The Investigation Related to the Role of ChatGPT in English

Language Teaching

Tony You Li

Computer Science Department, University of Victoria, 3800 Finnerty Rd, Victoria, BC, Canada

Keywords: ChatGPT, AI, Language Learning.

Abstract: As Chat Generative Pre-trained Transformer (ChatGPT) undergoes continuous iteration and development, its

impact on language learning is increasingly evident. This necessitates a comprehensive summary to

understand this evolving influence. The study employed qualitative research methods to evaluate ChatGPT's

influence on language learning across four areas: writing, reading, listening, and speaking. For writing, the

investigation centered on determining whether ChatGPT effectively aids non-native English speakers in

writing English papers. This involved exploring various aspects that ChatGPT can enhance. The study also

sought to understand the potential challenges that may arise. The assessment of reading skills centered on

ChatGPT's effectiveness in comprehension, particularly in breaking down complex texts and aiding

summarization, alongside concerns about the depth and accuracy of its interpretations. The combined

assessment of listening and speaking skills aimed to gauge ChatGPT's utility in oral communication. This

included evaluating the potential of ChatGPT to aid in developing conversational abilities, understanding

various accents, and enhancing fluency. The study revealed that ChatGPT significantly aids writing and

reading, substantially improving structuring ideas and understanding complex texts. The combined listening

and speaking skills assessment yielded less definitive results, primarily due to the relative novelty of this

application area for ChatGPT. However, despite ChatGPT's effectiveness, it also raises concerns, including

ethical issues, potential biases in training data, and challenges in interpretability. The study concludes that

careful integration of ChatGPT is essential, and future research should focus on evaluating a range of Artificial

Intelligence (AI) tools for language education.

1 INTRODUCTION

As of 2023, English dominates the global language

landscape, with approximately 1.456 billion speakers

worldwide (Dyvik 2023). It serves not only as a

crucial medium for communication among diverse

cultures but also as a gateway to vast opportunities in

education, business, and everyday life. Consequently,

numerous non-native English speakers embark on

their English learning journey each year. Achieving

proficiency involves overcoming challenges such as

complex grammar, extensive vocabulary, and cultural

differences from their native language. The variety of

accents and dialects further complicates mastering

listening and speaking skills. Many works, including

(Tambuskar 2022), have highlighted AI's role in

tailoring content to individual learning styles and

needs. This innovation enables more efficient English

comprehension and bridges cultural and linguistic

gaps, providing a customized and practical learning

experience.

The term "Artificial Intelligence" surfaced in the

1950s (McCorduck et al 1977), signifying the field's

birth. Initially focusing on basic problem-solving and

computational tasks, AI's scope expanded

dramatically in the 1990s, spurred by the internet and

computational advancements, heralding the era of

machine learning. This evolution culminated in one

of the most significant recent advancements: the

Transformer model. Introduced as a game-changer in

AI, particularly for Natural Language Processing

(NLP) (Kalyan et al 2021), its innovative language

processing approach laid the foundation for advanced

AI applications. A notable example is Chat

Generative Pre-trained Transformer (ChatGPT)

(Kocoń et al2023), developed by OpenAI.

ChatGPT, leveraging NLP, Supervised Learning,

and Reinforcement Learning (Roumeliotis and

Tselikas 2023), integrates these techniques to refine

Li, T.

The Investigation Related to the Role of ChatGPT in English Language Teaching.

DOI: 10.5220/0012807900004547

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Data Science and Engineering (ICDSE 2024), pages 429-433

ISBN: 978-989-758-690-3

429

its language responses based on user interactions,

establishing itself as a versatile AI tool. Since its

inception, ChatGPT has achieved rapid growth and

adoption in fields like accounting, human resources,

education, and health care (Fui-Hoon Nah et al 2023).

Its transformative impact in education is pronounced,

enhancing grammar accuracy, expanding vocabulary

(Hong 2023), and boosting learner engagement and

motivation (Zhai 2022). Considering its rapid

advancement and significant impact on English

language learning, a comprehensive evaluation of

ChatGPT in the educational sector is essential. This

review should not only gauge its effectiveness in

language skill enhancement but also explore its

potential to modify teaching methods. Understanding

these aspects is crucial to comprehending ChatGPT's

role in reshaping English language learning.

The remainder of the chapter is organized as

follows: Section 2 delves into the methodology

employed in this study, examining the approach and

tools used to assess ChatGPT's impact. Section 3

presents a discussion of the findings, analysing how

ChatGPT influences English language learning and

its implications in the educational landscape. Finally,

Section 4 concludes the paper, summarizing the key

insights and suggesting directions for future research.

2 METHODS

2.1 The Introduction of ChatGPT

ChatGPT, a product of OpenAI's extensive research

in artificial intelligence, represents a significant leap

in AI, particularly in natural language processing. The

core idea behind ChatGPT is to create a system that

can understand and generate human-like text based on

a given prompt, making interactions with machines as

natural and intuitive as talking to a human. At its

heart, ChatGPT is designed around the principle of

understanding context and generating responses that

are not only relevant but also coherent and

contextually appropriate. This involves a deep

understanding of language syntax, semantics, and

pragmatics (Ray 2023).

The technical framework of ChatGPT is built upon

the revolutionary Generative Pretrained Transformer

(GPT) architecture (Zhavoronkov 2022). This

involves training the AI on a massive corpus of text

data, enabling it to learn patterns, nuances, and the

structure of human language. The model is then fine-

tuned to generate accurate, informative, and

contextually relevant responses to the user's input.

The procedure of ChatGPT is illustrated in Fig. 1

(Medium 2022)

Figure 1: The working procedure of ChatGPT.

2.2 The Investigation of ChatGPT on

Writing

The integration of ChatGPT into language learning

has revolutionized how writing skills are developed.

As an advanced AI tool, ChatGPT offers substantial

assistance in enhancing various aspects of English

writing, such as grammar and syntax improvement,

revision, translation, and idea generation (Imran and

Almusharraf 2023).

Numerous studies, such as Schmidt-Fajlik’ study

(2023), demonstrated that ChatGPT is an ideal tool

for non-native English speakers to write English

papers. Despite its conveniences, many problems and

challenges still exist when using ChatGPT for essay

writing. One significant concern is the reliance on AI

for language construction, which might lead to a lack

of deep understanding and originality in the learners'

writing (Ningrum 2023). While ChatGPT can assist

with structure and grammar, overdependence on AI

tools could hinder the development of learners'

writing skills and critical thinking abilities.

Another problem is the ethical challenges.

ChatGPT is trained on a vast text corpus of existing

data, including published materials. Therefore, when

generating texts, ChatGPT may produce outputs that

closely resemble or inadvertently include excerpts

from articles published by other researchers (Li et al

2023). This becomes a critical issue in academic

settings, where plagiarism and the unauthorized use

of copyrighted materials are taken very seriously.

Moreover, there is a concern about accuracy. As

mentioned previously, since ChatGPT is trained on

existing data, it may occasionally process and use

poor-quality datasets. This reliance on pre-existing

data can be particularly problematic in fields that

ICDSE 2024 - International Conference on Data Science and Engineering

430

require high levels of precision, such as medical

education and clinical study (Jeyaraman et al 2023).

In these areas, the accuracy of information is not just

a matter of academic integrity but also of utmost

importance for patient safety and effective healthcare

practices.

2.3 The Investigation of ChatGPT on

Reading

The application of ChatGPT in enhancing reading

comprehension skills offers a new dimension to

language learning. ChatGPT's ability to analyse and

summarize text gives learners a unique tool to deepen

their understanding of complex reading materials.

The study of Chui (2023) and Young (2023) shows

that ChatGPT aids in breaking down challenging

texts, explaining complex concepts, summarizing

long passages, and rephrasing them. This mainly

benefits non-native English speakers struggling with

nuanced or dense academic texts. The AI's capacity

to clarify and elucidate content helps improve

comprehension and retention. ChatGPT also can

condense long passages into shorter summaries while

retaining the essential information and themes. This

text summarization feature is handy for learners

dealing with extensive reading materials and

improves learners' verbal ability (Hadwin et al 1999).

However, ChatGPT and similar systems in reading

comprehension have their challenges. For example,

they often encounter issues such as hallucination and

inaccuracies (Zhang et al 2023), where the generated

content may only sometimes be factually correct or

relevant. Imperfect evaluations are another concern

(Zhang et al 2023), as these systems are typically

optimized and assessed using reference summaries

that might not accurately capture the quality of the

generated content. Additionally, they need more

controllability (Zhang et al 2023), providing only

generic summaries instead of customizable ones

tailored to specific user needs or queries.

2.4 The Investigation of ChatGPT on

Listening and Speaking

On September 25, 2023, OpenAI introduced a

significant update to ChatGPT, equipping it with the

ability to speak and listen. This advancement has

opened new possibilities for practicing speaking and

listening skills. However, the effectiveness of

ChatGPT in enhancing these abilities is still a novel

concept, with only a limited number of studies

currently available to validate its impact on

improving learners' speaking and listening

proficiency.

From the insights provided in Young and

Shishido’s study (2023), it is evident that ChatGPT

has potential in this realm. This research highlights

ChatGPT's ability to engage learners in realistic

dialogues. This feature is particularly beneficial for

practicing conversational skills, as it allows learners

to experience and respond to spoken English in

various contexts, enhancing their speaking and

listening abilities. The AI's capacity to understand

and respond to voice inputs means that learners can

receive immediate and interactive feedback on their

pronunciation, fluency, and usage, which is crucial

for developing practical speaking skills.

Moreover, as explored in the study (Xing 2023),

integrating ChatGPT with convolutional neural

network models presents a promising approach for

enhancing English listening skills. By engaging in

real-time conversations with AI-driven systems,

learners can actively practice their listening

comprehension, getting exposed to various accents,

speech patterns, and vocabularies. This immersive

experience helps them become more familiar with

natural English language usage, significantly

improving their listening capabilities.

3 DISCUSSION

The application of ChatGPT in English language

learning has shown transformative impacts,

particularly in enhancing reading and writing skills.

ChatGPT's advanced NLP capabilities have

significantly improved comprehension, grammar

accuracy, vocabulary expansion, and idea generation

for language learning. The AI's proficiency in

analyzing and summarizing text has made it a

valuable tool for non-native English speakers

grappling with complex academic texts. ChatGPT's

assistance in structure, grammar, and even creative

idea generation has proven a significant boon in

writing. Yet, the impact of ChatGPT on listening and

speaking skills remains an emerging area of study.

While the recent update enabling ChatGPT to speak,

and listen has opened new possibilities for its

application in these areas, concrete studies and

extensive research to substantiate its effectiveness

still need to be completed. Early indications suggest

potential, but more empirical evidence is needed to

confirm its benefits in improving speaking and

listening proficiency.

Despite these advancements, there are significant

concerns about ChatGPT's application to language

The Investigation Related to the Role of ChatGPT in English Language Teaching

431

learning. While ChatGPT has shown proficiency in

aiding with English writing, over-reliance on it can

impede the development of writing skills. It is

recommended to use ChatGPT primarily as a tool for

refining and enhancing language skills rather than as

the primary avenue for content creation. Ethical

concerns also arise, especially regarding ensuring the

originality of content and preventing plagiarism.

Establishing clear guidelines and educating learners

about the importance of originality in their work is

crucial. Incorporating plagiarism detection tools and

emphasizing the ethical aspects of academic integrity

can help mitigate this risk. Given ChatGPT's reliance

on vast datasets, inaccuracies can be a concern.

Learners should be taught to verify AI suggestions

against trusted sources, especially when precision is

critical.

ChatGPT and other similar large language models

also have limitations that could affect language

learning. A fundamental limitation is the lack of

interpretability in ChatGPT's responses (Rimban

2023). In language learning, understanding the

reasoning behind specific linguistic choices is crucial.

Learners need clarity on why specific phrases or

structures are used to grasp important language rules

or cultural nuances. For instance, when ChatGPT

generates a complex sentence structure, learners may

need help understanding the grammatical rules

applied, leading to potential misunderstandings.

Moreover, due to privacy constraints surrounding

ChatGPT's training database, verifying the absence of

biased data in its training materials becomes

challenging. These biases can significantly impact

ChatGPT's language outputs. When language learners

interact with ChatGPT, they may encounter responses

that reflect unbalanced or skewed language use. This

issue becomes especially critical when learners try to

understand and use cultural nuances, colloquial

expressions, or diverse dialects. Without

transparency in the training data, it is difficult to

ensure that learners are exposed to a broad and

unbiased linguistic spectrum.

The future of ChatGPT and similar AI tools in

language learning is promising but requires careful

navigation. As technology evolves, it can be

anticipated more sophisticated AI models capable of

providing more nuanced and culturally aware

language learning experiences. Continued research

and development, ethical guidelines, and educational

best practices will be vital to unlocking the full

potential of AI in language education.

4 CONCLUSION

The study undertaken here assessed ChatGPT's

effectiveness in addressing these challenges across

four critical aspects of language learning: reading,

writing, listening, and speaking. While ChatGPT has

shown notable proficiency in improving reading and

writing skills through its advanced NLP capabilities,

its impact on listening and speaking skills requires

further exploration and empirical validation. Given

the concerns identified, such as over-reliance issues

and ethical concerns, it becomes evident that

ChatGPT should be regarded as a secondary learning

tool in language education. These concerns

underscore the need for a balanced approach, where

traditional learning methods are complemented, not

replaced, by AI assistance.

This study only focuses on the effectiveness of

ChatGPT. However, it is crucial to recognize that

numerous other tools, such as Grammarly and Rosetta

Stone, are available for English language learning,

each offering different approaches and capabilities.

While not covered in this study, these tools also play

a significant role in the landscape of language

learning. Future research plans include expanding the

scope to compare ChatGPT with these and other

educational tools to evaluate their effectiveness in

enhancing various aspects of English language

learning more comprehensively. Such comparative

studies will contribute to a broader understanding of

the optimal use of technology in English language

education.

REFERENCES

E. Dyvik. The most spoken languages worldwide in 2023,

Statista, (2023).

S. Tambuskar. Challenges and Benefits of 7 ways Artificial

Intelligence in Education, Review of Artificial

Intelligence in Education, (2022).

P. McCorduck, M. Minsky, O. Selfridge, H. Simon. History

of artificial intelligence. In IJCAI, pp. 951-954, (1977).

K. S. Kalyan, A. Rajasekharan, S. Sangeetha. Ammus: A

survey of transformer-based pretrained models in

natural language processing, arXiv preprint

arXiv:2108.05542, (2021).

J. Kocoń, I. Cichecki, O. Kaszyca, et al. ChatGPT: Jack of

all trades, master of none. Information Fusion, (2023).

K. Roumeliotis, N. Tselikas. ChatGPT and Open-AI

Models: A Preliminary Review. Future Internet, (2023).

F. Fui-Hoon Nah, R. Zheng, J. Cai, et al. Generative AI and

ChatGPT: Applications, challenges, and AI-human

collaboration. Journal of Information Technology Case

and Application Research, (2023).

ICDSE 2024 - International Conference on Data Science and Engineering

432

W. C. H. Hong. The impact of ChatGPT on foreign

language teaching and learning: opportunities in

education and research, Journal of Educational

Technology and Innovation, (2023).

X. Zhai. ChatGPT user experience: Implications for

education, Available at SSRN 4312418, (2022).

P. P. Ray. ChatGPT: A comprehensive review on

background, applications, key challenges, bias, ethics,

limitations and future scope. Internet of Things and

Cyber-Physical Systems, (2023).

A. Zhavoronkov. Rapamycin in the context of Pascal’s

Wager: generative pre-trained transformer perspective.

Oncoscience, (2022).

Medium ChatGPT-what is it and how does it work exactly?

https://medium.com/geekculture/chatgpt-what-is-it-

and-how-does-it-work-exactly-62e7010524d3 (2022).

M. Imran, N. Almusharraf. 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, (2023).

R. Schmidt-Fajlik. ChatGPT as a Grammar Checker for

Japanese English Language Learners: A Comparison

with Grammarly and ProWritingAid. AsiaCALL

Online Journal, 14(1): 105-119, (2023).

S. Ningrum. ChatGPT’s Impact: The AI Revolution in EFL

Writing. Borneo Engineering & Advanced

Multidisciplinary International Journal, 2(Special Issue

(TECHON 2023)): 32-37, (2023).

M. Li, Y. Zhang, Y. Sun, et al. AI-based ChatGPT Impact

on Medical Writing and Publication. Advanced

Ultrasound in Diagnosis & Therapy (AUDT), 7(2),

(2023).

M. Jeyaraman, S. Ramasubramanian, S. Balaji, et al.

ChatGPT in action: Harnessing artificial intelligence

potential and addressing ethical challenges in medicine,

education, and scientific research. World Journal of

Methodology, 13(4): 170, (2023).

H. Chui. ChatGPT as a Tool for Developing Paraphrasing

Skills Among ESL Learners. Journal of Creative

Practices in Language Learning and Teaching (CPLT),

11(2), (2023).

J. Young, M. Shishido. Evaluation of the Potential Usage

of ChatGPT for Providing Easier Reading Materials for

ESL Students. EdMedia+ Innovate Learning.

Association for the Advancement of Computing in

Education (AACE): 155-162, (2023).

A. Hadwin, J. Kirby, R. Woodhouse. Individual differences

in notetaking, summarization, and learning from

lectures. Alberta Journal of Educational Research,

45(1), (1999).

H. Zhang, X. Liu, J. Zhang. SummIt: Iterative Text

Summarization via ChatGPT. arXiv preprint

arXiv:2305.14835, (2023).

J. Young, M. Shishido. Investigating OpenAI’s ChatGPT

Potentials in Generating Chatbot's Dialogue for English

as a Foreign Language Learning. International Journal

of Advanced Computer Science and Applications,

14(6), (2023).

R. Xing. Advancements in English listening education:

Chat GPT and convolutional neural network

integration. Journal of Pedagogical Research, 7(5):

280-290, (2023).

E. L. Rimban. Challenges and limitations of ChatGPT and

other large language models. International Journal of

Arts and Humanities, (2023), 4(1): 147-152.

The Investigation Related to the Role of ChatGPT in English Language Teaching

433