The negotiation of authorial persona in dissertations literature review and discussion sections

Full Text

1. Introduction

Negotiation in academic writing is rooted in a social perspective to knowledge construction. Cherry’s (1988) perception of scientific debate illustrates the way knowledge is created; “scientific facts or knowledge are not ‘discovered’ by individuals in isolation but established through consensus-building discourse in scientific communities” (p. 266). In addition to factual content, scientific debates are about the individuals’ intersubjective exchange (Grossberg 1982). Dissertation writing as a type of scientific discourse is not about transmitting knowledge from an addresser to an addressee or audience, but about creating new knowledge through consensus-building with members of a community of interest. The function of a dissertation is to negotiate a distinctive authorial voice to convince the audience of the validity of the argument. The relation of a text with other texts is discussed as “intertextuality” (Bazerman 2004: 3), “dialogism” (Ewald 1998: 226), or “dialogicality” (Prior 2001: 63) to stress the interactive nature of any text. Interaction is associated with stance and voice expression– which represents highly valued features of academic writing (Alotaibi 2019, Alramadan 2020, Aull & Lancaster 2014, Barnawi 2011, Charles 2006, Hyland 2019, Street 2009).

It nevertheless represents a tricky facet of academic writing; the understanding of voice and stance as reflective of the writer’s subjectivity and personality may clash with the objectivity of academic writing (e.g., Gillett, Hammond & Martala 2009, Swales & Feak 1994). The balance between the observance of norms and making one’s contribution distinct bestows the success of an academic text (Lee 2011, Mei and Allison 2005, Sancho-Guinda & Hyland 2012). The task is further complicated for EFL writers, as is the case in the present research (Allison, Cooley, Lewkowicz & Nunan 1998, Bitchener & Basturkmen 2006, Can & Yuvayapan 2018, Paltridge & Starfield 2020). Researchers such as Elghoul (2016), Hajji (2012), Rouissi (2013), and Triki (Personal communication, June 17, 2019)¹ observe that Tunisian EFL researchers fail to craft a distinctive authorial persona in their dissertations. It makes their texts sound less effective. Hence the need to explore identity manifestation in this genre of academic writing regarding the high-stakes it represents for the writers.

The present study aims to unveil the way(s) postgraduate EFL researchers position themselves with reference to the authorities they refer to and to the audience they address through the systematic analysis of the engagement resources (Martin & White 2005) they rely on to establish a dialogue with the academic community. Engagement constitutes the heteroglossic nature of the text. Dialogic expansion and contraction account for the writer’s authorial manifestation by situating his/her stance regarding voices exterior to the texts. The rationale behind the cross-generic comparison is to understand the way(s) EFL postgraduate writers linguistically manifest their authorial persona as a persuasive strategy to negotiate a position within the academic community. MA and PhD dissertations represent substantial differences by virtue of the researchers’ place in the community and their experience in academic writing. The cross-degree differences represent areas of potential development that need to be foregrounded in order to help EFL writers master the intricacies of negotiating a personal voice. Besides, as a rhetorical construct, identity manifestation is not static. It differs depending on the dissertation section’s objective(s). The cross-sectional comparison is meant to assess the writer’s authorial presence in two rhetorically different sections– the Literature Review (henceforth LR) and the Discussion section (henceforth D). To this end, the following research questions are formulated:

1. In what ways, if any, is voice expression in MA and PhD dissertations manifested through the engagement system (i.e., dialogic expansion and contraction)?
2. In what ways, if any, does voice expression differ across the degrees (MA vs. PhD) and across the sections (LR and D)?
3. What do the differences/similarities reflect about the dialogic nature of writing, as demonstrated through expansion and contraction across the degrees and across the sections?
4. What do the differences/similarities reflect about the writer’s authorial presence and audience engagement as a persuasive strategy across the degrees and across the sections?

2. A social view of genre

Within the social view, genre is defined as “types of texts that are recognizable to readers and writers, and that meet the needs of the rhetorical situations in which they function” (Swales 1990: 467) [emphasis in the original]. Texts have well-defined features that are used to achieve specific rhetorical purposes. MA and PhD dissertations² represent a distinct genre of academic writing as they have a specific goal(s), follow a particular structure, and adhere to conventions and standards that distinguish them from other forms of written discourse. Gosling and Noordam (2011) see that in a dissertation, the researcher is expected to look into original data in order to test a hypothesis that he/she formulated about a specific phenomenon. Different formats are possible in a dissertation (Paltridge 2002: 131–132).

In the present study, the I(Lr)MRD organization (Johns & Swales 2002) is referred to as an example of dissertation structure. More specifically, a focus on the LR and the D sections as two distinct “(sub)-genre[s]” (Thompson 2009: 50) is done to understand their rhetorical functions and to determine what is expected from the writer in each section. Indeed, because they represent two distinct genres, the LR and D have different rhetorical purposes (Salager-Meyer 1994). They are expected to display different ways of meaning negotiation and different configurations of authorial presence as a rhetorical strategy. The purpose of the LR section is to place the work within a larger research context, highlight the gap in the literature, and emphasize the importance of the research perspective (Ellinger & Yang 2011, Imel 2011, Paltridge & Starfield 2007, Swales & Feak 1994). In the D, the researcher emphasizes his/her findings. He/she refers to the literature to support his/her claims. The writer thus endorses different roles. The aim behind the study of expansion and contraction strategies is to unveil the differing rhetorical maneuvers across the sections.

There is also evidence that the writing requirements vary by degree. There seems to be an agreement that the higher the degree, the more is expected from the researcher (e.g., Allison et al. 1998, Johns & Swales 2002, Nackoney, Munn & Fernandez 2011, Street 2009, Woodward-Kron 1999). The writers’ presumed role(s) and identity shift as they move from one level to the next. Although treated under the same label as postgraduate writing, MA and PhD texts present substantial differences (Paltridge & Starfield 2007, Richards & Miller 2005). According to Paltridge and Starfield (2007), there is a variation in the writer’s performance and the supervisors’ and examiners’ expectations depending on the degree of study. They emphasize that contribution to knowledge is especially vital for PhD research in comparison to MA writing. Critical thinking, openness to questioning, and reflective curiosity are abilities that researchers learn through their academic writing training. These features are expected to be present in the writings of PhD researchers. The comparison of expansion and contraction across the corpora is meant to assess EFL postgraduate writers’ skills to make their texts dialogic and stress their contribution to knowledge construction.

2.1. Cross-sectional differences: The Literature Review vs. Discussion sections

In a dissertation, the LR section is regarded as an “extensive”, “comprehensive” and “thorough” coverage of previous literature (Ellinger & Yang 2011: 117). Researchers (e.g., Ellinger & Yang 2011, Imel 2011, Paltridge & Starfield 2007, Gillett et al. 2009), however, warn against falling into the trap of making this part a mere summary of who said what. The LR is often transformed into a section where novice academic writers “regurgitate” what they found in previous literature, to borrow Gillett et al.’s (2009: 97) word. Rather, the LR is a place where the writer needs to synthesize what has been previously said. Synthesis is different from summary because it allows the researcher to come up with new knowledge and new perspectives (Imel 2011: 146–147) rather than repeating existing ones.

The researcher needs to express a personal stance and to construct an authorial voice to show the relevance of what he/she is reporting to his/her major claims and to foreground his/her contribution based on the works he/she reviewed and criticized (Bizzell 1992, Nackoney et al. 2011). Students at a postgraduate level are expected to add something to their field of study, not just to reproduce what has already been said. In evaluating other works, some footing (Triki 2019) is needed to show humility and avoid the reader’s possible disagreements and counterarguments. Humility should not, however, be understood as stifling one’s voice. Finding a balance between expressing a distinct persona and showing deference to and recognition of others’ voices is a challenging aspect of academic writing (Hyland 1998, Salager-Meyer 1994, Woodward-Kron 1999).

The D section is described as the “reverse” form of the LR (Paltridge & Starfield 2007: 146). The researchers qualify it as such because, in the LR, the focus is on previous research and the student’s work is at the periphery. On the contrary, in the D, the primary focus is on the student’s research, and the literature is at the periphery; it serves to back up the student’s work. The function of the D section is to convince the reader of the significance, validity, worthiness, and contribution of the research (Ellinger & Yang 2011, Flowerdew 2000, Saz-Rubio 2011). It is to be noted that the D section is often considered with the section that directly precedes it, which is the Results (henceforth R) section. Swales and Feak (1994), however, distinguish the two and insist that the D is “more than summaries” (p. 195) of the findings. After dealing with factual information in the R section, the writer is expected to move to the interpretation and evaluation of the outcomes in light of previous literature. Therefore, it is assumed that the D is about comparing, contrasting, and positioning one’s results with previous works. The difference with the LR is that the D is about claiming ownership, stressing personal contributions, and positioning oneself as an active agent in knowledge-making.

Authorial presence is hence needed to achieve the rhetorical function of the LR and D sections. Yet, because the writer has different roles and positions, it is expected that voice expression is achieved in different ways. The management of the voices used to construct the argument needs to serve the balance between acknowledging others’ contribution(s), asserting personal distinctiveness, and responding to audience potential disagreement. According to Hyland (2019), for example, hedges are usually used to open the dialogue and leave the space for other interpretations. They correspond to what Martin and White (2005) call dialogical expansion. On the contrary, boosters are used to stress the strength of the assertion and close the dialogue, according to Hyland (2019). Their use is an instance of dialogic contraction following the definition provided by Martin and White (2005). Other ways of opening and closing the dialogue are possible. This study is meant to spot the lexical and grammatical choices that MA and PhD EFL writers make to signal other voices in their dissertations either by opening or by closing the dialogue with the audience.

2.2. Postgraduate writing: differences across degrees

MA and PhD writing present two different subgenres within the dissertation category when it comes to the researcher’s roles and presence in the text. In effect, the academic discourse community is seen as a “hierarchical” structure (Hyland 2006: 20) where MA researchers are placed at the periphery. PhD writers are more established in the community. Yet, they occupy a lower hierarchical place in comparison to established scholars. The researchers’ statuses at different degrees give them specific roles to perform within the group. It is reflected in the way they manifest their authorial presence in the academic text. An academic writer is expected to develop higher-level thinking skills as he/she ascends the literacy ladder. Differences in critical thinking abilities are observed between MA and PhD researchers, with the former being considered as less experienced in academic writing.

Contribution to knowledge represents an important criterion that distinguishes the different stages of literacy (Street 2009, Nackoney et al. 2011). Paltridge and Starfield (2007) argue that “A good thesis or dissertation should tell the reader not just ‘what I have done’, but ‘why what I have done matters’” (p. 154). It is not enough to be knowledgeable of previous research; academic writing aims to evaluate the state of the art (Bizzell 1992, Nackoney et al. 2011). Evaluation involves the writer’s stance expression towards what is being reported and the audience’s reaction to it. As the researcher’s role(s) highly depends on the research stage he/she is accomplishing, evidence of his/her voice becomes necessary for the success of a piece of research.

The writer’s manifestation is especially important in the humanities and social sciences (Hyland 2008; Lee & Aitchison 2011; Myers 2001). The researcher guides the reader through text unfolding and factual interpretation. Dialogic expansion and contraction are therefore expected. Nevertheless, dissatisfaction with the postgraduate writer’s authorial presence in the text is reported. In the case of dissertations in social sciences written in the Tunisian EFL academic context, MA and PhD writers are often criticized for the lack of voice and stance, lack of personal contribution, and inability to position their work (e.g., Elghoul 2016, Hajji 2012, Rouissi 2013, Triki 2019). In the present study, we assume that problems of authorial voice expression are more salient in an EFL context. In effect, writing in a foreign language seems to pose further challenges for postgraduate writers, as will be explained in the following section.

2.3. Extra challenges for postgraduates writing in English as a Foreign Language:

The review of the literature about L1 and L2/EFL writing shows that writing in an L2/ EFL presents extra challenges to the writer in comparison to writing in the L1 (e.g., Connor 2008, Intaraprawat & Steffensen 1995, Mohan & Lo 1985, Paltridge 2002, Paltridge & Starfield 2020, Weigle 2002). Paltridge (2002), for example, asserts that “Thesis and dissertation writing is a difficult process for native speaker students and often doubly so for non-native speaker students” (p. 137). This is mainly due to some linguistic (e.g., Casanave & Hubbard 1992, Intaraprawat & Steffensen 1995), cultural (e.g., Connor 2008), and educational (e.g., Hyland 2003) factors that make the EFL text less effective. Chang (2015) notes that EFL writers “are often found to have weaker control of their academic writing, among which presenting an effective authorial stance has been reported as particularly challenging” (p. 1).

EFL writers’ mental schemata are constrained by their limited mastery of the foreign language’s formal and pragmatic features. MA and PhD researchers may be aware of the need to express a voice but miss the opportunity to appropriately establish a dialogue due to linguistic and cultural differences between the L1 and the EFL. As advanced by Contrastive Rhetoricians, what is accepted in one linguistic and cultural group might sound unfamiliar in another group. EFL writers need to develop, in addition to formal knowledge, intercultural communication skills that allow them to appropriately negotiate a position in a specific discourse community. It is also influenced by the educational environment, which favors, or not, such skills as critical thinking and openness to questioning. The inability to evaluate previous works and establish a dialogue with the audience might be due to inadequate knowledge of audience expectations at different educational levels.

3. Methodological framework

3.1. The Appraisal framework: Engagement

Authorial presence in Tunisian postgraduate EFL writings is examined using Martin and White’s (2005) Appraisal framework. It is concerned with the study of interpersonal meaning expressed through epistemic and affect modality. It involves three main subsystems; engagement, attitude, and graduation Fig. 1). The choice of the framework is motivated by its adequacy for the “analysis of stance as positioning in relation to values and voices in the text” (Hood 2004: 13)³.

Fig. 1. An overview of the Appraisal framework (Martin & White 2005: 38)

The present study is concerned with the engagement system, which deals with epistemic stance meanings (Gray and Biber, 2012, p. 18). Engagement is defined as “those meanings which in various ways construe for the text a heteroglossic backdrop of prior utterances, alternative viewpoints and anticipated responses” (Martin & White 2005: 97). It refers to the writers’ stance towards the material they refer to and the audience they address. More specifically, it corresponds to “epistemic stance” (Gray & Biber 2012: 17) or “evidentiality” (Chafe 1986, Chafe & Nichols 1986). It denotes “the status of the knowledge contained in propositions.” (Gray & Biber 2012: 16). Engagement is contrasted with “attitudinal stance” (ibid) or “affect” (Ochs, 1989) which rather refer to “personal feelings, emotions and attitudes rather than evaluations of knowledge.” (Gray &Biber, 2012: 17). We limited our exploration to the engagement system because according to Gray and Biber (2012), in academic research, epistemic stance is regarded as more relevant than attitudinal stance (attitude). Graduation has to do with the force and focus of the utterance. It does not support the dialogic nature of the exchange. It is therefore not taken into consideration in the present work.

The expression of stance can take two possible directions for the writer following the engagement framework (Fig. 2); the author can “make[s] allowances for dialogically alternative positions and voices” (Martin & White 2005: 102); he/she is thus said to open the dialogue with the audience and be dialogically expansive, or “act[s] to challenge, fend off or restrict the scope of” (ibid) other voices in the text. In this case, the writer is said to be dialogically contractive by closing the possibility of a dialogue with potential audience. The study of stance using this system not only unveils the writer’s authorial positioning but also foregrounds the resulting voice that the writer constructs throughout the text.

Fig. 2. The Engagement system (Martin & White 2005: 104)

The systems are further divided into more subsystems, but for the sake of the present research questions, the analysis of the corpus is going to be done based on the distinction between dialogically expansive and dialogically contractive expressions to see whether the writers’ voice is used to open up the dialogue or whether it is used to close up the discussion.

3.2. The corpus

The corpus under study consists of 20 LR and 20 D sections taken from ten MA and ten PhD dissertations written in English as an EFL by male and female Tunisian researchers in the field of applied linguistics. It was selected following a “random sampling” technique (Kothari 2004). Only dissertations written in applied linguistics were chosen to avoid differences due to disciplinary preferences (Becher & Trowler 2001). Besides, only dissertations that follow the I(Lr)MRD style were selected. It is to be noted that in both the MA and PhD dissertations, the R and D sections are not separated. Therefore, it was only possible to study both the R and the D together (henceforth RD). Finally, because interest is in the analysis of the writer’s linguistic choices, all quotes and labels are discarded from the LR and RD sections. The number of words per subcorpus is summarized in Table 1 below.

Table 1. Number of words in each subcorpus

3.3. Data analysis

The corpus was first qualitatively analyzed for the use of engagement resources. Expressions with dialogically expansive and dialogically contractive meanings were identified, then classified into the expand or contract categories. The UAM corpus tool was used for that purpose. The data were then exported to SPSS for quantitative analysis. A one-way ANOVA test was computed to study the similarities/ differences in the use of the expand and contract categories across the degrees and across the dissertation sections. To that purpose, the following hypotheses were formulated:

H_0: It states that the mean difference between the categories is not significant.

H_1: It states that the mean difference between the categories is significant.

Next, a significance level p≤ 0.05 was chosen as it is common to set in social sciences. It represents “the probability level below which the researcher is willing to treat his/her hypotheses as significant” (Triki & Sellami-Baklouti 2002: 56). It means that when the significance level is above 0.05, the Null Hypothesis cannot be rejected. If, on the contrary, p is below 0.05, H₀ is rejected, and H₁ is kept.

4. Results

4.1. The distribution of expansion and contraction across the corpora

The distribution of expansion and contraction in the corpus is summarized in Table 2 below. The difference in the number of words across the subcorpora is dealt with by computing engagement resources per 1000 words.

Table 2. Expansion and contraction mean occurrence in MA and PhD dissertations LR and RD sections

The exploration of the study results through the SPSS software shows that mean differences between expansion and contraction categories are not statistically significant when the rhetorical sections (LR and RD) are considered. The results are presented in Table 3. However, the difference is statistically significant when the corpus is studied taking the degree (MA and PhD) into consideration, as is shown in Table 4.

Table 3. ANOVA test for expansion and contraction categories across the sections

Table 4. ANOVA test for expansion and contraction categories across the degrees

The statistically significant difference between MA and PhD dissertations in the use of expansion and contraction in the present corpus suggests that the ability to establish interpersonal relationships in the text is related to the writers’ degree rather than to the rhetorical section. This might imply that voice and stance expression is a developmental feature of academic writing, as is advanced by Aull and Lancaster (2014). Researchers at different educational degrees have different strategies of authorial expression and positioning in their texts. The perception of their active role in knowledge construction might enhance their ability to create a dialogue and establish networks. It is reflected in their use of expansion and contraction to evaluate the literature they refer to, make assumptions, or address the reader. The findings echo those made by Barton (1993), Lancaster (2014), Nackoney et al. (2011), Paltridge and Starfield (2007), and Street (2009), who see that the writers’ ability to negotiate a position and to claim ownership depends on their literacy level.

To check the validity of these claims, a detailed analysis of expansion and contraction in each section is presented below. The mean difference of expansion and contraction in the MA and PhD LR and RD sections are illustrated in Fig. 3.

The ANOVA test is conducted to see if the mean difference is statistically significant. The results are summarized in Table 5.

The ANOVA results revealed that mean differences for the expansion and contraction categories in the MA and PhD dissertations are statistically significant (p ≤ 0.05) for all the sections except for contraction in the LR section, where p=0.051. In the latter case, H0 is kept. It shows that, quantitatively, contraction in the MA and PhD LR sections is used similarly. It is, however, more frequently used in the PhD RD section. Expansion is more frequently used in the PhD dissertations in both the LR and RD sections. The differences in the use of expansion and contraction across the sections are in line with the claim that LR and RD represent distinct subgenres (Thompson 2009) and have different rhetorical purposes (Salager-Meyer 1994). Therefore, the writer adopts different tones of voice across the sections.

Fig. 3. Expansion and contraction in MA and PhD LR and RD sections

Table 5. ANOVA test results for expansion and contraction  in MA and PhD dissertations LR and RD sections.

These results show that PhD dissertations present more instances where dialogicality is enhanced through expansion resources in both LR and RD. MA writers seem less inclined to open the possibility for other voices to be heard in both the LR and RD sections. It might be explained by the MA researchers’ limited experience in writing academic texts as long as the dissertation and their failure to see academic discourse as an intersubjective and interpersonal genre. Their limited use of expansion in comparison to their PhD peers reflects their inability to take a critical distance towards the voices they are using as evidence and towards the reader as an active part in argument construction.

The tendency to close the dialogue through contraction is similar in both the MA and PhD LR sections since no statistical difference is observed in means occurrences across the degrees. The tendency might be explained by the writers’ awareness of their role in this section regardless of their degree. Indeed, the function of the LR is to review the state of the art; the writer is expected to be open to different perspectives. Therefore, closing the dialogue through contraction is not the appropriate strategy to adopt to meet the LR rhetorical function.

A significant difference between the MA and the PhD writers is, however, observed in the use of contraction expressions in the RD sections. The rhetorical function of this section might explain this finding; the writer is expected to describe the study outcomes, compare and contrast them with those of previous research, and come up with conclusions concerning the issue being commented on. The writer is expected to highlight personal contribution and distinctiveness. Strong authorial presence is a strategy to cut clear an argument and to negotiate a voice among the panoply of voices present in the text. It is to be noted that the PhD subcorpus displays more instances of contraction use. Stronger authorial persona supports the claim made earlier that PhD researchers display greater mastery of the rhetorical strategies at hand to position themselves in the community (Barton 1993, Lancaster 2014, Nackoney et al. 2011, Paltridge & Starfield 2007; Street 2009).

4.2. Dialogicality and authorial presence in the Literature Review

The examples below are taken from the LR sections written by MA and PhD researchers. In example [1], the MA researcher deals with a key concept in his/her work which is “face-to-face political talk”. He/she refers to an authority figure to describe the concept. The use of the reporting verb is expansive as it opens the space for a dialogue between the different voices present in the text. The verb ‘contend’ actually conveys three voices– the source’s voice, the writer’s, and the reader’s. The reporting verb is multifunctional; it introduces the reference and signals the researcher’s detachment from what is being advanced, which might protect him/her from the reader’s potential disagreement. Anticipation of the audience reaction is also manifested through the use of the hedge ‘oftentimes’. The writer, then, strategically moves to contraction through the use of “much”, “more”, and “fervent” in the explanation of the concept being discussed to lead the reader into a specific evaluation. The voices in the text seem to converge towards the same interpretation. The problem is that the writer’s stance is not properly signaled. It is not explicitly distinguished from that of the original source. It weakens the MA researcher’s authorial presence and might be a source of criticism for him/her.

[1] Schudson (1997:299) contends that face-to-face political talk is “profoundly uncomfortable” as it is built on controversial issues that cause interactants to, oftentimes, withdraw from much of their civility and respect in favor of a more robust and fervent (MA LR)

Similarly, in example [2], the MA student does not establish clear boundaries between his/her personal stance and that of the reference he/she is using to comment on a key concept in the literature review. The use of the expression “One of the advantages” is expansive as it gives the reader the impression that the writer is evaluating the state of the art based on what came in the literature while the concept is being evaluated with reference to one authority (i.e., Martin (2000)). The writer merges his/her voice with that of the reference. This is also clear through the use of “mainly”, “dive”, and “therefore” which are also evaluative, but the origin of the evaluation is not clear. Their use is contractive and gives the writer a strong authorial presence. The MA writer hides behind an authority voice to make strong claims. The authorial persona he/she creates might therefore be criticized for sounding categorical and also for being ambiguous as to whose voice is being heard.

[2] One of the advantages of the appraisal framework lies in its accessibility to non-linguists, mainly due to the use of semantics rather than grammar as the way to dive into the analysis (Martin 2000; 143). It is, therefore, suited to an analysis aimed at potentially multiple readerships, e.g. linguists, critical discourse analysts, organizational behaviorists, and leadership rhetoric scholars. (MA LR)

The study of the PhD LR sections displays a more effective authorial presence. As is shown in example [3], the PhD researcher uses both expansion and contraction to report what has been said in the literature. By referring to more than one authority, the writer uses contraction (“is”, “widespread”, “especially”, and “typically”) to give force to his/her claims. He/she, however, uses expansion to attenuate his/her stance; the use of the hedge “often” twice in the sentence strengthens the writer’s authorial persona as a researcher by opening the possibility for other voices to be heard. It increases the dialogicality of his/her work which is in line with the heteroglossic nature of academic research.

[3] The use of false identities, often of a different sex, is widespread in electronic communities and in MUDs especially (Curtis, 1996; Mantovani, 1995; Spears and Lea, 1992). This is typically the case for unsolicited emails where spammers declared a false identity to fulfill their deceptive goals. Spam writers often plagiarize other identities to sustain the intention of the reader, convince him/her and eventually trap him/her. (PhD LR)

The analysis of these examples, though only illustrative, supports the conclusions made by researchers concerning the writer’s stance and voice expression as a criterion that distinguishes writers across the degrees. Indeed, the qualitative and quantitative analysis of the corpus shows that MA researchers’ LRs reflect the image of a writer who is “less committed and critically distant” (Lancaster 2014: 51) in contrast to the PhD researchers who are more committed in their stance expression and who “build a critically discerning reader in the text” (ibid). The qualitative analysis of the corpus shows that expansion and contraction are not always judiciously used to construct an effective authorial persona that is in line with the section’s rhetorical purposes. MA students seem to have more problems in the LR as they do not always succeed in manipulating the text polyphony to their advantage. Conversely, PhD researchers display a greater mastery of the expansion and contraction resources to manage the voices within the text.

4.3. Dialogicality and authorial presence in the Discussion

Quantitative analysis (Table 5) revealed significant differences between MA and PhD researchers in the use of expansion and contraction in the RD. To start with expansion, it is more frequently used by PhD researchers. The comparison between example [4] taken from the MA RD subcorpus and example [5] taken from the PhD RD subcorpus will illustrate how MA researchers’ use of expansion does not reinforce their personal contribution to knowledge.

[4] As pointed out in the literature review (see section 2.4.2.1), Culpeper (1996) makes clear that this super-strategy has as a purpose to damage the addressee’s positive face wants; e., his/her want to be approved of. Nevertheless, following Blitvich (2010), face damage is not simply limited to the addressee, but exceeds him/her to cover the collective face of those identifying with the out-group (see section 2.4.3). (MA RD)

[5] Among the one million words studied in the corpus, 129,796 words are prepositions, thus representing almost 13 percent of all the words in the corpus. This result supports the claim held by Mindt and Weber (1989) that every eighth word in English is a preposition [Chapter One: Syntactic Approach], and therefore, they can be thought of as important categories as far as the English language is concerned. (PhD RD)

In example [4], the MA researcher refers to what has been stated in the literature to explain his/her findings. Despite the force of the utterance (expressed through the use of contraction “as”, “makes clear”, “i.e.”), the researcher’s personal voice is not foregrounded. The use of previous literature to support one’s claim is a necessary rhetorical strategy in the RD section, but it seems that in the present example, there is a predominance of other researchers’ voices. The MA writer is restating what is already explained in the LR and implicitly aligns himself/herself with other voices. Reference to the literature does not serve the discussion's rhetorical purposes. It creates ambiguity as to whose voice is being heard. The writer’s authorial persona is thus counter-productive and jeopardizes his/her position.

Conversely, the PhD researchers’ use of expansions sounds more effective. In example [5], the writer opens the dialogue through the use of hedges (e.g., “almost”, “can”) to introduce their findings. It reflects the researchers’ personal contribution and distinctiveness, on the one hand, and acknowledges the possible criticism that can be directed at them, on the other hand. The use of expansion in the RD section in that way shows that PhD writers are aware that the scientific debate is not yet over and that other voices can be heard in addition to theirs. The use of expansion lubricates the interaction between the researcher and the reader. With dialogue being the essence of academic research, it can be assumed that the PhD researchers open the interpersonal exchange as a rhetorical strategy to position themselves in the community.

5. Discussion

The qualitative and quantitative exploration of a corpus of LR and RD sections written by MA and PhD EFL researchers for the use of engagement strategies foregrounded the role of the writer’s authorial persona in the persuasive act. Voice construction and the writer’s ability to manipulate the interpersonal relationships in the text are studied to understand the writer’s construction of an authorial presence as a tool to negotiate meaning. The triangulation of data analysis tools is particularly significant since it allows for the description of recurrent trends in the corpus and provides an illustration of the writers’ differing strategies in creating a persuasive authorial persona following the section rhetorical purpose, on the one hand, and the writer’s experience in writing academic texts, on the other hand.

The quantitative analysis shows that authorial persona, positioning, and the negotiation of meaning are related to the writer’s roles and responsibilities in the academic community. The significant differences in the use of engagement across the degrees demonstrate that MA and PhD researchers do not build text dialogicality in the same way. The analysis of actual utterances from the corpus in their context shows that the stance that MA and PhD researchers create reflects different perceptions of their roles in the academic community. At the MA level, the researchers display an ability to report previous literature. More is, however, needed from an academic writer. He/she needs to be able to problematize and to give a synthesis (Nackoney et al. 2011, Paltridge & Starfield 2007, Triki 2019). Paltridge and Starfield (2007) explain the writer’s evolution as a move from knowledge reproduction with a focus on correctness at secondary school– through “summarizing and describing” (p. 7) from other sources, to knowledge creation and “recombination” (ibid) at more advanced levels. It is achieved through the “questioning” and “judging” (ibid) of existent knowledge to come up with new and original ideas. Postgraduate writing is about personal contributions and distinctiveness. This is especially true for PhD writers, as noted by Nackoney et al. (2011), who see that “Doctoral students are expected to become more critically reflective and master the complexities of scholarly writing as they approach the task of writing their dissertations” (p. 27). These maneuvers involve the writer’s stance expression towards what is being reported and the audience’s reaction to it. The present study shows that authorial voice expression, positioning vis-à-vis other voices in the text, and the foregrounding of personal contribution sound less effective in the MA in comparison to the PhD dissertations.

The comparison of engagement resources across the sections shows that the writer's authorial persona is not static. It differs following the section’s rhetorical purposes. The analysis of expansion and contraction in light of the section’s rhetorical moves reveals that MA researchers do not always succeed in detaching their voice from that of the cited authors. The PhD researchers are more proficient in managing text polyphony. In the LR, expansion resources are used to open the dialogue between the different interactants involved in the persuasive act, i.e., the writer, the audience, and other texts. Contraction is used to narrow down the discussion and specify the study perspective. It is used in a similar way by MA and PhD researchers. In the RD, the writer adopts a different persona, one which situates the work in a wider research context through the use of expansion resources and stresses personal contribution and distinctiveness through the use of contraction. Expansion is also used to acknowledge other voices in the text to foster the researcher’s humility and anticipation of the reader’s different interpretation of the results.

The orchestration of the writer’s authorial voice, evidence from previous literature, and the audience’s potential reaction to content needs to be done in a way that fosters the writer’s position so that he/she is ushered into the discourse community. The writer’s responsibility in dissertation writing is not about stating facts in a neutral way. He/she is required to position himself/herself through engaging the audience in a dialogue to achieve the text’s persuasive ends. As Street (2009) emphasizes, “So, the writer is establishing who they are as a situated subject when they present their essay/dissertation, etc. They are not just presenting data in some supposedly objective way, as many stu­dents may have been led to believe up to this stage (and beyond) […]” (p. 12). The limited use of expansion resources in the MA subcorpora does not foster negotiation in knowledge creation as is advanced by Cherry (1988). Limited intertextuality undermines research quality as interaction represents a highly appreciated feature of academic writing. The study findings can thus be used in pedagogical contexts to assist EFL postgraduate writers.

6. Pedagogical implications

Knowledge creation is the result of the superposition of the writer’s individual choices, the voices coming from previous texts, and those which are anticipated from potential audience. Likewise, dissertation writing is about reporting other voices, stating and situating one’s own, and anticipating the reader’s potential reactions. The writer’s role is to orchestrate these voices and to situate his/her own stance towards them. Because these voices can be opposite or conflicting, the academic text is said to be heteroglossic (Brooke 2014, Lee 2011, Tang 2009). Within the polyphony of academic texts, the writer’s authorial stance needs to be clear to avoid ambiguity. The role(s) and responsibility(ies) of each agent of knowledge construction are to be acknowledged in the dissertation to secure the research ethics, avoid plagiarism, and highlight personal contribution(s). According to Ivanič (1998), the writer needs “to skillfully combine some characteristics of being an established member with those of being an apprentice.” (Ivanič 1998: 296), which is not an easy balance to attain. The researcher needs to demonstrate personality and avoid being pretentious. It is like walking on a tight rope; researchers can either fall into excessive tentativeness or into arrogance. At certain stages of dissertation writing, the writer is expected to open the possibility for other voices to be heard, and at other stages, the writer needs to demonstrate certainty and strength. This balance is challenging for novice researchers, especially non-native speakers.

The study has, thus, pedagogical implications as to the understanding of academic writing at a postgraduate level. The examination of expansion and contraction resources in authentic EFL academic discourse unveils the persuasive strategies that postgraduate researchers use to negotiate a position within a specific discourse community. We argue that, in many instances, and especially for MA researchers, some usages of engagement resources are ineffective. The investigation of the researchers’ linguistic choices across the degrees confirms the peripheral status of MA writers within the community (Hyland 2006) and the more established position of PhD researchers. It also supports what researchers such as Chang (2015), Hyland (2019), and Rouissi (2013) noted as to EFL writers’ failure to construct an effective voice that serves their rhetorical moves. The comparison between MA and PhD writings proves that the use of voice is a developmental feature of academic writing. Therefore, closer attention to authorial presence and to problems of voice negotiation through the study of engagement resources can help EFL writers produce more effective texts and can enhance their chances to get heard in their discourse communities. We aspire to foreground the linguistic behavior that fosters postgraduate researchers’ positioning in the community by pointing out these tendencies.

7. Conclusion

Authorial presence plays a pivotal role in argument construction as it affects the dialogue that the writer creates throughout the text. The role of the writer’s authorial persona in the persuasive act is highlighted through the qualitative and quantitative examination of a corpus of LR and RD sections authored by MA and PhD EFL researchers for the usage of engagement strategies. To appreciate the writer’s development of an authorial presence as a tool to negotiate meaning, the writer’s ability to create interpersonal relationships through the text is studied. The study confirms the cross-degree differences between MA and PhD researchers. These areas represent zones of proximal development (Vygotsky & Cole 1978) that need to be addressed in pedagogical contexts to help EFL postgraduate writers make their texts sound more appealing. The more frequent use of dialogic expansion by PhD writers fosters their ability to synthesize, critically evaluate, and question textual content in comparison to MA writers. The cross-sectional comparison of dialogicality shows that authorial persona is a dynamic construct that is meant to respond to specific rhetorical needs and to meet generic and community expectations. It is an aspect that needs careful consideration, especially for EFL writers who are said to experience extra challenges compared to their native speakers peers. Authorial voice manifestation can be accounted for by educational and individual preferences that could be further investigated to validate the study findings.

^{1 Training sessions about academic writing organized by the Doctoral School at the Faculty of Letters and Humanities of Sfax, Tunisia.}

^{2 The terms dissertation and thesis are used interchangeably to designate MA and PhD works.}

^{3 Hood, Susan. 2004. Appraising Research: Taking a Stance in Academic Research. Ph.D dissertation. Faculty of Education, University of Technology, Sydney.}

^{4 The significance level is below p < 0.05, which means that H1 is kept.}

About the authors

Emna Fendri

Author for correspondence.
Email: emna3000@yahoo.fr
Ph.D. in English Language and Linguistics Route de l’aéroport- Km 4.5 - Sfax, Tunisia

Mounir Triki

Email: mtriki2001@yahoo.com
Full Professor at the Faculty of Letters and Humanities Route de l’aéroport- Km 4.5 - Sfax, Tunisia

References