Cohesive Devices Chen, 2008
Cohesive Devices Chen, 2008
Jo-Ling Chen
National Tsing Hua University
Abstract
93
INTRODUCTION
In recent years, researchers have given considerable attention to how EFL and
ESL learners actually write and what problems they usually encounter in their writing.
One of the widely explored subfields of second language writing is the construct of
coherence. As one of the essential criteria for assessing writing, coherence has been
regarded as a subjective and hazy concept which is hard to learn and teach (Crewe,
1990; Lee, 2002). Several studies have indicated the problems that L2 writers have
while writing (Chen, 2007; Crewe, 1990; Kanno, 1989; Wu, 2006; Yasuko, 1989). For
example, Crewe (1990) identified two problems with regard to incoherent writing that
most Hong Kong college students had. One was using numerous connectives without
discerning the semantic differences among them, such as using “on the contrary” for
“however”. The other problem concerned overusing connectives. Inexperienced
writers tended to use more connectives to maintain surface logicality, but actually
there was no logicality in their writing. In addition, Lee (2002), as a writing teacher
and researcher, also admitted that the concept of coherence was not definite so that
writing teachers had difficulties in teaching and assessing students’ writing.
94
sentences. To provide a framework for studying and judging the cohesion and
coherence of writing, Halliday and Hasan (1976) introduced five different types of
cohesive ties: (1) reference (i.e., the indication of information from elsewhere such as
personals, demonstratives, and comparatives), (2) substitution (i.e., the replacement of
one component by another), (3) ellipsis (i.e., the omission of a component), (4)
conjunction (i.e., the indication of specific meaning which presupposes present items
in the discourse, such as additive, adversative, casual, and temporal), and (5) lexical
cohesion (i.e., the repetition of the same or relative lexical items). They contended
that through analyzing the use of cohesive device, one could evaluate or assess
writing quality from the perspective of coherence.
The cohesive theory proposed by Halliday and Hasan (1976), however, was
challenged by Carrell (1982) and Johns (1986) who argued for the importance of
readers’ background knowledge. Based on schema theory, Carrell (1982) contended
that “processing a text is an interactive process between the text and the prior
background knowledge or memory schemata of the listener or reader” (p. 482). In
other words, not only the structure and content of the text, but also the readers’
operation on the text should be taken into consideration. To support his criticisms of
the cohesive view of coherence, Carrell scrutinized three empirical studies and
claimed the relationship between the number of cohesive devices and coherence does
not actually exist.
In order to obtain further information for this study regarding the construct of
coherence and the relationship between cohesion and coherence, nine empirical
studies that investigated the correlation of the use of cohesive devices and the overall
quality of writing have been reviewed. This review found that these studies contained
controversial results. For example, some studies have contended that there is a
positive correlation between the number of cohesive devices and good writing (Ferris,
1994; Field & Oi, 1992; Jin, 2001; Liu & Braine, 2005). In contrast, other studies
have not shown a significant link between the number of cohesive features and the
quality of writing (Castro, 2004; Jafarpur, 1991; Johnson, 1992; Neuner, 1987; Zhang,
2000). Furthermore, among these nine studies, only two studies (i.e., Liu & Braine,
2005; Zhang, 2000) were conducted in China, and none of them were carried out in
Taiwan. The last fact indicates the need to carry out research in this area in Taiwan.
To gain more insights into the conflicting issues surrounding cohesive devices in
Taiwan, the present study investigated college students’ use of cohesive devices and
the relationship between the number of cohesive features and writing quality. Several
95
specific research questions addressed by this study are: (1) What kinds of cohesive
devices are used by college students? (2) How frequently are the cohesive devices
used? (3) What problems do the college students have while using cohesive features?
(4) Is there any relationship between the number of cohesive features and good
writing?
Data Analysis
Following Liu & Braine’s study (2005) and Zhang’s (2000)’s study, the present
study analyzed the data through two procedures: identifying the cohesive devices and
evaluating the overall quality of the essays. For the first phase of data analysis,
Halliday and Hasan’s (1976) cohesion framework was selected and adopted due to its
comprehensive and well-developed taxonomy. The number of cohesive features that
occurred in each category were counted, and descriptive statistics such as frequency,
mean, and standard deviation were computed by the SPSS statistical software package.
However, like Liu & Braine’s study (2005) and Zhang’s study (2000), two categories
of cohesive devices, substitution and ellipsis, were not analyzed since “they are
seldom used in formal writing” (Liu & Braine, 2005, p. 647).
The overall writing quality was determined by the researcher and an independent
rater who was also an experienced writing teacher. A holistic rating scale (ETS, 2004)
ranging from zero to five points was used. A well-organized essay with few errors and
good content could get five points, while an essay full of grammatical errors and
unable to express its main ideas could get zero. The inter-rater reliability was used to
check the consistency of grading scores. Pearson Correlation was also applied in order
to answer the question concerning the relationship between the number of cohesive
devices and the writing quality.
96
RESULTS AND DISCUSSION
Results of the Essay Evaluation
Forty-six essays were evaluated by two raters for the present study. The
inter-rater reliability was .895, which indicated the overall writing scores were
consistent and reliable. Therefore, the scores from two raters were averaged, and the
averaged scores were correlated with the number of cohesive devices used by students.
Table 1 provides the mean score, standard deviation, and the ranges of the 46 writing
scores.
Table 1
Mean, Standard Deviation, Range, and Other Statistics Related to the Essay Scores
Mean Standard deviation Minimum Maximum Range Median
3.53 0.5 2.25 4.5 2.25 3.5
The results reflected in Table 1 show that the mean score of the 46 essays is 3.53
and the standard deviation is .50. Therefore, essays with four points or above were
regarded as good essays and essays with three or below were considered poor ones. In
addition to distinguishing students’ writing quality, most of the scores ranging
between three and four also indicated that participants’ average writing proficiency
was high-intermediate, and they have a similar writing ability.
97
These findings are in line with previous studies (Liu & Brain, 2005; Neuner,
1987). Although they do not entirely correspond to Zhang (2000) in whose study the
percentage of conjunction devices are higher than the percentage of reference devices,
it is still evident that the number of lexical devices surpasses either reference devices
or conjunction devices. One explanation for this result may be that L2 learners tend to
use diverse lexical devices to express their ideas no matter which writing genre is
employed. To scrutinize the variety of cohesive devices used, a more detailed
examination of each category of device (i.e., reference, conjunction, and lexical
devices) is presented below.
In terms of the most frequently used cohesive items, it is interesting to find that
the participants in this study strongly preferred using “you” (414 counts) in their
essays (see Examples 1 & 2). It might reflect the fact that most of them did not take
their audience and their voice into consideration. With less use of third person
pronouns, students made their writing more subjective and personal. This observation
could help writing teachers become more aware of teaching voice and target audience
in their writing classes.
Example 1
The reason is that you may have lots of things to worry about and to deal with so
that you will need a friend to help you to let go your anxieties.
Example 2
Studying abroad offers you a chance to get alone with other people in foreign
country.... Although studying abroad will cost you a lot of money, high tuition
and high living expense … then you could consider about studying abroad.
98
Table 3
Reference Devices Used in Essays
Reference devices Pronominal Demonstratives The definite Comparative Total number of
devices article devices reference devices
Table 4
Conjunction Devices Used in Essays
Conjunction Additive devices Adversative Causal Temporal Total number of
devices devices devices devices conjunctive devices
Frequency 552 147 141 102 942
Mean per essay 12.00 3.20 3.07 2.22 20.48
Standard 5.41 1.78 1.79 1.43 6.41
deviation
Range 26 7 7 6 33
Percentage 59% 16% 15% 11% 100%
Most frequently And, or But, however Because, First, second,
used cohesive so in the end
items
99
conjunction device. For instance, they seldom employed phrases such as “in addition”,
“on the other hand”, “on the contrary”, “as a result”, and “all in all” in their writing.
Instead, they preferred using “and”, “but”, “so” and “because”. Writers’ preference for
using the simpler conjunctions could indicate difficulty in using phrases or a lack of
familiarity with other conjunctions. The following examples are extracted from the
students’ essays.
Example 3
Perhaps there are some terrible memories about argument before, but we all
believe we will get friendship forever. Because of this episode in my growing
process, I notice that …. However, one thing or event usually has two sides, a
positive side and a negative side.
Example 4
Firstly, it is no doubt that Taiwan is small and the world is huge… and this is
really helpful for learning more and deeply in your major field. But people may
wonder that being a student abroad really costs so much, is it worth? However,
you pay what you get.
The results corresponding to Liu & Brain (2005), Neuner (1987), and Zhang
(2000) indicated that L2 students preferred to use the same lexicon rather than
synonyms, collocations and antonyms to express ideas or describe objects. A possible
explanation for this tendency is that L2 students have a limited repertoire of
vocabulary and they have difficulty using a variety of words in their writing (Zhang,
2000). Furthermore, the most repeated words such as “friendship”, “friend”, “abroad”,
and “study” were tightly related to the topics or came from titles.
100
Table 5
Lexical Devices Used in Essays
Standard
22.09 2.41 5.13 2.75 2.55 26.99
deviation
Range 127 8 26 13 11 150
Percentage 70% 4% 11% 6% 10% 100%
Most frequently Friendship, Advantage- Study- Relationship-
used cohesive friend, disadvanta learn friendship-
items abroad, have, ge friend
study
Pronoun shift
One of the common problems found in the students’ writing was pronoun shift.
Pronoun shift refers to a grammatical error in which the writer uses a specific kind of
pronoun in a sentence or a paragraph and then suddenly shifts the pronoun to another.
Such errors not only cause the readers to be confused but also mix up the reference
use. The following examples taken from the participants’ essays demonstrate such
erroneous use.
Example 5
It is really hard to find their flam because they are always disingenuous, and the
worst is that you may be affect to leave your true friends. Just like someone said,
“better an open enemy than a false friend.” So I think the most important
character about true friendship is honesty.
101
Example 6
As a result, I really believe that we should not waste money studying abroad,
since you can receive all the same resources in Taiwan where it is much more
familiar and safer for you.
Run-on sentence
The second problem was run-on sentences (see Example 7 & 8). A “run-on
sentence” means that a sentence consists of two or more main clauses and the clauses
are joined without proper punctuation marks or conjunctions. Run-on sentences can
be attributed to a number of causes, including not knowing how to utilize
conjunctions appropriately and not understanding the difference between dependent
and independent clauses.
Example 7
Sometimes, it’s hard and tiresome to get along with someone because different
opinions and arguments exist wherever people are together, but from another
aspect, this is exactly the reason we can learn lots of things from friends,
including dealing with....
Example 8
When it comes to the issue of higher education, many people would choose to go
abroad and pursuit their dreams, but for me, the most ideal place to continue my
further studying is the place where I stared it., staying at the country where one
grow up and study is the best policy because….
Example 11
With friendship’s help, you recovered your lost energy…. (friends’)
Example 12
Like one day you find an amazingly delicious cheesecake, the first person burst
into your mind to share with must also be…. (comes into your mind)
102
Example 13
Of course, you will be homesick, but you can use internet technology, such as,
SKYPE, MSN or email to contact with your family. (communication technology;
contact your family)
One of the possible reasons for some of the errors in the above examples is that
students learned lexical items only through Chinese translation. They did not learn
vocabulary within the appropriate semantic context. Thus, it was difficult for them to
master the specific and correct use of the lexicon. In addition, the students were easily
influenced by their native language—Chinese. Some of misuse was a direct
translation from Chinese lexical patterns (Zhang, 2000).
Example 9
Aside from this , friendship can be describe as a wine , as it last longer it more
valuable , a true and worthy…, although you have long time do not contact with
each other , but it still will be there . Thus , a true and worthy friendship is not
simple… , the definition of it….
Example 10
Because oversea studying in top university helps students become a mature,
multi-culture, and resilient person, additionally, you also have excellent English
ability and interpersonally adept, and all these qualities are required in
international enterprise. Thus, you will have better job and promotion.
The Relationship Between the Number of Cohesive Devices and Writing Quality
In addition to understanding the kinds of cohesive devices used by college
students and the frequency of use, the main purpose of the study was to examine the
relationship between the number of cohesive devices and writing quality. Therefore,
the numerical essay scores and the number of each cohesive category (i.e., reference,
conjunction, and lexical) were correlated by Pearson.
103
Table 6
Correlation Between Essays Scores and Cohesive Devices
Essay scores Reference Conjunction Lexical Total number of
devices devices devices cohesive devices
Essay scores 1
Reference devices -0.005 1
Conjunction devices 0.242 0.285 1
Lexical devices 0.265 0.446** 0.747** 1
Total number of 0.208 0.718** 0.749** 0.936** 1
cohesive devices
*p<0.05
**p<0.01
104
inferred. Thus, the negative effect of overusing cohesive devices should be taken into
consideration when evaluating writing coherence.
Pedagogical Implications
According to the findings and discussions illustrated above, the present study
provides some suggestions for both writing teachers and EFL students. First, since the
number of cohesive devices has little influence on the quality of writing, “cohesion is
not coherence” (Carrell, 1982) could be realized. A composition with more cohesive
devices cannot be regarded as a coherent one. Therefore, when writing teachers
instruct students in employing cohesive devices, it is also important to remind their
students that a “proper” dose of cohesive devices makes writing better.
Second, teachers could help students enlarge their choice of vocabulary because the
findings indicated that the participants preferred repeating words rather than using
synonyms and antonyms to describe the main points of their topic. Writing teachers
could engage the students in some vocabulary activities such as word association
game before writing to elicit and build students’ vocabulary. Third, writing teachers
could introduce corpora (e.g., Collins COBULD Corpus) to students since corpora can
enhance learners’ awareness of lexico-grammatical patterning of texts (Thurstun &
Candlin, 1998; Yoon, 2008). Corpus-informed syllabi can be combined with writing
courses in which students can learn to solve their lexical problems through
concordances and collocation samplers. Last but not least, to enhance students’
awareness of coherence and cohesion, it is essential to incorporate reading into
writing (Heller, 1995; Hirvela, 2004). Students can acquire vocabulary, syntactic
structures, or features of genres through the process of reading to write. What’s more,
observing the use of cohesive devices can also enhance students’ awareness of the
characteristics of good English writing. It is expected that the suggestions briefly
mentioned above could improve students’ writing skills and promote writing teachers’
teaching quality.
Conclusion
The current study used Halliday and Hasan’s (1976) cohesive framework to
analyze students’ use of cohesive devices. The findings showed that the participants
employed a variety of cohesive devices in their essay writing. Among the three
cohesive devices examined, lexical devices were the most frequently used, followed
by reference devices and conjunctions. In addition to the frequency and percentage of
cohesive device use, the study also presented several writing problems that most of 23
EFL college students had: pronoun shift, run-on sentences, overuse of cohesive
devices, and misuse of lexical items. These problems can be regarded as common
105
errors in EFL students’ writing. Thus, such analysis is valuable in understanding
students’ writing problems.
The study also offered empirical evidence for distinguishing the construct of
coherence and cohesion. Since there was no significant correlation between the
number of cohesive devices and the holistic writing scores, the results of previous
studies (Castro, 2004; Jafarpur, 1991; Johnson, 1992; Neuner, 1987; Zhang, 2000)
were confirmed. However, the generalization of the results to other populations with
different learning backgrounds may be limited because the study merely involved 23
participants. Therefore, further research has to be carried out with larger sample sizes
to find out whether the results are consistent with previous studies.
REFERENCES
106
Heller, M. F. (1999). Reading-writing connections: From theory to practice. (2nd ed.).
NY: Longman.
Hirvela, A. (2004). Connecting reading and writing in second language writing
instruction. Ann Arbor, MI: University of Michigan Press.
Jafarpur, A. (1991). Cohesiveness as a basis for evaluating compositions. System,
19(4), 459-465.
Jin, W. (2001). A Quantitative Study of Cohesion in Chinese Graduate Students’
Writing: Variations across Genres and Proficiency Levels. Paper presented at the
Symposium on Second Language Writing at Purdue University (West Lafayette,
Indiana, September 15-16, 2000).
Johns, A. M. (1986). Coherence and academic writing: Some definitions and
suggestions for teaching. TESOL Quarterly, 20(2), 247-265.
Johnson, P. (1992). Cohesion and coherence in compositions in Malay and English.
RELC Journal, 23(2), 1-17.
Kanno, Y. (1989). The use of connectives in English academic papers written by
Japanese students. Psycholinguistics, 2, 41-54.
Kuo, C. H. (1995). Cohesion and coherence in academic writing: From lexical choice
to organization. RELC Journal, 26, 47-62.
Lee, I. (2002). Teaching coherence to ESL students: a classroom inquiry. Journal of
Second Language Writing, 11, 135-159.
Liu, M., & Braine G. (2005). Cohesive features in argumentative writing produced by
Chinese undergraduates. System, 33, 623-636.
Neuner, J. L. (1987). Cohesive ties and chains in good and poor freshman essays.
Research in the Teaching of English, 21(1), 92-105.
Thurstun, J., & Candlin, C. (1998). Concordancing and the teaching of the vocabulary
of academic English. English for Specific Purposes, 17(3), 267-280.
Wu, S. R. (2006). Connectives and topic-fronting devices in academic writing:
Taiwanese EFL student writers vs. international writers. 2006 International
Conference and Workshop on TEFL and Applied Linguistics, 417-425.
Yoon, H. (2008). More than a linguistic reference: The influence of corpus technology
on L2 academic writing. Language Learning and Technology, 12(2), 31-48.
Zhang, M. (2000). Cohesive features in the expository writing of undergraduates in
two Chinese universities. RELC Journal, 30(1), 61-95.
107