The Use of Cohesive Devices in Research Paper Conference to Achieve Texts Coherence

____________________ Abstract ___________________________________________________________________ This paper employs discourse analysis to investigate cohesive devices and coherence in research articles. The research was based on a purposive random sampling of 10 research articles from the 8 th ELTLT Conference 2019. The text analyzed in this paper includes the abstract, introduction, research methodology, findings and discussions, and conclusion. We seek to find out grammatical and lexical cohesion as conceptualized by Halliday and Hasan's (1976) concept, i.e., reference, substitution, ellipsis, conjunction, reiteration, and collocation. Furthermore, we discover the relation among research articles to achieve coherence by employing logical relations that highlight Thornbury's (2005) theory. According to the findings, the authors used cohesive devices in the creation of their writing product. From a total of 1955 clauses in 224 paragraphs, it was determined that almost all types of cohesive devices are used throughout the text. The conjunction of addition was the most frequently used cohesive device in the texts. Our findings demonstrate that authors engage the additional information to create a text without changing the


INTRODUCTION
People produce or make a text in expressing their thoughts in case to get a message across or to explain something to others either in spoken or written form. Written discourse has a texture to form a unified and sensible text. It produces people's feelings and thoughts that provide the context for creation and interpretation in writing. Those ways of text make a language to communicate with each other. It can be said that discourse mostly refers to people's language in use. Furthermore, it deals with Celce- Murcia's (2007) theory that proposed six communicative competencies. One of them is discourse competence as the core of communicative competence. It includes not only knowledge and the ability to linguistic resources, but also to create cohesion and coherence in spoken and written forms.
Furthermore, based on Yuniawan et al. (2017), using discourse as a social practice causes the dialectical connections between certain descriptive events with the situations, intuitions, and social structures that shape them. Moreover, the process of making English written text can be difficult for foreign language learners. So, they need components to make text connected within a sequence and can create a text based on the reader's expectation. Then, the need for cohesion is important as the element to create comprehensive text and coherence. These cohesive relationships between words and sentences have certain definable qualities that allow us to recognize the sentence.
Furthermore, to make the components of the text hang or hold together is by using cohesive devices. According to Gerot and Wignell's (1994, p. 170) statement, cohesive devices refer to the resources within language that provides continuity in a text. It gives us guidance on how the authors structure what they want to say and convey what is in their minds. Additionally, according to Halliday and Hasan (1976), the classification of cohesive devices is divided into two terms; grammatical cohesion and lexical cohesion. In grammatical cohesion proposed by Halliday and Hasan (1976), it consists of reference, substitution, ellipsis, and conjunction. Lexical cohesion contains reiteration and collocation. In addition, Halliday and Matthiessen (2014) assert that other exemplifies of lexical cohesion are not dependent on any general semantic relationship, but rather on a specific association between co-occurrence tendencies known as collocation.
In line with Crystal (2015), who claims that "the ties that bind a text together are frequently referred to under the heading of cohesion, and several types have been characterized, including conjunctive relations, coreference, substitution, ellipsis, repeated forms, lexical relationships, and comparison" (p. 261-262). So, a text must be coherent as well as cohesive. This indicates the concepts and relationships expressed should be relevant to each other.
However, the authors must generate and organize ideas at the level of wording by using appropriate word choices. Additionally, the text should be meaningful to turn such ideas into a readable text. Besides, they should employ a better combination of cohesive ties and coherent features in their product of writing. In the process of writing, the product or text production in terms of making interconnectedness between the sentences should be realized in meaning-making and lexical items to create unified text. Thus, a cohesive device is used to make the components of the text hang or hold the text together.
According to Eggins (2004), "lexical relations analysis is a way of systematically describing how words in a text relate to each other, how they cluster to build up lexical sets or lexical strings" (p. 42). It explains how lexical cohesion works between lexical content-encoding units. These refer to the open-class items such as nouns, main verbs, adverbs, and adjectives. Grammatical words or closed-class items include prepositions, pronouns, articles, and auxiliary verbs.
In addition, based on Halliday and Hasan's (1976) theory of cohesive devices, there are two types of text elements, i.e., internal and external. An internal element is called cohesion and an external element is called coherence. The concept of cohesion is an essential text element in creating well-organized and comprehensive text. These components allow people to combine texts in a sequence and create texts based on readers' expectations.
While Thornbury (2005) mentioned that coherence is a quality that the reader derives from the text: it is not simply a function of its cohesion. Moreover, he explained that conjuncts can express a variety of logical relations between parts of a text, i.e., the relation of addition, contrast or alternatives, cause and result, and sequence in time. Complete coherent texts can also be nonsense because the texts are linked not only to complete the use of cohesive devices but also through the context and the readership.
Research on this evaluative development of writing has been carried out with different goals and focuses. Here, some researchers are predominantly interested in investigating cohesion and coherence in the written text of students' academic writing. The following are Martinková (2009) Those researchers conducted studies focusing on text, newspaper, articles, and books. The findings show that the students' ability in writing increased for teaching writing that focused on cohesion and coherence devices. It is proved by the product of students' or writers' writing which is easy to understand and read.
Moreover, the components of cohesion in the text are needed in each sentence or paragraph. Then, the readers can understand well the content of the writing. Several studies highlighted the prominence of students' cohesion and the use of cohesive devices in the text (Nugraheni, 2016;Suwandi, 2016;Karadeniz, 2017;Wang & Zhang, 2018;Amperawaty & Warsono, 2019;Faizah et al., 2020;Lestari & Sutopo, 2020). Those studies concluded that grammatical cohesion and lexical cohesion could help transmit new information and the use of cohesive devices for English as a Foreign Language learners plays an important role in the text's cohesion and coherence, as well as in its persuasiveness. For academic purposes, past discoveries of study can be tested for future researchers and ought to be created with the growth of the times.
Moreover, based on the theme of the 8 th ELTLT Conference 2019, this study focused on the use of cohesive devices on the current issues and challenges of English language teaching, literature, and translation in the disruption era. The use of cohesive devices is intended to achieve the coherence of the text in this research article's analysis. This study has not been presented in any previous studies.
Concerning the importance of the purpose, this study enlarged and enhanced the readers' knowledge about how to write a good text by cohesive devices and coherent features properly in the text. In addition, a text has to be coherent as well as cohesive. The generalities, concepts, and relationships expressed should apply to each other, subsequently empowering us to make presumptive consequences about underlying meaning.
Following the background of the study, this study attempts to answer the research questions as follows.
How is the use of cohesive devices in the research articles on the 8 th ELTLT Conference 2019?
How is the use of references in the research articles at the 8 th ELTLT Conference 2019?
How is the use of substitution in the research articles at the 8 th ELTLT Conference 2019?
How is the use of ellipsis in the research articles at the 8 th ELTLT Conference 2019?
How is the use of conjunctions in the research articles at the 8 th ELTLT Conference 2019?
How is the use of reiteration in the research articles at the 8 th ELTLT Conference 2019?
How is the use of collocation in the research articles at the 8 th ELTLT Conference 2019?
How is the relation among cohesive devices to achieve coherence in the research articles at the 8 th ELTLT Conference 2019?

METHODS
In this study, a descriptive analysis approach was used in discourse analysis. It belongs to qualitative research that is fundamentally interpretive. In view from Creswell (2012), "it is concerned about data collection as well, the persuasiveness, and the self-awareness of the research" (p. 289). So, it means that we make an interpretation and description of the analyzed data.
Furthermore, simple qualification was used to demonstrate some tendencies in the types of cohesive devices used in the research articles. These tendencies are used to determine how cohesive devices are distributed. Besides that, observation sheets were used to examine the use of cohesive devices in research articles adapted from Halliday and Hasan (1976). All sections of research articles were examined in their entirety, starting from abstract, introduction, methodology, findings and discussion, and conclusions.
The objects of this study were 10 research articles which were written by authors from around Indonesia, Malaysia, and Bangladesh. The source of the research was at http://eltlt.proceedings.id.
Furthermore, purposive random sampling was employed to select the articles that were analyzed to meet the criteria.
The unit of analysis in this study is clause and sentence of research article at the 8th ELTLT Conference 2019. In categorizing cohesive devices, we used the table of cohesion adapted from Halliday and Hasan (1976). Each of the tables has the same column such as reference items (personal, demonstrative, and comparative), substitution (nominal, verbal, and causal), ellipsis (nominal, verbal, and causal), conjunction (additive, adversative, clausal, and temporal), reiteration (repetition, synonym, superordinate, general word), and collocation.
The procedures of data analysis were close reading the downloaded texts, breaking the text into clauses and sentences, identifying and classifying the use of cohesive devices based on the taxonomy proposed by Halliday and Hasan (1976), then tabulating the cohesive devices uses, calculating the occurrence of each categorizes of cohesive devices, then evaluating the coherence of the texts and interpreting the data. In addition, triangulation of analyst or expert and theory was used in this study as an investigator to validate the findings. The study's findings are then qualitatively described and interpreted in nonnarrative formats.

RESULTS AND DISCUSSION
The Use of Cohesive Devices in Research Article The first basic research question was how cohesive devices were in the research articles at the 8 th ELTLT Conference 2019. The cohesive devices are divided into two types; grammatical cohesion and lexical cohesion. In terms of grammatical cohesion, it is classified into several items: reference, ellipsis and substitution, and conjunction. Lexical cohesion is of reiteration and collocation. Reiteration consists of repetition, synonym, superordinate, and general words.
This segment examined all occurrences units of grammatical cohesion and lexical cohesion in authors' research articles. It could mainly be summarized that not all elements of grammatical cohesion were applied to associate the immediate sentences as the source of both cohesion and coherence. That occurred because some paragraphs demonstrated neither ellipsis nor substitution. These two units were only used a few times in this study. Moreover, not all the text in lexical cohesion contains reiteration, especially in superordinate.
There were ten texts from the authors' research articles that were chosen, reviewed, and analyzed in terms of cohesive devices. The number of occurrences of each unit in the texts was represented by the following finding. This can be summarized in Table 1:   Table 1 shows that the authors employed all types of cohesive devices. Concerning the reference, all types of reference are used. The authors were discovered to have used personal, demonstrative, and comparative references. As can be seen, demonstrative reference was the most frequently used by the authors when writing their research articles. Out of the 1955 clauses, there are 447 occasions of demonstrative reference. This indicates that the authors attempted to create coherence in a text by maintaining a pathway of information. It implies the use of demonstratives to refer to the extended text.
Furthermore, in terms of substitution and ellipsis, the results show that clausal, verbal, and nominal are represented in the combination of the total. The results showed that clausal substitution has mostly used in the authors' research article with a total of 108 clauses and followed by verbal substitution with a total of 43 clauses and the last was nominal substitution with a total of 28 clauses. Whereas, we only found two categories of ellipsis, i.e., clausal and nominal ellipsis. The least one is clausal ellipsis. It found that 9 items of clausal ellipsis and 81 items of nominal ellipsis out of 1955 clauses. In this case, the authors mostly used cardinal, ordinal, and indefinite quantifier words to develop their discussions. Further, since this study conducted text of proceedings so that ellipsis was rarely used by the authors. It was due to ellipsis being frequently found in spoken and text/speech. This indicates that several words and phrases created cohesiveness in the written text. In addition, the cohesive relationships between words and sentences have certain meaningful qualities by using substitution and ellipsis. Whereas we did not find any verbal ellipsis in ten texts. The authors of the research article did not present verbal ellipsis at all. It was due to the authors did not use both lexical verb and modal operators. Moreover, the structure of the verbal group does not represent its meaning in a direct and obvious way. However, most of the verbal ellipsis appears in spoken English since that is expressed by intonation.
The final type of grammatical cohesive relation is the conjunction. In the column of conjunction, most of the authors used all kinds of conjunction divided into additive, adversative, causal and temporal. The authors mainly used additive conjunction to give additional elements of explicitness and internal meaning in their sentences. It was found 541 items of additive conjunction out of 1955 clauses. Furthermore, it gives additional information without shifting any information in the previous clause or phrase in the authors' research article. So, the authors usually use it to develop their ideas and minds as support elements.
Moreover, lexical cohesion is classified into two types: reiteration and collocation. Regarding the reiteration that consists of repetition, synonymy, superordinate, and general word, the authors utilized each of the types of reiteration. Those are the repetition of 24 items, synonym 9 items, superordinate 5 items, and general word 17 items. Moreover, it is used to repeat words or word phrases that occurred within the text in the research article. Meanwhile, in terms of synonyms, it has the function to repeat the word by using another word that has the same or nearly the same meaning. Then, it plays an important role in the text because it not only created cohesion but also constructed the text to make it more amusing. Furthermore, the superordinate refers to any item whose meaning includes the earlier one, according to analysis. Briefly, the word comes first in the lexical structure, followed by the previous one. The last type of reiteration is the general word. It tends to be general when there is no detailed evidence and no-account explicit information.
The following part describes the use of collocation in texts. Through analyzing the use of collocation, it reveals that the authors used to write their research article. They used it to combine regular in which to fulfill the meaning, these words must occur together. That was found 85 items out of 1955 clauses.

The Relation Among Cohesive Devices to Achieve Coherence in the Research Articles
The analysis referred to Thornbury's, (2005) theory, that each sentence is linked to the previous one and connects text with a combination of identification and categorization, indicating that it has relation. Moreover, the findings show some clauses have logical relations in using cohesive devices. The use of cohesive devices has a huge impact to support text cohesion and making sense to the reader. It can be observed that the whole thing is logically organized and connected to the focus of the text. Therefore, the relation among cohesive devices to achieve coherence in the text was the direct involvement of all cohesive devices in making text.
However, the tendency to overuse cohesive devices should be controlled. It can impact the cohesion and coherence of the context. Even though the system is structured and contains cohesive devices, it does not mean that text is coherent and shows connectivity. Some cohesive device arrangements are positioned in the text inappropriately. It may reduce cohesion and coherence causing readers to lose track of the text. As a result, appropriate functional meaning relationships between sentences and paragraphs is required. By using various types of cohesive devices, it carries meaning within clause to sentence and the previous sentence into the next.
Here is an example of the logical relation among cohesive devices to achieve text coherence used by the authors of the research article at the 8 th ELTLT Conference 2019.
(1) So, the study projected in this article addresses the challenges and barriers of using digital content from Bangladeshi primary teachers' perspectives. (Akhter, 2019, p. 127) The example above shows the conjunction of "so" as a relation of result and conjunction of "and" as a relation of addition. Most of the authors often used "so" and "and" as internal linkers or conjuncts to connect one clause to the next or other clauses. However, there is no such structural relation. As long as the writing situation remains in a context that is not misleading, the communication process in creating text could build coherence. So, the use of cohesive devices is important to achieve coherence in the text.

CONCLUSIONS
Based on the research questions, we described seven conclusions. Referring to the first research question, the use of reference in the text found all kinds of reference, i.e., personal, demonstrative, and comparative reference. The total amount use of references was 1007 items appropriately used including 326 personal references, 447 demonstrative references, and 234 comparative references.
The second research question answers the use of substitution in the texts. Based on the data collected through documentation, there were three kinds of substitution used in the authors' research articles. The substitution reaching 179 items used. There were 108 items of clausal substitution, 43 items of verbal substitution, and 28 items of nominal substitution respectively. In applying substitution, the authors preferred to explain their ideas to be simpler.
The third research question answers the use of ellipsis in the texts. In applying ellipsis, we did not find any verbal ellipsis at all. In that case, the ellipsis is more frequent in conversation than in written text. It can be concluded that most of the authors used ellipsis suitably.
The fourth research question answers the use of conjunction in the texts. All types of conjunction are found in the texts. The authors used 779 items of conjunction. They dominantly used additive conjunction in the texts. As a result, they used all types of conjunction to link one clause to clause and one sentence to the following sentence.
The fifth research question answers the use of reiteration in the texts. Reiteration consists of repetition, synonym, superordinate, and general word. Based on the analysis, there were 55 items appropriately used reiteration. The authors used reiteration to make their product of writing more entertaining.
The sixth research question answers the use of collocation in the text. The analysis displayed those 85 items of collocation pattern. It examines the relationship between words based on the fact that they frequently occur in the same context. In addition, it is the regular combination of words to fulfill the meaning and the words must occur together. So, collocation produces a cohesive text.
The last research question answers the relation among cohesive devices to achieve coherence in research articles. The abstract, introduction, method, findings and discussion, and conclusion are all included in all research articles. Cohesive devices were found to be key aspects in achieving text coherence. In addition, cohesive devices also help a text hang together, make bridge the gap in the texts, and make sense.
As such, this research can be used as a reference for scholars, teachers, and future researchers to study the quality of written texts produced by authors or writers in other universities from various aspects of language around the world. Then, further studies need to explore the use of cohesive devices as a tool to improve the quality of writing. ecological views. International Journal of Language and Linguistics, 5(4), 67-75. Yuniawan, T., Rokhman, F., & Mardikantoro, H. B. (2017). The study of critical eco-linguistic in green discourse: Prospective eco-linguistic analysis. Humaniora, 29(3), 291-300.