Abstract

The international EFL teaching community has a variety of communication needs including the need to share unique experience with fellow professionals internationally. This paper will examine the last (March 2008) quarterly issue of AEJ to examine the extent to which our editorial process allows a first-person voice in accepted papers. This is intended to shed light indirectly on the way academic discourse communities may contribute towards suppressing unique individual and cultural voices

Key words: writing for academic journals, Generic Conventions

Authors are encouraged to conform with international standards of drafting, but every effort will be made to respect original personal and cultural voices and different rhetorical styles. (Asian EFL Journal (AEJ) Submission Guidelines)

   In this paper, I will attempt to defamiliarize the nature of writing for academic journals. Applied linguistics is a very broad area, and ELT in particular is an international practice which encompasses a culturally-diverse international community and a very broad range of individuals who engage in it. While we might all have become familiar with practices that allow us to participate as members of academic discourse communities, I will be asking whether we might sometimes be in danger of creating faceless, impersonal communities that suppress valuable, original voices.

Within academic discourse communities, what constitutes competent practice is always to some extent negotiable. As Eckert and Wenger (2005, p. 583) suggest, "What counts as competence and by whom is something that the community negotiates over time; indeed, it is this negotiation that defines the community." We might add that the lack or even the suppression of such dynamic negotiation would lead to a very static, rigid definition of competence which would run counter to many academics' view of academic inquiry.

In Nunn and Adamson (2007) voice was broadly defined as the means by which authors express their identity as scholars. In this paper, I will focus on just one much narrower aspect of 'voice': the use of first-person pronouns in relation to transitivity, community-pressure and genre. After briefly examining the way two expert philosophers use the first-person, I will contrast two brief examples from AEJ reviews, before analyzing the use of the first-person in one complete quarterly issue of AEJ. (March 2008 - see appendix.) My analysis will focus on the use of the first-person in both single- and joint-authored papers.

‘Successful’ and ‘Unsuccessful’ Academic Voices

Experts Have a Voice

A good example of the extent to which a first-person voice can be intimately integrated into theoretical thought processes appears in the preface to Tractatus Logico-Philosophicus (p. 4). I am citing a philosophical text on logic because I believe it can help challenge the notion that logical argumentation needs to adopt an impersonal voice. In this example, Wittgenstein appears to use the first person to commit himself totally and personally to the truth value of the ideas expressed, labelling them “unassailable and definitive”. 
        [T]he ‘truth’ of the thoughts that are here communicated seems to me unassailable and definitive. I therefore believe myself to have found, on all essential points, the final solution of the problems. 

   However, the use of the modals “seems to be” and “believe myself to have found”, and “on all essential points”, linked to the more vulnerable first-person (I therefore believe myself) considerably weakens his profession of absolute confidence. The whole statement can then be read as an interesting paradox between an immodest desire to make a very strong claim and a paradoxical desire to evaluate his own personal commitment to the ‘unassailable’ truth of his statement rather more modestly. Wittgenstein (p. 4) goes on to confirm the possibility of being mistaken in his personal belief in a conditional clause.

And if I am not mistaken in this belief, then the second thing on which the  value of this work consists is that it shows how little is achieved when these problems are solved.   

It is also interesting to note that what can, even temporarily, be believed with a high level of personal commitment, is then almost dismissed by Wittgenstein as insignificant. I believe that this kind of code switching between a first-person (If I am not mistaken) and impersonal passives (how little is achieved, these problems are solved) is part of what Wittgenstein might call the ‘game’ of engaging in academic discourse and code-switching is an important skill in this language game.
Similarly, in his introduction to the same work (p.xxv) Russell uses the first-person freely to take up a very personal, but no less reasonable, position towards logical and theoretical investigation:

         As one of long experience of the difficulties of logic and the deceptiveness of theories which seem irrefutable, I find myself unable to be sure of the rightness of a theory, merely on the ground that I cannot see any point on which it is wrong.

AEJ Examples

In our initial study on academic voice, (Nunn and Adamson, 2007), we discussed several successful and unsuccessful examples within the review system of AEJ. I will use two of these here as background to this current study.   In the first extract from a published PhD thesis, similar to the Wittgenstein examples already cited, there is code-switching between formal academic style, "PhD theses are not created in isolation", and the first-person voice of the author, "thanks to myself".  This extract was taken from the acknowledgement section of a successful PhD thesis published in our thesis section and is cited here because it directly addresses the notion of the author’s voice. More first-person uses are presented in Nunn and Adamson (2007) from other parts of the thesis.

Review Sample 1 ‘Successful’ example (Reinders, 2006, in Nunn and Adamson, 2007, pp.221-222)
Acknowledgements can be a bit boring. Mostly it’s the usual suspects: family, friends, children, one’s pet goldfish. I have yet to come across one that says ‘Thanks to MYSELF for being such a great person and hard worker’. Probably that is because PhD theses are not created in isolation. And even if one did do all the work by yourself, somewhere, somehow, the thing needs to relate to someone to have any value at all. My list of acknowledgements could be as long as this thesis. I would have never had the qualifications to embark on a PhD study, nor been at this particular University, and certainly not been able to complete my thesis, without the help of many good souls.

Review Sample 2‘Unsuccessful’ Example (From Nunn and Adamson, 2007, pp.221-222 – anonymous sample)

Whenever I asked them to say something about one topic, they, for the most part, would say, “How should I say it? What should I say?” and they would ask these questions in their native Thai.  I was determined to improve my students’ speaking ability and tried various techniques in order to do so.  But the results were not ideal.  At that time I thought maybe the teaching materials were not relevant to them, thus they did not want to say anything or could only say a little.

   This sample is an extract from a paper that received mixed reviews and was eventually rejected, although not for its first-person voice. Two review comments are cited below. The contradiction between them helps to explain why the first-person voice is significant:

*Review 1 comments: Well written paper. It is very easy to follow the author’s argument and s/he uses a very personal tone. (Major revisions requested based on methodological grounds.)
*Review 2 comments: The style of writing is not appropriate for an academic article. (Rejection was recommended based on language.)   

The second review comment indicates how a personal voice can at least contribute to the rejection of a paper. Examples like this one led to the conclusion that this aspect of academic writing in the journal needed more detailed attention given the journal’s policy of attempting to encourage alternative cultural and academic voices

Transitivity, Community-Pressure and Genre
Carter (1995, pp. 52-59) argues that 'genre' is a controversial topic, in particular in pedagogical contexts. "Holding up certain genres as models to a whole class is seen as rigid and deterministic" (p. 57) and therefore counters pedagogical goals of most writing courses. "A major concern is that genre-based writing practices can be inherently conservative and are designed to produce unreflective writers who will be able to do no more than sustain the genres…" (p. 55) This view is linked to what Carter calls "narrow vocationalism" and to a traditional transmission style of teaching. (See Nunn and Adamson, 2007 for a fuller discussion.) Texts then tend to be used as models that merely reproduce the structures in place and which lack the dynamism that is a characteristic of high-quality writing. Here I should point out that Wennerstrom (2003, p. 34) has also cited Carter (1996) to support a very different view that genre education and knowledge empowers "through learning to control and adapt a variety of conventional formats for written communication and on understanding the social contexts in which these conventions exist."

   Thompson (2004, p. 230) appears to suggest that the interpersonal function is de-emphasized in at least the more formal varieties of academic discourse. "One reason why nominalization is in harmony with the ideology of science, and of academic, formal writing in general, is that it makes it easy for processes to be objectified – to be expressed without the human doer." Suggesting that processes are 'objectified' is not the same as claiming 'objectivity'. Indeed, there might even be an implication that it is only the expression of the process that is 'objectified'. The question then is to determine whether this is just a linguistic ploy related to genre to disguise potentially subjective intrusions or whether the discourse can actually be characterized as 'objective'.

Generic Conventions and Bio Data
In AEJ all papers are preceded by bio data. An interesting point about bio data is that authors often send it in the first-person, but it is edited and published in the third person. This seems to be common practice and is certainly AEJ policy and one that I supervise and practise myself. This practice is of interest to this argument, because it is a genre-related practice that does not stand up to scrutiny. Authors write their own bio data, but it is presented as if someone else wrote it. The practice arguably gives more formal status to the author, but is also used to promote the journal. A refereed journal is apparently highlighting the achievements of its authors, but these are subjectively selected by the authors themselves. We would not, for example, state that author X previously submitted three papers to the journal which were rejected, and one that was rejected on screening and was therefore not considered suitable to be reviewed at all. We never state that a paper has been through two sets of major revisions: a fact that must have an impact on authorship after four reviewers have had some impact both on the contents and on the writing style. Most authors do not take us up on our offer to disagree with review comments (with their reasons), possibly assuming that if the reviewer wants the change, they would be wiser to make it whether they agree or not in order to get published. Some, but not all of these practices are there to protect an author's privacy.

Transitivity as a System of Choices

From a systemic perspective, authors have choices and the choice they make is partly about how to represent their voice whether the paper is describing a teaching approach, reporting a research process, rationalizing experience, expressing a theoretical position, summarizing the literature on a specialized topic, or, as is often the case, a combination of some or even all of these.

As this article is about first-person usage, a simple example might help to illustrate this distinction. If I am washing dishes it is difficult to describe even a trivial incident neutrally. To state that “the glass broke” or even “slipped out of my hand" may imply that it is a normal part of the tedious process of washing up for glasses to break. “I broke the glass” adds the agent, the glass is now the 'goal' of the verb and the agent might even be represented as responsible (“I broke that glass, it didn’t break itself”). Even the intransitive “the glass fell” is a choice from an ergative/ non-ergative system. Competent users of language therefore have to be able to make choices not just between transitive and an intransitive, but also between agentive and non-agentive uses of verbs, only some of which may be used transitively. 

   Grammatical ‘transitivity’ in its narrowest sense is defined in terms of whether the verb has or can or cannot have an object as 'goal', but transitivity in a systemic sense is much more than this. The examples in the previous paragraph indicate that this grammatical view only represents a limited range within a more complex system. To accommodate such common uses, Halliday and Matthiessen (2004) include the notion of 'ergativity' within the global transitivity system. “The glass dropped” is non-ergative, represented as self-engendered”, whereas “I dropped the glass” is an ergative use with some notion of agentive cause built into the representation. Such ergative/non ergative pairs are available very extensively in common English verbs. Halliday and Matthiessen estimate that 60% of English verbs, including those used most commonly in general English (and this point is important to this discussion), may be used in either an ergative or non-ergative way. For example, “I drove the car away”/ “The car drove away” opposes the agent to “the medium through which the process is actualized” (p. 284). 

   For Halliday and Matthiessen  “ ‘happening’ means that the actualization of the process is represented as being self-engendered. Something just happens as part of a process. On the other hand “ ‘doing’ is represented as being caused by a participant that is external to the combination of Process and Medium. This external cause is the agent”(p. 285). Some of the many examples of common verbs that have ergative/non ergative pairs available as choices in English are listed by Thompson (p.136) as follows: altered/ closed/ darkened/ deflated/ defrosted /drove/ melted/ narrowed/ ripened/ slowed down/ rang. Thompson opposes, for example, “they rang the bell” with “the bell rang”. Common depersonalized examples from academic discourse can include: "this data suggests", "these studies investigate", ''these results illustrate…" which have become conventional regardless as to whether they legitimately attribute agency to some kind of process rather than the agents of that process.

Impersonal Voice
As Hyland (2002, p. 351) points out, "style guides and textbooks commonly portray scholarly writing as a kind of impersonal, faceless discourse, and EAP teachers direct students to remove themselves from their text." Hyland (2002, p. 357) argues that "effective academic writing depends on appropriate language choices." He maintains (p. 355) that "an explicit writer presence is often an effective rhetorical option." This view matches my own understanding of systemic competence. A systemic view implies that competence is related to the knowledge of, and ability to use, the available linguistic choices appropriately within the context. Hyland claims that "self-mention" is not automatically inappropriate in academic discourse. 

   The following statement occurs in the submission guidelines of the Asian EFL Journal: "Authors are encouraged to conform with international standards of drafting, but every effort will be made to respect original personal and cultural voices and different rhetorical styles." By examining the use of the first-person pronoun, a further aim of this paper is to critically evaluate this claim in the light of the evidence available in one full issue. 

Hyland's suggests (2002, p. 352) that "most obviously, … a writer's identity is created and revealed through, the use or absence of the I pronoun." This paper will examine the use of first person subject pronouns in one full issue of the Asian EFL Journal (March 2008) in order to evaluate this claim.

Data Analysis: Agentive Use of the First Person in the March 2008 Issue
In the following section, I have recorded the use of the agentive first person for the March 2008 issue of the Asian EFL Journal. Only first-person subject pronouns in the authors' own text are counted. Pronouns in citations, in appendices or in recorded data of informants' discourse are excluded as they are not representative of the authors' own voice. In table 1 below, I have recorded the use of the first-person in the joint authored papers. Two things are noticeable from this table. Firstly, the number of first person subject pronouns is relatively low (slightly less than 1 in every 4 pages.) Secondly, there is a noticeable difference between the two sets of authors who use them very rarely (3 in 49 pages) and those who use them more frequently (20 in 38 pages.)

Table two below records the number of first-person pronouns used in single-authored papers. Here the average number of first-person singular subject pronouns is approximately 1 in every 3.5 pages. Rather like in the joint-authored papers, there is an easy division to make in terms of frequency with one difference: there is no singular first-person pronoun in 50% of the papers. In addition to the 4 authors who never use 'I' at all, 1 author almost never uses it (twice in 20 pages). 3 authors use it more frequently, the most frequent being 31 in a total of 31 pages in Chi Yen Chiu. This case will be examined below in some detail as this represents almost 60% of the total uses of 'I' in the 189 pages of single authored papers. The second most frequent usage is 9 in 15 pages by Nikolova (2008). This latter instance is interesting, but not surprising because this paper was included as an "alternative voice" paper (Nunn and Adamson, 2007) and would otherwise have been rejected as too personal.

There are also some single-authored papers in which the first-person plural subject pronoun 'we' is used. This occurs both for some authors who never use the first-person singular 'I' and also by some who do. When first-person singular and plural subject pronouns are considered together, the frequency (slightly less than 1 every 4 pages) is very similar to the joint-authored papers in which 'I' is not an option.

Purposes and Functions of 'We'
Joint-authored examples
An analysis of the examples of ‘we’ in the joint-authored papers allows a further classification into ‘we’ used to represent authorial agency and a more inclusive use of ‘we’. 

Qualitative Analysis and Discussion
In Lee and Oxford (2008), the 12 instances of 'we' can be divided into 9 with specific reference to the authors’ own agency. The 3 non-specific uses of 'we' appear to refer to a collective community of practitioners as in sample 1 below:

Sample 1
Though many researchers believe that learner awareness is a necessary feature of strategy use (e.g., Carrell, 1989; Cohen, 1995), we still can see situations like Baker and Brown (1984) pointed out: when a reader did not describe how to use a particular strategy but did in fact use it. (Lee and Oxford, p. 10)

This can be compared to examples like sample 2 below in which 'we' only refers to the authors (Lee and Oxford, p.13).
Sample 2

We wanted to ensure that the translation used in this study was as faithful as possible to the English version.

In sample 3 below (Lee and Oxford, p. 14), the four instances of 'we' indicate the authors' clear recognition of their own agency and the need to explain their own position unequivocally on a relatively controversial issue: the relative importance of self-rated versus externally evaluated proficiency levels of students.

Sample 3
We felt it was very important to obtain students’ metacognitive self-assessment of proficiency vis-à-vis their peers because we agree with motivational theorists like Harter (1986). They believe that “humans have complex perceptions of themselves and their competences. . . All of these self-perceptions have motivational properties, with competence on a task more directly influenced by task-limited self-perceptions than global self-esteem” (Pressley et al., 1989, p. 307). We consider that this self-perception on his/her own English proficiency will play a very important role in actual performances. In addition, one of the most important themes of this study is awareness of one’s own strategy use; therefore, we considered self-rated English proficiency to be more relevant to our study than standardized test scores.

In such examples we can start to observe some important aspects of agency. When an impersonal form is used to disguise agency, I would argue that the authors are doing one of two things. Either they are doing this because the agent is less important than the process or they are doing it to follow a genre-related convention to disguise a subjective judgment in more objective-like language.

We felt it was very important to obtain students’ metacognitive self-assessment of proficiency vis-à-vis their peers because we agree with motivational theorists like Harter (1986)." (Extracted from Sample 3.)

This sentence could have been depersonalized as follows:

It was very important to obtain students’ metacognitive self-assessment of proficiency vis-à-vis their peers because as motivational theorists like Harter have pointed out(1986)."… (Redrafted from sample 3 extract.)

My own feeling is that it is to the authors' credit, and in the interests of academic transparency that they did not choose this path. I do not believe that this is an entirely  subjective judgment, because it is supported by data analysis which is difficult to counter.

In sample 4 below (Lee and Oxford, p.23), the uses of impersonal and more subjective language are juxtaposed. The more scientific measurement is followed by clear subjective agency. 

Sample 4
In sum, gender, though reaching statistical significance on some interaction effects, had effect sizes that were negligible. We think that it is still good for learners because we cannot change nor manipulate gender. The smaller the effect size of gender, the more possible it is for teachers or learners to improve learning.

Again my view is that the choice of the first person agent instead of an impersonal form, such as a passive voice – "…gender cannot be changed or manipulated", is fully appropriate as it makes no attempt to disguise a well-supported interpretation as a logical necessity. 
In sample 5 below (Lee and Oxford, p.24), the first person used with speculate provides yet another example of the transparent use of agency by the authors which is more persuasive to me as a critical reader than any attempt to disguise this agency could have been. Am I convinced by this argument because it is expressed as a modest claim or because it corresponds to my own view of Japanese university entrance exams? 

Sample 5
Since 1969, Korean students have taken multiple-choice entrance examinations, equivalent to the Scholastic Aptitude Testin the U.S., and we speculate that such examinations might promote compensatory strategies for guessing the right choice from the context, even if the details are not fully understood.

In the final sample (Lee and Oxford, p. 28), agency is again undisguised even when used for the impersonal application of a statistical procedure such as factor analysis.

Sample 6
For the current Korean translation of Version 7.0, we conducted a new exploratory factor analysis to determine whether the underlying factor structure was similar to that found for the SILL in earlier studies. We found that the factor structure was very close to the SILL factors reported elsewhere.

It can of course be argued with some logical support that my interpretation is invalid because the agency is obvious in an academic article that carries the authors' name. (See for example, Rodman (1981, p.2), who states, "the counter-argument that the truncated passive obscures the identity of the agent is not valid in this case, it seems to me, for the agent is fully recoverable from the context.") This counter argument could easily use sample 6 to suggest that the first person is inappropriate here with some justification. I have little to counter this counter argument, but do not agree with it. My tentative counter-counter-argument is that any conventional use of a genre-related practice such as the conventional disguise of agency in academic texts needs to be scrutinized in a study of voice. I would like to determine whether depersonalized discourse is appropriately expressing a fully self-engendered process that would lead to the same conclusion whoever instigated it, or is inappropriately disguising personal agency. Is it not better to err on the side of transparent agency rather than the reverse? 
In response to a request for feedback on my analysis, one author, when asked whether use of the first person was a conscious and deliberate choice, stated unequivocally:

It was a deliberate choice. I personally dislike obfuscation, opaqueness, passivity, and general stuffiness in writing about research. I don't see why transparency and agency are not the norm. (Oxford: e-mail feedback)

In some cases, the use of 'we' is slightly more ambivalent. Otoshi and Heffernan (2008) appear to use first person ‘we’ inclusively to present themselves as members of the ‘community’ of language teachers, establishing a kind of solidarity with their audience, as seen in sample 7 below (p.75). However, this does not exclude a parallel interpretation that 'we' also specifically refers to the authors who are attempting to improve their own practice. This example serves to indicate that my quantitative classification has its limitations and requires qualitative interpretation which is more subjective:

Sample 7
Based on the results of this study, we can draw some conclusions about recommendations on how to best prepare our learners for oral presentations. As teachers, we can prepare our learners to be aware of the above criteria (Table 1 and Table 2) and the effect they have on the efficacy of oral presentations in EFL classrooms.

First-person plural 'we' in single authored papers
As we might expect, the 16 occurrences of 'we' in single-authored papers are classified as inclusive uses. Nonetheless, some of the examples indicate some interesting differences with the joint-authored papers, as the single author has another first person option that the joint authors do not have. In sample 8 below, the use of the inclusive, community-oriented 'we' is juxtaposed with a personal authorial voice 'me' and with his student's voice. 

Sample 8
As we can see, this counseling role not only empowered Jing to recognize her communicative needs but also helped me to interpret what she wished to gain from this class. (Chi Yen Chiu, 2008, p.101)

This example illustrates how an author can make full use of the systemic options of voice to good effect in the context of one particular study. This is not to suggest that all competent academic writers will select the same options. In sample 9 below different options were selected. The inclusive 'community' first person 'we' is juxtaposed with impersonal passive constructions. A statistical study into the 'C-test' is arguably more scientific. I interpret the author's choices here to mean that some interpretations "can be legitimately made" as part of a more objective research process while others cannot. This contrasts with the topic of the study from which sample 8 is drawn, which is a qualitative study of learner autonomy.

Sample 9
The primary concern for any test is that the interpretations and the uses we make from the test scores are valid. The evidence that we collect in support of the validity of a particular test can be of three general types: content relevance, criterion relatedness, and meaningfulness of construct (Bachman 1990). These categories have been separately discussed below with regard to the data presented in this study and the interpretations that can be legitimately made on their basis. (Rouhani, 2008, p. 162)

Rouhani (2008) never chooses 'I', but uses 'we' four times. Nonetheless, as a journal editor, while the authorial comment on interpretation is welcome, I would still maintain that it is the author who is ultimately responsible for these interpretations. Reviewers had to agree to the author's claim about the legitimacy of the interpretations.
To terminate this discussion on the use of 'we' in one complete journal issue, it is interesting to turn to the paper that uses the first-person singular and plural pronouns the second most frequently (Nikolova, 2008). The author uses the first-person 'I' 5 times in the 151-word abstract (p. 259), but still uses the option of a plural voice in support of her (single-authored) recommendations in the conclusion section: "We may enumerate our recommendations…" (p. 272).

First-person 'I' in single-authored papers
Given the relatively small number of first person pronouns in single-authored papers, the following example from Chi Yen Chiu (2008, p. 87) is interesting as it illustrates the deliberate avoidance of the first person. The author refers to himself in the third person.

Sample 10
After a few e-mail exchanges with EFI coordinators, the investigator became a volunteer teacher for cyber course of grammar and writing from February 12, 2001 to September 27, 2002. His contact with course participants was made possible by the coordinators who sent him the names and e-mail addresses of English learners from the basic to intermediate level. Since the EFI coordinators did not establish any requirements on how the course should be structured, the investigator was free to manage his own class.

Still more interesting is the fact that, in a 31-page paper, Chi Yen Chiu (2008, p. 93)  then code switches on page 16, using the first person for the first time in section (4.2.1.), a section discussing the discourse of teaching roles.

Sample 11
By using imperatives, I directed the learner in exactly what to do and showed her the focus of each task.

This is then followed by 30 more similar uses of 'I' in the second half of the paper. These 30 uses in just one section of the paper (in addition to 9 uses of 'me' and 10 uses of 'my', excluding first-person uses in the e-mail data samples used in the paper) warrant further inquiry as they represent more than half the total uses of 'I' in the 290-page issue within the space of just 13 pages. 
The 'teacher as investigator' is not now being discussed as a theoretical construct in the third person. Instead the first-person thought processes of the teacher are being described.

Sample 12
Mick sent his essay 3 two weeks later than the due date (see E-mail 103).  He apologized for the delay because he had been very busy at that time. Mick’s delay was unusual because he often turned in his written assignments sooner than I expected.  However, if I had had institutional power to enforce the due date, “being busy” would not have become a justifiable reason for Mick to miss it. (Chi Yen Chiu, 2008, (p. 96)

This interpretation is supported by the fact that the first person is then dropped again at the end of section 4. Section 5 discusses the more general implications and conclusions of the paper. For me, this does not represent inconsistency on the part of the author. It is more an example of code switching or rather 'voice switching' in different parts of the paper which are written for different purposes, possibly in relation to 'positioning'.

Having completed and written up the analysis above, I decided to compare the final published draft to the originally submitted draft of this same paper to check whether there was any difference caused by review. I should clarify at this stage that I had not been involved in the details of the review, which had been supervised by an associate editor. This led to potentially the most interesting finding in this particular case. In draft one, there is no example of 'I' even in section 4. All first-person uses in the published draft were written in the third-person in the original. For example, "I consulted my co-worker" in the final draft appears as "the teacher consulted his co-worker".  Further investigation reveals that these changes were done at the request of a reviewer who asked for more clarity about 'the teacher' as investigator. The identity of 'the teacher' (who is in fact the author) was disguised in the original manuscript. These comments amount to a desire for a more transparent voice.

Conclusions
While examples have been found of individual reviewers suppressing the author's first person voice, these cases are relatively rare and do not in themselves lead to rejection of papers. Some authors choose not to emphasize their own voice, but those who do use it tend to code-switch depending on the context and the topic or subtopic in different sections of a paper. The first-person can be used for a variety of purposes at different stages of the discourse, one of which is transparency. The author who used the first person the most did not do so in the original draft. The author chose to make this voice change in response to a review comment: "Especially, the researcher(s)'s (the author(s)'s) roles should be clearly explained." This comment did not specifically mention the first person, but using the first-person does clarify the author's ethnographic role as the teacher and investigator.
Any competitive selection process within an academic discourse community suppresses some voices. The question is then who is behind this process as reviewing is not a scientific self-engendered process. AEJ currently rejects more than 80% of all submissions. However, the current evidence suggests that AEJ does allow and sometimes even encourages first-person voices. Within the one AEJ quarterly issue analyzed in this paper, if success is measured in the ability to survive a rigorous review system, no clear pattern emerges as some successful authors use the first-person relatively frequently and other successful authors hardly use it at all. This aspect of the findings need not worry an international journal that values diversity, but it is clear that more attention needs to be given to promoting a greater range of diverse voices without compromising editorial standards.

References
Carter, R. (2004). Language and creativity: The art of common talk. London:   Routledge.

Carter, R. (1996). Politics and knowledge about language: The LINC project. In R.Hasan&G. Williams (Eds.),Literacy in society(pp.1-28). London: Longman

Carter, R. (1995). Keywords in language and literacy. London: Routledge.

Eckert P, &Wenger, E. (2005). What is the role of power in sociolinguistic variation? Journal of Sociolinguistics, 9(4), 582-589.

Flowerdew, J. (2000). Discourse community, legitimate peripheral participation, and    the nonnative-  English-speaking scholar TESOL Quarterly, 34(1), 127-150.

Halliday, M. & Matthiessen, C. (2004). An introduction to functional grammar. London: Arnold.

Hasan, R. & Williams, G., eds. Literacy in society. 1-28. London: Longman.

Holstein, J.A., &Jaber, F. G. (2000).The self we live by: Narrative identity in a postmodern world. New York: Oxford University Press.

Hunsten, S. (1994). Evaluation and organization in academic discourse. In M. Coulthard, (Ed.), Advances in written text analysis (pp. 191- 218). London: Routledge.
Hyland, K. (2002). Options of identity in academic writing, ELTJ, 56(4), 351-    358.

Lin, B. (2006).  Genre-based teaching and Vygotskian principles in EFL: The case  of a university writing course. Asian EFL Journal,8(3), 226-248.

Nunn, R. & Adamson, J. (2007). Towards the development of international criteria for journal paper evaluation. Asian EFL Journal, 9(4), 205-228

Reinders, H. (2006). The effects of different task types on L2 learners' intake and   acquisition of two grammatical structures. University Of Auckland, 2005
Department of Applied Language Studies and Linguistics. Asian EFL Journal,   Thesis Section, November 2006).

Ribeiro, B. (2006) Footing, positioning, voice, Are we talking about the same things? In A. De Fina, D. Schiffrin,  & M. Bamberg, (Eds.), (2006) Discourse and identity(pp.48-82). Cambridge: Cambridge University.

Rodman, L. (1981). The passive in technical and scientific writing. Paper presented at the Annual Meeting of the Conference on College Composition and Communication
(32nd, Dallas, TX, March 26-28, 1981) retrieved from
http://www.camlang.com/tsp001.cfm, on 22.08.2007.
Thompson, G. (2004). (2nd ed.): Introducing functional grammar. London: Arnold.

Wennerstrom, A. (2003). Discourse analysis in the language classroom. Volume 2. Genres of writing. Ann Arbor: The University of Michigan Press.

Appendix See PDF