Using Classical Rhetoric in the Age of AI - A Study on Waltzer, Cox, and Heyman’s Study - Testing the Ability of Teachers and Students to Differentiate between Essays Generated by ChatGPT and High School Students

Oct 15, 2026

Uploaded by Elsa

Analyzes essay writing by ChatGPT and high school students using Waltzer, Cox, and Heyman’s study, with a focus on specificity and clarity. Employs Aristotle’s rhetorical theory to examine how both AI-generated and student responses tend to omit detailed content, weakening clarity and argument strength. Argues that reinstating classical rhetoric pedagogy—especially specificity—can improve student writing and help educators detect AI-generated text.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages12

Reading Time30 min

AccessFree to download

Introduction

Literature Review

Using Classical Rhetoric in the Age ofAI: A Study on Waltzer, Cox, and Heyman’s Study“Testing the Ability of Teachers and Students to Differentiate between Essays Generated by ChatGPT and High School Students” By Daniel R. Fredrick ∗, Thomas P. Corbin Jr. ± & Gregory VanderPyl ° This paper analyzes essay writing in AI (ChatGPT) and high school students, focusing on their use of specific details. Discussing the writing examples from Waltzer, Cox, and Heyman’s study, we employ Aristotle’s rhetorical theory to explore how clarity is achieved through specificity in writing. The analysis reveals both ChatGPT and students commonly omit detailed content, affecting clarity and argumentative strength, which affects how humans may detect AI-generated text. The paper suggests reinstating elements of classical rhetoric pedagogy in writing courses, particularly the style device of specificity. This will, we believe, not only improve student writing but also help educators to detect AI-generated text. Introduction Waltzer, Cox, and Heyman—in their study “Testing the Ability of Teachers and Students to Differentiate between Essays Generated by ChatGPT and High School Students”—discovered that students and teachers alike faced difficulty in determining whether essays were written by real students or by ChatGPT, especially student essays which were ‘well-written’(2023). One of the writing prompts was the following: ‘Why does literature matter?’ This question is open-ended and subjective which gives students an opportunity to wrestle with and respond to a complex and qualitative idea. To answer such a question well requires originality of thought, some sophistication in understanding literature, and most importantly, specific statements that reveal a deep understanding of the topic. If the responses are based on platitudes or generalities, I am sure all educators would agree, then the answers fall short intellectually. In their essay, Waltzer, Cox, and Heyman show us various passages; some written by ChatGPT and the others written by students. Although the goal of their study is to discover to what extent teachers and students could determine what was written by AI and what was written by a human, the most striking observation one gathers from the sample passages they provide is how similarlyfaulty are the written responses from both students and ChatGPT. As we reviewed the sample writings from their research, we immediately identified which samples were written by ∗Associate Professor, American University of Sharjah, United Arab Emirates. ±Senior Lecturer, Stephen F. Austin State University-Nacogdoches, USA. ° Senior Instructor, American University of Sharjah, United Arab Emirates. ChatGPT because the responses were, as expected, very generic, tightly organized, and grammatical--without the slightest hint of sounding colloquial or awkward. We also noted that the ChatGPT excerpts were void of those qualities in the prose that feel so human, such as complexity or burstiness (D’Agostino, 2023). Unfortunately, the student prose was equally generic, suggesting a lack of sophistication in absorbing the material or appreciating the genre. In other words, we analyzed parallel deficiencies in both ChatGPT and student writing, namely the generic responses. If the student writing samples were quite similar to the writing samples of ChatGPT, then we can conclude that either the students are writing at least as well as a high powered AI text generator (which they are not) or that ChatGPT is unable to produce text beyond the level of a high school student (which it can). Because of its applicability to generate text in a multitude of fields and in various levels of sophistication and depth, we expect ChatGPT to surpass students. If it does not, then we need to know what is lacking in the features of its writing (or prompting)—not only to be able to increase our ability to detect ChatGPT but also to share this insight with students to imp

FAQs

What main writing feature does the paper identify as missing from both ChatGPT and student essays?

Both kinds of responses are described as missing specifics, which reduces clarity and argumentative strength.

How does Aristotle’s rhetorical theory connect to the study’s findings?

The paper uses Aristotle’s principle that style should be clear, explaining that specificity helps a speech or essay perform its function by indicating ideas clearly.

What pedagogical approach does the paper recommend for writing courses?

It recommends reinstating classical rhetoric pedagogy, particularly training students to use specificity to improve writing quality and support detection of AI-generated text.

Oct 15, 2026

Uploaded by Elsa