The problem the researchers set out to solve is a familiar one in science education. STEM project-based learning places students in authentic, open-ended tasks—designing a filtration system, testing how pollutants move through soil—that demand planning, evidence collection, analysis, and revision. But for primary school children, these open-ended structures can fragment attention, undermine motivation after failed attempts, and leave students unsure how to sequence the inquiry process. Prior research has consistently shown that inquiry learning benefits from scaffolding, especially for younger learners who depend on concrete, visible supports when grappling with abstract procedures. The question was not whether STEM projects should include inquiry, but how to make inquiry developmentally accessible to ten- and eleven-year-olds.

The team’s answer drew on conceptual metaphor theory, which holds that people understand abstract domains by mapping them onto familiar, concrete ones. Metaphors have long been used in science classrooms, but verbal or static metaphorical explanations often fail young students when the mapping between the familiar source and the abstract target is unclear. Digital games, the researchers reasoned, can instantiate those mappings interactively—through goals, rules, staged challenges, feedback, and progression—so that students experience the logic of inquiry rather than merely reading about it. Accordingly, they built Saving the Castle, a role-playing adventure developed in RPG Maker MV and deployed on Android tablets, in which the six-stage quest structure was deliberately aligned with six components of scientific inquiry: identifying questions, formulating hypotheses, planning, experimenting and data collection, analyzing and concluding, and communicating.

Crucially, the game was not a stand-alone activity. In the experimental condition, dubbed DMG-STEM PBL, the game served as a front-loaded scaffold: students played it before beginning hands-on project work, constructing an initial framework for inquiry through the adventure narrative. As project work progressed, the teacher deployed brief metaphor-recall prompts—asking students, for example, what the castle keys might represent in their current investigation, or how feedback in the game might inform revisions to their experimental strategy. Toward the end of each project, metaphor-supported reflective narration invited students to retell their work as an inquiry journey, casting their initial question as a mission goal, their decisions as route choices, their evidence as clues, and their revisions as strategy upgrades. This three-layer design was intended to let the inquiry framework be constructed, reactivated, and reflected upon across the entire project cycle.

The comparison group followed the same curriculum—two environmental STEM projects, Finding a Home for Waste and Cleaning Wastewater, delivered in six 35-minute sessions by the same experienced teacher in the same dedicated STEM classroom—but received conventional teacher-led explanation, procedural reminders, and reflective discussion instead of the metaphor-based supports. Implementation fidelity was checked by two independent observers using a 12-item checklist, and both conditions scored near the maximum, indicating that the metaphor scaffold was the main planned difference between the groups. Participants were 48 students in the experimental group and 42 in the control group, with class-level random assignment and no attrition over the study period.

The outcomes revealed a differentiated rather than uniformly favorable pattern. Students in the game-supported condition reported significantly higher overall perceived scientific inquiry competence, with the clearest gains concentrated in planning, experimenting and data collection, and analyzing and concluding—precisely the dimensions most closely aligned with the game’s procedural structure. No significant differences emerged for identifying questions, formulating hypotheses, or communicating, suggesting a boundary condition: practices that require epistemic creativity, theoretical reasoning, or social negotiation may need scaffolds beyond a narrative-driven metaphor. The authors caution that these findings rest on self-reports rather than performance-based assessments, so the results reflect stronger perceived competence rather than directly measured skill.

The learning experience told a similarly encouraging story. The experimental group reported significantly higher overall learning motivation, an effect driven by intrinsic rather than extrinsic motivation—a pattern consistent with the intervention’s design, which deliberately avoided points, badges, leaderboards, and ranking mechanisms in favor of meaning, competence, and engagement. Flow experience, the state of deep absorption associated with clear goals, immediate feedback, and matched challenge, was also significantly higher in the game-supported group, with a medium effect size. Supplementary exploratory analyses found no evidence that any of these differences varied by gender, which the researchers attribute to the game’s non-competitive, collaborative embedding within shared STEM projects rather than reliance on speed or reward accumulation.

One finding demands nuance: students in the game-supported condition reported higher overall cognitive load, driven by greater mental effort, though mental load—the perceived difficulty of the task itself—did not differ between groups. The researchers offer two compatible interpretations. The extra effort may reflect generative processing, as students actively connected the game’s staged progression and feedback to their ongoing inquiry tasks, reconsidering plans and strategies in ways consistent with their higher scores on planning and analysis. Alternatively, it may reflect the genuine coordination demands of juggling a game narrative, metaphorical meanings, teacher prompts, hands-on experimentation, and reflective narration. The authors frame this as a potential benefit-cost trade-off rather than unequivocal evidence of effectiveness, noting that instructional design should promote learning-relevant processing while minimizing extraneous demands.

The study’s limitations temper its promise. With only two intact classes, one per condition, class-level factors such as peer culture and group dynamics cannot be fully ruled out, and the three-week duration leaves open whether motivational gains persist after the novelty of gameplay fades—a well-documented concern in gamification research. Self-reported competence may also diverge from actual inquiry performance, and the design did not directly trace how students interpreted the metaphors or allocated cognitive resources during tasks. Still, the central lesson stands: the value of a digital metaphor game in STEM education lies less in the presence of game features themselves and more in the principled alignment among game design, inquiry practices, motivational mechanisms, and cognitive demands. For educators weighing whether to bring games into project-based science, the findings suggest the game should function not as a reward or a distraction but as a structural scaffold—one whose adventure, keys, and quests mirror the very process of doing science.

Subject of Research: The effectiveness of a digital metaphor game-mediated STEM project-based learning approach for primary students' scientific inquiry competence, learning experience, and cognitive load.

Article Title: Exploring the effectiveness of a digital metaphor game-mediated STEM PBL approach for primary students’ perceived scientific inquiry competence, learning experience, and cognitive load

Article References: Exploring the effectiveness of a digital metaphor game-mediated STEM PBL approach for primary students’ perceived scientific inquiry competence, learning experience, and cognitive load. (n.d.). https://doi.org/10.1186/s40594-026-00647-6

Image Credits: AI Generated

DOI: 10.1186/s40594-026-00647-6

Keywords: digital metaphor game, STEM project-based learning, scientific inquiry competence, primary education, cognitive load, learning motivation, flow experience, game-based learning, instructional scaffolding, conceptual metaphor theory, science education, quasi-experimental study

This innovative research, published in PLOS Complex Systems, advances the field by employing tools from complex systems science to map the intricate networks of metaphorical relationships. Conceptual metaphor theory traditionally posits that metaphors are not merely ornamental linguistic figures but foundational cognitive structures. The authors’ empirical data substantiate this claim by revealing stable, enduring metaphor networks that persist over time rather than fleeting rhetorical embellishments. Such an approach moves beyond subjective analyses towards objective quantification of metaphorical dynamics, marking a significant methodological breakthrough in cognitive linguistics.

At the core of their findings is the identification of a nuanced metaphor network characterized by distinct clusters that segregate abstract and concrete conceptual categories. This bifurcation allows for detailed examination of the processes through which metaphors operate, specifically highlighting two pivotal metaphorical mechanisms. The first involves mappings that extend from concrete, sensory-grounded domains to abstract, conceptual territories, enabling individuals to conceptualize intangible ideas through familiar physical experiences. The second mechanism uncovers the spontaneous emergence of novel metaphorical mappings that occur between concrete domains themselves, indicating that metaphorical creativity is not confined merely to bridging abstraction but also thrives within the sensory domain.

These metaphorical processes are not random; rather, the study’s network analysis reveals a pronounced concentration of metaphorical activity centered on a limited set of everyday topics. Within the concrete conceptual group, one category notably functions dually as a prolific source of metaphorical projections and as a frequent target, serving as a hub in the metaphor network. In contrast, within the abstract domain cluster, another category predominantly operates as a metaphorical target, suggesting an asymmetry in metaphorical flow and influence. This structural insight into metaphor networks provides a fresh lens through which the cognitive economy of metaphor usage can be understood, emphasizing metaphor as an inherently creative enterprise driven by the interplay of contrast and cognitive tension.

Delving deeper, the authors demonstrate how metaphors facilitate re-conceptualization by enabling the emergence of novel similarities that were previously unrecognized or unarticulated. Through mapping dissimilar domains laden with underlying tensions, metaphor activates cognitive processes that reorganize conceptual space, allowing new patterns of thought and understanding to surface. This dynamic underscores metaphor’s role not simply as a linguistic device but as a fundamental mechanism of cognitive innovation and semantic evolution.

Methodologically, the researchers employ diachronic data analysis—a temporal exploration of linguistic data spanning extensive periods—to trace the evolution and refinement of conceptual metaphors over time. This approach enables them to observe the stability and transformation of metaphorical networks, ensuring that conclusions are grounded in robust empirical evidence adorned with temporal depth. By applying statistical and data analysis techniques rooted in complex systems theory, the study achieves a synthesis of qualitative linguistic insights with quantitative rigor, charting new territory for metaphor research.

The implications of this study resonate across multiple disciplines. For cognitive linguistics and the philosophy of language, these findings invite reevaluation and refinement of conceptual metaphor theory, emphasizing the necessity of empirical methodologies to complement and ground theoretical constructs. The demonstrated longevity and structural complexity of metaphorical networks challenge reductionist views that treat metaphors as ephemeral or isolated language phenomena, instead supporting a model that sees them as dynamic cognitive architectures.

Beyond the humanities, the research opens promising avenues in artificial intelligence and machine learning, particularly in the domains of analogy-making and representation learning. Metaphors, as structured mappings between conceptual domains, share affinities with how AI systems model knowledge transfer and abstraction. The mathematical formalism and empirical insights offered by the study provide valuable tools for enhancing AI’s capability to simulate human-like figurative reasoning and conceptual flexibility, a key step toward more sophisticated natural language understanding systems.

Furthermore, the study’s use of complex systems methodologies advances the mathematics of cognition, contributing to formal epistemology by offering analytic frameworks that clarify how abstract meaning arises from structural preservation across conceptual domains. This cross-pollination of linguistic theory, cognitive science, and mathematical modeling exemplifies an integrative research paradigm essential for grappling with intricate phenomena such as metaphor.

Importantly, the study reveals the centrality of tension and contrast in metaphor formation, framing metaphor not as a mere decorative feature of language but as a process fueled by cognitive disequilibrium. Through harnessing disparities between concrete and abstract domains, metaphorical thought promotes the generation of meaning by reconciling and reframing these contrasts, highlighting human cognition’s remarkable capacity for adaptive creativity.

The data-driven approach adopted by Teich, Leal, and Jost further demonstrates the efficacy of blending qualitative theoretical frameworks with quantitative empirical analytics. The scale and granularity of their diachronic corpus enable unprecedented insight into metaphor formation, evolution, and resilience. Such research paves the way for future investigations to expand the scope and detail of metaphor networks, potentially integrating multimodal and cross-cultural datasets to deepen understanding of metaphor as a universal cognitive and linguistic phenomenon.

Ultimately, this study marks a seminal step in bridging conceptual metaphor theory with formal and empirical sciences. By elucidating the structural properties and dynamic intricacies of metaphor networks through sophisticated data analysis, it equips researchers with novel methodologies and conceptual tools to explore the cognitive underpinnings of figurative language. The implications extend far beyond academia, touching fields such as AI, cognitive computing, and epistemology, and herald a new era where the once elusive realm of metaphor becomes increasingly accessible to scientific inquiry.

This research underscores the profound creativity embedded in human language, where metaphors serve as cognitive engines fueling semantic innovation and conceptual expansion. As science continues to unravel the fabric of thought, such interdisciplinary collaborations affirm that understanding metaphor is not only a matter of linguistic interest but a gateway to comprehending the very nature of human intelligence and creativity.

Subject of Research: Conceptual Metaphor Theory and the cognitive structures underlying metaphorical language.

Article Title: Diachronic data analysis supports and refines conceptual metaphor theory

News Publication Date: 5-Aug-2025

Web References: 10.1371/journal.pcsy.0000058

Keywords: Conceptual metaphor theory, metaphor networks, cognitive linguistics, complex systems, diachronic analysis, abstract and concrete domains, figurative language, machine learning, artificial intelligence, representation learning, cognitive innovation, semantic structure