Lexical Cohesion Pattersn in NS and NNS Dissertation Abstracts in Applied Linguistics: A Comparative Study

| January 7, 2014

December 2008. Volume 3 Issue 3
| December home | PDF version | SWF version |

Title
Lexical Cohesion Patterns in NS and NNS
DissertationAbstracts in Applied Linguistics:
A Comparative Study

Author
JIN Kai
College of Foreign Languages
Chengdu University of Technology, China

Bio-Data
JinKai is associate professor of Applied Linguistics at Chengdu University of Technology, China. She has taught English for 15 years. He has an MA in Applied Linguistics from Southwest Jiaotong University. Her research interests include discourse analysis, text linguistics & language pedagogy.

Abstract
Using Halliday & Hasan’s (1976)theory of text cohesion and coherence as the back-up and based on Hoey’s(1991) model for our analysis, the present thesis takes the genre of dissertation abstracts in the discipline of applied linguistics as the subject of our study. Fifteen abstracts are randomly selected as NS samples andNNS samples respectively. The study attempts to examine the similarities and differences in lexical cohesion patterns between a native speaker of English (NS)and a non-native speaker of English (NNS)dissertation abstracts and to account for their similarities and especially for their differences. It is found that NS abstracts tend to use more complex repetitions than NNS ones, which have a high rate of using simple repetitions. Another finding is that the patterning of lexical repetition in the sample texts could take a central place in the organization and understanding of dissertation abstracts.

Key words: cohesion;lexical cohesion patterns;dissertation abstract

1. Introduction
As one of the most important cohesive mechanisms for the actualization of textual coherence, lexical cohesion has recently led to much debate in the field of applied linguistics and text linguistics. Halliday & Hasan define it as the cohesive relationship achieved by the selection of vocabulary, which is realized by certain cohesive devices and cohesive ties(Halliday & Hasan, 1976:275). The earliest study dealing with lexical cohesion can be traced back to Halliday (1962), who, for the first time, introduces the notion of cohesion and attempts to root cohesion in the framework of semantics. In his model, two general categories of cohesion are defined: grammatical cohesion and lexical cohesion. Halliday and Hasan, (1976), develop lexical cohesion to the perspective of texture. According to their categorization (1976, 1985), lexical cohesion is composed of reiteration, synonymy, hyponymy and collocation. From then on, with the development of functional linguistics and text linguistics, large quantities of remarkable achievements have been made, and studies of lexical cohesion have been in multi-level patterns, among which Hoey’s (1991) model is the most influential. Since the1990s, the emphasis on research into lexical cohesion has gradually shifted from theoretical exploration to genre-based practical analysis and to contrastive studies across languages. However, in the literature, the vast majority of research is based on Halliday & Hasan’s framework, and little research has been done to examine Hoey’s model.

(1) How are lexical cohesion patterns realized in NS and NNS abstracts? Which types of lexical cohesive devices are most frequently used?

(2) Are there any similarities and differences in the patterns of lexical cohesion between NS and NNS dissertation abstracts? What are the characteristics of lexical cohesion patterns in mature native speaker abstract writing?

(3) How may their similarities and differences be interpreted?

2. Hoey’s Lexical Cohesion Patterns
Hoey shows how it is possible to record lexical cohesion and other forms of repetition in matrix form in order to identify the bonded sentences. He introduces in his study a number of potentially computable notions such as links, ties, bonds, and bond networks in relation to lexical cohesion and to text organization. According to Hoey, links falls into the following ten categories which are numbered in decreasing order of importance (Hoey, 1991: 83):

(1) Simple Lexical Repetition
(2) Complex Lexical Repetition
(3) Simple Mutual Paraphrase
(4) Simple Partial Paraphrase
(5) Antonymous Complex Paraphrase
(6) Other Complex Paraphrase
In addition to those lexical links, there are, however, other types of textual connections that serve the same function.
(7) Substitution
(8) Co-reference
(9) Ellipsis
(10) Deixis

It will be seen that Hoey’s model is a way of investigating the patterning effects of lexical cohesion, therefore priority is given to lexical links over grammatical links. Every lexical selection affects or creates cohesive links that, as we have seen, help organize the text. On the other hand, each textual selection constrains the lexical choices possible, and it is in the combination of the lexical and textual choices that writers or speakers make that their creativity is expressed (Hoey, 1991: 219). Therefore, it is crucial to investigate how lexical cohesion and text organization affect each other.

3. Research Design
The corpus contains 30 naturally occurring abstracts in dissertations from China Doctor/Master’s Dissertation Databases (CDMD) and the Linguistlist, the world’s largest online linguistic resource. The 15 writers of NS MA dissertation abstracts were solicited through an email sent via the Linguistlist server for native speakers of English. The 15 NNS MA dissertation abstracts were written by the Chinese graduates to form a comparison group. Abstracts of the above-mentioned dissertations, which are written in English, are considered to be successful instances of this particular genre. For the NS graduates, English is their first language (L1), while for the NNS, English is their second language (L2).

In this paper, the qualitative study will go hand in hand with the quantitative study. The former will allow us to say interesting things about the relationships existing between the sentences of which it is composed and also to conduct a more detailed investigation into the way repetition organizes texts, while the latter will enable us to find whether there are striking differences between NS and NNS abstracts in terms of lexis patterns. The data analysis in the present study is carried out with SPSS (version 11.0). T-tests for small samples are used to test the null hypothesis that dissertations from native speakers (NS) will not be different from dissertations from non native speakers (NNS) with regard to their lexical cohesion patterns.

4. Analysis and Discussion
We deploy Hoey’s repetition matrix and his symbolic shorthand to analyze the data with his categories of cohesion. Vertical and horizontal columns in the matrix are numbered. By choosing any vertical column and following it down, we learn if a sentence is linked with the following sentences. By choosing any horizontal column and following it across, we learn whether a sentence is linked with the preceding sentences Tables 4.1 and 4.2 will show all the links in the abstracts of NNS-14 and NS-5 respectively.
In order to make this research more convincing, two of the 30 sampled abstracts (NNS-14 and NS-5) are studied in detail. Guided by the theories and data obtained under the criteria set above, the two abstracts are analyzed in a detailed way.

A matrix such as Table 4.1can be used for interpretation in a variety of ways. Most relevantly from our point of view, we can use it to trace a sentence’s connections with other sentences in the text, whether adjacent or non-adjacent. Horizontal columns identify a sentence’s connections with its predecessors in a text; vertical columns identify its connections with later sentences. Thus, for example, sentence 2 is shown by its horizontal column to have two links with sentence 3 and one with sentence 4Its vertical column shows it to have no links with sentences 5, 9, 10 and three links with sentence 11.

We note above the variability of density of links reflected in Table 4.1. Useful as the matrix is that we have been using, it is too complex to allow detailed investigation of this phenomenon. To explore the matter further requires that we present a matrix that shows links counted rather than itemized. Following this course of action, we redraw Table 4.1 as Table 4.2; while the former presented the links with greater delicacy, the latter allows us to compare directly closeness of connection between sentences.

 

(1)

                   

(2)

1

(2)

                 

(3)

4

2

(3)

               

(4)

1

1

2

(4)

             

(5)

1

Φ

2

Φ

(5)

           

(6)

2

1

3

1

Φ

(6)

         

(7)

1

Φ

2

1

1

2

(7)

       

(8)

2

1

3

1

1

Φ

Φ

(8)

     

(9)

2

Φ

2

1

Φ

1

5

1

(9)

   

(10)

3

Φ

4

1

2

1

2

3

3

(10)

 

(11)

2

3

1

2

Φ

1

Φ

1

Φ

Φ

 

Table 4.2: Matrix for links counted in NNS-14
Using the ten categories of Hoey’s patterns of lexis, a matrix of another abstract (NS-5) is drawn in Table 4.3 and Table 4.4 presents a matrix that shows links counted rather than itemized.

 

(1)

                   

(2)

4

(2)

                 

(3)

4

5

(3)

               

(4)

2

3

4

(4)

             

(5)

2

3

5

3

(5)

           

(6)

1

3

4

2

3

(6)

         

(7)

1

2

1

1

2

1

(7)

       

(8)

3

4

3

2

2

3

2

(8)

     

(9)

2

4

4

3

1

1

1

1

(9)

   

(10)

3

5

4

4

2

2

2

3

2

(10)

 

(11)

2

2

3

3

3

1

1

3

2

1

 

Table 4.4: Matrix for links counted in NS-5

Two sentences linked by at least three lexical items in each sentence are said to form a bond. The number three is relative, not absolute. Repetition matrices enable us to have a mode of representing the links that will accurately reflect their non-linear complexity and, at the same time, permit us to handle and interpret them conveniently. It will be seen from our analyses that sentences bonded (pairs of sentences linked by more than three links) are mostly non-adjacent pairs

4.1 Comparison of Frequency of Lexical Cohesive Devices
It will be seen from Table 4.1 and Table 4.2 that there is considerable variation in the number of repetitions a sentence may have with others. A good proportion of sentences are connected by no repetitions of any kind. Out of 55 cells in the matrix, 13 (24%) are empty; of the remainder, a further 20 show one repetition only. This means that less than two-thirds (60%, i.e. 33 out of 55) of sentence pairs are not significantly connected by repetition. The remaining third, however, show marked variation in density of linkage. There are 9 pairs of sentences linked by three or more repetitions, and only one of these is adjacent pairs. This would seem to warrant careful investigation.

Similarly, it may be seen from Table 4.3 and Table 4.4 that out of 55 cells in the matrix, 14 pairs show one repetition only. This means that 25% of sentence pairs are not significantly connected by repetition. The remaining pairs, however, show marked variation in density of linkage. There are 28 pairs of sentences linked by three or more repetitions, and only five of these are adjacent pairs.

4.1.1 Similarities
The most significant similarity lies in the fact that simple lexical repetition enjoys a higher frequency in both texts. The number in NS-5 amounts to 46% whereas NNS-14 amounts to 83%, far outnumbering the total number of the other patterns of lexis. This seems to suggest that NS abstracts tend to make use of fewer simple lexical repetitions in contrast with NNS in which more simple lexical repetition is used, which is in accordance with Zhu (2001) and Miao (2002)’s findings that simple repetition is a common phenomenon in English.

Secondly, there appear to be few instances of substitution, co-reference and ellipsis in both sample texts, which reveals that the dissertation abstracts we have selected are relatively formal.

4.1.2 Differences
The striking difference lies in the fact that the percentage of complex lexical repetition decreases from NS abstract to NNS abstract (46% and 4% in NS and NNS abstracts respectively), which shows that the NS abstract demonstrates a relatively higher level of sophistication than that of the NNS abstract because Western writers tend to use more sophisticated means of lexical cohesion than mere simple lexical repetition.

Complex lexical repetition occurs when two lexical items share a lexical morpheme, but are not formally identical, or when they are formally identical, but have different grammatical functions. It is commonplace for an English word to have inflections and derivatives, which might contribute to the higher frequency of complex repetitions in English.

In addition, the frequency of simple mutual paraphrases between the two abstracts shows some differences. Simple mutual paraphrases in NNS-14 amount to 9% in comparison to 3% in NS-5. The difference, which is not great, seems to suggest that simple mutual paraphrase is used more frequently in NNS abstracts than in NS ones. Nevertheless, it might have something to do with Zhu (2001)’s conclusion that Chinese speakers tend to use more simple mutual paraphrases, especially in literary works.
Thirdly, it seems that simple partial paraphrase, co-reference, substitution and ellipsis are not frequently used in both NNS and NS abstracts Thus texts of different genres may to some degree influence the frequency of lexical choices.
Lastly, in terms of other complex paraphrase, NNS-14 covers 4% in comparison with 2% in NS-5.

4.2 Statistical Analysis of the 30 Abstracts
In order to make this research more convincing, quantitative analysis is employed in order to find out whether there are statistically significant differences between NS and NNS abstracts in terms of lexical cohesion patterns. The data obtained is analyzed with T-tests (computations with SPSS, version 11.0)
The result shows that (1) there is no statistically significant difference between the two groups’ use of simple lexical repetition (t= 1.114, d.f.=24.273, P=.276); (2) there is a statistically significant difference between the two groups’ use of complex lexical repetition (t=2.682, d.f.=23.320, P=.013); (3) there is a statistically significant difference between the two groups’ use of simple mutual paraphrase (t=2.107, d.f.=22.636, P=.044); (4) there is no statistically significant difference between the two groups’ use of simple partial paraphrase (t=.584, d.f.=25.530, P=.565), antonymous complex paraphrase (t=.877, d.f.=17.208, P=.393) and other complex paraphrase (t=1.384, d.f.=25.630, P=.178); (5) there is no statistically significant difference between the two groups’ use of substitution (t=.845, d.f.=27.676, P=.406), co-reference (t=.449, d.f.=19.085, P=.659), ellipsis (t=.887, d.f.=27.843, P=.382) and deixis (t=.345, d.f.=27.939, P=.733).

5 Conclusion
5.1 Major Findings
The intention of our research is to explore lexical cohesion patterns of dissertation abstracts, but the motivation is to examine and account for the similarities and especially the differences in the lexical cohesion patterns between NS and NNS dissertation abstracts. Our major findings are as follows:

Firstly, Hoey’s matrix-oriented approach with its attendant categories of lexical cohesion patterns accurately reflect their non-linear complexity and, at the same time, permit us to be aware that bonded sentences are, more often than not, non-adjacent pairs and simple lexical repetition is used most frequently by both groups. But it can be seen that NNS abstracts are more dependent on simple lexical repetition compared to NS ones. This suggests that non-native speakers of English tend to make use of simple lexical repetitions to compensate for their limitedness of vocabulary and thinking in English whereas native speakers might be inclined toward variety.

Secondly, there are similarities and disparities between the frequency of lexical cohesive devices between NS and NNS dissertation abstracts. There is no statistically significant difference between the two groups’ rate of using simple partial paraphrases, antonymous complex paraphrases and other complex paraphrases. However, there is a statistically significant difference between the two groups’ rate of using complex lexical repetition and simple mutual paraphrase. Mother tongue Chinese influence might be the biggest cause of this, plus students’ vocabulary learning habits.
The finding that simple mutual paraphrase enjoys a higher frequency in NNS abstracts is compatible with Zhu (2001)’s claim that the frequency of simple mutual paraphrase in Chinese texts (esp. literary works) is higher than that in English texts, for which the juxtaposition of synonyms is the main reason (e.g. 称三赞四in Chinese vs. compliments in English,恨五骂六in Chinese vs. curses in English, etc.).

Complex lexical repetition is characteristic of mature native speaker dissertation abstracts. In English, care is usually taken to avoid the clumsy juxtaposition of the same lexical item in a passage. In addition, it is commonplace for an English word to have inflections and derivatives, which might contribute to the higher frequency of complex repetitions in NS samples. Furthermore, writers should not be encouraged to say the same thing over and over again, but they should be advised to make connections between what they are currently saying and what they said before. One of the most important ways for a writer to avoid clumsiness is by means of complex repetition, which may be an important index of growth (Stotsky, 1983).

Thirdly, in such a genre as formal as the dissertation abstracts, there appear to be few instances of substitution, co-reference and ellipsis.

Lastly, the research seems to indicate that the major lexical strings in an abstract reflect the essential messages of the article and some of them may also appear in the title of the article.

5.2 Pedagogical Implications for EFL Abstract Writing
It can be seen from our analysis that NS abstracts tend to use far more complex lexical repetition and relatively few simple lexical repetition than NNS writings. This is because NNS students of English are likely, consciously or unconsciously, to translate their ideas from Chinese into English in the process of writing. As a result, they tend to use the same words over and over again without the awareness of different cohesive devices. Therefore the number of cohesive devices used is limited and the few that are used are of a monotonous type and lacking variety. Furthermore, especially among EFL students, limitations of vocabulary and ignorance of the means whereby one can repeat in English may lead a learner to juxtapose the same lexical item clumsily in adjacent sentences, so the most important way to avoid clumsiness is by means of complex lexical repetition. Stotsky (1983) comments that “an increase in the use of morphologically complex words [i.e. complex repetition], rather than repetition of a simple word or the use of a cumbersome paraphrase, may be an important index of growth.” (as cited in Hoey, 1991: 244) What NNS university students lack above all for writing is complexity of vocabulary, which is of great importance to produce native style fluency in composition.
Therefore, words should be learnt in the morphological forms in which they are encountered, both so that the morphological range of the language in question may be acquired and because they may be the most frequent forms in use (ibid: 240). The adjective clumsy will be encountered more frequently in the forms clumsily and clumsiness; it is, therefore, these forms, not the base form (or the full range of forms), that should be learnt first. If a learner acquires their vocabulary in the most frequent forms rather than merely memorizing and manipulating the lexical items, they may quickly produce cohesion in writing.

This article has shown that EFL abstract writing in China is faced with two problems. Firstly, there is too much emphasis on simple lexical repetition and not enough focus on complex lexical repetition. Secondly, the ways learners memorize and manipulate the lexical items needs improvement. Educators might introduce lists of stems and affixes with their meanings for students to memorize, a practice which is as popular for first-language texts as it is for second. For learning vocabulary, the study of morphological features might not be a desirable end in itself, but knowledge of basic affixes helps learners decode words and for that reason has long been a part of vocabulary teaching texts. Thus, students are not only taught to recognize basic forms of words and how they combine with certain commonly occurring affixes but also taught to learn how to make good use of these related words. EFL abstract writing should raise learners’ consciousness of the significant lexical items, and encourage learners toanalyze the language and learn from their findings.

5.3 Recommendation of Further Studies
The small sample size and its selective nature may limit the generalizability of the results; thus, the findings should be confirmed with a larger sample of participants. It is also important to confirm the results with different groups of students such as those with higher or lower L2 proficiency levels. In particular, the study would need to be extended to validate its findings with non-English major Chinese graduates who may constitute a more appropriate participant group to investigate the effects of use of patterns of lexis on L2 abstract writing.

Despite the limitations, the present study suggests several directions for future research. First, the relation between use of patterns of lexical cohesion and the generic analysis of abstracts needs further investigation. Besides conducting case studies, it is equally important for studies within the contrastive rhetoric tradition to conduct large-scale research in much greater depth. These issues remain for further study.

The above research results are merely the first step in our current project. To deepen our study, we hope to conduct further in-depth analyses of use of patterns of lexis in the participants’ abstract writing samples from different levels. The results, we believe, will shed more light on the nature of Chinese EFL students’ abstract writing and contribute to the ongoing study of EFL students’ academic writing as a whole.

References
Halliday, M. A. K. 1985/1994.An introduction to functional grammar. London: Edward Arnold.

Halliday, M.A.K. 1962. The linguistic study of literary texts. In H. Lunt (Ed.), Proceedings of the Ninth International Congress of Linguists.(pp.302-307). The Hague: Mouton.

Halliday, M. A. K. and Hasan, R. 1976. Cohesion in English. London: Longman.

Hasan, R. 1984.Coherence and cohesive harmony. In J. Flood (Ed.), Understanding reading comprehension: Cognition, language and the structure of prose. (pp.139-154).Newark, DE: International Reading Association.

Hasan, R. 1985.The texture of a text. In M. A. K. Halliday & R. Hasan (Eds.).Language, text and context. (pp. 70-96). Geelong, Vic.: Deakin University.

Hoey, M. P. 1983. On the surface of discourse. London: Allen and Unwin.

Hoey, M. P. 1991. Patterns of lexis in text. Oxford: Oxford University Press.

Hui, Miao. 2002. Patterns of lexis in Chinese and English texts. Unpublished M.A. thesis. Northest Normal University, Changchun, China.

Salager-Meyer, F. 1990.Discoursal movements in medical English abstracts and their linguistic exponents: A genre analysis study, Interface4 (2), 107-124.

Stotsky, S. 1983. Types of lexical cohesion in expository writing: implications for developing the vocabulary of academic discourse. College Composition and Communication,34(4), 430-446.

Wolfersberger, M. 2003.L1 to l2 writing process and strategy transfer: a look at lower proficiency writers. TESL-EJ. 7(2), A6.

Tags:

Category: 2008