Proofreading student writing: a research-based stakeholder tool focused on ethical practice

Abstract

This article presents a research-based stakeholder tool informed by a study of the various types of changes proofreaders may make when proofreading a student text. Whilst the tool can be used to advise higher education students, (non-)professional proofreaders/editors, English for Academic Purposes (EAP) lecturers, writing centre tutors, and university policy makers as to appropriate forms of third-party intervention, it is primarily intended to be used as a framework that assists academics in deciding which kinds of third-party intervention are un/ethical when allowing their students to have work proofread. As such, the stakeholder tool serves a heuristic purpose in that it: 1) displays the various types of interventions proofreaders could make for an academic to consider which interventions they wish to allow, from the lightest-touch (e.g., correcting typos or spelling errors) to the heaviest-touch (e.g., substantial, wholesale rewriting of the text by the proofreader at the level of content); 2) advises academics about un/ethical forms of proofreading intervention; 3) confirms an academic has given permission for a student to seek out proofreading support whilst specifying what in the academic’s opinion is un/ethical in terms of proofreading intervention; and 4) educates and trains stakeholders in academically sound proofreading practices. This stakeholder tool is timely when considering that some current university proofreading policies are poorly worded and ambiguous, and when considering the lack of clarity about or knowledge of university proofreading guidelines by stakeholders such as staff and students.

Introduction

This article describes a research-based stakeholder tool which enables academics to make informed decisions as to whether they permit their students to consult a third party. The tool sets out a range of proofreader interventions informed by an empirical study of the various types of changes proofreaders may make when working on student writing. Using our tool, academics are then able to clearly specify the types of intervention they are or are not happy for a proofreader to make. In addition, the tool serves a formative purpose by educating and training stakeholders in responsible proofreading practices. The stakeholder tool evolved from a doctoral study conducted by Author a. Author a’s doctoral study investigated un/ethical forms of third-party intervention made to student texts for assessment. The stakeholder tool was devised in response to her doctoral supervisor’s claim, here her co-author, that many British universities do not have proofreading policies, with those that do providing little information as to ethically (in)appropriate forms of intervention (Harwood (2018, p. 477); see also Davis 2024, for further evidence of vague UK proofreading policies). In support of such an assertion, Harwood (2022, p. 117) cites the University of Sheffield policy, which on the one hand does not forbid proofreading but on the other neglects to furnish clear definitions of permissible third-party interventions and emphasises that students must submit work of which they are the sole author.

In order to devise a research-based stakeholder tool, Author a began by consulting various taxonomies of the types of interventions student proofreaders have been found to make by drawing upon a number of empirical studies which either surveyed reported proofreader practices (Kruger & Bevan-Dye 2010; Harwood et al. 2009, 2010, 2012) or compiled their taxonomies by providing proofreaders with an authentic essay written by a second language speaker of English to proofread (Harwood, 2018). Having identified these different forms of third-party intervention, the second phase of Author a’s research, which further informed the stakeholder tool, involved conducting her own textual analysis of the proofreading changes made to a native Spanish-speaking doctoral student writer’s thesis who was completing a four-year EdD degree at a UK university, as well as interviewing four stakeholders: 1) the aforementioned EdD student writer; 2) the EdD student writer’s freelance proofreader who in this case acted on a non-commercial basis, undertaking the proofreading free of charge as she was friends with the writer; 3) a senior lecturer based in the field of English Language and Linguistics at the same UK university as the student; and 4) a retired senior lecturer, also formally employed at the same institution who was based in the area of Urban Studies and Planning and currently proofreads tertiary level students’ written work for assessment. The purpose of the interviews was to present the types of changes that a proofreader can make based on Author a’s textual analysis as well as other forms of amendment which were not made, in order to raise the stakeholders’ awareness as to the different forms of intervention that a third party could undertake. In turn, Author a investigated her stakeholders’ views on the ethical (in)appropriateness of the different types of third-party intervention. In collaboration with Author b, Author a’s textual analysis and exploration of stakeholders’ views regarding the ethicality of various types of proofreader interventions allowed her to present a tool which academics can use to decide for themselves how far proofreaders of their students’ work should be permitted to intervene.

Having discussed the evolution of our stakeholder tool, the following section explains the importance of said tool with regard to current proofreading policies and practices.

Literature review

Before discussing the importance of and the need for a stakeholder tool, we firstly explain our definition of the term ‘proofreading’ which follows Harwood et al. (2009, p. 167):

Types of help (whether voluntary or paid) that entail some level of written alteration to a “work in progress” (i.e. work that will contribute towards an undergraduate or postgraduate degree, or which may be published).

The all-encompassing nature of this definition was selected as research has shown that so-called ‘proofreaders’ vary widely in their practices and while some restrict themselves to a narrower range of interventions, others are willing to go much further, engaging in more substantial rewriting or reworking of students’ texts (Harwood, 2018). Indeed, by following Harwood et al.'s (2009) definition, we allow for our stakeholder tool to define proofreading in broad terms, reflecting the reality that some ‘proofreaders’ engage in heavy-touch interventions, by asking academics whether they agree to their students receiving third-party interventions on a very wide range of issues, some of which (e.g., rewriting at the level of content and argumentation) are far beyond the traditional proofreader remit described by Harwood et al. (2009, p. 168), i.e., “a final run-through” before submission, “changing a misspelling here or putting in a punctuation mark there”.

Moving onto the importance of a proofreading stakeholder tool, such a research-based instrument is timely when considering: 1) the lack of policy and uniformity in British universities concerning third-party support; 2) uncertainty amongst students and academics as to university proofreading policies; 3) ethical concerns regarding students consulting proofreaders; and 4) the varying conceptualisations of a proofreader, as will now be discussed.

Regarding the lack of proofreading policies in British universities, this is especially apparent when considering that some are rather ambiguous and therefore problematic and confusing. For instance, the University of Manchester’s (n.d.) proofreading statement provides only two examples of prohibited “text editing,” i.e., “adding or rewriting of phrases or passages within a piece of student’s work”, and students are advised that a third party is not to change “content or meaning,” both of which are clearly large areas open to interpretation. The University of Sheffield’s (2024) proofreading stance is even less comprehensive, stating that the use of paid proofreading services is not endorsed but if students do choose to consult third parties, they do so at their own risk, thus making it unclear exactly what proofreaders are permitted to do. Furthermore, viewing proofreading policies across universities exposes a lack of uniformity amongst institutions, with some permitting proofreading (e.g., Oxford Brookes University 2015; Sheffield Hallam University, n.d.; University of Essex 2024) and others completely forbidding the practice or adopting a less permissive approach (e.g., Manchester Metropolitan University, n.d.; Swansea University 2021; University of Leeds 2015). Additional to problematic policies and a lack of uniformity, proofreading guidelines also need to be disseminated more effectively, as exemplified by Cottier (2017) in a student context. Cottier’s research, which included an assessment of guidelines regarding the expectations and experiences of postgraduate students at two universities in Queensland, found that even though 79% of her student participants were aware of their universities’ editing guidelines, only 36% had in fact consulted them. This led Cottier to the clear conclusion that such policies need to be communicated more effectively to both students and supervisors. Indeed, some academics as well as students are clearly unaware of university proofreading guidelines, as highlighted by Salter-Dvorak’s (2019, p. 124) research concerning the extent to which current language policies create social inequality for L2 master’s students in British universities. Salter-Dvorak mentions that one of her lecturer participants encourages his students to seek out informal proofreading help which is contrary to the course handbook that stipulates writers should self-edit rather than contact a proofreader.

Regardless of the clarity or otherwise of various universities' proofreading guidelines, there is evidence of considerable discord and disquiet with reference to proofreading, meaning that debates about its ethicality are needed, and that stakeholder tools which demystify proofreading and empower academics to discuss and delimit its scope are needed. There are numerous arguments against third-party proofreading support, which include:

• 1) Proofreading being a form of “spoon-feeding gone mad” that depends on students’ ability to afford such services (Baty 2006);

• 2) The apparent unfairness of students who can afford proofreading services seemingly obtaining higher marks (see De Oliveira 2020, pp. 249-250; McKie 2019; Turner 2018, p. 95);

• 3) Proofreading breeding writer dependency as writers may be unable to reproduce under exam conditions the same level of work submitted with the aid of a third party (Harwood et al. 2010, p. 56);

• 4) The potentially far-reaching consequences of proofreading when considering that a large proportion of students, especially those studying towards a doctorate, later progress to academic supervisory roles themselves (De Oliveira 2020, p. 246), meaning that such writers who are overly dependent on proofreaders may not be equipped to act as advisors to their own students about how to write well;

• 5) Concerns that students may accept all proofreader interventions made to their text via Microsoft Track Changes without paying attention to persistent errors. This lack of attention could make proofreading ethically questionable because the student’s objective may be only for their text to be ‘fixed’ by the proofreader, rather than to gain a formative or pedagogic experience (Harwood, 2019, p. 19); and

• 6) Some proofreaders may not be sufficiently trained to differentiate between “editing and substantive editing”, and even qualified proofreaders may be unaware that such forms of intervention are contrary to university procedures (Lines 2016, pp. 375-376).

Ethical concerns surrounding proofreading are compounded when considering the varying conceptualisations of a proofreader’s role amongst and between students, proofreaders, and academics. For instance, Conrad’s (2021) study concerning (un)acceptable forms of academic behaviour highlighted that some students found it appropriate to have a parent or classmate help proofread their work, citing it as a form of peer review in which the writer could decide to accept or reject a reviewer’s comments, whereas another student claimed it would be unacceptable to produce one’s own work by using the input of a third party or even consulting a website. From a proofreader viewpoint, Harwood et al.'s (2009) investigation of 16 proofreaders’ proofreading beliefs, practices, and experiences found that whilst grammatical and spelling corrections were viewed by the proofreaders as acceptable and accuracy and content level changes were not, a noticeable lack of agreement occurred regarding the delineation between language and content forms of intervention. Further, from an academic’s perspective, Salter-Dvorak’s (2019, pp. 124-128) research cited above highlighted how two L2 graduate students, Lijuan and Farideh, studying at the same institution received conflicting advice regarding the use of proofreading services from their lecturers. Lijuan’s lecturer, Rob, advocated informal proofreading services for language issues as his own feedback focused solely on content. In contrast, Farideh did not receive any advice concerning the use of a proofreader from her lecturers. Further, when Farideh asked her course leader and personal tutor, Martina, to view a draft, the latter stated that this course of action would be contrary to university policy and instead advised forms of self-editing such as reading an essay aloud to identify errors. The consequence of such different approaches to proofreading resulted in Lijuan being a far more confident writer as language issues had been addressed pre-submission, whereas Farideh was naturally very upset when receiving feedback on work in which the marker highlighted that language issues impeded understanding. Further evidence of contradictory understandings of the ethicality of proofreading emerges from Harwood's (2023) study of the conceptualisations of proofreading by 32 lecturers, 34 English language tutors, and 56 students, and from Harwood's (2024) more recent foregrounding of all three parties’ uncertainty as to what types of interventions are ethically (un)acceptable for proofreaders to engage with.

From the above survey of literature focused on a lack of proofreading policies, stakeholder ambiguity and uncertainty with regard to students consulting a third-party, ethical concerns, and the varying conceptualisations of the proofreader’s role, it is apparent that a research-based stakeholder tool which raises academics’ awareness of the range of potential proofreading interventions, and subsequently enables academics to prescribe interventions they are happy for a proofreader to make, could be of value. Such a heuristic is especially important when considering the potentially educative benefits of proofreading for student writers, enabling them to develop their academic literacy skills. In this vein, McNally and Kooyman (2017, p. A148) emphasise that if universities are to progress by accommodating L2 students from diverse cultural backgrounds with varying writing needs, a more flexible approach should be adopted rather than expecting learners to succeed solely by their own efforts. Similarly, Haggis (2006) argues that universities should question areas of the curriculum which prevent students’ progress, rather than assuming learners from diverse backgrounds are already equipped with the skills required to complete traditional forms of assessment. This view is supported by Shaw (2014), who states that many L2 students are not provided with adequate services at their universities to support language needs. Importantly, Harwood et al. (2012, p. 575) advise that proofreaders can help student writers by adopting a “leveller” role in which they reduce the disadvantages that L2 students experience. As such, leveller proofreaders assist students in meeting university standards and accordingly put L2 speakers on a level playing field with that of the L1 cohort, thus allowing markers to assess L1 and L2 students’ work in terms of content rather than being distracted by the latter’s possibly faulty grammar. Indeed, proofreading can be educative, as discussed by Harwood (2022, pp. 123-124) with regard to:

• 1) promoting reflection and learning when students analyse errors;

• 2) drawing students’ attention to persistent mistakes;

• 3) writers learning language and its rules as well as language learning strategies;

• 4) proofreaders teaching academic writing conventions;

• 5) proofreading providing personalised opportunities to learn which complement other forms of support such as that provided by writing tutors; and

• 6) proofreaders using educative self-correction techniques that include underlining errors or correction symbols.

We conclude our literature review with some thoughts about the rapid emergence of AI-generated writing tools, such as ChatGPT. Some readers may feel the rise of this software will rapidly render debates about the rights and wrongs of proofreading obsolete, and that discussions around academic integrity should now focus fully on AI and LLMs, rather than on the impact of proofreaders on student writers’ texts. While we do not dismiss the undoubtedly profound influence AI software is already exerting on academic writing (cf. Allen & Mizumoto 2024; Naghdipour 2022) (and not just on the writing of novice academic writers), we believe that writers will continue to seek out critical friends and collaborators to help enhance their texts and their products, and that therefore debates around the ethics of proofreading remain pertinent. Further, when considering the benefits of proofreading coupled with the previously mentioned issues surrounding its practice, a tool which informs academics of the range of interventions proofreaders can potentially make, as well as providing the means for stakeholders to reflect upon the ethical acceptability of these interventions will enable lecturers to reflect on the extent to which they are content for third party proofreaders to intervene on students’ work for assessment.

Methods

Designing the stakeholder tool

The methods employed in our study focused primarily on designing a stakeholder tool in order to measure the types and amount of (non-)interventions made by a proofreader. The following subsections provide a brief overview of the stakeholder tool, explain how it was devised through a textual analysis taxonomy, and conclude with a presentation of the stakeholder tool itself. In designing the stakeholder tool, our academics were presented with a range of possible interventions to raise their awareness of the different types of change proofreaders can make. Having raised academics’ awareness of the range of possible interventions, our tool then allows them to decide which changes they would be happy for a proofreader to undertake and to confirm permission for a student to seek out proofreading support. The following sections now discuss the design of our stakeholder tool in more detail with regard to: 1) a textual analysis taxonomy we devised to determine different forms of lighter and heavier touch proofreading intervention that would be included in the tool; and 2) the presentation of the stakeholder tool itself.

Textual analysis taxonomy

In order to compile our stakeholder tool, we firstly devised a taxonomy showing the full range of possible intervention types, from lighter to heavier touch. The taxonomy is presented in concise form in Table 1; full details, together with examples of each type of intervention in the taxonomy, are provided in Appendix A. Our taxonomy was devised primarily based on those of Harwood (2018) and Kruger and Bevan-Dye (2010). Harwood's (2018) framework, adapted from a taxonomy devised by Willey and Tanimoto (2012) which included elements of Luo and Hyland’s (2016) instrument, was selected based on the fact that it focuses specifically on the proofreading of student writing rather than on the proofreading of other genres (e.g., monographs) and accordingly deemed suitable for our own study. Kruger and Bevan-Dye’s (2010) framework was also incorporated into our taxonomy due to its very finely grained nature, with 66 tasks fitting into four categories of copyediting, stylistic editing, structural editing, and content editing. Kruger and Bevan-Dye based these four categories on those proposed by Mossop (2007, cited in Kruger and Bevan-Dye 2010, p. 159) as they believed that each category clearly distinguishes the varying types of editing task. With further regard to Mossop, his book is designed for editors working with texts for publication, and for some, the terms ‘editing’ and ‘proofreading’ overlap, whereas others find them to be discrete. However, our broad definition of ‘proofreading’ encompasses both traditional conceptualisations of proofreading (e.g., punctuation corrections) and editing (e.g., structural and content changes), and in this article we use ‘proofreading’ as a catch-all as it is the most commonly employed term in our UK context to describe third-party interventions on student writing. Further, our taxonomy features styles of intervention identified by Cottier (2017) with regard to editing methods and methods of raising queries and comments. In addition, when devising the taxonomy, we analysed the proofreading changes made to a Spanish-speaking doctoral student’s texts which allowed us to further refine our framework.

Table 1 Concise textual analysis taxonomy. This Table presents in concise form a taxonomy showing the full range of possible intervention types, from lighter to heavier touch. Full details of the taxonomy, together with examples of each type of intervention, are provided in Appendix A

Having completed our analysis of the proofreader’s interventions, we then presented our participants with the different types of intervention that a proofreader can make as will now be discussed.

Presenting the stakeholder tool

Having first raised lecturers’ awareness of the full range of possible intervention types, from lighter to heavier touch, in order to then advise academics of potentially un/ethical third-party intervention types, we displayed types of proofreading intervention which the reader can view under the Stakeholder Tool Questionnaire in Section 3 of the full version of our stakeholder tool shown in Appendix B. For ease of reference, Fig. 1 shows a more concise excerpt from Section 3 to illustrate how academics are advised of potentially un/ethical forms of intervention which a proofreader could make for the academic to then consider whether they would wish to permit such a change. The figure itself shows that we firstly denote the type of intervention, i.e., Minor Addition, and acknowledge the source of the descriptor (point 1: ‘Minor addition involves the proofreader adding one to five words to a student’s text’), which in this case was from Harwood (2018, p. 519). Following this, points 2 and 3 provide a brief explanation as to why both types of change may generally be viewed as acceptable (‘the meaning is not altered and the text flow or cohesion is improved’). The figure itself shows two examples obtained through our textual analysis which provide academics with clear examples of what constitutes Minor Addition. Having presented the intervention and explained why, in this case, most stakeholders would deem such a change to be ethically acceptable (for evidence of this judgement of ethical acceptability of Minor Additions, see Harwood, 2023), the supervisor is still provided with the option of deciding whether they do or do not approve of the proofreader making the type of intervention in focus.

Fig. 1

Less Ethically Contentious Interventions. This figure shows a more concise excerpt from Section 3 to illustrate how academics are advised of potentially un/ethical forms of intervention which a proofreader could make for the academic to then consider whether they would wish to permit such a change

Having advised academics as to empirically-grounded evidence of stakeholder consensus around un/ethical forms of intervention and soliciting academics’ (non-)approval for the various types of intervention, Section 4 of the stakeholder tool feeds into a ‘Stakeholder Agreement Permitting Proofreading’ document as shown in Fig. 2, based on that of the Society of English-language Professionals in The Netherlands (SENSE) (SENSE 2016) which can be seen in context by referring to Appendix B.

Fig. 2

Stakeholder Agreement Permitting Proofreading. Section 4 of the stakeholder tool feeds into a ‘Stakeholder Agreement Permitting Proofreading’ document as shown in Fig. 2, based on that of the Society of English-language Professionals in The Netherlands (SENSE) (SENSE 2016) which can be seen in context by referring to Appendix B

In order to serve an educative purpose, the Stakeholder Tool Questionnaire could be used to train stakeholders in responsible forms of proofreading practice. For instance, student writers could receive training in academic skills classes, and workshops could be created for various audiences: proofreaders, EAP lecturers, writing centre tutors, academics, and university policy makers. A possible lesson/workshop plan adapted from Conrad (2021) that could be used to educate stakeholders as to appropriate forms of proofreading practice would proceed as follows. First, stakeholders read two concise, accessible texts which argue in favour of and against the ethical acceptability of proofreading practices taken from the higher education press. Stakeholders then consult their institution’s proofreading policy to discuss the extent to which the policy addresses the ethical issues raised by the two texts. Such attention to institutional policy is urgently needed, given that Harwood (2023) found that only a minority of the lecturers, English language teachers, and students in his sample were aware of what their institutions’ policies were regarding the ethical (un)acceptability of proofreading. Next, stakeholders’ attention is drawn to the consequences of academic dishonesty to highlight ethically unacceptable forms of proofreading practice. Following this, stakeholders are presented with example proofreader interventions extracted from the Stakeholder Tool Questionnaire and discuss whether and to what extent they find each intervention example ethically (un)acceptable. After their discussion, the views of lecturers, English language tutors, and students on the ethical acceptability of these same intervention types are shown to participants, this research evidence being taken from Harwood's (2023) findings. The workshop ends with further discussion by workshop attendees to compare/contrast participants’ views with the views of Harwood's (2023) participants, and a wrap-up summary by the workshop trainer relating to academic integrity and proofreading. This pedagogical cycle is included in Section Five of Appendix B (‘Using the Stakeholder Questionnaire to Educate and Train Stakeholders’).

Discussion

We have presented a stakeholder tool focusing on academic integrity within the context of proofreading. Our tool has four purposes, as follows:

First, we wish to alert academics to the full range of potential proofreader interventions. Lecturers who will be assessing essays, dissertations, or theses may erroneously believe, when they see adverts for ‘proofreaders’ around their universities, that all these proofreaders confine their interventions to light-touch grammar, spelling, and syntax interventions associated with traditional conceptualisations of proofreading (“a final run-through” before submission, “changing a misspelling here or putting in a punctuation mark there”; see Harwood et al. 2009, p. 168). As a consequence, they may believe that proofreading student writing is ethically unproblematic and does not surface issues relating to academic integrity. However, a number of studies over nearly two decades (e.g., Richards, 2024; Harwood, 2018, 2023; Harwood et al. 2009; Lines 2016; Turner 2011) have demonstrated that although some proofreaders may take a narrower view of their remit, others are prepared to engage in much more substantial (e.g., content- and organizational-level) interventions, which many lecturers and university policy makers would find ethically problematic, believing such interventions to violate expectations around academic integrity and student authorship. Alerting academics to the full palette of possible proofreader interventions, then, including those associated with both lighter- and heavier-touch styles, is a crucial first step achieved by our stakeholder tool.

Second, we seek to raise stakeholders’ awareness of the views of various parties (lecturers, English language tutors, and students) as to the un/ethical acceptability of these different styles of intervention, as evidenced by Harwood's (2023) empirical study, while giving academics the latitude to disagree with these views, to determine the extent to which they will permit their students’ work to be proofread.

Third, our instrument serves to act as confirmation of an academic’s approval for students to consult third-parties whilst delineating un/ethical forms of intervention, creating a written record of the forms of ‘proofreading’ which are to be allowed or debarred. This written record also importantly reminds student writers that it is their responsibility to carefully examine (and hopefully learn from) the proofreader’s interventions and make the necessary changes to their texts before submission; they are not permitted to abdicate responsibility for the final version of their text to the proofreader.

Fourth, our tool serves an educative purpose by means of stakeholder workshops which can be targeted at different groups (e.g., student writers, academics, university policy makers). These workshops seek to: i) enhance stakeholders’ awareness of the academic integrity issues associated with the proofreading of student writing through debating texts from the higher education press which argue about the rights and wrongs of proofreading; ii) raise stakeholders’ awareness of their institutions’ proofreading policies; and iii) provide a safe space in which the ethics of proofreading can be debated and reflected upon, and institutional proofreading policies can be critiqued.

With regard to the first three objectives, by raising stakeholders’ awareness of the different forms of ‘proofreading’ taking place, by alerting stakeholders to views regarding un/ethical forms of intervention, and by asking all stakeholders to read and sign the Stakeholder Agreement Permitting Proofreading document, our tool aims to ensure shared responsibility for proofreading and for what is done in the name of proofreading, and to encourage communication about ethical proofreading amongst and between stakeholder groups. An awareness of appropriate forms of proofreading intervention and ensuring that all parties consent to the use of a third-party proofreader are important issues when considering the numerous ethical concerns associated with proofreading mentioned in the literature review. Concerning the fourth and final objective of the stakeholder tool serving an educative purpose, the information gleaned from the questionnaire could allow departments to determine what type of proofreading support academics feel is ethical and what they are comfortable with. Such information could then inform the stakeholder workshops mentioned above so as to eventually standardise proofreading practices and avoid issues that Salter-Dvorak’s (2019) two students (Lijuan and Farideh) encountered at their university. Further, should the questionnaire reveal that academics are happy for students to consult third-parties, decisions could be made by policy makers as to whether students can consult a professional proofreader (i.e., an established, commercial service) or an informal proofreader (e.g., family, friends, classmates), or whether they would want to adopt a more involved approach by managing proofreading practices institutionally—for instance through the provision of in-house services or outsourcing to external proofreading/editing agencies (see Harwood, 2019, p. 39 for arguments concerning in-house/external proofreaders). With further regard to the stakeholder tool serving an educative purpose, the workshops described and discussed above were devised in accordance with Bretag and Mahmud’s (2016, pp. 467-469) argument concerning the importance of educating staff and students regarding academic integrity. Indeed, the authors advise that rather than focusing on negative aspects regarding practices which should be avoided, greater emphasis should be placed on educating and promoting values that institutions wish to nurture. This is very much the ethos of our stakeholder tool, in that greater emphasis should be placed on educating and training with regard to ethically appropriate forms of proofreading.

Conclusion

At the heart of our approach is an emphasis on encouraging ethically appropriate forms of proofreading and communication amongst stakeholders to encourage good practice, whilst acknowledging that challenges would undoubtedly arise in establishing a stakeholder tool in the teaching and learning cycle of an institution. In such an environment, the experience of using the tool and reflecting upon its outcomes could be used to inform an eventual proofreading policy in a contextually-sensitive manner which involves all parties. As such, the stakeholder tool is designed to facilitate rather than to prescribe or to proscribe. Our tool should enhance academics’ knowledge of the range of possible proofreading interventions, raise awareness of how other parties view the ethicality of these interventions, and allow them to make informed decisions concerning suitable forms of third-party intervention which they are happy for proofreaders to make to their students’ texts. Policymakers could also consider adapting a tool like ours to enable stakeholders to reflect upon the ethicality of the use of AI and LLMs as writing aids. There are, of course, important differences between proofreading and AI use—not least i) that no third party need be consulted when using AI; and ii) that AI will correct/rewrite/supply text, rather than merely flagging it up as potentially problematic or commenting on it. Nevertheless, the ethical questions we have raised throughout this article remain with reference to LLMs, and stakeholders must debate them to properly inform policy. A tool like ours can enable such debate, in concert with stakeholder discussion groups and workshops.

With regard to the limitations of our study, Author a’s doctoral study was undertaken during COVID-19 in which she encountered considerable difficulty in recruiting participants, meaning that the textual analysis taxonomy was piloted on a small sample. As such, future studies could address such concerns through a greater range of stakeholder involvement that would allow for a more robust analysis as to the effectiveness of said taxonomy and the findings could be used to better inform the resulting stakeholder tool. The tool also requires extensive testing in order to validate it and to ensure that it is sufficiently user-friendly with reference to the diverse range of stakeholder groups it encompasses.