Language Awareness

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/rmla20

Use and understanding of connectives: an

embedded case study of ESL learners of different
proficiency levels

Angus Cheng & Art Tsang

To cite this article: Angus Cheng & Art Tsang (2022) Use and understanding of connectives:
an embedded case study of ESL learners of different proficiency levels, Language Awareness,
31:2, 155-174, DOI: 10.1080/09658416.2021.1871912

To link to this article: https://doi.org/10.1080/09658416.2021.1871912

Published online: 24 Feb 2021.

Submit your article to this journal

Article views: 872

View related articles

View Crossmark data

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rmla20
Language Awareness
2022, VOL. 31, NO. 2, 155–174
https://doi.org/10.1080/09658416.2021.1871912

Use and understanding of connectives: an embedded case

study of ESL learners of different proficiency levels
Angus Cheng and Art Tsang
Faculty of Education, The Chinese University of Hong Kong, Shatin, Hong Kong

ABSTRACT ARTICLE HISTORY

Appropriate use of cohesive devices is an important writing skill. Received 31 August 2019
Recently, there has been growing research on cohesiveness in ESL stu- Accepted 18 December
dents’ texts at university level. However, research into secondary school 2020
students’ writing and learners’ perceptions of cohesive devices has KEYWORDS
remained underexplored. This case study investigated the use of con- Connectives; ESL writing;
nectives, a particular type of cohesive device, by nine ESL secondary cohesion; language
school learners. In this study, qualitative data were collected from class- proficiency; case study
room observations, students’ compositions, and semi-structured group
interviews. Techniques of corpus-assisted discourse analysis were
applied using UAM CorpusTool. The findings show that the main prob-
lem faced by high-proficiency learners was the overuse of connectives,
while average- and low-proficiency learners tended to overuse and
misuse them. The findings also reveal that inadequate knowledge of
connectives, low awareness of genre, and poor understanding of writ-
ing requirements were possible reasons for these problems. Based on
the findings, practical implications for language teachers and sugges-
tions for further research are discussed.

Introduction
Writing is often regarded as the most challenging language skill for English-as-a-Second-
Language (ESL) or English-as-a-Foreign-Language (EFL) learners as it entails producing text
which is both coherent and meaningful (Nunan, 1999). Coherence is a multi-faceted construct
which is essential for effective writing (Lee, 1998). It can be achieved by various mechanisms
including cohesion, macrostructure, and propositional development (Lee, 1998). The use of
cohesive devices is one of the key elements for creating coherence in writing, as these devices
are linguistic features that give a text its texture (Halliday & Hasan, 1976).
A cohesion framework proposed by Halliday and Hasan (1976), which gave a compre-
hensive outline of the linguistic elements that connect the ideas within a text, is arguably
one of the most influential references in the area of cohesion. It has been adopted or adapted
in various studies (e.g. Crossley et al., 2016; Yasuda, 2019). According to the framework,
connectives are cohesive devices which explicitly mark major relationships between ele-
ments within a text. Connectives have been found to be especially challenging for learners
since even more experienced writers have difficulties using them correctly (Crewe, 1990).

CONTACT Art Tsang [email protected]

© 2021 Informa UK Limited, trading as Taylor & Francis Group
156 A. CHENG AND A. TSANG

Connectives are highly relevant to all ESL/EFL students as this type of cohesive device is
omnipresent in writing. Their appropriate use is one key criterion for producing coherent
essays and reports. While many studies of cohesion have focussed on ESL/EFL writing at
tertiary level, there has been little research on secondary-level learners, a target group which
is ‘the locus of most language learning in the world’ (Paran, 2008, p. 490). It is, therefore,
important to examine this topic at the secondary school level. By analyzing authentic stu-
dents’ writing with reference to the Halliday and Hasan (1976) framework, this research study
shed light on ESL secondary-level students’ use and understanding of connectives to express
various relationships in writing and to provide language teachers with insights about teach-
ing this aspect of writing more effectively. It is important to note that the present study
focussed on the cohesion framework, which is different from the topical structure analysis
(i.e. examining how meaning is created by the progression of sentence topics) employed in
studies such as those by Kılıç et al. (2016) and Schneider and Connor (1990).

Types of connectives
Connectives, also known as conjunctions, linkers, or linking adverbials, are often used as
cohesive devices that establish different logical relationships within a text (Crewe, 1990;
Thornbury, 2005). According to Halliday and Hasan (1976), conjunctions can be classified
into four major categories based on the logical relationships, namely additive, adversative,
causal, and temporal (see Table 1). It is noteworthy that this classification system is different
from the general classification of grammatical categories as it looks into how ideas are con-
nected at the discourse level. For instance, in conclusion and however are classified generally
as a prepositional phrase and an adverb respectively, but they are conjunctions in the model
in question. To avoid confusion, we will use the term connectives instead of conjunctions.
Early studies that examined cohesion and coherence in ESL/EFL students’ writing also
used this framework for cohesion analysis (e.g. Connor, 1984; Khalil, 1989). Connor (1984)
compared the argumentative writing of a class of advanced university ESL students with
that of a group of native English speakers. It was found that the ESL learners used more
connectives and fewer lexical cohesive devices than the native speakers did. Khalil (1989)
investigated a group of Arab EFL freshmen’s expository essays. Cohesive devices used in
these students’ essays were categorised and counted. It was found that the learners unduly
relied on ‘reiteration of the same word’ (p. 366) and the connectives ‘and’ and ‘also’. Both
studies concluded that the limited variety of cohesive devices used by ESL/EFL learners
would affect the overall coherence and writing quality (Connor, 1984; Khalil, 1989).

Table 1. Four major types of connectives (Adapted from Halliday and Hasan, 1976 and Thornbury,
2005).
Logical relationships Meaning Examples
Additive Add new ideas; give examples; express and, for example, moreover
similarity or emphasis
Adversative Signal contrastive relations but, however, yet
Causal Express cause and effect relationship as a result, because, so
Temporal Indicate the sequence of events and in conclusion, then, to sum up
ideas; Express conclusive relations
 Language Awareness 157

Research into connectives in ESL/EFL writing

In the existing literature, there are two major approaches to exploring how ESL/EFL learners
use connectives in writing. One approach focuses on the distribution of the four types of
connectives and their inappropriate use qualitatively (e.g. Crewe, 1990; Yin, 2015), while the
other approach compares the use and quantity of connectives in native speakers’ and
non-native speakers’ writing (e.g. Bolton et al., 2002; Chen, 2006; Lei, 2012).
In previous studies, ESL/EFL learners’ problems with connectives have usually been clas-
sified into three major types: misuse, overuse, and underuse (Bolton et al., 2002; Crewe, 1990;
Yang & Sun, 2012; Yin, 2015). Quantitative evidence shows that ESL learners do not have any
significant problems with underuse in their writing (hence resulting in, for instance, run-on
sentences), but that misuse and overuse are common (Bolton et al., 2002; Chen, 2006; Crewe,
1990; Lei, 2012). Misuse of connectives occurs when a learner uses an inappropriate one
(Crewe, 1990). For instance, some ESL learners tend to misuse ‘on the contrary’ for ‘however’
or ‘on the other hand’ when they are showing a simple contrastive relationship (Crewe, 1990).
The connective ‘on the contrary’ has an implied corrective meaning of something mentioned
previously not being true (Crewe, 1990). It has also been found that learners often use ‘on
the other hand’ wrongly, which has a contrastive meaning, as an additive connective (Field
& Yip, 1992). Another connective, ‘besides’, is also identified to be problematic as ESL students
tend to use it to add new viewpoints (Field & Yip, 1992; Lei, 2012; Yeung, 2009). The use of
‘besides’ is often discouraged in writing because it is usually used in spoken language ‘to
introduce a final and entirely different point to back up a given explanation’ (Field & Yip,
1992, p. 26). These misuses were further confirmed in a corpus study which compared
Chinese and British university students’ writing in the United Kingdom (Leedham & Cai,
Leedham & Cai, 2013). It was identified that Chinese students had a higher use of ‘on the
other hand’ and ‘besides’ as they might have ignored issues of formality (Leedham & Cai,
2013). These examples demonstrate that misuse of connectives may convey different mean-
ings from the intended ones, which in turn may mislead the reader.
Another obvious problem with connectives found in learners’ writing is their overuse. Earlier
studies investigated this from a qualitative approach (e.g. Crewe, 1990; Tang & Ng, 1995). These
studies provided examples from students’ writing that showed a large number of connectives
being packed into short paragraphs. For instance, a student in Tang and Ng (1995) study used
four connectives ‘However’, ‘In fact’, ‘But’, ‘So’ in a paragraph consisting of merely five sentences.
Crewe (1990) provided a similar example of more than ten connectives packed into three short
paragraphs. This overuse hinders the coherence of the writing, as readers may be confused
by the establishment of too many logical relationships. More recent studies have approached
the same topic differently by using quantitative findings from corpus analyses (e.g. Chen, 2006;
Lei, 2012). For instance, by comparing the ratios of frequencies of a range of connectives in
ESL learners and native English students’ writing, it was found that ‘so’, ‘and’, ‘also’, and ‘moreover’
were overused by Hong Kong learners (Bolton et al., 2002; Ma & Wang, 2016).
Similar patterns were observed in writing produced by ESL/EFL learners, seemingly irre-
spective of their L1 backgrounds. A study which compared the use of connectives in Korean
university students’ writing with a subcorpus of American students’ argumentative essays
found that the former tended to overuse additive connectives ‘moreover’ and ‘furthermore’
and temporal connectives ‘firstly’ and ‘secondly’ (Ha, 2016). Another study which compared
Chinese and English students’ research articles in four disciplines found that Chinese students
158 A. CHENG AND A. TSANG

overused ‘moreover’ and ‘furthermore’ even though their overall use of connectives was lower
than that of English learners (Gao, 2016). Recently, some attempts were made to focus on
the learners’ interlanguage. Moving beyond native and non-native learners’ writing com-
parison, Appel and Szeib (2018) compared Arabic, Chinese, and French L1 learners’ academic
English essays to identify how different L1 groups use connectives. It was identified that
Arabic L1 learners tended to overuse ‘in addition’ and ‘also’, Chinese L1 learners overused
‘however’ and French L1 learners overused ‘in fact’ and ‘indeed’ (Appel & Szeib, 2018).
All in all, the literature is replete with evidence suggesting the prevalence of errors involv-
ing the use of connectives by ESL/EFL learners of diverse L1 backgrounds. While different
groups of learners tend to make different errors, the error examples found in various studies,
especially the recurring ones in different learner corpora, offer valuable insights into some
potential problems that may also be prominent in learners’ writing. In addition to uncovering
and analyzing errors, it is also important to understand what causes them, which is dis-
cussed next.

Possible reasons for learners’ errors

Summarizing the relevant literature, we found three common possible factors causing misuse
or overuse of connectives in ESL learners’ writing. First, pedagogical factors such as design
of teaching materials and teaching methods may mislead learners into using connectives
erroneously when they write. Some textbooks or teachers provide lists of connectives so as
to encourage students to apply them in writing (Crewe, 1990; Field & Yip, 1992; Yeung, 2009).
However, some learners tend to simply transfer the connectives on the list to their writing
without a thorough understanding of their meanings and usage (Field & Yip, 1992).
Sometimes, non-equivalents are listed together as alternatives (e.g. ‘In contrast’ and ‘On the
contrary’) and this also tends to cause errors (Crewe, 1990; Leedham & Cai, 2013). Apart from
connective lists, some textbooks provide mechanical exercises related to connectives and
this may encourage overuse (Crewe, 1990; Kwan, 2017). Textbook exercises drill students to
connect isolated sentences with connectives but may fail to promote their awareness of the
appropriate ones for various contexts (Kwan, 2017; Yeung, 2009).
Learners’ perceptions of how connectives should be used may also lead to errors. Some
learners think that using more connectives can make their writing better. For example, some
do not have a clear idea of how to develop their arguments, so they rely heavily on connec-
tives to join sentences together in order to give a seemingly complete structure (Crewe,
1990; Tang & Ng, 1995). Overusing connectives is a way for some learners to disguise or
mask poor writing performance (Crewe, 1990). Since the ideas are not connected appropri-
ately or elaborated thoroughly, such a use of connectives only gives surface logicality and
does not facilitate readers’ understanding of the logical development (Chen, 2006; Crewe,
1990). In other words, some learners overuse connectives to create superficial connections
between sentences, but fail to achieve coherence in their writing by doing so (Lei, 2012).
Finally, learners’ L1s may also lead to their problems with connectives. Some of the mis-
takes can be explained by L1 transfer. For example, it has been found that Chinese ESL
learners tend to transfer Chinese connectives wrongly, such as interpreting the meaning of
‘From another aspect’ as equivalent to ‘On the other hand’ or ‘Besides’ in English (Field & Yip,
1992; Yeung, 2009). Moreover, the Chinese language allows double conjunctions in complex
 Language Awareness 159

sentences, such as the ‘because…, so’ and ‘although… but’ constructions, and some learners
transfer these into English sentences (Chan, 2010). These examples show that language
transfer may explain some cases of misuse and overuse of connectives.

The study
As shown above, most studies have examined ESL/EFL learners’ writing in university settings.
This study attempted to address the gap of the lack of studies for secondary-level ESL learn-
ers. Various studies have also suggested that it would be useful to conduct qualitative studies
(Gao, 2016; Leedham & Cai, 2013), or in-depth case studies ‘to find out what actually occurs
in the minds of the writers’ (Lee, 2002, p.154). However, previous studies mainly reported
statistical results from corpus analyses. The present study is unique in that it went beyond
textual analyses by incorporating ethnographic elements. All the participants participated
in group interviews, the findings of which cast light on various related topics such as their
views about the use of connectives and experiences of learning connectives. Also, one of
the authors was the participants’ English teacher at the time of the study, so he was able to
observe them regularly in a naturalistic context. It was hoped that important implications
could be drawn from this study to help different learners to enhance their cohesion and
coherence in ESL writing.
A multi-methods qualitative approach was adopted for this study to investigate ESL learn-
ers’ use of connectives. It was an embedded single-case design in which there were three
embedded units of analysis (Yin, 2018), namely high-, average-, and low-proficiency groups
(n = 3 in each group). The data were collected via observations, students’ writing, and
semi-structured group interviews. This article reports relevant findings from this case study.
It was guided by the following research questions, classified into two groups:
Group I: Use of connectives
RQ1: What are the features of connectives (in terms of quantity and types) in ESL writing
by secondary-level students?
RQ2: What are the problems regarding the use of connectives in writing by these learners?
Group II: Perceptions of connectives and experiences in learning them
RQ3: What are high-, average-, and low-proficiency learners’ perceptions of the use of
connectives in writing?
RQ4: What are the different approaches these learners adopt in learning connectives?

Methods
The participants
Nine native-Cantonese-speaking ESL learners (Mage= 13.33; SD = 0.47) from a secondary
school in Hong Kong participated in the study. Purposive sampling was used to recruit
participants who represented learners of different proficiency levels (high, average, and low)
in ESL writing in a mixed-ability class of 19 students. The proficiency levels were determined
by their previous performances in writing and consultation with an English teacher who had
taught the class. Their proficiency in ESL writing was expressed in band scores in the
International English Language Testing System (IELTS), which ranges from band one (lowest)
to band nine (highest). There were three groups, each with three participants (see Table 2).
160 A. CHENG AND A. TSANG

Data collection
Participants’ essays
The participants were asked to write an expository essay as this genre involves relatively
more logical connections (Field & Yip, 1992; Rose, 2013). They were given the topic ‘Facebook’
and were required to write around 200 words (Appendix 1). Three 55-minute lessons were
dedicated to this writing module. The first lesson focused on the pre-writing stage in which
the teacher/researcher discussed a sample expository essay and analysed the features of
this genre with the participants. The focus was on the general structure (i.e. introduction,
main body, and conclusion) and content, with no explicit discussion on cohesion/cohesive
devices. The students were taught to give background information about a topic in the
introduction and to elaborate on the topic in the body paragraphs using examples, facts,
or quotes. They were also guided to elaborate on their ideas by including different perspec-
tives, such as the pros and cons of an issue. The second and the third lessons were the
while-writing stage, during which the students completed mind maps and the writing task
individually. Their compositions were collected for analyses.

Group interviews
The interviews aimed at examining the participants’ learning experiences and perceptions
of connective use. This instrument is more interactive and could elicit richer data as it allows
both valid and flexible responses (Burns, 2009; Hyland, 2009). In the consent form and before
each interview session, the participants were assured that the interview sessions were not
any part of their schoolwork and their responses would be confidential. This, together with
the rapport established, helped encourage the participants to express their real thoughts
and minimise response bias (McKay, 2006). To avoid language barriers and to maximise their
ease of expression, the participants were allowed to use either Cantonese (their L1) or English.
Three group interviews were conducted, each with the three participants in each profi-
ciency group. Each interview involved two sections: (1) text evaluation and (2) guided dis-
cussion. Inspired by Lee’s (1998) suggestion for using coherence awareness-raising tasks, in
section (1), the interviewees were given three versions of the same text (Appendix 2). The
text was adapted from the resource banks provided by their school. It was chosen for its
suitability in terms of proficiency level and topic. Each version differed only in the amount
of different connectives at the beginning of a sentence. The rationale behind this was to
make the task less cognitively challenging for the young participants. Version A was a para-
graph with minimal use of connectives (only one, despite), version B showed an excessive
use of multi-word or long connectives (first and foremost; as a matter of fact; in spite of; more-
over; to this end) and version C used a few simple connectives (firstly; despite; in short). The
interviewees were asked to read the three versions and choose the best version in terms of
its writing quality. The interviewer then asked follow-up questions to elicit the learners’
perceptions of the use of connectives. In section (2), the interviewees were engaged in a

Table 2. The three groups of participants.

High (n = 3) Average (n = 3) Low (n = 3)
Code* #1, #2, #3 #4, #5, #6 #7, #8, #9
Proficiency in writing 7 to 7.5 6 to 6.5 5 to 5.5
(IELTS Banding)
*These codes will be referred to when reporting interview findings.
 Language Awareness 161

discussion with each other and the interviewer, based on the interview guide that had been
created (Appendix 3). The topics covered included the participants’ perceptions of good
writing and connectives, and their experiences in learning/using connectives.

Observations
Classroom observations were conducted to complement the interview findings. As the first
author was the participants’ English teacher, he had ample opportunities to observe the
nine participants’ habits and behaviour in all three writing lessons and outside class. Being
their teacher, he had already developed a good rapport with them, which enabled him to
capture their habits and behaviour in a naturalistic setting from an emic perspective (Harbon
& Shen, 2015; Mackey & Gass, 2015). This, together with the prolonged observation period,
facilitated data collection from more authentic participant behaviour. The observations
focused on recording the participants’ behaviour holistically (Fraenkel et al., 2015). Reference
was made to previous studies. For instance, regarding paying attention, the participants:

“…are not displaying any inattentive or disruptive behavior; they are looking at the teacher and
following his or her movements, looking at visual stimuli, turning to watch another student
who is contributing to the task, following the text being read, or making read, or making appro-
priate nonverbal response” (Guilloteaux & Dörnyei, 2008, p. 62)

The researcher kept a teacher diary in which field notes were written immediately after
each of the three writing lessons. To allow the participants to behave naturally, covert obser-
vations were used (Cohen et al., 2018). The participants were unaware that they were
observed and no special seating arrangements were made for the participants (i.e. they were
sitting at different places in the classroom; some with non-participants). The observation
results were discussed with another English teacher of the class to ensure validity.

Data analysis
The participants’ essays were analyzed both quantitatively and qualitatively. They were typed
verbatim and loaded into O’Donnell’s (2013) UAM CorpusTool to form a learners’ corpus. The
manual tagging function assisted the coding process, which helped quantify the data more
efficiently and accurately (Baker, 2010). The connectives used in each composition were
identified and categorised based on the Halliday and Hasan cohesion framework (see Table
1) by the first author. A second coder, specialising in English Education and having experience
in genre analysis, was recruited to conduct the same process to determine the inter-coder
reliability. The first author briefed the coder about the coding scheme in a training session.
They analyzed some samples together to ensure that the coder gained sufficient familiarity
with the scheme before conducting the actual coding.
The inter-coder agreement for all of the analyses by the two coders was 89%. All differ-
ences were resolved after a discussion between the coders. The connectives used in the
writing scripts were also calculated in terms of numbers of tokens per 100 words, since the
average composition length varied with the students’ proficiency levels, ranging from 188
words to 498 words.
The use of connectives was also evaluated qualitatively to identify the learners’ problems,
including misuses and overuses. This study adopted an interpretive approach similar to that
162 A. CHENG AND A. TSANG

used by Crewe (1990). Misuse refers to connective use that ‘causes comprehension problems’,
while overuse refers to situations when connectives ‘clutter up the text unnecessarily, and
at worst cause the thread of the argument to zigzag about’ (Crewe, 1990, p. 324). The dis-
tinction between the two is that misuse signals an inappropriate relationship semantically,
whereas overuse refers to redundant connectives that make the text tortuous, such as placing
two or more of the same categories in adjacent positions or using them unnaturally at the
beginning of almost every sentence (Crewe, 1990).
The audio recordings of the group interviews were transcribed verbatim. The transcripts
were then examined via content analysis techniques (Burns, 2009). The first author read
through the transcripts and identified preliminary codes using an inductive approach which
allowed important themes to emerge from the data (Burns, 2009). Then, he explained the
codes to the second coder and they discussed them, making reference to some sample
scripts. After the list of codes was finalised, the interview transcripts were then coded inde-
pendently. The analysis showed an inter-coder agreement of 81%. The coders immersed
themselves by reading and re-reading the transcripts and discussed the codes to reach
mutual agreement (Taber, 2007).

Findings
Observations
It was evident from the observations that the high and average achievers were relatively
more attentive than the low achievers in English lessons. In the writing lessons, the high and
average achievers responded to the teacher’s questions and paid attention to the class
discussion of sample compositions. The average students demonstrated the highest observ-
able engagement as they were the only ones who copied notes and highlighted phrases
(e.g. useful expressions) in the sample texts. The low-achievers were inattentive as they
chatted with their neighbours during the lessons. It was observed that they did not attempt
the exercise about the sample text and simply copied the answers when the teacher
explained them. During the while-writing stage, the low achievers also chatted with their
neighbours at times and hence needed more time to complete their task.
Another issue of interest was the participants’ exposure to English input in addition to
English lessons. The high-achievers were found to read books in English during their leisure
time. This was probably a form of entertainment rather than part of any assignment, and
they had chosen different books to read based on their interests. This was not observed in
the other groups. It was also observed that the high-achievers communicated with other
students in English at times, whereas the average- and low-achievers always communicated
with their peers in Chinese. Table 3 shows the observation findings succinctly.

Participants’ essays
Distribution of connectives used
The overall use of connectives is summarised in Table 4. The average number used was 6.59
per 100 words. All three groups seemed to have very similar numbers of connectives per
100 words (ranging from 5.76 to 7.37).
 Language Awareness 163

All four types of connectives were found in the participants’ texts. Table 5 presents the
frequencies and percentages of the connectives used in the learners’ essays by category.
Additive devices accounted for the greatest proportion of use across all proficiency levels,
with over 45% in the high- and average-proficiency groups and almost 40% in the low-pro-
ficiency group. It is noteworthy that the high- and average-proficiency participants used 3
to 4 temporal devices, while the low-proficiency participants used 9, around 3 times more
than the other groups. The low-proficiency group used fewer causal devices, with only 11
instances.

Overuse of connectives
From the writing samples collected, the high-proficiency participants mainly had the prob-
lems of overusing connectives, whereas cases of overuse and misuse were identified in both
the average and low-proficiency groups.
Regarding overuse, some participants used two or more connectives to express one single
relationship. As shown in Example 1 in Figure 1, a student from the average-proficiency
group started the final paragraph with one causal and two temporal devices to show a
conclusive relationship. In this case, using one temporal connective, i.e. in conclusion, would
have sufficed.
Also, the use of double conjunctions was identified in the writing of the low-proficiency
students. Example 2 shows a typical error resulting from the ‘although…but…’ double con-
junction structure common in Chinese, extracted from Participant 9’s writing. In this case,
he should have used either ‘even though’ or ‘but’.

Table 3. Relevant findings from the observations.

High (n = 3) Average (n = 3) Low (n = 3)
Observable - Attentive - Attentive - Inattentive, chatty
behaviour in - Actively responded to - Actively responded to teacher’s - Only copied answers of the
English writing teacher’s questions questions exercise when discussing
lessons when discussing the - Made notes/ highlighted the sample texts
sample texts phrases in the sample texts
Observable - Read English books during - No such reading habits - No such reading habits
English input leisure time (e.g. recess) - Always conversing in Chinese - Always conversing in
beyond the - Conversing in English at Chinese
English lessons times

Table 4. Overall frequencies of connectives used.

High Average Low Total
Total number of words 1198 963 660 2821
Total number of connectives 69 71 46 186
Number of connectives per 100 words 5.76 7.37 6.97 6.59

Table 5. Frequency of connectives used by categories.

Additive Adversative Causal Temporal
Frequency % Frequency % Frequency % Frequency %
High 32 46.4 12 17.4 22 31.9 3 4.3
Average 31 43.7 9 12.7 27 38.0 4 5.6
Low 18 39.1 8 17.4 11 23.9 9 19.6
Total 81 43.5 29 15.6 60 32.3 16 8.6
164 A. CHENG AND A. TSANG

Both the high and average groups also tended to overuse the additive ‘and’ in writing.
Example 3 illustrates this overuse in a short paragraph written by a high-proficiency student.
It can be seen that the student used the additive ‘and’ in almost every sentence to connect
his ideas.

Figure 1. Eight excerpts from the participants’ essays.

 Language Awareness 165

Misuse of connectives
Misuse of connectives was found to be the most serious in the low-proficiency participants’
texts. There were some cases of confusion between similar connectives. In Example 4,
Participant 7 started the final paragraph of his essay with ‘at last’ instead of ‘lastly’ to give a
conclusion about his ideas; while in Example 5, Participant 8 misused ‘so that’ to express the
result or effect. ‘In Example 6, Participant 4 confused the connective ‘as if’ with ‘if’.
Students with low proficiency also faced the problem of using non-existing expressions
and connectives that do not establish effective logical relationships. In Example 7, the partic-
ipant probably intended to use ‘For these reasons’ while in Example 8, the student wrongly
used temporal devices to express contrastive ideas between paragraphs.

Interviews
Three major themes emerged from the interviews. These were perceptions of connective
use, perceptions of good writing and the students’ learning experiences.

Perceptions of connective use

The learners, in general, understood that the use of connectives has a positive effect on
organisation in writing. For example, the interviewees from the high-proficiency group
expressed that connectives ‘help you organise the passage’ (Participant 1) and ‘smoothen
the passage out’ (Participant 3). Participant 2 further elaborated that the use of connectives
helps make the ideas clear to readers:
“… you are discussing this point, and you suddenly switch to another point. To me, this isn’t
actually that good because you need to show readers whether you are talking about this point
or that point”.

Similarly, interviewees from the average group also expressed that connectives ‘make the
points clearer’ (Participant 5) and ‘it will be confusing if connectives are not used as the
readers won’t know which point you are talking about’ (Participant 6). These responses sug-
gest that both the high and average-proficiency groups were aware that connectives have
positive effects on organisation and clarity in writing. The low-proficiency participants, how-
ever, did not seem to be aware of the functions of connectives. They provided vague
responses such as ‘connectives make the writing look better and stronger’ (Participant 7)
and ‘the use of connectives in writing is important’ (Participant 9).
It is also interesting to note that the average and low-proficiency groups believed that
the use of connectives would improve their writing scores. For instance, Participant 4 men-
tioned that ‘if you don’t use connectives, you will lose marks as you can’t link your ideas’.
Such perceptions correspond with the text evaluation (i.e. Appendix 2) results elicited in
the interviews. As shown in Table 6, as many as two-thirds of the students generally perceived
that the use of multi-word or long connectives signified better writing quality, even though
some devices in the text were used erroneously as they did not establish any logical rela-
tionships. The participants in all groups believed that the use of first and foremost (as in
version B), for instance, is less common and ‘may impress the teacher to give higher marks’
(Participants 2, 3, 6, 7). The same applied to as a matter of fact (in version B) (Participants 6, 7).
166 A. CHENG AND A. TSANG

Perceptions of good writing

The participants seemed to suggest that they prioritise content and language over organi-
sation when they write. When asked about their opinions of good writing, their responses
included ‘no run-on sentences, good vocabulary, right grammar’ (Participant 1), ‘use some
difficult vocabulary items’ (Participant 6), ‘correct tenses’ (Participant 8), ‘don’t go off topic’
(Participant 9), and ‘your writing should be entertaining to attract the readers’ (Participant 3).
These show that the participants were more concerned about the content and language
domains when writing and less concerned about the effects of organisation of the ideas in
their essays.
Regarding organisation, the participants tended to focus on the macrostructure instead of
cohesion. For instance, Participant 1 expressed the belief that ‘the most important part is to
have a good introduction. Introduction shows what the organisation of a piece of writing is
like’. Participants 7 and 9 also shared similar opinions as they suggested that they would pay
attention to the order of ideas and paragraphing. Overall, it seemed that the participants did
not have a high awareness about issues related to cohesion at the micro-level (i.e. phrasal or
sentential rather than paragraphing) that play an important role in connections between ideas.

Learning experience
The participants had diverse experiences of learning connectives. All three high-proficiency
participants suggested that they had acquired them incidentally through their reading. In
the interviews, the high-proficiency group stated that they had greater exposure to English
input as they read novels and news articles during their free time and occasionally commu-
nicate in English among themselves. These responses support the observation findings
summarised in Table 3.
Both the average and low-proficiency learners shared that they learnt connectives pri-
marily through the teaching materials provided by their teachers. For example, Participant
4 stated that he learnt connectives from ‘sample essays.’ Participant 5 concurred and added
that he was given lists of connectives from which he ‘copied the important ones into [his]
notebook and used some of them in [his] writing’. These responses also echo the observation
finding (e.g. the average learners highlighting and taking notes). The low-proficiency par-
ticipants recalled that they had experienced transferring some connectives from the lists
provided by their teacher directly into their own writing.

Discussion
The study set out to explore learners’ use of, perceptions of, and experiences in learning to
use connectives. Guided by the four research questions, we conducted observations, inter-
views, and content analysis of the participants’ essays. The research questions will be

Table 6. Participants’ choices of the best version in text evaluation.

Text A Text B Text C
Minimal use of Excessive use of multi-word Use of a few simple
connectives or long connectives connectives
Number of students (%) 2 (22.2%) 6 (66.7%) 1 (11.1%)
Participants #4, #8 #1, #2, #3, #5, #6, #7 #9
 Language Awareness 167

answered here, followed by a succinct discussion of implications, limitations, and directions

for further research.

[RQ1] What are the features of connectives (in terms of quantity and types) in ESL
writing by secondary-level students?
As the data revealed, the participants used all four types of connectives in their writing.
Similar to the findings of previous studies (Chen, 2006; Ha, 2016), additive connectives were
found to be the most frequent in writing across all proficiency levels. The study also provides
corroborating evidence that ESL learners tend to rely more on simple connectives like ‘and’
and ‘but’ than multiword ones (Chen, 2006; Liu & Braine, 2005). Also, since the average num-
bers of connectives used by different proficiency groups were similar, this may add support
to the idea that it is not the quantity but the quality of how cohesive devices are used that
affects the writing quality (Khalil, 1989). This idea can be further examined or substantiated
by future studies.

[RQ2] What are the problems regarding the use of connectives in writing by these
learners?
Although researchers such as Bolton et al. (2002) and Chen (2006) have mentioned ESL
learners’ common problems of overusing and misusing connectives, this study furthered
our understanding by analyzing these problems, taking into account the participants’ pro-
ficiency levels. This study shows that high-proficiency learners mainly have the problem of
overuse of additives but no misuse of connectives. This may imply that these learners prob-
ably have a better understanding of the logical relationships established by various connec-
tives than the two other groups. The average and low-proficiency participants had problems
with both overuse and misuse of connectives. Misuse of connectives was found to be espe-
cially prevalent in the low-proficiency participants’ essays. Inadequate knowledge of the
connectives is probably a reason for the misuses. As illustrated in Example 4, Participant 7
intended to express a conclusive relationship but confused ‘at last’ with ‘lastly’. The student
probably did not have the knowledge that ‘at last’ carries an implied meaning of impatience
after long delays (Carter et al., 2016). Likewise, both ‘so’ and ‘so that’ are causal connectives,
but the latter expresses a purpose rather than a result (Swan, 2005). The learners in this study
assumed these seemingly alike connectives convey the same meaning and failed to use
them accurately. Similar to the case of ‘on the contrary’ discussed by Field and Yip (1992),
the likely reason behind these errors was their lack of awareness of the implied meaning
and subtle differences between similar connectives.

[RQ3] What are high-, average-, and low-proficiency learners’ perceptions of the
use of connectives in writing?
In general, the findings from the interviews and text evaluations indicated that most students
believed using more connectives, especially the complex ones, could improve the organisa-
tion and/or overall quality of their writing. Such views, particularly from those who believed
168 A. CHENG AND A. TSANG

more connectives would make their writing better, somewhat echo previous studies that
some learners use connectives to disguise poor writing (Crewe, 1990). As shown in Examples
1 and 8, the participants attempted to use the connectives to connect their ideas, but their
overuse and misuse only created surface logicality as effective cohesion could not be estab-
lished (Chen, 2006; Crewe, 1990). An interesting finding is that the participants’ perceptions
of complex connectives did not necessarily correspond to their writing performances (i.e.
participants from different proficiency groups had similar perceptions regarding the use of
multi-word or long connectives). The students’ voices also provided insights into the learners’
use of connectives. When viewing overall writing quality, they seemed to prioritise content
and language domains over organisation. When directed to talk about organisation, they all
focussed on aspects of macro-level coherence, such as macrostructure and propositional
development. None of them talked about micro-level cohesion or the use of connectives. In
short, they seemed to have a low awareness of the contribution of cohesion or the use of
connectives to overall writing quality.

[RQ4] What are the different approaches these learners adopt in learning
connectives?
Based on the participants’ sharing and their observable behaviours, it seems that learners
of different proficiency levels relied on different learning strategies. The participants from
the average and low-proficiency groups stated that they copied connectives from lists or
sample writing provided by their teachers while the high-proficiency learners believed they
acquired them through leisure reading rather than from lists or sample writing. As discussed
in the literature, connective lists and textbook exercises may have been the reasons for some
of the errors (Kwan, 2017; Tseng & Liou, 2006). The learning experiences of these participants
may indicate that high-proficiency learners tend to make better use of implicit learning
strategies than other students in mastering connectives. The reliance on connective lists
may be less effective than incidental learning in fostering students’ awareness of the appro-
priate use of conjunctive devices in writing.

Implications for ESL teachers

The use of connectives is related closely to cohesion and writing quality (Lee, 1998). The
findings suggest that ESL learners may neglect or not be aware of the importance of cohe-
sion. Teachers are advised to adopt writing pedagogies that can raise learners’ awareness of
cohesion and contexts, for example by using the genre-based approach as it provides a
systematic way for learners to integrate both language and context in the writing classrooms
(Hyland, 2007; see also Henry & Roseberry, 1999) or by the languaging approach, which can
facilitate generation of both ideas and expressions for writing (see Yang, 2016). As discussed,
neglecting the context may result in inappropriate use of connectives like ‘besides’ (Kwan,
2017; Yeung, 2009). It is noteworthy from Table 5 that the low-proficiency learners used a
lot more temporal connectives than the other groups. This may suggest that they had not
mastered the genre or text type of the writing task thoroughly. The text type of the task
should have been an exposition rather than a narrative involving more temporal relationships
(Turbill & Bean, 2006). Similarly, the overuse of additives in Example 3 (see Figure 1) is indeed
 Language Awareness 169

a typical feature of lengthy coordination in the spoken genre (Crystal, 2006). Although spoken
and written genres fall in continua rather than dichotomies, it may not be appropriate for
learners to overuse speech-like features in written discourse. The participants’ scripts exem-
plify the common challenge of moving from a more colloquial to a written model of language
use for junior secondary learners (Christie & Derewianka, 2008). Some learners seemed to
have low genre awareness in writing, which led to problems with the use of connectives.
Ideally, teachers should discuss and highlight the use of connectives in authentic writing
scripts to increase learners’ awareness of how cohesion is achieved at the discourse level.
Collaborative writing, which can result in peer-scaffolded awareness of what good writing
entails (Chen & Yu, 2019), is another promising method (and a non-teacher-centred one too).
Simply giving connective lists to students should be avoided, as studies have shown that
some errors are caused by inadequate knowledge of the subtle differences in meanings of
similar expressions. To make good use of connective lists, especially for average and low-pro-
ficiency learners who rely more on teaching materials, teachers should explain and provide
examples for differentiating similar connectives. Teachers should also make it clear to all
learners (including the high achievers; #1-#3 all chose text B with long and multi-word con-
junctions as the best text) that the length of a connective and the quantity of connective
use are in no way connected with the writing quality.

Conclusion
This case study revealed that the participants had knowledge of various types of connectives.
Regarding actual usage, the high-proficiency learners were found to have the problem of
overuse, whereas the average- and low-proficiency learners had both problems of overuse
and misuse. Few investigations of connectives have examined learners’ different proficiency
levels, their views and how they master connectives as this study has done. The observations
and interview findings suggest that, although different proficiency groups may learn con-
nectives differently, their perceptions of them do not seem to be related to their proficiency
levels with their proficiency. Understanding learners’ voices is helpful when developing
learner-centered activities and curricula. However, there were some limitations to the study.
It was a case study with a small number of participants. While an analytic generalisation was
appropriate, there is a need to recruit more participants in future studies in order to collect
data suitable for making statistical generalisations (Yin, 2018). The use of only one written
text – expository essays – was rather limited. A wider range of written genres could have
been collected from the ESL learners for a more in-depth analysis of the use of connectives.
Also, this study only shed light on a few problems related to the connectives the ESL learners
had encountered; it did not attempt to explore the effectiveness of the suggested approaches
in combating these problems. Further research using the intervention (e.g. genre-based and
awareness-raising tasks) approach is needed to study the effects of these interventions on
learners’ awareness and use of connectives. Finally, the relationship between learners’ per-
ceptions of complex connectives and writing performance also deserves further inquiry.

Disclosure statement
No potential conflict of interest was reported by the author(s).
170 A. CHENG AND A. TSANG

Notes on contributors
Angus Cheng is a graduate from the Chinese University of Hong Kong majoring in English and
English as a second language education. His research interests are ESL writing and language
awareness.
Art Tsang is an assistant professor in the Faculty of Education at the Chinese University of Hong
Kong. He takes an avid interest in research into ESL/EFL education. His articles have been pub-
lished in international peer-reviewed journals such as TESOL Quarterly, Language Teaching
Research, and Journal of Multilingual and Multicultural Development.

ORCID
Angus Cheng https://doi.org/0000-0001-5547-317X
Art Tsang http://orcid.org/0000-0003-2507-157X

References
Appel, R., & Szeib, A. (2018). Linking adverbials in L2 English academic writing: L1-related differences.
System, 78, 115–129. https://doi.org/10.1016/j.system.2018.08.008
Baker, P. (2010). Corpus methods in linguistics. In L. Litosseliti (Ed.), Research methods in linguistics
(Research methods in linguistics) (pp. 93–113). Continuum.
Bolton, K., Nelson, G., & Hung, J. (2002). A corpus-based study of connectors in student writing:
Research from the International Corpus of English in Hong Kong (ICE-HK). International Journal of
Corpus Linguistics, 7(2), 165–182. https://doi.org/10.1075/ijcl.7.2.02bol
Burns, A. (2009). Doing action research in English language teaching: A guide for practitioners.
Routledge.
Carter, R., McCarthy, M., Mark, G., & O’Keeffe, A. (2016). English grammar today: An A-Z of spoken and
written grammar. Cambridge University Press.
Chan, A. (2010). Toward a taxonomy of written errors: investigation into the written errors of Hong
Kong Cantonese ESL learners. TESOL Quarterly, 44(2), 295–319. https://doi.org/10.5054/
tq.2010.219941
Chen, C. W. Y. (2006). The use of conjunctive adverbials in the academic papers of advanced Taiwanese
EFL learners. International Journal of Corpus Linguistics, 11(1), 113–130. https://doi.org/10.1075/ij-
cl.11.1.05che
Chen, W., & Yu, S. (2019). Implementing collaborative writing in teacher-centered classroom contexts:
student beliefs and perceptions. Language Awareness, 28(4), 247–267. https://doi.org/10.1080/096
58416.2019.1675680
Christie, F., & Derewianka, B. (2008). School discourse: Learning to write across the years of schooling.
Continuum.
Cohen, L., Manion, L., & Morrison, K. (2018). Research methods in education (5th ed.). Routledge.
Connor, U. (1984). A study of cohesion and coherence in English as a second language students’ writ-
ing. Research on Language & Social Interaction, 17(3), 301–316.
Crewe, W. (1990). The illogic of logical connectives. ELT Journal, 44(4), 316–325. https://doi.org/10.1093/
elt/44.4.316
Crossley, S. A., Kyle, K., & McNamara, D. S. (2016). The development and use of cohesive devices in L2
writing and their relations to judgments of essay quality. Journal of Second Language Writing, 32(C),
1–16. https://doi.org/10.1016/j.jslw.2016.01.003
Crystal, D. (2006). Language and the internet (2nd ed.). Cambridge University Press.
Field, Y., & Yip, L. (1992). A comparison of internal conjunctive cohesion in the English essay writing of
Cantonese speakers and native speakers of English. RELC Journal, 23(1), 15–28.
Fraenkel, J., Wallen, N., & Hyun, H. (2015). How to design and evaluate research in education (9th ed.).
McGraw-Hill Education.
 Language Awareness 171

Gao, X. (2016). A cross-disciplinary corpus-based study on English and Chinese native speakers’ use of
linking adverbials in academic writing. Journal of English for Academic Purposes, 24, 14–28. https://
doi.org/10.1016/j.jeap.2016.08.002
Guilloteaux, M., & Dörnyei, Z. (2008). Motivating Language Learners: A Classroom-Oriented
Investigation of the Effects of Motivational Strategies on Student Motivation. TESOL Quarterly,
42(1), 55–77. https://doi.org/10.1002/j.1545-7249.2008.tb00207.x
Ha, M. J. (2016). Linking adverbials in first-year Korean university EFL learners’ writing: A corpus-in-
formed analysis. Computer Assisted Language Learning, 29(6), 1090–1101. https://doi.org/10.1080/
09588221.2015.1068814
Halliday, M., & Hasan, R. (1976). Cohesion in English (A Longman paperback). Longman.
Harbon, L., & Shen, H. (2015). Researching language classrooms. In B. Paltridge & A. Phakiti (Eds.),
Research methods in applied linguistics. A practical resource (pp. 457–470). Bloomsbury.
Henry, A., & Roseberry, R. L. (1999). Raising awareness of the generic structure and linguistic features of
essayintroductions.LanguageAwareness,8(3-4),190–200.https://doi.org/10.1080/09658419908667128
Hyland, K, (2007). Genre pedagogy: Language, literacy and L2 writing instruction. Journal of Second
Language Writing, 16(3), 148–164.
Hyland, K. (2009). Teaching and researching writing. (2nd ed.). Routledge.
Khalil, A. (1989). A study of cohesion and coherence in Arab EFL college students’ writing. System,
17(3), 359–371. https://doi.org/10.1016/0346-251X(89)90008-0
Kılıç, M., Genç, B., & Bada, E. (2016). Topical structure in argumentative essays of EFL learners and im-
plications for writing classes. Journal of Language and Linguistic Studies, 12(2), 107–116.
Kwan, Y. H. (2017). The teaching of additive connectives: An evaluation of current English grammar
textbooks in Hong Kong. The Asian Journal of Applied Linguistics, 4(2), 161–172.
Lee, I. (1998). Enhancing ESL students’ awareness of coherence-creating mechanisms in writing. TESL
Canada Journal, 15(2), 36–49. https://doi.org/10.18806/tesl.v15i2.700
Lee, I. (2002). Teaching coherence to ESL students: A classroom inquiry. Journal of Second Language
Writing, 11(2), 135–159. https://doi.org/10.1016/S1060-3743(02)00065-6
Leedham, M., & Cai, G. (2013). Besides… on the other hand: Using a corpus approach to explore the
influence of teaching materials on Chinese students’ use of linking adverbials. Journal of Second
Language Writing, 22(4), 374–389. https://doi.org/10.1016/j.jslw.2013.07.002
Lei, L. (2012). Linking adverbials in academic writing on applied linguistics by Chinese doctoral stu-
dents. Journal of English for Academic Purposes, 11(3), 267–275. https://doi.org/10.1016/j.
jeap.2012.05.003
Liu, M., & Braine, G. (2005). Cohesive features in argumentative writing produced by Chinese under-
graduates. System, 33(4), 623–636. https://doi.org/10.1016/j.system.2005.02.002
Ma, Y., & Wang, B. (2016). A corpus-based study of connectors in student writing: a comparison be-
tween a native speaker (NS) corpus and a non-native speaker (NNS) learner corpus. International
Journal of Applied Linguistics and English Literature, 5(1), 113–118.
Mackey, A., & Gass, S. M. (2015). Second language research: Methodology and design (2nd ed.).
Routledge.
McKay, S. (2006). Researching second language classrooms (ESL and applied linguistics professional se-
ries). Lawrence Erlbaum.
Nunan, D. (1999). Second language teaching & learning. Heinle & Heinle.
O’Donnell, M. (2013). UAM corpus tool. http://www.corpustool.com/
Paran, A. (2008). The role of literature in instructed foreign language learning and teaching:
An evidence-based survey. Language Teaching, 41(4), 465–496. https://doi.org/10.1017/
S026144480800520X
Rose, D. (2013). Genre in the Sydney school. In The Routledge handbook of discourse analysis (pp. 235–
251). Routledge.
Schneider, M., & Connor, U. (1990). Analyzing topical structure in ESL essays: Not all topics are equal.
Studies in Second Language Acquisition, 12(4), 411–427. https://doi.org/10.1017/S0272263100009505
Swan, M. (2005). Practical English usage. Oxford University Press.
Taber, K. (2007). Classroom-based research and evidence-based practice: A guide for teachers. Sage
Publications.
172 A. CHENG AND A. TSANG

Tang, E., & Ng, C. (1995). A study on the use of connectives in ESL students’ writing. Perspectives, 7(2),
105–121.
Thornbury, S. (2005). Beyond the sentence: Introducing discourse analysis. Macmillan Education.
Tseng, Y. C., & Liou, H. C. (2006). The effects of online conjunction materials on college EFL students’
writing. System, 34(2), 270–283. https://doi.org/10.1016/j.system.2006.01.006
Turbill, J., & Bean, W. (2006). Writing Instruction K-6: Understanding process, purpose, audience. Richard
C Owen Publishers, Inc.
Yang, L. (2016). Languaging in story rewriting tasks by Chinese EFL students. Language Awareness,
25(3), 241–255. https://doi.org/10.1080/09658416.2016.1197230
Yang, W., & Sun, Y. (2012). The use of cohesive devices in argumentative writing by Chinese EFL learn-
ers at different proficiency levels. Linguistics and Education, 23(1), 31–48. https://doi.org/10.1016/j.
linged.2011.09.004
Yasuda, S. (2019). Children’s meaning-making choices in EFL writing: The use of cohesive devices and
interpersonal resources. System, 85, 102108. https://doi.org/10.1016/j.system.2019.102108
Yeung, L. (2009). Use and misuse of ‘besides’: A corpus study comparing native speakers’ and learners’
English. System, 37(2), 330–342. https://doi.org/10.1016/j.system.2008.11.007
Yin, R. K. (2018). Case study research and applications: Design and methods (6th ed.). SAGE Publications.
Yin, Z. (2015). The use of cohesive devices in news language: overuse, underuse or misuse? RELC
Journal, 46(3), 309–332. https://doi.org/10.1177/0033688215597578
 Language Awareness 173

Appendix 1. Writing prompt

Appendix 2. Text evaluation in the interviews

174 A. CHENG AND A. TSANG

Appendix 3. Interview guide