https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
386
SECOND LANGUAGE ACQUISITION THEORIES AND THEIR
RELATIONSHIP TO AI TOOLS.
Polina Golovachyova
Uzbekistan State World Languages University, Bachelor Student, English
Philology
Email: polinagolovachyova03@gmail.com
Phone: +998 940572638
Abstract:
This article studies the interconnection between second language acquisition
theories (SLA) and Artificial Intelligence(AI) tools in modern language education.
By analyzing key theories SLA -
including Krashen’s input hypothesis, Long’s
interaction hypothesis, Swain’s output hypothesis, Schmidt’s noticing hypothesis,
and Vygotsky’s sociocultural theory, it discusses how AI technologies can be
designed to improve the process of language learning. The main focus is on
adaptive educational environments, AI-mediated interactions, feedback
mechanisms, and socio-cognitive support tools. As AI becomes more integrated
into educational process, there is a growing need for theoretical aspects of
education for ensuring ethical, effective, and pedagogically sound tool
development.
Keywords
: second language acquisition, artificial intelligence, input
hypothesis, interaction, adaptive learning, language learning technologies
Introduction
Second Language Acquisition - it is a field, which investigates how people
acquire languages, that differ from their native ones. For the last decades, the
https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
387
number of theoretical models, which explain cognitive, social and emotional
factors influencing language learning. At the same time, the recent achievements
in the sphere of AI stimulate appearing of sophisticated educational tools,
providing more opportunities for teaching languages and personalized learning.
Integration AI tools to SLA theories provides a foundation for designing systems,
which correspond to natural process how humans acquire languages.
The use of AI in second language teaching raises important questions: How
can these tools support What kind of feedback should they offer? How can they
adapt to individual differences? The article answer these questions, by connecting
prominent SLA theories with practical AI applications.
Main Part
Krashen’s
input hypothesis and AI personalization
Stephen Krashen’s Input Hypothesis (1985) с
laims that learners acquire
language when they get a chance to comprehensible input
—
language that is
slightly beyond their current level (i+1). This theory also emphasizes the
importance of a low-anxiety environment, as stress can block input from being
processed effectively. AI tools such as Duolingo, Rosetta Stone, and LingQ
implement this theory by adjusting level of difficulty due to learne
r’s
performance.
Natural Language Processing (NLP) algorithms assess lexical and grammatical
complexity to provide learners with optimal content. For instance, adaptive reading
platforms like Newsela modify articles to suit learners
’
reading levels, aligning
with the "i+1" model.
Moreover, chatbots and AI tutors can lower the affective filter by offering
private, non-judgmental environments. Virtual agents like ChatGPT or Google's
Bard provide opportunities for safe practice, reducing performance anxiety and
enabling learners to engage more freely in spontaneous language use.
https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
388
Long’s
interaction hypothesis and AI conversational agents
Michael Long (1996) argued that interaction, particularly negotiation of
meaning, is key to acquisition. When communication breaks down, strategies such
as clarification requests, recasts, and confirmation checks facilitate understanding
and promote linguistic development.AI-powered conversational tools can simulate
these interactions. For example, AI speaking partners like ELSA Speak or Replika
engage users in dialogues where errors prompt contextual feedback. When learners
make mistakes, these systems might offer recasts rather than explicit corrections,
keeping the conversation natural while providing learning opportunities.
These AI systems replicate authentic interaction patterns found in
communicative classrooms, enabling consistent practice regardless of time or
teacher availability. Some AI agents are even capable of multimodal feedback
—
combining text, speech, and visual cues
—
to enhance comprehension, aligning with
multimodal learning theories.
Swain’s
output hypothesis and AI writing and speaking tasks
Merrill Swain’s Output Hypothesis (1985)
states that language production
skills
—
speaking and writing is important for acquisition. Producing language
encourages learners to notice gaps in their knowledge, test hypotheses, and start to
understand the language deeper. AI platforms, such as Grammarly or Write &
Improve by Cambridge University assist learners by finding errors and offering
explanations, it makes people to think how language works.Voice recognition tools
such as Google's Speech-to-Text or speech feedback in apps like Mondly help
learners improve pronunciation and grammar.AI systems also support hypothesis
testing through open-ended writing or speaking tasks where learners try new
constructions and get feedback. This matches with formative assessment principles
in education.
https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
389
Schmidt’s
noticing hypothesis and AI-enhanced attention
Richard Schmidt (1990) suggested that conscious attention to language forms
is essential for acquisition. It is important to notice linguistic features in coming
information to analyse language effectively.AI technologies use visual highlights
(e.g., bolding verb forms or highlighting prepositions) to draw attention to specific
structures. Systems like Netex Learning or Edmodo integrate such features into
learning materials. Additionally, intelligent systems can interrupt a session with a
mini-lesson if this system finds a consistent pattern of learner error
—
a just-in-time
feedback approach.Visual analytics and heatmaps are include in AI platforms that
track learner attention and help educators understand where learners focus most,
offering data-driven insights into what learners notice or ignore.
Vygotsky’s
sociocultural theory and AI mediation
Vygotsky’s
sociocultural theory (1978) emphasizes that learning occurs
through social interaction and mediation by more knowledgeable others. Learning
is most effective within the Zone of Proximal Development (ZPD) tasks learners
can perform with guidance.AI systems act as mediators by providing scaffolding
through hints, prompts, and examples. For instance, Write & Improve offers
example responses for writing prompts, enabling learners to model their answers.
In virtual collaborative platforms like Classcraft or Microsoft Teams for
Education, AI tracks group dynamics, supports cooperative learning, and provides
personalized support based on group performance.While AI cannot fully replicate
human mentorship, its ability to adjust in real time allows it to perform a similar
scaffolding role. Some researchers (Lantolf & Thorne, 2007) argue that such tools
represent a new class of cultural artifacts mediating second language development.
Ethical considerations and limitations
https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
390
Despite the promise of AI, several challenges remain. Ethical concerns
include data privacy, potential biases in AI feedback, and overreliance on
technology. Moreover, while AI can simulate interaction, it lacks true empathy and
cultural understanding that human teachers provide.It is also crucial to maintain
teacher presence and human interaction in AI-enhanced classrooms. AI should
support, not replace, human educators. Theories of SLA help ensure that AI design
remains learner-centered, pedagogically sound, and responsive to real
developmental needs.
The role of AI in learner autonomy and self-regulation
One significant contribution of AI tools to SLA is their support for learner
autonomy and self-regulated learning
—
concepts rooted in sociocognitive theories
of language development. According to Benson (2011), learner autonomy involves
the capacity to take control of one's learning process, including goal setting,
strategy use, and self-assessment. AI-enhanced language platforms offer
personalized learning paths, immediate feedback, and progress tracking, enabling
learners to make informed decisions about their study patterns.For example,
applications like Lingvist and Busuu analyze learners’ weak points and adapt
vocabulary review accordingly, encouraging metacognitive reflection. AI systems
also allow for just-in-time learning, where learners can access explanations or
translations as needed, promoting independence and fostering strategic learning
behaviors. This aligns with Zimmerman’s (2002) model of self
-regulated learning,
where monitoring and reflection are key stages facilitated by adaptive technology.
AI and data-driven language pedagogy
Another emerging field at the intersection of SLA and AI is data-driven
learning (DDL), where learners explore corpora or large text data sets to discover
patterns in language use. Tools such as Sketch Engine or AntConc are increasingly
incorporating AI-based enhancements, including semantic analysis and automatic
https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
391
error tagging. These tools encourage learners to analyze real-life usage examples,
aligning with inductive approaches to grammar and vocabulary learning.According
to Boulton and Cobb (2017), DDL fosters learner noticing and hypothesis
formation, key aspect
s of Schmidt’s Noticing Hypothesis and Swain’s Output
Hypothesis. When AI facilitates pattern recognition or highlights collocations and
grammatical structures, it effectively bridges SLA theory and practical application.
Moreover, teachers can use AI-generated learner data to inform lesson design,
creating a formative feedback loop supported by SLA research.
Conclusion
Second language acquisition helps evaluate and guide the designing of AI-
tools in language teaching.The input hypothesis is crucial in adapting material to
the level of the learner. The interaction and output hypotheses build the foundation
for communicative and productive tasks. The noticing hypothesis explains how
important it is to pay attention to language forms. The sociocultural theory
emphasizes the importance of real communication.
AI is able to improve the foreign language learning process, creating a
comfortable, calm and productive environment. However, it is significant to ensure
that these technologies are ethically acceptable and reliable.
In the future, researchers have to keep searching for approaches how AI can
assist for better understanding and developing the process of second language
acquisition.
References
1.
Boulton, A., & Cobb, T. (2017). “Corpus use in language learning: A
meta-
analysis.”
Language Learning, 67(2), 348
–
393.
2.
Benson, P. (2011). Teaching and Researching Autonomy in Language
Learning. Routledge.
https://scientific-jl.com/luch/
Часть
-46
_ Том
-5_
июнь
-2025
392
3.
Chapelle, C. A. (2020). English Language Learning and Technology:
Lectures on Applied Linguistics in the Age of Information and Communication
Technology. John Benjamins.
4.
Godwin-
Jones, R. (2018). “Using Mobile Technology to Develop
Language Skills and Cultural Understanding.”
Language Learning &
Technology, 22(3), 4
–
17.
5.
Krashen, S. D. (1985). The Input Hypothesis: Issues and Implications.
Longman.
6.
Kukulska-
Hulme, A. (2020). “Mobile
-assisted language learning
[Revisited].”
The Encyclopedia of Applied Linguistics.
7.
Long, M. H. (1996). "The Role of the Linguistic Environment in
Second Language Acquisition." In W. Ritchie & T. Bhatia (Eds.), Handbook
of Second Language Acquisition.
8.
Swain, M. (1985). "Communicative competence: Some roles of
comprehensible input and comprehensible output in its development." In S.
Gass & C. Madden (Eds.), Input in Second Language Acquisition.
9.
Schmidt, R. (1990). "The role of consciousness in second language
learning." Applied Linguistics, 11(2), 129
–
158.
10.
Vygotsky, L. S. (1978). Mind in Society: The Development of Higher
Psychological Processes. Harvard University Press.
11.
Lantolf, J. P., & Thorne, S. L. (2007). Sociocultural Theory and the
Genesis of Second Language Development. Oxford University Press.
12.
Zimmerman, B. J. (2002). "Becoming a self-regulated learner: An
overview." Theory Into Practice, 41(2), 64
–
70.
