1. Introduction

[1] In his thorough demythologizing of English monolingualism in composition, Paul Kei Matsuda (2006) helped contextualize for those of us who work in writing centers what we experience daily: multilingual writers are the norm. Many scholars have pointed out a need to move beyond traditional writing center (WC) pedagogies to better serve English learners (Hewett & Lynn, 2007; Matsuda, 2012; Myers, 2003, Powers, 1993; Salem, 2016). Indeed, there has been no shortage of scholarship on the prominence of multilingual (ML) writers in writing centers. So why is the problem of how to respond to the writing of multilingual students such an intractable issue for writing centers? In a word: training. Encouraging tutors to be more accepting of language differences is essential but not enough; tutors need training on how to prompt language learning with their feedback. Perhaps nowhere is this more evident than in asynchronous consulting, where the tutor feels alone with the writer’s text, and where questions about how to best provide feedback inevitably arise. Cut off from immediate interaction with the writer, a tutor’s effectiveness hinges on their training and support resources. Though each writer’s text is unique, the kinds of linguistic challenges encountered by tutors are documentable if not quite predictive. Experienced tutors tend to develop a repertoire of feedback gleaned through trial and error, and writing centers can capture these experiences for training purposes. In addition to the insights of experienced tutors, writing centers can better prepare tutors to provide quality language-focused feedback by drawing on the field of applied linguistics. Though sustained training of peer tutors in responding to the linguistic challenges of ML writers is a significant investment of time and resources, writing centers cannot avoid the following facts:

• multilingual writers are a significant number of students who seek writing center consultations (Matsuda, 2012; Myers, 2003; Rafoth, 2015)
• multilingual writers frequently ask for and benefit from feedback on lexis and grammar (Myers, 2003; Nakamaru, 2010)
• writing centers, with their individual focus, are ideal sites for language learning (Hewett & Lynn, 2007; Leki, 2009; Myers, 2003; Severino & Deifell, 2011)

In this article, I will describe how I compiled model language-focused written feedback to ML writers into a reference manual for tutors. The manual described in these pages draws from applied linguistics research findings that are essential for asynchronous tutoring sessions in which multilingual writers seek feedback on grammar and vocabulary in their writing. But before we turn to the development of such a reference manual, let us first consider the role of peer tutors in teaching language; views on asynchronous feedback for ML writers; scholarship on training tutors to work with ML writers; and a brief overview of applied linguistics research germane to asynchronous feedback.

2. The Role of Peer Tutors in Language Teaching

[2] How tutors should respond to multilingual (ML) writers gets straight to the heart of the limitations of the peer tutoring model. Writing center practitioners have not reached consensus on the role of the peer tutor when it comes to providing feedback to ML writers on grammar, vocabulary, and other related language concerns. What tutors should know about English, and how to teach it within a tutorial, is notably absent in much of the writing center literature about multilingual writers.Traditional writing center pedagogies, such as non-directive tutoring, tend to reflect compositionist process-oriented priorities, and remain influential in many writing centers. These pedagogies privilege the writing process over student language concerns, meaning students who request and benefit from lexicogrammatical feedback are sometimes poorly served when writing centers shy away from it (Salem, 2016).

[3] Administrators and scholars have, quite justifiably, been skeptical of peer tutors’ effectiveness in taking on grammar instruction within tutorials. With the exception of the odd grammar enthusiast on staff, peer tutors do not typically come equipped with the requisite knowledge of grammar and language teaching to provide effective feedback on grammatical concerns. Without training, feedback that they do provide tends to veer uncomfortably close to the much dreaded editing and proofreading. As Paul Kei Matsuda (2012) observed:

peer tutors, who are by definition sympathetic readers but not experts in the teaching of writing or language, may not be able to meet the needs of clients who have an advanced knowledge of the subject and discipline-specific genres yet are struggling to express their ideas in the second language. (p. 48)

​One of the first scholars to call for a new understanding of the role of peer tutors when consulting multilingual writers was Judith Powers (1993), who suggested that tutors see themselves as “cultural informants” (p. 370), an approach best suited for working with international students. For Judith Powers, this shift entailed becoming less Socratic and putting aside hesitation about being more directive. However, Jane Cogie, Kim Strain, and Sharon Lorinskas (1999) found that the way they had been practicing Judith Powers’s cultural informant approach in their writing center ultimately led tutors to editing. Cogie et al. advocated that tutors teaching of self-editing strategies to clients in order to help tutors resist the pull of corrective editing. For Cogie et al., the onus for language learning remained with the student as part of self-editing. Cynthia Linville (2009) largely echoed this perspective, emphasizing the need for ESL students to learn to edit their own texts. Though Linville shifted some responsibility for identifying errors to tutors, proposing a list of six common errors that she maintained tutors should learn to identify, she asserted that “a tutor is not a grammar teacher. [Their] ability to help is limited” (p. 117).

[4] Yet for some practitioners, a clear delineation between tutors who help ML writers identify patterns of error--but no more than that--and teachers who can demonstrate how to better communicate with academic interlocutors is a problematic binary. Myers (2003) questioned the usefulness of such distinctions:

I think it is both possible and desirable for writing center staff to fill the role of “foreign/second-language teachers” as well as writing instructors. In fact, writing tutors are perfectly positioned to facilitate the language learning these students need in order to develop their ability to write in English. The central insight in foreign language pedagogy in the last thirty years is that, in fact, language acquisition emerges from learners wrestling with meaning in acts of communicating or trying to communicate. That is exactly what ESL students are doing in writing centers, person to person. (pp. 296-297)

​​Here, Myers indicated that writing centers can better understand and adapt to the language learning needs of students. Often overlooked but central to these needs is vocabulary and its overlap with grammar as an essential language learning need. Sarah Nakamaru (2010) and Carol Severino and Elizabeth Deifell (2011) have urged writing centers to teach vocabulary to English learners within the context of their own writing. This is a further break from previous conceptions of a writing tutor’s role, especially given concerns about appropriation of student writing (see for example Severino, 2009).

[5] As long as we recognize that ML students will remain a key writing center demographic into the foreseeable future, it seems expedient to reorient tutors toward more active language teaching. At some institutions, writing centers are the only obvious resource for multilingual students to learn more about the language they need in order to progress through academia. Well-meaning attempts to cordon off student language needs as “editing” or “proofreading,” then recommending additional coursework, or referring a student to links or worksheets, are missed opportunities to teach students within the language context that is most immediately relevant to them: their own. Rather than referring writers to ancillary resources (e.g. worksheets, coursework, friends), WCs need to embrace the fact that tutors themselves are often the best resources a language learner can access.

3. Asynchronous Feedback and ML Writers

[6] Asynchronous feedback for ML writers seems to be one area of broad agreement among writing center scholars and practitioners. Regardless of how one sees the role of a tutor, most practitioners agree that asynchronous feedback for ML writers should not simply entail the tutor editing or correcting the writer’s text (Rafoth, 2009; Hewett & Lynn, 2007; Myers, 2003). Ben Rafoth (2009) represented the prevailing view that asynchronous feedback for ML writers should focus on higher-order concerns (HOCs), though any language issues that cause confusion for readers are a worthy priority. He further advised tutors to keep the focus on HOCs because writers tend to prioritize language feedback over advice on organization and other rhetorical concerns, even when they receive feedback on both. Lee-Ann Kastman Breuch and Linda Clemens (2009) expanded on Ben Rafoth’s advice by suggesting that tutors “provide ESL students with feedback on higher-order concerns as well as error feedback (rather than error correction) on lower-order concerns—especially when ESL students request such feedback” (pp. 134-135).

[7] Much has been made of the differing natures of face-to-face and online tutoring (both synchronous and asynchronous). Despite the potential pitfalls in asynchronous modes for tutors and writers to focus on grammar and mechanics, several scholars have noted distinct advantages afforded to language learners by the asynchronous format. Most obvious is the posterity of the feedback. Rather than occurring in a passing, unrecorded moment as in a face-to-face session, the feedback is available to writers for as long as they care to save it (Breuch & Clemens, 2009), which could lead to increased opportunities for acquisition of the language on which the feedback was provided (Williams & Severino, 2004). Additionally, access and social justice are also strong arguments for asynchronous OWLing. The material conditions of many students’ lives, such as work, commuting, and family obligations (Inoue & Keown, 2015), mean that asynchronous modes of feedback are sometimes a student’s best or only option.

[8] Overall, the prevailing practice in asynchronous feedback is to treat ML writers like any other writer. This approach seems motivated by a desire to keep tutors and writers alike focused on HOCs. While this is a legitimate concern for novice tutors working asynchronously, it still skirts the issue of how tutors should respond to language breakdowns that do in fact cloud reader understanding. Additionally, telling tutors to focus on HOCs does not prepare them to respond to writers who have done a satisfactory job fulfilling an assignment’s expectations but request linguistic feedback for their own learning goals. Furthermore, a focus on HOCs assumes that lexicogrammatical errors are rarely a high priority. On the contrary, a writer consistently using a word or grammatical form outside the conventions of a discourse will continue to do so in subsequent drafts unless a tutor identifies it and provides useful feedback (Myers, 2003; Nakamaru, 2010)

3.1. Writing Center Training on ML Writers

​[9] Virtually all writing center professionals would likely agree that tutors need training to prepare them to work with ML writers. Though at one time this preparation might have boiled down to what one might call proofreading resistance training, the field has clearly recognized that we must do more to serve our students. Despite this recognition, discussions of training tutors to work effectively with ML writers have rarely offered practical guidance on how to attend to lexicogrammatical issues.

[10] Perhaps a dearth of actionable suggestions for working with ML writers in writing centers is understandable when we consider that the field of composition, from which writing centers tend to draw their intellectual lineage, has persisted in its suspicion of grammar instruction because of its association with remediation and a recognition that “good grammar” does not equate with good writing. While an undue obsession with prescriptive grammar can smack of elitism, classism, and racism, and can inflict real harm on students, this is far from the only option for understanding grammar. Thus, as writing centers serve ML students, we must be able to accept language differences even as we teach learners of English to use language effectively in the growing number of ways they need it in the academy. Tutors cannot accomplish this task without more robust training.

[11] Ben Rafoth (2015) devotes the last chapter of his book Multilingual Writers and Writing Centers to tutor preparation as it concerns ML writers. He nods approvingly at the growing diversity of writing center tutors, lauds the growing acceptance of directive tutoring approaches for some situations, and urges reaching out to faculty to help them develop greater tolerance for accented writing. But when a tutor is confronted with a breakdown in meaning in a student text, and the writer isn’t there to ask, how should the tutor respond? In fairness to Ben Rafoth, he notes that his book does not “provide a cupboard of strategies to open for tomorrow morning’s tutoring session” (p. 129). But maybe a big cupboard is precisely what tutors need when wrestling with questions about language learning and teaching.

[12] Judith Powers and Jane Nelson (1995) called for L2 specialists to collaborate with writing centers and get involved in tutor training. They point out that while a workshop or a visit from an outside specialist can help, it barely scratches the surface, and so they urged “better, perhaps less traditional, solutions to the problem of L2 tutor training” (p. 129). Continuing in this vein, Hewett and Lynn (2007) contended that tutors should attain a working knowledge of English grammar from a pedagogical perspective (i.e. the ability to show how the language works) and that tutors receive regular mentoring on the feedback they give to students. Hewett and Lynn underscored that there is no substitute for understanding how English works: “...we think that correctness of advice ought to be considered the first priority” (para. 13) [emphasis in original]. To this end, we must challenge assumptions about correctness. Predictably, much of the incorrect advice I have observed tutors (and faculty!) give originates from prescriptive rules of thumb (e.g. stick with the same tense, never use passive voice, don’t end a sentence with a preposition, etc.). To discourage tutors from defaulting to prescriptive myths about English usage, tutor education should take a descriptive approach to grammar, drawing on sociolinguistics and corpus linguistics studies. To briefly illustrate this distinction, I offer Table 1, which I frequently use in tutor training to begin challenging assumptions about the value and purpose of grammatical correctness.

[13] ​Finally, none other than the Conference on College Composition and Communication ([CCCC], 2009) has called for better training for writing center tutors in our work with ML writers: “It is imperative that writing centers model and discuss effective approaches for working with second language writers in tutor training...” I concede that it would be naïve to overlook the significance of a commitment to training tutors to be effective language teachers; training requires time and expertise in an environment where the bulk of the labor is done by students, some of whom we hope are developing professionals in the field, and all of whom we know are likely to graduate and move on. That’s why it’s easy to despair, to throw up our hands and say “Impossible!” But I hope to persuade you that the feedback we give–especially in asynchronous modes–holds the key to developing tutors into informed, strategic, purposeful providers of linguistic feedback.

4. Relevance of Written Corrective Feedback Research

[14] Before I describe the tutor training project I have developed, I want to briefly examine the current consensus in the field of Written Corrective Feedback (WCF), because my approach to teaching language through asynchronous feedback is rooted in an applied linguistics understanding of error and second language acquisition research. (For a richly detailed survey of WCF from both Composition and Applied Linguistics perspectives, see John Bitchener & Dana Ferris, 2012.) As a teacher of English to speakers of other languages, I proceed from the assumption–though not always the case–that learners of English want to make fewer errors in their writing, and one of my objectives as a language teacher is to help them achieve that. It just so happens that I work in a writing center, not a classroom. To teach a language means that the default target language I teach is a “standard,” even as I am aware that any language “standard” is a social construction bound up in the power and prestige of its speakers, is never inherently superior, and is always changing (Lippi-Green, 2012). While delving fully into this dilemma is beyond the scope of this article, it should not be dismissed. (Writing Centers and the New Racism, edited by Laura Greenfield and Karen Rowan, is essential reading.)

[15] In their erudite survey of WCF research, John Bitchener and Dana Ferris (2012) note that though there is much researchers cannot yet definitively say about WCF, they have established that:

the long-term effectiveness of providing a single treatment of unfocused written CF is uncertain: as the findings are conflicting, but the long-term effectiveness of providing a single treatment of focused written CF on discrete, rule-based linguistic categories of error is clear and compelling for the limited linguistic environments investigated so far. It is unclear whether focused or unfocused written CF is the more effective. (pp. 73-74)

​In other words, when learners are provided with limited, useful linguistic information that they can understand and act on, they can learn from it. Here, unfocused feedback describes commenting on any number of errors without establishing a clear pedagogical priority, whereas focused feedback entails the commenter identifying and prioritizing specific types of errors for the writer to grapple with. In the short term, a writer’s accurate repairs of identified errors is not necessarily evidence of learning nor a guarantee that the identified type of error will be learned and thus avoided in the next draft (Bitchener & Ferris, 2012, p. 73). Language acquisition is not on a quarter (or semester) schedule, so it behooves writing centers to take the long view in their approach to language-focused feedback (Myers, 2003). The efficacy of direct versus indirect feedback is another topic of inquiry in WCF that should interest practitioners of asynchronous tutoring. Indirect feedback alerts the writer to the existence of an error without making the correction, whereas direct feedback entails the instructor making the correction. John Bitchener and Dana Ferris note that (as of 2012) “recent studies report a clear advantage for direct forms of feedback” (p. 74). It should also be noted that several variations of direct feedback exist, one of which is coupling metalinguistic explanation with direct feedback, and some researchers have suggested that this might be the most useful kind of direct feedback for learners (Bitchener & Ferris, 2012, p. 74). For learners who have received explicit instruction in English, metalinguistic feedback could (re)activate knowledge about the grammatical concept at issue. Though investigation into many other variables on the effectiveness of varied WCF approaches are just beginning, we can be reasonably confident that quality, strategic WCF that affords learners the opportunity to use it does lead to language learning over time.

[16] Before moving on, one additional applied linguistics concept is worth consideration for asynchronous tutoring. In Second Language Acquisition research, noticing a discrepancy between how language learners use a particular aspect of the language (e.g. past tense) and how other speakers use it can lead a learner to acquire that form (Schmidt, as cited in Mitchell & Myles, 2004; also see Rafoth, 2015, pp. 112-114). One way to facilitate noticing is through enhanced input, which I have just demonstrated by bolding and italicizing the term. Enhanced input helps draw learners’ attention to differences in language forms and some evidence suggests that enhanced input can facilitate a learner’s acquisition of a form when they have prior knowledge of it (Han, Park, & Combs, 2008). This strategy can be particularly helpful in asynchronous tutoring, where tutors must endeavor to demonstrate the difference between language as used by the writer and language suggested by the tutor. (For a helpful summary of key second language acquisition concepts relevant to writing center work, see Rafoth, 2015.)​

[17] Though WCs have recognized a need for varied pedagogies for ML writers and validated language learning as a legitimate goal of a writing center session, the literature on WCs evinces little implementation of established WCF research in asynchronous modes of tutoring. To address this need for change, I argue that the expedient approach is to teach tutors to respond to writers’ lexicogrammatical errors with feedback aimed at language teaching. I further argue that asynchronous modes of tutoring are ideal contexts for tutors to practice metalinguistic explanations and feedback, and that these lessons can transfer to synchronous online and face-to-face tutoring. One approach to reimagining training to meet this goal, described in detail in the following section, draws on authentic samples of student writing submitted for asynchronous feedback, in addition to model feedback that embodies both WCF research and the pedagogical priorities specific to each writing center.

5. Origin of the Written Feedback Manual

[18] One of the central challenges in writing center staffing is training tutors to provide effective feedback to peer writers. Where do writing center administrators find the time and budget to train tutors? Though on occasion an experienced tutor lands on the doorstep, skilled writing tutors take time to develop through training and experience. Bringing tutors to the point where they meet basic writing center competencies and retaining them for a significant period of time before they graduate is often half the battle of running a writing center.

[19] All of this traditional writing center training typically eschews grammar--not merely as a topic worthy of a writing consultation but also as a key to understanding language. Now imagine including applied linguistics in this training. For years, working as a specialist for students whose first language is not English in a writing center, I would try to share important knowledge about applied linguistics/TESOL with peer writing tutors. My colleagues and I allotted three hours to pre-service training on working with multilingual writers, set aside some tutor meetings for follow-up training, and I implored tutors to ask me questions. On occasion, tutors would describe a problem to me, but they rarely wrote it down, leaving us working with hypotheticals and generalities for a writing consultation that had already concluded. Although the work of applied linguists like Eli Hinkel (2009) could help me anticipate common grammatical issues in academic English that cause breakdowns in communication, and though writing tutors always said they wanted more training in working with multilingual writers, I struggled to share my knowledge and experience in ways that would stick with tutors in future writing consultations. It was clear to me that the training was insufficient. What tutors really needed, I realized, was the “cupboard of strategies” alluded to by Ben Rafoth (2015, p. 129).

5.1. Institutional Context

[20] The writing center where I developed this training manual and reference is located on a campus of a major research university in the Pacific Northwest. The campus aims to provide increased local access to higher education in a medium-sized city in a region undergoing dramatic socioeconomic change. Three factors distinguish the campus from a residential public research university: 1) The student body includes a significant number of students returning to college to complete their studies or earn new degrees, and many of these students work full-time in addition to their studies; 2) many students commute to class and do not stay on campus once their classes have concluded; and 3) about one-third of incoming first-year students speak a language other than English with their family. Notably, most of these students are community residents, not international students. The campus neither requires nor offers specific multilingual sections of writing courses, nor does it require or offer English language courses with the aim of improving fluency, grammatical accuracy, or vocabulary size, meaning the writing center is the go-to language learning resource for ML students. In 2010, the writing center responded to calls from faculty and added a full-time, professional staff tutor to specialize in working with ML writers.

[21] In 2014, the writing center where I work began offering WCOnline’s eTutoring, an asynchronous mode of soliciting and delivering written commentary and feedback. Whereas students previously had to send a paper to an email address and wait up to two business days, now any appointment could be asynchronous. Our internal data on tutoring sessions show that 50% are currently conducted asynchronously, and half of those asynchronous consultations are sought by ML students. This written mode of feedback, within an appointed time and without the peer writer’s immediate presence, create new challenges for tutors.

[22] Despite the disadvantages of asynchronous writing consultations, they are much more conducive to creating a record than are face-to-face consultations, as noted by Lee-Ann Kastman Breuch and Linda Clemens (2009). Our writing center saves documents with consultant feedback on a server; if a peer consultant forgets to upload the paper to the appropriate appointment, we can find it later. Gradually, it dawned on me that we were creating a record of how each of us--peer and professional--provided feedback. What choices were we making? What did we excel at? What did we need to work on? We were also creating a set of practices that we could point to and hold up (hopefully) as an example to new hires, laying down a baseline of competence for tutors to reach before they could conduct written asynchronous consultations on their own. Reviewing asynchronous consultations also became an effective way for supervisors of peer tutors to periodically check to make sure that they remained on track. Written asynchronous feedback provided a new opportunity to share my practices with my colleagues, though it would take me several years before I hit upon how to do it effectively. Eventually, I began showing examples of feedback to tutors in the weekly meeting. Instead of trafficking in hypothetical grammar issues plucked from a grammar book, we were working with authentic student texts, so the challenges in them were empirically derived and tailored to the needs of our students.

[23] In my own feedback practices, I noticed a strong sense of déjà vu as I commented repeatedly on many of the same issues in different student papers, so I began saving some of my responses as templates for easy access in future asynchronous tutorials. Rod Ellis’s (2008) detailed typology on written corrective feedback inspired me to approach both error in asynchronous tutorials and my own feedback typologically. Once I had begun banking my responses, I realized I could compile them into a single document that I could share as a resource for other tutors. But that process would require identifying, coding, and organizing the most common types of errors in student papers. That, too, would be a valuable insight. And so out of curiosity, and a desire to find the most effective way to share my language teaching feedback practices with my colleagues, I started combing through ML student writing recently submitted to our center.

5.2. The Design Process

[24] Our writing center uses Microsoft Word commenting features to provide feedback in the margins of student documents. Though we also provide summative feedback that we place at the beginning of the document, margin comments are a common practice for us. I started with student papers I had recently worked on. As I worked on new papers, I kept an eye out for issues that seemed to be grammatical or lexical in nature. Whether an error was common, uncommon, or just a challenge to explain, I cut and pasted both the paragraph and the feedback provided by the writing tutor into a separate document. I recorded the identities of the students, and made a separate folder for student papers that I drew from, but I ultimately removed identifying information to keep each writer anonymous.

[25] Initially, I organized the document by categories that I expected to find, knowing that I would revise the categories and add new ones as I went. In an effort to be empirical, I purposefully did not consult a grammar text to create the typology in the document. Broadly, the categories included verb issues (tense, subject-verb agreement, passive voice), sentence structure (fragments, run-ons, relative clause problems), and vocabulary issues. In the course of my daily tutoring duties in the writing center, when I found an issue that I had not previously documented, or an issue that stretched my explanation skills, I copied an excerpt of that text with my response and preserved it in the emerging typology.

[26] As I continued working on student texts, I asked myself if the errors I encountered fit in categories I had already identified. To my surprise, I identified several common issues across student texts that I had not previously recognized as common. I also began to see relationships between grammatical and rhetorical issues that I previously hadn’t considered (e.g. writers creating dueling subject nouns in sentences where they quoted a complete sentence).

[27] Once I was satisfied that I had collected an empirical sample of the most common grammatical, lexical, and rhetorical issues across student papers at our institution--admittedly I did this impressionistically--I began to revise the document. At this point, I began to think more explicitly about the end users, who I intended to be other writing tutors in our writing center. I analyzed the items in the document to ensure they were appropriately categorized, eliminating redundancies, moving some, and creating subcategories for others. The typology that ultimately emerged is evident from the manual’s table of contents (Figure 2).

[28] I then turned my attention to the feedback attached to each item. I had a choice to make: should I preserve each comment as it was, or should I edit some comments? I chose the latter, because I reasoned that demonstrating ideal ways to respond was one of the purposes of the manual. So while the errors in the document are all authentic excerpts of student writing, the feedback is edited to ensure that it models the practices we expect our tutors to use.

[29] 5.2.1. Defining Model Feedback. Editing my own feedback prompted me to reflect on my feedback principles, not simply within each piece of feedback but across them. For me, model language-focused written feedback increases opportunities for language learning. Yet I began to worry that tutors would take from the manual that they were to comment on each error that they encountered, which was not my purpose for creating the manual. So I decided to codify a set of response strategies to promote language learning, all of which could be identified and demonstrated by the feedback in the manual (Figures 3 & 4). Each of the comments that demonstrate a response strategy also appear later in the manual. 

[30]​ Next I attended to the look of the document. As writing consultants, we get the sense that clogging the margins with comments from top to bottom overwhelms or discourages writers. I did not want to endorse this practice, but my margins in the document initially appeared this way. As in Figure 5 below, I chose to move some of the feedback from the margins and place it below some of the sample text, particularly if that explanation took up a lot of space. (For a thoughtful discussion of comment placement and volume, see Hewett, 2015.)

[31] As the document came together, I sought feedback from my colleagues, both professional and peer. After making revisions, I wrote an introduction, a user guide, and described a process for adding to the document. I wanted to be sure that tutors saw the document as something always in progress, something that they were encouraged to contribute to.

5.3. Integrating the Manual into Training and Practice

​[32] I announced the manual in a weekly tutor training meeting. For tutors’ convenience, I made it available in the same digital folder in which tutors save eTutoring work. A laminated version of the manual is also available in our eTutoring room, which we have set aside for tutors to work on asynchronous consultations. I presented the document as one part reference, one part manual, and I encouraged tutors to add vexing examples they encountered to the digital version. But as the academic year gathered steam, the document sat in its folder. Finally, in Spring quarter, I realized that without intentional efforts to engage tutors in its use, it would languish. When our director asked me to lead a series of trainings, I assigned the document as reading and developed activities that corresponded with each section, stretching them out over successive weeks. And when I provided feedback to tutors on their eTutoring work as part of ongoing observations, I referred to the manual and the principles within it. It was then that I started to get feedback from tutors on how they were referring to it during asynchronous consultations. Even better, I started to see the feedback strategies I advocated in the manual show up in the feedback of several tutors. It was working!

[33] Yet an important caveat is in order. Examples of model feedback, such as those in the manual, cannot replace the body of scholarship tutors need to engage with in their development. In our writing center, the manual is just one part of a series of models that include readings on ML writers in writing centers (Shanti Bruce & Ben Rafoth, 2009; Ben Rafoth, 2015; Sharon A. Myers, 2003), sociolinguistics (Laura Greenfield, 2011; Vershawn Ashanti Young, 2011), and challenging traditional writing center pedagogies (Anis Bawarshi & Stephanie Pelkowski, 1999; Nancy Grimm, 1999). Any manual modeled after the one described in these pages will likely be most effective when it supplements–not supplants– the invaluable work of scholars across disciplines that inform the practice of asynchronous tutoring.

6. Why Create a Feedback Manual for Your Own Center?

[34] Writing centers, like the people they serve, have unique qualities. Perhaps your writing center serves a high number of international students with the same native language, or maybe the student body is richly multilingual. Maybe your writing center works with a much higher volume of engineering students, or nursing students, or is visited by a broad cross-section of students from each program on your campus. Your writing center could be entirely online, or housed within a specific program. The iterations are many. Each of these situations creates a learning context all its own, especially when overlaid with the priorities of writing center administrators. That is why I recommend that each writing center that conducts some of its work asynchronously undertake its own study of submitted student writing and the feedback provided on it. The writing and language of the students will likely reflect the languages and discourses most prominent in each context. The feedback that we observe will reflect our practices as they stand and provide us with the opportunity to reflect on the following:

• Do our practices match our priorities?
• Which practices do we excel at? Which need reinforcing or change?
• How can we explain what is happening in this piece of writing to better help this student?

[35] Another reason to create a feedback manual that is unique to your center is that the ultimate product can reflect how your center reads ML texts (Severino, 1993; see also Matsuda & Cox, 2009) and responds to varieties and dialects of English in an academic space, a sociolinguistic question with serious social justice concerns. Dilemmas that we encounter in our reflections on our practices present us with opportunities to change.

[36] For writing center professionals, inquiry into our own asynchronous practices offers professional development opportunities. In all likelihood, we will encounter language phenomena that we do not understand or cannot explain. The project of creating a feedback manual is an opportune time to research answers to both new and persisting questions. In doing so, we can plan how to respond, potentially changing feedback practices in our centers to better serve language learners who seek our feedback.

6.1. Defining and Assessing Effectiveness of Feedback

[37] But is our feedback effective? How do we assess that? I define effective written feedback on lexicogrammatical errors as feedback that affords ML writers the opportunity to learn about that part of the language and the rhetorical task they are trying to accomplish. I put in the effort to make sure that my feedback is noticeable to students. I also want the feedback to prompt writers to take an action beyond reading it. By examining subsequent iterations of that writing, we see if the writer acted on specific pieces of feedback, but we cannot definitively determine if the writer has fully learned a lesson we tried to teach through the feedback. To be confident that a language learner has learned that lesson, we would need to look to new writing produced by that writer. Even then, as research on WCF has shown, we cannot expect total accuracy (Bitchener & Ferris, 2012). Moving an aspect of a learner’s language from error to mistake takes time and iterative practice on the part of tutor and student. Observing the same pattern of error in subsequent writing does not mean that a tutor or learner has failed; it likely means additional work is ahead.

[38] Another means of assessing the effectiveness of our feedback is to ask writers if they feel it helped them. Sometimes they tell us directly in an email, or in the appointment form when they seek feedback on a new assignment. As seasoned writing center professionals know, students often vote with their feet--students who don’t return to work with us are sometimes saying that they didn’t get what they wanted or needed. If what they expected was comprehensive editing without a learning component, they will not find it with other tutors if our training has been sufficient. Fully assessing feedback effectiveness would likely entail surveying writers and observing multiple pieces of writing authored by the same writers at different time intervals. To assess the effectiveness on tutors of a manual like the one described here, a director or other trainer would need to review feedback provided by peer tutors. The manual’s impact would be successful if peer tutor practices generally aligned with the model feedback demonstrated in the manual. Tutor engagement with the manual, in the form of questions and suggested revisions, might be another way to gauge effectiveness.

7. Final Thoughts

[39] The inquiry I undertook, which ultimately culminated in a manual for asynchronous feedback with a focus on language teaching, does not add to research about tutoring or written corrective feedback. It was neither an experiment nor a case study. However, I hope that I have offered a novel process for conducting an inquiry into a writing center’s asynchronous feedback practices, as well as helpful guidelines and sources for suffusing asynchronous feedback practices with key findings from the field of applied linguistics. It’s time writing center practitioners get beyond imagining our centers as places where tutors help multilingual writers. For students who are multilingual, the path toward becoming better writers winds through language learning, which includes not only making rhetorical moves, fulfilling the prompt, and considering the audience, but also gaining insight into the interactions between lexicogrammar and the discursive features of academic writing. Some might say, what if I don’t know enough about grammar to undertake this project? This is perhaps the most important reason to do it: help your students by adding to your expertise. (Celce-Murcia & Larsen-Freeman’s The Grammar Book is an indispensable reference.) For skeptics who contend that a peer, by definition, is not an expert, and cannot be fairly expected to know enough about language to teach it, I empathize. Though studying languages other than English can provide invaluable insight into the trials and triumphs of writing in a second language (or third, or fourth...), just as essential for tutors is acknowledging to ourselves that we are always language learners, adding to our knowledge of languages we use every day. Can we truly be peers of multilingual writers if we do not understand why key aspects of English are difficult to learn, or if we are unable to explain, to demonstrate, to teach why we communicate in our discourses as we do? I hope that my presentation of this manual will function as a springboard for further interdisciplinary inquiry into feedback--asynchronous and more--so that writing centers can better serve the language learning needs of ML writers.

7.1. Acknowledgments

I’d like to thank my colleague, Rebecca Disrud, for encouraging me to write about this manual. In addition to her excellent feedback, I’m grateful for feedback I received from others, including my colleague Tori Olive and ROLE special issue editors Beth Hewett and Megan Boeshart. Finally, I’m indebted to Asao Inoue, Kathi Bailey, Jean Turner, John Hedgcock, and Lynn Goldstein for their mentorship and support.

8. References

Bawarshi, Anis & Pelkowski, Stephanie. (1999). Postcolonialism and the idea of a writing center. The Writing Center Journal, 19(2), 41-58.

Bitchener, John & Ferris, Dana. (2012). Written corrective feedback in second language acquisition and writing. New York: Routledge.

Breuch, Lee-Ann K. & Clemens, Linda S. (2009). Tutoring ESL students in online hybrid (synchronous and asynchronous) writing centers. In S. Bruce & B. Rafoth (Eds.) ESL writers: A guide for writing center tutors (2nd ed.) (pp. 132-148). Portsmouth, NH: Boynton/Cook.

Bruce, Shanti & Rafoth, Ben. (2009). ESL writers: A guide for writing center tutors (2nd ed.). Portsmouth, NH: Boynton/Cook.

​Celce-Murcia, Marianne & Larsen-Freeman, Diane. (1999). The Grammar Book: An ESL/EFL teacher’s course (2nd ed.). Boston, MA: Heinle & Heinle.

Cogie, Jane, Strain, Kim, & Lorinskas, Sharon. (1999). Avoiding the proofreading trap: The value of the error correction process. The Writing Center Journal, 19(2), 7-32.

​Conference on College Composition and Communication. (2009). Statement on second language writing and writers. Retrieved from http://cccc.ncte.org/cccc/resources/positions/secondlangwriting

Ellis, Rod. (2008). A typology of written corrective feedback types. ELT Journal, 63(2), 97-107. doi:10.1093/elt/ccn023

Greenfield, Laura & Rowan, Karen. (2011). Writing centers and the new racism: A call for sustainable dialogue and change. Logan, UT: Utah State University Press.

Greenfield, Laura. (2011). The “standard English” fairy tale: A rhetorical analysis of racist pedagogies and commonplace assumptions about language diversity. In Laura Greenfield & Karen Rowan (Eds.), Writing centers and the new racism: A call for sustainable dialogue and change (pp. 33-60). Logan, UT: Utah State University Press.

Grimm, Nancy M. (1999). Good intentions: Writing center work for postmodern times. Portsmouth, NH: Boynton/Cook.

Han, Zhaohong, Park, Eun Sung, & Combs, Charles. (2008). Textual enhancement of input: Issues and possibilities. Applied Linguistics, 29(4), 597-618. doi:10.1093/applin/amn010

Hewett, Beth L. (2015). The online writing conference: A guide for teachers and tutors. Boston, MA: Bedford/St. Martin’s.

​Hewett, Beth L. & Lynn, Robert. (2007). Training ESOL instructors and tutors for online conferencing. The Writing Instructor. Retrieved from http://www.writinginstructor.org/hewett-lynn-2007-09

Hinkel, Eli. (2009). Teaching academic ESL writing: Practical techniques in vocabulary and grammar. New York: Routledge.

Inoue, Asao & Keown, Kelvin. (2015, November). Where is the feedback?: Considering the material conditions of feedback to L2 writers in non-writing classrooms. In J. Bitchener (Chair), The 14th Symposium on Second Language Writing, Auckland, New Zealand.

Linville, Cynthia. (2009). Editing line by line. In S. Bruce & B. Rafoth (Eds.) ESL writers: A guide for writing center tutors (2nd ed.) (pp. 116-131). Portsmouth, NH: Boynton/Cook.

Lippi-Green, Rosina. (2012). English with an Accent: Language ideology, and discrimination in the United States (2nd ed.) (pp. 7–40). New York City: Routledge.

Matsuda, Paul K. (2006). The myth of linguistic homogeneity in U.S. college composition. College English, 68(6), 637-651.

Matsuda, Paul. K. & Cox, Michelle. (2009). Reading and ESL writer’s text. In S. Bruce & B. Rafoth (Eds.) ESL writers: A guide for writing center tutors (2nd ed.) (pp. 42-50). Portsmouth, NH: Boynton/Cook.

Matsuda, Paul K. (2012). Teaching composition in the multilingual world. In K. Ritter & P. K. Matsuda (Eds.), Exploring composition studies: Sites, issues, and perspectives (pp. 36-51). Boulder, CO: Utah State University Press.

Mitchell, Rosamund & Myles, Florence. (2004). Second language learning theories. (2nd ed.). New York: Oxford University Press.

Myers, Sharon. (2003). Reassessing the “Proofreading Trap”: ESL tutoring and writing instruction. In C. Murphy & S. Sherwood (Eds.), The St. Martin’s sourcebook for writing tutors (pp. 284-302). Boston, MA: Bedford/St. Martin’s.

Nakamaru, Sarah. (2010). Lexical issues in writing center tutorials with international and US-educated multilingual writers. Journal of Second Language Writing, 19(2), 95-113. doi: 10.1016/j.jslw.2010.01.001

Powers, Judith. K. (2008). Rethinking writing center conferencing strategies for the ESL writer. In R. W. Barnett & J. S. Blumner (Eds.), The Longman guide to writing center theory and practice (pp. 368-375). (Original work published in 1993)

Powers, Judith. K & Nelson, Jane. V. (1995). L2 writers and the writing center: A national survey of writing center conferencing at graduate institutions. Journal of Second Language Writing, 4(2), 113-138. doi: 10.1016/1060-3743(95)90003-9

Rafoth, Ben. (2009). Responding online. In S. Bruce & B. Rafoth (Eds.) ESL Writers: A guide for writing center tutors (2nd ed.), (pp. 149-160). Portsmouth NH: Boynton/Cook.

Rafoth, Ben. (2015). Multilingual writers and writing centers. Boulder, CO: Utah State University Press.

Salem, Lori. (2016). Decisions...decisions: Who chooses to use the writing center. The Writing Center Journal, 35(2), 147-171.

Severino, Carol. (1993). The sociopolitical implications of response to second language and second dialect writing. Journal of Second Language Writing, 2(3), 44-62.

Severino, Carol & Deifell, Elizabeth. (2011). Empowering L2 tutoring: A case study of a second language writer’s vocabulary learning. The Writing Center Journal, 31(1), 25-55.

Williams, Jessica & Severino, Carol. (2004). The writing center and second language writers. Journal of Second Language Writing, 13, 165-172.

Young, Vershawn Ashanti. (2011). Should writers use they own English? In Laura Greenfield & Karen Rowan (Eds.), Writing centers and the new racism: A call for sustainable dialogue and change (pp. 61-72). Logan, UT: Utah State University Press.