Engagement Resources in Project Works Written by Selected Students at St. Bosco’s College of Education, Ghana

Writing a thesis in a second language (L2) remains a considerable challenge. By drawing on the Appraisal Theory (Martin & White, 2005), this study explored the way students of St. John Bosco's College of Education engage with the academic discourse community in their undergraduate (diploma) project writing. The data comprised introduction and conclusion sections of thirty (30) purposively sampled project works. Engagement items in the project works were examined manually. The results revealed that expand engagement markers were more frequently used than contract engagement markers, that deny was the dominant type of contract engagement markers , and that acknowledge was the most favored subtype of the expand engagement markers. This study confirms that an engagement system is a key tool to help novice writers align authorial voices with readers, thereby achieving promotional and persuasive purposes. The findings are of potential interest to other novice research writers and their examiners in future project work writing and examinations.


Introduction
Academic writing entails expressing a position on issues critical to a discipline, and this position must be presented within the conventions of a specific community (Hyland, 2004;Swales, 2004). These conventions both constrain the genre and allow for the carving out of a personal standpoint, referred to as "stance" and "voice" (Biber, 2006). According to Hyland (2012), these two concepts address the interpersonal dimensions of language, more precisely how writers and speakers present themselves and their ideas to specific interlocutors. The implication is that the ability to analyze one's material and establish rapport with readers is increasingly recognized as a necessary characteristic of successful academic writing. As a result, it is critical to emphasize the importance of writing in both the development of a scholarly identity and the project work experience.
Writing a thesis in a second language (L2) has always been challenging, as L2 writers must develop academic literacy skills that enable them to read and write complex materials. Additionally, this process requires considerable time and effort. Students must master the conventions and practices involved with participating in their discipline's discourse in order to become successful scholarly writers. As Paltridge and Starfield (2007) noted, novices must demonstrate their ability to conduct original research, possess specialized knowledge of research materials, and be capable of producing high-quality writing to be accepted into their disciplinary communities. The purpose of project/research writing is to engage readers not only by presenting an external reality but also by challenging accepted knowledge and validating new claims (Hood, 2004;Hyland, 2005;Xu & Nesi, 2019). Xu and Nesi (2019) added that the choices writers make to mitigate potential image damage may be influenced by the culture of the country and the expectations of the local discourse communities.
In Ghana, final year students in various colleges of education are required to conduct action research, and present a dissertation in writing. Conducting a successful research, termed "Project work," is a partial requirement for the award of Diploma of Education by the Institute of Education, University of Cape Coast, Ghana. Stated differently, like theses and dissertations, a project work is an advanced student-written research genre that many graduate students must complete prior to receiving their diploma, bachelor's, master's, or doctoral degrees. Such a writing task places enormous demands on students, particularly second language (L2) writers. Non-native writers of English frequently struggle to meet supervisors' and examiners' rhetorical needs and expectations, causing significant anxiety and frustration (Lee & Casal, 2014). Strategies that guide readers toward acceptance of these claims are particularly prevalent in the introductions of research articles/project works, where argumentation is used to convince the audience about the need for future investigations, and in the conclusions, where argumentation is used to convince the audience that the new research questions have been addressed (Berkenkotter & Huckin, 2016).
Many studies have examined research articles from European countries [for example, Denmark (Shaw, 2003), Finland (Mauranen, 1993), France (Vassileva, 2000), Germany (Sanderson, 2008), Italy (Molino, 2010), and Spain (Sheldon, 2013)], focusing on a variety of linguistic elements, including personal pronouns, hedges, negations, counterclaims, proclaims, conjunctions, and reporting verbs. These studies have generally advanced our understanding of the formal properties of these texts, following the Swalesian (1990) genre approach. As a result, Hyland (2004) believes that investigating the interpersonal components of academic writing is advantageous because they show the norms and expectations of specific cultural and professional communities. Furthermore, existing research indicates that writers' decisions may be influenced by national culture and the expectations of local discourse communities. It is, therefore, imperative that further studies be conducted to examine how engagement resources are deployed in undergraduate/diploma project writing in a non-native environment, particularly Ghana. Thus, the current study contributes to this line of investigation by analyzing a broader range of engagement resources in student-written project works.at St. John Bosco's College of Education. The study was guided by these research questions: 1. What are the types of engagement resources and their frequencies employed in the project works written by selected students from St. John Bosco's College of Education? 2. How do the students use individual engagement resources to articulate claims and engage with the research community in their project works?

Theoretical Framework: The Engagement Domain within the Appraisal Theory
We employed Martin and White's (2005) Appraisal Theory (specifically, the Engagement domain) in this study. "Engagement" encompasses a broad range of locutions that enable an author's voice to connect with the other voices and alternative positions that are considered relevant in a communicative event. The main reason for this choice is that the theory demonstrates interest in writer-reader relationships and context. As Xu and Nesi (2019) explained, the Engagement domain was created to assess writers' solidarity with their readers and their ability to anticipate potential objections to claims. Bakhtin's (1981) concept of Heteroglossia serves as the foundation for the Engagement domain in Appraisal Theory. The assertions: This information is likely to be accurate, or I believe this information is accurate, or Emmanuel believes this information is accurate; demonstrate heteroglossia. Martin and White (2005) define monoglossia as a contrastive phenomenon in which a speaker or writer makes unsupported claims without acknowledging or bargaining with opposing ideas and thus without providing dialogic space for the listener, as in the statement: This is accurate data.
Martin and White (ibid) coined the terms Contract and Expand to describe the extent to which a heteroglossic claim facilitates dialogue. Contraction occurs when the speaker/writer does not permit the expression of a diverse range of viewpoints. It requires writers to make a greater effort to guide the reader toward their own conclusion while acknowledging that others may hold opposing viewpoints. When the speaker/writer allows for multiple points of view, Expansion occurs. See Figure 1. The Heteroglossia system is depicted in Figure 1. The italicized examples in the figure show how each sub-system is realized via words. CONTRACT includes the DISCLAIM and PROCLAIM elements. The DISCLAIM option contains language resources that place the textual voice as opposed to, or rejecting, some opposite perspectives (Martin & White, 2005). It includes resources for DENY and COUNTER. The DENY resource contains negations (e.g., "no," "don't," and "never") that acknowledge alternative assertions but exclude them from the current discussion. The COUNTER resource (for example, "although," "however," and "even") has a counter-expectation connotation and refers to a statement that supplants another that would have been expected in the situation. The PROCLAIM category covers resources that contradict, conceal, or rule out alternative perspectives by portraying the claim as extremely justifiable (Martin & White, 2005). It is made up of three components: CONCUR, PRONOUNCE, and ENDORSE. The author's voice is portrayed as agreeing with or exchanging knowledge with the projected readers in the CONCUR resource. (e.g., "absolutely," "of course," and "not surprisingly"). The PRONOUNCE resource includes authorial emphasis on the proposition's correctness (e.g., "in fact," "the truth is," and "I thus suggest"). The ENDORSE resource validates the proposition that has been sourced externally (e.g., "display," "demonstrate," and "discover").
EXPAND includes the ENTERTAIN and ATTRIBUTE elements. The ENTERTAIN resource (e.g., "may," "could," "it appears," and "it's feasible") withholds writers' complete commitment to the concept being referenced, whereas the ATTRIBUTE resource (e.g., "say," "report," "argue," and "according to") connects the proposition to other sources without revealing the author's position.

Empirical review
The available literature indicates that, as regards the use of interpersonal resources in academic writing, genre, discipline, and culture all vary significantly. For example, Xu and Nesri (2019) investigated how two academics from different cultural backgrounds interact with their discourse community. Thirty applied linguistics research papers were evaluated using the Engagement system and the UAM Corpus Tool. It was revealed both academics in China, and the United Kingdom employed a range of engagement strategies, and the various combinations of engagement items resulted in a variety of interactive effects.
Other studies have examined the structure, function, and language characteristics of various chapters of thesis/project/research articles, including the introductions to Ph.D. dissertations (Bunton, 2002), literature reviews (Kwan, 2006), conclusions (Bunton, 2005), and discussion chapters of master's theses (Dudley-Evans, 1994), as well as variations in master's thesis introductions across disciplines (Samraj, 2008). These examples demonstrate that theses and dissertations employ distinct rhetorical frameworks in accordance with their objectives, requirements, expectations, and intended audience. It has also been demonstrated that theses and dissertations should not be viewed as deficient iterations of specialist research articles (RAs) but as distinct genres in their own right that are successful iterations of such genres due to their compliance with institutional requirements and examiners' expectations (Hyland, 2008, Koutsantoni, 2006. The review of available literature indicates that, although there are numerous studies on the RAs and theses Asia, Europe, and America, little can be said about Africa, particularly at the undergraduate or diploma level (in colleges of education), where such writers are seen as novice or apprentice writers, who face significantly more obstacles and require additional writing assistance (Bitchener & Basturkmen, 2006). Thus, in order to gain a better understanding of how apprentice writers of project works utilize Engagement resources, this Appraisal-based study reports on an Engagement in undergraduate (Diploma) project works written by non-native students of the St. John Bosco's Colleges of Education, Ghana, with a particular emphasis on the introduction and conclusion chapters.

Methods
We adopted the qualitative design. Some descriptive statistics were used to supplement the quality content analysis of the data.

Data Source, Size, and Sample
The undergraduate project works were chosen for this study. The sampled project works were collected from St. John Bosco's College of Education Resource Centre. This site was selected because of its proximity to the researchers. We purposively sampled the most current ones, numbering thirty (30) undergraduate-authored project works completed in 2019.

Data Collection Procedures
The researchers wrote a letter to the Principal via the Head of the Department of Languages, requesting permission to use the project works solely for academic purposes. We were granted formal access to the hard copies of the project works kept at the Education Resource Centre after lengthy discussions with the Principal about the necessity of the current study.

Data Analysis Procedures
We eliminated figures, tables, captions, and footnotes from the corpus. Additionally, we omitted all portions other than the introductions (which included, but were not limited to, sections describing the data and methodology) and conclusions (all sections following the section containing the research results). Thus, we considered all sections preceding the presentation of data and methods to be Introductions (including a literature review if it was a separate section), and all sections after the presentation of data and methods to be Conclusions. These two sections were chosen because they typically include the most arguments, necessitating more Engagement strategies.
The data analysis employed the Engagement domain (Martin & White, 2005). The researchers initially met to become acquainted with the engagement resources. As a group, we read one project work numerous times and analyzed the engagement resources contained within. We then granted ourselves three months to examine the data on our own. After three months, we reconvened and reviewed each instance in which we disagreed or had doubts. Eventually, we agreed on all of the data. Table 1 shows the make-up of the data.

Results and Discussion
The section discusses the results in terms of the types of engagement resources, their frequencies, and how the students used the resources to represent claims and engage with the research community in the data. Table 2 presents the types of engagement resources and their frequencies employed in the project works written by selected students from St. John Bosco's College of Education.  Table 2 illustrates the frequency of occurrence of engagement strategies in project work written by selected students of St. John Bosco's College of Education. The results indicate that the writers used the engagement strategies differently in interacting with their audience. This difference is particularly marked in the greater use of acknowledge (42.0%), deny (19.3%), entertain (17.0%), and endorse (15.0%) engagement strategies, indicating the students attempted to express a clear personal stance to their supervisors/readers. This finding is consistent with Xu and Nesi's (2019) assertion that the writers seek to conform to discourse community or genre conventions in order to gain acceptance for their claims or to persuade readers of the need for the new study. The implication is that engagement strategies act as a mitigating factor, and that subtle and complicated strategies are required to reduce this potential harm. The present study suggests the students' awareness of their supervisors' potential power and expertise. The greater use of acknowledge, deny, entertain, and endorse engagement strategies is critical for two reasons: (1) it closes down the power dynamics, expertise, and alternative interpretations; and (2) it signals the recognition of possible alternatives to their (writers') own viewpoint. As Cheng and Unsworth (2016) noted, writers risk ruining their relationship with their discourse community if they advocate for their views over established ones.
The subsequent sections focused on how the students used the engagement resources to represent claims and engaged with the research community in the data. We first began with the contract engagement resources and, finally, the expand resources.

Contract Engagement Resources
Contraction, as an engagement resource, occurs when the speaker/writer restricts the room for alternative views to be addressed. When writers utilize Contraction, they still allow the possibility that others might hold alternative perspectives but make major efforts to sway the reader(s) towards their own opinion. Table 3 presents the frequencies of contract engagement resources as utilized in the data. As illustrated in Table 3, deny engagement markers were most frequently used (48.2% of the total contract engagement resources) in the students' project works, indicating that some dialogic alternatives were directly rejected or supplanted. In the words of Martin and White (2005), these markers are intended to eliminate or limit the scope of specific dialogic choices in subsequent communicative interactions. Our findings indicated that concur, which comprises formulations that explicitly declare the addresser's agreement with, or possession of, some anticipated dialogic partner, was underutilized. Typically, this dialogic partner is the supposed addressee of the work, in which case the writer cannot claim the same knowledge as their superiors. Hence, the few uses of concur contractive resource.
The finding that deny the contract engagement markers was most frequently used by the students is not surprising: it serves to negotiate interpersonal meaning. The students used the resources (strategies) that evoke a preceding utterance or an alternative perspective in order to be directly rejected, substituted, or declared unsustainable. Obviously, denying or rejecting a position is maximally contractive in the sense that, while the alternative position is acknowledged, it is judged to be inapplicable. Consider, for example, the extracts below: Example 1: The research was limited to only reading as a skill, but not the other aspects: such as writing, speaking, and listening. Example 2: Unfortunately, most of the teachers do not have adequate knowledge on how to help such learners with special needs. The kind of training that teachers get does not adequately prepare them to handle all pupils.
In Example 1, not which falls under the disclaim of Deny, indicates a misalignment with some part genres of the language. Here, obviously, the writer indicated a misalignment with the other skills (e.g., "writing, speaking and listening") and, hence, aligned the supervisor//reader into a position of opposition to writing, speaking, and listening part genres.
Similarly, the Deny marker in the second example signaled that the writer disagreed with the claim relating to the expertise of the teachers to deal with special issues regarding the reading ability of the learners with special needs. The writer, here, employed "implicit negation." Pagano (1994) observes that when writers make such denials, they seek to project "existential paradigms" onto their intended audience. Denials such as those depicted in Examples 2 imply that the addresser is more knowledgeable than the addressee in some area and is acting to correct the addressee's mistake or misconception. As a result, it is viewed as corrective rather than confrontational, portraying the addresser as gently assessing the addressee's level of expertise and adjusting their speech accordingly. As long as the reader is not resistant to having this particular lack of knowledge projected onto them, and as long as they have no cause to reject the particular position offered, such writers will foster solidarity. (Martin & White, 2005).
Another relevant contractive marker worth discussing is Endorse (with 52 occurrences in the data). These formulations, rather than directly rejecting or overruling a contrary position, act to limit the scope of dialogistic alternatives in the ongoing colloquy (Martin & White, 2005). See Examples 3, 4, 5, and 6 below: Example 3: The study revealed that form one pupils of Doba Junior High School have talents and potential to learn integrated science and that their interests in the subject depend largely on a number of factors. Example 4: In conclusion, the study also revealed a lot of approaches by different authors on how to conduct an experiment to compare the water holding capacity of the three types of soil. This came to light during the review of the related literature. Example 5: The results showed that eighty percent (80) of the pupils could not read well. Example 6: The table shows that one (1) pupil representing four percent (4%) scored six (6) marks, four (4) pupils representing sixteen percent (16%) scored five (5) marks.
All four (4) extracts are dialogistic in nature, referring openly to empirical data and perspectives from other approaches. The first and second extracts [3, 4] demonstrate a formulation that makes use of a special type of reporting verb (revealed)one that takes a viewpoint on the ascribed claim, believing it to be true. Kiparsky and Kiparsky (1977) had extensively examined this form of reporting verb in relation to concepts of "factivity." By employing such "endorsing" phrases, the authorial voice establishes the claim as "true" or "legitimate," aligning itself with the external voice introduced as the source of the proposition. According to Martin and White (2005), such formulations demonstrate the author's increased engagement in the current rhetorical cause and co-opt an authoritative second party. As a result, they position themselves against, or at the very least ward off, real or possible opposing positions. The reporting verbs (e.g., revealed and demonstrate) are used in the preceding extracts to relate the statement with individual subjectivity and, more specifically, with the authorial voice's subjectivity. Given that particular subjectivities are continuously at odds with and opposed to other subjectivities, endorsements help to construct a heteroglossic backdrop of various alternative viewpoints on the concept. The analysis revealed that the other contractive resources (such as Pronounce and Counter) were marginally utilized in the data. See the Table 3. Pronounce resources concentrate on formulations that contain explicit authorial emphasis, interventions, or interpolations. For instance, I assert..., The facts of the situation are as follows..., The reality is that..., We can only conclude that..., intensifiers having clausal scope such as really, indeed, etc. See the extract below: Example 7: The many benefits of integrated science cannot be overemphasized. As evinced in the data, Cannot be overemphasized involves authorial interpolations and emphasis directed against some assumed or directly referenced counter position. Such formulations are dialogistic in that they acknowledge the presence of this counter view in the current communicative setting, and are contractive in that they challenge, confront or resist this particular dialogistic alternative (Martin & White, 2005).
Counter, on the other hand, focuses on formulations that represent the current proposition as replacing or supplanting, and thereby 'countering,' a proposition that would have been expected in its place. The countering is typically conveyed via conjunctions and connectives such as although, however, yet, and but. An example is the extract below.
Example 8: However, after the intervention process, it was again observed that about eighty-four percent (84%) of the pupils could experiment to demonstrate the water holding capacity of the three main types of soil using simple apparatus while the rest have some few challenges with the topic. We turn our attention to expansive engagement resources in the next discussion.

Expansive Engagement Resources
This dialogic expansion is the degree to which an utterance, by dint of one or more locutions, actively makes allowances for dialogically alternative positions and voices by a speaker/writer. It is usually achieved through the use of locutions such as Entertain, Acknowledge, and Distance (Martin & White, 2005). Table 4 presents the frequency of occurrence of the expansive engagement markers in the selected project works written by students of St. John Bosco's College of Education. As evidenced in Table 4, the writers of the project works tended to utilize the dialogic expansion of Acknowledge (147, representing 70.3%) the most, followed by Entertain (59, indicating 28.2%), and finally, with least preference to Distancing (03, indicating 1.5%). We discuss the individual dialogic expansions in detail in the subsequent sections.

Dialogistic expansion of "Entertain"
The Entertain option implies that its stance is only one of several conceivable positions and, thus, creates dialogic room for those possibilities to a greater or lesser extent. It includes the meanings of modal auxiliaries (e.g., may, might, could, must, etc. ), modal adjuncts (e.g., perhaps, probably, definitely, etc. ), modal attributes (it's possible that..., it's likely that..., etc. ), and certain mental verb/attribute projections (I suspect that..., I believe, I'm convinced that..., I doubt, etc.) (Pinying, 2018). The extract below exemplifies the use of Entertain locutions in the data.
Example 9: Piaget, Brunner, Zoltan, and Skemp cited in Martin (1994), stated that "in learning, the mental processes of the child must be taken into account, children cannot learn the same content as adults and individual children will differ in the way they learn.
The above example encompasses meanings by which the speaker/writer makes assessments of likelihood via modal auxiliaries. The writer seemed to be siding with Piaget, Brunner, Zoltan, and Skemp on their claims. According to Martin and White (2005), this is dialogistic because such locutions are viewed as actively constructing a heteroglossic background for the text by overtly grounding the assertion in the writer's individual subjectivity and thus recognizing that the assertion is only one of several available in the current communicative context. The modal must, as a dialogistic perspective, focuses our attention on "epistemic status" and "reliability of knowledge" because they are seen as the primary, determining communicative motive for ensuring that the appropriate conditions and modalities are met in accordance with the empirical views advanced by various authorities (Piaget, Brunner, Zoltan, and Skemp) regarding the appropriate conditions of learning. Additionally, it directly bases the demand in the speaker's subjectivity -as an appraisal of obligation by the speaker (Martin & White, 2005).
Apart from the use of modals as indicated above, the expansion of entertain is more diverse as it also includes 'evidentials.' See the Example 10: Example 10: The random sampling method was, however used to select pupils from the class for the study because they all seem to have the same characteristics.
As evidenced in Example 9, seem presents one proposition from a set of possible alternatives and, hence, creates dialogic space for any of them. Martin and White (2005) assert that an Entertain evidential (seem) provides a proposition that is interpreted as contingent and subjective as a result of being obtained through a process of deduction or supposition on the part of the writer.

Dialogistic expansion of "Acknowledge"
Acknowledge refers to those locutions in which there is no overt indication, at least through the choice of framer, of the authorial voice's viewpoint on the proposition. This is where reporting verbs such as say, report, state, declare, announce, believe, according to, and think belong. Find Examples 11, 12, and 13: Example 11: Tsadidey (2003) defines reading as "the ability to converse with written materials or get a message that has been set out in a specific code." Example 12: According to Stauffer (1975), "reading is recognition, ability to call to mind particular meaning and ability to shift or re-associate meaning until the concepts presented are clearly grasped, critically evaluated, accepted, applied or rejected." Example 13: Piaget, Brunner, Zoltan and Skemp cited in Martin (1994), stated that "in learning, the mental processes of the child must be taken into account, children cannot learn the same content as adults and individual children will differ in the way they learn.
In Examples 11, 12, and 13, the highlighted Acknowledge resources (defines, according to, and stated) convey neutral position and portray the authorial voice as interacting with those voices. They, thus, explicitly interpret the communicative environment as heteroglossic (Martin & White, 2005). The reader interprets the writers in such instances as having nothing invested in the position being advanced in the reported material. According to Martin & White (2005), such acknowledgements enable the writer to maintain a distance from any aligned or misaligned relationships. The resources portray the writer as an "informational fair trader" who merely relays the ideas of others and is, thus, uninvolved in any solidarity bond the reader may build with the quoted source. Therefore, it suffices to say that students of St. John Bosco's College of Education use the expansive engagement marker of 'acknowledge' to overtly construe the communicative setting as heteroglossic and bring in external sources to lend support to their argument.

Dialogistic expansion of "Distance"
Distance involves formulations in which, via the semantics of the framer employed, there is an explicit distancing of the authorial voice from the attributed material (Martin & White, 2005). It is most typically realized by means of the reporting verb, claim, and certain uses of 'scare' quotes. See the Example 14 below: Example 14: They claimed that teachers must demonstrate how language items are used and in what situation they are appropriate As in Example 14, Distance signals that the writer does not take responsibility for what is being reported. This finding concurs with the view of Caldas-Coulthard (1994) that the writer/speaker "detaches him/herself from responsibility for what is being reported" (p. 295). Additionally, we agree with Martin and White (2005) that it contributes to the very solidarity that is constructed and assists the writer in signaling recognition of possible alternatives to their own viewpoint.

Conclusion
By drawing on the Appraisal Theory (Martin & White, 2005), this study explored the way students of St. John Bosco's College of Education engaged with the academic discourse community in their undergraduate (diploma) project writing. The data comprised introduction and conclusion sections of thirty (30) purposively sampled project works. Engagement items in the project works were examined manually. The results revealed that expand engagement markers were more frequently used than contract engagement markers, that deny was the dominant type of contract engagement markers, and that acknowledge was the most favored subtype of the expand engagement markers. This study confirms that an engagement system is a key tool to help novice writers align authorial voices with readers, thereby achieving promotional and persuasive purposes.
The findings of this study are especially useful to individuals who provide writing assistance to novice research writers, as well as those who review and edit international research, because they shed light on how apprentice writers in a non-native setting employ engagement resources in their work.