[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-581098-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-581098-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","co-adaptation-in-learnerchatgpt-dyadic-interaction-a-multi-leveled-linguistic-analysis","Co-Adaptation in Learner–ChatGPT Dyadic Interaction: A Multi-Leveled Linguistic Analysis","","Examines how co-adaptation unfolds in learner–ChatGPT dyadic interaction within an applied linguistics and TESOL context, addressing limitations of prior work that emphasized outcomes or global metrics. Using a seven-week interaction dataset and Complex Dynamic Systems Theory, the study operationalizes alignment across two levels: linguistic form (lexical synchrony and syntactic synchrony) and content (topic initiation, sustainment, and shifting). Process tracing reconstructs micro-level causal chains and adjacent-turn co-occurrence to evidence mutual linguistic modification.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/co-adaptation-in-learnerchatgpt-dyadic-interaction-a-multi-leveled-linguistic-analysis/581098/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/co-adaptation-in-learnerchatgpt-dyadic-interaction-a-multi-leveled-linguistic-analysis/581098.png","ImageObject",300,407,{"name":42,"@type":43},"Rowan","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-11-15","2026-10-15",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",8,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What problem does the study address about LLM–learner research?","Question",{"text":63,"@type":64},"It argues that interactional processes in LLM–learner dyads are under-examined, with prior studies often focusing on outcomes or global metrics rather than turn-by-turn coordination.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"How is co-adaptation defined and analyzed in this work?",{"text":68,"@type":64},"Co-adaptation is treated as dynamic alignment that emerges over time as participants modify linguistic behavior. The analysis uses Complex Dynamic Systems Theory and examines language (form) and content (meaning).",{"name":70,"@type":61,"acceptedAnswer":71},"What two levels does the analytic framework use to detect alignment?",{"text":72,"@type":64},"Language alignment is captured through lexical synchrony and syntactic synchrony, while content alignment is captured through how topics are initiated, sustained, and shifted across turns and weeks.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},581098,1792062470,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,111,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":25,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":55,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":144,"read_time":145,"preview_page_count":4},1099514067415,"https://ap-avatar.wpscdn.com/avatar/100002539d78ffe74a7?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779092875211072502","Studies in Applied Linguistics & TESOL at Teachers College, Columbia University, Vol. 25, No. 2, pp. 36-50 The Forum  \nCo-Adaptation in Learner–ChatGPT Dyadic Interaction: A Multi-Leveled Linguistic Analysis  \nAshley Beccia, Sue Min Park, Jill Williams  \nTeachers College, Columbia University  \nINTRODUCTION  \nDespite the rapid uptake of large language models (LLMs) like ChatGPT in second language (L2) learning environments, the interactional dynamics of LLM–learner dyads remain under-examined. Existing research has primarily focused on the products of LLM–learner interactions, while the interactional process is rarely a central concern. For example, Sok and Shin (2025) compared learners’ task performance before and after interacting with ChatGPT, emphasizing the importance of outcome gains rather than the turn-by-turn exchanges with the LLM. Kusumaningrum et al. (2024) analyzed the degree of conceptual, lexical, and structural overlap between ChatGPT-generated text and learners’ final email drafts, focusing on learners’appropriation of AI output rather than the dynamics of learner–ChatGPT interaction.  \nWhen interactional data have been examined, researchers tend to employ global metrics rather than fine-grained analyses of the ongoing linguistic coordination. For instance, Sim, Kim and Ku (2025) compared ChatGPT–learner and learner–learner interaction using pre/post gainsand aggregate counts of pragmatic-related episodes and exposure to appropriate forms. As a result of this limited research scope, little is known about how learners and LLMs mutually adapt on a granular level over time.  \nUnderstanding these dynamics requires recognizing that LLMs represent a fundamentally different type of interlocutor. While humans co-adapt through embodied perception, affective attunement, and lived experience (Larsen-Freeman, 2020), LLMs generate language probabilistically based on large-scale pretraining and alignment procedures such as instruction-following and reinforcement learning from human feedback (Kulkarni et al., 2023; Ouyang et al., 2022). These systems display stylistic defaults, such as elaborated and expository responses, and typically operate within bounded context windows that limit durable,  \ncross-session memory (Zhong et al., 2024) . This creates an interactional ecology in which alignment may occur without symmetry, and responsiveness reflects pattern-matching rather than experiential attunement. Even so, research shows that humans can align linguistically with computational partners, producing priming effects comparable to human–human interaction (Branigan et al., 2010) . Therefore, co-adaptation does not require shared human cognition or  \n© 2025 Beccia, Park, & Williams. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits the user to copy, distribute, and transmit the work provided that the original authors and source are credited.  \nStudies in Applied Linguistics & TESOL at Teachers College, Columbia University, Vol. 25, No. 2, pp. 36-50 The Forum  \nintention, but the mechanisms and constraints that shape alignment in human–LLM interaction remain largely unexplored.  \nANALYTIC APPROACH  \nTo investigate how co-adaptation unfolds in a human–LLM dyad, we conducted a multi-level analysis of the seven-week interactional dataset. Guided by a Complex Dynamic Systems Theory (CDST) perspective, co-adaptation was conceptualized as the dynamic alignment that emerges as interlocutors modify their linguistic behavior over time. Because alignment can occur on multiple levels, the analysis focused on two dimensions: language (form) and content (meaning) .  \nTo capture how these adjustments unfolded moment-by-moment, we employed process tracing (Hiver & Al-Hoorie, 2019). Rather than correlating variables, process tracing reconstructs causal chains by iteratively locating observable manifestations of theorized mechanisms and examining their sequencing and co-occurrenc","cbCaimSW1EWhI7OH","https://ap.wps.com/l/cbCaimSW1EWhI7OH","pdf",241224,15,"English","# Introduction\n# Analytic Approach\n# Findings\n## Lexical Items","[{\"question\":\"What problem does the study address about LLM–learner research?\",\"answer\":\"It argues that interactional processes in LLM–learner dyads are under-examined, with prior studies often focusing on outcomes or global metrics rather than turn-by-turn coordination.\"},{\"question\":\"How is co-adaptation defined and analyzed in this work?\",\"answer\":\"Co-adaptation is treated as dynamic alignment that emerges over time as participants modify linguistic behavior. The analysis uses Complex Dynamic Systems Theory and examines language (form) and content (meaning).\"},{\"question\":\"What two levels does the analytic framework use to detect alignment?\",\"answer\":\"Language alignment is captured through lexical synchrony and syntactic synchrony, while content alignment is captured through how topics are initiated, sustained, and shifted across turns and weeks.\"}]","Co-Adaptation in Learner–ChatGPT Dyadic Interaction: A Multi-Leveled Linguistic Analysis | PDF",1792038841,38]