The doctoral writing conversation: establishing a generic doctoral writing programme

ABSTRACT Over the past few decades, the number of people enrolled in doctoral study has increased dramatically across the world. In practical terms, this has meant that universities now receive increasingly diverse students with regard to ethnicity, age, language, culture, and background preparedness for higher degree study. Students can, and often do, begin their doctorates with scant understanding of the precise expectations and rigorous demands of thesis writing. Yet, regardless of academic discipline, successful completion of a doctorate requires a written thesis. To help students master thesis writing requirements, a proliferation of self-help writing books, blogs, specific writing techniques, and programmes have emerged. This paper describes an approach developed at a New Zealand university where a generic doctoral writing programme, the Doctoral Writing Conversation, has evolved to make explicit to students the implicit language understanding that accomplished academic writers use to produce text. Utilising the idea of language as a tool to mediate understanding, the paper will explore how the programme is structured and functions but will also describe some of the insights I have gained along the way.


Introduction and background
In the past two decades there has been an increasing (and increasingly diverse) population of doctoral students globally (Norgrove & Scott, 2017;OECD, 2017). In New Zealand the number of international doctoral students has risen so dramatically over the past ten years that by 2015, 45% of total PhD enrolments were from international students. This has characterised New Zealand as having the strongest growth rate of international PhD students among the OECD countries (Berquist, 2017).
Accommodating large numbers of doctoral candidates from different cultural, educational, and linguistic backgrounds within a relatively short time frame has challenged 'traditional' approaches to doctoral supervision in which one or two disciplinary experts have mentored students who, for the most part, have then worked independently. Coupled with this, many New Zealand and Australian universities face an aging workforce, retirements, and the casualisation of the work-force, all of which have contributed to a shrinking pool of eligible supervisors being responsible for increased numbers of doctoral students.
In addition to these factors, the research doctorate is characterised by the production of a single extended piece of independent writingthe thesis. Its purpose is to outline a focus, issue, or problem, review the related work of other scholars, explore the problem (usually, but not always accompanied by data collection, analysis, and synthesis), and present an overall story or argument within an accepted (recognisable) structure and format. Ultimately the research thesis is expected to make an original contribution to human knowledge. Yet, only two or possibly three external examiners read (judge) the finished written thesis. As a piece of formal assessment, then, the written research thesis must be characterised as very high stakes indeed.
There are other essential aims of the doctorate beyond the completion of a written thesis. It is through the process of writing that a sense of self as an independent academic scholar emerges from the concept of 'self' as student, dependent on others to signal approval or rejection of ideas. However, Paré (2011) in his insightful discussion of thesis writing requirements comments that supervisors are often at a loss as to how to assist students with their written work, which can result in writing advice that is vague or suggestions for improvement that are provided without explanatory justifications. Aitchison and Guerin (2014) believe that the most useful assistance rendered by supervisors is the provision of sample sentences or suggested passages, assistance with re-writes, or co-authorship of publications. However, within the current doctoral environment, such timeconsuming approaches to writing development could prove onerous or impossible for supervisors to manage.
To help address the demands of thesis writing, there has been a rapid growth in the number of 'DIY'/ self-help guidebooks, but as Kamler and Thomson (2008) critiqued, such books usually present doctoral writing practice as a series of linear steps to be mastered rather than acknowledging the reality of research writing with its drafting, redrafting, and refinement of argument as meaning emerges. All of these concerns, then, argue for more systematic, supportive, institutionally-based approaches to thesis writing development.

The case for generic writing advice
Given the increasingly large, diverse cohort of doctoral students, additional pressure on academic staff to supervise students, and the rhetorical demands of thesis writing, a case can be made for generic writing development to be offered by specialists with expertise in writing. Carter (2011) argues convincingly that doctoral writing is its own genre and that knowledge of it derives from 'our participation in the communicative activities of daily and professional life' (Carter, 2011, p. 731). Carter (2011) also discusses the essential role of interdisciplinary conversation to build conceptual understanding, as epistemology and ontology are most visible at cultural boundaries. Citing Cuthbert, Spark and Burke (Carter, 2011, p. 732) she states that in multidisciplinary conversations participants' sense of self as disciplinary exponents is enhanced. 'Discussion at the borderlands illuminates discipline-specific practice ' (2011, p. 733). Through having students converse with each other about their own conceptual struggles, they can provide insights to each other about how to become 'unstuck' when they encounter writing problems (Kiley, 2009, p. 302). Importantly, approaching writing and meaning-making as social activity can utilise effectively the diversity and massification of higher education (Aitchison, 2009).
Diversity in the group can support the acceptance of a variety of views on a person's work and can also help students develop awareness of the variety of audiences who might encounter their writing (Ferguson, 2009;Guerin et al., 2013). Multidisciplinary groups can add to discipline-based doctoral education by providing breadth in students' development; in this sense it complements rather than compromises disciplinary education (Cuthbert, Spark, & Burke, 2009).
Supervision or facilitation of thesis writing groups can take a range of forms. In much of the research on generic thesis writing programmes, groups have been initiated by the institution, according to a more or less structured format, and have been strongly guided by a staff member (Aitchison, 2003;Cuthbert et al., 2009;Ferguson, 2009;Larcombe, McCosker, & O'Loughlin, 2007). Aitchison (2003) defends such an approach when she states that 'the role of the facilitator is crucial in this small group approach as students must actively and sensitively participate as developing writers as well as learn how to critically appraise the work of others' (p. 97). Of importance is that the facilitator assists students within a low-stakes and non-judgmental environment (Larcombe et al., 2007).
Additional factors that support the success of thesis writing groups are participants' experience of success, for example through the accomplishment of writing goals (publications, thesis chapters, for example) (Maher, Fallucca, & Mulhern Halasz, 2013) and the participation in a structured environment. Such structure could be just time and space that has been set aside (Maher et al., 2013), or it could involve a more elaborate plan, which includes expectations, performance benchmarks, accountability, methods of working, or enabling conditions, including a facilitator (Holmes, Birds, Sealy, Smith, & Wilson, 2010).
A strong case has been made across many publications for the value of pan-university generic doctoral writing programmes, but what is seldom included in academic literature is discussion of how to structure, organise, and develop them as learning environments. Kiley (2009) in her overview of threshold concepts in the doctorate briefly outlines some strategies for supporting students, but without extensive detail. Lee (2018) discusses the role of frameworks as a way to visually map the elements of thesis writing, and does include some helpful suggestions for how to address different parts of the framework, but again, without a great deal of specific detail. Yet, drawing upon ideas from cognition, activity, and mediation, there is much to guide the development of a generic thesis writing approach.

The Thesis Writing Circles (TWC)
Conceptualising writing support In January 2008, I became the director of Student Learning an educative service unit at the University of Waikato, New Zealand. It provided learning development for students across all faculties and at all levels of study from first year undergraduate to the doctorate. However, at that time, there was scant centralised thesis writing support for doctoral students anywhere at the university (although there were some Faculty-specific workshops), and I undertook an investigation of what programmes were provided elsewhere in New Zealand. All university websites provided descriptions of their centralised thesis writing support (particularly in at the doctoral level)except ours; we provided no formal, institution-wide writing assistance at all for thesis students. I believed that there was significant room for improvement, which led to the development of a variety of cross-campus initiatives to support doctoral candidates, including workshops for students at different stages of their thesis (beginning, middle, and (almost) completion). We also introduced a weekly, two-hour, cohort-based writing and research discussion forumthe Thesis Writing Circles (TWC)a name adopted from a successful doctoral writing programme elsewhere (Aitchison, 2003).
Based on our investigation of other programmes, we wanted the TWC to be generic (appealing to students from any Faculty), facilitated by writing specialists (in Student Learning), and structured (weekly). Beyond that, in the beginning stages of the TWC initiative, we were planning 'on the fly'. Also, although I was not entirely sure how to structure the programme on a weekly basis, I knew what I wanted to avoidthe lecture. Certainly theme-based workshops are valuable for distributing information, but I wanted to create a physical and conceptual environment in which students worked together to understand and produce text. We did not entirely abandon workshops though; we ran a few sessions to provide some background information about writing, but our main goal was to have students bring focused pieces of their own writing to share and for which they would provide peer feedback. We also encouraged students to start their own writing / peer editing groups outside of the times when we met them.
In July 2009 the TWC was launched, but it quickly became evident that a programme in which students worked on their own writing, coupled with opportunities for peer editing, was not what they wanted. During peer writing sessions attendance was extremely low (3-5 students), and on some occasions no one at all attended. The workshop sessions proved to be more popular, but our initial themes were somewhat 'hit and miss' and even for them, attendance varied. As for student self-managed peer writing groups, they did not occur at all. By the end of 2009, we concluded that although the programme had not been a failure, it had certainly not been a great success either. As a result, I conducted a small-scale qualitative evaluation of students' and staff perceptions of doctoral writing in general and the TWC in particular with the aim of better understanding how the programme could be reorganised to better meet student need. I remained convinced that a cross-disciplinary, peer-based, doctoral writing forum was a worthy and achievable goal.

The TWC becomes the DWC
Two key findings emerged from the study (Johnson, 2014a) and which influenced our ongoing generic thesis writing development approach. One was the essential role of cross-disciplinary conversation in helping students clarify their ideas before they commenced writing. This insight, the most important from the study, relates to the idea of language as a tool to mediate and develop academic scholarly identity. I would summarise this finding as 'talking to think; thinking to write'. Secondly, while it was apparent that most supervisors believed that student cohort groups could be valuable for alleviating social isolation, they were unsure as to whether or not a 'Thesis Writing Circle' could provide an appropriate environment for writing development. Upon reflection, we realised that the programme's name was problematic and that it had led to various misunderstandings about the programme's intention and activities. At least one supervisor imagined his students sitting around in circles editing each other's work, which she felt would be time-consuming and not particularly valuable. As a result the programme name was changed to Doctoral Writing Conversations, which I believed more closely reflected the perspective and functions of a collaborative, discourse-based doctoral writing environment. A revised DWC format was shaped by what students said they wanted and by what supervisors believed could be valuable.

Building institutional acceptance
Beyond developing strategies to provide generic writing support for thesis students, it was also clear that we needed supervisory 'buy-in' for our programme. Clearly, helping students understand and manipulate the rhetorical features of text to develop argument and shape conceptual insights to inform their thesis can complement supervisors' work. Yet, during the TWC evaluation, one supervisor related that her student (who had attended a TWC session) had argued a methodology (at variance with her approach) during their supervisory meeting. Apparently the student's insight had developed through conversation with other students at the TWC, prompting the supervisor to state that, 'their way of thinking about things is not our way of thinking about things'; he stated that the cross-disciplinary approach in the TWC was 'nonsense'.
Similarly, I encountered suspicion when discussing the TWC initiative at an international conference in New Zealand. An audience member assured me that generic writing groups were fine as an idea, but challenged me with the question, 'What are you telling our students?'. He was skeptical about the value of sending his students along to such a programme. Certainly, I was aware that unless supervisors encouraged their students to attend our programme and supported it, we would never reach the numbers of students who could potentially benefit from writing development. In spite of publicising our sessions to students, I realised that the programme was unlikely to achieve its full potential without supervisor endorsement and active support.
A strategy that I adopted was to invite groups of supervisors (between 2 and 4 people) to attend a DWC session at the beginning of each month; we called these 'invited guest conversations'. I established a conversational theme, which could relate to writing or the research process (for example, 'working in interdisciplinary teams', 'research ethics in the field', or 'presenting the thesis in an engaging, readable format'). Supervisors were asked to briefly share their thoughts on the topic, and then the session was opened for student questions. The format worked well in that students could interact, on a collegial basis, with experienced researchers (other than their supervisory panel) and (often) from outside their discipline. As for the supervisors, they met and interacted with a much larger group of thesis students, from various disciplines, than they would normally encounter, plus they met other disciplinary experts from outside their own field. Finally, supervisors had an opportunity to experience first-hand the DWC programme and gain insight into how it functioned and what it aimed to achieve. Over the next three years, this approach extended understanding of the DWC across Faculties and into a wide variety of disciplines through the interactions with a range of supervisors. The academic staff that participated always reported how much they had enjoyed meeting and conversing with a multidisciplinary group of doctoral students.
In addition to the guest conversations we introduced two, one-day, off-campus writing retreats (mid-winter and end of year). There is a substantial literature on the merits of writing retreats and many suggestions for how to organise them. These range from highly prescriptive, timed, task-oriented sessions to more flexible approaches in which students set their own learning goals and work at their own pace. We adopted the latter style (a loose structure), but also included a 'break-out' room if a student wanted to discuss a particular writing issue.

Reassessing the DWC approach
At the end of 2014, we reassessed the DWC format, which I felt had become 'stale'. Student participation in the 'invited guest conversations' had declined, and when I sought evaluative feedback, students reported that the guest sessions, while interesting enough, did not provide focused guidance about thesis writing. Students reported that they wanted more opportunities to encounter and discuss language in use. They also wanted increased opportunities to assemble and write. Student reflections supported findings emerging from my two-year participation in a research project to examine threshold concepts in doctoral writing (Johnson, 2013(Johnson, , 2014b. Two key findings emerged from my case study students did not possess deep conceptual understanding of what research 'writing' actually means, and they did not have an agentic view of themselves as scholars. Their understanding that meaning emerges from the repetitive and cyclical practices of writing was often poorly developed. Yet, in all academic research understanding emerges as new ideas are discussed, clarified, written, and refined. Meyer and Land (2005) refer to this as time in 'liminal' space before an crossing intellectual threshold is crossed, leading to deep understanding.
The guest conversations were abandoned, and the weekly DWC sessions became much more focused on activities to exemplify the particulates of thesis writing. We ceased offering any sessions on research processes. A seminal work by Brown, Collins, and Duguid (1989) provided invaluable guidance on how to shape learning activities within our situated environment to stimulate learning and develop knowledge. Encountering rules or new ideas, even if a student can recite them by rote, does not ensure that they can be adopted or adapted (used) in new situations. In the case of writing a thesis, written language is the tool by which the task is accomplished.
For me, these ideas reinforced that workshops in which information was presented to students, even if there was discussion, were less useful than activities in which meaningmaking was the explicit goal. Students also requested more, and longer, off-campus writing opportunities. Brown et al. (1989) discuss the use of language as a tool to mediate understanding and contribute to learning. Of interest to this discussion of how to help thesis students become independent researchers and scholars is the idea that 'activity, concept, and culture are interdependent' (Brown et al., 1989, p. 33). If we accept the notion that writing a thesis is at least as much about developing an independent sense of self as a scholar, part of a larger academic culture, then questions of how to structure activity (in this case thesis writing), concept (emerging from research), and culture (that of the academy) become critical. Learning must encompass all three components, yet as Brown et al. (1989) aptly state 'teaching methods often try to impart abstracted concepts as fixed, well-defined, independent entities that can be explored in prototypical examples and textbook exercises' (p. 33). Brown et al. (1989) also discuss the idea of language as a tool to mediate understanding, a concept echoed by Bakhurst (2009) who postulates that mediation acknowledges that human behaviour is far more complex than a simple reaction to a stimulus, but that all human activity is shaped by 'artifacts that are created to prompt or modulate action' (Bakhurst, 2009, p. 199). The question then becomes how can we best shape learning environments through the use of conceptual and physical artifacts, such as language, text, group discourse, classroom space, and activities (for example) to maximise learning? Further how can such learning contribute to the shaping of students' emerging academic scholarly identity?

The reinvigorated DWC
These ideas shaped the design of the DWC so that sessions are dialogic, active, and reflective of scholarly work within the academy. What follows are three brief examples of talking to think, thinking to understand, and understanding to write in the genre of the doctoral thesis.

The writing requirements of different parts of the thesis
In my own supervisory practice, I frequently found that students had great difficulty writing some thesis chapters, which I considered straightforward. This included the Introduction and the Methodology. Having difficulty with other chapters such as the Literature Review or Conceptual Framework did not surprise me, and there are excellent examples to guide supervisors and students in those tasks (Bitchener, 2018a(Bitchener, , 2018b. Upon reflection, I realised that students have trouble with the Introduction and Methodology chapters because at a deep level, they do not understand their rhetorical functions, and thus they cannot write them. They cannot imagine what to include. This activity is group-based and is guided by initial questions and discussion to elicit what understanding students already have about the purpose of the different chapters (they consider the two chapters separately and sequentially). Then a second set of questions guide students through critical examination of text examples. Each group is given a set of Tables of Contents (ToC) from nine completed doctoral theses from different disciplines across the university. Using a ToC as a mediating tool for discussion is a powerful device, as the ToC presents a concise visual display of the entire thesis structure from its Chapters, to its main themes, to all its thematic sub-headings. In short, the ToC presents a clear snap-shot of thesis structure around which discussion can be based.
A final set of questions is designed to raise student awareness of the common features of the chapter, regardless of discipline. From the examples, students discuss such variables as chapter length, key headings, and the sequence of presentation of the Introduction text; they also discuss who should decide what material to include. Through discussion, students realise that there is scant variation in Introduction structure across disciplines, but there is considerable variation between introductory chapter content in a thesis with publication as opposed to a thesis as monograph. As the practice of thesis with publication becomes more popular in our university environment, students find it very useful to view and discuss the distinctions in how chapters are structured.
In contrast to the Introductions, there is considerable variation in Methodology chapters, across thesis types, academic disciplines, and of course, research approach (quantitative, qualitative, mixed method, historical or philosophical investigation, for example). Again the ToC texts function as visual aids (mediating tools) to critical reading and discussion to help raise student awareness of expected thesis writing conventions.

Responding to supervisor feedback
A key aspect of supervision is providing written feedback to students, and there has been extensive research and publication about how, when, how much, and in what form feedback should be given (Bitchener, 2018a;Bitchener & Storch, 2016;Taylor, Kiley, & Humphrey, 2017). Within the research literature however, there is scant advice to students about how they should respond to feedback, although there are some blog discussions of how to respond to overly negative feedback (Hayton, 2015). Anecdotally, from my conversations with DWC students, they have very few ideas for how to interpret supervisor feedback, much less how to respond to it in a systematic manner. In this 'talking to think' exercise, we attend to both issuesthat of interpreting the nature of feedback, what to do with that feedback, and then how to structure a response for discussion with the supervisor.
The mediating tool around which we build discussion in this exercise is actual supervisor feedback (mine), which I share with students. With written permission from one of my successfully completed doctoral students, I anonymise the thesis text from one of her drafts and extract about five pages of that text with my feedback. Again, working in groups, the students read through the feedback and categorise the comments as being surface-level errors, indicating missing ideas, or questioning the flow of the argument. Other comments are requests for clarification of meaning or are general remarks on overall thesis structure. Through the activity of reading, discussing, and categorising feedback comments, students gain insight into the range of feedback that a supervisor could provide. They then discuss which comments should be addressed first and which would be the more time-consuming to address.
Finally I provide example grids to show how students can document supervisor comments, the actions they have taken to address comments, and any questions that remain outstanding. These synopsis grids can then form the basis of the student's discussion with their supervisor at their next meeting. In this example, we have mediating text (real student writing and supervisor feedback), discussion of the meaning of the feedback, and the development of a plan for how to respond to that feedback.

Writing retreats / focused structured writing sessions
The final examplethe writing retreatis somewhat different from the previous two, although the ideas informing the retreats are rooted in activity and social mediation for writing development. Growing from a single-day, off-campus activity held twice yearly, the writing retreats have expanded to four, two-day, off-campus sessions. This change has been in response to student request. The retreats follow a similar model described by Murray (2015), in which social time (morning and afternoon teas and a light lunch) is interspersed amongst periods of silent writing. Some opportunities for group reflection on goal setting and progress are provided at the start and end of each of the two-day retreats. The physical space itself consists of a large room with tables, WiFi access, kitchen facilities, and a small 'break-out' room. Students bring their own technology and work materials.
Although it might appear that the writing retreat activity is solitary, this is not the case. There is an important social aspect to writing, with others, in a physical space that is removed from the distractions of one's usual environment. Students describe the atmosphere as inspirational because although everyone is engaged in silent writing, they are aware of the social presence of the others, which serves to stimulate their thinking. Students also state that being with peers helps them stay focused on their own writing. If breaks from writing or assistance from a learning developer are needed, then the break-out room provides the space and opportunity for talking without disturbing the group.
The social meal times provide opportunities to mix and mingle, build contacts, and establish social networks, all of which are vital aspects of academic life. Also, students really appreciate that they are the centre of attentionmade to feel specialand that their writing success is important to the university. This activity, then, is an example of writing mediated through shared physical space, scheduled discussion, and the social presence of others.

Stocktake of the DWC: is the approach effective?
Considering the early days of infrequent or low student attendance, the DWC now attracts between 18 and 35 thesis students each week (depending on the topic being covered). Given that thesis students lead busy lives and there is no requirement for them to attend the sessions, their ongoing support of a generic thesis writing programme provides evidence of its efficacy. Simply stated, students vote with their feet, and if they did not find the DWC useful, they would not attend.
The DWC with its activity framework, in which talking, social interaction, and working with texts, mediates student understanding of thesis writing requirements has become a well-known resource for both doctoral and research Masters students at our university. One of its key advantages is that it makes explicit many aspects of the internalised writing knowledge that academic supervisors have already developed. The pan-university nature of the DWC also contributes to writing development, as students learn about academic approaches and paradigms from disciplines outside of their own. Student feedback indicates that the DWC provides them with the skills to more productively engage with their supervisors. Alerting students to important aspects of not only what to discuss with supervisors, but how to structure their discussions, can lead to more productive supervisory meetings. In addition, better understanding of thesis writing requirements can lead to increased student confidence. Another key strength of the DWC is that it can disrupt the hierarchical nature of supervision (in engaging, productive, and positive ways) to complement the supervisory process. This productive disruption can restore students' confidence in their writing ability and maintain their writing productivity.
Another interesting outcome is that the dialogic approach of the DWC raises awareness that ambiguity is normal during the research process. Students realise that they are not alone if they feel lost during their doctorate; everyone at some point feels intellectually inadequate and unable to write. Writing together and talking about writing helps students live with uncertainty and manoeuvre through liminal space to make writing progress. Students report that the DWC is their place; they find strength and acceptance within the group, which helps them to continue writing even when they are unclear about what they are discovering in their research.
Where to next with the DWC?
We are always seeking to improve what we can provide for students through the DWC.
One key shortcoming, which we know exists, is that (like many other such programmes) it cater for on-campus and often full-time students only. We have on several occasions tried to address this deficit, and students have certainly requested online access to the DWC. Some sessions have been voice-recorded and made available to students, and for some months we trialled a Skype link with students on another campus. We do post session notes and text-based discussion documents in our LMS, but all attempts to broaden the DWC into an online space have so far proved unsatisfactory.
There are a few examples successful programmes elsewhere. Maher et al. (2013) describe two writing retreat approaches (one extended and the other brief) in which participants would write, but from wherever they were located on the scheduled day. Virtual contact strategies were created so that participants could provide and receive mutual support during this time. Haas (2014) in her overview of standalone writing groups refers to a variety of technology-mediated asynchronous and synchronous forms. In our case, however the DWC with its activity-based, tool mediated, discussion format has yet to be established in any meaningful way in an online format. We could do so synchronously, but this would still not address the needs of part-time students with daytime responsibilities. But, in our next DWC 'upgrade', more innovative ways to support off-campus students will be our goal.