is lexical; conjunction stands on the border line between the two categories. In other words, It is mainly grammatical but sometimes involves lexical selection.

Constructionalists view language comprehension as an interactive process between the text and the person using the text. They assume that meaning does not exist in the text but becomes available to the reader as a result of his own contribution. Language users employ text in comprehension as a set of guidelines to the active (re)creation of meaning.

Jonz (1987) in his explanation of the advantage(s) of adopting a constructionist point of view says: 

... one is able to speculate on the structure of language knowledge and on the various stages in the acquisition of such structures as well as their application to the cognitive tasks involved in comprehending.

From the above statement, it follows that constructionists emphasize the role of background knowledge as a feature of a text; and the cognitive tasks involved in the comprehension process. Below we will discuss these key points i.e. background knowledge and cognitive tasks at large.

Coady (1979) presents us with a psycholinguistic model of reading in which he illustrates the interaction of cognitive tasks with background knowledge in a reading task.

He defines the term conceptual ability as general intellectual capacities, and process strategies as various subcomponents of reading skills which also apply to oral language. Regarding background knowledge, he believes that it will become an important variable when we notice students with Western backgrounds of some kind learn English faster, on average, than those without such kind of background.

Carrel and Eisterhold (1983) consider language background knowledge an important factor in comprehending a text; they express this importance as follows: 

Efficient comprehension requires the ability to relate the textual material to one's own knowledge. Comprehending words, sentences, and entire texts involves more than just relying on one's linguistic knowledge.

Further in their article, Carrel and Eisterhold (1983) talk of two types of background knowledge: formal and informal. Formal knowledge refers to the reader's knowledge of the rhetorical organizational structures of different types of texts; content knowledge refers to the content area of a text. They also believe that reader's failure to provide the proper formal and, particularly, content knowledge (schema) would result in various degrees of non-comprehension.

Farhady (1982), in an attempt to examine the importance of learner characteristics (i.e. his schema) in relation to learner performance on ESL tests, comes up with significant differences between his subjects with different major fields. He also points out that this difference, as a variable, should be esteemed in the tests that are designed in such a way as to refrain from pushing any sort of injustice against learners in a heterogeneous class.

Discussing the cognitive processes involved in reading a text, Eisterhold (1983) distinguishes two basic modes of information processes: bottom-up and top-down. He further elaborates on how these two modes function in a schema theory model. He says: 

Schemata are hierarchically organized, from most general at the top to most specific at the bottom. As these bottom-level schemata converge into higher level, more general schemata, these, too, become activated. Top-down processing, on the other hand, occurs as the system makes general predictions based on higher level, general schemata and then searches the input for information to fit into these partially satisfied higher order schemata.

From the above quotation one may infer that these two modes function separately. However, both these modes function simultaneously at all levels: the data needed to instantiate the schemata become available through bottom-up processing; top-down processing facilitates their assimilation if they are anticipated on the part of the listener or reader's conceptual expectations.


METHOD

160 university students (80 English majors and 80 non-English majors) served as the subjects of this study. The English majors, all taking "Advanced Translation" course in the Azad University of Meybod (in Yazd province) were normally supposed to be of higher proficiency level, than their non-English major counterparts in the same university. The non-English major subjects were all engineering students, taking "General English II." The only criteria for the assignment of subjects to the two groups were their major fields and the above-mentioned courses they were taking.


INSTRUMENT

Two cloze tests were designed out of a passage of  750 words length. The passage was chosen from a reading textbook. Then every fifth word was deleted. The first and the last sentences remained intact, resulting a passage of which 40 words were left out. In the second version of the test, first all the cohesive ties were identified according to the taxonomy proposed by Halliday and Hasan (1976). Then one member of each pair of cohesive ties was deleted. The cohesive ties were either of lexical or referential type. Again leaving the first and last sentences of the text intact, we came up with a passage of which 40 words were left out.


PROCEDURE

Testing took place during the spring of 1994. In order for the test to be taken seriously, students were told that the test was part of their course requirements. To make the subjects familiar with the test-taking procedure, the instruction was orally given both in English and Farsi. For the sake of eliminating any sort of probable misunderstanding, illustrative examples were given prior to the test-taking procedure. The subjects' performance was scored using the acceptable word method. The data thus obtained were subjected to a two-way ANOVA. Tables I and II (in appendices) show the descriptive statistics and the two-way analysis of variance.


ANALYSIS

To determine the effects of each of the two variables (i.e. test format and language proficiency), a two-way ANOVA was applied to the data. In both versions of the test, English-major students outperformed non-English-major subjects. The two-way ANOVA main effect for language proficiency was F[1, 38] = 31.21, P 0.05. The difference due to the format of the test was also significant. The scores on the cohesive ties format were significantly higher than those on standard format test. The two-way ANOVA main effect for the test format was F[1, 38] = 9.93, P 0.01. All the computations were done by the employment of the SPSS Computer Software.


RESULTS AND DISCUSSION

Regarding language proficiency, the results were neither new nor interesting because it is quite obvious that English-major subjects would normally perform better than their non-English-major counterparts. The differences due to test format, however, were very interesting. All subjects, regardless of their majors, performed better on the cohesive ties format (although, again,  the English-major subjects did better). This difference in performance can be accounted for with a consideration of the fact that, in standard fixed-ratio format, deletions with regular intervals may be crucial to the meaning of the text and may sometimes leave no clue to the meaning and consequently to the words to be supplied. In the cohesive ties format, since one member of any pair of cohesive ties is left intact, enough context is provided for the testee to supply the correct words.


REFERENCES

Carrel, P. L. and J. C. Eisterhold (1983). "Schema Theory and ESL Reading Pedagogy" in TESOL Quarterly. 17, 553-573.

Farhady, H. (1982). "Measures of Language Proficiency from the Learner's Perspective" in TESOL Quarterly. 16, 43-59.

Halliday, M. A. K. and R. Hasan (1976). Cohesion in English. London: Longman UK group Limited.

Halliday, M. A. K. and R. Hasan (1980). "Text and Context: Language in a Social-Semiotic Perspective" in Sophia Linguistica. VI. Tokyo: Sophia University Graduate School of Languages and Linguistics.

Jonz, J. (1987). "Textual Cohesion and Second Language Comprehension" in Language Learning. Vol. 37, 30.

Perkins, K. and J. P. Angelis (1985). "Schematic Concept Formation: Concurrent Validity for Attained English as a Second Language Reading Comprehension?" in Language Learning. Vol. 32, 2.


 

Web Counter
LCD TV Reviews


Mohammad Hossein Parvaz and Dr Muhammad Ali Salmani-Nodoushan, http://www.znu.ac.ir/Members/nodushan.htm, 2006.   All rights reserved.