The Impact of Attitude Markers on Enhancing Evaluation in the Review Article Genre

Academic review writing is not merely about providing an overview but taking a stance and evaluating other scholars’ views in evaluation-loaded texts. In academic review genres, the authors describe, analyse, and evaluate the developments of a research. Perhaps the most noticeable way of such an evaluation in academic review genres is the use of attitudinal lexicon and its categories. The main aim of this research is (1) to investigate what type of attitude markers are frequently used to enhance evaluation in the review articles, and (2) to analyse their functions in the different analytical sections of the review articles. The data, drawn from a randomly selected corpus of thirty-two review articles, was analysed using Wordsmith tools (Scott, 2012) to investigate how evaluation was enhanced by the use of attitudinal lexicon. The results indicated that the attitude markers were more frequent in the Conclusion section than other analytical sections in the corpus. Four types of attitude markers were identified, however only two types (i.e. attitudinal adjectives and adverbs) were the most frequent markers. They appeared more frequently in Move2 and Move3 of the Conclusion sections. This study revealed how the authors professionally communicate with their readers to clarify their evaluation through attitude markers and express importance, limitations and gaps, compare and contrast, and praise and criticize the developments of a research in applied linguistics. The findings of this research can be drawn on in EAP courses for novice writers to facilitate their achievement in academic writing.


INTRODUCTION
Evaluation is one of the distinctive rhetorical strategies in academic writing.It is noteworthy that authors strategically engage their immediate audience in their evaluation and use persuasive strategies in their academic discourses.It is also stated that "journal article authors persuade readers by deploying discourse markers to present their arguments" (Yang, 2016, p. 89); they also present themselves explicitly or implicitly to bound with readers.The authors often need to take a stance associated with their evaluative texts.They evaluate other scholars' views, their works, the methods used in their works and the results indicated in their works through employing stance metadiscoursal features including attitude markers, selfmention resources, hedges and boosters.analytical sections.In other words, we attempted to clarify how the author's stance was professionally curved in this type of genre.The attitudinal lexicon in the review article genre was specifically highlighted to investigate how the authors were involved in evaluation and how they took a stance.These elements are more personal, unlike the interactive resources, and concerned with the writers' explicit interventions to evaluate material and judgments.

REVIEW OF RELATED LITERATURE
It is worth here noting that writers deal with the judgments and evaluations in the interactional dimension.As it has been proposed, the interactional dimensions can be categorized as stance and engagement (Hyland, 2005a).Interactions are accomplished in academic writing by making choices from the interpersonal systems of stance and engagement.These two terms can be defined simply as following: (a) stance refers to the writer's textual voice.This is an attitudinal and writer-oriented function.It deals with the ways the writers present themselves and express their views and judgment (Hyland, 2008).(b) Engagement, on the other hand, is the reverse side of the interaction.This is an alignment function addressing the ways writers rhetorically recognize the presence of their readers to actively pull them along with the argument, include them as discourse participants, and guide them to interpretations (Hyland, 2008, p. 5).Hyland believes that stance and engagement are two sides of the same coin because they contribute to the interpersonal dimension of discourse.More importantly, these two resources (stance and engagement) have a dialogic purpose in the discourse and they can take up the actual or anticipated voices and positions of potential readers (Bakhtin, 1981).Hyland and Tse (2004) investigated the use of metadiscourse in L2 postgraduate students' dissertations in different disciplines.They found that the frequency of metadiscourse in soft knowledge disciplines (e.g.social sciences) is higher than the frequency of hard knowledge disciplines (e.g.natural sciences and engineering disciplines) which is the indicator that metadiscourse plays a significant role in explicating a context for interpretation.They believe that metadiscourse patterns indicate how the resources of language mediate the contexts in which they are used and reflect the knowledge domain and argument form of disciplines that create them.
In a similar vein, Hyland (2005a) has conducted a cross-disciplinary study.Based on the analysis of 240 published research articles taken from eight disciplines belonging to the soft sciences and the hard sciences, Hyland explored two key resources of academic interaction: stance and engagement.The results revealed that interactional resources tended to be more common in the soft-knowledge disciplines than the hard sciences.Hyland (2005a: 145) claimed that it is mainly because the soft-knowledge fields are typically more interpretive and less abstract than the hard sciences and that their form of arguments rely more on a dialogic engagement and a more explicit recognition of alternative voices.
In another definition, Hyland (2008) believes that stance can be referred to "the writer's textual voice or community recognized personality" (p.5).Stance can indicate the writer's attitude, evaluation and authorial voice in academic argument happening in academic discourses.As it has been contended, "stance is something of a catch-all yet elusive concept" (Crosthwaite, Cheung & Jiang, 2017, p. 3).The authorial voice can be publicised and concretised through the use of attitude markers, self-mentions, boosters, and hedges.The use and distribution of the stance features depend on the author's cultural options and disciplines.The authors intrude to present their authorial voice to the content and readers, however this identity is limited and subject to the variability of culture and disciplines (Bruce, 2016).
According to Biber (2006), the concept of stance can be presented in academic discourses through three parameters including "grammatical devices, value-laden word choice, and paralinguistic devices" (p.99).The first device is 'grammatical stance markers' including adverbials (e.g.hopefully, unfortunately, luckily) or clauses (e.g.'I am sure that…', 'I doubt that…') which are employed to indicate stance.The second device is 'evaluative words' which are totally different from grammatical devices and involve just a single proposition (e.g.I like/ dislike this food) to indicate the writer's attitude and feeling.The third device is 'paralinguistic devices' which can be classified into spoken and written language devices.The devices showing stance in spoken language can be divided into linguistic (i.e., 'pitch', 'intonation', 'stress', and 'duration') and non-linguistic devices (i.e., 'gesture', 'facial expression', 'body language').Paralinguistic devices used in written language can be devices like 'italics', 'bold typed words', 'brackets', and 'underlining'.
One of the well-known stance studies focused on pragmatic aspect of stance is Hyland's (2005b) work.He has considered stance as one of the ways that the writers achieve interaction.According to Hyland (2008, p. 7), there are three components for stance including 'evidentiality', 'affect' and 'presence'.Based on this notion, he has offered a taxonomy for stance markers including 'hedges', 'boosters', 'attitude markers', and 'selfmention'.We focused only on one feature of stance that was analyzing and identifying only attitude markers associated with the review article genre.
In another study, Jalilifar and Moazen (2014) analyzed the use of attitudinal language in 169 discussion sections of applied linguistic research articles published in ISI and non-ISI journals.They analyzed their data for the use of evaluative expressions by following Appraisal model introduced by Martin and Rose (2003).Their findings indicated that there was significant variation between ISI and non-ISI papers regarding the use of evaluative expressions.The higher use of Affect and Judgment resources in ISI journals was noticed, whereas the higher use of Appreciation in non-ISI papers.Finally, they suggested that more studies are needed to substantiate the results of their study.Adel (2006, p. 174) also defines attitude markers as "the importance of something, the interest of something, its appropriateness, and the personal emotional concomitants of linguistic material".Similarly, in another study, Koutsantoni (2004, p. 179) analyzed some pragmatic functions of attitude markers employed in electronic and electrical engineering research articles.The results of his study indicated that attitude markers were employed in the research papers to present the significant of research area, justify the authors' work, emphasize the originality of their work, indicate gaps in the developments of a research and evaluate the previous works in a related field.The current study was also in line with his research and Koutsantoni (2004, p. 179) further adds that attitude markers "create a research space for engineers, assert their learned authority and expertise, solicit readers' acceptance of claims, and reach consensus".However, such studies remain narrow in focus dealing only with attitude markers in the experimental and qualitative research articles.They have not focused on the analysis of metadiscoursal stance markers in the review article genre.
At the same time, Hyland (2005b, p. 180) state that the authors employ attitude markers in their texts to present a position and take a stance.They use these markers to make readers agree with their points of view and pull the readers into a conspiracy of agreement.The current research uses Hyland's notion and it is mainly focused on analyzing and discussing the role and function of attitude markers in the review article genre.
Thus far, previous studies have confirmed the effectiveness of metadiscoursal stance features in academic genres.They have revealed a correlation between attitude markers and enhancing evaluation and projecting the authors into their texts to professionally communicate their evaluation and credibility with their readers.Considering all of this, it seems that a great deal of research has focused on different aspects of metadiscourse in different genres including research articles (Dahl, 2004;Duenas, 2010;Harwood, 2005;Hyland, 2001Hyland, , 2005a;;Swales, 1990;), school textbooks (Crismore, 1989), and postgraduate dissertations (Hyland, 2004;Swales, 1990), however there has been scant attention to the analysis of metadiscoursal stance markers, particularly attitude markers in the review article genre.

RESEARCH QUESTIONS
Based on the justification and objectives of this research, the following questions were formulated: 1. Which attitude markers do the authors use to indicate evaluation in the review articles in applied linguistics?2. What is the overall frequency of use and distribution of the markers in the corpus?3. How do the attitude markers function in the different analytical sections of the review articles?

METHODOLOGY
The central aim of the current research was to draw on metadiscoursal stance features analysis to identify how attitude markers were employed to enhance evaluation in the review article genre.Evaluation can be highlighted and enhanced through the use of attitude markers appearing in the form of attitudinal verbs (e.g.expect, prefer, agree), adverbs (e.g.only, dramatically, essentially, unfortunately, importantly), adjectives (e.g.important, problematic, remarkable, difficult), and attitudinal nouns (e.g.importance, significance, lack, issue, limitation).They can express positivity or negativity.They can also show significance, agreement, disagreement, surprise and a few others to name.As Hyland (2005b, p. 180) believes, attitude markers make "writers both express a position and pull readers into a conspiracy of agreement so that it can often be difficult to dispute these judgments".It is important to note that the review article authors can use attitude markers as one of the persuasive strategies to indicate their evaluation and attitude towards the proposition in their texts to seek the readers' agreement.Following Hyland's (2005b) notion, these markers can fulfill several functions which can be determined for attitude markers.The functions are as follows: (1) attitude markers express the significance of the proposition, (2) they can justify the research, (3) they can judge and evaluate the researchers' works positively or negatively (i.e.praise and criticize), (4) they may indicate limitations and niche, and (5) they can emphasize the originality of the researchers' works by comparing and contrasting.In the following sub-sections, the corpus of this research is described, then the data analysis procedure is presented, and eventually the rhetorical move patterns of the review article genre are explained.

THE CORPUS
The research was conducted with a set of written academic review texts to support its objectives.One of the regulations to collect a corpus for genre-based studies is that they have to follow the same characteristics (Conrad, 2002).The first characteristics of the corpus for this research was that it consisted of review articles from one established journal which focused only on applied linguistics, namely Annual Review of Applied Linguistics (ARAL printed by Cambridge University Press).The data was drawn from a randomly selected corpus of thirty-two review articles, published between 2000-2007.ARAL is a disciplinerelated key journal in the field of applied linguistics.The prestige and reputation of this journal in publishing review articles have been taken into account.
Another criterion involved in the sampling procedure of our study is to consult the specialist informants in that field, this is known as 'informant nomination' the established tradition in sampling and targeting the corpus-based studies (Hyland, 2000(Hyland, , 2007;;Kuhi & Behnam, 2010;Sorayyaei Azar, 2012).They were asked to name the most prestigious journals with higher degree of reputation among academics in which their review articles were published.Our informants recommended us to select the review articles in applied linguistics from ARAL.They put much more emphasis on targeting ARAL in this study due to a clear understanding of review articles in applied linguistics.One of the pieces of feedback obtained was: 1) …The purpose is to further discussion about a topic or about a scholar's contribution in order to determine the current state of the art.Contributions to ARAL are typically reviews of my first kind and a good way of arriving at an understanding of what a review article is would be to read a number of such articles.

THE CORPUS TOOL
The corpus tool, Wordsmith tools version 6 (Scott, 2012) was used to identify and extract explicit attitudinal lexicon automatically by setting the scan options.The corpus tool identifies and reads plain text files which end with .txtdirectory.Wordsmith tool Version 6 extracts lists of linguistic features in n-grams using statistical measures.In order to analyze, identify, and extract these features in the corpus, the computer-readable review texts were carefully scanned and analyzed in search of the attitudinal lexicon which were previously developed in Hyland's (2000) study.

DATA ANALYSIS PROCEDURE
The study on the use of stance features, attitude markers in particular, based on Hyland's (2005b) model was observed.The analysis of attitude markers in review articles was conducted in the subsequent steps accordingly.First, 32 review articles in the corpus of this research were analyzed using Wordsmith tools.Second, a list of 85 attitude markers was selected and developed based on previous works and lists in literature especially Hyland (2000Hyland ( , 2008)).The choice of attitude markers in the current study matches Hyland's list.The focus of this stage of the study was on the investigation of explicit attitudinal lexicon used in the corpus of this study.Next, a very careful investigation of the context of attitude markers was manually carried out to make sure that they really stand for the attitudinal feature.At this stage of analysis, the main step was focusing on form, frequency, type, and function of these markers.In other words, to get clearer understanding of their functions, a rigorous analysis of the context was emphasized.
Several important points and steps were taken into consideration at this stage of analysis.First, during the analysis several irrelevant cases were found by the concordance so they were deleted from the results when the items were double checked by the researchers (e.g.'Major' was found in the review texts as a proper name, or 'New' was found in the review texts standing for the first word of 'New York' as a proper name).
Second, all evaluative items in integral and non-integral citations carrying and denoting other writers' ideas were deleted from the results.Third, we normalized the frequency counts at 1,000 words and it was applied for the entire of this study.Finally, to get higher reliability in the findings of the current study, not only the researchers' analysis but also the second-rater's analysis who was a Ph.D. holder in applied linguistics and possessed extensive experience and expertise in doing genre analysis on academic discourses was applied.The four analytical sections of review articles were read word by word to make sure that the features really stood for evaluative markers.The items were double checked by the second rater.The inter-rater reliability was found to be above 85%.In spite of some differences in identification of the attitude markers and their roles, this percentage suggests high overall reliability in this research.
In this section, the rhetorical move patterns of the analytical sections in the review article genre were explained.The coding criteria proposed were used for analyzing the rhetorical move structures employed in the analytical sections of the corpus in the present research.There are four moves for the Abstract section in the review articles in applied linguistics namely 'situating the review' (i.e.laying out the situation for the current review), 'presenting an issue/ indicating a gap' (i.e.identifying a gap/ an issue/ a need for further research/ or posing a question in the field of the review), 'presenting the review' (i.e.stating the author's main purpose and view, and presenting the structure of the review), and 'discussing the review' (i.e.interpreting the discussions/ or giving suggestions, recommendations, and implications of the study).This coding system for the Abstract section of review articles fits partially Santos' (1996) model, with two moves missing namely 'describing methodologies' and 'summarizing the results'.These two moves do not usually appear in the review article genre.
There are usually three moves for the Introduction section in the review articles in applied linguistics namely 'establishing background' (i.e.setting the scene for the current review to show the centrality and significance of an Issue/ or an idea/ or a theme), 'narrowing down the scope of the review' (i.e.posing a question, identifying an idea/ or an issue, indicating a gap/ or a need for further research), and 'presenting the review' (i.e.establishing the author's position by restating the issue or idea to emphasize the value system and stating the purpose of the review, and outlining the relevant areas of the field being reviewed).This coding system for the Introduction section of review articles fits Swales' (1990) CARS model.
The moves and move elements in the thematic units often appear in more complex and cyclical structures than linear ones.Thus, it was found that the thematic units of the Body section were very complex and recursive.They were full of cyclical structures.Due to these reasons, the accurate frequency of the strategies in the thematic units was not counted.It should also be added that the two-layer analysis of moves and steps appears to be inadequate because of the complex and cyclical structures in theme-bound units of the Body section.
There are three moves for the Conclusion section in the review article genre namely 'summarizing the review' (i.e.summarizing the purpose of the review and the contentious issue/ or the author's position), 'evaluating the review' (i.e.presenting the strength/ or weak points of the review), and 'giving suggestions' (i.e.giving suggestions to solve the identified issues, to do further research, or to draw pedagogical implications).This coding system for the Conclusion section of review articles in applied linguistics fits Yang and Allison's (2003) model.

RESULTS AND DISCUSSION
In the following sections, the results of investigating attitude markers are presented.The first section discusses the frequency of use and distribution of attitude markers in the corpus.The second section highlights how these markers enhance evaluation and fulfil their different functions in the four prevalent analytical sections of the review articles (i.e.Abstract, Introduction, Body, and Conclusion sections).

DISTRIBUTION OF ATTITUDE MARKERS IN THE CORPUS
The analysis showed that attitude markers appeared in the four analytical sections with different frequency.For example, we identified 7.18 attitude markers per 1,000 words in the Introduction sections which was less frequent than the attitude markers in the Conclusion sections (9.75 attitude markers per 1,000).Table 1 shows the frequency of attitude markers in the four different analytical sections and their occurrences per 1,000 words in the corpus.The findings revealed two important points.The first point is that most of the attitude markers used in the four analytical sections were attitudinal adjectives (almost 69% of the total attitude markers in this research).In contrast, the rest of the attitudinal markers such as adverbs, verbs, and nouns were scarce.The most frequent attitude markers in the four analytical sections, for example, were 'only', 'important', and 'issue'.
The published research in applied linguistics included a high number of explicit attitudinal lexicon in research articles.Other researchers' results strategically indicated that authors' voice was clearly visible in their works and researches.This voice can be identified by stance features particularly attitude markers and self-mentions.These attitude markers were found to express the authors' judgment and attitude towards the proposition they discussed or argued.
In consensus with other researchers' findings regarding the analysis of attitude markers in academic discourses (Duenas, 2010;Hyland, 2005b;Koutsantoni, 2004;Swales & Burke, 2001, Yang, 2016), explicit adjectives were found to be more frequent than any other attitudinal lexicon such as adverbs, nouns, and verbs in the corpus.It was also found that attitudinal adverbs (such as 'only', 'necessarily', 'significantly', 'completely', 'essentially, 'appropriately', and 'unfortunately') were used more frequently than verbs in the review articles in applied linguistics, very closely followed by nouns.Attitude verbs were the least frequent attitudinal lexicon in this study.Adverbs and almost nouns were the second and third most frequent attitude markers in the entire corpus of this study respectively.
All in all, attitudinal lexicon such as adjectives, adverbs, nouns, and verbs were considered as evaluative items expressing value, significance and importance, indicating a need or a lack of sources, expressing evaluation, showing emotion, and indicating weaknesses and shortcomings.
The second point is that the attitude markers in the Introduction sections appeared with the frequency of 7.18 items per 1,000 words, while they were found in the Conclusion sections with the frequency of 9.75 items per 1,000 words (see Table1).Although the number of words in the Introduction sections (11983 words) were more than the Conclusion sections (9947 words), the attitude markers in the Conclusion sections (97 hits) appeared more frequently than the Introduction sections (86 hits).One main reason for the different frequency may be referred to the authors' evaluation moves and move elements being involved in the Conclusion sections.In consensus with other researchers' (Crosthwaite, Cheung & Jiang, 2017;Yang, 2016) findings regarding the analysis of attitude markers in academic discourses, attitude markers were found to be more frequent in the Discussion and Conclusion sections where the authors intrude into the argument, take a stance and express their attitude towards the findings and views.The attitude markers were used more frequently in these sections than other analytical sections.In the following section, various types of the review article genre were explained.

DIFFERENT TYPES OF THE REVIEW ARTICLE GENRE
As it was discussed somewhere else (Sorayyaei Azar & Azirah, 2014), the review articles in the field of applied linguistics can be classified into three types namely: (1) critical evaluative review (i.e. its focus is on critical evaluation of published work and it encompasses the subject-oriented approach), (2) bibliographic review (i.e. it gives readers a comprehensive descriptive record of annual works in that field and it encompasses the literature-oriented approach), and (3) mixed-mode review (i.e. it has the twin roles).Almost fifty percent (47%) of the review articles in this research involved mixed-mode reviews and the rest of the reviews belonged to both classes, that is to say, the rate of variability for the two classes was very close (28% and 25% critical and bibliographic reviews respectively).The findings clarified the functionality variation among review articles whether they functioned as critical evaluative or bibliographic reviews.
The critical evaluative review articles were written with the authors' critical and evaluative perspectives and its value relied on the writers' expertise.The attitude markers were employed in this review genre not only to reflect the writers' own position and seniority, but also to show the flow of their evaluation.At the same time, the writers created a room to interact with their readers, indicate a gap, evaluate the developments of a research, and seek their agreement.Therefore, it is necessary for the novice writers and junior researchers to master this type of genre due to its significant analysis and evaluation conducted by the scholar.The critical evaluative review article usually serves four main purposes, they intend to: (1) become subject-oriented, (2) survey the selected works and to raise an issue in the related field, (3) announce the strong point of view in order to evaluate the opponent's position and opinions in the related field, and (4) suggest a new direction or trend.The intended audiences of this type of review were the researchers, experts, and parent members of the discourse community in the field.
The bibliographic review article usually provides an overview of the research developments and intend to: (1) become literature-oriented, (2) raise awareness of researchers and graduate students in their own present activity and related field, (3) inform the missing work in that field, and (4) give an overview of the subject.
The mixed-mode review article usually attempted to play twin roles and written by a solicited scholar in the field.It encompassed both literature-oriented and subject-oriented approaches.The writers attempted to provide an overview of the related field and then evaluated the developments.It was found a degree of intensity between narrative text types and critical evaluative text types in the mixed-mode review articles.The authors usually intended to present an overview of recent developments and then they discussed the recent issues in that field, or vice versa.

HOW ATTITUDE MARKERS FUNCTION DIFFERENTLY IN THE ANALYTICAL SECTIONS ABSTRACT SECTION IN THE REVIEW ARTICLE GENRE
As discussed before, the Abstract section is one of the prevalent sections in the review article (i.e. this section often includes four Moves 'situating the review', 'presenting an issue/ indicating a gap', 'presenting the review', and 'discussing the review').
It is important to note that evaluative items in the Abstract sections are not rare.The findings indicated that attitude markers were frequently used in the Abstract sections particularly in Move 1 ('situating the review').In other words, the authors used attitudinal lexicon as evaluative triggers in Move1 of the Abstract sections to situate the review, provide background knowledge on previous research, and define the idea or theme of the review Examples of attitudinal lexicon in the Abstract sections taken from the corpus are shown as follows: 1) Over four decades ago the so-called Chomskyan revolution appeared to lay the foundation for a promising new partnership between linguistics and psychology.(RevA.5,Abstract, Critical Evaluative Review Article) 2) The omnipresence of English in Europe has led to numerous discussions about its widespread functions and special status compared to all other European languages.(RevA.25,Abstract, Critical Evaluative Review Article) 3) A few studies have continued theoretical threads from previous work…we have also seen the beginnings of promising new lines of research…(RevA.8,Abstract, Bibliographic Review Article) 4) …Neurobiology and psychology have become more closely integrated in recent years as evidenced by the emergence and development of such disciplinary interfaces as biopsychology and…is widely accepted… (RevA.6,Abstract, Mixed-mode Review Article) 5) …a critical overview of the issues and research conducted since the most recent state-of-the-art article published in ARAL… (RevA.28,Abstract, Critical Evaluative Review Article) Move 1 was used to lay out the situation for the review texts either by providing the background on previous research or elaborating the idea or theme.This move can also be recognized through clusters of evaluative items in the review texts.Attitude markers were also frequently used in Move 2 of the Abstract sections ('presenting an issue/ indicating a gap').The attitudinal lexicon such as adjectives, adverbs, verbs, and nouns (e.g.'limited', 'critical', 'overwhelming', 'many', 'few', 'new', 'current', 'unfortunately', 'only', 'thoroughly', 'difficult', 'neglect', 'issues') were identified in Move 2 of the Abstract sections indicating a gap or raising an issue.
1) Many have now concluded, however, that the hopes originally expressed for this partnership were not realized.(RevA.5,Abstract, Critical Evaluative Review Article) 2) Unfortunately, all studies focus only on English discourse, which suggests that analyses of discourse in other languages are clearly needed; moreover, it was extremely difficult to find published discourse analytic studies which employ only quantitative research methods.(RevA.10,Abstract, Mixed-mode Review Article) 3) There is empirical support for each of these implications, although at the same time, additional research related to many is needed to further identify… (RevA.19,Abstract, Mixed-mode Review Article) 4) After years of neglect, political theorists in the last few years have started to take an interest in issues of language policy, and to explore the normative issues they raise.(RevA.13,Abstract, Mixed-mode Review Article) 5) Although there are many studies on the new international norms…, there are limited discussions… (RevA.26,Abstract, Critical Evaluative Review Article) 3) To make our task manageable, we have limited ourselves to looking at published work _specifically, refereed journal articles, book chapters, and books; we have not addressed here scholarship in the form of dissertations, ERIC…We have attempted to be comprehensive, but inevitably things get left out: We apologize in advance for any omissions.(RevA.20,Introduction, p.71) 4) In a chapter of this size it is impossible to cover all facets of corpus linguistics…I have neglected much work with cross-linguistic comparisons… (RevA.12,theme-bound unit, p.85) As shown in the excerpts (2, 3 and 4), one of the strategies to save the face and tone down the negativity of the criticism is to use the praise-criticism strategy.This strategy was drawn from Hyland's (2000: 55) investigation in book reviews and tried to pair praise and criticism.First, the writer often praised the work, and second he/she highlighted the limitations of the work or vice versa.This strategy was employed by the writers to claim that they were aware of the gap or weakness so that this strategy could soften the threat of a critical comment in the review article genre.In the first excerpt, the writer also emphasized his/ her uncertainty about the inference of methodological preoccupations.

THEME-BOUND UNITS IN THE BODY SECTION
The third analytical section was Theme-bound units in the Body section.The number of these theme-bound units was not predictable.Due to highly cyclical and recursive moves it was decided to focus on argumentation across the thematic units (i.e.how argumentative patterns were formulated and shaped in these thematic units of the Body section. In the Body sections it was interesting to notice that adjectives occurred more frequently than other attitudinal lexicon such as adverbs, nouns, and verbs because the authors attempted to employ these evaluative markers to express their value, attitude, and judgment.Like the Introduction and Conclusion sections, adverbs (e.g.'interestingly', 'obviously', 'significantly', 'surprisingly', 'unfortunately', and 'usefully') were the second most frequent attitude markers and verbs (e.g.'agree', 'prefer', and 'wish') are less frequent than attitudinal adjectives and adverbs in the Body sections.Mainly adjectives and attitudinal verbs, adverbs and nouns were considered as evaluative items expressing value, significance and importance, showing critique and praise, expressing evaluation, and referring to strength and weakness.Examples of attitudinal lexicon in the thematic units of the Body section taken from the corpus are shown as follows: 1) The description that corpus analysis provides, then, is necessarily partial in that it privileges one aspect of language.It deals with the attested message forms, but not with the possible or potential of the code from which they are drawn…Now, it is important to stress that this view does not deny the validity of such text description, but only to recognize that…its validity is bound to be limited.(RevA.1,themebound unit 2, p. 24) 2) On the other hand, it has very few points of contact with SLA.Indeed, in this context it is probably not a chance event that last year's ARAL 19 had an opening section…two intellectual worlds thus continue to be socially constructed poles apart…However, if these kinds of field-imposed restrictions are a cause of regret to the LSP movement, it is also true that LSP has been insufficiently concerned with… (RevA.3,theme-bound unit 2, p. 61) As shown in the examples (examples 1 and 2), one of the strategies to tone down criticism is to pair it with praise (Hyland & Diani, 2009).In the first example, the author tried to evaluate the corpus analysis and referred to its incomplete descriptions about only one aspect of language, however at the same time he emphasized its validity.Or in the second example, the author referred to one of the weaknesses of LSP which had fallen apart from SLA.However, at the same time the author emphasized that although field-imposed restrictions could be considered as a sort of regret to the LSP movement, LSP had always involved students' learning matters.In these excerpts, we could see that the authors employed these strategies to evaluate others' works and views by criticizing and praising, representing themselves and acknowledging alternative views.
Thus, it was found that the authors in the theme-bound units have often used the praise-criticism strategy to soften the threat of a critical comment (Sorayyaei Azar & Azirah, 2017b).The authors attempted professionally to position their views by applying these evaluative strategies in their recursive and complicated review article genre.They, in fact, tried to take advantage of evaluative items such as opposing adjectives and noun phrases to pair criticism with praise (e.g.'interesting methodology' against 'mistaken assumption'; 'success' against 'fail'; 'interesting story' against 'too complex').Some more examples taken from the corpus are as follows: 3) It is important to note, however, that ESP is not the sole issue in LSP concerns in North Africa.Issues involving other languages also need to be reviewed… (RevA.4,theme-bound unit 1, p.79) 4) The story of how psychologists attempted to test this idea is interesting, but too long and complex to relate here; the important point is that after some early successes, most psychologists failed to be convinced… (RevA.5, theme-bound unit 1, p. 5) 5) Although this research perspective is still relatively new, during the past decade it has been adopted by a growing number of scholars…partly driven by recognition that, by accounting for the dynamic evolution of motivation… (RevA.7,theme-bound unit 1, p.46) 6) A small but increasing amount of CA and CA-informed research on talk in educational institutions directly addresses issues of interest to applied linguists.(RevA.9,theme-bound unit 3, p. 14) 7) This problem seems unsolvable within the LHR framework.Its attraction_ namely, that its standards apply universally to all individuals regardless of history, numbers, or nationhood_ is also its weakness.
The only sorts of language rights that can be defined in this universal way are minimal rights…In this way, LHR are insufficient to ensure linguistic justice…In the end, both the benign neglect approach and the LHR approach suffer from the flow of attempting to avoid the unavoidable… (RevA.13,themebound unit 3, p. 11) 8) Over the years there have been various empirical attempts to demonstrate the influence of L1 concepts…one study especially interesting because of its methodology (Ijaz, 1986), considered cases…Ijaz assumed that meaning transfer was the same as conceptual transfer, and as the discussion in the preceding section suggests, such an assumption is mistaken.Furthermore, the cloze tests Ijaz used cannot really show… (RevA.21,theme-bound unit 2, p.9) The rhetorical move pattern for the Conclusion section and how the attitude markers enhanced evaluation in the Conclusion section of the review articles are explained in the following section.

CONCLUSION SECTION IN THE REVIEW ARTICLE GENRE
The last analytical section was the Conclusion section (i.e. it consists of three moves including 'summarizing the review', 'evaluating the review', and 'giving suggestions').The rhetorical move pattern in the Conclusion section was similar to Yang and Allison's (2003) model.Their model is able to explain the move structure identified in this study.Although the findings of this study conforms to Yang and Allison's model to the actual practice of concluding in the review article genre, there was a variation in the rhetorical move pattern for the Conclusion section in the three types of review articles (for more details see Sorayyaei Azar & Azirah, 2017a).
Move 2 in the Conclusion section dealt with the evaluation of the research or other scholars' view.The authors evaluated the significance and limitations of the developments in a research by employing attitude markers.These markers conveyed not only positivity (such as praising the works or views) but also negativity of the propositions (such as criticizing the works and views) in the text.
Move 2 exemplified various evaluative items employed by the authors.The authors used attitudinal lexicon to indicate the significance of the developments in the related field (e.g.'new advances ', 'lively area', 'recent developments', 'growing', and 'advancing').
Apart from that, the authors could present the limitations of the developments in the research (e.g.'pressing problems', 'major challenges', 'unable to do full justice ', 'disappointing', 'limited', 'unfortunately', 'issues', 'criticizing', and 'difficulty').Some examples dealt with the evaluation in the Conclusion sections (Move2) are as follows: 1) …not succeeded in developing a unified view of language that is broadly accepted by researchers…There are signs, however, that future prospects for collaboration are brighter.Specifically, new advances in the brain sciences… (RevA.5,Summary and conclusion, p.16) 2) …ESP/ LSP has played its full part in the emergence of Applied linguistics as a discipline…I have not been able to do full justice to the lively area that usually goes by the name of Contrastive Rhetoric…some other recent developments.The four types of attitude markers like adjectives, adverbs, nouns, and verbs were found to express the authors' views, judgment, evaluation, and attitude towards their propositional content and views.The attitudinal lexicon explicitly clarified the authors' view and perspective of a particular idea.They refer to the authors' "affective attitude" towards a certain matter (Hyland, 2008).
Attitude items fulfill various functions expressing the views positively or negatively, indicating a niche or an issue in the review texts.Their functions can be presented in various types (see Appendix 'A' for more details).The findings showed that the first highest hit referred to 'only' and the second highest referred to 'important'.Overall, there were 1520 hits/ 7.66 attitude markers in 1,000 words in the entire of the corpus.Appendix B summarizes the frequency of use for the different attitudinal lexicon in the review articles.It shows the total frequency of attitude markers (1520) and their occurrences per 1,000 words in the entire of the corpus.It indicates that the most frequent attitude markers in review articles were attitudinal adjectives (69%) and the least frequent attitude markers were attitudinal verbs.Appendix B represents 51 cases of those markers used in the current research.
The results of this research are in line with Hyland's (1999bHyland's ( , 2005b) ) results.Similarly, Hyland's study and the current study, analyzed the attitude markers in the whole research articles and review articles to indicate the authors' evaluative strategies of the developments in the field of 'applied linguistics'.However, Hyland investigated stance features (i.e.attitude markers, hedges, boosters, self-mention markers) in applied linguistic research articles.Although, the occurrence of attitude markers was investigated in the four analytical sections of review articles, the main focus of the current study was on types, frequency, and functions of these markers in the corpus.He identified 8.6 attitude markers per 1,000 words were used in the research articles.Unlike this study (7.66 per 1,000 words), the attitude markers in his investigation were more frequently used.Hyland (1999Hyland ( , 2005b) ) did not mention the number of attitude markers that he had employed in his research, whereas the number of attitude markers investigated in this study were 85 attitude markers.Moreover, he analyzed the four stance features in 30 applied linguistic research articles, whereas in this study only one stance feature (i.e.attitude markers) was investigated in 32 applied linguistic review articles.In other words, while the number of words were not specified and mentioned in Hyland's work, the number of words in the entire corpus of this study were 198426 overall words.
Similarly, Crosthwaite, Cheung & Jiang's (2017, p. 30) research revealed that attitude markers were primarily employed in the Discussion section of professional dentistry research reports and dentistry student reports where the writers express their attitude towards their claims and results.At the same time, they also discussed that attitude markers were relatively rarely used in other prevalent sections of the research reports; the professional writers produced more attitude markers in the discussion sections than the dentistry students.
In short, it can be contended that the review article genre creates the room for the authors to interact with their readers.The authors try to communicate with their readers to take a stance and emphasize their evaluation of the developments of a research and raise an issue or indicate a gap, argue ideas, and give suggestions and offer possible solutions.The given functions for this type of genre often clarify that evaluation is one of the essential roles for the review article genre and we can claim that the review article genre acts as an evolutionary genre.

CONCLUSION
In the current study, the findings indicated that the authors used attitude markers to enhance evaluation in the different analytical sections of the review texts with different frequency.It indicated certain attitude markers like 'only' and 'important' were the most frequent markers in the review articles but with different frequency.Interestingly, attitude markers were more common in the Conclusion sections than other sections.They appeared more frequently in Move 2 and Move 3 of the Conclusion sections namely 'evaluating the review' and 'giving suggestions'.The authors explicitly presented themselves, expressed their attitude towards their own claims and results, and evaluated other scholar's views.Moreover, they presented the significance and limitations of the works and developments in a research.They also gave suggestions for further research and offered possible solutions for the raised issues.It was noticed that the authors employed a cluster of attitude markers along with inclusive pronouns in the review texts to interact with the readers and seek their agreement.The attitudinal lexicon was more frequent in these moves and strategies of the Conclusion sections.Four types of attitude markers (i.e.adjectives, adverbs, nouns, and verbs) were identified, however only two types of attitudinal lexicon including attitudinal adjectives and adverbs were the first and second most frequent attitude markers, respectively.Attitudinal nouns and verbs were the least attitudinal lexicon in the review articles.
It is noticed that the authors often argue ideas and leave an authorial footprint of argumentation in their review genre.The authors usually apply different evaluative strategies to present their own viewpoints, evaluation, and attitude towards the propositional content and others' claim.The attitude markers often interact with the readers as agents that they can detect the writers' footmarks in their review texts.These footprints are alike the writers' ideas and attitude towards the propositions they argue.The authors use attitude markers not only to reflect their own position and seniority, but also to show the flow of their evaluation and argument.
Consequently, it was illustrated that how the authors in the corpus of this research used the attitude markers to present their evaluation, interact with their readers, and take their stance in the field.The chain of attitudinal markers in the review article genre was professionally used to identify ideas, indicate gaps, awaken and encourage the discourse community members to think about these issues in the development of the research.Attitudinal lexicon as one of the most interesting elements of stance clarifies the authors' judgment and evaluation in the interactional dimension in the review article genre.
While it is acknowledged that English for Academic Purposes (EAP) course offered in the university levels have focused on academic writing skills, the current research suggests that writing academic review genres can be considered an area of emphasis to provide the instructive guidance for EFL postgraduates and novice writers how to review the research developments critically and get them ready to write the review of literature sections efficiently.Syllabus designers need to provide and prepare the related material and textbooks (e.g.Swales & Feak, 2012) to be presented in EAP classes.
Moreover, the lecturers in these classes should inform the postgraduate students about the norms and conventions of the discourse community such as the rhetorical move structures and the textual strategies the authors employed to indicate significance, present evaluation, compare and contrast views, show limitations of the developments, and pair critique with praise in comparing and contrasting strategies.The syllabus designers in their materials and lecturers in their EAP classrooms need to instruct and practice with EFL postgraduates about these rhetorical structures and the textual strategies employed in the academic review genre.They can also inform them about the variation, cyclical move elements, and recursion in the academic review genres.These instructions can be taught the postgraduates or practitioners when they read the review texts, they may encounter such textual strategies and variation in the review article genre.This type of instruction can raise the postgraduates' awareness about the different kinds of rhetorical strategies in the review article genre.Now, it is vital to acknowledge limitations to our study.It would have been better if a different discipline had been chosen to do a comparative study.The review articles in this corpus (published between 2000-2007 due to being a part of PhD program at University of Malaya) are all from applied linguistics (not from non-applied areas), therefore the findings of this research also need to be tested on recent review articles from other disciplines particularly hard sciences so that EAP or ESP instructors can take advantage and they may use findings in their 'Seminar' and 'Research Writing' classes for hard science postgraduate students or the practitioners.In terms of analyzing stance features, this research focused only on the frequency, type and functions of one of the stance features.Further research is needed to focus on the functions of all stance features (i.e.hedges, boosters, attitude markers, and self-mentions) in review articles within one discipline or multi-disciplines to have a comparative analysis.

TABLE 1 .
Overall distribution of attitude markers in the analytical sections One is the tremendous interest in corpus linguistics and its great potential for LSP work…Overall, we can see that LSP has a number of structural problems such as weaknesses…uncertain provision of professional training.(RevA.3,Final Considerations, p.67) 3) …there is a significant amount of practitioner knowledge built up in programs and classrooms around the world in support of specific instructional approaches…In many cases, this knowledge works well and supports students' reading development.(RevA.19,Conclusions, p.59) 4) This research will be useful in further determining where such learners' strengths and weaknesses lie…it will take time and effort for SLA researchers to study illiterate adults.It may be challenging for some SLA researchers…there are issues of access to illiterate adult learners (RevA.24,Conclusion, p.89) 5) …create significant challenges for applied linguistics.There is still a bias in the field…the ELF research discussed above need to be taken seriously… (RevA.26,Conclusion, p.211) 6) However, for early FL programs to be useful, certain conditions must be met: (1) learners need to have positive attitudes… (RevA.28,Conclusions and Implications, p.251) Table 2 classifies attitude markers into different types of function:

TABLE 2 .
Four functions of attitude markers