Wei Xu
The University of Arizona
This assignment asks students to use provided templates of prompts to feed ChatGPT for performing a recontextualization task named “Bilingual Genre Redesign,” meaning rewriting their existing text-based literature review into an open genre in two language versions. The ChatGPT assignment creates opportunities for students to critically engage with the language dynamics involved in the recontextualization process by evaluating whether and how the GPT-rewritten texts align with their understanding of shifted rhetorical context. Through seeing the limitations of AI in catching the fluidity needed in redesigned texts, students may develop a keener and critical awareness of AI use in their writing.
Learning Goals
Original Assignment Context: Undergraduate-level first-year writing class for English as an additional language (EAL) learners
Materials Needed: A completed academic text-based literature review and generative AI (GAI) tools (i.e., ChatGPT) for students to use
Time Frame: One class varying from 50-75 minutes. Depending on the length of work time that the instructor offers to students, the implementation of this activity may not take the entire class session time.
Overview
This AI-facilitated activity was designed to answer students’ frequently asked questions about a writing project of writing across genres and languages in a genre-based first-year writing course for English as an additional language (EAL) writers. The original project asks students to analyze the features of a public genre in two languages and re-write their academic literature review into this public genre, preparing a version of it in two languages. Some questions from students about this project include: “Why cannot the two language versions be literal translations of each other if the content is the same?” “Can I copy one paragraph from my literature review and use it for the chosen public genre after just simplifying the language of that paragraph?” In this AI activity, students were guided to select one short paragraph from their literature review that they plan to test, ask ChatGPT to generate examples of genre redesign and bilingual redesign using the prompts provided by the instructor, and critically examine the language in the AI-generated samples in small groups, thus making sense of the concepts such as target audience and linguistic context.
This activity has been experimented once in the first-year writing course for English as an additional language (EAL) writers in a previous semester. The outcome is that students gained an idea of the complications involved in cross-genre and cross-language writing. Through examining the AI-generated samples, students also realized the limitations of AI in addressing dynamics in specific rhetorical contexts, such as target audiences’ background knowledge and specific communicative purposes.
This activity can be used to support a writing project of writing across genres and languages that asks students to redesign a text-based literature review into an open genre (i.e., blog or infographic) in two language versions. The students who took the course had literature review as the target learning genre prior to the genre redesign assignment. Redesigning a text-based academic genre into an open multimodal genre in two languages helps tap into multilingual writers’ powerful resources of multilingualism to foster their rhetorical flexibility in shuttling across genres and languages.
Alternatively, this activity may be used as a standalone low-stakes activity for multilingual writers to teach basic writing concepts, such as rhetorical situations and target audience. The activity can be either a group activity or an individual activity. The time frame required is flexible, depending on whether the instructor expects it to be only a practice or a formal assignment with final deliverables.
Step 1: Use AI to produce a genre redesign sample
Firstly, I divided students into groups. It is suggested that students with the same first language (L1) who selected the same target public genre for redesign be grouped together. Each group should not have more than 4 people. For certain students who do not share the same L1 with any others, they could experiment this activity individually.
Next, the group members decided on the purpose and target audiences for bilingual genre redesign. Group members designated one member to copy one short paragraph/section from their literature review that they would like to rewrite into a section for the target public genre (e.g., blog post / infographics). They then used the following prompt to ask ChatGPT to produce a rewrite sample.
Step 2: Use AI to produce a bilingual redesign sample
The students copied the AI-generated text that they received from Step 1. They then used the following prompt to ask ChatGPT to produce a bilingual translated sample.
Step 3: Focused group analysis and retry
On a Google doc, the students copied the AI-generated texts from Step 1 and Step 2. The group members looked at the same Google doc and discussed the following guiding questions:
In this step, most students would find some places that they do not feel satisfied about in the AI-generated samples. I then encouraged them to re-try using the following suggested prompts. Of note, students could also develop their own prompts for this step, based on their evaluation of the texts.
Step 4: Review the analysis result and plan for next stage
In groups or individually, the students used their linguistic, cultural, and rhetorical knowledge to manually revise the AI-generated texts to fix the places they still feel unsatisfied about, especially the L1 version. I also asked them to document the rationale why they decided to make a change there.
After manual revisions, individually, the students reviewed their rationale of revision and the finalized samples. Then, they planned for the next step of the Bilingual Genre Redesign project. Suggested guiding question are provided below:
I would like to acknowledge the Writing Program at the University of Arizona for developing the writing project of Bilingual Genre Redesign. More specifically, the idea of this cross-genre and cross-language writing project comes from “Tardy, C. M., Sommer-Farias, B., & Gevers, J. (2020). Teaching and researching genre knowledge: Toward an enhanced theoretical framework. Written Communication, 37(3), 287-321.”