By Jun Akiyoshi, The Pennsylvania State University, and Rajwan Alshareefy, University of Delaware
Both of us, Jun and Rajwan, have similar backgrounds. We worked as EFL/ESL teachers, studied in an interdisciplinary area of Composition and Applied Linguistics, enjoyed talking about research and practice of writing education, and most importantly, we worked together at the same writing center when we were graduate students. Even after we earned our Ph.D.s, we continued to engage with, learn about, and research (writing) education. Throughout the years, we often talked about the theory and praxis of (college) writing, second language education, among many others. Our conversations became more heated when AI-powered tools became available and free to the public in 2022. Over the past few months, we continued to read and learn about AI and writing in both composition and language education fields, and we wanted to share our thoughts in a way that is accessible to a broad audience, something that a blog post can offer.
From our academic communities, we sensed that many practitioners (including writing center tutors) were concerned (and sometimes anxious) about how AI (most prominently ChatGPT) may shape the future of writing education as a field of study and a profession. When Jun attended the 2023 NCPTW in Pittsburgh, he witnessed people’s strong concerns about ChatGPT. Rajwan found it worth investigating that his colleagues had mixed feelings and conflicted ideas about ChatGPT. As with every new technology, ChatGPT had both problems and benefits for writers. One major concern in our academic circles was whether jobs such as composition teachers and writing center tutors would be affected by this new technology. In this blog post, we respond to this concern by reviewing the negative and the positive aspects of ChatGPT and argue that writing center tutors can only be assisted but not replaced by this new technology. As Composition and Applied Linguistics scholars, our argument is based on studies published in the area of (second) language education as well as writing studies.
Technology Tools and Language/Writing Education
Reviewing published articles in the language education field, we noticed that research on technology in relation to language and writing education is not necessarily new. The wide use of (AI) technology in language and writing education is well documented in Alharbi’s (2023) review in which he analyzed 104 peer-reviewed empirical studies and found that those studies focused on certain technologies like automated writing evaluation tools, automated written corrective feedback tools, and automated machine translation tools. For many of us who work either as writing tutors or writing teachers, Grammarly is a well-known automated written corrective feedback tool.
Let us introduce one research study that investigated Grammarly in the writing center context. Zhang and colleagues (2020) examined 43 ESL writers’ perceptions of Grammarly in comparison to in-person tutorial sessions. They found that Grammarly and in-person tutoring have different strengths and weaknesses and suggested a “both/and” stance instead of an “either/or” stance with Grammarly and the writing center tutorials. For example, writing center tutorials are helpful when ESL students want to receive more individualized comments and more accurate feedback on their papers’ lower-order concerns (e.g., grammar, spelling, word choice, etc.). On the other hand, when restrictions such as time, space, and lack of resources exist, ESL writers can use Grammarly, because it’s relatively more available and easy to access. We found Zhang and colleagues’ suggestion quite valuable as they indicated that “we encourage ESL writers to explore all writing resources available to them and incorporate them into their writing process to their advantage, be them automated checkers or human tutors” (p. 41). Following their suggestion, we can tell that the existence of the electric writing assistant tool does not necessarily replace writing tutors or teachers. As can be seen from these brief examples, writing researchers, instructors, and tutors have worked, coped, studied, and utilized (AI) technologies for many years without being negatively affected by it as professionals.
ChatGPT and Writing: Some Issues at This Point
Regarding ChatGPT and writing, let us first talk about some issues that writing tutors and teachers have to consider. First, as many other researchers, we strongly encourage ChatGPT users to critically evaluate the generated texts. For example, Imran and Almusharraf (2023) warned that chatbots would show “challenges of lack of originality, inaccuracies, too generalized responses, and poor logic flow in writing prompts” (p. 8). Buriak and colleagues (2023) also recommended writers to “not use text verbatim from ChatGPT” mainly because the chatbot might have “reused text from other sources, leading to inadvertent plagiarism” (p. 4092). Kohnke and colleagues (2023) showed a similar concern and indicated that “its responses are not entirely original but paraphrases of sources that have not been appropriately cited (i.e., plagiarism)” (p. 544). ChatGPT is said to hallucinate and what this means is that ChatGPT can generate fake author information (Kim, 2023) and fake paper titles and publication IDs (Alkaissi & McFarlane, 2023). With this in mind, we should encourage writing tutors to make students aware of the potential problems of using this technology.
Secondly, as another potential problem, we also want to acknowledge that human tutors’ knowledge of genre-specific convention (or their knowledge of writing broadly) cannot be replaced by ChatGPT yet. To give an example of the importance of a human audience, in Johnson-Eilola and colleagues’ (2024) study, a set of AI-generated instructions (COVID-19 home testing kit) were tested and examined by human professional writers (e.g., experts in technical writing) and found to be lacking on multiple levels. Specifically, the instructions (1) failed to follow “several common conventions that make instructions easier to use” (p. 205) (e.g., lacking numbers for the steps to follow), (2) missed some pieces of information and also included “inaccurate information” (p. 205) (e.g., lacking a detailed explanation about the depth of swab insertion), and (3) failed to include any illustrations or images that should help users better understand the instruction. Even though AI-generated instructions may “[afford] a rudimentary usability” and guide readers with certain instructions (Johnson-Eilola et al., 2004, p. 205), the abovementioned problems cannot be ignored. Johnson-Eilola and colleagues concluded that while ChatGPT might be used as a collaborator for writers, it would not replace technical writers yet. This implies that even when ChatGPT may help students brainstorm and initiate their writing as a collaborator, the role and contribution of a human tutor (who is well-versed in a given genre conventions) is essential in the writing process and cannot be substituted by AI tools.
Thirdly, when it comes to audience awareness, ChatGPT may not necessarily be able to read students’ texts and offer a realistic audience perspective. Tutors can make writers aware of both explicit and implicit audiences (i.e., readers who physically read the text and readers that writers target during their writing) (Fontaine-Iskra, 2017), something that ChatGPT may fail to offer. Helping students identify the potential reactions and impressions of the audience in the tutoring sessions is a valuable experience for them to enhance the quality of their writing.
Some Benefits of ChatGPT as a Writing Assistant Tool
After reviewing some problems that writing tutors and teachers should keep in mind for their students, let us look at some positive aspects of ChatGPT in writing. One of the primary benefits of ChatGPT in the language learning context is its ability to simulate “authentic interactions” in which learners can ask the chatbot to “identify the meaning of words in context, correct and explain language mistakes, create texts in various genres” (Kohnke et al., 2023, p. 539). ChatGPT can also play a role as a writing assistant (Imran & Almusharraf, 2023). For example, referring to its performance in editing English grammar, Kim (2023) claimed that ChatGPT’s “ability to produce more refined sentences surpass those of the unedited ones” (p. 2) and Kim also emphasized that ChatGPT can edit a paragraph instantaneously and is “available for free to everyone” unlike other kinds of language editing services (p. 2). ChatGPT can help writers create more accurate and consistent texts (Imran & Almusharraf, 2023). Considering these benefits, we think that ChatGPT can also potentially be beneficial to students with various linguistic backgrounds (e.g., Kwon and colleagues’ study in which they found an AI-powered chatbot was efficient in improving EFL writers’ writing performance). One way writing center tutors can use this editing capacity is by letting AI tools take care of students’ common requests of “checking grammar” and “accuracy issues” while giving tutors an opportunity to focus on higher order concerns such as organization, creating thesis statements, etc. Although Younis and colleagues (2023) claimed that Grammarly Premium did a better job in grammar and style correction, ChatGPT can still be a good option because (1) it is free to use (Kim 2023) and (2) its interactive feature can help writers spontaneously ask questions related to writing (Kohnke et al., 2023).
By reading published articles, we also found that ChatGPT can be used in students’ research writing. Let’s see what Chauke and colleagues (2024) reported in their study. They investigated 10 South African graduate students and examined their perceptions of ChatGPT in their research process. The study findings showed that there were four perceived benefits: (1) refining research topics by discussing with ChatGPT, (2) using ChatGPT “for editing and paraphrasing their academic research” (p. 52) to avoid errors in writing to meet the academic standard of written English, (3) using ChatGPT to brainstorm and consider a researchable topic, and (4) saving time for the literature review process (i.e., searching relevant sources and writing reviews) by using ChatGPT.
Final Thoughts: ChatGPT in the Writing Center Context
Although the articles we cited in this blog post are limited in number, we were able to obtain some basic information about the drawbacks and benefits of AI in writing education. Based on the articles we reviewed, ChatGPT’s ability to generate human-like interactions can be helpful when the writing center is not available for students (either due to space restrictions or schedule conflicts). Students can use ChatGPT to (1) identify grammatical mistakes in their writing, (2) obtain some explanations about English grammar (if needed), (3) brainstorm ideas for their writing projects, and/or (4) explore and organize information sources needed for their projects. That being said, thinking about the potential drawbacks of ChatGPT, student writers should not depend on ChatGPT only. They are encouraged to visit the writing center at some point in their writing process to obtain some feedback from human tutors to enhance the quality of writing.
Tutors are valuable resources thanks to their roles as explicit and/or implicit audience, with critical eyes and deep knowledge and experience in the writing of different genres. We see the relationship between AI and human tutors as a “both/and” relationship instead of an “either/or” one. ChatGPT can be introduced to the writing center context only as an additional resource. One way to utilize this technology as a resource in writing centers is to experiment with, create, test, revise, and utilize ChatGPT prompts that can offer the maximum benefit for the writing center purposes. For example, Deans and colleagues (2023) introduced some real-case scenarios where ChatGPT was used in tutoring sessions. Writing center directors and tutors can share those kinds of scenarios and prompts that they used. If they share those scenarios and prompts within and across writing centers, writing center directors and tutors can collaboratively accumulate knowledge and information about how to use ChatGPT in practical settings.
Works Cited
Alharbi, W. (2023). AI in the foreign language classroom: A pedagogical overview of automated writing assistance tools. Education Research International, 2023. https://doi.org/10.1155/2023/4253331
Alkaissi, H., & McFarlane, S. I. (2023). Artificial hallucinations in ChatGPT: Implications in scientific writing. Cureus, 15(2), e35179.
Buriak, J. M., Akinwande, D., Artzi, N., Brinker, C. J., Burrows, C., Chan, W. C. W., Chen, C., Chen, X., Chhowalla, M., Chi, L., Chueh, W., Crudden, C. M., Di Carlo, D., Glotzer, S. C., Hersam, M. C., Ho, D., Hu, T. Y., Huang, J., Javey, A., … Ye, J. (2023). Best practices for using AI when writing scientific manuscripts. ACS Nano, 17(5), 4091–4093.
Chauke, T., Mkhize, T. R., Methi, L., & Dlamini, N. (2024). Postgraduate students’ perceptions on the benefits associated with artificial intelligence tools on academic success: In case of ChatGPT AI tool. Journal of Curriculum Studies Research, 6(1), 44–59.
Deans, T., Praver, N., & Solod, A. (2023, August 1). AI in the writing center: Small steps and scenarios. Another Word. https://dept.writing.wisc.edu/blog/ai-wc/
Fontaine-Iskra, A. (2017). Tutors’ column: Making audience visible: Readership and audience in writing centers. WLN: A Journal of Writing Center Scholarship, 41(7), 25–28.
Imran, M., & Almusharraf, N. (2023). Analyzing the role of ChatGPT as a writing assistant at higher education level: A systematic review of the literature. Contemporary Educational Technology, 15(4), ep464.
Johnson-Eilola, J., Selber, S. A., & York, E. J. (2024). Can artificial intelligence robots write effective instructions? Journal of Business and Technical Communication, 38(3), 199–212.
Kim, S. G. (2023). Using ChatGPT for language editing in scientific articles. Maxillofacial plastic and reconstructive surgery, 45(1), 13.
Kohnke, L., Moorhouse, B. L., & Zou, D. (2023). ChatGPT for language teaching and learning. RELC Journal, 54(2), 537–550.
Kwon, S. K., Shin, D., & Lee, Y. (2023). The application of chatbot as an L2 writing practice tool. Language Learning & Technology, 27(1), 1–19.
Younis, H. A., Mohammed, O., Muthmainnah, Sahib, T. M., Akhtom, D., Hayder, I. M., Salisu, S., & Shahid, M. (2023). ChatGPT evaluation: Can it replace Grammarly and Quillbot tools? British Journal of Applied Linguistics, 3(2), 34–46.
Zhang, J., Ozer, H., & Bayazeed, R. (2020). Grammarly vs. face-to-face tutoring at the writing center: ESL student writers’ perceptions. Praxis: A Writing Center Journal, 17(2), 33-47.
Jun Akiyoshi is currently teaching Rhetoric and Writing courses at the Pennsylvania State University. Before starting his current job, Jun was working as a writing tutor at Kathleen Jones White Writing Center at Indiana University of Pennsylvania. His research interests in relation to the Writing Center involve “writing center as a resource for writing teacher education.” In addition to his teaching, Jun also works as a social media editor of Writing Commons to contribute to teachers and students in writing classrooms.
Rajwan Alshareefy is a postdoctoral researcher in the First Year Writing program. He currently teaches technical and First-year writing courses at the University of Delaware. Rajwan has a background in teaching English and composition in and beyond the United States. His research interests include social identity, critical discourse analysis, academic writing and transnationalism. Rajwan is interested in the way writing practices can be influenced by the writers’ sense of belonging to places beyond nation-state borders.