Contemporary societies are increasingly characterized by hybrid phenomena in which boundaries traditionally taken for granted—such as those between the real and the virtual, the human and the non-human, and nature and technology—are becoming blurr...
Contemporary societies are increasingly characterized by hybrid phenomena in which boundaries traditionally taken for granted—such as those between the real and the virtual, the human and the non-human, and nature and technology—are becoming blurred and reconfigured. Complex global challenges such as pandemics, climate crises, and digital transformation unfold as complex systems that cannot be explained through linear causality, positioning humans as part of heterogeneous networks entangled with non-human actors including technologies, tools, information, and policies. In this context, education is no longer understood as a stable site of knowledge transmission but is increasingly reconfigured as a dynamic and interactive field of practice that responds sensitively to socio-technical change. In mathematics education in particular, ongoing reforms—including the strengthening of digital and AI components, the introduction of Artificial Intelligence Mathematics courses, the emphasis on mathematical modelling, and connections to climate change and sustainability—reflect attempts to reposition mathematics as a core discipline for structuring and interpreting problems in digital societies.
Against this backdrop, this study focuses on the ways in which non-human entities, including digital technologies, implement unexpected forms of agency in mathematics education. The study aims to explore how dynamic networks of human and non-human actors are configured in digital tool–based mathematics teaching and learning. Rather than being completed by a single subject, the design and implementation of instruction are conceptualized as sociomaterial practices constituted through ongoing processes of translation, negotiation, connection, and disconnection among heterogeneous actors. Drawing on posthumanism and Actor–Network Theory (ANT) as ontological, epistemological, and methodological frameworks, this study analyzes how networks are formed and transformed in the design and enactment of digital tool–based mathematics instruction. The guiding research questions are as follows: (1) How are networks configured in the design process of digital tool–based mathematics instruction? (2) How is the designed program translated and prepared for implementation within a school context? (3) How are networks configured during the enactment of digital tool–based mathematics instruction?
This study was conducted based on posthuman qualitative research methodologies, ANT methodologies, and object interview heuristics proposed by Adams and Thompson (2025). During the design phase, two mathematics teachers collaboratively designed a 20-session program encompassing digital tool–based data analysis, digital content creation, and civic engagement activities through ten online meetings. Using interpretative phenomenological analysis (IPA), the study explored how diverse human and non-human actors—including policy documents, budgets, official documents, school leadership support, undergraduate and graduate disciplinary backgrounds, scientific experimental knowledge, digital devices, mathematical concepts, climate change, and big data—were negotiated and aligned to produce the final program. Subsequently, processes of teacher recruitment, attention-building strategies, and role enrollment for classroom enactment were analyzed according to Callon’s (1986) four moments of translation: problematization, interessement, enrollment, and mobilization. The designed program was enacted by a teacher with 20 students in a general academic high school in Incheon Metropolitan City, with the researcher and an assistant teacher participating as observers. Using core ANT concepts and the H6 and H8 heuristics, networks of non-human actors during classroom enactment were identified, and cases of formation, transformation, and dissolution of group-level actor-networks were examined.
The first findings concerns the actor-network configured during the instructional design phase. The two teachers translated the same policy documents differently based on their personal backgrounds and coordinated their respective actors in order to pass through obligatory passage points (OPPs). Throughout this process, various non-human actors—such as scientific experiments (MBL, Arduino), generative AI, big data, loss functions, and mean squared error (MSE)—emerged, disappeared, and were reconfigured. For example, the previously concealed tetrad of generative AI became visible, “scientific experiments” were detoured into “digital content creation,” and strong connections were established between the program and digital literacy. These findings suggest that instructional design is not a linear decision-making process but a site of translation requiring continuous negotiation among heterogeneous human and non-human elements.
The second findings analyzes the preparation and translation processes for classroom enactment using Callon’s four moments of translation. Teachers translated the designed program to fit the school context and students, revising instructional materials and constructing learning environments. First, problematization began as teachers’ beliefs, digital tools, data, climate change, sustainability, and statistical software entered the instructional network, and obligatory passage points for enactment were established. Alliances were strengthened by articulating the benefits actors could gain by passing through these passage points. Second, during interessement, attention-building strategies were enacted by researchers, teachers, and non-human actors. Researchers attempted to detach teachers from existing instructional routines and align them with the new program, while non-human actors such as digital tools, climate time-machine websites, and AI-generated songs attracted students into the network. Third, during enrollment, the roles of researchers, teachers, students, and digital technologies were defined and stabilized, although actors occasionally attempted betrayal when their interests were not sufficiently satisfied. To minimize such risks, the study highlights the importance of avoiding excessive OPPs and flexibly revising them when actors encounter difficulties. Fourth, during mobilization, new actors acquired representativeness and authority, and actor-networks were stabilized through inscriptions and prescriptions functioning as immutable mobiles. The macro-scale phenomenon of climate change was translated into a micro-scale classroom issue and subsequently retranslated to connect with the local community through the program, which itself functioned as an immutable mobile.
The third findings addresses the sociomaterial dynamics of non-human actors during classroom enactment. First, connections among data, graphs, and affect produced different assemblages across student groups. For instance, disparities in carbon dioxide emissions between developed and developing countries generated climate-related emotions such as injustice and resentment, suggesting that affect itself functions as an actor in learning. Second, different digital tools articulated divergent interpretations of data analysis: Desmos and Algeomath were strongly connected to optimal trend lines based on mean squared error, Python emphasized correlation, and textbook-based mathematical concepts emerged, detoured, or were replaced. Third, generative AI exhibited agency through prompt engineering during digital content creation, as structured prompts developed by one group circulated across groups, amplifying the power of digital content. Fourth, digital content functioned as an immutable mobile, spreading to other schools and local communities. Although the program was enacted in a single school, it demonstrated significant potential for expansion and reproduction through teacher professional development and informal teacher networks. At the same time, cases of network disruption and recovery were observed, including restrictions on generative AI use, tensions between written examinations and the program, and shifts in group roles following the emergence of student AI literacy. Some actors weakened or dismantled networks, while others aligned with project goals and strengthened connections.
Through these analyses, this study demonstrates that digital and AI technologies function not merely as auxiliary tools in mathematics teaching and learning but as sociomaterial actors endowed with agency. Classroom practice is empirically shown to be a non-linear, dynamic network constituted through ongoing negotiation and translation among human and non-human elements. Mathematics teachers’ practices are re-conceptualized not as isolated competencies or black boxes but as provisional actor-networks entangled with institutions, policies, technologies, materials, and student emotions. By foregrounding the sociomaterial constitution of digital tool–based mathematics teaching and learning from a posthuman perspective, this study contributes to rethinking theory and practice in mathematics education. Posthumanism and ANT function not only as analytical tools but also as philosophical resources for reimagining educational phenomena from ontological and methodological perspectives. This exploratory study positions non-human agency at the center of analysis and proposes a relational, multi-layered understanding of digital tool–based mathematics teaching and learning as complex networks of human and non-human actors, offering a point of departure for future research on the ontological status of non-human actors in mathematics education.