Executive Summary

McKinsey projects that by 2030, roughly two-thirds of today’s HR activities could be fully automated or automated in delivery. This piece argues the more consequential question is not how much HR work agents absorb, but who benefits from that shift and who bears its costs. The McKinsey report frames agentic transformation as a governance design problem; this response reframes it as a distributive one.

• Redefined value: As transactional work shifts to agents, procedural expertise is devalued even as HR claims to elevate itself toward “strategic” work.

• Apprenticeship gap: Entry-level tasks train future judgment. Losing them requires deliberate, funded reinvestment, not passive hope.

• Cognitive load: Time freed by automation is not neutral; someone decides how it gets spent, and by whom.

• Global translation: A North Star built around one enterprise’s platform and norms may not transfer across labor markets with different protections and expectations.

McKinsey analysis indicates that, by 2030, roughly two-thirds of today’s HR activities could be fully automated or fully automated in delivery.

Yet greater automation may not reduce the need for human HR work. As AI agents assume transactional and administrative responsibilities, HR professionals may face growing demands for judgment, agent governance, exception management, employee development, and the reconciliation of management priorities with employee interests.

This tension raises a fundamental question for the “North Star” of agentic HR: Automation for whom, and toward what ends?

Should HR automation primarily reduce costs and increase managerial control, expand HR’s strategic capacity, improve employees’ work and career experiences, or contribute to broader social value? More importantly, how should its benefits, risks, and decision-making authority be distributed among management, HR professionals, employees, and society?

From Governance Design to Distributive Question

Komm, Hancock, and Durth (2026), in their recent McKinsey report on building HR for the agentic era, frame this shift primarily as a design problem. Organizations that succeed, they argue, do not simply scale isolated pilots; they define a human-agent operating model first and work backward to implementation. The authors identify three tensions that any credible North Star must confront: who governs the orchestration platform on which agents run, where personalized HR support tips over into surveillance, and which human capabilities, judgment, relationships, coaching, persuasion, deserve deliberate investment as routine work is absorbed by agents.

These are useful distinctions, but they are framed largely as internal governance questions for HR functions and their C-suite sponsors. The deeper question, and the one this piece wants to press further, is distributive rather than architectural. Automation reallocates work, but it also reallocates power. When two-thirds of HR’s transactional load shifts to agents, someone decides what happens to the freed capacity, and someone decides what happens to the workers whose entry-level tasks disappear.

What Remains for Humans, and Who Decides

Consider the report’s own account of what remains for humans: strategic engagement, agentic capability management, exception handling, and, notably, cognitive load management, the mental effort employees now spend integrating AI into their daily work. That last item deserves more attention than a single line affords it. If a redesigned HR function shifts fifty percent of human time toward “strategic engagement and innovation,” as the report projects, this is not simply a productivity gain. It is a redefinition of what counts as HR work, and by extension, who is qualified to do it. Employees whose value was rooted in procedural expertise, running payroll accurately, managing onboarding smoothly, may find that expertise devalued precisely as the function claims to be elevating itself.

There is also the apprenticeship problem the report raises but does not fully resolve. If agents absorb entry-level tasks, junior professionals lose the on-the-job training ground where judgment has traditionally been built. Organizations will need to construct that development deliberately, the report says, but deliberate construction requires investment, and investment requires someone to decide it is worth making rather than deferring to short-term efficiency gains.

A North Star Built Elsewhere

For diasporic and international HR professionals in particular, these questions carry additional weight. Governance structures built around a Western enterprise’s orchestration platform, its skills graph, its definitions of what personalization “serves” an employee, may not translate cleanly across labor markets with different regulatory protections, different norms around workplace surveillance, and different assumptions about the employer-employee relationship. A North Star defined in Hamburg, Washington, or Cologne is not automatically a North Star everywhere the function operates.

Rebuilding HR, and For Whom

None of this argues against the transformation the McKinsey authors describe. The pilot trap they name, speed at the task level, drift at the level of design, is real, and the organizations that avoid it will likely outperform those that do not. But a North Star worth following should answer not only how HR should be rebuilt, but for whose benefit, and who gets a voice in deciding.

Reference

Komm, A., Hancock, B., & Durth, S. (2026, August). Escaping the pilot trap: Building HR for the agentic era. McKinsey & Company.

https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/escaping-the-pilot-trap-building-hr-for-the-agentic-era

Prof. Dr. Jeonghwan (Jerry) Choi (Managing Editor), University of Maine at Presque Isle

Jeonghwan (Jerry) Choi, PhD is an Associate Professor of Business at the University of Maine at Presque Isle and Editor-in-Coordination of K-GSP Forum (contact: jeonghwan.choi at gmail.com). With over 25 years of industry and consulting experience, he specializes in leadership development, human resource management, organizational behavior, and social entrepreneurship. His research focuses on workforce resilience, organizational health, and self-directed leadership — bridging rigorous scholarship with practical insight to cultivate leaders who create meaningful, sustainable, and humane organizations.

한글 요약

맥킨지 분석에 따르면 2030년까지 오늘날 HR 업무의 약 3분의 2가 완전 자동화되거나 전달 단계에서 자동화될 수 있다. 하지만 자동화가 확대된다고 해서 HR 인력의 필요성이 줄어드는 것은 아니다. 에이전트가 거래적·행정적 업무를 맡게 되면서 HR 전문가에게는 판단력, 에이전트 거버넌스, 예외 상황 관리, 직원 개발, 그리고 경영진과 직원 이해관계의 조율이라는 더 높은 수준의 역할이 요구된다.

맥킨지 보고서는 이를 주로 거버넌스 설계 문제로 다루지만, 이 글은 더 근본적인 질문을 제기한다. 자동화는 업무만이 아니라 권력도 재분배한다. 절차적 숙련도에 기반했던 직원들의 가치는 오히려 폄하될 수 있고, 신입 직원이 판단력을 기르던 도제식 훈련 기반도 사라질 위험이 있다.

특히 서구 기업의 플랫폼과 기준을 중심으로 설계된 북극성 비전은 다른 노동시장에 그대로 적용되기 어렵다. 따라서 진정한 북극성 비전은 HR을 어떻게 재구축할 것인가뿐 아니라, 누구를 위한 것이며 누가 그 결정에 참여하는가라는 질문에도 답해야 한다.

EXECUTIVE SUMMARY

We live in a connected world where decisions and disruptions can move rapidly across individual, team, firm, industry, nation, and global-system levels. This article introduces the Six-Level Connected World Framework, integrating these six levels with interdisciplinary inquiry, cumulative learning, and AI-enabled exploration to explain how local actions can develop into wider patterns of resilience, innovation, value creation, and systemic change. By combining supply chain thinking with insights from literature, philosophy, history, and networked research, the framework offers a broader and more human-centered way to understand complex global challenges while retaining human judgment and intellectual ownership in the AI era.

Keywords: Connected world; Six-Level Connected World Framework; Global supply chains; Interdisciplinary inquiry; AI-enabled research

1. Introduction

When I was in India as a Fulbright-Nehru Scholar in 2017, I saw more clearly how global geopolitical dynamics shape the everyday choices of individuals, firms, and nations as they seek opportunity, security, and livelihood. Questions of trade, infrastructure, technology, employment, policy, and international relationships were not abstract forces operating somewhere in the distance; they were visible in how people worked, how companies competed, and how countries positioned themselves in a changing global environment. That experience reinforced a simple but powerful idea: in a connected world, actions at one level rarely remain at that level. A decision made by one person can move through a team, reshape a firm, disrupt an industry, influence a nation, and eventually produce global consequences.

My research had begun much earlier with new product development in the automotive industry, especially with the question of how firms organize people, processes, technologies, and relationships to bring new products successfully to market. As this work developed, my attention increasingly shifted toward supplier relations because product development could not be fully understood within the boundaries of a single firm. Trust, coordination, knowledge sharing, quality, cost, speed, and joint problem solving showed that suppliers were not peripheral participants but essential partners in innovation and competitive performance.

This focus on supplier relationships gradually expanded into global supply chain management. Automotive firms and their suppliers were increasingly connected across national borders, technologies, industries, and institutional environments, requiring a wider understanding of global sourcing, manufacturing networks, logistics, innovation partnerships, and strategic alliances. Over time, it became increasingly clear that global supply chains must be understood across multiple levels: the individual, team, firm, industry, nation, and global system, because decisions and disruptions move continuously across these levels. This integrated approach is referred to here as the Six-Level Connected World Framework.

This broader perspective also opened the door to literature, history, philosophy, and contemporary developments such as artificial intelligence, demographic change, trade restructuring, geopolitical competition, and industrial transformation. Multi-level, systems, and network perspectives are already well established in supply chain scholarship, which has increasingly emphasized interdependence, complex adaptive systems, network relationships, and interactions across organizational and institutional boundaries (Pathak et al., 2007; Wieland, 2021). The contribution of this article therefore does not lie in proposing multi-level analysis itself as new, but in integrating six connected levels—individual, team, firm, industry, nation, and global system—with interdisciplinary inquiry, cumulative learning across multiple studies, and AI-enabled exploration while retaining human judgment and intellectual ownership. This integrated perspective provides a broader way to examine how capability, relationships, resilience, value creation, transformation, and responsible action develop across an increasingly connected world.

2. Two Approaches to Understanding a Complex World and Their Synergy

Supply chain scholarship increasingly views supply networks as complex adaptive systems in which relationships, capabilities, and disruptions interact across organizational boundaries (Pathak et al., 2007; Wieland, 2021), while resilience and performance unfold across interconnected micro-, meso-, and macro-level structures (Azadegan & Dooley, 2021), supporting an integrated view of individual → team → firm → industry → nation → global system, where local actions can expand into system-wide consequences. This cumulative logic is central to the article because the six-level framework is intended not as a static model, but as a way to connect many focused inquiries into larger patterns of understanding.

2.1. The Concentrated Major-Work Approach

The first approach seeks to understand complexity through a few large works organized around major questions, as illustrated by Tolstoy’s War and Peace and the sustained scholarship of Edward Gibbon, Arnold Toynbee, Immanuel Kant, and David Hume. Its major strength is depth and integration, because such works can connect many events, levels, ideas, and relationships within one coherent structure. Its limitation is that this level of concentration requires substantial time and selectivity, which can reduce responsiveness to emerging issues, unexpected observations, and new interdisciplinary possibilities.

2.2. The Cumulative Small-Work Approach

The second approach begins with smaller, more focused contexts, as illustrated by Edgar Allan Poe, Guy de Maupassant, and O. Henry, whose many short works accumulated into distinctive bodies of insight without any single story needing to explain the whole human condition. Poe produced roughly 70 tales and sketches, while Maupassant wrote about 300 short stories and O. Henry about 300, demonstrating how sustained creation of smaller works can itself generate an enduring intellectual and literary legacy. A similar pattern appears in research through numerous mini-projects examining topics such as artificial intelligence, industries, nations, historical episodes, and literary or philosophical questions, offering flexibility and rapid learning through observation, comparison, and experimentation while carrying the risk of fragmentation if the individual studies fail to connect into a recognizable larger contribution.

Figure 1 contrasts two ways of understanding a complex world and shows how they can work together. The concentrated major-work approach, illustrated by Tolstoy’s War and Peace and the works of Gibbon, Toynbee, Kant, and Hume, seeks depth, coherence, and lasting impact through a relatively small number of large integrative works, while the cumulative small-work approach, illustrated by Edgar Allan Poe, Guy de Maupassant, O. Henry, and numerous empirical or conceptual studies, develops insight through many shorter and more focused explorations. The figure places synergy between these two approaches, suggesting that repeated small-scale investigations can serve as intellectual laboratories from which larger ideas and frameworks gradually emerge. The central process is therefore one of repeated exploration → comparison → integration → broader understanding → enduring contribution, showing how focused studies and large integrative thinking can reinforce each other over time.

Figure 1. Two Approaches to Understanding the Complex World and their Synergy

2.3. Synergy: From Mini-Projects to Larger Understanding

The most promising approach combines the Tolstoyan scale of a large integrative canvas with the Poe-like discipline of examining specific situations closely, allowing smaller projects to serve as intellectual laboratories while periodic synthesis connects recurring findings into broader concepts and frameworks. Global supply chain management is especially suited to this approach because focused studies of individual decisions, supplier relationships, industry disruptions, national policies, or geopolitical conflicts can each reveal part of a larger system spanning individual, team, firm, industry, nation, and global levels. The central challenge is therefore to connect many focused inquiries to the larger picture so that repeated exploration gradually develops into deeper, more enduring integrative understanding. In practice, the Tolstoyan or Gibbonic model of producing only a few major books over a lifetime is difficult to sustain in academic systems that evaluate scholars periodically through multiple, visible, and measurable outputs; only after reaching a senior stage of their careers may scholars have greater freedom to undertake a long-term project that integrates their most serious thematic ideas into a major work.

3. Why Multiple Levels Matter in Global Supply Chain Management

The practical importance of multiple levels can be seen in everyday life. A factory disruption may affect automobile availability, a shortage of medical inputs may influence healthcare access, energy constraints may raise household costs, and government trade restrictions may alter jobs, prices, and investment. The framework therefore matters not only to firms and policymakers, but also to consumers, workers, families, and communities whose daily lives are increasingly shaped by interconnected supply chains.

3.1. Individual and Team Levels: Judgment, Trust, and Coordination

At the individual level, decisions by buyers, engineers, and managers concerning price, quality, information sharing, and risk can influence trust, product quality, delivery performance, safety, and long-term supplier relationships, making judgment, competence, character, learning, and responsibility fundamental to global supply chain performance. At the team level, complex products and services depend on cross-functional and cross-organizational collaboration among engineers, purchasing specialists, suppliers, analysts, production managers, and international partners who must communicate openly, share goals, resolve disagreements, and combine complementary expertise. Supply chain performance therefore depends not only on technology and formal systems but also on how effectively individuals and teams coordinate across professional, organizational, cultural, and national boundaries.

3.2. Firm and Industry Levels: Capability, Competition, and Ecosystem Relationships

At the firm level, organizations differ in how effectively they build and combine capabilities in supplier relationships, flexible production, alternative sourcing, digital visibility, innovation, procurement, manufacturing, logistics, and risk management, which helps explain why some recover quickly from disruption while others struggle. At the industry level, firms operate within complex ecosystems that include suppliers, technology providers, universities, governments, logistics partners, and other supporting actors whose interdependencies can amplify bottlenecks, reshape standards, and alter competitive positions. Global supply chain analysis therefore requires attention to both firm capabilities and industry-wide relationships, dependencies, power structures, capability concentrations, and value flows that cannot be understood by examining one organization in isolation.

3.3. National and Global Levels: Institutions, Geopolitics, and Systemic Transformation

At the national level, decisions about factory location, talent, financing, technology development, and critical resources are shaped by education systems, infrastructure, industrial policy, labor markets, demographics, regulations, technological capability, financial institutions, and political stability. At the global level, national systems interact through trade, technological rivalry, strategic alliances, geopolitical conflict, demographic change, climate pressures, and artificial intelligence, allowing disruptions or breakthroughs in one area to reshape sourcing, investment, industry structures, and markets across borders. Global supply chain analysis therefore requires continual movement across individual → team → firm → industry → nation → global system, because each level both shapes and is shaped by the others within an interconnected and continuously changing world.

Figure 2. The Six-Level Connected World Framework for Global Supply Chains

Figure 2 builds on established systems and network perspectives in supply chain research by presenting global supply chains as a connected six-level system in which influences move continuously across individuals, teams, firms, industries, nations, and the global system. Its distinctive purpose is not to claim that interdependence across supply networks is new, but to make explicit how human judgment, collective coordination, organizational capability, industry ecosystems, national institutions, and global forces can be examined within one integrated framework.

4. Why Literature, Philosophy, and History Matter for Global Supply Chain Management

Multi-level supply chain analysis explains where influences operate and how change moves across individual → team → firm → industry → nation → global system, while literature illuminates people and relationships, philosophy sharpens judgment and responsibility, and history reveals recurring patterns across time, thereby extending the six-level framework beyond quantitative analysis to include the human, ethical, and historical forces shaping complex supply chain decisions.

4.1. Literature: Understanding Human Character, Relationships, and Consequences

Literature helps illuminate the human side of global supply chains by showing how ambition, fear, loyalty, trust, pride, sacrifice, and misjudgment shape decisions and relationships, as seen in works such as Tolstoy’s War and Peace and the short stories of Edgar Allan Poe. It also encourages readers to see the same supply chain decision from multiple viewpoints, since managers, suppliers, employees, and communities may experience its consequences very differently. Literature therefore deepens supply chain analysis by revealing human motives, competing perspectives, and unintended consequences that formal procedures and performance measures alone may not fully capture.

4.2. Philosophy: Clarifying Judgment, Responsibility, and Meaning

Philosophy deepens global supply chain management by asking not only what organizations can do, but what they should do when decisions involve suppliers, workers, customers, communities, and future generations. Thinkers such as Kant, Hume, and Aristotle provide useful perspectives on duty, reason, character, judgment, and moral responsibility, especially when AI, automation, and global sourcing make action possible faster than institutions can establish clear rules. Philosophy therefore extends supply chain analysis beyond efficiency and feasibility toward responsibility, asking what kind of value is being created, for whom, and at what human or social cost.

4.3. History: Recognizing Patterns of Rise, Decline, Adaptation, and Overextension

History provides a practical lens for global supply chain management by showing how industrial leadership shifts over time, as seen in the changing roles of Britain, Japan, South Korea, and China in global production and trade. It also warns that firms, industries, and nations can lose advantage through complacency, excessive concentration, overextension, technological disruption, weak institutions, or failure to adapt to changing conditions. History therefore helps supply chain scholars and managers recognize recurring patterns of rise, decline, vulnerability, adaptation, and renewal that can inform how they interpret present challenges and anticipate future change.

5. Why Networked Interdisciplinary Research Matters for Global Supply Chain Management

Consider the global semiconductor shortage, where operations management, engineering, economics, information systems, sociology and psychology, political science, history, and philosophy each illuminate different dimensions of the problem, from production bottlenecks, technology, incentives, and digital coordination to human behavior, institutions, geopolitics, historical patterns, and ethical responsibility, showing why complex global supply chain challenges require interdisciplinary inquiry that connects technical, behavioral, institutional, historical, and ethical perspectives.

5.1. Networks Expand Knowledge, Capability, and Perspective

Imagine a research team studying the future of electric vehicles. One researcher understands battery technology, another knows automotive supply chains, another studies consumer markets, another examines government policy, and industry participants contribute practical knowledge about sourcing, production, cost, and implementation. When these perspectives are connected, the research question becomes much richer because the team can examine not only whether a technology works, but also whether it can be produced at scale, supplied reliably, accepted by markets, supported by policy, and sustained competitively.

Research networks also expose scholars to viewpoints they might never encounter working alone. A researcher in the United States may interpret supply chain resilience differently from a colleague in South Korea, Japan, India, Germany, or another country because each works within different industrial structures, institutions, technologies, and historical experiences. Collaboration therefore does more than add participants; it creates complementary capability by bringing together different evidence, assumptions, methods, industries, and national perspectives that can challenge narrow interpretations and improve the credibility of the final explanation.

5.2. Interdisciplinary Research Connects Multiple Levels and Complex Problems

Interdisciplinary research is essential because challenges such as artificial intelligence affect the individual, team, firm, industry, national, and global levels simultaneously, influencing decision making, coordination, productivity, business models, strategic capability, and international competition. Different disciplines contribute complementary insights, with engineering addressing technical possibilities, economics examining incentives and resource allocation, organizational and behavioral research explaining capabilities and human responses, and political science and history clarifying institutional and geopolitical conditions. Interdisciplinary research therefore helps scholars trace how complex problems move across individual → team → firm → industry → nation → global system, while recognizing that influence can also flow in the opposite direction.

5.3. From Research Networks to Cumulative Intellectual Contribution

Research networks create value not simply by producing more papers, but by allowing focused studies across industries, technologies, countries, disciplines, and time periods to function as connected intellectual experiments that reveal recurring patterns of capability, partnership, disruption, competition, and transformation. One study may generate an initial insight, another may examine it at a different level or in another country or industry, and repeated exploration → comparison → integration can gradually transform specialized findings into broader conceptual understanding. The distinctive contribution proposed here is therefore the combination of six-level analysis with interdisciplinary and cumulative inquiry, allowing smaller projects to feed progressively into larger frameworks of understanding. In the AI era, this process can be accelerated through powerful search, comparison, synthesis, and visualization tools, but it should remain supported by technical safeguards, institutional governance, continuing human oversight, and intellectual ownership of the research questions, interpretations, and conclusions (Choi et al., 2026).

6. Conclusion

The connected world is too complex to be understood from one level, one discipline, or one point of view, and this article therefore integrates six connected levels—individual, team, firm, industry, nation, and global system—with interdisciplinary and cumulative inquiry. AI gives researchers extraordinary power to search, compare, connect, visualize, and develop ideas across these levels and disciplines, but responsibility for asking meaningful questions, interpreting evidence, and making sound judgments must remain human. In the AI era, meaningful scholarship will depend on curiosity that crosses boundaries, collaboration that connects different forms of expertise, and judgment that keeps technology in its proper role as a tool while researchers retain ownership of their questions, interpretations, arguments, and conclusions. A guiding principle for future research is therefore: use powerful tools, cross intellectual boundaries, connect the levels, and never surrender ownership of the thinking.

References

Azadegan, A., & Dooley, K. (2021). A typology of supply network resilience strategies: Complex collaborations in a complex world. Journal of Supply Chain Management, 57(1), 17–26.

Bakhtin, M. M. (1984). Problems of Dostoevsky’s poetics (C. Emerson, Ed. & Trans.). University of Minnesota Press.

Choi, Y. B., Hong, P. C., & Park, Y. S. (2026). Layered control architectures for AI safety: A cybersecurity-oriented systems framework. Systems, 14(4), 447.

Christopher, M. (2016). Logistics & supply chain management (5th ed.). Pearson.

Gibbon, E. (1776–1789). The history of the decline and fall of the Roman Empire. W. Strahan & T. Cadell.

Hong, P., Kim, S.-C., Lee, A., & Kang, H. (2024). The entrepreneurial transformation process of BTS: Initiation, development, growth and expansion. Journal of Enterprising Communities: People and Places in the Global Economy.

Hong, P. C., Park, S., Wert, J. L., & Kim, M. K. (2026). Implementing dynamic capability driven innovation model: A case study of South Korea’s defence industry. Technology Analysis & Strategic Management, 1–17.

Hong, P. C., Park, Y. S., Hwang, D. W., & Jalali Sepehr, M. (2024). A growth theory perspective on the competitive landscape of shipbuilding: A comparative study of Japan, Korea, and China. Maritime Economics & Logistics.

Hume, D. (1748). An enquiry concerning human understanding. A. Millar.

Kant, I. (1781). Critique of pure reason. Johann Friedrich Hartknoch.

Nair, A., Yan, T., Ro, Y. K., Oke, A., Chiles, T. H., & Lee, S.-Y. (2016). How environmental innovations emerge and proliferate in supply networks: A complex adaptive systems perspective. Journal of Supply Chain Management, 52(2), 66–86.

National Academy of Sciences, National Academy of Engineering, & Institute of Medicine. (2005). Facilitating interdisciplinary research. National Academies Press.

Pathak, S. D., Day, J. M., Nair, A., Sawaya, W. J., & Kristal, M. M. (2007). Complexity and adaptivity in supply networks: Building supply network theory using a complex adaptive systems perspective. Decision Sciences, 38(4), 547–580.

Poe, E. A. (1840). Tales of the grotesque and arabesque. Lea and Blanchard.

Simatupang, T. M., & Sridharan, R. (2005). An integrative framework for supply chain collaboration. The International Journal of Logistics Management, 16(2), 257–274.

Toynbee, A. J. (1934–1961). A study of history (Vols. 1–12). Oxford University Press.

UNESCO. (2022). Recommendation on the ethics of artificial intelligence. United Nations Educational, Scientific and Cultural Organization.

Wieland, A. (2021). Dancing the supply chain: Toward transformative supply chain management. Journal of Supply Chain Management, 57(1), 58–73.

Zahra, S. A., & Newey, L. R. (2009). Maximizing the impact of organization science: Theory-building at the intersection of disciplines and/or fields. Journal of Management Studies, 46(6), 1059–1075.

ABOUT THE AUTHOR

Distinguished Professor, Dr. Paul Hong (Editor in Chief) — University of Toledo

Paul C. Hong is Distinguished University Professor at the University of Toledo. Drawing on more than four decades of leadership and missional service and three decades of scholarship in leadership and supply chain management, his research examines how individuals, organizations, and societies develop the capabilities necessary for resilience, adaptation, and long-term impact. His recent work integrates history, management, philosophy, and faith-based narratives to explore leadership formation through the development of identity and purpose, interests and priorities, insight and perspective, integrity and principles, and influence through stewardship—qualities that remain essential in the age of artificial intelligence.

© K-Global Scholars and Professionals Forum. All rights reserved. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

Original Article:

8 20 26 Exploring The Dynamics Of A Connected World Hong

2.11MB ∙ PDF file

Download

Download

Suggested Citation

Hong, P. C. (2026, August 20). Exploring the dynamics of a connected world: The Six-Level Connected World Framework for global supply chains. K-GSP Forum, pp. 1–13.

한글요약

연결된 세계의 역학: 글로벌 공급망을 위한 6단계 연결 세계 프레임워크의 제안

Paul C. Hong은 미국 University of Toledo의 Distinguished University Professor로, 정보시스템과 공급망 관리를 중심으로 사람·조직·산업·국가가 복잡성과 변화에 어떻게 대응하는지를 연구해 왔다. 그의 연구는 자동차 신제품 개발과 공급업체 관계에서 출발하여 글로벌 공급망 관리와 다층적·학제적 연구로 확장되어 왔다.

오늘날 글로벌 공급망은 단순히 기업과 기업을 연결하는 물류체계가 아니라 개인, 팀, 기업, 산업, 국가, 그리고 글로벌 시스템이 서로 영향을 주고받는 살아 있는 네트워크이다. 이 글은 이러한 상호연결성을 설명하기 위해 개인 → 팀 → 기업 → 산업 → 국가 → 글로벌 시스템을 연결하는 ‘6단계 연결 세계 프레임워크’를 제안한다. 한 공장의 생산 차질, 공급업체와의 관계 변화, 조직 내부의 협업 문제, 정부의 산업정책과 같은 다양한 사건들은 여러 수준을 거치며 누적·상호작용·확산되어 예상보다 훨씬 큰 파급효과를 만들어 낼 수 있다. 따라서 글로벌 공급망 관리는 개별 사건만을 분석하는 데 머물지 않고, 특정 행동과 변화가 여섯 수준을 통해 어떻게 시스템 전체의 결과로 이어지는가를 추적해야 한다.

복잡한 세계를 이해하는 지적 접근에도 서로 다른 두 가지 방식이 있다. 톨스토이의 전쟁과 평화, 기번, 토인비, 칸트, 흄의 저작처럼 몇 개의 거대한 질문을 오랫동안 집중적으로 탐구하는 방식이 있는 반면, 에드거 앨런 포, 모파상, 오 헨리처럼 수많은 작은 작품과 구체적 상황을 통해 점진적으로 인간과 세계를 이해하는 방식도 있다. 이 글은 두 방식 가운데 하나만 선택할 필요가 없으며, 많은 작은 연구를 지적 실험실로 활용하고 반복적인 탐색 → 비교 → 통합을 통해 더 큰 개념과 지속적인 지적 기여로 발전시키는 것이 더욱 생산적인 접근이 될 수 있다고 제안한다.

글로벌 공급망을 깊이 이해하기 위해서는 경영학과 공학뿐 아니라 문학, 역사, 철학의 통찰도 중요하다. 문학은 인간의 성격·욕망·두려움·신뢰와 관계를 보여주고, 역사는 산업과 국가의 흥망·적응·과잉확장과 변화의 반복적 패턴을 보여주며, 철학은 무엇이 가능한가를 넘어 무엇을 해야 하는가, 누구를 위해 가치를 창출하는가, 결과에 대해 누가 책임져야 하는가를 묻게 한다. 따라서 기술혁신과 지정학적 경쟁, 인공지능, 인구변화가 동시에 공급망을 변화시키는 시대에는 문학은 사람을 이해하게 하고, 역사는 패턴을 보게 하며, 철학은 판단을 날카롭게 한다는 점에서 이 세 분야의 통찰은 글로벌 공급망 관리에 더욱 큰 가치를 갖는다.

마지막으로 이 글은 복잡한 공급망 문제를 해결하기 위해 운영관리, 공학, 경제학, 정보시스템, 심리·사회학, 정치학, 역사, 철학을 연결하는 네트워크형 학제연구가 필요하다고 강조한다. AI는 연구자가 더 넓게 검색하고 비교하며 여러 수준과 학문을 연결하는 강력한 도구가 될 수 있지만, 연구 질문의 선택, 의미 해석, 판단, 논증과 결론에 대한 책임은 여전히 인간 연구자에게 남아 있어야 한다. 결국 미래의 중요한 혁신적 통찰은 수준과 학문, 시간을 가로질러 사건과 패턴을 연결하고, AI로 탐구의 범위를 넓히면서도 인간의 판단과 지적 소유권을 지키는 연구에서 나타날 가능성이 크다.

K-Global Scholars and Professionals Forum

Guide for Authors

Manuscript preparation, submission, and editorial standards

This guide tells prospective authors how to prepare and submit work to the Journal of K-GSP. It supplements the Journal’s editorial policies and procedures, which govern review authority and decision-making. Where a question of scholarly …

Read more

It was more than 30 years ago when representatives of various countries first recognized the fact that extreme weather caused by global warming could endanger the survival of humanity and resolved to take action. With little time left to respond and the realization that missing this opportunity might make it impossible to return the climate to its previous state, they decided to reduce the amount of greenhouse gases emitted in daily life and began implementing it 2021. However, the plan is not being properly performed due to various reasons.

There was a research report that if gas emissions continue at the current rate, Korea where most of the region is temperate will soon turn subtropical. Uljin and Gangneung became subtropical between 2016 and 2025. There is also a prediction that all coastal cities across Korea will become subtropical by 2040.

A severe heatwave warning was issued in Gyeongsan and Pohang in early July 2026. Before the intense heat season started, the Tibetan high-pressure system and the North Pacific high-pressure system covered Korean Peninsula at once, and hot humid southwest wind blew underneath.

It was the result of record-breaking heatwaves hitting various parts of the world due to the influence of Super El Niño. In Europe and North America, the deadly heatwave took lots of lives recently. Omega Heat Dome trapped Eiffel tower, making its operation time shortened. America was suffering from a severe abnormal high temperature since 1950s, and heatwave warnings have been issued for about two-thirds of the country.

Among the natural disasters such as heavy rain, typhoons, heatwave, strong winds, and heavy snow, heatwave is the worst, and named ‘silent killer’. I live in Albuquerque, New Mexico, in which June is the hottest month of the year. There were many days when the maximum temperature exceeds the normal body temperature. Days when the temperature exceeded 90°F (32.2°C) continued for an entire week. On top of the extreme heat, a drought is also tormenting people. Therefore, various mass media frequently advise that we must prepare to withstand the severe heat which strikes unexpectedly.

Find a way to stay cool while spending most of the day at home. Wear loose clothes made of light-colored, moisture-absorbent natural fibers. Leave the window open late at night or early in the morning when the temperature drops. Place a bowl filled with ice cubes in front of the fan. Draw the curtain, screen or blind in the room facing the sun.

Set the air conditioner at 24 to 25°C. If it is hot at home, try to spend time in a nearby public building that is well-air-conditioned and cool. Fill water in several 500mL bottles and drink from time to time so that you know how much you drink. People who sweat a lot have to replenish their losses by consuming sports drinks or electrolyte powder. To lower your body temperature, place a towel soaked in cold water or an ice pack on your wrists, neck, armpits, or groin. Wrap your head with a large handkerchief or scarf soaked in water.

Take a bath with lukewarm water. Be careful, because if the water is too cold the body reacts and the core temperature is sustained for a longer period. It is best to avoid alcohol or beverages high in caffeine due to their diuretic effects. As much as possible, avoid foods high in protein or fat that increase the body’s metabolic heat, and eat fresh vegetables, salad, watermelon, cucumbers.

The sunlight is strongest between 11am and 3pm. When sweltering heat persists, anyone can be affected by prolonged exposure to sunlight. Many people realize that heatwaves are dangerous to humans, but they do not think that they themselves could be included in that danger. Plan to do strenuous exercise or outdoor activities early in the morning or late in the afternoon to avoid exposure to strong sunlight. Healthy people also should avoid the mid-summer heat.

Since one cannot stay at home all the time, more thorough preparation is required when engaging in outdoor activities. Pay attention to real-time emergency health alerts and air quality programs. Wear light, loose clothes and a wide-brimmed hat. Cool place for regular rest should be provided at the workplace. It must also be prepared to prevent dehydration. Seniors aged 65 or older, children, patients taking medication, pregnant women, outdoor workers, and athletes must pay special attention to those things.

It is time to be well aware of the symptoms of diseases caused by high temperatures. Heat exhaustion, which occurs when there is a lack of fluids and salts due to excessive sweating, manifests symptoms such as extreme fatigue, dizziness, nausea, and cold, wet skin. Heatstroke, which occurs when the body is unable to release heat while in a hot and humid environment, manifests a body temperature exceeding 40°C, a lack of sweating, loss of consciousness, and collapse. Those symptoms indicate that one must immediately move to a cool place and rest while slowly drinking water or a sports drink to replenish fluids and electrolytes, or take the patient to the emergency room immediately.

While treating symptoms is important, it is even more crucial to take precautions in advance to prevent the problem from occurring. The unprecedented heatwave is an emergency caused by climate change. It raises the possibility that this could become a social disaster, placing humanity in grave danger. All of us must accurately recognize the reality and respond swiftly and decisively to climate change.

폭염

지구의 온난화가 초래하는 극단적인 날씨가 인류의 생존을 위태롭게 할 수 있다는 사실을 인식하고 여러 나라 대표들이 그에 대처하기로 결의한 것이 30년도 더 전이었다. 대처할 시간이 얼마 남지 않아 기회를 놓치면 기후를 이전의 상태로 되돌리기가 불가능할 수도 있다는 생각에, 2021년에는 일상에서 배출되는 온실 가스의 양을 줄이기로 결정하고 실행에 들어갔다. 하지만 그 계획은 여러 가지 이유로 제대로 실행되지 못하는 실정이다.

가스의 배출이 지금처럼 계속되면 대부분의 지역이 온대인 우리 나라가 아열대로 바뀐다는 연구 보고가 있었다. 경북 울진과 강원 강릉은 2016~2025년에 이미 아열대가 되었다. 전국 해안의 모든 도시가 2040년에는 아열대가 될 것이라는 예측도 나왔다. 2026년 7월 초에는 경북 경산시와 포항시에 폭염 중대 경보가 발령되었다. 본격적인 무더위가 시작되기도 전에 티베트 고기압과 북태평양 고기압이 상공을 겹치기로 덮고, 뜨겁고 습한 남서풍이 그 아래로 불어 한반도가 ‘찜통’에 갇힌 것이었다. ‘슈퍼 엘니뇨’의 영향으로 세계 곳곳에 기록적인 폭염이 찾아온 결과였다고 한다.

최근 유럽과 북미 등에서도 살인적인 폭염이 수천 명의 목숨을 앗아갔다. ‘오메가 열돔’에 갇혀 에펠 탑 등 관광 명소의 운영이 단축되었고, 사망자가 속출하고 있다. 미국 역시 1950년대 이후 가장 심각한 이상 고온 현상이 나타나 국토의 3분의 2가량에 폭염 경보가 내려졌다. 단순한 재난을 넘어 일상을 위협하는 사회적 재난으로 번졌음을 보여준다. 자연 재난에 드는 호우, 태풍, 강풍, 폭염, 대설 가운데 가장 많은 사망자를 내는 폭염은 ‘조용한 살인자’로 알려져 있다.

내가 살고 있는 뉴 멕시코 주 앨버커키는 연중 유월이 가장 더운 달이다. 최고 기온이 성인의 정상 체온을 웃도는 날이 많고, 해가 진 다음에도 섭씨 20도가 넘는다. 기온이 화씨 90도 (섭씨 32.2도)를 넘는 날이 한 달 내내 계속된다. 더위가 극성을 부리는 데다가 가뭄까지 들어 사람을 괴롭힌다. 그래서 각종 대중 매체에서는 예상치 못하게 닥쳐와 사람을 방심하게 만드는 심한 더위를 이겨낼 준비를 해야 한다고 자주 일러준다.

하루 중 대부분의 시간을 집에서 보내면서는 시원한 몸으로 사는 방법을 모색해야 한다. 헐렁하고, 가볍고, 옅은 색에 흡습성을 가진 천연 섬유로 지은 옷을 입는다. 기온이 내리는 밤 늦게 아니면 아침 일찍 창문을 열어놓는다. 선풍기 앞에 얼음이 담긴 그릇을 놔둔다. 해가 비치는 쪽의 방에 휘장을 치고, 햇빛 가리개나 발 또는 커튼으로 볕이 들지 않게 한다.

에어컨을 24~25°C에 맞춰 가동한다. 집이 더우면 가까운 곳의 냉방이 잘 되어 시원한 공공 건물에서 시간을 보내는 방법을 모색한다. 몇 개의 500mL 들이 병에 물을 담아 놓고 수시로 조금씩 마셔서 자기가 마시는 양을 어림한다. 땀을 많이 흘리는 사람은 스포츠 음료나 전해질 분말을 섭취하여 손실을 보충한다. 체온을 낮추려면 찬물에 적신 수건이나 아이스 팩을 목, 손목, 겨드랑이, 사타구니 등에 넣는다. 물에 적신 큰 손수건이나 스카프로 머리를 감싼다.

미지근한 물로 목욕한다. 물이 너무 차가우면 몸이 반응하여 신체 중심부의 온도가 더 오래 지속되기 되기 때문에 조심한다. 알코올이나 카페인이 많은 음료는 이뇨 작용이 있어서 피하는 것이 좋다. 몸의 대사열을 증가시키는 단백질이나 지방질이 많은 음식은 되도록이면 피하고, 신선한 야채, 샐러드, 수박, 오이 등을 먹는다.

햇볕은 아침 11시에서 오후 3시 사이에 가장 강하다. 찜통 더위가 계속되는 때는 햇볕에 오래 노출되면 누구든 영향을 받을 수 있는데, 많은 이들이 폭염은 사람에게 위험하지만 정작 자신이 거기에 포함될 수 있다고는 생각하지 않는다. 힘드는 운동이나 야외 활동은 이른 아침이나 오후 늦은 시간에 하도록 계획을 짠다. 건강한 사람도 한더위는 피하는 것이 좋다.

집에만 있을 수는 없어서 야외 활동을 할 때는 더 철저한 준비가 요구된다. 실시간 비상 건강 경보나 대기의 질 프로그램 등에 주의를 기울인다. 가볍고 헐렁한 옷에 챙이 넓은 모자를 쓴다. 일터에는 규칙적으로 쉴 수 있는 시원한 곳이 마련되고, 탈수를 막을 준비도 되어있어야 한다. 65세 이상의 노인, 어린이, 약을 복용하는 환자, 임신부, 야외 노동자, 운동 선수 등은 각별한 관심을 기울여야 한다.

고온이 초래하는 질환의 증상을 잘 알고 있어서 나타나면 바로 대처해야 한다. 땀을 많이 흘린 나머지 수분과 염분이 부족하여 발생하는 열 탈진은 극심한 피로, 어지럼증, 메스꺼움, 차고 젖은 피부 등의 증상을 보인다. 고온 다습한 곳에 있을 때 몸의 열을 밖으로 내보내지 못하여 생기는 열사병은 40°C가 넘는 체온에 땀이 나지 않거나 의식을 잃고 쓰러지는 증상을 나타낸다. 즉시 시원한 곳으로 이동해 휴식을 취하면서 물이나 스포츠음료를 천천히 마셔 수분과 전해질을 보충하거나, 곧바로 응급실로 데려가야 하는 상태임을 말해준다.

증상의 치료도 중요하지만 문제 자체가 발생하지 않도록 미리 조심하는 것이 더 중요하다. 역대에 없었던 폭염은 기후의 변화가 초래하는 비상 사태다. 사회적 재난이 되어 인류가 심각한 위험에 처할 수 있음을 예상하게 만든다. 우리 모두가 현실을 정확하게 인식하고, 기후의 변화에 신속하고 단호하게 대처해야 할 때다.

Original Article:

Heatwave Original

189KB ∙ PDF file

Download

Download

ABOUT THE AUTHOR

Dr. Chung Gil Lee is Professor Emeritus of Veterinary Medicine at Chonnam National University, where he served on the faculty until 2007. He holds an M.Sc. and Ph.D. from James Cook University, Australia, and began his career as an Interpreter Officer in the Korean Army. A graduate of Chonnam National University (1963), he is currently a member of the Korean Literature Society of America.

© K-Global Scholars and Professionals Forum. All rights reserved. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

Gani Choi Rip’s Philosophy of Calligraphy and Painting Connoisseurship: Korean Art Criticism and Its Significance in the Age of Generative AI

Abstract

Gani Choi Rip (崔岦, 1539–1612), one of the major literary intellectuals of the mid-to-late Joseon Dynasty, has generally been studied as a writer, diplomat, classical prose stylist, and scholar rather than as an important participant in the history of Korean art criticism. Yet his collected works, Gani-jip (簡易集), contain a substantial body of writings on calligraphy, painting, illustrated albums, commemorative pictures, and the preservation and interpretation of visual objects. These writings reveal a sophisticated mode of connoisseurship that cannot be reduced to judgments of technical skill. Choi Rip repeatedly connected calligraphic form with the moral and intellectual character of the writer; interpreted painting through poetry, memory, and lived experience; distinguished visual resemblance from artistic value; examined provenance and attribution cautiously; and understood artworks as repositories of personal, historical, and cultural memory.

This article reconstructs five distinctive principles of Choi Rip’s art connoisseurship: (1) reading character through calligraphy, (2) interpreting painting as a dialogue between image and poetry, (3) distinguishing formal resemblance from artistic vitality, (4) evaluating provenance and authenticity through critical comparison, and (5) reading artworks as vessels of memory and historical experience. It then identifies five broader contributions to the history of Korean art appreciation: the integration of moral cultivation and visual judgment, the strengthening of the scholar’s role as connoisseur and interpreter, the interaction of calligraphy, poetry, and painting, the development of historically conscious attribution, and the preservation of artworks as cultural memory. Finally, the article compares Choi Rip’s philosophy with contemporary debates concerning generative artificial intelligence (GenAI), creativity, authorship, authenticity, and human agency. Recent research shows that GenAI can increase artistic productivity and expand exploration while simultaneously raising concerns about declining average novelty, authenticity, agency, and authorship. Choi Rip’s philosophy therefore offers a useful humanistic framework for AI-era art criticism: evaluate not merely what an image looks like, but who shaped it, why it was made, what cultural memory it carries, and what kind of human judgment gives it meaning.

Keywords: Choi Rip; Gani-jip; Joseon Dynasty; Korean art history; calligraphy; literati painting; art connoisseurship; authenticity; generative AI; human-AI creativity

1. Introduction: Rediscovering an Art Critic Hidden Inside a Literary Scholar

The history of Korean art criticism is often narrated through painters, collectors, theorists, albums, painting manuals, and later literati masters. Yet the history of connoisseurship also resides in a less obvious body of sources: the writings of literary scholars who encountered paintings and calligraphy as part of everyday intellectual life. Gani Choi Rip belongs to this category. Although he was not a professional painter, his Gani-jip preserves numerous texts written in response to calligraphic works, landscape paintings, bamboo paintings, commemorative albums, and ancient paintings. The surviving collection was first printed in 1631 and consists of nine volumes; the modern Han’guk Munjip Ch’onggan edition derives from that early textual tradition (Choi, 1996). (MK Study)

This evidence requires a reconsideration of Choi Rip’s place in Korean cultural history. The appropriate question is not whether he should be labeled a “professional art theorist.” He should not. Rather, the more productive question is: What principles of visual judgment emerge when his writings on calligraphy and painting are read together?

This distinction is important because Joseon visual culture was deeply embedded in literary culture. Scholar-officials could function simultaneously as writers, collectors, patrons, calligraphers, viewers, and interpreters. Modern scholarship has demonstrated that Joseon literati played significant roles in the production, collection, and patronage of painting, while the conceptual boundaries of “literati painting” itself developed historically rather than appearing as a fully formed category from the beginning (Park, 2018; Yungman, 2012). (KCI)

Choi Rip’s writings therefore provide an especially valuable window into how a sixteenth- and early seventeenth-century Korean intellectual actually looked at visual objects. His criticism is not organized as a formal treatise. It is embedded in inscriptions, notes, prefaces, and poetic responses. Precisely because of this, his writings reveal connoisseurship as a practice of attentive reading rather than merely a theoretical system.

2. Historical and Intellectual Context

Choi Rip lived through one of the most turbulent periods in Korean history. Born in 1539 and dying in 1612, he experienced the intellectual transformations of the late sixteenth century, the Imjin War, and the political and cultural reconstruction that followed it. He was renowned for literary composition and served the state in positions requiring sophisticated documentary and diplomatic writing. His literary reputation has consequently tended to overshadow his engagement with visual culture. (한국민족문화대백과사전)

The intellectual environment in which Choi Rip operated was characterized by an intimate relationship among literature, calligraphy, painting, historical learning, and moral cultivation. In East Asian literati culture, visual art was rarely isolated from textual culture. A painting might contain an inscription; a poem might respond to a landscape; calligraphy could function as an expression of the writer’s mind; and an album could become a repository of relationships among scholars.

Joseon intellectuals also inherited Chinese traditions of art criticism, including the emphasis on spirit resonance, brushwork, learning, personality, historical lineage, and the relationship between painting and poetry. Yet these concepts were not simply imported unchanged. Korean scholars interpreted them through their own social, ethical, and literary circumstances. Research on Joseon literati painting emphasizes that the category of “literati painting” itself has a complex historical genealogy and that the terminology and conceptual framework changed over time (Park, 2018). (KCI)

Choi Rip’s writings belong to this transitional intellectual environment. They are particularly valuable because they show an active viewer thinking through artworks rather than merely reproducing a fixed theoretical vocabulary.

3. Five Distinctive Principles of Choi Rip’s Art Connoisseurship

3.1 Reading Character Through Calligraphy

Perhaps the clearest statement of Choi Rip’s aesthetic philosophy appears in his “A Note on a Small Folding Screen of Toegye’s Calligraphy” (退溪書小屛識).

Writing about calligraphy by the celebrated scholar Yi Hwang (Toegye), Choi Rip states that calligraphy is “heart-painting” (心畫): the visible form of writing allows the viewer to imagine the person who produced it. He then examines Toegye’s script, emphasizing its dignity, strength, discipline, and consistency even when written in cursive form (Choi, 1996, Vol. 3). (Wikisource)

This passage provides a fundamental principle of Choi Rip’s connoisseurship:

The visual form of calligraphy is evidence of the cultivated mind behind it.

This does not mean that Choi Rip mechanically assumed that beautiful handwriting always indicates moral virtue. Rather, he reads formal characteristics—regularity, restraint, strength, and disciplined movement—as traces of sustained self-cultivation. The artwork becomes a record of character.

This approach is particularly significant because it expands calligraphic criticism beyond questions of handwriting technique. Line quality is not merely a formal property; it becomes evidence of intellectual and ethical formation.

In contemporary terms, Choi Rip’s method resembles a form of embodied aesthetic interpretation. The brush records physical movement, but physical movement is shaped by habit, training, intention, temperament, and discipline. His criticism therefore moves from visible form to invisible character.

3.2 Painting as a Dialogue Between Image and Poetry

A second distinctive principle appears in Choi Rip’s “A Note on Yi Eokye’s Inscribed Landscape Painting” (李鵝溪題詩山水圖識).

Here he offers a striking formulation:

“Painting is a silent poem; poetry is a painting with sound.”

(Choi, 1996, Vol. 3).

For Choi Rip, painting and poetry are not competing forms. They complete one another. A landscape can communicate visually what poetry communicates verbally, while poetry can articulate what the visual image evokes but cannot literally say. (Wikisource)

This principle is important for understanding Korean literati culture. The painting is not simply an image of mountains and rivers. It is an intellectual space in which landscape, memory, emotion, poetry, and moral reflection converge.

Choi Rip’s approach therefore represents a multimodal theory of appreciation centuries before the term “multimodal” became common. He does not isolate the visual object from the literary ecosystem surrounding it.

This perspective also anticipates a major challenge in contemporary digital art. AI-generated images are frequently evaluated as isolated visual outputs, while the prompt, conversation, editing process, training data, cultural references, and human intention remain invisible. Choi Rip’s approach reminds us that the image alone may never constitute the complete artwork.

3.3 Distinguishing Formal Resemblance from Artistic Vitality

Choi Rip’s writings also demonstrate a sophisticated distinction between merely resembling something and successfully conveying its essence.

In his note on a landscape painting owned by Lord Geumgye, he describes the pleasure of mountains and water as something deeply rooted in human experience. He observes that people are repeatedly attracted to particular landscape forms and that paintings can preserve an experience that cannot always be physically entered or possessed (Choi, 1996, Vol. 3). (Wikisource)

This suggests a third principle:

A successful artwork does not simply reproduce an object; it reconstructs an experience.

Choi Rip’s criticism consequently moves beyond literal representation. A landscape painting becomes successful when the viewer can recover something of the experience, mood, memory, or spiritual resonance associated with landscape.

This distinction is especially relevant to generative AI. Contemporary image generators can produce technically convincing landscapes within seconds. But technical resemblance and experiential meaning are not identical. The most visually accurate mountain image may not communicate the memory of a particular mountain, the experience of exile, the emotional significance of home, or the cultural meaning of a landscape.

Choi Rip therefore offers a useful criterion for evaluating AI art: Does the work merely look convincing, or does it communicate a meaningful human experience?

3.4 Critical Attribution: Authenticity Requires Comparison and Doubt

Perhaps the most surprisingly modern aspect of Choi Rip’s connoisseurship appears in his “A Note on an Ancient Painting Owned by Yi So-yun” (李少尹所有古畫識).

The painting had been acquired after the devastation of war. Its owner suspected that it might have been painted by An Gyeon, one of Korea’s most celebrated early painters. Choi Rip does not simply accept this attribution. He acknowledges that he is not an expert painter, but he explains that he had seen authentic works associated with An Gyeon and that certain details seemed consistent with that hand. At the same time, he examines the painting’s deterioration, provenance, stylistic features, clothing, landscape structures, and possible historical chronology. He explicitly expresses uncertainty concerning some aspects of the attribution (Choi, 1996, Vol. 3). (Wikisource)

This is an important example of critical connoisseurship.

Choi Rip does not treat attribution as an act of intuition alone. He compares:

• stylistic characteristics,

• historical chronology,

• material condition,

• provenance,

• iconographic details,

• and regional conventions.

His willingness to acknowledge uncertainty is equally important.

In other words, expertise does not mean claiming certainty where evidence is incomplete. A trustworthy connoisseur knows the difference between observation, inference, and speculation.

This principle has direct relevance to AI-based art authentication. Modern computer vision can compare brush patterns, composition, pigments, signatures, and stylistic features at extraordinary scale. Yet algorithmic confidence is not equivalent to historical truth. Provenance, material evidence, restoration history, and contextual knowledge remain essential.

Choi Rip’s skeptical method therefore provides an early humanistic model of evidence-based authentication.

3.5 Artworks as Vessels of Memory and Historical Experience

A fifth principle is visible throughout Choi Rip’s writings on paintings and illustrated albums: artworks preserve relationships and memories.

His note on Ha Dae-i’s bamboo painting begins with admiration for the deceased artist’s natural talent. Yet the passage quickly becomes an elegy. The painting has acquired significance because its creator is gone. The object becomes a material trace of a human relationship and a lost world (Choi, 1996, Vol. 3). (Wikisource)

Similarly, Choi Rip’s writing about commemorative albums and paintings repeatedly emphasizes the value of preserving visual records of people and events. In one passage he observes that without paintings, people separated by time could no longer confront one another visually; painting allows past faces and relationships to remain present.

This leads to a fifth principle:

The value of art increases when it preserves human memory across time.

The artwork is therefore not merely an aesthetic commodity. It is a form of cultural memory.

This principle is especially important in the digital age, when images can be generated infinitely but may also become infinitely disposable. Choi Rip reminds us that cultural value is partly determined by what an object enables later generations to remember.

4. Five Contributions to the History of Korean Art Appreciation

4.1 Integrating Moral Cultivation and Visual Judgment

Choi Rip’s interpretation of Toegye’s calligraphy demonstrates a distinctive Joseon tendency to connect artistic form with intellectual and moral cultivation. His approach reinforces a broader literati culture in which calligraphy was not simply handwriting but a visible manifestation of cultivated character.

This should not be interpreted as a uniquely Choi Rip invention. Rather, his contribution lies in providing a particularly clear example of how this principle operated in actual acts of appreciation.

4.2 Strengthening the Role of the Scholar as Connoisseur

Choi Rip demonstrates that meaningful art criticism did not require professional artistic production. A learned scholar could evaluate paintings through literature, history, ethics, memory, and visual observation.

This is significant for Korean art history because Joseon painting culture depended heavily upon networks of scholar-viewers, collectors, patrons, and writers. Research on early Joseon literati painting confirms the importance of collecting and writing as components of art culture (Yungman, 2012). (KCI)

4.3 Reinforcing the Unity of Poetry, Calligraphy, and Painting

Choi Rip’s “silent poem” formulation captures the integrated aesthetic characteristic of East Asian literati culture. Painting, poetry, and calligraphy could operate as mutually reinforcing forms of expression.

His writings therefore provide valuable evidence for studying the intellectual environment from which later Korean literati painting developed.

4.4 Introducing Historical Consciousness Into Connoisseurship

His discussion of ancient painting demonstrates that attribution requires historical reasoning. The viewer must ask when a work could have been produced, what visual conventions belong to a particular period, and whether the material evidence is compatible with the proposed attribution.

This historical consciousness is particularly valuable because later Joseon art theory increasingly developed sophisticated distinctions among different modes and lineages of painting. Modern scholarship has shown that Joseon literati gradually developed more differentiated conceptions of painting, including distinctions associated with literati and court painting (Jang, 2014). (KCI)

4.5 Preserving Art as Cultural Memory

Finally, Choi Rip’s inscriptions demonstrate that writing about art could itself become part of the preservation process. An artwork accompanied by an inscription was transformed into a richer cultural artifact containing visual, literary, biographical, and historical information.

This principle is particularly important today. Digital reproduction can preserve an image while simultaneously detaching it from provenance and context. Choi Rip’s model suggests that preservation should include not merely pixels but metadata, biography, provenance, interpretation, and cultural context.

5. Choi Rip and the Problem of Creativity in the AI Era

The emergence of generative AI has transformed the meaning of artistic production. Text-to-image systems can now generate visually sophisticated works from natural-language prompts. Research based on millions of artworks has found that generative AI can increase human creative productivity and expand the exploration of possible ideas, while average novelty can decline as creators increasingly converge around similar visual patterns (Zhou & Lee, 2024). (OUP Academic)

This produces a paradox.

AI can make more images than any previous technology, but abundance does not automatically produce cultural significance.

Choi Rip’s philosophy offers a useful response. He encourages us to distinguish production from meaning.

A machine can produce an image.

A human must still decide:

• Why this image?

• Why this subject?

• Why this style?

• What experience does it represent?

• What tradition does it engage?

• What does it remember?

• What should be preserved?

• What should be rejected?

Recent scholarship similarly emphasizes the continuing importance of human agency in AI-supported creativity. Research on human-AI co-creativity argues for moving beyond a simple “AI versus human” model toward a collaborative framework in which human judgment and machine generation interact (Medeiros & Cropley, 2025). (DOI)

Thus, Choi Rip’s connoisseurship can be transformed into a model for AI-era critical judgment.

6. From Human Connoisseurship to AI-Augmented Connoisseurship

AI can potentially improve art criticism in four major ways.

First, it can conduct large-scale visual comparison. An AI system can compare thousands of works against a candidate painting far faster than a human scholar.

Second, AI can assist provenance research by connecting names, dates, collections, auction records, inscriptions, and digitized manuscripts.

Third, multimodal AI can connect image and text, enabling simultaneous analysis of paintings, inscriptions, poems, biographies, and historical documents.

Fourth, AI can identify patterns that humans might overlook.

Yet Choi Rip’s philosophy suggests that AI should function primarily as an assistant to judgment, not as the final judge.

This is consistent with recent research emphasizing human agency in generative-AI systems. Human-complementary approaches can enhance creativity and innovation, but human agency remains central to determining how AI benefits are interpreted and distributed (Baptista et al., 2025). (ScienceDirect)

The ideal future connoisseur may therefore be neither purely human nor purely algorithmic.

It may be AI-augmented human connoisseurship.

The machine supplies scale.

The scholar supplies context.

The algorithm identifies patterns.

The historian evaluates evidence.

The AI generates possibilities.

The human decides which possibilities deserve cultural meaning.

7. Authenticity: Choi Rip’s Most Important Lesson for Generative AI

Perhaps the strongest connection between Choi Rip and contemporary AI art concerns authenticity.

Research on AI-assisted visual art suggests that audiences may perceive AI-generated or AI-assisted works as innovative while simultaneously judging them less authentic or less labor-intensive, with perceived authenticity influencing valuation (Co-creating Art with Generative Artificial Intelligence, 2024). (ScienceDirect)

This tension is remarkably close to the problem implicit in Choi Rip’s criticism.

For him, a work’s value cannot be reduced to surface appearance. The history behind the object matters.

Who made it?

Under what circumstances?

How was it transmitted?

Who preserved it?

What does it reveal about its creator?

The same questions should be asked of AI-generated art.

A future museum label may need to disclose:

Human creator → Concept → Prompt → Model → Training context → Selection → Editing → Final work → Provenance.

Such transparency would transform AI art from an opaque output into a documented creative process.

This is where Choi Rip’s philosophy becomes more than historical curiosity. It becomes a possible foundation for AI art provenance ethics.

8. Toward a Choi Rip Humanistic Framework for AI Art

Based on the five principles reconstructed above, a Choi Rip-inspired framework for AI-era art criticism can be expressed through five questions.

Character

Who is the human agent behind the work, and what values or intentions shape the creation?

Vitality

Does the work communicate something beyond technical correctness and visual novelty?

Context

What cultural, historical, literary, and artistic traditions does the work engage?

Evidence

Can its provenance, authorship, production process, and claims of originality be independently examined?

Memory

What human experience, relationship, event, or cultural meaning does the work preserve?

This framework does not reject AI.

On the contrary, it welcomes AI while refusing to confuse computational capability with cultural judgment.

Recent studies indicate that human-AI co-creation can expand artistic possibilities, but questions of agency and authorship become increasingly complex as AI becomes more deeply embedded in the creative process (Bomba & De Angeli, 2025; Rafner et al., 2025). (ScienceDirect)

A Choi Rip framework therefore places human interpretation at the center of technological augmentation.

9. Conclusion: From the Brush to the Algorithm

Gani Choi Rip’s writings on calligraphy and painting reveal an intellectual tradition considerably richer than a simple appreciation of beautiful objects. His connoisseurship was a form of cultural interpretation.

He looked at calligraphy and saw character.

He looked at painting and heard poetry.

He looked at landscape and recovered experience.

He looked at an ancient painting and questioned attribution.

He looked at a preserved image and remembered the person who was gone.

These five practices constitute a coherent philosophy of art appreciation: character, dialogue, experience, evidence, and memory.

Choi Rip should not therefore be presented as the sole founder of Korean art criticism or as a systematic art theorist in the modern sense. His importance is subtler and, in many respects, more interesting. His scattered writings demonstrate how an accomplished Joseon intellectual practiced connoisseurship at the intersection of literature, ethics, history, friendship, memory, and visual perception.

That insight becomes especially powerful in the age of generative AI.

AI can now generate thousands of images in the time required for a human artist to sketch one. It can imitate styles, reconstruct historical appearances, produce landscapes, and create calligraphic simulations. It can also assist scholars in comparing enormous collections of visual material.

But abundance creates a new problem.

When images become almost free, meaning becomes scarce.

When imitation becomes effortless, authenticity becomes difficult to establish.

When generation becomes automatic, human judgment becomes more—not less—important.

Choi Rip’s philosophy offers an answer to this problem. The future of art should not be evaluated solely by asking whether an image is beautiful, realistic, innovative, or technically sophisticated. We must also ask whether it expresses meaningful human intention, whether its provenance is credible, whether its cultural context is understood, whether its creative process is transparent, and whether it contributes something worth remembering.

Recent research supports the idea that the most productive relationship between AI and creativity may be collaborative rather than substitutive. GenAI can increase productivity and expand the space of possible ideas, but human ideation, selection, interpretation, and agency remain decisive (Zhou & Lee, 2024; Medeiros & Cropley, 2025). (OUP Academic)

The lesson of Choi Rip can therefore be reformulated for the twenty-first century:

The algorithm may generate the image, but the human being must still determine what the image means.

The brush has become an algorithm.

The studio has become a human-AI interface.

The album has become a digital archive.

Yet the fundamental question remains unchanged:

What does this work tell us about the human being who created, selected, preserved, and interpreted it?

In that sense, Choi Rip’s philosophy has not become obsolete.

It has become newly necessary.

References

Bomba, F., & De Angeli, A. (2025). Agency and authorship in AI art: Transformational practices for epistemic troubles. International Journal of Human-Computer Studies, 205, 103652. https://doi.org/10.1016/j.ijhcs.2025.103652 (ScienceDirect)

Choi, R. [崔岦]. (1996). Gani-jip [簡易集; Collected works of Gani Choi Rip]. Minjok Munhwa Ch’ujinhoe. (Original work published 1631). (MK Study)

Jang, E.-Y. (2014). The literati’s point of view on painting in the late Joseon Dynasty. Kyujanggak, 44, 1–28. https://doi.org/10.22943/kyujg.2014..44.001 (KCI)

Medeiros, K., & Cropley, D. H. (2025). Human-AI co-creativity: Does ChatGPT make us more creative? The Journal of Creative Behavior. https://doi.org/10.1002/jocb.70022 (DOI)

Park, E. S. (2018). Yuhwa, literati painting in the Joseon period: The origin and meaning of the term. Korean Bulletin of Art History, 50, 245–266. https://doi.org/10.15819/rah.2018..50.245 (KCI)

Rafner, J., Zana, B., Hansen, I. B., Ceha, S., Sherson, J., Benedek, M., & Lebuda, I. (2025). Agency in human-AI collaboration for image generation and creative writing: Preliminary insights from think-aloud protocols. Creativity Research Journal. https://doi.org/10.1080/10400419.2025.2587803 (DOI)

Yungman, B. (2012). Collecting and writing about literati painting during the early Chosŏn period. Journal of East Asian Art History, 1, 181–220. https://doi.org/10.22978/aarth.2012.1..181 (KCI)

Zhou, E., & Lee, D. (2024). Generative artificial intelligence, human creativity, and art. PNAS Nexus, 3(3), pgae052. https://doi.org/10.1093/pnasnexus/pgae052 (OUP Academic)

Chu, W., Baxter, D., & Liu, Y. (2025). Exploring the impacts of generative AI on artistic innovation routines. Technovation, 143, 103209. https://doi.org/10.1016/j.technovation.2025.103209 (ScienceDirect)

Garcia, M. B. (2025). The paradox of artificial creativity: Challenges and opportunities of generative AI artistry. Creativity Research Journal, 37(4), 755–768. https://doi.org/10.1080/10400419.2024.2354622 (Taylor & Francis Online)

Guo, W., Liang, Z., Wang, C., Li, X., Hu, H., Chen, S., Yu, Q., & Zhao, Q. (2025). Student-AI collaborative creative problem-solving: The role of human agency. Computers & Education, 239, 105433. https://doi.org/10.1016/j.compedu.2025.105433 (ScienceDirect)

Sobrinho, A. B. F. (2026). Exploring how generative AI reshapes artistic creativity. The Journal of Creative Behavior, 60(2), e70113. https://doi.org/10.1002/jocb.70113 (Wiley Online Library)

Ukpaka, P. M. (2025). The creative agency of large language models: A philosophical inquiry. AI and Ethics, 5, 2455–2466. https://doi.org/10.1007/s43681-024-00557-9 (DOI)

Zhou, E., Lee, D., & Gu, B. (2025). Who expands the human creative frontier with generative AI: Hive minds or masterminds? Science Advances, 11(36), eadu5800. https://doi.org/10.1126/sciadv.adu5800 (PubMed)

Publication note: The strongest primary-source basis for the paper is Choi Rip’s Gani-jip, especially Volume 3. The text explicitly contains the calligraphy, landscape, bamboo, landscape-collection, scattered-painting, and ancient-painting notes used above. (Wikisource) The paper’s central interpretive contribution is therefore not the unsupported claim that Choi Rip was a formal “art theorist,” but the more defensible argument that his dispersed writings constitute an identifiable method of Joseon-period visual connoisseurship. Recent scholarship on Joseon literati painting and GenAI creativity provides the historical and contemporary framework for that argument. (KCI)

간이 최립의 서화감평 철학: 한국 미술비평과 생성형 인공지능 시대의 의미

초록

간이 최립(崔岦, 1539–1612)은 조선 중·후기의 대표적인 문인 지식인으로서 문장가, 외교가, 학자, 교육자로 주로 연구되어 왔다. 그러나 그의 문집인 『간이집(簡易集)』에는 서예와 회화, 화첩, 기념화, 고화(古畫), 서화의 보존과 해석 등에 관한 상당한 분량의 글이 수록되어 있다. 이러한 글들은 최립의 서화감평이 단순히 기술적 완성도나 외형적 아름다움을 평가하는 데 머물지 않았음을 보여준다. 그는 서예의 형식을 서자의 인격과 학문적 수양과 연결하고, 회화를 시·기억·삶의 경험과의 대화로 해석했으며, 단순한 형태적 유사성과 예술적 생명력을 구별하고, 작품의 전승과 작가 귀속을 신중하게 검토하였다. 또한 서화를 개인적·역사적·문화적 기억을 보존하는 매체로 이해하였다.

본 논문은 최립의 서화 관련 글을 종합하여 그의 서화감평에 나타나는 다섯 가지 특징을 재구성한다. 첫째, 서예를 통해 작가의 인격을 읽는 방법, 둘째, 회화를 이미지와 시의 대화로 이해하는 방법, 셋째, 형태적 유사성과 예술적 생명력을 구별하는 방법, 넷째, 비교와 의심을 통해 진위와 작가 귀속을 검토하는 방법, 다섯째, 서화를 기억과 역사적 경험을 담는 그릇으로 이해하는 방법이다. 이어서 이러한 감평 방식이 한국 미술사에서 갖는 다섯 가지 의의를 검토한다. 즉, 도덕적 수양과 시각적 판단의 결합, 문인 지식인의 감식가적 역할 강화, 시·서·화의 통합, 역사적 의식을 갖춘 고증적 감평, 그리고 서화를 문화적 기억으로 보존하는 전통의 강화이다.

마지막으로 본 논문은 최립의 서화감평 철학을 생성형 인공지능(Generative AI)의 창의성·저작권·진정성·인간 주체성에 관한 최근 논의와 연결한다. 최근 연구는 생성형 AI가 인간의 창작 생산성과 아이디어 탐색의 범위를 확대할 수 있는 반면, 평균적 참신성, 진정성, 창작 주체성 및 저자성에 새로운 문제를 제기한다고 지적한다. 따라서 최립의 서화감평 철학은 AI 시대의 인간 중심 미학을 위한 유용한 틀을 제공한다. AI 시대의 미술감상은 단순히 “이 이미지가 어떻게 보이는가?”만을 묻는 것이 아니라, “누가 만들었는가, 왜 만들었는가, 어떤 문화적 기억을 담고 있는가, 그리고 어떤 인간적 판단이 그것에 의미를 부여하는가?”를 함께 물어야 한다.

주제어: 최립, 『간이집』, 조선시대, 한국 미술사, 서예, 문인화, 서화감평, 진정성, 생성형 인공지능, 인간-AI 창의성

1. 문인학자 속에 숨겨진 미술감식가의 발견

한국 미술비평의 역사는 일반적으로 화가, 화론가, 화첩, 회화 이론서, 그리고 후대의 문인화가들을 중심으로 서술되어 왔다. 그러나 미술감평의 역사는 전문 화가의 저술에만 존재하는 것이 아니다. 그림과 글씨를 일상적인 지적 활동의 일부로 접했던 문인 학자들의 글에도 중요한 미술비평의 흔적이 남아 있다. 간이 최립은 바로 이러한 범주에 속한다.

최립은 전문 화가가 아니었지만, 『간이집』에는 서예 작품, 산수화, 묵죽화, 기념화, 화첩 및 고화 등에 대한 글이 상당수 남아 있다. 현전하는 『간이집』은 1631년에 처음 간행되었으며, 모두 9권으로 구성되어 있다. 현재 전하는 『한국문집총간』 계열의 판본 역시 이러한 초기 문헌 전통에 기반한다(Choi, 1996).

이러한 자료는 최립의 한국 문화사적 위치를 새롭게 검토하게 한다. 중요한 질문은 그를 현대적 의미에서 “전문 미술이론가”라고 부를 수 있는가가 아니다. 그러한 표현은 지나치게 확대된 해석일 수 있다. 보다 생산적인 질문은 다음과 같다.

그의 서화 관련 글을 함께 읽을 때, 어떠한 미술감평의 원칙이 드러나는가?

이 질문은 매우 중요하다. 조선시대 시각문화는 문학문화와 깊이 연결되어 있었기 때문이다. 문인 관료는 동시에 작가, 수장가, 후원자, 감상자, 서예가, 그리고 작품의 해석자가 될 수 있었다. 최근 연구 역시 조선시대 문인들이 회화의 제작·수집·후원에 중요한 역할을 담당했음을 보여주며, “문인화”라는 개념 자체도 처음부터 완성된 범주로 존재한 것이 아니라 역사적으로 형성되었음을 밝히고 있다(Park, 2018; Yungman, 2012).

따라서 최립의 글은 16세기와 17세기 초의 한국 지식인이 실제로 어떻게 서화를 바라보고 판단했는지를 보여주는 귀중한 자료이다. 특히 그의 글은 하나의 체계적인 화론으로 정리되어 있지 않다. 오히려 제발(題跋), 식(識), 서문, 시문 등에 분산되어 있다. 바로 이러한 특성 때문에 그의 글은 미술감평을 고정된 이론이라기보다 주의 깊은 읽기의 실천으로 보여준다.

2. 역사적·지적 배경

최립은 한국 역사에서 매우 격동적인 시기에 살았다. 1539년에 태어나 1612년에 세상을 떠난 그는 16세기 후반의 지적 변화와 임진왜란, 그리고 전쟁 이후의 정치·문화적 재건을 모두 경험하였다. 그는 뛰어난 문장가로 이름을 떨쳤으며, 국가의 공식 문서와 외교 문서를 작성해야 하는 직책에서도 활동하였다. 이러한 문학적·외교적 명성이 그의 서화 관련 활동을 상대적으로 가려 왔다.

그러나 최립이 활동했던 지적 환경에서는 문학, 서예, 회화, 역사학습, 도덕적 수양이 긴밀하게 연결되어 있었다. 동아시아 문인문화에서 시각예술은 독립된 영역으로 존재하는 경우가 드물었다. 그림에는 제화시와 발문이 붙었고, 시는 산수화를 해석했으며, 서예는 글쓴이의 마음과 인격을 표현하는 수단이 되었다. 화첩은 여러 문인의 인간관계와 학문적 교류를 보존하는 문화적 공간이 되기도 했다.

조선의 문인들은 사혁(謝赫)의 기운생동(氣韻生動)을 비롯하여 중국의 다양한 화론과 서예론을 계승하였다. 그러나 이러한 개념을 단순히 그대로 수입한 것은 아니었다. 조선의 문인들은 자신들의 사회적·윤리적·문학적 상황에 맞추어 이를 재해석하였다. 조선 문인화의 역사에 관한 연구 역시 “문인화”라는 개념과 용어가 시대에 따라 변화했음을 강조한다(Park, 2018).

최립의 서화 관련 글은 이러한 지적 환경 속에 위치한다. 특히 그는 정형화된 이론을 반복하기보다 실제 작품을 바라보면서 판단하고 질문한다. 바로 이 점에서 그의 글은 조선시대 서화감평의 실제 모습을 이해하는 데 중요한 자료가 된다.

3. 최립 서화감평의 다섯 가지 독특한 원리

3.1 서예를 통해 인격을 읽다

최립의 서화미학을 가장 분명하게 보여주는 글 가운데 하나가 「퇴계서소병지(退溪書小屛識)」, 즉 “퇴계 선생의 글씨를 쓴 작은 병풍에 대한 글”이다.

그는 퇴계 이황의 서예를 논하면서 서예를 “심화(心畫)”, 즉 마음의 그림이라고 표현한다. 글씨의 시각적 형태를 통해 그 글씨를 쓴 사람의 정신과 인격을 짐작할 수 있다는 의미이다. 그는 퇴계의 서체를 분석하면서 초서에서도 품격과 힘, 절제와 규율이 유지되고 있음을 강조한다(Choi, 1996, Vol. 3).

여기에서 최립의 첫 번째 서화감평 원칙이 드러난다.

서예의 형태는 그 글씨를 만든 사람의 수양된 마음을 보여주는 흔적이다.

물론 이것을 “아름다운 글씨는 언제나 훌륭한 인격에서 나온다”는 단순한 등식으로 이해해서는 안 된다. 최립은 글씨의 형식적 특성, 즉 안정성, 절제, 힘, 필선의 규율 등을 오랜 수양의 흔적으로 읽는다.

따라서 그의 서예비평은 단순한 필법 평가를 넘어선다. 선의 질감은 단순한 시각적 속성이 아니라 지적·윤리적 수양이 표현된 결과가 된다.

현대적인 표현으로 바꾸면 최립의 방식은 일종의 ‘체화된 미학적 해석(embodied aesthetic interpretation)’이라고 볼 수 있다. 붓은 신체의 움직임을 기록하지만, 그 움직임은 훈련, 습관, 의도, 기질과 수양에 의해 형성된다. 따라서 최립의 서예비평은 보이는 형식에서 보이지 않는 인격으로 이동한다.

4. 회화를 이미지와 시의 대화로 이해하다

두 번째 원리는 최립의 「이아계제시산수도지(李鵝溪題詩山水圖識)」, 즉 “이아계가 시를 써 넣은 산수화에 대한 글”에서 잘 드러난다.

여기에서 최립은 회화와 시의 관계를 매우 인상적인 표현으로 설명한다.

“그림은 소리 없는 시이고, 시는 소리가 있는 그림이다.”

(Choi, 1996, Vol. 3)

최립에게 회화와 시는 경쟁하는 예술이 아니다. 서로를 완성하는 관계이다. 산수화는 시가 언어로 전달하는 것을 시각적으로 표현하며, 시는 그림이 환기하지만 직접적으로 말할 수 없는 것을 언어로 표현한다.

이러한 관점은 한국 문인문화의 특징을 이해하는 데 매우 중요하다. 산수화는 단순히 산과 강의 모습을 재현하는 이미지가 아니다. 그것은 풍경, 기억, 감정, 시, 철학적 성찰이 만나는 지적 공간이다.

최립의 관점은 현대의 용어로 말하면 일종의 ‘다중양식적 미학(multimodal aesthetics)’이라고 할 수 있다. 그는 시각적 대상을 그것을 둘러싼 문학적·문화적 맥락과 분리하지 않는다.

이 관점은 디지털 예술의 중요한 문제와도 연결된다. AI가 생성한 이미지는 흔히 독립적인 시각적 결과물로 평가된다. 그러나 그 이미지 뒤에는 프롬프트, 대화 과정, 수정 과정, 학습 데이터, 문화적 참조, 인간의 의도가 존재한다. 최립의 관점에서 보면 이미지 자체만으로는 완전한 작품의 의미를 설명하기 어렵다.

5. 형태적 유사성과 예술적 생명력을 구별하다

최립의 글에는 단순히 대상을 닮게 그리는 것과 예술적으로 대상의 본질을 전달하는 것은 다르다는 인식도 나타난다.

그는 금계공이 소장한 산수화에 대한 글에서 산수의 즐거움이 인간의 경험 속에 깊이 자리하고 있음을 설명한다. 사람은 특정한 산수의 형상에 반복적으로 끌리며, 회화는 실제로 그 장소에 가서 경험하지 않고도 그 경험의 일부를 보존할 수 있다(Choi, 1996, Vol. 3).

여기에서 세 번째 원리가 도출된다.

좋은 예술은 대상을 단순히 재현하는 것이 아니라 경험을 재구성한다.

따라서 최립의 미술감평은 사실적인 재현만으로 작품의 성공 여부를 판단하지 않는다. 산수화가 성공하려면 관람자가 그 안에서 풍경과 관련된 경험, 분위기, 기억 또는 정신적 울림을 회복할 수 있어야 한다.

이것은 생성형 AI의 미술을 평가하는 데 특히 중요한 기준이다.

AI는 몇 초 만에 매우 사실적인 산수화를 만들 수 있다. 그러나 시각적으로 정확한 산의 이미지와 고향에 대한 기억, 유배의 경험, 특정 산천에 대한 문화적 의미를 전달하는 작품은 동일하지 않다.

최립의 관점은 AI 예술에 다음과 같은 질문을 던지게 한다.

“이 작품은 단지 그럴듯하게 보이는가, 아니면 의미 있는 인간 경험을 전달하는가?”

6. 고증과 의심을 통한 진위 판단

최립의 서화감평에서 가장 현대적인 특징 가운데 하나는 「이소윤소유고화지(李少尹所有古畫識)」, 즉 “이소윤이 소장한 고화에 대한 글”에서 발견된다.

전쟁으로 혼란스러운 상황에서 한 고화를 입수한 소장자는 그 그림이 유명한 조선 초기 화가 안견의 작품일 가능성을 의심하였다. 최립은 이를 무조건 받아들이지 않았다. 자신이 전문 화가는 아니라고 밝히면서도 안견과 관련된 진작을 본 경험을 토대로 몇몇 형식적 특징을 비교하였다. 동시에 그림의 훼손 상태, 전승 경위, 의복의 형태, 산수 표현, 시대적 양식과 연대 등을 검토하면서 작가 귀속에 대한 불확실성을 인정하였다(Choi, 1996, Vol. 3).

이는 ‘비판적 감식(critical connoisseurship)’의 중요한 사례이다.

최립은 작가 귀속을 단순한 직관에 맡기지 않고 다음과 같은 요소를 비교한다.

• 양식적 특징

• 역사적 연대

• 작품의 보존 상태

• 전승 경위

• 도상적 특징

• 시대적 관습

더 중요한 것은 그가 자신의 판단에 불확실성이 있음을 인정했다는 점이다.

즉,

전문가란 모든 것을 확신하는 사람이 아니라 증거가 부족한 곳에서 확신과 추측을 구별할 줄 아는 사람이다.

이 원리는 AI 기반 미술품 감정에도 직접 적용할 수 있다. 현대 컴퓨터비전 시스템은 붓질, 구성, 안료, 서명, 양식적 특징을 대규모로 비교할 수 있다. 그러나 알고리즘의 높은 확률값이 곧 역사적 진실을 의미하는 것은 아니다. 작품의 소장 경위, 재료, 복원 기록, 역사적 맥락 역시 중요하다.

최립의 태도는 따라서 초기 형태의 증거 기반 미술품 감정(evidence-based authentication) 모델로 해석할 수 있다.

7. 서화를 기억과 역사적 경험의 그릇으로 이해하다

다섯 번째 원리는 최립의 서화 관련 글 전반에서 나타난다. 작품은 인간관계와 기억을 보존한다.

그의 「하대이묵죽지(河大而墨竹識)」에서는 세상을 떠난 화가 하대이의 묵죽에 대한 평가가 단순한 예술적 찬탄에 그치지 않는다. 작가의 죽음으로 인해 그림은 한 인간과의 관계를 기억하게 하는 유물이 된다. 작품은 사라진 인간세계의 물질적 흔적이 된다(Choi, 1996, Vol. 3).

또한 최립은 기념화와 화첩에 관한 글에서 그림이 사람과 사건을 시각적으로 보존하는 역할을 강조한다. 시간이 흐른 뒤 사람들은 직접 만날 수 없지만, 그림을 통해 과거의 얼굴과 관계를 다시 마주할 수 있다.

따라서 다섯 번째 원리는 다음과 같이 정리할 수 있다.

예술의 가치는 인간의 기억을 시간 너머로 전달할 때 더욱 커진다.

이러한 관점에서 작품은 단순한 미적 상품이 아니다. 그것은 문화적 기억의 매체이다.

오늘날 디지털 시대에는 이러한 원리가 더욱 중요하다. 이미지는 무한히 생성될 수 있지만 동시에 무한히 쉽게 소비되고 버려질 수도 있다. 최립은 우리에게 문화적 가치란 단순히 이미지의 존재 여부가 아니라 후대 사람들이 무엇을 기억할 수 있게 해주는가에 의해 결정될 수 있음을 일깨운다.

8. 한국 미술사에서 최립의 서화감평이 갖는 다섯 가지 공헌

8.1 도덕적 수양과 시각적 판단의 결합

최립의 퇴계 서예 해석은 예술적 형식과 학문적·도덕적 수양을 연결하는 조선 문인문화의 특징을 잘 보여준다. 서예는 단순한 필기법이 아니라 수양된 인격을 드러내는 시각적 표현이었다.

이를 최립 개인이 창안한 독창적인 이론이라고 단정할 수는 없다. 오히려 그의 중요성은 이러한 원리가 실제 서화감평에서 어떻게 작동했는지를 보여주는 구체적인 사례를 제공한다는 데 있다.

8.2 문인 지식인의 감식가적 역할 강화

최립은 전문 화가가 아니었음에도 문학, 역사, 윤리, 기억, 시각적 관찰을 결합하여 그림을 평가하였다.

이는 조선의 서화문화에서 학자와 문인이 감상자·수장가·후원자·해석자로 중요한 역할을 했음을 보여준다. 조선 전기 문인화의 수집과 글쓰기 문화에 관한 연구 역시 문인들의 이러한 역할을 확인해 준다(Yungman, 2012).

8.3 시·서·화의 통합적 예술관 강화

“그림은 소리 없는 시”라는 최립의 표현은 동아시아 문인예술의 통합적 미학을 압축적으로 보여준다.

그림과 시, 서예는 서로 독립된 예술이 아니라 하나의 지적·정신적 세계를 구성하는 상호보완적 표현이었다.

따라서 최립의 글은 후대 한국 문인화의 지적 기반을 연구하는 데 중요한 자료가 된다.

8.4 미술감평에 역사적 의식을 도입

최립의 고화 감평은 작가 귀속을 위해 역사적 판단이 필요함을 보여준다. 작품이 어느 시대에 제작될 수 있는지, 특정 시대의 양식과 어떤 관계가 있는지, 물질적 증거가 제시된 귀속과 양립하는지를 물어야 한다.

이러한 역사적 의식은 후대 조선 회화론에서 발전한 다양한 화풍과 계보에 대한 구분과도 연결된다. 조선시대 문인들이 회화를 바라보는 관점이 시대에 따라 점차 세분화되었다는 연구 역시 이러한 발전을 보여준다(Jang, 2014).

8.5 서화를 문화적 기억으로 보존

최립의 제발과 기록은 예술작품에 대한 글 자체가 작품의 보존 과정에서 중요한 역할을 할 수 있음을 보여준다.

그림에 기록이 더해지면 작품은 시각적 이미지를 넘어 문학적·역사적·전기적 정보를 함께 담은 복합적인 문화유산이 된다.

이 원리는 오늘날에도 매우 중요하다. 디지털 복제는 이미지를 보존할 수 있지만 작품의 전승 과정과 맥락을 분리할 위험도 있다. 최립의 관점에서 볼 때 보존은 단순히 픽셀을 저장하는 것이 아니라 메타데이터, 전기, 소장 경위, 해석, 문화적 맥락을 함께 보존하는 것이어야 한다.

9. 최립과 AI 시대의 창의성 문제

생성형 AI의 등장으로 예술 창작의 의미가 크게 변화하고 있다. 텍스트-이미지 생성 시스템은 자연어 프롬프트만으로 상당히 정교한 이미지를 만들어 낸다.

수백만 개의 작품을 대상으로 한 최근 연구는 생성형 AI가 인간의 창작 생산성을 높이고 가능한 아이디어의 탐색 범위를 넓힐 수 있음을 보여준다. 동시에 창작자들이 유사한 AI 생성 결과에 수렴하면서 평균적인 참신성이 감소할 가능성도 제시한다(Zhou & Lee, 2024).

여기에는 중요한 역설이 존재한다.

AI는 과거 어떤 기술보다 더 많은 이미지를 만들 수 있지만, 이미지의 풍부함이 곧 문화적 의미의 풍부함을 의미하지는 않는다.

최립의 철학은 이 문제에 대한 중요한 답을 제공한다. 그는 생산과 의미를 구별한다.

기계는 이미지를 생산할 수 있다.

그러나 인간은 여전히 다음을 결정해야 한다.

• 왜 이 이미지를 만드는가?

• 왜 이 주제를 선택하는가?

• 왜 이 양식을 사용하는가?

• 어떤 경험을 표현하는가?

• 어떤 전통과 대화하는가?

• 무엇을 기억하게 하는가?

• 무엇을 보존할 가치가 있는가?

• 무엇을 거부해야 하는가?

최근 연구 역시 인간과 AI의 관계를 단순히 “AI 대 인간”의 경쟁구도로 보는 것보다 인간의 판단과 기계의 생성이 상호작용하는 협업적 모델로 전환할 필요성을 강조한다(Medeiros & Cropley, 2025).

따라서 최립의 서화감평은 AI 시대의 비판적 미적 판단을 위한 모델로 재구성될 수 있다.

10. 인간 감식에서 AI 증강 감식으로

AI는 미술감평을 네 가지 측면에서 크게 향상시킬 수 있다.

첫째, 대규모 시각적 비교가 가능하다. AI는 하나의 작품을 수천, 수만 개의 작품과 비교하는 데 인간보다 훨씬 빠르다.

둘째, 전승 및 소장 경위 연구를 지원할 수 있다. 인명, 연대, 소장처, 경매 기록, 비문, 디지털 문헌을 연결하여 작품의 계보를 추적할 수 있다.

셋째, 멀티모달 AI는 이미지와 텍스트를 동시에 분석할 수 있다. 그림, 제발, 시, 작가의 생애, 역사적 기록을 하나의 분석 체계 안에서 연결할 수 있다.

넷째, AI는 인간이 쉽게 발견하지 못하는 시각적 패턴을 찾아낼 수 있다.

그러나 최립의 철학은 AI가 판단의 보조자이지 최종 심판자가 되어서는 안 된다는 원칙을 제시한다.

최근 연구도 생성형 AI 시대에 인간의 주체성이 여전히 중요함을 강조한다. 인간의 판단과 AI의 능력을 상호보완적으로 결합할 때 창의성과 혁신을 강화할 수 있지만, AI의 결과를 어떻게 해석하고 어떤 의미를 부여할 것인지는 여전히 인간의 주체적 판단에 달려 있다(Baptista et al., 2025).

따라서 미래의 미술감식가는 순수한 인간도, 순수한 알고리즘도 아닐 가능성이 크다.

그는 ‘AI로 증강된 인간 감식가(AI-augmented human connoisseur)’가 될 것이다.

기계는 규모를 제공한다.

학자는 맥락을 제공한다.

알고리즘은 패턴을 발견한다.

역사가는 증거를 평가한다.

AI는 가능성을 생성한다.

인간은 그 가능성 가운데 어떤 것이 문화적 의미를 가질 것인지 결정한다.

11. 진정성: 생성형 AI 시대에 최립이 주는 가장 중요한 교훈

최립과 현대 AI 예술을 연결하는 가장 강력한 주제는 ‘진정성(authenticity)’이다.

AI 보조 시각예술에 관한 연구는 관람자들이 AI가 생성하거나 AI의 도움을 받아 만들어진 작품을 혁신적으로 평가할 수 있는 동시에, 인간이 직접 많은 노동을 투입한 작품에 비해 진정성이나 노동집약성을 낮게 평가할 가능성도 있음을 보여준다. 또한 인식된 진정성이 작품의 가치평가에 영향을 미친다는 연구 결과도 제시된다(Co-creating Art with Generative Artificial Intelligence, 2024).

이 문제는 최립의 미술감평과 놀랍도록 유사하다.

최립에게 작품의 가치는 표면적 형태만으로 결정되지 않는다. 작품 뒤에 있는 역사가 중요하다.

누가 만들었는가?

어떤 상황에서 만들었는가?

어떻게 전승되었는가?

누가 보존했는가?

그 작품은 창작자의 삶에 대해 무엇을 말하는가?

AI 예술에도 동일한 질문을 던질 필요가 있다.

미래의 미술관에서는 AI 작품에 다음과 같은 제작 정보가 표시될 수도 있다.

인간 창작자 → 개념 → 프롬프트 → AI 모델 → 생성 과정 → 선택 → 수정 → 최종 작품 → 전승 및 소장 정보

이러한 투명성은 AI 예술을 불투명한 결과물이 아니라 기록된 창작 과정으로 변화시킬 것이다.

바로 이 지점에서 최립의 철학은 단순한 역사적 관심사를 넘어선다.

그것은 AI 예술의 출처와 진정성을 위한 윤리적 모델이 될 수 있다.

12. AI 예술을 위한 ‘최립 인간중심 프레임워크’

앞서 재구성한 최립의 다섯 가지 원리를 토대로 AI 시대의 미술감평을 위한 새로운 프레임워크를 다음과 같이 제안할 수 있다.

인격(Character)

작품을 만든 인간적 주체는 누구이며, 어떤 가치와 의도가 창작을 형성했는가?

기운(Vitality)

작품은 기술적 정확성과 시각적 새로움을 넘어 어떤 생명력과 경험을 전달하는가?

맥락(Context)

작품은 어떤 역사적·문화적·문학적·예술적 전통과 대화하는가?

증거(Evidence)

작품의 출처, 저자성, 제작 과정, 독창성에 관한 주장을 독립적으로 검증할 수 있는가?

기억(Memory)

작품은 어떤 인간적 경험, 관계, 사건 또는 문화적 의미를 보존하는가?

이 프레임워크는 AI를 거부하지 않는다.

오히려 AI를 적극적으로 활용하면서도 계산 능력과 문화적 판단을 동일시하지 않는다.

최근 연구 역시 인간과 AI의 공동창작이 예술적 가능성을 확대할 수 있지만, AI가 창작 과정에 깊이 개입할수록 주체성과 저자성의 문제가 더욱 복잡해진다는 점을 강조한다(Bomba & De Angeli, 2025; Rafner et al., 2025).

따라서 최립적 관점에서 인간은 기술적 증강의 중심에 위치한다.

13. 결론: 붓에서 알고리즘으로

간이 최립의 서화 관련 글은 아름다운 작품에 대한 단순한 감상을 훨씬 넘어선다. 그의 감평은 문화 해석의 한 형태였다.

그는 서예에서 인격을 보았다.

그는 그림에서 시를 들었다.

그는 산수에서 경험을 회복했다.

그는 고화에서 작가 귀속을 의심하고 검증했다.

그는 보존된 이미지에서 사라진 인간을 기억했다.

이 다섯 가지는 하나의 통합된 미학으로 정리될 수 있다.

인격, 대화, 경험, 증거, 기억.

최립을 한국 미술비평의 유일한 창시자나 현대적 의미의 체계적인 미술이론가로 규정해서는 안 된다. 그러한 주장은 역사적으로 지나친 확대해석이 될 수 있다. 오히려 그의 중요성은 더 미묘하고 흥미롭다. 그의 여러 산문과 제발은 문학, 윤리, 역사, 우정, 기억, 시각적 관찰이 교차하는 지점에서 한 조선 지식인이 어떻게 서화를 감상하고 평가했는지를 보여준다.

이러한 통찰은 생성형 AI 시대에 더욱 강력한 의미를 갖는다.

AI는 이제 수천 장의 이미지를 인간 화가가 한 장을 스케치하는 데 걸리는 시간보다 훨씬 빠르게 만들어 낼 수 있다. AI는 화풍을 모방하고, 역사적 모습을 재구성하며, 산수화를 만들고, 서예의 형태를 시뮬레이션할 수 있다. 또한 학자들이 방대한 시각자료를 비교하고 분석하는 것을 도울 수 있다.

그러나 풍요로움은 새로운 문제를 만든다.

이미지가 거의 공짜가 될수록 의미는 희소해진다.

모방이 쉬워질수록 진정성은 확인하기 어려워진다.

생성이 자동화될수록 인간의 판단은 오히려 더욱 중요해진다.

최립의 철학은 이러한 문제에 대한 하나의 답을 제공한다. AI 시대의 예술은 단순히 이미지가 아름다운지, 사실적인지, 혁신적인지, 기술적으로 정교한지만으로 평가해서는 안 된다. 우리는 동시에 그 작품이 의미 있는 인간의 의도를 표현하는지, 그 출처가 신뢰할 만한지, 문화적 맥락이 이해되고 있는지, 창작 과정이 투명한지, 그리고 기억할 가치가 있는 무언가를 인류에게 남기는지를 물어야 한다.

최근 연구 역시 AI와 인간의 가장 생산적인 관계가 대체가 아니라 협업일 가능성을 보여준다. 생성형 AI는 생산성을 높이고 아이디어의 영역을 확대할 수 있지만, 인간의 발상, 선택, 해석, 주체성은 여전히 결정적인 역할을 한다(Zhou & Lee, 2024; Medeiros & Cropley, 2025).

따라서 최립의 철학은 21세기를 위해 다음과 같이 다시 표현할 수 있다.

알고리즘은 이미지를 생성할 수 있지만, 그 이미지가 무엇을 의미하는지를 결정하는 것은 여전히 인간이다.

붓은 알고리즘이 되었다.

작업실은 인간-AI 인터페이스가 되었다.

화첩은 디지털 아카이브가 되었다.

그러나 근본적인 질문은 변하지 않았다.

“이 작품은 그것을 만들고, 선택하고, 보존하고, 해석한 인간에 대해 무엇을 말해 주는가?”

그런 의미에서 최립의 서화감평 철학은 낡은 것이 아니다.

오히려 AI 시대에 새롭게 필요한 철학이 되었다.

참고문헌

Baptista, J., et al. (2025). Human agency and artificial intelligence: Implications for creativity, innovation, and work. Technological Forecasting and Social Change.

Bomba, F., & De Angeli, A. (2025). Agency and authorship in AI art: Transformational practices for epistemic troubles. International Journal of Human-Computer Studies, 205, 103652. https://doi.org/10.1016/j.ijhcs.2025.103652

Choi, R. [崔岦]. (1996). Gani-jip [簡易集; 간이집]. 민족문화추진회. (원전은 1631년에 간행됨.)

Jang, E.-Y. (2014). The literati’s point of view on painting in the late Joseon Dynasty. Kyujanggak, 44, 1–28. https://doi.org/10.22943/kyujg.2014..44.001

Medeiros, K., & Cropley, D. H. (2025). Human-AI co-creativity: Does ChatGPT make us more creative? The Journal of Creative Behavior. https://doi.org/10.1002/jocb.70022

Park, E. S. (2018). Yuhwa, literati painting in the Joseon period: The origin and meaning of the term. Korean Bulletin of Art History, 50, 245–266. https://doi.org/10.15819/rah.2018..50.245

Rafner, J., Zana, B., Hansen, I. B., Ceha, S., Sherson, J., Benedek, M., & Lebuda, I. (2025). Agency in human-AI collaboration for image generation and creative writing: Preliminary insights from think-aloud protocols. Creativity Research Journal. https://doi.org/10.1080/10400419.2025.2587803

Yungman, B. (2012). Collecting and writing about literati painting during the early Chosŏn period. Journal of East Asian Art History, 1, 181–220. https://doi.org/10.22978/aarth.2012.1..181

Zhou, E., & Lee, D. (2024). Generative artificial intelligence, human creativity, and art. PNAS Nexus, 3(3), pgae052. https://doi.org/10.1093/pnasnexus/pgae052

Zhou, E., Lee, D., & Gu, B. (2025). Who expands the human creative frontier with generative AI: Hive minds or masterminds? Science Advances, 11(36), eadu5800. https://doi.org/10.1126/sciadv.adu5800

Garcia, M. B. (2025). The paradox of artificial creativity: Challenges and opportunities of generative AI artistry. Creativity Research Journal, 37(4), 755–768. https://doi.org/10.1080/10400419.2024.2354622

Ukpaka, P. M. (2025). The creative agency of large language models: A philosophical inquiry. AI and Ethics, 5, 2455–2466. https://doi.org/10.1007/s43681-024-00557-9

Sobrinho, A. B. F. (2026). Exploring how generative AI reshapes artistic creativity. The Journal of Creative Behavior, 60(2), e70113. https://doi.org/10.1002/jocb.70113

2026년 8월 16일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026.

Choi, Y. B. (2026, August 16). Gani Choi Rip’s Philosophy of Calligraphy and Painting Connoisseurship: Korean Art Criticism and Its Significance in the Age of Generative AI. K-GSP.

Joseon’s Civil Service Examinations, Korea’s Passion for Education, and Generative AI /

From the Age of Examinations to the Age of Questions

Abstract

The civil service examination system of the Joseon Dynasty was more than a mechanism for selecting government officials. It was a central social institution connecting education, social mobility, family prestige, and the selection of national talent. This historical experience resonates in important ways with Korea’s enduring passion for education. This essay examines the Joseon examination system and contemporary Korean educational culture through the broader civilizational transformation brought about by generative artificial intelligence (AI). Particular attention is given to the fact that Joseon examinations were not merely tests of memorization; they also required the interpretation of classical texts, argumentation, writing, and policy-oriented reasoning. This historical perspective raises a fundamental question for contemporary education: What should examinations measure when generative AI can rapidly retrieve information, organize knowledge, and produce sophisticated written responses? The essay argues that education in the AI era should place greater emphasis on the quality of questions, verification of information and AI-generated outputs, critical thinking, human judgment, and the ability to discover and formulate new problems. Rather than suppressing Korea’s strong educational passion, the essay proposes redirecting it—from a competition over who knows more to a competition over who can ask better questions. It further envisions a new form of “civil service examination for the AI era,” in which students are allowed to use AI but are evaluated on their ability to identify errors, verify evidence, justify judgments, and generate questions that AI cannot readily provide. Ultimately, if the Joseon examination asked, “What do you know?”, education in the age of generative AI must ask a more challenging question: “If AI already knows all the answers, what will you ask?” This transformation could turn Korea’s longstanding educational passion into a powerful educational asset for creativity and human-centered learning in the AI era.

Keywords: Joseon civil service examinations, Korean educational passion, generative AI, AI in education, educational assessment, critical thinking, questioning, creativity, AI-era examination

An Age-Old Passion for Examinations: Joseon’s Civil Service Examinations

Imagine a Joseon scholar sitting at his desk from before dawn. The Four Books and Five Classics are open before him. He reads the passages again and again. Ink stains his brush, while his mind is filled with the language and arguments of the Confucian classics.

Why does he study so intensely?

It is not simply out of intellectual curiosity. Passing the civil service examination could open the door to government service, changing not only an individual’s future but also the social standing of an entire family.

This scene inevitably reminds us of Korea’s educational passion today.

The civil service examination system was first introduced during the reign of King Gwangjong of Goryeo in 958 and continued throughout the Joseon period until it was abolished during the Gabo Reforms of 1894. During Joseon, the examination system included the literary examination (mungwa), military examination (mugwa), and various technical examinations (japgwa). The saengwon and jinsa examinations also played important roles in the educational and social system (Academy of Korean Studies, 2024).

The examination system was not simply a single test. Joseon maintained government educational institutions such as the National Academy (Seonggyungwan), the Four Schools (Sahak), and local Confucian schools (hyanggyo), alongside private educational institutions such as village schools (seodang). Education and the examination system were closely interconnected (National Institute of Korean History, n.d.).

Education in Joseon therefore functioned as more than the acquisition of knowledge. It was an institutional pathway connecting social mobility, personal achievement, and the selection of talent for government service.

In this respect, Joseon’s examination culture has a remarkable historical resonance with Korea’s contemporary educational culture.

The Examination as Joseon’s Educational Pipeline

For Korean students today, elementary school, middle school, high school, and university entrance examinations can be understood as parts of a long educational pathway.

Joseon had its own educational ladder.

Students educated at the Four Schools and local Confucian schools could advance to higher levels of education. Seonggyungwan, the highest national educational institution, also served as an important center for students preparing for the literary civil service examination (National Institute of Korean History, n.d.).

Education at the Four Schools was likewise closely associated with preparation for the saengwon and jinsa examinations.

In contemporary language, education and competitive examinations formed a single ecosystem.

For this reason, Korea’s strong passion for education cannot be explained simply as a product of modern competition. The historical experience of connecting education, examinations, government service, and social status also deserves consideration.

Of course, contemporary Korean educational competition cannot be reduced to the legacy of the Joseon examination system. Industrialization, the expansion of higher education, labor-market competition, urbanization, and experiences of social mobility have all played major roles.

Yet the two eras share one powerful cultural belief:

Education can change the future of an individual and a family.

Was the Joseon Examination Really Just a Test of Memorization?

Here, another misconception about the Joseon examinations needs to be corrected.

The examinations should not be understood simply as tests of memorization.

For example, the saengwon examination included questions concerning the meaning and interpretation of the Four Books and Five Classics, while the jinsa examination evaluated literary ability. The literary civil service examination also required knowledge of the classics, interpretation, writing, and policy-oriented thinking (Academy of Korean Studies, 2024).

In other words, Joseon examinations required not only memory, but also interpretation, argumentation, writing, and policy reasoning.

This point is particularly important in the age of AI.

Generative AI can already organize information and produce polished prose far more quickly than humans.

What, then, should education evaluate in the age of AI?

This question connects directly with the historical lesson of Joseon’s examination culture.

What If ChatGPT Entered the Joseon Examination Hall?

Let us imagine something extraordinary.

One day, generative AI enters the Joseon examination hall.

The examiner asks:

“Discuss how wise and capable people should be selected to govern the state.”

Before picking up his brush, a scholar asks AI for help.

Within seconds, AI produces an elegant response. It contains quotations from the classics, persuasive reasoning, and historical examples.

Is the person who submits this answer a truly talented scholar?

This hypothetical question is remarkably similar to a problem facing universities today.

Generative AI can significantly assist students with writing and information organization. At the same time, it raises fundamental questions about whether traditional assignments and examinations accurately measure what students actually know and can do. Recent research on generative AI and higher-education assessment has therefore called for the redesign of conventional assessment practices (Xia et al., 2024; Weng et al., 2024).

Xia et al. (2024) found that generative AI is transforming assessment in higher education and argued that assessment practices need to evolve to measure learning, responsibility, and academic integrity more appropriately.

Weng et al. (2024) likewise examined the limitations of conventional assessment in generative-AI environments and emphasized the need to rethink both learning outcomes and assessment practices.

The fundamental question, therefore, is not simply:

“Should AI be banned from examinations?”

The deeper question is:

“What should an examination measure?”

The Machine That Produces Answers and the Human Who Asks Questions

Before generative AI, students needed to demonstrate their ability to produce answers.

After generative AI, some of that ability can be automated by machines.

The center of education therefore needs to shift.

From finding information to asking questions.

From producing answers to verifying answers.

From remembering correct answers to exercising judgment.

From reproducing model answers to discovering new problems.

Hao et al. (2024) analyzed the transformative effects of large language models and generative AI on assessment. They noted that these technologies create new opportunities while also raising important concerns about validity, reliability, transparency, and fairness.

Kaldaras et al. (2024) similarly argued that developing valid assessments in the age of generative AI requires educators to reconsider the purposes of assessment and the abilities they actually want to measure.

The examination of the AI era, therefore, should not be a competition against AI.

It should be an examination that uses AI while measuring distinctly human judgment.

Korea’s Educational Passion Does Not Need to Disappear

Should Korea’s intense educational competition be restrained in the age of AI?

Not necessarily.

In fact, Korea’s passion for education could become an important national asset in the AI era.

The problem is not the intensity of educational ambition. It is the direction in which that ambition is directed.

Education should move:

From memorizing more to understanding more deeply.

From solving problems faster to creating better problems.

From pursuing higher scores to making better judgments.

From submitting AI-generated answers to verifying, challenging, and improving AI-generated answers.

This direction is consistent with emerging research on assessment. Bearman and Ajjawi (2024), for example, discuss the growing importance of students’ ability to critically evaluate AI-generated outputs—a capacity closely related to evaluative judgment.

If Korea’s educational passion moves in this direction, a culture historically centered on competition can evolve into a culture centered on creativity.

Imagining a New Examination for the AI Era

What, then, might a new civil service examination look like in the age of AI?

There is no need to ban AI from the examination hall.

We might actually invite AI into the examination hall.

Students could be asked to use AI to solve a problem and then identify errors in the AI-generated response. They could verify sources, compare competing answers, and finally formulate a new question that AI itself failed to raise.

For example:

“Ask a generative AI system to explain the Joseon civil service examination system. Identify three historical errors or oversimplifications in its response. Correct them using reliable historical sources and scholarship. Finally, formulate one new research question that the AI failed to identify.”

AI would not be prohibited.

Students would be encouraged to use it.

But their grades would not depend on how beautifully AI writes.

They would be evaluated on:

The quality of their questions.

Their ability to evaluate the reliability of sources.

Their ability to identify AI errors.

Their ability to justify their judgments with evidence.

Their ability to formulate new questions.

This could become the new civil service examination of the generative-AI era.

Learning from the Joseon Scholar

Here we can discover an unexpected lesson from the Joseon scholar.

The scholar read books.

But reading was not the end of learning.

He questioned, interpreted, argued, and expressed his understanding in his own words.

Today, AI can read and summarize enormous quantities of information.

What, then, should humans do?

Read more deeply.

The more AI provides answers, the better our questions must become.

The more AI writes, the more carefully humans must examine its evidence.

The more choices AI provides, the more important human judgment becomes.

And no matter how many answers AI can generate, humans must continue to formulate questions that no machine has thought to ask.

This may be one of the most meaningful ways to carry the intellectual spirit of the Joseon scholar into the age of AI.

From the Joseon Examination Hall to the AI Examination Hall

The Joseon civil service examination was more than a test. It was a national mechanism for evaluating what a person knew, how well that person could interpret knowledge, and how persuasively he could express his ideas. At the same time, it was a vast cultural system that connected education with personal achievement, family honor, and social mobility.

Modern Korea’s educational passion has a similar character. The belief that good education, high test scores, and admission to a prestigious university can lead to a better future remains powerful.

But generative AI is fundamentally challenging this traditional educational formula.

AI can now find, organize, summarize, and generate information far faster than humans.

If so, examinations that ask only “How much do you know?” may no longer adequately measure human capability.

A new educational question therefore emerges:

If AI can generate the answer, what should we test humans for?

The first answer is the ability to ask questions.

In the past, people with more knowledge often had an advantage. Now that AI can access enormous amounts of knowledge, a more important human ability is to determine what knowledge is needed and formulate a precise question.

A good question comes before a good answer.

The outstanding student of the AI era, therefore, is not simply the student who finds the correct answer quickly.

It is the student who can ask AI a question that no one had previously thought to ask.

The second answer is the ability to verify.

AI can produce answers that sound convincing but are incorrect. The AI-era intellectual must therefore be able to check sources, compare evidence, identify logical errors, and challenge the machine’s conclusions.

If the Joseon scholar read the classics and debated their meaning, the AI-era learner must read and verify the knowledge produced by AI.

If AI creates the first draft, humans must make the final judgment.

The third answer is the ability to create.

Korea’s educational passion can become a strength rather than a weakness. But the goal of competition must change.

Instead of competing over who has memorized the most, students should compete over who can formulate the best questions.

Instead of competing over who can write an answer fastest, they should compete over who can discover the most important problem.

Instead of competing over who can complete an assignment without AI assistance, they should compete over who can use AI most responsibly and creatively to create new value.

This is the evolution of Korea’s educational passion for the age of AI.

The Final Examination of Education

Seen from this perspective, the Joseon examination is not merely an obsolete institution.

It leaves us with fundamental questions:

What kind of person does a society call talented?

What abilities should education cultivate?

What should examinations measure?

And what is the ultimate purpose of education?

Joseon selected people who could read, interpret, and express the meaning of the classics.

Industrial society needed people with specialized knowledge.

The AI era demands a new kind of human being:

Someone who knows how to use AI but does not surrender judgment to AI.

Someone who can obtain information while questioning its reliability.

Someone who can find an answer while also discovering the problem beyond the answer.

Someone who can use machines without surrendering human imagination and values.

This may be the new scholar of the AI era.

In the Joseon examination hall, the most important tool was the brush.

Today, the keyboard has taken its place, and now generative AI sits beside it.

Yet changing the tool does not eliminate the fundamental abilities that make us human.

Indeed, the more powerful AI becomes, the clearer those human abilities become:

The ability to question.

The ability to doubt.

The ability to verify.

The ability to judge.

And above all, the ability to formulate questions that no one else has thought to ask.

Korea’s educational passion, therefore, does not need to disappear.

It needs to change direction.

From education for memorization to education for deeper thinking.

From competition for higher scores to competition for better questions.

From trying to write better than AI to learning to think more creatively with AI.

At that moment, the Joseon examination becomes more than a historical artifact. It becomes a mirror for the future of education.

The Joseon examination asked:

“What do you know?”

The new examination of the AI era must ask something much more difficult:

“If AI already knows all the answers, what will you ask?”

The person who can answer that question may be the true talent of the AI era.

And perhaps this is the most important civil service examination facing Korean education today.

The age of examinations is not over. Only the questions on the examination are changing.

References

Academy of Korean Studies. (2024). Gwageo [Civil service examination]. Encyclopedia of Korean Culture.

Bearman, M., & Ajjawi, R. (2024). Learning to become an AI assessor: The implications of generative artificial intelligence for assessment. Assessment & Evaluation in Higher Education. https://doi.org/10.1080/02602938.2024.2335321

Hao, J., von Davier, A. A., Yaneva, V., Lottridge, S., von Davier, M., & Harris, D. J. (2024). Transforming assessment: The impacts and implications of large language models and generative AI. Educational Measurement: Issues and Practice, 43(2), 16–29. https://doi.org/10.1111/emip.12602

Kaldaras, L., Akaeze, H. O., & Reckase, M. D. (2024). Developing valid assessments in the era of generative artificial intelligence. Frontiers in Education, 9, 1399377. https://doi.org/10.3389/feduc.2024.1399377

National Institute of Korean History. (n.d.). Education and the civil service examination system in early Joseon. Our History Net.

Weng, X., Xia, Q., Gu, M., Rajaram, K., & Chiu, T. K. F. (2024). Assessment and learning outcomes for generative AI in higher education: A scoping review on current research status and trends. Australasian Journal of Educational Technology, 40(6), 37–55. https://doi.org/10.14742/ajet.9540

Xia, Q., Weng, X., Ouyang, F., Lin, T. J., & Chiu, T. K. F. (2024). A scoping review on how generative artificial intelligence transforms assessment in higher education. International Journal of Educational Technology in Higher Education, 21, 40. https://doi.org/10.1186/s41239-024-00468-z

조선의 과거, 한국의 교육열, 그리고 생성 AI

시험의 시대에서 질문의 시대로

시험을 향한 오래된 열망: 조선의 과거

조선의 어느 선비가 새벽부터 책상 앞에 앉아 있다. 사서삼경을 펼쳐 놓고 문장을 읽고 또 읽는다. 붓끝에는 먹물이 묻어 있고, 머릿속에는 수많은 경전의 문장과 논리가 들어 있다. 그가 공부하는 이유는 단순한 학문적 호기심 때문만은 아니다. 과거에 합격하면 관료로 진출할 수 있고, 그것은 개인의 미래뿐 아니라 가문의 사회적 위상과도 연결되었다.

이 장면은 오늘날 한국의 교육열을 생각하게 한다.

조선의 과거제도는 고려 광종 9년(958)에 처음 실시되었고 조선시대에도 계승되다가 1894년 갑오개혁으로 폐지되었다. 조선의 과거는 문과·무과·잡과로 구성되었으며, 문과와 함께 생원시·진사시가 중요한 시험으로 기능했다(한국학중앙연구원, 2024).

과거는 단순한 시험 하나가 아니었다. 조선은 성균관·사부학당·향교 등의 관학과 서당 등의 사학을 운영하면서 교육과 과거제도를 긴밀하게 연결하였다(국사편찬위원회, n.d.).

따라서 조선에서 교육은 개인적인 지식 습득을 넘어 사회적 이동과 국가 인재 선발을 연결하는 제도적 통로였다.

이 점에서 조선의 과거는 오늘날 한국의 입시문화와 놀라운 역사적 공명을 갖는다.

과거는 조선의 입시 시스템이었다

오늘날 한국 학생에게 초등학교, 중학교, 고등학교, 대학입시는 하나의 긴 교육 경로로 인식된다.

조선에도 이와 유사한 교육의 계단이 있었다.

사부학당과 향교에서 교육받은 학생들이 더 높은 수준의 교육으로 진입하고, 성균관은 국가의 최고 교육기관이자 문과시험을 준비하는 중요한 교육기관이었다(국사편찬위원회, n.d.).

사부학당의 교육도 생원·진사시 준비와 밀접하게 관련되어 있었다. 국사편찬위원회의 자료에 따르면 사부학당에서는 일정한 자격을 갖춘 학생들이 교육을 받고 생원·진사시를 준비할 수 있었다.

오늘날의 표현으로 바꾸면 교육기관과 선발시험이 하나의 생태계를 형성하고 있었던 셈이다.

그러므로 한국인의 강한 교육열을 단순히 현대 한국 사회의 경쟁심으로만 설명하는 것은 충분하지 않다. 조선시대부터 교육과 시험, 관직, 사회적 지위가 긴밀하게 연결되어 온 역사적 경험 역시 중요한 배경으로 살펴볼 필요가 있다.

물론 현대 한국의 교육열을 조선의 과거제 하나만으로 설명할 수는 없다. 산업화, 대학 확대, 노동시장 경쟁, 도시화, 사회이동의 경험 등 수많은 현대적 요인이 함께 작용한다.

그럼에도 두 시대에는 하나의 강력한 공통점이 있다.

교육이 개인과 가족의 미래를 바꿀 수 있다는 믿음이다.

조선의 과거는 정말 암기시험이었을까

여기에서 조선의 과거에 대한 또 하나의 오해를 바로잡을 필요가 있다.

과거를 단순한 암기시험으로만 이해해서는 안 된다.

생원시에서는 사서의와 오경의를 시험했고, 진사시에서는 문예적 능력을 평가했다. 문과에서는 경전 이해와 정책적 식견, 문장 작성 능력 등이 중요한 평가 대상이었다(한국학중앙연구원, 2024).

즉 조선의 과거에는 기억(memory)만이 아니라 해석(interpretation), 논변(argumentation), 문장력(writing), 정책적 사고가 요구되었다.

이 점은 AI 시대에 특별히 중요하다.

왜냐하면 생성 AI가 이미 인간보다 훨씬 빠르게 정보를 정리하고 문장을 작성할 수 있기 때문이다.

그렇다면 AI 시대의 교육은 무엇을 평가해야 하는가?

이 질문은 조선의 과거가 우리에게 던지는 역사적 질문과 연결된다.

ChatGPT가 과거시험장에 들어온다면

상상해 보자.

어느 날 조선의 과거시험장에 생성 AI가 등장한다.

시험관이 묻는다.

“어진 인재를 선발하여 나라를 다스리는 방법을 논하라.”

한 선비가 붓을 들기 전에 AI에게 질문한다.

AI는 몇 초 만에 유려한 문장을 만들어낸다. 고전의 인용도 풍부하고 논리도 그럴듯하다.

그 답안을 제출한 사람은 훌륭한 인재인가?

이 질문은 오늘날 대학이 직면한 문제와 정확히 연결된다.

생성 AI는 학생의 글쓰기와 정보 정리 능력을 크게 보조할 수 있지만, 동시에 기존의 과제와 시험이 실제로 학생의 능력을 얼마나 정확하게 측정하는지에 대한 의문을 제기한다. 생성 AI가 고등교육 평가에 미치는 영향을 검토한 연구들은 전통적인 평가방식을 재설계할 필요성을 제시하고 있다(Xia et al., 2024; Weng et al., 2024).

Xia et al. (2024)은 생성 AI가 고등교육의 평가를 변화시키고 있으며, 평가가 학생의 학습과 책임, 학업적 진실성을 더욱 적절하게 측정할 수 있도록 변화할 필요가 있다고 분석하였다.

Weng et al. (2024) 역시 생성 AI 환경에서 기존 평가방식의 한계를 검토하면서 학습성과와 평가방식을 새롭게 설계할 필요성을 제시하였다.

따라서 핵심 문제는 단순히 “AI를 시험에서 금지할 것인가?”가 아니다.

더 근본적인 질문은 이것이다.

“시험은 무엇을 측정해야 하는가?”

답을 만드는 기계와 질문하는 인간

생성 AI 이전에는 학생에게 답을 쓰는 능력이 중요했다.

그러나 생성 AI 이후에는 답을 만드는 능력의 일부가 기계에 의해 자동화될 수 있다.

따라서 교육의 중심축은 이동해야 한다.

정보를 찾는 능력에서 질문하는 능력으로.

답을 작성하는 능력에서 답을 검증하는 능력으로.

정답을 기억하는 능력에서 판단하는 능력으로.

모범답안을 재현하는 능력에서 새로운 문제를 발견하는 능력으로.

Hao et al. (2024)은 대규모 언어모델과 생성 AI가 평가의 효율성과 새로운 가능성을 제공하는 동시에 타당도, 신뢰도, 투명성, 공정성 등의 중요한 문제를 제기한다고 분석하였다.

Kaldaras et al. (2024) 역시 생성 AI 시대에 타당한 평가를 설계하기 위해 평가의 목적과 측정하고자 하는 능력을 근본적으로 재검토해야 한다고 지적한다.

따라서 AI 시대의 시험은 AI와 경쟁하는 시험이 되어서는 안 된다.

AI를 활용하면서도 인간의 판단력을 측정하는 시험이어야 한다.

한국의 교육열은 사라질 필요가 없다

그렇다면 한국의 교육열을 AI 시대에 억제해야 하는가?

그럴 필요는 없다.

오히려 한국의 교육열은 AI 시대의 중요한 국가적 자산이 될 수 있다.

문제는 교육열의 크기가 아니라 교육열의 방향이다.

더 많이 외우는 교육에서 더 깊이 이해하는 교육으로.

더 빨리 문제를 푸는 교육에서 더 좋은 문제를 만드는 교육으로.

더 높은 점수를 얻는 교육에서 더 정확하게 판단하는 교육으로.

AI가 만들어준 답을 제출하는 교육에서 AI의 답을 검증하고 개선하는 교육으로.

이러한 변화는 교육평가 연구에서 강조되는 방향과도 일치한다. Bearman과 Ajjawi는 생성 AI 시대에 학습자가 AI가 만들어낸 결과물을 비판적으로 평가할 수 있는 능력, 즉 평가적 판단의 중요성이 더욱 커진다고 논의하였다(Bearman & Ajjawi, 2024).

한국의 교육열이 이러한 방향으로 이동한다면, 과거의 경쟁문화는 미래의 창조문화로 전환될 수 있다.

새로운 과거시험을 상상하다

그렇다면 AI 시대의 새로운 과거시험은 어떤 모습이어야 할까?

시험장에서 AI를 무조건 쫓아낼 필요는 없다.

오히려 AI를 시험장 안으로 데려올 수도 있다.

학생에게 AI를 사용하여 문제를 해결하게 하고, 그 과정에서 AI가 제시한 오류를 찾아내게 할 수 있다. 출처를 검증하게 하고, 서로 다른 답변을 비교하게 하며, 마지막에는 AI가 만들지 못한 새로운 질문을 스스로 제시하게 할 수 있다.

예를 들어 다음과 같은 시험문제를 생각해 볼 수 있다.

“생성 AI에게 조선의 과거제도에 관한 설명을 작성하게 하라. 그 답변에서 세 가지 역사적 오류 또는 지나친 일반화를 찾아라. 신뢰할 수 있는 역사자료와 연구를 이용하여 수정하라. 그리고 AI가 제시하지 못한 새로운 연구 질문을 하나 만들어라.”

이 시험에서는 AI 사용을 금지하지 않는다.

오히려 AI를 적극적으로 사용하게 한다.

그러나 학생의 점수는 AI가 얼마나 아름다운 문장을 만들어냈는지가 아니라 다음과 같은 능력에 따라 결정된다.

질문의 질.

자료의 신뢰성을 판단하는 능력.

AI의 오류를 발견하는 능력.

근거를 가지고 자신의 판단을 설명하는 능력.

새로운 질문을 만들어내는 능력.

이것이야말로 생성 AI 시대의 새로운 과거시험이 될 수 있다.

조선의 선비에게서 AI 시대의 인간을 배우다

여기에서 조선의 선비에게서 의외의 지혜를 발견할 수 있다.

선비는 책을 읽었다.

그러나 단순히 읽는 데 그치지 않았다.

질문하고, 해석하고, 논변하고, 자신의 문장으로 표현했다.

오늘날 AI는 엄청난 양의 정보를 읽고 요약할 수 있다.

그렇다면 인간은 무엇을 해야 하는가?

더 깊이 읽어야 한다.

AI가 답을 만들어줄수록 인간은 더 좋은 질문을 해야 한다.

AI가 글을 써줄수록 인간은 그 글의 근거를 검증해야 한다.

AI가 수많은 선택지를 제시할수록 인간은 어떤 선택이 옳은지 판단해야 한다.

그리고 AI가 아무리 많은 답을 제공하더라도 인간만이 발견할 수 있는 새로운 질문을 만들어야 한다.

이것이 조선의 학문정신을 AI 시대에 계승하는 한 방법일 것이다.

과거시험장에서 AI 시험장으로

조선의 과거시험은 단순한 시험이 아니었다. 그것은 한 사람이 무엇을 알고 있는지, 무엇을 해석할 수 있는지, 그리고 자신의 생각을 얼마나 설득력 있게 표현할 수 있는지를 평가하는 국가적 장치였다. 동시에 교육을 개인의 성공, 가문의 명예, 사회적 이동과 연결하는 거대한 문화 시스템이었다.

오늘날 한국의 교육열도 이와 닮아 있다. 좋은 교육을 받고, 좋은 시험 성적을 얻고, 좋은 대학에 진학하는 것이 더 나은 미래로 이어진다는 믿음은 여전히 강력하다.

그러나 생성 AI의 등장은 이 오래된 교육공식을 근본적으로 흔들고 있다.

AI는 이제 인간보다 훨씬 빠르게 정보를 찾고, 정리하고, 요약하고, 문장을 만들 수 있다.

그렇다면 “누가 더 많이 알고 있는가?”를 묻는 시험만으로는 인간의 능력을 충분히 평가하기 어렵다.

여기에서 교육의 새로운 질문이 시작된다.

AI가 답을 만들어준다면, 인간에게 무엇을 시험할 것인가?

첫 번째 답은 질문하는 능력이다.

과거에는 많이 아는 사람이 유리했다. 이제는 AI가 방대한 지식에 접근할 수 있기 때문에, 인간에게 더욱 중요한 능력은 어떤 지식이 필요한지를 판단하고 정확한 질문을 만드는 능력이 된다.

좋은 질문은 좋은 답보다 먼저 온다.

따라서 AI 시대의 뛰어난 학생은 단순히 정답을 빨리 찾는 학생이 아니다.

AI에게 이전에는 생각하지 못했던 질문을 던질 수 있는 학생이다.

두 번째 답은 검증하는 능력이다.

AI는 매우 그럴듯하지만 틀린 답을 만들 수도 있다. 따라서 AI 시대의 지식인은 AI의 답을 그대로 받아들이는 사람이 아니라 자료를 확인하고 출처를 비교하며 논리적 오류를 발견할 수 있는 사람이다.

조선의 선비가 경전을 읽고 그 뜻을 논변했다면, AI 시대의 학습자는 AI가 만들어낸 지식을 다시 읽고 검증해야 한다.

AI가 첫 번째 초안을 만든다면, 인간은 마지막 판단을 내려야 한다.

세 번째 답은 창조하는 능력이다.

한국의 교육열은 약점이 아니라 강점이 될 수 있다. 다만 경쟁의 목표를 바꾸어야 한다.

누가 더 많이 외웠는가를 경쟁하는 대신 누가 더 좋은 질문을 만드는가를 경쟁해야 한다.

누가 더 빨리 답을 쓰는가를 경쟁하는 대신 누가 더 중요한 문제를 발견하는가를 경쟁해야 한다.

누가 AI의 도움 없이 과제를 완성하는가를 경쟁하는 대신 누가 AI를 가장 책임 있고 창의적으로 활용하여 새로운 가치를 만들어내는가를 경쟁해야 한다.

이것이 한국 교육열의 AI 시대적 진화다.

교육의 마지막 시험

이렇게 보면 조선의 과거는 낡은 제도가 아니다.

오히려 우리에게 중요한 질문을 남긴다.

한 사회는 어떤 사람을 인재라고 부르는가?

어떤 능력을 교육해야 하는가?

시험은 무엇을 측정해야 하는가?

그리고 교육의 궁극적인 목적은 무엇인가?

조선은 경전을 읽고 해석하고 글로 표현하는 사람을 선발했다.

산업사회는 전문지식을 가진 사람을 필요로 했다.

AI 시대는 새로운 인간상을 요구한다.

AI를 사용할 줄 알면서도 AI에게 자신의 판단을 맡기지 않는 사람.

정보를 얻으면서도 정보의 진위를 검증하는 사람.

정답을 찾으면서도 정답 너머의 문제를 발견하는 사람.

기계를 활용하면서도 인간만의 상상력과 가치판단을 포기하지 않는 사람.

바로 이러한 인간이 AI 시대의 새로운 선비일 것이다.

조선의 과거시험장에서 가장 중요한 도구는 붓이었다.

오늘날에는 키보드가 그 자리를 차지했고, 이제는 생성 AI가 그 옆에 앉아 있다.

그러나 도구가 바뀌었다고 해서 인간에게 필요한 가장 근본적인 능력까지 사라지는 것은 아니다.

오히려 AI가 강력해질수록 인간에게 필요한 능력은 더욱 선명해진다.

질문하는 능력.

의심하는 능력.

검증하는 능력.

판단하는 능력.

그리고 아무도 묻지 않았던 질문을 만들어내는 능력.

따라서 한국의 교육열은 사라질 필요가 없다.

이제 방향을 바꾸면 된다.

더 많이 외우기 위한 교육에서 더 깊이 생각하기 위한 교육으로.

더 높은 점수를 얻기 위한 경쟁에서 더 좋은 질문을 만들기 위한 경쟁으로.

AI보다 잘 쓰기 위한 교육에서 AI와 함께 더 창의적으로 생각하기 위한 교육으로.

그때 조선의 과거는 단순한 역사적 유물이 아니라 미래교육의 거울이 된다.

조선의 과거시험이 물었던 질문은 “무엇을 알고 있는가?”였다.

AI 시대의 새로운 과거시험은 더 어려운 질문을 던져야 한다.

“AI가 모든 답을 알고 있다면, 당신은 무엇을 물을 것인가?”

이 질문에 답할 수 있는 사람이야말로 AI 시대의 진정한 인재다.

그리고 어쩌면 이것이 오늘날 한국 교육이 마주한 가장 중요한 과거시험일 것이다.

시험의 시대는 끝나지 않았다. 시험의 질문이 바뀌고 있을 뿐이다.

참고문헌

Bearman, M., & Ajjawi, R. (2024). Learning to become an AI assessor: The implications of generative artificial intelligence for assessment. Assessment & Evaluation in Higher Education. https://doi.org/10.1080/02602938.2024.2335321

Hao, J., von Davier, A. A., Yaneva, V., Lottridge, S., von Davier, M., & Harris, D. J. (2024). Transforming assessment: The impacts and implications of large language models and generative AI. Educational Measurement: Issues and Practice, 43(2), 16–29. https://doi.org/10.1111/emip.12602

Kaldaras, L., Akaeze, H. O., & Reckase, M. D. (2024). Developing valid assessments in the era of generative artificial intelligence. Frontiers in Education, 9, 1399377. https://doi.org/10.3389/feduc.2024.1399377

Weng, X., Xia, Q., Gu, M., Rajaram, K., & Chiu, T. K. F. (2024). Assessment and learning outcomes for generative AI in higher education: A scoping review on current research status and trends. Australasian Journal of Educational Technology, 40(6), 37–55. https://doi.org/10.14742/ajet.9540

Xia, Q., Weng, X., Ouyang, F., Lin, T. J., & Chiu, T. K. F. (2024). A scoping review on how generative artificial intelligence transforms assessment in higher education. International Journal of Educational Technology in Higher Education, 21, 40. https://doi.org/10.1186/s41239-024-00468-z

국사편찬위원회. (n.d.). 신편 한국사: 조선 초기의 정치구조—교육제도와 과거제도. 우리역사넷.

한국학중앙연구원. (2024). 과거. 한국민족문화대백과사전.

2026년 8월 16일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026.

Choi, Y. B. (2026, August 16). Joseon’s Civil Service Examinations, Korea’s Passion for Education, and Generative AI: From the Age of Examinations to the Age of Questions. K-GSP.

Reading the Logic of Change Through the Brush

Choi Rip’s Juyeok Bonui Gugyeol Buseol (주역본의구결부설 (周易本義口訣附說)) and the Wisdom of Interpretation in the Age of Generative AI

When we think of Gani (簡易) Choi Rip (崔岦, 1539–1612), we often first remember him as one of the great literary figures of the Joseon Renaissance. His prose was powerful, his classical Chinese writing was sophisticated, and his diplomatic writings demonstrated the intellectual and literary capabilities of Joseon Korea. Yet to remember Choi Rip merely as an accomplished writer would be to overlook a significant part of his intellectual world.

Choi Rip also possessed a profound interest in the I Ching (Book of Changes), and this interest produced an important scholarly work, Juyeok Bonui Gugyeol Buseol (주역본의구결부설 (周易本義口訣附說)). The work reveals Choi Rip not simply as a literary master but also as an independent interpreter of classical texts and a scholar of I Ching learning.

According to the Kyujanggak Institute for Korean Studies at Seoul National University, the work applies Korean gugyeol reading marks to Zhu Xi’s Zhouyi Benyi (The Original Meaning of the Book of Changes) and incorporates Choi Rip’s own explanations while consulting Cheng Yi’s Yizhuan (Commentary on the Book of Changes). Several seventeenth-century woodblock editions are preserved in Kyujanggak, and the accompanying preface records Choi Rip’s participation in the Joseon government’s effort to revise and establish a standardized I Ching text (Kyujanggak Institute for Korean Studies, n.d.).

The book therefore allows us to see another Choi Rip.

He was not merely a man who wrote beautiful sentences.

He was a scholar who asked a deeper question:

How should a difficult text be read?

Returning to the Original Text

The I Ching is a book of change.

Heaven and earth change. Yin and yang change. Human circumstances change. Individuals, societies, and nations constantly move from one condition to another.

For this reason, the deepest question of the I Ching is not simply, “What will happen in the future?” It can also be understood as the more fundamental question:

How should human beings understand and respond to change?

Choi Rip approached the classical text with this intellectual seriousness. Rather than merely repeating established commentaries, he worked through Zhu Xi’s Zhouyi Benyi, added gugyeol readings, consulted Cheng Yi’s interpretations, and developed his own explanations (Kyujanggak Institute for Korean Studies, n.d.).

This point is crucial.

The gugyeol provided a method of reading; the supplementary explanations provided a space for thinking.

A classical Chinese sentence can change meaning depending on how it is segmented, how grammatical relationships are understood, and how individual expressions are interpreted within their larger context.

Consequently, Choi Rip’s work should not be viewed merely as an attempt to make a classical Chinese text easier to pronounce in Korean. It can also be understood as an effort to establish a systematic approach to textual interpretation.

He combined reading and interpretation into a single intellectual process.

Choosing Questions Over Authority

One of the most important features of Choi Rip’s scholarship is his courage to question.

Scholarship does not end with memorizing what previous authorities have said.

A great scholar respects tradition but asks again:

“Is this really so?”

“Does the original text actually say this?”

“Could another interpretation be possible?”

“What evidence supports this interpretation?”

Joseon I Ching scholarship developed within a rich tradition of commentaries, particularly those associated with Cheng Yi and Zhu Xi. Recent scholarship on the history of I Ching thought in late Joseon Korea has also examined how interpretations of the I Ching evolved through intellectual lineages involving figures such as Zhu Xi, Yi Hwang, and Jeong Yak-yong (Kim, 2025).

Choi Rip’s Juyeok Bonui Gugyeol Buseol should be understood within this broader intellectual history.

His approach was not necessarily a rejection of tradition. Rather, it represented a form of dialogue with tradition.

Intellectual originality does not require ignoring everything that came before.

It begins by understanding previous knowledge deeply and then asking:

“How do I understand this?”

Reading the I Ching Through the Eyes of a Literary Master

Here we find an important connection between Choi Rip’s literary and philosophical activities.

Choi Rip was a distinguished classical writer. He paid close attention to the structure, rhythm, and subtle meanings of sentences. It is therefore significant that the same sensitivity to language informed his engagement with the I Ching.

He looked at the text itself.

Where does a sentence begin and end?

How does one expression relate to another?

What meaning does a particular word acquire within its context?

How does a particular interpretation correspond to—or differ from—earlier commentaries?

This approach has an intriguing resonance with contemporary hermeneutics.

Recent research on generative AI has begun to examine AI systems not merely as information-retrieval tools but as cultural technologies that participate in processes of interpretation. Kommers et al. (2026), for example, describe computational hermeneutics as an emerging approach that takes situatedness, plurality, and ambiguity seriously when evaluating generative AI.

Across four centuries, the technological environments are radically different, but the underlying question is surprisingly similar:

Where does meaning come from?

From Choi Rip’s Gugyeol to Modern Prompt Engineering

This is where Choi Rip’s concept of gugyeol offers an intriguing conceptual bridge to generative AI.

Gugyeol provided a framework for approaching a complex classical text.

Generative AI likewise responds differently depending on how a question is structured and what contextual instructions are provided.

Consider the difference between asking an AI system:

“Explain the I Ching.”

and asking:

“First analyze the original passage in its historical and linguistic context. Then compare Zhu Xi’s Zhouyi Benyi with Cheng Yi’s interpretation, identify their differences, and provide the textual evidence supporting each interpretation.”

These are fundamentally different intellectual tasks.

The second prompt provides the AI with a reading procedure and interpretive framework.

In this limited methodological sense, one might think of it as a modern “AI gugyeol.”

This comparison should not be understood as claiming that Choi Rip’s gugyeol and modern prompt engineering are historically identical. They emerged in entirely different intellectual and technological environments. Yet the methodological analogy is valuable: both involve establishing structured procedures for approaching complex textual meaning.

Research on generative AI has likewise demonstrated that structured prompting can influence performance on complex reasoning tasks. Wei et al. (2022), for example, showed that chain-of-thought prompting can elicit improved performance on a range of reasoning problems.

The lesson for the AI age is therefore more profound than simply “ask better questions.”

It is:

Design the question itself as an interpretive structure.

From Verifying the Original Text to Verifying AI

Yet one principle must accompany interpretation:

Interpretation requires evidence.

Generative AI can produce remarkably fluent language. But fluency and truth are not the same thing.

Liusie and Gales (2023) demonstrated the importance of detecting hallucinations in large language models—cases in which models generate plausible but unsupported information.

This problem becomes particularly serious in classical-text scholarship.

An AI system might invent a citation, mix different editions, incorrectly attribute an interpretation to a scholar, or produce a quotation that does not actually exist.

Even an elegant interpretation becomes academically dangerous if its textual foundation is false.

This leads to a first principle that modern AI can learn from Choi Rip’s approach:

Before interpreting, verify the original text.

A second principle follows:

Distinguish evidence from interpretation.

A third is:

Identify the source and provenance of the evidence.

These principles resonate with retrieval-augmented generation (RAG). Lewis et al. (2020) proposed RAG as a way of combining language generation with retrieved external knowledge for knowledge-intensive tasks. The approach illustrates a fundamental problem in generative AI: reliable generation requires access to relevant and verifiable sources.

Choi Rip’s return to the original text can therefore be reinterpreted, in the AI age, as a principle of source-grounded interpretation.

Learning from Tradition Without Becoming Trapped by It

Another important lesson from Choi Rip is the balance between tradition and originality.

Ignoring tradition produces shallow scholarship.

But merely repeating tradition produces little intellectual innovation.

Choi Rip consulted Zhu Xi’s Zhouyi Benyi and Cheng Yi’s Yizhuan while adding his own interpretations (Kyujanggak Institute for Korean Studies, n.d.).

This is also highly relevant to contemporary AI.

Generative AI has been trained on enormous quantities of existing knowledge. Its great strength is therefore the ability to synthesize and recombine existing information at extraordinary speed.

But creativity does not consist merely of repeating what already exists.

It emerges through:

comparison,

tension,

recombination,

and the creation of new perspectives.

The foundation-model literature has likewise emphasized that powerful general-purpose AI systems bring not only extraordinary capabilities but also risks involving bias, reliability, misuse, and social impact (Bommasani et al., 2021).

The AI age therefore needs more than more knowledge.

It needs better frameworks for interpreting knowledge.

The Kind of Person Needed in an Age of Change

Why should we read the I Ching today?

Perhaps because we live in an age defined by change.

Artificial intelligence writes.

AI creates images.

AI writes software.

AI translates.

AI assists research.

AI increasingly performs forms of intellectual labor once reserved for humans.

Yet even as technology changes, fundamental human questions remain.

What matters?

What should we believe?

Which decision is right?

How far should technology be used?

And perhaps the deepest question is:

What should human beings become in a changing world?

At this point, the I Ching becomes remarkably contemporary.

Five Principles for Reading in the Age of AI

Choi Rip’s Juyeok Bonui Gugyeol Buseol can be translated into five principles for the age of generative AI.

First, Return to the Original Text

Even when an AI system speaks with confidence, verify the original text and its source.

Second, Read the Context

Do not interpret an isolated sentence without considering its linguistic, historical, and intellectual context.

Third, Compare Multiple Interpretations

Do not automatically treat the first AI-generated answer as the correct one. Compare competing perspectives.

Fourth, Distinguish Evidence from Interpretation

Separate what the source actually says from what a scholar—or an AI system—infers from it.

Fifth, Keep Final Judgment Human

AI can provide information and possible interpretations, but human beings must remain responsible for value judgments and consequential decisions.

Together, these five principles could become the foundation of a Humanistic AI Hermeneutics inspired by Choi Rip’s approach to classical interpretation.

Why Choi Rip Matters Again

Choi Rip matters today not because he predicted artificial intelligence.

He did not know computers.

He did not know the internet.

He could not have imagined generative AI.

Yet he confronted a problem that technology cannot eliminate:

How should human beings read, understand, evaluate, and interpret knowledge?

That question remains.

Indeed, it may be more important now than ever.

AI can process more information than any individual human.

But the quantity of information does not guarantee the quality of judgment.

AI can generate sentences faster than humans.

But fluency does not guarantee truth.

AI can present multiple perspectives.

But deciding which perspective deserves our trust remains a human responsibility.

This is where Choi Rip returns.

He seems to tell us:

“Read—but do not simply believe.”

“Learn from tradition—but think for yourself.”

“Interpret—but never abandon evidence.”

And above all:

“Read change without being swept away by change.”

More than four centuries ago, Choi Rip used a brush to read the I Ching and search for the meaning of change.

Today, we can read the I Ching with generative AI.

The tools have changed completely.

But the essence of responsible scholarship has not.

Return to the original text.

Understand the context.

Compare interpretations.

Ask new questions.

Verify the evidence.

And keep final judgment within the realm of human responsibility.

This may be the enduring wisdom that Choi Rip’s Juyeok Bonui Gugyeol Buseol offers the age of generative AI:

The ancient foundation of a new humanism—one in which intelligence is not merely the ability to generate answers, but the ability to interpret, question, verify, and judge wisely.

References

Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., Brynjolfsson, E., Buch, S., Card, D., Castellon, R., Chatterji, N., Chen, A., Creel, K., Davis, J. Q., Demszky, D., … Liang, P. (2021). On the opportunities and risks of foundation models. arXiv. https://doi.org/10.48550/arXiv.2108.07258

Kim, Y. (2025). [Article on the historiography and intellectual development of late Joseon I Ching thought]. Korean Journal of Philosophy, 85, 75–101. https://doi.org/10.35504/kph.2025..85.003

Kommers, C., Ahnert, R., Antoniak, M., Benetos, E., Benford, S., Bunz, M., et al. (2026). Computational hermeneutics: Evaluating generative AI as a cultural technology. Frontiers in Artificial Intelligence, 9, Article 1753041. https://doi.org/10.3389/frai.2026.1753041

Kyujanggak Institute for Korean Studies, Seoul National University. (n.d.). Juyeok Bonui Gugyeol Buseol (周易本義口訣附說) [Woodblock editions, Kyujanggak collections]. Seoul National University.

Lee, C. (2012). Two examples of the composition of the method of divination in the I Ching: The divination method of Zhu Xi and Jeong Yak-yong. Semiotic Inquiry, 33, 93–123.

Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.-t., Rocktäschel, T., Riedel, S., & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in Neural Information Processing Systems, 33. https://doi.org/10.48550/arXiv.2005.11401

Liusie, A., & Gales, M. J. F. (2023). SelfCheckGPT: Zero-resource black-box hallucination detection for generative large language models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (pp. 9004–9017). Association for Computational Linguistics. https://doi.org/10.18653/v1/2023.emnlp-main.557

Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E. H., Le, Q. V., & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35. https://doi.org/10.48550/arXiv.2201.11903 +++

{Solti}

August 1, 2026

붓끝에서 변화의 원리를 읽다

간이 최립의 『주역본의구결부설 (周易本義口訣附說)』과 생성 AI 시대의 해석 지혜

조선의 문인 간이(簡易) 최립(崔岦, 1539–1612)을 생각하면 우리는 먼저 뛰어난 문장가를 떠올린다. 그러나 최립을 단지 ‘글 잘 쓰는 문인’으로 기억한다면 그의 학문 세계를 충분히 이해하기 어렵다. 최립에게는 문학만큼이나 깊은 학문적 관심이 있었고, 그 중심에 『주역(周易)』이 있었다.

그 결실 가운데 하나가 바로 『주역본의구결부설(周易本義口訣附說)』이다. 서울대학교 규장각한국학연구원 소장본의 서지자료에 따르면 이 책은 최립이 주희(朱熹)의 『주역본의(周易本義)』에 한글 구결을 붙이고, 『주역본의』와 정이(程頤)의 『역전(易傳)』을 참고하여 자신의 의견을 덧붙인 저술이다. 규장각에는 17세기에 간행된 목판본이 여러 종 전해지고 있으며, 「진주역본의구결소(進周易本義口訣疏)」에는 최립이 1602년 주역 교정청의 작업에 참여한 사실과 1605년에 이르기까지 이 작업을 이어간 과정이 기록되어 있다(서울대학교 규장각한국학연구원, n.d.).

이 책을 바라보는 순간 최립이라는 인물의 모습이 조금 달라진다.

그는 단순히 문장을 잘 쓰는 사람이 아니었다.

그는 텍스트를 어떻게 읽어야 하는가를 고민했던 사람이었다.

원문을 다시 읽는 사람

『주역』은 변화의 책이다.

하늘과 땅이 변하고, 음과 양이 변하고, 인간의 처지가 변한다. 개인의 삶도, 사회도, 국가도 끊임없이 변한다. 그러므로 『주역』의 핵심적인 질문은 단순히 “미래에 무슨 일이 일어날 것인가?”가 아니라, “변화하는 상황에서 우리는 무엇을 어떻게 판단해야 하는가?”라는 질문으로 읽을 수 있다.

최립은 이러한 고전을 읽으면서 기존의 주석을 단순히 반복하지 않았다. 그는 주희의 『주역본의』를 중심으로 구결을 붙이고, 동시에 정이의 『역전』을 참고하면서 자신의 견해를 덧붙였다. 규장각의 해제도 이 책이 바로 이러한 방식으로 구성되었다고 설명한다(서울대학교 규장각한국학연구원, n.d.).

이 점은 중요하다.

구결은 읽기의 방법이고, 부설은 생각의 공간이었다.

한문 원문에 어떤 토를 붙이는가에 따라 문장의 의미가 달라질 수 있다. 문장을 어디에서 끊는가, 어떤 단어를 어떤 관계 속에서 이해하는가에 따라 하나의 구절이 전달하는 의미가 달라진다.

따라서 최립의 구결 작업은 단순히 한문을 한국어로 읽기 쉽게 만드는 작업으로만 이해하기 어렵다. 그것은 텍스트의 의미에 접근하는 하나의 해석 방법이었다.

그리고 그 위에 자신의 설명을 덧붙임으로써 최립은 읽기와 해석을 하나의 학문적 과정으로 결합하였다.

권위보다 질문을 선택하다

최립의 학문적 태도에서 우리가 주목해야 할 것은 질문하는 용기이다.

학문은 앞선 학자의 말을 암기하는 데서 끝나지 않는다.

위대한 학자는 권위를 존중하면서도 다시 묻는다.

“정말 그러한가?”

“원문은 정말 그렇게 말하는가?”

“다른 해석은 가능한가?”

“이 해석은 어떤 근거를 가지고 있는가?”

조선시대 역학은 주희의 『주역본의』와 정이의 『역전』 등 여러 전통적 주석을 중심으로 발전하였다. 최근 조선후기 역학사 연구에서도 주희, 이황, 정약용 등의 사상적 계보와 역학 해석의 변화가 중요한 연구 주제로 다루어지고 있다(김영우, 2025).

최립의 『주역본의구결부설』도 이러한 조선 역학의 흐름 속에서 읽어야 한다.

특히 그는 기존의 해석을 참고하면서도 자신의 의견을 덧붙였다. 이것은 전통을 거부하는 태도라기보다 전통과 대화하는 태도라고 보는 것이 더 적절하다.

학문적 독창성은 아무것도 참고하지 않는 데서 나오지 않는다.

오히려 기존 지식을 충분히 읽은 뒤,

“나는 이것을 어떻게 이해하는가?”

라고 묻는 데서 시작된다.

문장가의 눈으로 『주역』을 읽다

여기에서 최립의 문학과 역학이 만난다.

최립은 뛰어난 한문 문장가였다. 그는 문장의 구조와 표현의 미묘한 차이를 중시했다. 그러한 문장가의 눈으로 『주역』을 읽었다면, 경전의 문장을 단순한 철학적 명제로만 보지 않았을 가능성이 크다.

그는 문장 자체를 보았다.

어떤 단어가 어디에 놓였는가.

앞 문장과 뒤 문장이 어떻게 연결되는가.

하나의 표현이 전체 문맥에서 어떤 의미를 가지는가.

그리고 그 의미가 기존 주석과 어떻게 연결되거나 달라지는가.

이러한 태도는 현대의 해석학과도 흥미롭게 만난다.

최근 생성AI 연구에서는 AI 시스템 자체를 단순한 정보검색 도구가 아니라 ‘맥락을 다루는 문화적 기술’로 바라보는 계산적 해석학(computational hermeneutics)이 제안되고 있다. Kommers 등(2026)은 생성AI의 해석에서 상황성(situatedness), 복수성(plurality), 모호성(ambiguity)을 중요한 문제로 제시하면서, AI 평가 역시 단순한 정답률을 넘어 문화적 맥락과 의미를 고려해야 한다고 주장한다.

수백 년의 시간 차이를 두고 있지만 질문은 놀랍도록 비슷하다.

“문장의 의미는 어디에서 오는가?”

최립의 구결과 오늘날의 프롬트

여기에서 최립의 ‘구결’을 생성AI와 연결해 볼 수 있는 흥미로운 가능성이 열린다.

구결은 복잡한 고전 텍스트를 어떻게 읽을 것인지에 대한 일정한 지침을 제공한다.

생성AI에서도 질문의 형식과 맥락을 어떻게 설계하느냐에 따라 결과가 크게 달라질 수 있다.

예를 들어 AI에게 단순히

“『주역』을 설명하라.”

라고 요청하는 것과,

“『주역』 원문의 문맥을 먼저 분석하고, 주희의 『주역본의』와 정이의 『역전』의 관점을 비교한 뒤, 각각의 해석에 대한 근거와 차이점을 제시하라.”

라고 요청하는 것은 전혀 다른 작업이다.

두 번째 질문은 AI에게 읽기의 규칙과 해석의 절차를 제공한다.

이것은 현대적 의미에서 하나의 ‘AI 구결’이라고 볼 수 있다.

물론 이러한 비교는 역사적 동일성을 주장하는 것이 아니다. 최립의 구결과 현대의 프롬프트는 서로 다른 시대와 기술 환경에서 나온 것이다. 그러나 복잡한 텍스트를 일정한 해석 절차를 통해 이해하려 한다는 방법론적 유사성은 충분히 탐구할 가치가 있다.

생성AI 연구에서도 구조화된 프롬트가 복잡한 추론 과제의 성능에 영향을 줄 수 있음이 연구되어 왔다. Wei 등(2022)은 단계적 추론을 유도하는 chain-of-thought prompting이 여러 추론 과제에서 성능 향상을 가져올 수 있음을 보여주었다.

따라서 최립에게서 우리가 배울 수 있는 현대적 교훈은 단순히 “좋은 질문을 하라”는 정도에 그치지 않는다.

질문 자체를 하나의 ‘해석 구조’로 설계하라.

이것이 중요하다.

원문 확인에서 AI의 사실성 검증으로

그러나 여기에는 반드시 하나의 조건이 따라온다.

해석에는 ‘근거’가 있어야 한다.

생성AI는 매우 유창한 문장을 만들어 낼 수 있다. 하지만 유창함과 진실성은 같은 것이 아니다.

Liusie와 Gales(2023)는 대규모 언어모델이 유창한 답변을 생성하면서도 사실이 아닌 내용을 만들어낼 수 있다는 문제를 지적하고, 이러한 환각(hallucination)을 탐지하기 위한 SelfCheckGPT 방법을 제안했다.

이 문제는 고전문헌 연구에서 더욱 심각해질 수 있다.

AI가 존재하지 않는 문헌을 인용하거나, 서로 다른 판본의 내용을 섞거나, 특정 학자의 견해를 잘못 귀속시키거나, 원문에 없는 구절을 만들어 낸다면 아무리 아름다운 해석이라도 학술적으로는 위험하다.

그러므로 최립의 독법에서 현대 AI가 배워야 할 첫 번째 원칙은 다음과 같다.

“해석하기 전에 원문을 확인하라.”

두 번째는,

“해석과 사실을 구별하라.”

세 번째는,

“근거와 해석의 출처를 밝혀라.”

이는 현대의 retrieval-augmented generation(RAG) 연구가 강조하는 문제와도 연결된다. Lewis 등(2020)은 지식집약적 작업에서 외부 지식을 검색하여 생성 과정에 결합하는 RAG 방식을 제안하면서, 생성 결과의 사실성과 지식의 출처(provenance)를 확보하는 문제가 중요하다고 지적했다.

최립의 ‘원문으로 돌아가기’라는 태도는 이렇게 AI 시대의 source-grounded interpretation으로 다시 읽힐 수 있다.

전통을 읽되 전통에 갇히지 않는 지혜

최립의 또 하나의 중요한 교훈은 ‘전통과 독창성의 균형’이다.

전통을 무시하면 깊이가 사라진다.

그러나 전통을 그대로 반복하면 독창성이 사라진다.

최립은 『주역본의구결부설』에서 주희의 『주역본의』와 정이의 『역전』을 참고하면서 자신의 의견을 덧붙였다(서울대학교 규장각한국학연구원, n.d.).

이것은 현대 AI 연구에서도 중요한 원칙이다.

생성AI는 이미 방대한 기존 지식을 학습하고 있다. 따라서 AI의 강점은 기존 지식을 빠르게 종합하는 데 있다.

하지만 진정한 창의성은 기존 지식을 단순히 반복하는 것이 아니라,

비교하고, 충돌시키고, 재구성하고, 새로운 관점을 만들어 내는 것

에서 나온다.

Foundation model 연구 역시 대규모 사전학습 모델의 강력한 일반화 능력과 함께 편향, 오류, 오용, 사회적 영향 등 다양한 위험을 함께 지적한다(Bommasani et al., 2021).

따라서 AI 시대에 필요한 것은 ‘더 많은 지식’만이 아니다.

‘더 좋은 해석 체계’가 필요하다.

변화의 시대에 필요한 사람

『주역』을 오늘날 다시 읽어야 하는 이유는 무엇일까?

아마도 ‘변화’ 때문이다.

우리는 지금 인류 역사상 가장 빠른 기술적 변화의 시대 가운데 하나를 살아가고 있다.

AI가 글을 쓴다.

AI가 그림을 그린다.

AI가 코드를 작성한다.

AI가 번역한다.

AI가 연구를 돕는다.

AI가 인간의 지적 노동을 보조하고 때로는 대체하기 시작한다.

그러나 기술이 아무리 빨리 변해도 인간에게 남는 질문은 있다.

무엇이 중요한가?

무엇을 믿어야 하는가?

어떤 판단이 옳은가?

기술을 어디까지 사용해야 하는가?

그리고 가장 중요한 질문은 이것이다.

“변화하는 세계에서 인간은 어떤 존재가 되어야 하는가?”

이 질문 앞에서 『주역』은 다시 현대적인 책이 된다.

최립에게서 배우는 AI 시대의 다섯 가지 독법

최립의 『주역본의구결부설』을 오늘날의 생성AI 시대에 적용한다면 다음과 같은 다섯 가지 원칙으로 정리할 수 있다.

첫째, 원문으로 돌아가라

AI가 아무리 확신에 차서 말하더라도 원문과 출처를 확인해야 한다.

둘째, 맥락을 읽어라

한 문장을 고립해서 이해하지 말고 앞뒤 문맥과 역사적 상황을 함께 살펴야 한다.

셋째, 여러 해석을 비교하라

하나의 답변만 정답으로 받아들이지 말고 서로 다른 관점을 비교해야 한다.

넷째, 근거와 해석을 구별하라

“문헌이 말하는 것”과 “연구자가 해석하는 것”을 구분해야 한다.

다섯째, 최종 판단은 인간이 하라

AI가 정보를 제공할 수는 있지만 가치 판단과 책임을 AI에게 완전히 넘겨서는 안 된다.

이 다섯 가지는 최립의 역학적 태도와 현대 AI의 신뢰성 문제를 연결하는 하나의 ‘Humanistic AI Hermeneutics’로 발전시킬 수 있다.

최립은 왜 오늘 다시 필요한가

최립이 오늘날 우리에게 다시 의미를 갖는 이유는 그가 미래의 기술을 예견했기 때문이 아니다.

그는 AI를 알지 못했다.

컴퓨터도 알지 못했다.

인터넷도 알지 못했다.

그러나 그는 인간이 지식을 어떻게 읽고 이해하고 판단해야 하는지에 관한 중요한 질문을 던졌다.

그 질문은 기술이 바뀌어도 사라지지 않는다.

오히려 AI 시대에는 더욱 중요해진다.

오늘날 AI는 우리보다 훨씬 많은 정보를 처리한다.

그러나 정보의 양이 판단의 질을 보장하지는 않는다.

AI는 우리보다 훨씬 빠르게 문장을 만든다.

그러나 문장의 유창함이 진실성을 보장하지는 않는다.

AI는 여러 관점을 동시에 제시할 수 있다.

그러나 어떤 관점을 받아들일 것인지는 여전히 인간의 책임이다.

바로 이 지점에서 최립이 다시 나타난다.

그는 우리에게 말한다.

“읽어라. 그러나 그대로 믿지는 마라.”

“전통을 배워라. 그러나 스스로 생각하라.”

“해석하라. 그러나 근거를 잊지 마라.”

그리고 무엇보다,

“변화를 읽되, 변화에 휩쓸리지는 마라.”

400여 년 전 최립이 붓으로 『주역』을 읽었다면, 오늘 우리는 생성AI와 함께 『주역』을 읽을 수 있다.

그러나 도구가 달라졌다고 해서 독법의 본질까지 달라질 필요는 없다.

원문을 확인하고,

맥락을 이해하고,

기존 해석을 비교하고,

새로운 질문을 만들고,

근거를 검증하고,

마지막 판단은 인간의 몫으로 남겨두는 것.

이것이야말로 간이 최립의 『주역본의구결부설』이 생성AI 시대에 다시 들려주는 오래된 지혜이자 새로운 인간학(Humanistic AI)의 출발점일 것이다.

참고문헌

Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., Brynjolfsson, E., Buch, S., Card, D., Castellon, R., Chatterji, N., Chen, A., Creel, K., Davis, J. Q., Demszky, D., … Liang, P. (2021). On the opportunities and risks of foundation models. arXiv. https://doi.org/10.48550/arXiv.2108.07258

김영우. (2025). 조선후기 역학사상사 서술을 어떻게 할 것인가? 한국철학논집, 85, 75–101. https://doi.org/10.35504/kph.2025..85.003

Kommers, C., Ahnert, R., Antoniak, M., Benetos, E., Benford, S., Bunz, M., et al. (2026). Computational hermeneutics: Evaluating generative AI as a cultural technology. Frontiers in Artificial Intelligence, 9, Article 1753041. https://doi.org/10.3389/frai.2026.1753041

Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.-t., Rocktäschel, T., Riedel, S., & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in Neural Information Processing Systems, 33. https://arxiv.org/abs/2005.11401

Lee, C. (2012). 주역 설시법의 두 가지 구성 사례: 주자의 설시법(揲蓍法)과 다산 정약용의 설시법에 대한 기호적 함의 [A example of two kinds of the method of divination in I Ching]. 기호학 연구, 33, 93–123.

Liusie, A., & Gales, M. (2023). SelfCheckGPT: Zero-resource black-box hallucination detection for generative large language models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (pp. 9004–9017). Association for Computational Linguistics. https://doi.org/10.18653/v1/2023.emnlp-main.557

서울대학교 규장각한국학연구원. (n.d.). 주역본의구결부설(周易本義口訣附說) [규장각 소장본, 奎4296, 奎4582, 奎5234].

Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E. H., Le, Q. V., & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35. https://doi.org/10.48550/arXiv.2201.11903 +++

2026년 8월 15일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026.

Choi, Y. B. (2026, August 15). Reading the Logic of Change Through the Brush: Choi Rip’s Juyeok Bonui Gugyeol Buseol (주역본의구결부설 (周易本義口訣附說)) and the Wisdom of Interpretation in the Age of Generative AI. K-GSP.

Reading Is the Art of Asking

What Choi Rip’s Reading Practice Teaches Us About Prompt Engineering in the Age of AI

Abstract

This essay reinterprets the intellectual practices of Choi Rip (崔岦, 1539–1612), a prominent Korean scholar and literary writer of the mid-Joseon period, through the lens of Prompt Engineering in the age of generative artificial intelligence. For Choi Rip, reading was not merely the accumulation of knowledge. It was an active process of closely examining classical texts, selectively absorbing their structures and meanings, reflecting upon them, and reconstructing them through one’s own writing. His principle that reading should not become excessive—expressed in the idea of “reading should not be diffuse” (讀書貴不汎)—and his practice of deep, deliberate reflection and repeated revision, associated with gakui damsa (刻意湛思), offer valuable intellectual principles for an era characterized by information abundance and AI-generated content. This essay translates Choi Rip’s reading practice into five principles of Prompt Engineering: purposeful reading, selective context construction, structured questioning, iterative prompt refinement, and independent verification. It further argues that Prompt Engineering should not be understood merely as a technical skill for instructing AI, but as a humanistic method for structuring and clarifying human thought. Just as Choi Rip refined his ideas through the brush, contemporary users refine their questions and thinking through prompts while interacting with AI. The central argument is that the most valuable human capability in the AI era is not the ability to generate more information, but the ability to read deeply, ask precise questions, construct meaningful context, and critically verify AI-generated results. Choi Rip’s reading philosophy therefore offers a historically grounded and humanistic foundation for responsible and intellectually productive engagement with artificial intelligence.

Keywords: Choi Rip, Gani-jip, reading practice, Prompt Engineering, generative AI, artificial intelligence, humanistic AI, critical thinking, Context Engineering, Joseon literature

There Is a Difference Between Reading Much and Reading Well

We live in an age in which humanity possesses more information than ever before. With a smartphone, we can access thousands of books and an almost unimaginable volume of digital text. With generative AI, information can now be retrieved, summarized, translated, and reorganized within seconds.

Yet a paradox has emerged.

As information has become abundant, the ability to ask the right question has become more important.

We can ask an AI system almost anything and receive an answer immediately. But a vague question often produces a vague answer. A poorly framed question can even produce a confident but inaccurate response. The central intellectual skill of the AI era, therefore, is no longer simply knowing more. It is knowing what to ask, how to ask it, and how to evaluate the answer.

This is where we should rediscover the sixteenth-century Korean scholar and writer Choi Rip (崔岦, 1539–1612).

The Veritable Records of the Joseon Dynasty describe Choi Rip as a man who devoted himself to learning from an early age and never became negligent in his studies. He was not merely a gifted literary writer. He studied classical prose deeply, examined literary models, refined his writing, and cultivated a broad intellectual world.

For Choi Rip, reading was not simply an act of collecting information.

It was a process of reading, thinking, comparing, reconstructing, and rewriting.

That is precisely why his reading practice has something important to teach us about Prompt Engineering today.

Choi Rip Read to Build Questions, Not Merely to Collect Information

One of the distinctive characteristics of Choi Rip’s scholarship was his close study of classical texts. He did not simply admire ancient writings. He examined their structures, styles, rhetorical techniques, and intellectual meanings and then transformed what he learned into his own writing.

His works include Hansayeoljeoncho (漢史列傳抄), Sipgagunchesi (十家近體詩), and the extensive collection known as Gani-jip (簡易集), which contains epitaphs, inscriptions, prefaces, essays, letters, commemorative writings, and other literary forms.

His engagement with classical prose helped reshape the literary environment of his time.

In contemporary language, this resembles what we might call document-based Prompt Engineering.

A skilled prompt designer does not simply ask an AI system random questions. The user first defines the problem, identifies the necessary background, establishes the desired output, supplies relevant examples, and specifies constraints.

Choi Rip did something remarkably similar.

To write well, he first read good writing. He studied its structure and intellectual purpose and then reconstructed it through his own voice.

In other words,

he read good writing in order to learn how good thinking becomes good writing.

This provides the first principle connecting Choi Rip’s reading practice with AI:

Read for structure, not merely for information.

“Do Not Let Reading Become Overflow”

One of the most revealing statements associated with Choi Rip’s literary thought is:

“In reading, it is important not to become excessive; in writing, it is important not to become superficial.”

This principle is remarkably relevant in the age of generative AI.

Providing an AI system with more information does not automatically produce a better answer.

Too much information can become too little understanding.

Modern Prompt Engineering similarly emphasizes the importance of selecting and organizing context rather than simply increasing the amount of information supplied to a model. Different prompting techniques can produce different outcomes depending on the task, model, and context (Sahoo et al., 2024; Vatsal & Dubey, 2024).

Choi Rip’s warning against excessive reading can therefore be translated into a modern principle:

Do not prompt by accumulation; prompt by selection.

A good prompt is not necessarily good because it is long.

It is good because every important piece of information has a purpose.

Choi Rip’s “Deep Thinking” and Prompt Refinement

Perhaps one of the most modern aspects of Choi Rip’s writing practice was his willingness to revise.

Historical accounts describe him as repeatedly revising his drafts, carefully contemplating his intended meaning—a process associated with the expression gakui damsa (刻意湛思), meaning to concentrate one’s intention and think deeply.

This resembles one of the fundamental principles of Prompt Engineering.

The first prompt does not have to be perfect.

A sophisticated AI user examines the response and asks:

What is missing?

What was ambiguous?

What additional context is necessary?

What evidence should be requested?

What assumptions did the model make?

What should be changed in the output format?

The process becomes:

Prompt → Response → Evaluation → Refinement → Better Prompt

This is remarkably similar to:

Draft → Review → Revision → Rewriting

Research on prompt engineering likewise emphasizes iterative improvement, evaluation, and the analysis of why a particular prompt succeeds or fails (Ye et al., 2024).

Thus, Choi Rip’s practice of deep contemplation can be translated into an AI-era principle:

Prompt Engineering is the art of thinking deeply before asking again.

From Classical Reading to Context Engineering

Choi Rip did not read texts as isolated objects. His writings reveal an intellectual world in which literature, history, philosophy, art, diplomacy, and classical learning were interconnected.

Hansayeoljeoncho demonstrates his engagement with historical biographies and classical prose. Sipgagunchesi reflects his interest in poetic traditions. Gan-jip preserves a wide range of literary and intellectual encounters.

This provides an important connection to a rapidly developing concept in AI: context engineering.

A prompt is an instruction.

Context is the larger intellectual environment in which that instruction operates.

For an AI system, context may include documents, retrieved sources, previous interactions, structured data, examples, memory, tools, and constraints.

Seen from this perspective, Choi Rip’s reading practice can be represented as:

Reading → Context Building → Question → Analysis → Revision

In the AI era, it is therefore not enough to ask a good question.

We must create the intellectual context in which a good question can produce a good answer.

Just as Choi Rip built his intellectual world through extensive and selective reading, modern AI users must construct reliable contexts from carefully selected sources.

Five Prompt Engineering Principles We Can Learn from Choi Rip

Read With a Purpose

Choi Rip’s reading was not passive. He read with intellectual purpose.

The same principle applies to AI.

Instead of asking:

“Explain this.”

we can ask:

“Explain the central argument of this study to undergraduate students, identify three supporting pieces of evidence, and discuss one important limitation.”

The second prompt begins with a purpose.

Select, Don’t Accumulate

Choi Rip’s warning against excessive reading has become even more relevant in the information age.

The goal is not to give AI everything.

The goal is to give it what matters.

Ask for Structure

Choi Rip studied the structures of classical prose.

Likewise, AI users can specify the structure they need:

Claim → Evidence → Counterargument → Conclusion

Structure transforms a vague request into a defined intellectual task.

Iterate

Choi Rip revised his writing repeatedly.

The same principle should govern AI interaction.

The first response should often be treated as a draft rather than a final product.

Prompt Engineering is therefore not a one-time command. It is an iterative process of experimentation and refinement.

Verify Before Trust

Choi Rip worked within a scholarly culture in which texts, sources, and traditions mattered.

Modern AI users need the same intellectual discipline.

An AI system can produce fluent prose without producing accurate facts.

Therefore, important historical claims, statistics, quotations, citations, and scientific assertions should be independently verified against authoritative sources.

Fluency is not evidence of truth.

Deep Reasoning and Choi Rip’s Intellectual Discipline

One of the influential developments in modern Prompt Engineering has been the use of prompting techniques that encourage models to approach complex tasks through intermediate reasoning steps. Wei et al. (2022), for example, demonstrated that Chain-of-Thought prompting could improve performance on various complex reasoning tasks.

This offers an interesting parallel with Choi Rip’s emphasis on deep thought.

Of course, we should not claim that Choi Rip practiced modern Chain-of-Thought prompting.

The historical comparison is conceptual, not technological.

The underlying intellectual principle, however, is similar:

Read → Analyze → Compare → Judge → Express

Instead of rushing immediately toward a conclusion, the thinker develops an intellectual path toward the conclusion.

That discipline remains essential even when AI performs much of the linguistic work.

The Future Is Not Simply Asking AI Questions

We often define Prompt Engineering as “the skill of giving instructions to AI.”

Choi Rip’s reading practice suggests a deeper definition.

Prompt Engineering is the discipline of structuring one’s own thinking so that an AI system can meaningfully assist.

To construct a good prompt, the human user must first determine:

What do I want to know?

What evidence do I need?

What perspective should be used?

What assumptions should be avoided?

What output should be produced?

What must be verified?

The process of writing a good prompt therefore forces the human being to become a better reader, researcher, and thinker.

This is why Prompt Engineering should not be regarded merely as a technical skill.

It can also become a form of humanistic intellectual training.

From Choi Rip’s Brush to the AI Prompt

Choi Rip lived in an age when books were among the most important instruments of knowledge.

We live in an age in which AI increasingly functions as an interface to knowledge.

The tools are radically different.

But the intellectual process remains surprisingly similar.

Choi Rip read books.

We read data, documents, and digital information.

Choi Rip analyzed the structure of classical prose.

We analyze the structure and reliability of AI responses.

Choi Rip revised his drafts.

We refine our prompts.

Choi Rip thought deeply before writing.

We construct context before asking AI to generate.

Choi Rip examined and refined his writing.

We verify and revise AI-generated responses.

Thus, Prompt Engineering is not an entirely new human capability.

It is a digital extension of an ancient intellectual tradition:

reading, thinking, questioning, revising, and verifying.

The AI Era Needs Deep Readers More Than Prompt Technicians

The future of Prompt Engineering is particularly interesting because AI systems themselves are increasingly capable of interpreting user intent and optimizing prompts.

Research into automatic prompt optimization and prompt compression already points toward a future in which some aspects of prompt construction may be increasingly automated (Chang et al., 2024).

What, then, remains uniquely important for humans?

The ability to decide what is worth asking.

As AI becomes better at generating language, humans will need more—not less—of the following:

reading ability, judgment, contextual understanding, questioning ability, verification, and intellectual responsibility.

This is where Choi Rip becomes surprisingly contemporary.

He did not stop at reading.

He thought about what he read, transformed thought into writing, and refined that writing.

That principle remains valid in the AI era.

Where the Brush Ends, the Prompt Begins

Choi Rip used a brush to shape sentences.

We use keyboards and interfaces to shape prompts.

The tool has changed.

The intellectual responsibility has not.

Choi Rip read deeply.

We must read deeply.

Choi Rip selected and interpreted.

We must select and contextualize.

Choi Rip revised.

We must iterate.

Choi Rip cared about the quality of expression.

We must care about the accuracy and reliability of AI output.

His intellectual discipline can therefore be condensed into a simple principle:

Read widely, but do not become overwhelmed. Think deeply, do not rush, and revise what you write.

In the language of the AI era:

Read deeply.
Select carefully.
Prompt clearly.
Iterate relentlessly.
Verify independently.

Ultimately, Prompt Engineering is not merely the art of asking a computer a question.

It is the art of making the human question clear.

Choi Rip refined his thinking through the brush.

We refine our thinking through prompts.

The technology has changed, but the fundamental human task has not.

In the age of AI, the person who reads deeply will remain indispensable—not despite artificial intelligence, but because of it.

And perhaps that is the most important lesson we can rediscover from Choi Rip more than four centuries after his death:

The future of AI depends not only on smarter machines, but on humans who know how to read, think, ask, and verify.

Choi Rip’s approach to reading is an ancient wisdom for Prompt Engineering in the age of AI. A good AI user is not simply someone who knows a great deal, but someone who reads deeply, asks precise questions, and continually verifies the answers.

References

Chang, K., Xu, S., Wang, C., Luo, Y., Liu, X., & Xiao, T. (2024). Efficient prompting methods for large language models: A survey. arXiv. https://doi.org/10.48550/arXiv.2404.01077

맹영일. (2011). 간이 최립의 화론 연구. 동양학, 50, 21–42. https://doi.org/10.17320/orient.2011..50.21

Sahoo, P., Singh, A. K., Saha, S., Jain, V., Mondal, S., & Chadha, A. (2024). A systematic survey of prompt engineering in large language models: Techniques and applications. arXiv. https://doi.org/10.48550/arXiv.2402.07927

Vatsal, S., & Dubey, H. (2024). A survey of prompt engineering methods in large language models for different NLP tasks. arXiv. https://doi.org/10.48550/arXiv.2407.12994

Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q. V., & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35. https://doi.org/10.52202/068431-1800

Ye, Q., Ahmed, M., Pryzant, R., & Khani, F. (2024). Prompt engineering a prompt engineer. Findings of the Association for Computational Linguistics: ACL 2024, 355–385. https://doi.org/10.18653/v1/2024.findings-acl.21

한국민족문화대백과사전. (2023). 최립(崔岦). 한국학중앙연구원.

조선왕조실록. (1612). 광해군일기[정초본], 광해 4년 7월 11일: 전 동지중추부사 최립의 졸기.

읽는 법이 곧 묻는 법이다

간이 최립의 독서법에서 배우는 AI 시대의 Prompt Engineering

초록

본 에세이는 조선 중기의 문인·학자 간이 최립(崔岦, 1539–1612)의 독서와 문장 수련에 나타난 지적 방법을 오늘날 생성형 인공지능 시대의 Prompt Engineering과 연결하여 재해석한다. 최립에게 독서는 단순한 지식 축적이 아니라 고전의 구조와 의미를 깊이 읽고, 선택적으로 수용하며, 사유를 거쳐 자신의 문장으로 재구성하는 과정이었다. 특히 ‘독서귀불범(讀書貴不汎)’이라는 독서관과 깊이 생각하고 반복적으로 글을 수정하는 ‘각의담사(刻意湛思)’의 태도는 정보 과잉과 생성형 AI의 자동화가 공존하는 오늘날 중요한 지적 원리를 제공한다. 본 에세이는 이러한 최립의 독서법을 목적 지향적 읽기, 선택적 맥락 구성, 구조화된 질문, 반복적 프롬트 개선, 독립적 검증이라는 다섯 가지 Prompt Engineering 원리로 확장한다. 또한 Prompt Engineering을 단순한 AI 명령 작성 기술이 아니라 인간 자신의 사고를 명료하게 구조화하는 인문학적 방법론으로 재정의한다. 최립이 붓으로 문장을 다듬었다면, 오늘날의 인간은 프롬트를 통해 AI와 대화하면서 자신의 질문과 사고를 다듬는다. 결국 AI 시대에 가장 중요한 역량은 더 많은 정보를 생산하는 능력이 아니라 깊이 읽고, 정확하게 질문하며, 결과를 비판적으로 검증하는 능력이다. 최립의 독서철학은 이러한 인간 중심의 AI 활용을 위한 역사적·인문학적 토대를 제공한다.

주제어: 최립, 간이집, 독서법, Prompt Engineering, 생성형 AI, 인공지능, 인문학적 AI, 비판적 사고, Context Engineering, 조선 문학

책을 많이 읽는 사람과 잘 읽는 사람

오늘날 우리는 역사상 어느 시대보다 많은 정보를 가지고 있다. 스마트폰 하나만 열어도 수천 권의 책과 수십억 개의 문장이 눈앞에 나타난다. 생성형 AI까지 등장하면서 인간이 정보를 얻는 데 필요한 시간은 더욱 짧아졌다.

그런데 이상한 일이 벌어지고 있다.

정보는 폭발적으로 증가했는데, 제대로 질문하는 능력은 오히려 더 중요해졌다.

AI에게 질문하면 거의 즉시 답을 얻을 수 있다. 그러나 질문이 모호하면 답도 모호하다. 질문이 잘못되면 AI는 매우 그럴듯한 잘못된 답을 만들어낼 수도 있다. 따라서 AI 시대의 핵심 능력은 단순히 “많이 아는 것”이 아니라 “무엇을 어떻게 물어야 하는가를 아는 능력”이 되었다.

이 지점에서 16세기 조선의 문인 간이(簡易) 최립(崔岦, 1539–1612)을 다시 만나 배움을 구할 필요가 있다.

조선왕조실록의 최립 졸기는 그가 “글을 읽을 줄 알면서부터 힘써 배우며 게을리하지 않았다”고 기록한다. 그는 뛰어난 문장가였을 뿐 아니라 고문을 깊이 연구하고, 시문을 다듬고, 다양한 문헌을 읽고 분석했던 학자였다.

최립에게 독서는 정보를 모으는 행위가 아니었다.

읽고, 생각하고, 비교하고, 다시 쓰는 과정이었다.

바로 이 점에서 그의 독서법은 오늘날 Prompt Engineering과 놀랍도록 닮아 있다.

최립의 독서는 ‘정보 수집’이 아니라 ‘질문을 만드는 독서’였다

최립의 학문적 특징 가운데 중요한 것은 고문(古文)을 단순히 감상하는 데 그치지 않고 그 구조와 표현 방식을 분석하여 자신의 문장으로 재구성했다는 점이다.

그는 『한사열전초(漢史列傳抄)』를 편찬하고 『십가근체시(十家近體詩)』를 남겼다. 또한 『간이집(簡易集)』에는 비문·묘지·서발·기문·서간 등 다양한 문체의 글이 수록되어 있다. 한국민족문화대백과사전은 그가 고문의 형식미를 추구했고, 윤근수와 함께 고문을 숭상하여 당대 문단의 분위기를 바꾸었다고 설명한다.

이것은 단순한 독서가 아니다.

오늘날의 표현으로 바꾸면 문헌 기반의 Prompt Engineering에 가깝다.

좋은 프롬트를 만드는 사람은 AI에게 무작정 질문하지 않는다. 먼저 문제를 정의하고, 필요한 배경을 제공하며, 원하는 결과의 형식을 지정하고, 필요한 경우 사례를 제시한다.

최립 역시 좋은 글을 만들기 위해 먼저 좋은 문장을 읽었다. 그리고 그 문장의 구조와 정신을 파악한 뒤 자신의 문장으로 다시 만들어냈다.

즉, 좋은 글을 쓰기 위해 좋은 글을 읽고, 좋은 질문을 하기 위해 좋은 질문의 구조를 연구한 것이다.

이것은 인간의 독서법과 AI의 Prompt Engineering을 연결하는 첫 번째 원리다.

Read for structure, not merely for information.

정보를 얻기 위해서만 읽지 말고, 사고의 구조를 배우기 위해 읽어라.

“읽되 범람하지 말라”

최립의 문장론을 연구한 자료에는 매우 중요한 구절이 남아 있다.

그는 “독서에는 범람하지 않는 것이 중요하고, 글을 지음에는 뜨지 않는 것이 중요하다(讀書貴不汎, 而爲文貴不浮)”고 했다. 또한 학문을 뜻한다면 단순히 책과 글에만 머무르지 않고 “조존함양(操存涵養)”을 해야 하며, 마음을 다스리고 조급하지 않는 것이 중요하다고 보았다. (操存涵養의 전체적인 의미는: “마음을 잘 지키고 덕성을 깊이 기르는 것”)

이 말은 AI 시대에 더욱 의미가 커진다.

생성형 AI에게 수많은 정보를 넣는다고 해서 반드시 좋은 결과가 나오는 것은 아니다.

Too much information can become too little understanding.

Prompt Engineering에서도 핵심은 정보를 무조건 많이 넣는 것이 아니다. 과업에 필요한 맥락을 선택하고, 목표를 분명히 하고, 불필요한 정보를 제거해야 한다. 최근의 Prompt Engineering 연구들도 다양한 프롬트 기법이 작업의 성격과 모델에 따라 서로 다른 효과를 보인다는 점을 보여준다 (Sahoo et al., 2024; Vatsal & Dubey, 2024).

따라서 최립의 “不汎”은 오늘날 다음과 같이 번역할 수 있다.

Do not prompt by accumulation; prompt by selection.

좋은 프롬트는 길기 때문에 좋은 것이 아니다. 좋은 프롬트는 필요한 것을 정확히 선택하기 때문에 좋다.

최립의 ‘각의담사’와 Prompt Refinement

최립의 글쓰기에서 특히 현대적인 면모는 초고를 그대로 내놓지 않았다는 점이다.

한국민족문화대백과사전은 최립이 글을 지을 때 초고를 서너 번씩 바꾸며 ‘각의담사(刻意湛思)’, 즉 ‘뜻을 새기고 깊이 생각하는 과정’을 거쳤다고 설명한다.

이것은 오늘날 Prompt Engineering의 핵심 과정과 거의 정확하게 대응한다.

첫 번째 프롬프트가 완벽할 필요는 없다.

오히려 좋은 AI 사용자는 첫 번째 답을 보고 다시 묻는다.

“무엇이 부족했는가?”

“어떤 조건을 추가해야 하는가?”

“어떤 부분이 모호했는가?”

“어떤 근거를 요구해야 하는가?”

“결과의 형식을 어떻게 바꾸어야 하는가?”

그리고 프롬트를 다시 작성한다.

Prompt → Response → Evaluation → Refinement → Better Prompt

이것은 최립의

초고 → 검토 → 수정 → 재작성

과 놀라울 정도로 유사하다.

오늘날 Prompt Engineering이 단순한 질문 작성 기술이 아니라 반복적인 실험과 평가의 과정이라는 연구도 이러한 관점을 뒷받침한다. Ye et al. (2024)은 프롬트 엔지니어링을 효과적인 지시를 만드는 일뿐 아니라 모델의 오류를 분석하고 무엇이 부족한지를 추론하며 프롬프트를 개선하는 복합적인 작업으로 설명한다.

따라서 최립의 “각의담사”는 AI 시대에 이렇게 다시 정의할 수 있다.

Prompt Engineering is the art of thinking deeply before asking again.

고전을 읽는 최립과 Context Engineering

최립은 책 한 권을 독립적으로 읽는 사람이 아니었다. 그의 저술을 보면 역사·문학·철학·예술·외교 등 다양한 지식 영역이 서로 연결되어 있다.

그의 『간이집』에는 다양한 인물과 문헌, 역사적 사건과 문화적 대상이 등장한다. 『한사열전초』에서는 중국 역사 속 인물과 문장을 선별하고, 『십가근체시』에서는 시학적 전통을 정리했다. 그의 역학 연구 역시 이러한 학문적 관심의 폭을 보여준다.

오늘날 AI 연구에서 중요성이 커지는 개념 가운데 하나가 context engineering이다. Prompt가 AI에게 직접 전달하는 명령이라면, context는 AI가 문제를 해결할 수 있도록 제공되는 자료·배경·검색 결과·기억·도구 등의 전체 환경에 가깝다.

따라서 최립의 독서법을 현대적으로 확장하면 다음과 같은 구조가 된다.

Reading → Context Building → Question → Analysis → Revision

AI 시대에는 단순히 “좋은 질문”을 만드는 것만으로 충분하지 않다.

좋은 질문이 작동할 수 있는 좋은 문맥을 만들어야 한다.

최립이 고전을 읽고 자신의 지식 체계를 구축했던 것처럼, 오늘날 우리는 AI에게 필요한 자료를 선별하고 구조화해야 한다.

최립에게서 배우는 다섯 가지 Prompt Engineering

최립의 독서와 문장 수련을 오늘날의 AI 활용법으로 번역하면 다음과 같은 다섯 가지 원칙을 얻을 수 있다.

첫째, 목적을 먼저 정하라 — Read with a purpose.

무엇을 읽느냐보다 왜 읽느냐가 중요하다. AI에게도 마찬가지다. “이것을 설명해 줘”보다 “이 연구의 핵심 주장을 학부생에게 설명하기 위해 세 가지 근거와 한계점을 제시하라”가 훨씬 강력하다.

둘째, 핵심을 선별하라 — Select, don’t accumulate.

최립의 “讀書貴不汎”은 정보 과잉 시대의 중요한 교훈이다. 모든 정보를 넣는 것이 아니라 문제 해결에 필요한 정보만 선택해야 한다.

셋째, 구조를 요구하라 — Ask for structure.

최립은 고문의 문체와 형식을 연구했다. AI에게도 결과의 구조를 명확히 제시할수록 원하는 결과를 얻기 쉽다. 예컨대 “주장–근거–반론–결론”과 같은 구조를 지정하는 것이다.

넷째, 한 번에 끝내려 하지 말라 — Iterate.

최립이 초고를 여러 차례 수정했던 것처럼 AI의 첫 답변도 초안으로 생각해야 한다. Prompt Engineering은 일회성 명령이 아니라 반복적 개선 과정이다 (Ye et al., 2024).

다섯째, 답보다 사고의 질을 검증하라 — Verify before trust.

AI는 유창한 문장을 만들 수 있지만 유창함이 진실성을 보장하지는 않는다. 따라서 중요한 사실·인용·역사적 주장·통계는 반드시 원자료와 대조해야 한다.

Chain-of-Thought와 최립의 ‘깊이 생각하기’

현대 Prompt Engineering에서 중요한 발전 가운데 하나는 복잡한 문제를 단계적으로 다루도록 유도하는 방법이다. Wei et al. (2022)의 연구는 Chain-of-Thought prompting이 복잡한 산술·상식·기호 추론 문제에서 성능을 향상시킬 수 있음을 보여주었다. 이 연구는 이후 LLM reasoning 연구에 큰 영향을 미쳤다.

여기에서 최립의 “湛思”, 즉 깊이 생각한다는 태도와 흥미로운 평행관계를 발견할 수 있다.

물론 최립이 현대적 의미의 Chain-of-Thought prompting을 사용했다는 뜻은 아니다.

중요한 것은 방법이 아니라 원리다.

복잡한 문제를 바로 결론으로 뛰어넘지 않고,

읽고 → 분석하고 → 비교하고 → 판단하고 → 표현하는 것.

최립의 독서법은 바로 이러한 지적 훈련을 강조했다.

AI 시대에도 인간이 반드시 보존해야 할 능력이 바로 이것이다.

AI에게 질문하는 시대가 아니라 AI와 함께 읽는 시대

우리는 흔히 Prompt Engineering을 “AI에게 명령하는 기술”이라고 생각한다.

그러나 최립의 독서법을 통해 보면 조금 다른 정의가 가능하다.

Prompt Engineering은 AI에게 명령하는 기술이 아니라, 자신의 생각을 정교하게 구조화하는 기술이다.

좋은 프롬트를 만들려면 사용자가 먼저 자신의 목적을 알아야 한다.

무엇을 알고 싶은가?

어떤 자료가 필요한가?

어떤 관점에서 보고 싶은가?

어떤 근거가 필요한가?

어떤 결과 형식을 원하는가?

무엇을 검증해야 하는가?

결국 좋은 프롬트를 만드는 과정에서 인간 자신이 더 좋은 독자가 되고 더 좋은 연구자가 된다.

이것이 AI 시대에 Prompt Engineering이 단순한 기술적 스킬을 넘어 인문학적 사고 훈련이 될 수 있는 이유다.

최립의 붓에서 AI의 프롬트로

16세기의 최립은 붓으로 문장을 만들었다.

21세기의 우리는 키보드로 프롬트를 만든다.

도구는 완전히 달라졌다.

그러나 지적 과정은 놀랍도록 비슷하다.

최립은 책을 읽었다.
우리는 데이터를 읽는다.

최립은 고문의 구조를 분석했다.
우리는 AI의 응답 구조를 분석한다.

최립은 초고를 고쳤다.
우리는 프롬트를 고친다.

최립은 깊이 생각한 뒤 문장을 만들었다.
우리는 맥락을 구성한 뒤 AI에게 질문한다.

그리고 최립은 자신의 문장이 제대로 되었는지를 다시 살폈다.
우리는 AI의 답변을 검증한다.

따라서 Prompt Engineering은 전혀 새로운 인간 능력이 아니다.

그것은 오래된 인간의 지적 전통, 즉 읽고, 생각하고, 질문하고, 고치고, 검증하는 능력의 디지털 확장이라고 볼 수 있다.

AI 시대에 필요한 것은 ‘프롬트 기술자’보다 ‘깊이 읽는 인간’이다

Prompt Engineering의 미래는 흥미롭다. AI 모델 자체가 점점 더 사용자의 의도를 이해하고 프롬트를 자동으로 최적화할 가능성이 높기 때문이다. 실제로 연구자들은 자동 프롬트 최적화와 프롬트 압축 같은 방법을 연구하고 있다 (Chang et al., 2024).

그렇다면 인간에게 남는 것은 무엇인가?

바로 무엇을 질문할 것인가를 결정하는 능력이다.

AI가 문장을 더 잘 만들수록 인간에게 필요한 것은 문장 생성 능력만이 아니다.

오히려

독서력, 판단력, 맥락 이해력, 질문력, 검증력, 그리고 지적 책임감

이 더 중요해진다.

이 점에서 최립은 오늘날 우리에게 매우 현대적인 메시지를 던진다.

그는 많이 읽는 것에서 멈추지 않았다.

읽은 것을 생각하고, 생각한 것을 문장으로 만들고, 그 문장을 다시 다듬었다.

AI 시대에도 이 원칙은 변하지 않는다.

붓을 내려놓은 자리에 프롬트가 있다

최립이 살았던 시대에는 책이 지식의 중심이었다. 오늘날에는 AI가 지식의 인터페이스가 되고 있다.

그러나 인터페이스가 바뀌었다고 해서 인간의 지적 책임까지 AI에게 넘길 수는 없다.

오히려 AI가 강력해질수록 인간은 더욱 정확하게 읽고, 더욱 깊이 생각하고, 더욱 좋은 질문을 만들어야 한다.

최립의 독서법을 한 문장으로 압축한다면 이렇게 말할 수 있을 것이다.

많이 읽되 범람하지 말고, 깊이 생각하되 조급하지 말며, 쓴 글은 다시 고쳐라.

그리고 이것을 AI 시대의 언어로 바꾸면 다음과 같다.

Read deeply.
Select carefully.
Prompt clearly.
Iterate relentlessly.
Verify independently.

결국 AI 시대의 진정한 Prompt Engineering은 컴퓨터에게 질문하는 기술이 아니다.

인간이 무엇을 알고 싶은지 스스로 분명하게 만드는 기술이다.

최립이 붓으로 문장을 다듬었다면, 우리는 프롬트로 사고를 다듬는다.

도구가 바뀌어도 깊이 읽는 인간의 가치는 사라지지 않는다. 오히려 AI 시대에는 그 가치가 더욱 커진다.

그리고 바로 그 지점에서 우리는 400여 년 전의 최립에게서 미래를 배울 수 있다.

최립의 독서법은 AI 시대의 Prompt Engineering을 위한 오래된 지혜다. 좋은 AI 사용자는 단순히 많이 아는 사람이 아니라, 깊이 읽고 정확하게 질문하며 끊임없이 답을 검증하는 사람이다.

참고문헌

Chang, K., Xu, S., Wang, C., Luo, Y., Liu, X., & Xiao, T. (2024). Efficient prompting methods for large language models: A survey. arXiv. https://doi.org/10.48550/arXiv.2404.01077

맹영일. (2011). 간이 최립의 화론 연구. 동양학, 50, 21–42. https://doi.org/10.17320/orient.2011..50.21

Sahoo, P., Singh, A. K., Saha, S., Jain, V., Mondal, S., & Chadha, A. (2024). A systematic survey of prompt engineering in large language models: Techniques and applications. arXiv. https://doi.org/10.48550/arXiv.2402.07927

Vatsal, S., & Dubey, H. (2024). A survey of prompt engineering methods in large language models for different NLP tasks. arXiv. https://doi.org/10.48550/arXiv.2407.12994

Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q. V., & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35. https://doi.org/10.52202/068431-1800

Ye, Q., Ahmed, M., Pryzant, R., & Khani, F. (2024). Prompt engineering a prompt engineer. Findings of the Association for Computational Linguistics: ACL 2024, 355–385. https://doi.org/10.18653/v1/2024.findings-acl.21

한국민족문화대백과사전. (2023). 최립(崔岦). 한국학중앙연구원.

조선왕조실록. (1612). 광해군일기[정초본] 55권, 광해 4년 7월 11일: 전 동지중추부사 최립의 졸기.

2026년 8월 14일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

Choi, Y. B. (2026, August 14). Reading Is the Art of Asking: What Choi Rip’s Reading Practice Teaches Us About Prompt Engineering in the Age of AI. K-GSP.

From mid- to late June 2006, I stayed in South Korea on my way to a conference at Qingdao University in China. My schedule was packed. I first visited Handong Global University in Pohang as a guest lecturer and toured Gyeongju, a historic city in southeastern Korea. I then presented a paper at an international conference at Chungbuk National University. A few days later, I traveled to Gwangyang, South Jeolla Province, to visit poet Song Su-kwon (1940-2016), then a professor of poetry writing at Suncheon National University. I first visited his writing shack along the Seomjin River, and then he and I crossed the river to tour sites associated with Park Kyoung-ni’s celebrated novel The Land. Our day together ended at Ssanggye Temple, a place mentioned in Kim Dong-ri’s short story “The Post Horse” (1947). It was a pleasure to spend the entire day with him, sharing lunch and dinner and talking about poetry and regional culture. Because I felt a personal connection with him as a budding poet from the same province, I promised to translate some of his poems for publication in a book. The project took longer than I had anticipated, but I eventually published the collection a year before his death.

Although I have read and studied Korean poetry since childhood, no other poet has influenced me as directly as Song. Perhaps our shared roots—growing up in rural areas of the same province, speaking the same dialect, and viewing life as a sea of suffering without surrendering to despair—explain my deep attraction to his work. Through his poetry, I rediscovered my native culture and learned to draw poetic ideas and images from my own life and remain true to my feelings. His sensory imagery, East Asian sensibility, empathy with nature, and celebration of rural life taught me how poetry can resonate with readers. Although I write mostly haiku and other Japanese-style poetry, Song has influenced me more than even the four Japanese haiku masters I greatly admire: Basho, Buson, Issa, and Shiki. Translating his poetry has been one of the most gratifying experiences of my professional life.

It was equally meaningful to explore with Song the settings of Park Kyoung-ni’s The Land, only about thirty minutes from his writing shack across the Seomjin River. This was my first visit to South Gyeongsang Province, and I was delighted to see the bridge connecting the two provinces as a symbol of harmony and shared prosperity. Hwagae Market in Hadong likewise celebrates the cultural kinship between the people of the two provinces, despite their distinctive regional dialects. Park Kyoung-ni spent a quarter century completing her epic novel, making it especially meaningful for me to visit the landscape where its opening section unfolds. The journey brought together three things that have shaped my literary life: my native region, Korean literature, and my own development as a poet.

This photo essay presents photographs from Gwangyang, Hadong, and Ssanggye Temple, three places closely connected to modern Korean literature. Together, they reveal how the landscape along the Seomjin River has nurtured literary imagination and brought writers, readers, and neighboring communities together.

Map of Gwangyang and Hadong: Gwangyang in South Jeolla Province and Hadong in South Gyeongsang Province are separated by the Seomjin River, which flows southward into the South Sea. Composite image based on map data © OpenStreetMap contributors.

Namdo 2 Bridge: Completed across the Seomjin River, Namdo 2 Bridge connects Akyang-myeon in Hadong County, South Gyeongsang Province, with Dap-myeon in Gwangyang, South Jeolla Province. The bridge provides a new link between the two municipalities and facilitates travel between the communities on opposite sides of the river. It also serves as a symbol of harmony and friendship between the Jeolla and Gyeongsang regions, which have often been divided by political rivalries and animosities fueled by self-serving politicians. All the photos in this essay were taken on June 15, 2006.

Poet Song Soo-kwon’s Writing Shack in Gwangyang, South Jeolla Province

Poet Song Soo-kwon’s writing shack in Gwangyang along the Seomjin River. Its serene, lush surroundings provide an ideal setting for poetic imagination.

Above the entrance to the writing shack hangs a wooden plaque inscribed 漁樵莊 (Korean: 어초장), meaning “Fisherman and Woodcutter’s Retreat.”

Poet Song Soo-kwon checks on his vegetable garden. To the left, the Seomjin River flows southward; across the river lies South Gyeongsang Province.

Greenery surrounds poet Song Soo-kwon’s writing shack.

Two of Song’s poems, accompanied by illustrations, adorn the walls of his otherwise modest and functional living room. They are titled “아침강” (“Morning River”) and “섶다리” (“Brushwood Bridge”), a temporary bridge made of pine branches and earth.

Drinking tea with poet Song Soo-kwon (right) in his living room.

Poet Song Soo-kwon’s wife left written instructions on the wall explaining how to set the electric switches to the proper temperatures, what to do when leaving the house, and what to do upon returning. Poetry was his entire world, so he often needed reminders about some of life’s practical details. Fortunately, his wife was a loyal, proud, and supportive companion throughout his life.

After leaving the writing shack, poet Song Soo-kwon and I headed toward Hadong County, across the Seomjin River. We were still in Gwangyang at this point.

One year before his death in 2016, I translated forty-two of Song Su-kwon’s poems, with an introduction and notes, and published the collection under the title Eating Alone and Other Poems by Song Soo-kwon (Cyberwit). He was delighted to receive copies of the book. Below are three poems from the collection, each of which is set in his writing shack (Eochojang) or its surrounding landscape.

Yi Cheong-jun’s Hat: A Reply to Yi Cheong-jun’s “Song for Eochojang’s Old Man”

Anyone who has seen the river wind
blow away a hat knows this.
Tangled in the wind, his silver hair looks desolate
as he watches helplessly as his hat floats down the river.
With no chance of retrieving it from the water,
he realizes that his life, too, is waning like the disappearing hat.
Startled, the old man stops casting his line into the river.
He watches the hat drift farther away
until it vanishes without a trace.
The lone sandbank on the other side of the river
is ablaze in the setting sun.

*

Causality

My beloved Zhong [Song’s pet] died.
I buried him under the persimmon tree on the hill of Eochojang.

The following spring, verdant persimmon leaves barked,
“Bow-wow!”

*

My Love

Dogwood blossoms in the mountain village yonder—
let them fall.
Plum blossoms in the river village down yonder—
let them fall.
Apricot, peach, and cherry blossoms—
let them all fall.
Pear blossoms at the mouth of the river—
let them all fall.

This spring, meeting these flowers along the river does not sadden me.
This spring, I have something I really want to rub my cheeks against—
the persimmon buds in the sun-drenched orchard.
This spring, I have something I trust.

Light green buds swell, shy beneath the sunlight.
Likewise, my love swells, and my heart beats faster.
The sun shines on the persimmon buds.
I gasp, rubbing my cheeks against them and whispering to them.
My love amounts to the joy I share with those buds.

Hadong: Made Famous by Park Kyoung-ni’s Epic Novel The Land (1969-1994)

We have arrived in Hadong. Outside our car, rice fields are filled with young rice seedlings, with the harvest season about three months away.

A tourist map of Hadong County, known for its cherry blossoms and the Seomjin River.

The first part of Park’s The Land is set in Korea and northeast China between 1897 and 1908. One of the novel’s main storylines traces the decline of the prominent and wealthy Choi family in Pyeongsa-ri, Hadong. The map is titled “Guide to Vice Minister Choi and His Family’s House.”

Poet Song Soo-kwon passes a mom-and-pop shop on the way to the Choi family’s house and the locations used in the film adaptation of The Land.

A view of the rice fields on the way to the Choi family’s house and the settings of the film adaptation of The Land. The Choi family is portrayed as a manseokji, a wealthy landowner whose fields produced 10,000 seok of rice annually. This would have amounted to roughly 1,440 metric tons of rice, requiring an enormous estate of approximately 2-4 million pyeong (about 1,650-3,300 acres).

This wooden panel depicts selected scenes from the film adaptation of Park’s novel The Land.

Houses used as filming locations for the screen adaptation of Park Kyoung-ni’s The Land, recreating the rural village setting of the novel.

The Choi family’s house.

Poet Song Soo-kwon (right) and I rest in the main wooden hall of the Choi family’s house.

A local man impersonates Vice Minister Choi inside the Choi family’s house.

A peek into one of the rooms in the Choi family’s house.

A sign in Hadong reads “Hwagae Market.” Inscribed beneath it are lyrics from a song written and sung by Jo Young-nam, portraying the market as a symbol of harmony between the people of South Jeolla and South Gyeongsang provinces.

A stone monument at Hwagae Market features a traditional Korean traveling merchant.

A bronze statue commemorating Kim Dong-ri’s short story “The Post Horse” (역마, 驛馬, Yeokma) stands in Hwagae, Hadong, where the story is set. Published in 1947, the story explores the Korean folk belief in yeokmasal, a fate that compels a person to wander rather than settle in one place. To rid her son Seong-gi of his wandering fate, his mother sends him to Ssanggye Temple nearby, but her efforts ultimately fail. Unable to marry the woman he loves, Gye-yeon, Seong-gi eventually accepts his destiny and leaves Hwagae Market to begin a life of wandering.

The entrance to Ssanggye Temple in Hadong, one of the settings in Kim Dong-ri’s short story “The Post Horse” (1947).

With poet Song Soo-kwon (right) in front of the Main Buddha Hall at Ssanggye Temple.

About the Author

Professor of English and Creative Writing and Associate Dean of the School of Humanities and Theology, Missouri Baptist University

John J. Han, PhD, is the author, editor, co-editor, or translator of 36 books. His forthcoming book, Echoes from the Hills: Critical Essays on Ozarks Literature (co-edited with C. Clark Triplett), will be published by the University of Arkansas Press. He has also published more than 3,000 poems, primarily haiku, in a wide range of journals and anthologies. In addition to his long tenure at Missouri Baptist University, he has taught at Kansas State University, the University of Nebraska–Lincoln, Nebraska Wesleyan University, and Washington University in St. Louis. He has also served as a visiting scholar at Georgia College & State University and the University of Ljubljana in Slovenia.

© K-Global Scholars and Professionals Forum. All rights reserved. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

A Summer Festival in the Shenandoah Valley: Discovering the Real America at the Rockingham County Fair

What comes to mind when you hear the word “fair” in America?

A giant amusement park? A farmers’ market? Or perhaps a small community festival where local residents gather to listen to music, eat hamburgers, and spend a summer day together?

Visit the Rockingham County Fair in Virginia, and you will discover all of these things at once. But stay a little longer, and you begin to realize that it is much more than a festival. It is a small version of America itself—a place where agriculture, family, community, youth education, and entertainment come together in one setting.

The 2026 Rockingham County Fair is being held from August 10 through August 15 at 4808 South Valley Pike in Harrisonburg, Virginia. Established in 1949 as a nonprofit organization, the Fair has grown into one of Virginia’s major agricultural fairs.

Its 2026 theme is particularly evocative: “Stars, Stripes & Summer Nights—Rooted in America’s Traditions.”

Stars and stripes, summer nights, and American traditions. These words alone tell us something about what this festival is trying to preserve.

The Olympics of Rural America

One of the most fascinating things about the Rockingham County Fair is that agriculture is not merely a backdrop. It is at the heart of the event.

Cattle, sheep, goats, dairy cattle, poultry, and rabbits are exhibited and judged. Residents also compete in agricultural products, horticulture, flowers, photography, art, cooking, and homemaking.

Youth organizations such as 4-H and FFA (National FFA Organization) play an especially important role in the Fair, giving young people hands-on experience in agriculture, animal care, leadership, and community service.

In this sense, the Fair can almost be called the Olympics of rural America.

For one participant, a carefully raised calf may represent years of dedication. For another, a tomato, pumpkin, flower, photograph, or homemade cake may be the work of a season.

Behind every entry is a simple but powerful message:

“I made this. I grew this. I cared for this.”

That is very different from modern consumer culture.

We purchase perfectly packaged food at supermarkets and consume finished images online. At a county fair, however, producers bring the results of their own work and place them before the community.

That may be one of the most human aspects of American agricultural culture.

For Children, the Fair Is a Classroom

One of the most meaningful aspects of the Rockingham County Fair is the presence of young people.

Through 4-H and FFA programs, children learn to care for animals, cultivate crops, prepare projects, present their work, compete with others, and accept the results of competition.

The Fair includes junior competitions in areas such as homemaking, art, photography, farm crops, flowers, and horticulture.

This is not simply entertainment.

It is a place where book learning and hands-on learning meet.

And that may be especially important in the age of artificial intelligence.

A child can ask ChatGPT, “How do I grow tomatoes?” But no AI system can substitute for the experience of actually planting a tomato, watering it, observing its growth, dealing with pests, harvesting the fruit, and finally bringing it to the Fair.

Children learn something fundamental here:

Good results do not come from good questions alone.
They require time, effort, failure, observation, and repetition.

Perhaps this is exactly the kind of education human beings need most in the age of AI.

And Then the Evening Changes Everything

During the day, you may find yourself watching cattle and sheep.

But as the sun begins to set, the Fair suddenly transforms into a huge entertainment venue.

The 2026 program features musical performances by artists including Wilson Fairchild, Flatland Cavalry, and Clay Walker, along with truck and tractor pulls, rodeo events, and the Demo Derby.

The 2026 schedule includes Bull Bucking & Barrels Rodeo on August 13, Demo Derby on August 14, and the Interstate Tractor Pull on August 15.

Here we discover an interesting American paradox.

In the morning, farmers and children are caring for livestock.

In the afternoon, families are looking at agricultural products, artwork, flowers, and homemade creations.

In the evening, thousands of spectators cheer as tractors, trucks, and cars roar across the arena.

Agriculture and entertainment are not separate worlds.

They coexist within the same community.

That is one of the distinctive strengths of rural American culture.

A Place Where “Free” Still Matters

Another appealing feature of the Fair is its effort to make the event accessible to families and community members.

In 2026, children age six and under can enter free, while special admission programs recognize veterans, first responders, school staff, and healthcare providers on designated days. Wednesday is Youth Day, while Thursday is Sunshine & Senior Day.

These policies suggest that the Fair is not simply a commercial event designed to maximize revenue.

It is a community gathering.

Of course, concerts and certain special events require separate tickets. But the heart of the Fair remains the same: families and neighbors coming together to share time and experience.

Why the Fair Helps Us Understand America

This is the main reason I want to introduce the Rockingham County Fair to K-GSP Forum readers.

To understand America, we should not look only at Washington, D.C.

Nor should we look only at Wall Street in New York.

And we certainly should not look only at the technology companies of Silicon Valley.

America is also created in small communities like this one.

A childcare for a calf.

A farmer exhibits the crops he or she has grown.

A grandmother displays food she has prepared.

A teenager submits a photograph.

A family attends a concert together.

Thousands of people cheer as a tractor race across a dirt track.

All of this is America.

And at the center of it is not a grand government institution, but something much simpler:

community.

A Little Strange—and Therefore More Interesting—to Koreans

For someone who grew up in Korea, an American county fair may initially seem a little strange.

Why do people go to a festival to see cows?

Why do tractors compete in races?

Why would someone bring a homemade cake or a flower from the garden and proudly compete for a ribbon?

But once we understand the meaning behind the Fair, it begins to look completely different.

Korea also has many local festivals and agricultural fairs. But American county fairs often have a particularly strong sense of grassroots community participation.

Someone is a participant.

Someone else is a volunteer.

Another person is a spectator.

Someone else is a sponsor.

Ultimately, everyone becomes part of the Fair.

In that sense, the Fair is also a small laboratory for understanding how American civil society works.

What Becomes More Valuable in the Age of AI

Interestingly, the Rockingham County Fair becomes even more meaningful when viewed through the lens of artificial intelligence.

We now live in an age in which almost everything can be done through a screen.

AI can write an essay.

AI can generate an image.

AI can create a travel itinerary.

But at the Fair, people return to the physical world.

They touch the soil.

They see animals.

They taste food.

They talk with other people.

They listen to music.

They laugh.

They cheer.

Perhaps the more advanced AI becomes, the more humans will need not additional digital experiences, but richer human experiences.

In that sense, the county fair is not a relic of the past.

It may actually be a cultural asset for the future.

Discovering America on a Summer Night

The official Rockingham County Fair program reveals a remarkable combination of agricultural exhibition, youth education, family entertainment, music, motorsports, and community activity.

But more important than its size, awards, or statistics is something you can only understand by walking through the Fairgrounds.

You begin to realize:

America’s real strength does not exist only inside enormous buildings.

Sometimes it is found in a dusty arena.

Sometimes it is found in the proud smile of a child leading a calf.

Sometimes it is found in the face of a farmer looking at a pumpkin he or she has grown.

And when the sun goes down, the lights come on and the music begins.

Families unfold their chairs.

Children run around.

Adults talk with one another.

Somewhere nearby, a tractor engine roars.

Somewhere else, a guitar begins to play.

And in that moment, we realize something important.

A fair is not simply a place where things are bought and sold.

It is a place where people show one another how they live.

And the Rockingham County Fair continues to tell that old American story today.

Stars. Stripes. Summer Nights.

And beneath all of them,

there is still one enduring foundation:

Community. +++

{Solti}

August 14, 2026

Shenandoah Valley의 여름 축제—Rockingham County Fair에서 만나는 미국의 진짜 얼굴

미국에서 “Fair”라고 하면 무엇이 떠오를까?

거대한 놀이공원일까, 아니면 농산물 시장일까? 혹은 카운티 주민들이 모여 음악을 듣고 햄버거를 먹으며 하루를 보내는 작은 지역축제일까?

버지니아의 Rockingham County Fair에 가보면 이 모든 이미지가 동시에 떠오른다. 그런데 조금 더 오래 머물러 보면, 이곳은 단순한 축제가 아니라 미국의 농업·가족·지역공동체·청소년 교육·엔터테인먼트가 한자리에 모이는 작은 미국이라는 사실을 발견하게 된다.

2026년 Rockingham County Fair는 8월 10일부터 15일까지 Harrisonburg의 4808 South Valley Pike에서 열린다. 공식 자료에 따르면 이 Fair는 1949년 설립된 비영리 조직에서 출발했으며, 오늘날 버지니아를 대표하는 농업 박람회 가운데 하나로 성장했다.

그리고 올해의 주제는 재미있게도 “Stars, Stripes & Summer Nights—Rooted in America’s Traditions”이다. 별과 성조기, 여름밤, 그리고 미국의 전통. 이 네 단어만으로도 이 축제가 무엇을 지키고 있는지 짐작할 수 있다.

농촌의 올림픽

Rockingham County Fair의 가장 흥미로운 점은 농업이 단순한 배경이 아니라는 것이다.

소, 양, 염소, 젖소, 가금류, 토끼 등이 전시되고 경쟁한다. 농작물과 원예, 꽃, 사진, 미술, 요리와 가정예술 분야에서도 주민들이 자신의 솜씨를 겨룬다. 특히 4-H와 FFA(National FFA Organization, 미국 농업교육 청소년 조직)의 청소년들이 참여하는 가축 경진대회와 판매 행사는 이 Fair의 중요한 부분이다.

4-H와 FFA와 같은 청소년 단체들은 청소년들에게 농업, 동물 돌봄, 리더십, 지역사회 봉사에 관한 실질적인 현장 경험을 제공한다.

그래서 이곳을 ‘농촌의 올림픽’이라고 불러도 과장이 아니다.

누군가에게는 잘 키운 소 한 마리가 평생의 자랑이고, 다른 누군가에게는 자신이 정성껏 기른 토마토나 호박, 꽃 한 송이가 최고의 작품이다.

사진 한 장, 케이크 하나, 정원에서 기른 채소 하나에도 “내가 이것을 만들었다”는 이야기가 들어 있다.

바로 이 점이 현대의 거대한 소비문화와 다르다.

우리는 슈퍼마켓에서 완벽하게 포장된 식품을 사고, 인터넷에서 완성된 이미지를 소비한다. 그러나 Fair에서는 생산자가 자신의 결과물을 직접 들고 나온다.

그것은 미국 농업문화의 가장 인간적인 모습이다.

아이들에게 Fair는 교실이다

Rockingham County Fair에서 특히 눈여겨볼 것은 청소년들이다.

4-H와 FFA 프로그램을 통해 아이들은 동물을 돌보고, 농작물을 재배하고, 자신의 작품을 준비하고, 다른 사람들 앞에서 발표하고, 경쟁의 결과를 받아들이는 법을 배운다. Fair의 경쟁 부문에는 Junior Homemaking, Junior Art, Junior Photography, Farm Crops, Flowers, Horticulture 등이 포함되어 있다.

이것은 단순한 놀이가 아니다.

책으로 배우는 교육과 손으로 배우는 교육이 만나는 곳이다.

AI 시대에는 더욱 중요한 교육이다.

ChatGPT에게 “토마토를 어떻게 재배하나요?”라고 물어볼 수는 있다. 하지만 실제 토마토를 심고, 물을 주고, 병충해를 관찰하고, 마침내 수확해서 Fair에 출품하는 경험은 AI가 대신할 수 없다.

아이들은 여기서 한 가지 중요한 사실을 배운다.

좋은 결과는 좋은 질문만으로 만들어지지 않는다.
시간, 노력, 실패, 관찰, 반복이 필요하다.

어쩌면 이것이 AI 시대에 인간에게 가장 필요한 교육인지도 모른다.

그런데 저녁이 되면 분위기가 완전히 달라진다

낮에는 소와 양을 구경하던 Fair가 해가 지기 시작하면 갑자기 거대한 공연장이 된다.

2026년 프로그램에는 Wilson Fairchild, Flatland Cavalry, Clay Walker 등의 음악 공연이 마련되었고, 트럭과 트랙터 풀, 로데오, Demo Derby 같은 강렬한 프로그램도 이어진다.

특히 8월 13일에는 Bull Bucking & Barrels Rodeo, 14일에는 Demo Derby, 15일에는 Interstate Tractor Pull이 예정되어 있다.

여기에서 미국 문화의 재미있는 역설을 발견할 수 있다.

오전에는 농부와 아이들이 가축을 돌보고,

오후에는 가족들이 농산물과 미술품을 구경하고,

저녁에는 관중들이 환호하며 트랙터와 자동차의 굉음을 즐긴다.

농업과 엔터테인먼트가 서로 다른 세계가 아니다.

둘은 같은 공동체 안에 존재한다.

이것이 미국 지방문화의 독특한 힘이다.

“공짜”라는 단어가 아직 살아 있는 곳

Fair의 또 다른 매력은 가족 단위 방문객을 위한 다양한 할인이다.

2026년에는 성인 일반 입장권을 사전 구매하면 $5 수준이고, 현장 입장은 $8이다. 6세 이하 어린이는 무료이며, 특정 요일에는 Veterans, First Responders, School Staff, Health Care Providers 등을 위한 무료입장 혜택도 있다. 수요일은 Youth Day, 목요일은 Sunshine & Senior Day로 운영된다.

이런 정책을 보면 Fair가 단순히 수익을 극대화하는 상업적 이벤트가 아니라 지역사회 전체가 함께 즐기는 행사라는 점을 알 수 있다.

물론 콘서트와 특별 행사는 별도 티켓이 필요하다. 그러나 Fair의 핵심은 여전히 가족과 공동체가 함께 시간을 보내는 데 있다.

왜 이곳이 미국을 이해하는 데 중요한가

K-GSP Forum 독자들에게 Rockingham County Fair를 소개하고 싶은 가장 큰 이유가 바로 이것이다.

미국을 이해하려면 워싱턴 D.C.의 정치만 보아서는 안 된다.

뉴욕의 월스트리트만 보아서도 안 된다.

실리콘밸리의 AI 기업만 보아서도 안 된다.

미국이라는 나라는 이런 작은 지역공동체에서도 만들어진다.

한 아이가 송아지를 돌보고,

한 농부가 자신이 기른 농작물을 출품하고,

한 할머니가 정성껏 만든 음식을 선보이고,

한 청소년이 사진을 출품하고,

한 가족이 저녁에 함께 콘서트를 보고,

수천 명의 사람들이 트랙터가 흙먼지를 일으키며 질주하는 모습을 보며 환호한다.

이 모든 것이 미국이다.

그리고 그 중심에는 거창한 국가기관이 아니라 community가 있다.

한국인에게는 조금 낯설지만, 그래서 더 재미있다

한국에서 성장한 사람에게 미국의 County Fair는 처음에는 조금 낯설 수 있다.

“왜 사람들이 소를 보러 축제에 가지?”

“왜 트랙터가 경주를 하지?”

“왜 누군가가 직접 만든 케이크나 꽃을 가지고 와서 상을 받으려고 하지?”

그러나 한 번 그 의미를 이해하면 Fair가 전혀 다르게 보인다.

한국에도 지역축제와 농산물 축제가 많다. 그러나 미국의 County Fair에는 특히 강한 지역공동체의 자발성이 있다.

누군가는 참가자이고,

누군가는 자원봉사자이며,

누군가는 관객이고,

또 누군가는 후원자다.

결국 모두가 Fair의 일부가 된다.

이것은 미국의 시민사회가 작동하는 방식을 보여주는 작은 실험실이기도 하다.

AI 시대에 오히려 더 소중해지는 것

흥미롭게도 Rockingham County Fair를 AI 시대의 관점에서 바라보면 더욱 특별해진다.

우리는 지금 화면 속에서 거의 모든 것을 할 수 있는 시대를 살고 있다.

AI에게 글을 쓰게 할 수 있고,

사진을 만들게 할 수 있으며,

여행계획도 짜게 할 수 있다.

그러나 Fair에서는 사람들이 다시 현실의 세계로 돌아온다.

흙을 만지고,

동물을 보고,

음식을 맛보고,

사람과 이야기하고,

음악을 듣고,

웃고,

환호한다.

AI가 발전할수록 인간에게 필요한 것은 더 많은 디지털 경험이 아니라 더 풍부한 인간적 경험일지도 모른다.

그런 의미에서 County Fair는 과거의 유물이 아니다.

오히려 미래를 위한 문화적 자산이다.

여름밤에 만나는 미국

Rockingham County Fair의 공식 프로그램을 살펴보면 이 축제는 농업 박람회이면서, 청소년 교육의 장이고, 가족축제이며, 음악공연이고, 스포츠 이벤트이며, 지역경제의 장이다. 공식 관광 자료는 이 Fair를 버지니아 최대 규모의 County Fair로 소개하며, 농업 경쟁 프로그램으로 국제적인 인정을 받아온 행사라고 설명한다.

그러나 숫자나 수상 경력보다 더 중요한 것이 있다.

Fairground를 걷다 보면 알게 된다.

미국의 진짜 힘은 거대한 건물 안에만 있는 것이 아니라는 것을.

때로는 흙먼지가 날리는 경기장에 있고,

때로는 어린아이가 자랑스럽게 끌고 가는 송아지에게 있으며,

때로는 직접 키운 호박 하나를 바라보는 농부의 미소에 있다.

그리고 해가 지면 조명이 켜지고 음악이 시작된다.

사람들은 의자를 펴고 앉는다.

아이들은 뛰어다니고,

어른들은 이야기를 나눈다.

어딘가에서는 트랙터 엔진이 굉음을 내고,

어딘가에서는 기타 소리가 들려온다.

그 순간 우리는 깨닫는다.

Fair란 단순히 물건을 사고파는 장소가 아니다.

사람들이 자신의 삶을 서로에게 보여주는 장소다.

그리고 Rockingham County Fair는 그 오래된 미국의 이야기를 오늘도 계속 들려주고 있다.

Stars. Stripes. Summer Nights.

그리고 그 모든 것의 뿌리에는,

여전히 community가 있다.

2026 Rockingham County Fair: 8월 10–15일, 4808 S. Valley Pike, Harrisonburg, Virginia. 공식 행사 일정과 티켓 정보는 Rockingham County Fair 공식 웹사이트에서 확인할 수 있다. +++

2026년 8월 14일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

EXECUTIVE SUMMARY

Hangul, the Korean alphabet created in the fifteenth century, is increasingly recognized as one of the world’s most scientific and accessible writing systems. Its simplicity, phonetic accuracy, digital adaptability, and educational efficiency have led some linguistic communities and language planners to explore its use for literacy development and language preservation. As Korean cultural influence expands globally through media, technology, and education, Hangul’s role is evolving beyond a national script into a potential model for communication, language documentation, and cultural empowerment.

Keywords: Hangul, Literacy, Language Preservation, Korean Wave, Writing Systems

1. Introduction

Hangul occupies a distinctive position among the world’s writing systems because it was intentionally designed to expand literacy. Developed in 1443 under the leadership of King Sejong the Great and promulgated in 1446, Hangul provided ordinary people with a comparatively simple and logical means of representing spoken Korean. Its systematic construction, close relationship between sounds and symbols, and accessibility for learners distinguished it from many writing systems that had evolved gradually over long historical periods.

For centuries, Hangul remained primarily associated with the Korean language, national identity, education, and cultural continuity. In recent decades, however, its international visibility has expanded substantially as interest in Korea has grown through education, trade, migration, digital media, K-pop, films, dramas, and other expressions of Korean culture. Hangul is therefore increasingly encountered not only as the script used to learn Korean, but also as a symbol of Korea’s broader cultural presence and growing global connectivity.

At the same time, interest in Hangul is beginning to extend beyond cultural attraction. Indigenous communities, language revitalization projects, educators, and linguistic researchers have explored its potential relevance for literacy development, documentation of under-recorded languages, digital communication, and preservation of endangered linguistic traditions. These developments suggest that Hangul’s global significance may depend on the interaction of several reinforcing characteristics: scientific design, learning accessibility, digital compatibility, expanding cultural visibility, and its possible usefulness in practical linguistic applications.

This article therefore asks: Why is global interest in Hangul growing, and under what conditions might that interest develop from cultural curiosity into practical linguistic application? Figure 1 summarizes the article’s central logic by tracing the progression from scientific design → learning accessibility → digital compatibility → global cultural visibility → global applications. The resulting applications may take several forms, including language learning, literacy and documentation, and language preservation, providing the conceptual foundation for the sections that follow.

2. Why Hangul Attracts Global Interest

The growing international interest in Hangul reflects three interconnected strengths—its scientific and phonetic design, accessibility for learning and language preservation, and compatibility with digital communication and Korea’s expanding soft power—which together attract diverse audiences ranging from language learners and educators to researchers, cultural communities, and literacy-oriented institutions, thereby reinforcing Hangul’s expanding relevance as both a Korean writing system and a potential global linguistic resource.

2.1 Scientific Design and Phonetic Precision

One of Hangul’s most distinctive characteristics is its intentionally systematic design. Its consonant symbols were developed to reflect aspects of the speech organs involved in pronunciation, while its vowels follow simple and consistent organizing principles, creating a close relationship between visual form and spoken sound. This structure gives Hangul a comparatively transparent sound-symbol system and distinguishes it from writing systems that developed gradually through long historical processes (Kim-Renaud, 1997, 2000; King, 1996; Lee & Ramsey, 2011).

This phonetic orientation also supports precise representation of spoken sounds and allows learners to recognize recurring patterns and connect letters with pronunciation relatively directly. Such characteristics are particularly relevant for language learning and documentation because a practical writing system must represent meaningful sound distinctions while remaining accessible to ordinary users (Coulmas, 2003; Sampson, 1985; Sohn, 1999). Hangul’s significance, however, should not be interpreted as universal superiority over other writing systems; its value lies in demonstrating how scientific design, phonetic precision, logical organization, and practical usability can be combined in ways that merit broader international attention.

2.2 Learning Accessibility and Language Preservation

A second source of global interest is Hangul’s relative accessibility for learners, supported by its limited number of basic symbols, regular combinations, and close sound-symbol correspondence, which can lower initial barriers to literacy and facilitate early reading development (Kim-Renaud, 1997, 2000; Sohn, 1999; Song, 2005). This accessibility is especially significant when considered in relation to language documentation and preservation. Minority and indigenous languages that remain primarily oral, lack standardized orthographies, or face declining intergenerational transmission may benefit from adaptable phonetic systems for recording oral histories, producing educational materials, and documenting vocabulary and pronunciation (Coulmas, 2003; Sampson, 1985). At the same time, any use of Hangul should respect identity, cultural memory, and community ownership by supporting locally defined literacy and preservation goals without displacing established scripts or traditions.

Table 1. Hangul and Selected Writing-System Traditions: Comparative Characteristics

Source: Authors’ comparative synthesis based on Coulmas (2003), Kim-Renaud (1997, 2000), King (1996), Lee and Ramsey (2011), Sampson (1985), and Sohn (1999).

Note: The comparisons in Table 1 are descriptive rather than hierarchical evaluations. Characteristics such as accessibility, sound representation, adaptability, and digital suitability depend partly on the language being represented, educational context, user population, and purpose of use. The table therefore does not imply that Hangul is universally superior to, or should replace, established writing systems.

Table 1 compares Hangul with selected writing-system traditions—Latin alphabet conventions in English and Spanish, Chinese characters, and Devanagari as used for Hindi—across learning accessibility, sound representation, cross-linguistic adaptability, digital suitability, and structural characteristics. The comparison emphasizes structural differences rather than assigning universal rankings, since the effectiveness of any writing system depends on the language, historical context, community practices, and purpose for which it is used. Hangul is distinctive for its systematic design, relatively transparent relationship between symbols and speech sounds, syllable-block organization, and adaptability to contemporary digital environments, while the other systems possess their own important linguistic, historical, and cultural strengths. Accordingly, Hangul is presented here not as a universally superior or replacement script, but as an intentionally designed writing system whose characteristics may offer useful possibilities for language learning, documentation, and locally chosen linguistic experimentation.

2.3 Digital Compatibility and Korean Soft Power

A third source of global interest is Hangul’s compatibility with contemporary digital communication, as it functions efficiently across computers, smartphones, search systems, social media, online education, and other networked platforms. This digital readiness strengthens Hangul’s international accessibility and connects its linguistic structure with broader patterns of technological modernization and global communication. At the same time, the rapid expansion of Korean cultural visibility through K-pop, dramas, films, gaming, education, tourism, and international business exposes global audiences to Korean words and written expressions, allowing Korean soft power to convert cultural attraction into educational and linguistic engagement (Nye, 2004; Choy & Hong, 2026). The interaction of digital compatibility and cultural visibility therefore helps explain why Hangul increasingly symbolizes not only Korean language and identity, but also creativity, modernity, resilience, and global connectivity (Hong & Hyun, 2025).

3. Where Is Global Interest Emerging?

Global interest in Hangul is not concentrated in one region, nor does it arise from a single motivation. In some countries, interest is primarily connected to Korean language learning, employment, trade, and popular culture; in others, Hangul attracts attention because of its phonetic characteristics, educational accessibility, or possible usefulness for documenting languages that lack standardized writing systems. This diversity can be understood by distinguishing cultural, educational, economic, and language-preservation motivations across countries and organizations.

3.1 Asia and Central Asia

Asia and Central Asia provide some of the clearest examples of international engagement with Hangul, although the forms and motivations of that engagement vary substantially across countries and communities. The Cia-Cia community of Baubau, Indonesia, offers the best documented case of experimental adaptation of Hangul to a non-Korean language, showing both its possible usefulness for minority-language documentation and preservation and the importance of local acceptance, education, and community use for long-term viability, while also cautioning against describing Hangul as a formal replacement for Indonesia’s established writing system (Cho, 2012; Song, 2013). Mongolia and Uzbekistan illustrate a different pattern in which interest in Hangul is linked primarily to Korean-language education, employment, tourism, cultural exchange, and expanding institutional ties with Korea, as reflected in Mongolia’s early connection to the King Sejong Institute network and Uzbekistan’s rapid growth in Korean-language learning and related educational participation (King Sejong Institute Foundation, 2025). Taken together, these cases show that the regional pattern is not a single movement toward adopting Hangul as a local script, but a set of diverse pathways through which linguistic need, cultural attraction, educational opportunity, and relationships with Korea expand interest in Hangul.

3.2 North America and Europe

In North America and Europe, interest in Hangul has grown primarily within countries that already possess established writing systems and high literacy levels, making replacement of existing scripts neither necessary nor likely. The principal drivers are Korean-language education, university-based Korean studies, diaspora communities, cultural consumption, academic research, and expanding economic and technological relationships with Korea, within a broader global increase in Korean-language learning. This pattern highlights an important distinction between learning Hangul and adopting Hangul, since countries such as the United States, Canada, the United Kingdom, France, and Germany engage with Hangul mainly through language study, culture, education, business, and international exchange. Cultural visibility and institutional support can therefore sustain substantial and continuing interest in Hangul without creating a functional need or policy rationale for replacing established local writing systems.

3.3 International Institutions and Language Communities

A third area of interest extends beyond individual nations to international organizations, minority-language communities, indigenous groups, and language-preservation initiatives, where attention shifts from cultural learning toward literacy, documentation, multilingualism, and language survival. Communities lacking standardized orthographies or facing declining intergenerational transmission may find Hangul’s phonetic structure, learnability, and digital compatibility useful for recording speech, developing educational materials, and preserving linguistic heritage, although neither UNESCO nor the United Nations officially promotes Hangul as a universal or global writing system (Brouillette, 2019; Limage, 1999). Major Korean Studies institutions in the United States, Canada, the United Kingdom, and South Korea—including leading centers at Harvard, USC, Columbia, George Washington University, the University of Michigan, SOAS, the University of British Columbia, The Ohio State University, the University of Hawai’i at Mānoa, and the Academy of Korean Studies—provide durable infrastructure for language education, scholarship, academic exchange, and public engagement. Collectively, these organizations and institutions suggest that Hangul’s international significance lies not in replacing established scripts, but in supporting locally defined linguistic needs while sustaining broader educational, cultural, and scholarly engagement with Korea.

4. A 2×2 Framework for Understanding Global Interest

The international diffusion of Hangul reflects more than cultural popularity, as some countries and communities are drawn by Korean culture, education, business, and media while others recognize potential value in its phonetic structure for literacy, documentation, and language preservation; a 2×2 framework distinguishes these motivations by combining cultural and educational interest in Korea with functional linguistic need; this framework clarifies where engagement is likely to remain primarily cultural, where practical applications may emerge, and where future opportunities may develop. Countries and communities differ substantially in their functional need for Hangul and their cultural and educational interest in Korea, and the interaction of these two dimensions helps explain varying patterns of global engagement.

4.1 Four Patterns of Global Engagement

Integrated Engagement represents settings in which strong functional linguistic needs coincide with substantial cultural and educational engagement with Korea. Countries or communities in this group have practical reasons to explore Hangul and sufficient familiarity with Korea to support experimentation, learning, and possible adaptation. The Cia-Cia community in Indonesia is the clearest example because Hangul has been explored in connection with language documentation and preservation, while Mongolia, Uzbekistan, and Kazakhstan illustrate broader combinations of educational cooperation, language learning, and cultural engagement. This pattern provides the most favorable conditions for sustained practical use because linguistic need and cultural openness reinforce each other.

Cultural Engagement is characterized by high cultural and educational interest in Korea but relatively low functional need for an alternative writing system. Countries such as the United States, France, Germany, Thailand, Vietnam, and Japan already possess established writing systems, so their engagement with Hangul is driven primarily by Korean-language education, K-pop, K-drama, tourism, business, academic exchange, and other dimensions of Korean soft power. In these settings, Hangul functions mainly as a gateway to Korean language and culture rather than as a practical replacement for existing scripts. This pattern is likely to remain an important source of international Hangul learners and cultural engagement.

Figure 2. Global Engagement with Hangul: A 2×2 Classification Framework

Source: Authors’ Compilation

Figure 2 identifies four analytical patterns—Integrated Engagement, Cultural Engagement, Specialized Engagement, and Emergent Engagement—based on different combinations of functional linguistic need and Korea-related cultural and educational interest. These patterns are dynamic rather than permanent, as cultural exposure, educational relationships, linguistic needs, and institutional conditions may shift over time.

Specialized Engagement includes communities with meaningful functional needs but comparatively limited cultural or educational engagement with Korea. These may include minority or indigenous-language communities facing challenges such as weak orthographic standardization, language endangerment, limited literacy resources, or insufficient documentation. Indigenous-language communities in parts of Africa, Oceania, Papua New Guinea, the Pacific Islands, and other regions illustrate settings in which Hangul’s phonetic structure could potentially offer practical value. The emphasis here is on future possibility and community-defined need, especially where accessible tools are sought for documenting and preserving local languages.

Emergent Engagement describes settings in which both functional need for Hangul and cultural or educational interest in Korea remain relatively low. Countries in this group generally possess established linguistic infrastructures and have comparatively limited educational, cultural, or institutional interaction with Korea. Under these conditions, interest in Hangul is likely to remain modest. Movement toward greater engagement would depend on changes such as stronger Korean cultural visibility, expanded educational or economic relationships, or the emergence of new linguistic needs.

4.2 From Cultural Interest to Sustainable Application

The framework suggests that the future international role of Hangul will depend on how different forms of engagement develop from the interaction between functional linguistic need and Korea-related cultural and educational interest. Cultural visibility can create awareness and initial curiosity, while practical linguistic needs may create opportunities for literacy, documentation, language learning, and preservation. The four engagement patterns—Integrated Engagement, Cultural Engagement, Specialized Engagement, and Emergent Engagement—therefore provide different pathways through which interest in Hangul may develop into more sustained forms of international application.

Proposition 1. Integrated Engagement. Sustained engagement with Hangul is most likely when functional linguistic needs and Korea-related cultural or educational interest are both strong.

Integrated Engagement reflects situations in which practical linguistic need and substantial familiarity with Korea reinforce one another. Communities are more likely to experiment with, adapt, and sustain the use of Hangul when they have both a functional reason to consider the script and sufficient educational, cultural, or institutional relationships to support learning and implementation. The Cia-Cia experience in Indonesia provides the clearest illustration, while Mongolia, Uzbekistan, and Kazakhstan represent emerging or prospective forms of such engagement.

Proposition 2. Cultural Engagement. Strong Korea-related cultural and educational interest can sustain substantial engagement with Hangul even when functional need for an alternative local writing system remains limited.

Cultural Engagement emphasizes the role of Korean language education, K-culture, academic exchange, tourism, business, and other dimensions of Korean soft power in generating international interest in Hangul. In countries such as the United States, Canada, France, Germany, the United Kingdom, Vietnam, Thailand, and Japan, Hangul primarily functions as a gateway to Korean language and culture rather than as a substitute for established writing systems. Cultural visibility can therefore produce meaningful and durable engagement without necessarily resulting in local script adoption.

Proposition 3. Specialized Engagement. Hangul may have particular value where functional linguistic needs are strong even when Korea-related cultural and educational interest remains comparatively limited.

Specialized Engagement focuses on communities facing specific needs in literacy development, language documentation, orthographic development, or language preservation. In these settings, Hangul may be evaluated primarily for its phonetic structure, learning accessibility, and digital compatibility rather than because of broader cultural interest in Korea. Minority and indigenous-language communities may therefore represent settings in which the practical characteristics of Hangul become more important than Korean soft power itself.

Proposition 4. Emergent Engagement. Engagement with Hangul may develop over time even where both current functional need and Korea-related cultural or educational interest remain relatively low.

Emergent Engagement highlights the dynamic nature of international interest in Hangul. Countries or communities with limited present engagement may move toward stronger forms of involvement as Korean cultural visibility expands, educational and economic relationships deepen, digital access increases, or new linguistic needs emerge. This pattern suggests that the international position of Hangul should be viewed as evolving rather than fixed, with changing circumstances potentially creating new pathways toward cultural, specialized, or integrated engagement.

5. Conclusion

Hangul began as a practical Korean solution for expanding literacy, but its scientific design, phonetic precision, learning accessibility, and digital compatibility have increasingly extended its relevance beyond Korea. Korea’s growing cultural influence has further expanded global awareness of Hangul, while educational, linguistic, and language-preservation needs have created additional pathways for more substantive engagement. Its future international contribution appears especially promising where cultural interest intersects with locally defined needs for language learning, literacy, documentation, and preservation. In this sense, Hangul’s broader significance lies not in replacing established writing systems, but in demonstrating how an intentionally designed script can serve as an accessible, adaptable, and practical linguistic resource in an increasingly connected world.

References

Brouillette, S. (2019). UNESCO and the fate of the literary. Stanford University Press.

Cho, T.-Y. (2012). Cia-Cia language: From the era of oral to the era of writing. Humaniora, 24(3), 324–332. https://doi.org/10.22146/jh.1374

Choy, B. G., & Hong, P. (2026). Soft power of South Korea in the digital age: Reimagining K-pop, K-movies, and cultural diplomacy through AI-enabled diffusion. In S. Kang, W. Heo, W. H. Lee, J. Lee, & Y. B. Choi (Eds.), Korean studies in the artificial intelligence era: Volume 2: Adoption insights in hospitality, culture, communication, music, and philosophy (pp. 178–209). CRC Press.

Coulmas, F. (2003). Writing systems: An introduction to their linguistic analysis. Cambridge University Press.

Hong, P. C., & Hyun, Y. R. (2025). Impact of diverse orientations in human emergency responses: Illustrations of Korean national experiences in modern history. In K. A. Thomas & J. V. Kureethara (Eds.), Education and pedagogical experiences: Coping with human emergencies and exploring resilience strategies (pp. 277–301). Routledge India.

Kim-Renaud, Y.-K. (Ed.). (1997). The Korean alphabet: Its history and structure. University of Hawai’i Press.

Kim-Renaud, Y.-K. (2000). Sejong’s theory of literacy and writing. Studies in the Linguistic Sciences, 30(1), 13–46.

King, R. (1996). Korean writing. In P. T. Daniels & W. Bright (Eds.), The world’s writing systems (pp. 218–227). Oxford University Press.

King Sejong Institute Foundation. (2025, June 30). KSIF designates 11 new KSIs in 9 countries for 2025. https://www.ksif.or.kr/cop/bbs/selectBoardArticle.do?nttId=9220000007293&bbsId=BBSMSTR_000000000141

Lee, K.-M., & Ramsey, S. R. (2011). A history of the Korean language. Cambridge University Press.

Limage, L. J. (1999). Literacy practices and literacy policies: Where has UNESCO been and where might it be going? International Journal of Educational Development, 19(1), 75–89.

Nye, J. S., Jr. (2004). Soft power: The means to success in world politics. PublicAffairs.

Sampson, G. (1985). Writing systems: A linguistic introduction. Stanford University Press.

Sohn, H.-M. (1999). The Korean language. Cambridge University Press.

Song, J. J. (2005). The Korean language: Structure, use and context. Routledge.

Song, S.-W. (2013). Being Korean in Buton? The Cia-Cia’s adoption of the Korean alphabet and identity politics in decentralised Indonesia. KEMANUSIAAN: The Asian Journal of Humanities, 20(1), 51–80.

ABOUT THE AUTHORS

Distinguished Professor, Dr. Paul Hong (Editor in Chief) — University of Toledo

Paul C. Hong is Distinguished University Professor at the University of Toledo. Drawing on more than four decades of leadership and missional service and three decades of scholarship in leadership and supply chain management, his research examines how individuals, organizations, and societies develop the capabilities necessary for resilience, adaptation, and long-term impact. His recent work integrates history, management, philosophy, and faith-based narratives to explore leadership formation through the development of identity and purpose, interests and priorities, insight and perspective, integrity and principles, and influence through stewardship—qualities that remain essential in the age of artificial intelligence.

Peter C. Park — Environmental Engineer, City of Toledo

Peter C. Park is an Environmental Engineer with more than 25 years of experience serving the City of Toledo and holds graduate degrees in Chemical Engineering and Occupational Health from the University of Toledo. He is currently pursuing certification as a Korean language instructor and has a strong interest in expanding access to Korean language and culture for both heritage learners and non-Korean speakers.

© K-Global Scholars and Professionals Forum. August 2026. All rights reserved.

Original Article:

8 18 26 The Growing Global Interest In Hangul Hong&park

2.23MB ∙ PDF file

Download

Download

Suggested Citation

Hong, P. C., & Park, P. C. (2026, August 18). The growing global interest in Hangul: Why nations and communities are exploring the Korean writing system. K-GSP Forum, pp. 1–12.

한글요약

한글의 세계적 관심 확대

Paul C. Hong: 톨레도대학교 석좌교수로서 글로벌 공급망, 혁신, 회복탄력성 및 한국의 글로벌 문화·기술 확산을 연구하고 있다.
Peter C. Park: 톨레도시 환경엔지니어로서 한국어 교원 자격을 준비하며 한국어와 한국문화의 교육 및 확산에 관심을 갖고 있다.

한글은 15세기 세종대왕의 주도 아래 백성들의 문해력 향상을 목적으로 창제된 문자이지만, 오늘날에는 한국의 국가 문자를 넘어 세계적 관심을 받는 언어 자원으로 확장되고 있다. 한글의 체계적인 자음과 모음 구조, 높은 음성 표기 능력, 비교적 쉬운 학습 접근성, 디지털 환경에서의 효율성은 다른 문자 체계와 구별되는 중요한 특징이다. 특히 한글은 과학적 설계와 실용성을 동시에 갖춘 문자라는 점에서 언어 학습뿐 아니라 문해 교육, 언어 기록, 소수언어 및 위기언어 보존과 같은 다양한 가능성을 보여 준다. 이러한 세계적 관심은 한글 자체의 구조적 장점뿐 아니라 스마트폰, 컴퓨터, 검색, 소셜미디어, 온라인 교육과 같은 디지털 환경의 확산, 그리고 K-pop, 한국 드라마, 영화, 게임, 관광, 교육, 비즈니스 등 한국의 소프트파워 확대와도 밀접하게 연결되어 있다.

지역별로 보면 한글에 대한 관심의 성격은 매우 다양하다. 인도네시아의 찌아찌아 공동체는 한글을 지역 언어 기록과 보존의 가능성과 연결해 왔으며, 베트남과 태국에서는 한국어 학습, 취업, 무역, 관광, 문화적 관심이 주요 동기가 되고 있다. 몽골, 우즈베키스탄, 카자흐스탄에서는 교육 교류, 노동 이동, 경제 협력, 디아스포라 및 외교 관계가 한글 학습과 관심을 촉진하는 반면, 미국, 캐나다, 프랑스, 독일, 영국 등에서는 기존 문자 체계를 대체하려는 필요보다 한국어 교육, 대학의 한국학 프로그램, 문화 소비, 학술 교류, 비즈니스 관계가 중심적 역할을 한다. 이러한 차이는 한글의 세계적 확산이 하나의 동일한 경로를 따르는 것이 아니라 각 지역의 문화적 관심, 교육적 기반, 경제적 관계, 그리고 실제 언어적 필요에 따라 서로 다른 형태로 나타남을 보여 준다.

이 논문은 이러한 차이를 설명하기 위해 한글의 기능적 필요성과 한국에 대한 문화·교육적 관심이라는 두 차원의 2×2 개념 틀을 제시하고, 국가와 공동체를 통합적 참여, 문화적 참여, 전문적 참여, 초기적 참여로 구분한다. 이 틀은 한글에 대한 관심이 단순한 한류 현상이 아니라 문화적 매력과 실제 언어적 필요가 서로 다른 방식으로 결합하는 현상임을 보여 준다. 따라서 한글의 세계화는 다른 문자 체계를 대체하는 방향으로 이해되어서는 안 되며, 특히 소수언어와 위기언어 공동체에서는 외부에서 강요되는 대체 문자가 아니라 지역 공동체가 스스로 필요성을 인정하고 선택할 수 있는 보조적 언어 자원으로 접근되어야 한다. 이러한 관점에서 한글의 장기적 국제적 가치는 한국 문화에 대한 관심이 문해 교육, 언어 학습, 기록, 보존과 같은 실제 필요와 결합하고, 과학적 설계, 교육적 접근성, 디지털 적응성, 지역 공동체의 자율적 활용 가능성이 함께 작동할 때 더욱 커질 수 있다.

Choi Rip’s Four-Point Memorial on Abolishing Factionalism in the Age of Artificial Intelligence

Political Unity, Remonstrance, and Popular Trust in Joseon Korea during the Imjin War

Abstract

This essay examines a memorial submitted by Choi Rip (崔岦, 1539–1612), known by his literary name Gani (簡易), to King Seonjo in 1596, during the Imjin War. Recorded in the Annals of King Seonjo, the memorial presents four interconnected political questions concerning Heaven, the state, ministers, and the people. Choi Rip explicitly refers to them as “four points” (四條), making the document an important primary source for understanding his political thought and his response to the crisis of wartime Joseon (National Institute of Korean History [NIKH], 1596/2026).

The second point is especially significant. Choi Rip asks whether there was “偏黨, 不相和睦” — factionalism and lack of harmony — within the royal court. Drawing on the Analects, the Book of Changes (Yijing), and Zhu Xi’s interpretation, he invokes “渙其群” and explains it as “散其朋黨,” or dispersing factional associations. He therefore regarded factionalism not simply as personal disagreement but as a potential threat to political cohesion and national survival during wartime.

The other three points complete his political framework. First, the ruler must practice self-discipline, particularly 慎獨 (sindok), or moral vigilance when alone. Second, ministers must not merely obey the king outwardly while criticizing him privately; they must practice honest remonstrance (直諫). Third, the people must possess a willingness to endure hardship and defend the state because they trust their government. Choi Rip thus presents an integrated theory of crisis governance based on ethical leadership, political unity, truthful counsel, and popular trust.

This essay argues that Choi Rip’s memorial should not be interpreted anachronistically as an anticipation of modern democracy or party politics. Rather, it represents a distinctly Confucian theory of political self-correction within the institutional framework of the Joseon Dynasty. Yet its central question remains remarkably relevant: How can a political community preserve unity, hear uncomfortable truths, correct its own mistakes, and maintain public trust when confronted by existential crisis?

Keywords: Choi Rip, 崔岦, Gani, 簡易, King Seonjo, Joseon, factionalism, remonstrance, political unity, public trust, Imjin War, Yijing, Confucian political thought

A Political Question in the Middle of War

In 1596, Joseon Korea was still struggling with the devastating consequences of the Imjin War. The war was not confined to battlefields. Building fortifications, transporting provisions, mobilizing soldiers, training troops, and defending coastal areas placed enormous burdens upon the population.

It was in this environment that Choi Rip asked a fundamental question about the survival of the state.

He began his memorial by describing the situation as one in which:

“國家存亡, 間不容髮 / 국가존망, 간불용발”

The survival or destruction of the state was separated by no more than the width of a hair. He further observed:

“君臣同憂 / 군신동우”

Ruler and ministers shared the nation’s anxiety and were exhausting themselves in search of solutions (NIKH, 1596/2026).

Yet Choi Rip did not limit his proposal to military measures.

He asked a deeper question:

Why does a state become vulnerable to external enemies?

His answer was not simply a matter of military strength. A nation’s survival depended upon the moral condition of its leadership, political cohesion within the court, the willingness of ministers to speak honestly, and the confidence of the people in their government.

This perspective is particularly important when viewed against the political background of early Joseon factional politics under King Seonjo. The growing political influence of the Sarim literati was accompanied by the emergence of the Eastern and Western factions. Modern scholarship has connected this factional development with structural tensions within the political order of Seonjo’s reign (Lee, 2012).

Choi Rip’s 1596 memorial should therefore be understood at the intersection of two crises:

the external crisis of the Imjin War and the internal crisis of political factionalism.

Choi Rip and the Historical Significance of the Memorial

Choi Rip (1539-1612)’s courtesy name was Ipji (立之), and his literary name was Gani (簡易). He was a scholar-official, diplomat, and celebrated literary figure of the late sixteenth-century Joseon Dynasty.

His literary reputation rested particularly on his mastery of classical prose and poetry. His collected writings were preserved in the Gani chip (簡易集).

The 1596 memorial is significant not only as a political document but also as an example of Choi Rip’s sophisticated classical prose and his ability to construct political arguments through extensive references to the Confucian classics.

The Annals of King Seonjo records the memorial and reports that King Seonjo ordered the matter to be referred to the Border Defense Council (Pibyŏnsa) for discussion and recommendation (NIKH, 1596/2026).

The response demonstrates that Choi Rip’s argument was treated as a matter of substantive state policy rather than merely literary commentary.

The First Point — Heaven and the Ruler’s Self-Cultivation

Choi Rip’s first question concerns the relationship between Heaven and political leadership:

“何謂天之欲禍, 未欲禍也? / 하위천지욕화, 미욕화야?”

What distinguishes a situation in which Heaven intends calamity from one in which it does not?

Choi Rip draws upon classical Confucian texts to connect the fate of the state with the moral condition of political leadership.

At the center of his argument is the concept of 慎獨 (sindok), moral vigilance when one is alone.

He writes:

“先儒以謹獨一事, 爲事天之事 / 선유이신독일사, 위사천지사”

Earlier Confucian scholars regarded careful self-discipline when alone as a way of serving Heaven (NIKH, 1596/2026).

The political implication is clear:

Political reform begins with the ruler’s own moral self-examination.

The state cannot be repaired merely through military mobilization or administrative commands. The ruler must first discipline himself.

Although this idea may be compared metaphorically with modern concepts such as leadership integrity or ethical self-audit, Choi Rip’s original framework remained firmly grounded in Confucian theories of Heaven, kingship, and self-cultivation.

The Second Point — The State: Overcoming Factionalism

The second point is the most striking part of Choi Rip’s memorial.

He discusses historical examples and argues that when evaluating another state, one should not look only at its soldiers, provisions, or fortifications. One should also examine whether the ruler and ministers are harmonious.

He then turns directly to the Joseon court:

“今日朝廷之下, 得無有偏黨, 不相和睦之弊 / 금일조정지하, 득무유편당, 불상화목지폐”

“Is there not, within today’s royal court, the problem of factionalism and mutual discord?” (NIKH, 1596/2026).

This statement provides direct evidence that Choi Rip perceived 偏黨 (pyeondang), factional division, as a serious political problem.

He strengthens his argument through classical sources.

From the Analects, he invokes Confucius’ warning:

“吾恐季孫之憂, 不在顓臾而在蕭墻之內也 / 오공계손지우, 불재전유이재소장지내야”

The danger may not lie outside the walls; it may lie within them.

The metaphor of 蕭墻之憂 (소장지우) therefore becomes particularly powerful in the context of wartime Joseon.

The country was under attack from abroad, yet Choi Rip warned that internal division could also be destructive.

He then turns to the Yijing, citing the fourth line of the Huan hexagram:

“渙其群, 元吉 / 환기군, 원길”

and drawing upon Zhu Xi’s interpretation:

“散其朋黨 / 산기붕당”

— dispersing factional associations.

This is the textual foundation for Choi Rip’s call to overcome factionalism (NIKH, 1596/2026).

Factionalism as a National-Security Problem

Choi Rip did not treat factionalism merely as a matter of personal disagreement.

His deeper argument was that:

political fragmentation weakens the state’s ability to respond to external threats.

This interpretation is consistent with modern scholarship on factional politics in Joseon. Political factionalism emerged during the sixteenth century as competing groups of scholar-officials struggled over political influence, legitimacy, and policy.

Lee (2012), for example, examines the causes and development of factional conflict during King Seonjo’s reign and situates the conflict within the transformation of Joseon’s political order.

Jackson (2026) likewise identifies political factionalism as a major feature of Joseon political life that influenced royal decision-making over an extended period.

Choi Rip’s argument is therefore remarkable because he connected factional division directly with the question of national survival during wartime.

For him:

Political fragmentation was not merely an administrative inconvenience. It was a strategic vulnerability.

Who Was Responsible for Ending Factionalism?

Choi Rip does not simply identify the problem. He assigns responsibility.

After discussing the relevant classical texts, he states:

“是皆責在大臣者 / 시개책재대신자”

The responsibility lies particularly with the senior ministers (NIKH, 1596/2026).

This is an important element of Choi Rip’s political philosophy.

He did not simply ask the king to command everyone to cooperate. He placed particular responsibility upon the political leadership.

Senior officials had to rise above factional interests and restore a sense of common purpose.

This may resemble a modern theory of political accountability when viewed retrospectively. Historically, however, Choi Rip’s position remained firmly within the Confucian political order.

He was not proposing competitive party politics or a modern pluralistic parliamentary system.

Rather, he wanted officials to subordinate factional interests to the moral and political interests of the state.

The Third Point — Ministers Must Speak Truth to Power

Choi Rip’s third question concerns the responsibilities of ministers:

“何謂臣之毋以亡道事之也? / 하위신지무이망도사지야?”

What does it mean for a minister not to serve the ruler in a way that leads him toward destruction?

His answer centers on remonstrance (ganjaeng, 諫諍).

A loyal minister must tell the ruler what he needs to hear rather than merely what he wants to hear.

Choi Rip invokes the traditional maxim:

“國將興, 必賞諫臣; 國將亡, 必殺謀臣 / 국장흥, 필상간신; 국장망, 필살모신”

When a state is about to prosper, it rewards those who remonstrate; when a state is about to perish, it destroys those who give counsel.

The principle is straightforward:

A government that cannot hear criticism cannot correct its mistakes.

This becomes especially important in wartime, when a single erroneous decision can produce enormous consequences.

The Danger of “面從退言”

Choi Rip identifies a particularly dangerous political habit:

“面從退言之習 / 면종퇴언지습”

the habit of agreeing with the ruler in his presence but speaking differently after leaving him.

The problem is not merely hypocrisy.

If officials remain silent before the ruler, the ruler cannot receive accurate information. Errors remain uncorrected. Policies continue despite their weaknesses. The political system loses its capacity for self-correction.

Choi Rip therefore regards honest remonstrance as an essential component of good government.

The loyal minister is not the minister who always agrees.

The loyal minister is the one who speaks when speaking is necessary.

Remonstrance and National Security

This argument becomes even more significant in the context of the Imjin War.

During a national emergency, the quality of information reaching political leaders can determine the outcome of policy.

If leaders suppress unpleasant information, officials will learn to conceal problems. If officials fear punishment for disagreement, political institutions become informationally blind.

Choi Rip’s criticism of 面從退言 can therefore be read as an early Confucian reflection on the relationship among truthful communication, political accountability, and national security.

Lee’s (2012) analysis of the political structure during King Seonjo’s reign is relevant here because it highlights the political tensions and institutional dynamics that characterized this period.

Choi Rip’s solution was not institutional engineering in the modern sense.

His solution was ethical:

leaders must create conditions in which truthful counsel can be heard.

The Fourth Point — The People and Political Trust

The fourth question concerns the people:

“何謂民之有與之俱存之心也? / 하위민지유여지구존지심야?”

What does it mean for the people to possess the will to exist together with the state?

He again turns to the Yijing, citing the T’ai hexagram:

“說以先民, 民忘其勞; 說以犯難, 民忘其死. / 열이선민, 민망기노; 열이범난, 민망기사.”

When the people willingly follow a leader into hardship, they can forget their labor; when they willingly face danger, they can even forget death (NIKH, 1596/2026).

Choi Rip connects this classical principle with the examples of Yu and King Tang.

The people endured enormous burdens because they believed that their leaders were acting for the common good.

This leads to one of the most important arguments in the memorial:

A government cannot demand sacrifice indefinitely without maintaining public trust.

The People under the Burden of War

Choi Rip applies this principle directly to the conditions of his own time.

He describes the burdens placed upon the population, including:

• constructing and repairing fortifications;

• defending coastal areas;

• training soldiers;

• transporting military supplies; and

• performing various forms of wartime labor.

Such policies could be necessary for national defense. But when the people no longer believed that the government could protect them, necessary mobilization could be experienced as exploitation.

Choi Rip therefore recognizes an important distinction:

The same sacrifice can be understood either as patriotic service or as oppressive labor, depending upon the level of trust between government and people.

This insight is remarkably relevant to the study of wartime governance.

National Security Begins with Public Trust

For Choi Rip, public sentiment is not merely a moral ideal.

It is a practical component of national power.

When people trust the government, they are more willing to:

• transport provisions;

• construct fortifications;

• participate in military training;

• accept taxation and labor obligations; and

• endure the hardships of war.

When that trust collapses, the same policies become sources of resentment and resistance.

Choi Rip’s argument can therefore be represented as a cycle:

Responsible governance → public trust → collective participation → national mobilization → stronger national defense

The opposite cycle is equally dangerous:

Political factionalism → policy distrust → popular alienation → weakened mobilization → national vulnerability

In this sense, Choi Rip’s memorial can be interpreted not merely as a statement of Confucian benevolence but also as an early reflection on state capacity during national crisis.

The Four Points as an Integrated Political Philosophy

At the conclusion of the memorial, Choi Rip explicitly refers to:

“凡臣所列四條 / 범신소열사조”

— “the four points that I have presented” (NIKH, 1596/2026).

These four points can be summarized as follows:

Domain Choi Rip’s Core Principle Central Question Political Function

Heaven 慎獨 self-discipline How should the ruler govern himself? Moral foundation of leadership

State 散朋黨 overcoming factionalism How can political division be overcome?Political cohesion

Ministers 直諫 honest remonstrance How should officials correct the ruler?Political self-correction

People 得民心 winning popular trust Why should people sacrifice for the state? Social mobilization

These four points are not independent.

The ruler’s self-discipline produces moral credibility.

Political unity produces coherent government.

Honest remonstrance enables correction of mistakes.

Public trust transforms government policy into collective action.

Thus, Choi Rip’s political philosophy can be understood as a chain:

Moral leadership → political cohesion → truthful communication → public trust → national survival

Choi Rip and Yi I: Convergence and Difference

Choi Rip’s argument can be compared with the political thought of Yi I (李珥), another major Confucian thinker of the Joseon period.

Modern scholarship has examined Yi I’s efforts to overcome political factionalism through concepts of political reconciliation and institutional reform (Lee, 2012).

Choi Rip shared Yi I’s concern that political division threatened effective government. Yet the circumstances were different.

Yi I addressed the developing political tensions of the sixteenth-century court, whereas Choi Rip wrote his memorial amid the continuing crisis of the Imjin War.

For Choi Rip, therefore, factionalism had become inseparable from the question of national survival.

His use of “渙其群 (환기군)” and “散其朋黨 (산기붕당)” was not simply a call for people to get along.

It was a demand that political actors recognize a larger common purpose when the survival of the state itself was at stake.

Choi Rip’s Critique Should Not Be Equated with Modern Anti-Party Politics

There is an important historical qualification.

Choi Rip’s call to “散其朋黨 (산기붕당)” should not be interpreted as a call for the abolition of political parties in the modern democratic sense.

Modern democratic political parties serve important functions by organizing competing interests, representing citizens, and institutionalizing political disagreement.

Choi Rip lived in a fundamentally different political order.

His target was not political pluralism as such but factional loyalty that subordinated public interests to private group interests.

The historical question he raises nevertheless remains relevant:

What happens when political groups become more committed to defeating one another than to solving the problems of the state?

This question transcends the political institutions of any single historical period.

Choi Rip and the Philosophy of Political Self-Correction

Perhaps the most interesting feature of Choi Rip’s memorial is that he does not present a single policy solution.

Instead, he emphasizes the capacity of government to recognize and correct its own mistakes.

The ruler must examine himself.

Senior ministers must rise above factional interests.

Officials must speak honestly.

The people must have confidence that government serves their welfare.

Taken together, these principles create a political system capable of self-correction.

This can be compared analytically with contemporary concepts such as:

accountability, dissent, feedback, transparency, and institutional trust.

Such comparisons should be understood as analytical analogies rather than claims that Choi Rip anticipated modern political science.

His intellectual framework remained Confucian.

Nevertheless, the underlying problem is strikingly familiar:

How can a political system prevent itself from becoming blind to its own failures?

Re-reading Choi Rip in the Age of Artificial Intelligence

Choi Rip’s memorial is particularly interesting in the age of artificial intelligence because AI systems have intensified questions about information, authority, decision-making, and trust.

Powerful AI does not automatically produce wise decisions.

If leaders receive biased information, if organizations become trapped in informational echo chambers, if dissenting experts are ignored, or if citizens do not trust AI-assisted government, technological capacity may actually amplify existing weaknesses.

Choi Rip’s four principles can therefore be used as a modern interpretive framework, without claiming that he anticipated AI.

慎獨 / 신독 — Leadership Integrity

Leaders should critically examine AI-generated information rather than accepting it automatically.

散朋黨 / 산붕당 — Breaking Information Echo Chambers

Organizations should prevent factional or algorithmic information bubbles from determining policy.

直諫 / 직간 — Institutionalized Dissent

Human experts should be able to challenge AI-generated recommendations, especially when those recommendations affect public safety or national security.

得民心 / 득민심 — Human Trust

AI-based government must earn public confidence. Technical efficiency alone cannot produce legitimate governance.

These applications are contemporary interpretations rather than historical claims about Choi Rip.

Conclusion — The Enemy Within

In 1596, while the Imjin War continued, Choi Rip did not simply tell King Seonjo to build more fortifications or recruit more soldiers.

He asked a more fundamental question:

Why does a state become weak?

His answer had four dimensions.

When the ruler fails to examine himself,

when the court becomes divided by faction,

when ministers are afraid to speak truthfully,

and when the people lose trust in the government,

the state becomes vulnerable even before the enemy breaches its borders.

This is the central message of Choi Rip’s memorial.

His reference to “渙其群 (환기군)” and “散其朋黨 (산기붕당)” was therefore much more than an appeal for political harmony. It was a question about whether a nation could still act as a unified political community.

His criticism of 面從退言 (면종퇴언) was more than a complaint about hypocrisy. It was a question about whether those in power could hear uncomfortable truths.

And his emphasis on the people was more than a conventional expression of Confucian benevolence. It was an argument that public trust can be transformed into real national capacity during a crisis.

Nearly four and a half centuries later, the question remains remarkably powerful:

The strongest state is not necessarily the state with the most weapons. It is the state capable of overcoming internal division, hearing uncomfortable truths, correcting its own mistakes, and maintaining trust between government and people.

In the age of artificial intelligence, that lesson may be more important than ever.

The most dangerous enemy of a nation may not always be waiting beyond its borders. Sometimes it begins with the loss of trust within them.

References

Jackson, A. D. (2026). Political factionalism in Chosŏn Korea. In A. D. Jackson, S. Nguyen, & E. Richards (Eds.), Snapshots and soundbites of Korea (pp. 250–253). Routledge. https://doi.org/10.4324/9781003591689-100

Lee, J.-C. (2012). 선조 대 당쟁의 원인과 전개양상—이이를 중심으로 [The causes and development of factional strife during King Seonjo’s reign: Focusing on Yi I]. 藏書閣, 28, 72–107. https://doi.org/10.25024/jsg.2012..28.72

Liu, S. J. (2019). Su Shi’s (蘇軾) influence on Choi Yip’s (崔岦) poetry. Journal of Chinese Language and Literature, 81, 87–112. https://doi.org/10.15792/clsyn..81.201908.87

National Institute of Korean History. (1596/2026). The Veritable Records of King Seonjo, Vol. 74, Seonjo 29, fourth month, fifth day (辛丑), third article: Memorial of Haenghogun Choi Rip (崔岦). Annals of the Joseon Dynasty. Annals of King Seonjo — Choi Rip’s Memorial

Song, J.-H. (2017). 왜란·호란 양란기 구천(句踐) 담론과 재건의 정치 [The discourse of Goujian and the politics of reconstruction during the Imjin and Manchu Wars]. 한국동양정치사상사연구, 16(1), 91–122.

한국학중앙연구원. (n.d.). 최립(崔岦) [Choi Rip]. 한국민족문화대백과사전. Korean Historical Dictionary — Choi Rip

최립. (1631). 간이집(簡易集) [Collected Works of Gani]. Initial woodblock edition.

인공지능 시대의 간이 최립(崔岦)의 파당(派黨) 해소를 위한 네 가지 조목의 상소(上疏) 고찰

임진왜란기 조선의 정치적 통합, 직간, 그리고 민심

초록

본고는 간이(簡易) 최립(崔岦, 1539–1612)이 임진왜란이 계속되던 선조 29년(1596)에 선조 임금에게 올린 네 가지 조목의 상소(上疏, 상소)를 중심으로 그의 정치사상을 고찰한다. 『선조실록(宣祖實錄)』에 수록된 이 상소에서 최립은 천(天), 국가(國家), 신하(臣下), 백성(民, 민)의 관계를 네 가지 조목으로 나누어 논하면서, 국가적 위기를 극복하기 위해서는 군주의 자기수양, 조정의 파당 해소, 신하의 직간, 백성의 신뢰가 함께 이루어져야 한다고 주장하였다.

특히 두 번째 조목에서 최립은 당시 조정에 편당(偏黨)을 만들고 서로 화목하지 못하는 폐단이 있지 않은지를 직접 묻는다. 그는 『논어(論語)』와 『주역(周易)』 및 주희(朱熹)의 해석을 인용하면서 『주역』 환괘(渙卦)의 “환기군(渙其群)”을 “산기붕당(散其朋黨)”, 즉 붕당을 흩어 없애는 것으로 해석하였다. 이는 최립이 파당을 단순한 개인적 갈등이 아니라 국가의 정치적 결속과 존립을 위협할 수 있는 문제로 인식했음을 보여준다.

나머지 세 조목은 이러한 정치적 통합론을 보완한다. 첫째, 군주는 근독(謹獨)을 통해 홀로 있을 때에도 자신을 삼가고 수양해야 한다. 둘째, 신하는 임금 앞에서는 순종하고 뒤에서는 불평하는 면종퇴언(面從退言)의 폐단을 버리고 직간(直諫)해야 한다. 셋째, 백성은 국가가 자신들을 보호하고 있다는 믿음을 가질 때 국가의 어려움과 희생을 함께 감당할 수 있다.

따라서 최립의 네 가지 조목은 도덕적 지도력 → 정치적 결속 → 정직한 소통 → 백성의 신뢰 → 국가의 생존이라는 하나의 정치적 자기교정 체계를 형성한다. 본고는 이러한 최립의 주장을 현대 민주주의나 정당정치의 선구적 형태로 해석하기보다는 조선(朝鮮)의 유교적 군주정치라는 역사적 맥락에서 이해해야 한다고 본다. 그러나 정치적 분열을 극복하고, 권력자에게 불편한 진실을 전달하며, 백성의 신뢰를 국가역량으로 전환해야 한다는 그의 문제의식은 오늘날에도 중요한 의미를 갖는다.

핵심어: 최립(崔岦), 간이(簡易), 선조(宣祖), 조선(朝鮮), 붕당(朋黨), 파당(派黨), 직간(直諫), 민심(民心), 임진왜란(壬辰倭亂), 주역(周易), 유교 정치사상

전쟁 한가운데서 던진 정치의 질문

1596년 조선은 아직 임진왜란(壬辰倭亂)의 참혹한 충격에서 벗어나지 못하고 있었다. 전쟁은 전쟁터에만 존재한 것이 아니었다. 성을 쌓고, 군량을 운송하고, 군사를 동원하고, 군사훈련을 실시하고, 해안을 방어하는 모든 일이 백성의 삶을 무겁게 짓누르고 있었다.

바로 이러한 상황에서 최립은 국가의 생존에 관한 근본적인 질문을 던졌다.

그는 상소의 첫머리에서

“국가존망(國家存亡), 간불용발(間不容髮)”

이라고 하였다.

이는 “국가의 존망(存亡)이 머리카락 하나가 들어갈 틈도 없을 정도로 위급하다”는 뜻이다. 이어

“군신동우(君臣同憂)”

라고 하여 임금과 신하가 함께 국가의 위기를 걱정하며 해결책을 찾고 있다고 하였다.

그러나 최립은 단순히 더 많은 군사를 모집하고 성을 더 튼튼하게 만들자는 데서 그치지 않았다.

그는 더 근본적인 질문을 던졌다.

왜 국가는 외적의 공격에 취약해지는가?

그의 대답은 군사력에만 있지 않았다.

국가의 생존은 통치자의 도덕적 상태, 조정의 정치적 결속, 신하의 정직한 간언, 그리고 백성의 국가에 대한 신뢰에 달려 있다는 것이었다.

이러한 관점은 선조(宣祖)대 조선 정치의 상황과 함께 살펴볼 필요가 있다. 선조 초에는 사림(士林) 세력이 정치적 영향력을 확대하는 과정에서 동인(東人)과 서인(西人)의 분열이 발생하였다. 이러한 정치적 분열은 선조대 정치질서의 변화와 정치세력 간의 긴장 속에서 발전하였다.

따라서 최립의 1596년 상소는 두 가지 위기가 교차하는 지점에서 읽을 수 있다.

밖에서는 임진왜란이라는 국가적 위기가 진행되고 있었고, 안에서는 정치적 파당과 분열이라는 위기가 존재했다.

최립과 이 상소의 역사적 의미

최립은 1539년에 태어나 1612년에 세상을 떠났다. 자(字)는 입지(立之)이고 호(號)는 간이(簡易)이다. 그는 조선 중기의 문신(文臣)이자 외교관이며 뛰어난 문장가였다.

그의 문학적 명성은 특히 한문(漢文) 산문과 시에서 두드러졌다. 그의 저술은 후대에 『간이집(簡易集)』으로 정리되어 전해졌다.

1596년의 상소는 정치문서인 동시에 최립 특유의 고전적 문장과 논증 능력을 보여주는 중요한 산문이다.

최립은 단순히 자신의 의견을 주장하지 않았다. 그는 『논어(論語)』, 『주역(周易)』 등 유교 경전(經典)과 역사적 고사(故事)를 끌어와 당시의 현실을 분석하였다.

『선조실록(宣祖實錄)』은 이 상소를 기록하고, 선조는 그 내용을 비변사(備邊司)에 내려 논의하도록 하였다.

이는 최립의 주장이 단순한 문인의 개인적인 의견이 아니라 당시 국가정책의 차원에서 검토되었음을 보여준다.

첫 번째 조목 — 하늘과 군주의 자기수양

최립의 첫 번째 질문은 천(天)과 통치자의 관계에 관한 것이다.

그는 묻는다.

“하위천지욕화 미욕화야(何謂天之欲禍, 未欲禍也)”

즉, 하늘이 재앙을 내리려는 경우와 그렇지 않은 경우는 무엇인가라는 질문이다.

최립은 유교 경전을 통해 국가의 흥망(興亡)과 통치자의 도덕적 상태를 연결한다.

그 중심에 있는 개념이 바로 근독(謹獨)이다.

그는 다음과 같이 말한다.

“선유이근독일사 위사천지사(先儒以謹獨一事, 爲事天之事)”

즉, 선대 유학자들은 홀로 있을 때 자신을 삼가는 것을 하늘을 섬기는 일로 보았다는 뜻이다.

정치적 의미는 분명하다.

정치의 개혁은 군주 자신의 자기성찰에서 시작되어야 한다.

국가의 문제를 해결하기 위해서는 군사력이나 행정명령만으로는 부족하다. 통치자 자신이 먼저 자신의 마음과 행동을 살펴야 한다.

두 번째 조목 — 국가: 파당을 해소하라

최립의 상소에서 가장 주목해야 할 부분은 두 번째 조목이다.

그는 역사적 사례를 통해 국가의 힘을 평가할 때 단순히 군사의 수, 군량, 성곽만 보아서는 안 된다고 주장한다.

오히려 임금과 신하가 서로 화목한가를 살펴보아야 한다.

그리고 마침내 자신의 시대의 조정을 향해 직접 묻는다.

“금일조정지하 득무유편당 불상화목지폐(今日朝廷之下, 得無有偏黨, 不相和睦之弊)”

이는 “오늘날 조정에 편당을 짓고 서로 화목하지 못하는 폐단이 있지 않은가?”라는 뜻이다.

여기서 ‘조정(朝廷)’은 왕과 신하들이 국가의 정사를 논의하는 정치공간을 뜻한다.

‘편당(偏黨)’은 특정한 정치집단이나 사적인 세력에 치우치는 것을 의미한다.

최립은 이어 『논어(論語)』의 유명한 구절을 인용한다.

“오공계손지우 불재전유 이재소장지내야(吾恐季孫之憂, 不在顓臾而在蕭墻之內也)”

뜻은 “나는 계손(季孫)의 근심이 전유(顓臾)에 있는 것이 아니라 도리어 자기 집 담장 안에 있을까 두렵다”는 것이다. (여기서 ‘전유(顓臾)’는 계손씨가 공격하려고 했던 외부의 작은 나라를 가리킵니다. 반면 ‘소장(蕭牆)’은 궁궐이나 집 안쪽의 담장을 뜻합니다.)

이 말은 외부의 적보다 내부의 분열이 더욱 위험할 수 있다는 정치적 비유로 읽을 수 있다.

전쟁 중인 조선에서 이 말은 특별한 의미를 갖는다.

외부에는 일본군이라는 명백한 적이 있었지만, 최립은 내부의 정치적 분열 역시 국가를 위태롭게 만들 수 있다고 경고하였다.

“환기군”과 “산기붕당”

최립은 이어 『주역(周易)』 환괘(渙卦)의 구절을 인용한다.

“환기군 원길(渙其群, 元吉)”

여기에서 환(渙)은 흩어지다, 풀어지다라는 뜻이며, 군(群)은 무리 또는 집단을 뜻한다.

최립은 주희(朱熹)의 해석을 통하여 이를

“산기붕당(散其朋黨)”

이라고 설명한다.

즉,

“사적인 붕당(朋黨)을 흩어 없애야 한다”

는 의미이다.

이것이 최립의 파당 해소론에서 가장 중요한 경전적 근거 가운데 하나이다.

파당은 국가안보의 문제였다

최립은 파당을 단순한 인간관계의 불화로 보지 않았다.

그가 더 깊이 우려한 것은 정치적 분열이 외적에 대한 국가의 대응력을 약화시킬 수 있다는 사실이었다.

여기에서 중요한 것은 ‘붕당(朋黨)’과 현대적 의미의 ‘정당(政黨)’을 동일시하지 않는 것이다.

최립이 문제 삼은 것은 특정 집단에 대한 정치적 충성이 국가 전체의 공익(公益)보다 앞서는 상황이었다.

따라서 그의 논리는 다음과 같이 요약할 수 있다.

외부의 적이 강해서 국가가 약해지는 것만이 아니다.

내부의 분열 때문에 국가가 약해질 수도 있다.

이러한 의미에서 최립에게 정치적 화합은 단순한 도덕적 미덕이 아니라 국가안보(國家安保)의 조건이었다.

파당을 없애는 책임은 누구에게 있는가

최립은 문제를 지적하는 데 그치지 않고 책임의 주체까지 명확하게 제시한다.

그는 다음과 같이 말한다.

“시개책재대신자(是皆責在大臣者)”

즉, “이 모든 책임은 대신(大臣)에게 있다”는 뜻이다.

이 대목은 최립의 정치사상에서 매우 중요하다.

그는 단순히 임금에게 모든 사람에게 화합하라고 명령하라고 요구하지 않았다.

오히려 고위 정치지도자들이 먼저 책임을 져야 한다고 보았다.

대신들은 당파적 이해관계를 넘어 국가 전체의 이익을 바라보아야 했다.

이것은 현대적 의미의 정치적 책임성(政治的責任性)과 비교할 수 있는 측면이 있지만, 최립의 사상은 현대 민주주의의 책임정치가 아니었다.

그는 공익을 위해 사적인 당파적 이해를 내려놓아야 한다는 유교적 정치윤리를 주장한 것이다.

세 번째 조목 — 신하는 권력자에게 진실을 말해야 한다

최립의 세 번째 질문은 신하(臣下)의 역할에 관한 것이다.

그는 묻는다.

“하위신지무이망도사지야(何謂臣之毋以亡道事之也)”

이는 “신하가 임금을 망하는 길로 섬기지 않는다는 것은 무엇인가?”라는 질문이다.

최립이 강조하는 것은 간언(諫言)과 직간(直諫)이다.

충성스러운 신하는 임금이 듣고 싶어 하는 말이 아니라 임금에게 필요한 말을 해야 한다.

그는 다음과 같은 전통적인 정치론을 인용한다.

“국장흥 필상간신 국장망 필살모신(國將興, 必賞諫臣; 國將亡, 必殺謀臣)”

즉,

“나라가 장차 흥하려 하면 반드시 간언하는 신하를 포상하고, 나라가 장차 망하려 하면 반드시 계책을 내는 신하를 죽인다.”

는 뜻이다.

최립이 강조하는 원리는 명확하다.

비판을 들을 수 없는 정부는 자신의 잘못을 고칠 수 없다.

“면종퇴언”의 위험

최립은 당시 정치에서 매우 위험한 습관을 지적한다.

“면종퇴언지습(面從退言之習)”

‘면종(面從)’은 사람의 면전에서는 순종하는 것을 말하고, ‘퇴언(退言)’은 그 사람이 물러간 뒤 다른 말을 하는 것을 뜻한다.

따라서 ‘면종퇴언지습(面從退言之習)’은

“임금 앞에서는 순종하는 척하면서 물러난 뒤에는 다른 말을 하는 습관”

을 의미한다.

이것은 단순한 위선의 문제가 아니다.

신하들이 임금 앞에서 사실을 말하지 않으면 임금은 정확한 정보를 얻을 수 없다.

잘못된 정책이 계속된다.

정책의 문제점이 교정되지 않는다.

결국 정치체제 전체가 스스로의 오류를 수정할 능력을 잃게 된다.

따라서 최립에게 ‘직간(直諫)’은 단순한 신하 개인의 용기가 아니었다.

그것은 국가의 자기교정 능력을 유지하기 위한 필수 조건이었다.

충성스러운 신하는 항상 동의하는 신하가 아니다.

필요한 순간에 진실을 말하는 신하가 진정으로 충성스러운 신하다.

네 번째 조목 — 백성과 정치적 신뢰

네 번째 질문은 민(民), 즉 백성에 관한 것이다.

최립은 묻는다.

“하위민지유여지구존지심야(何謂民之有與之俱存之心也)”

이는 “백성이 국가와 함께 존재하고자 하는 마음을 갖게 한다는 것은 무엇인가?”라는 뜻이다.

그는 다시 『주역(周易)』의 원리를 인용한다.

“열이선민 민망기노 열이범난 민망기사(說以先民, 民忘其勞; 說以犯難, 民忘其死)”

여기에서 ‘說 (열)’은 기쁘게 하다 또는 기꺼이 따르게 하다는 뜻이다.

전체적인 의미는 지도자가 백성을 앞세우고 함께 어려움을 감당하면 백성은 자신의 수고를 잊으며, 함께 위험을 무릅쓰면 죽음까지도 잊는다는 것이다.

최립은 이를 우(禹)와 탕(湯)의 역사적 사례와 연결한다.

백성은 엄청난 노동과 희생을 감당했지만, 지도자가 자신들을 위해 행동한다는 믿음이 있었기 때문에 그 고통을 견딜 수 있었다.

전쟁의 부담을 짊어진 백성

최립은 이 원리를 자신의 현실에 적용한다.

그는 당시 백성들이 부담하고 있던 여러 가지 일을 열거한다.

성곽(城郭)의 축조와 보수, 해안 방어, 군사훈련, 군량(軍糧)과 군수물자의 운송, 각종 전쟁 관련 부역(賦役) 등이 그것이다.

이러한 정책 자체는 국가방어를 위해 필요할 수 있었다.

그러나 백성이 국가가 자신들을 보호할 것이라고 믿지 못한다면, 같은 정책도 애국적 봉사가 아니라 억압적인 부역으로 느껴질 수 있다.

최립은 바로 이 차이를 포착한다.

동일한 희생도 정부와 백성 사이의 신뢰 수준에 따라 애국적 봉사가 될 수도 있고 강제적인 고통이 될 수도 있다.

국가안보는 민심에서 시작된다

최립에게 민심(民心)은 단순한 도덕적 가치가 아니었다.

그것은 실제적인 국가역량(國家力量)의 기반이었다.

백성이 정부를 신뢰하면 그들은 군량을 운반하고, 성을 쌓고, 군사훈련에 참여하며, 전쟁의 고통을 함께 견딜 수 있다.

반대로 신뢰가 무너지면 같은 정책이 원망과 저항을 낳는다.

최립의 논리는 다음과 같은 순환구조로 정리할 수 있다.

책임 있는 통치 → 백성의 신뢰 → 공동체적 참여 → 국가동원력 → 강한 국방력

반대 방향의 악순환도 가능하다.

정치적 파당 → 정책에 대한 불신 → 민심 이탈 → 국가동원력 약화 → 외적에 대한 취약성

이러한 의미에서 최립의 상소는 단순한 애민사상(愛民思想)을 넘어 국가적 위기에서 국가역량이 어떻게 형성되는가에 대한 역사적 성찰로 읽을 수 있다.

네 가지 조목은 하나의 정치철학이었다

상소의 마지막 부분에서 최립은 명확하게 다음과 같이 말한다.

“범신소열사조(凡臣所列四條)”

즉 “신이 열거한 네 가지 조목”이라는 뜻이다.

이를 정리하면 다음과 같다.

영역핵심 원리중심 질문 정치적 기능

천(天, 천) 근독(謹獨) 군주는 어떻게 자신을 다스려야 하는가? 통치의 도덕적 기반

국(國, 국) 산붕당(散朋黨) 정치적 분열을 어떻게 극복할 것인가? 정치적 결속

신(臣, 신) 직간(直諫) 신하는 어떻게 임금의 잘못을 바로잡을 정치적자기교정 것인가?

민(民, 민) 민심(民心) 백성은 왜 국가를 위해 희생해야 하는가? 사회적 동원력

이 네 가지는 서로 독립되어 있지 않다.

군주의 자기수양은 도덕적 신뢰를 만든다.

정치적 화합은 일관된 국정운영을 가능하게 한다.

신하의 직간은 정책의 오류를 수정한다.

백성의 신뢰는 국가의 정책을 실제 공동행동으로 전환한다.

따라서 최립의 정치철학은 다음과 같은 연쇄구조로 이해할 수 있다.

도덕적 지도력 → 정치적 결속 → 정직한 소통 → 백성의 신뢰 → 국가의 생존

최립과 이이 — 공통점과 차이

최립의 정치사상은 같은 시대의 대표적인 유학자 율곡 이이(栗谷 李珥)의 정치적 문제의식과 비교하면 더욱 선명해진다.

이이 역시 정치적 분열을 극복하고 정치적 조화를 이루어야 한다고 주장하였다.

최립 역시 정치적 분열이 국정 운영을 위협한다고 보았다는 점에서 이이와 공통점이 있다.

그러나 두 사람의 문제의식에는 차이가 있다.

이이가 선조대 정치적 갈등이 심화되는 상황에서 정치적 조화를 모색했다면, 최립은 임진왜란이라는 전쟁의 한복판에서 정치적 분열을 바라보았다.

따라서 최립에게 파당 문제는 단순한 정치적 비효율이 아니라 국가 존망(國家存亡)의 문제로 더욱 직접적으로 다가왔다.

그가 『주역』의

“환기군(渙其群)”

과

“산기붕당(散其朋黨)”

을 강조한 이유도 여기에 있다.

최립의 파당론을 현대의 반정당주의와 혼동해서는 안 된다

여기에는 중요한 역사적 주의가 필요하다.

최립의 산기붕당(散其朋黨)을 오늘날의 정당 해산이나 정치적 반대파 제거와 동일하게 이해해서는 안 된다. 현대 민주주의에서 정당(政黨)은 서로 다른 정책과 이해관계를 조직하고 시민의 의견을 정치과정에 반영하는 중요한 제도이다.

최립이 살았던 정치체제는 근본적으로 달랐다. 그가 비판한 것은 정치적 다양성 그 자체가 아니라,

공적 국가이익보다 사적인 당파적 이해를 앞세우는 정치적 파벌주의였다.

따라서 최립이 던진 질문은 오늘날에도 다음과 같은 형태로 되살아날 수 있다.

정치집단이 국가의 문제를 해결하는 것보다 상대편을 이기는 데 더 많은 관심을 갖게 되면 국가에는 어떤 일이 발생하는가?

정치적 자기교정의 철학

최립의 상소에서 특히 주목할 부분은 그가 단 하나의 정책적 해법을 제시하지 않았다는 점이다.

그는 오히려 국가가 자신의 잘못을 발견하고 수정할 수 있는 능력을 강조했다.

군주는 스스로를 성찰해야 한다.

대신은 당파적 이해를 내려놓아야 한다.

신하는 정직하게 말해야 한다.

백성은 국가가 자신들을 위해 존재한다는 믿음을 가져야 한다.

이 네 가지가 결합되면 국가는 스스로를 교정할 수 있다.

이러한 관점은 오늘날의 책임성, 이견, 피드백, 투명성, 제도적 신뢰와 비교해 볼 수 있다.

물론 이는 최립이 현대 정치학을 예견했다는 뜻이 아니다.

그의 사상적 기반은 철저히 유교적이었다.

그러나 그가 던진 근본적인 질문은 오늘날에도 그대로 살아 있다.

정치체제가 자신의 실패를 스스로 발견하고 수정하지 못하게 되는 것을 어떻게 막을 것인가?

인공지능 시대에 다시 읽는 최립

최립의 상소가 오늘날 특별히 흥미로운 이유는 인공지능 시대에 정보, 권위, 의사결정, 신뢰의 문제가 더욱 중요해지고 있기 때문이다.

강력한 인공지능(AI)이 등장했다고 해서 현명한 의사결정이 자동으로 이루어지는 것은 아니다.

지도자가 편향된 정보를 받아들이거나, 조직이 정보의 ‘에코체임버(echo chamber)’에 갇히거나, 전문가의 반대 의견이 무시되거나, 시민이 AI를 이용한 행정을 신뢰하지 않는다면 기술은 기존의 약점을 오히려 확대할 수 있다. (원래 echo chamber는 소리가 벽에 반사되어 같은 소리가 반복해서 들리는 공간을 뜻합니다. 이를 사회·정치·미디어 환경에 비유하여, 자신과 비슷한 생각이나 정보만 반복적으로 접하면서, 자신의 기존 생각이 더욱 강화되는 현상을 의미)

따라서 최립의 네 가지 원리를 오늘날의 관점에서 해석적 틀로 활용할 수 있다.

근독(謹獨) — 지도자의 자기성찰과 AI 윤리

지도자는 AI가 생성한 정보를 무비판적으로 받아들여서는 안 된다.

AI의 판단을 다시 검토하고 자신의 판단과 책임을 성찰해야 한다.

산붕당(散朋黨) — 정보 에코체임버의 해체

조직은 특정 집단이나 알고리즘이 만들어내는 정보 편향에 갇히지 않도록 해야 한다.

서로 다른 관점과 독립적인 검증을 보장해야 한다.

직간(直諫) — 제도화된 이견

AI가 제시한 권고에 대해 인간 전문가가 자유롭게 반론을 제기할 수 있어야 한다.

특히 국가안보와 공공안전에 관련된 AI 의사결정에서는 이러한 이견 제도가 중요하다.

득민심(得民心) — 인간의 신뢰

AI 기반 행정과 정책은 시민의 신뢰를 얻어야 한다.

기술적 효율성만으로는 정당한 통치를 만들어낼 수 없다.

이러한 해석은 현대적 적용이며, 최립이 AI 시대를 예견했다는 역사적 주장은 아니다.

결론 — 가장 위험한 적은 내부에서 시작될 수 있다

1596년 임진왜란이 계속되는 상황에서 최립은 단순히 선조에게 더 많은 성을 쌓고 더 많은 군사를 모집하라고 말하지 않았다.

그는 더욱 근본적인 질문을 던졌다.

국가는 왜 약해지는가?

그의 답은 네 가지였다.

군주가 스스로를 성찰하지 않을 때,

조정이 파당으로 갈라질 때,

신하가 진실을 말하기 두려워할 때,

백성이 국가를 신뢰하지 않을 때,

국가는 외적이 국경을 넘어오기 전부터 이미 약해지기 시작한다.

이것이 최립 상소의 핵심 메시지다.

그가 ‘환기군(渙其群)’과 ‘산기붕당(散其朋黨)’을 강조한 것은 단순히 정치인들이 사이좋게 지내라는 뜻이 아니었다.

그것은 국가가 하나의 공동체로 행동할 수 있는가라는 질문이었다.

그가 ‘면종퇴언(面從退言)’을 비판한 것도 단순히 위선을 비난한 것이 아니었다.

그것은 권력자가 불편한 진실을 들을 수 있는가라는 질문이었다.

그리고 그가 ‘민심(民心)’을 강조한 것도 단순한 애민사상이 아니었다.

그것은 위기 상황에서 백성의 신뢰가 실제 국가역량으로 전환되는가에 대한 질문이었다.

약 430년이 지난 오늘날에도 최립의 질문은 강력하다.

강한 국가는 반드시 가장 많은 무기를 가진 국가가 아니다. 내부의 분열을 극복하고, 불편한 진실을 들으며, 자신의 오류를 수정하고, 정부와 국민 사이의 신뢰를 지킬 수 있는 국가가 진정으로 강한 국가다.

특히 인공지능 시대에는 이 메시지가 더욱 중요해질 수 있다.

AI가 아무리 강력해져도 분열된 인간 사회, 비판을 받아들이지 못하는 권력, 그리고 신뢰를 잃은 정부를 대신하여 국가를 구해줄 수는 없다.

결국 최립의 오래된 상소가 오늘 우리에게 던지는 가장 강력한 질문은 이것이다.

“국가를 무너뜨리는 가장 위험한 적은 과연 국경 밖에만 있는가?”

참고문헌

국사편찬위원회. (1596/2026). 『선조실록(宣祖實錄, 선조실록)』 권74, 선조 29년 4월 5일 신축(辛丑, 신축) 3번째 기사: 행호군 최립(崔岦, 최립)이 네 가지 조목으로 상소하다. 『조선왕조실록(朝鮮王朝實錄, 조선왕조실록)』. 『선조실록』—최립의 상소 원문

이정철. (2012). 선조 대 당쟁의 원인과 전개양상—이이를 중심으로. 『장서각(藏書閣, 장서각)』, 28, 72–107. https://doi.org/10.25024/jsg.2012..28.72

송재혁. (2017). 왜란·호란 양란기 구천(句踐, 구천) 담론과 재건의 정치. 『한국동양정치사상사연구(韓國東洋政治思想史硏究, 한국동양정치사상사연구)』, 16(1), 91–122.

Liu, S. J. (2019). Su Shi’s (蘇軾, 소식) influence on Choi Yip’s (崔岦, 최립) poetry. Journal of Chinese Language and Literature, 81, 87–112. https://doi.org/10.15792/clsyn..81.201908.87

Jackson, A. D. (2026). Political factionalism in Chosŏn Korea. In A. D. Jackson, S. Nguyen, & E. Richards (Eds.), Snapshots and soundbites of Korea (pp. 250–253). Routledge. https://doi.org/10.4324/9781003591689-100

최립. 『간이집(簡易集, 간이집)』. 초간 목판본.

한국학중앙연구원. 최립(崔岦, 최립). 『한국민족문화대백과사전(韓國民族文化大百科事典, 한국민족문화대백과사전)』. 한국민족문화대백과사전—최립

2026년 8월 14일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

Choi, Y. B. (2026, August 14). Choi Rip’s Four-Point Memorial on Abolishing Factionalism in the Age of Artificial Intelligence: Political Unity, Remonstrance, and Popular Trust in Joseon Korea during the Imjin War. K-GSP.

Cover image: Finding My Voice by Marie Myung-Ok Lee. Image reproduced for purposes of review and commentary.

Finding My Voice is the debut novel of Korean American writer Marie Myung-Ok Lee (originally published under the name Marie G. Lee). Originally published in 1992, it is recognized as one of the first novels by a major American publisher written by a Korean American author featuring a Korean American teenage protagonist. While the novel is accessible to and aimed at young adult readers, it deals with serious issues of racism, identity, assimilation, family expectations, and the search for self. Much of the novel draws on Lee’s own experiences growing up as the only Korean American student in a small town in Minnesota. It is a coming-of-age novel that not only explores the struggles of adolescence but also the psychological costs of being perceived as a perpetual foreigner in one’s own country.

Marie Myung-Ok Lee grew up in Hibbing, Minnesota, a small mining town. Both of her parents fled North Korea before eventually immigrating to the United States. After finishing high school, Lee attended Brown University, where she earned a Bachelor of Arts degree. While Finding My Voice was her first novel, she has written several other young adult works, including If It Hadn’t Been for Yoon Jun (1993), Saying Goodbye (1994), Necessary Roughness (1996), and F Is for Fabulous (1999). In 2020, Finding My Voice was reissued by Soho Press. The novel received the American Library Association’s “Best Books for Reluctant Young Adult Readers” recognition, among many other awards and honors. Lee has also published works beyond her young adult fiction. Her most recent novel, The Evening Hero, was published by Simon & Schuster in 2021. Lee has taught fiction writing at Yale University, served as a visiting lecturer in American Studies at Brown University, and been an adjunct professor at Columbia University. She is also a founder of the Asian American Writers’ Workshop (AAWW).

The novel focuses on seventeen-year-old Ellen Sung, a Korean American high school senior living in the small town of Arkin, Minnesota. Ellen’s family is the only Korean family in town, making her someone who was constantly noticed which made her painfully aware of her differences. As the story begins, Ellen is preparing for her senior year in high school, hoping that graduation will finally provide an escape from a place she has never felt accepted.

From almost the first page of the book, Ellen struggles with the tension between her Korean heritage and her desire to fit into American culture. For instance, her mother lovingly packs her lunch with traditional Korean food, like litchi nuts, and on the way to the bus stop Ellen surreptitiously slips them into a garbage can. She is continuously fearful that her classmates will make fun of her. Throughout the novel, Lee reveals how even ordinary acts become emotionally charged when one feels like an outsider.

Interestingly, unlike many earlier immigrant novels, Finding My Voice is not directly focused on immigration itself. Although Ellen’s parents experienced the typical linguistic and cultural transitions of first-generation immigrants, Ellen is a second-generation immigrant whose struggle is not learning a new language or finding opportunities to interact with American citizens. Her issue is how to negotiate two identities at the same time. She is in the difficult position of being neither completely Korean nor completely accepted as an American. The feeling of living between two worlds is a major part of her experience as an adolescent.

Perhaps the major plot of the story is Ellen’s developing relationship with Tomper Sandel, a popular athlete admired by almost everyone in school. It is difficult for Ellen to imagine someone like Tomper really liking her because of an internalized doubt about interracial relationships, yet Tomper genuinely likes her for who she is rather than relating to her as some exotic curiosity. However, the relationship demonstrates the prejudice and othering embedded in a small-town community. It becomes an opportunity for gossip and racial comments disguised as jokes. Lee uses this romance to expose the social boundaries that racism creates.

The romance with Tomper ultimately serves less as a fairytale relationship than as a component of Ellen’s emotional maturity. Although their relationship does help her gain confidence, the novel’s real resolution is Ellen’s increasing acceptance of herself. As the school year progresses, Ellen gradually becomes less willing to remain quiet when there is evident discrimination. Earlier, she ignored the racist remarks of schoolmates, but by the end, she recognizes that silence only reinforces prejudice. Learning to speak honestly about her experiences with prejudice reveals the meaning of the title Finding My Voice.

The relationship with Tomper generates several problems within Ellen’s family, and perhaps Tomper’s as well. As is often the case, her Korean family has done a lot to establish themselves within America. So, they place great value on education, hard work, and respect for Korean culture and traditions. Like many first-generation parents, they want their daughter to preserve her identity as a Korean and not to stray too far from the nest. Even though they love their daughter, they do not understand why she wants so much to be accepted within American society. Ellen often interprets her parents’ strictness and controlling communication as attempts to control her just as she begins to find her way. It is not as though either side is completely wrong in their wishes; they simply look at the world from different historical experiences and find it difficult to know how to make the other understand.

Lee makes it clear through Ellen’s experiences that her attempts to erase her Korean heritage will not make her whiter. At the same time, she no longer feels entirely Korean. It is as though she is split in two, so that at times she fits nowhere in particular. This position of trying to live in two worlds is a common theme in many Korean American novels. Like Suzy Park in Suki Kim’s The Interpreter or Jane Re in Patricia Park’s Re Jane, these characters discover that the struggle for a new life in America comes at a cost, and they must somehow find a balance between the two cultures. They must resolve to accept who they really are within these contrasting cultures to finally feel at home.

Lee offers a close look at everyday racism toward immigrants in America. Rather than portraying dramatic acts of violence or discrimination, she tells a story of persistent, repeated forms of prejudice that seem to pile on top of each other. Ellen’s schoolmates continually cross boundaries of appropriateness through their insensitive comments and casual racial remarks, which are initially ignored but over time become much more hurtful. Together, these comments and casual remarks create an unwelcome environment that makes Ellen feel it is necessary to continually defend who she is. Lee’s awareness of this form of racism became a precursor to later writers who would become more open about subject in their work.

Lee’s work is also about exploring the process of assimilation. What happens when an immigrant wishes to fit into American society but discovers that, no matter how hard one tries, some people are still going to see them as a foreigner? Trying to cover up Korean traditions, like eating litchi nuts, will not necessarily make a person acceptable to peers. Lee does not portray assimilation as either a preservation or rejection of Korean traditions. Eventually, Ellen learns that peer acceptance cannot be achieved by hiding her Korean identity. It is only when she accepts herself as she is that she begins to gain the confidence to defend her identity.

The novel’s title encapsulates Ellen’s struggle to find her place. Her developing voice symbolizes her self-acceptance. This means more than having confidence; it means a willingness to speak out when prejudicial attitudes and inappropriate behaviors are obvious. Through most of the novel, Ellen avoids confrontation but finally realizes that staying quiet only allows prejudice to thrive. Her emerging voice makes it possible for Ellen to embrace her real identity rather than having to choose between her Korean and American identities. In the end, she demonstrates that it is quite possible for her to accept both as parts of her authentic self.

Ultimately, Finding My Voice is important within the Korean American genre because it turns an individual coming-of-age story into a broader discussion of race, identity, and belonging in America. Lee shows that finding one’s voice is inseparable from authentically finding a place of belonging. Ellen’s journey toward self-worth illustrates that genuine belonging cannot be achieved by erasing culture or traditions. It only comes by embracing the sometimes-difficult situation of trying to live between cultures. This means accepting both parts as equally valid, without apology. Lee brings Ellen close to this more fulfilling place by the end of the story.

Lee’s work speaks powerfully to the question of how different cultures encounter one another and how these encounters are reflected in the lives of immigrants to America. Like so many other Korean American writers, Lee treats this issue as a deeply ethical concern. Her stories attempt to portray the unique position of the other, the stranger, or the outsider and to encourage readers to consider what it means to encounter someone whose experiences and cultural background differ from their own. This ethical concern recalls the work of the Jewish philosopher Emmanuel Levinas and his concept of “facing,” in which the encounter with the face of the other creates an ethical obligation to recognize that person’s uniqueness and respond to his or her appeal. Such an encounter also changes the position of the representative of the dominant culture. Rather than expecting the other to change or assimilate, it requires a willingness to listen, to acknowledge differences, and to accept responsibility for the ways in which one’s own actions affect those who are not the same as the dominant culture.

In this story of a young second-generation woman from an immigrant family, Lee creates a persuasive vehicle for understanding the nature of racism and the suffering that immigrants and their children can experience in their search for a means of belonging in America. In one sense, the story places the reader in the position of either welcoming or judging Ellen and her situation. We are forced to confront the assumptions and prejudices that shape her experiences and to recognize how deeply such attitudes can affect a young person’s developing sense of self. Ellen certainly has not solved every problem in her life, but she has made significant progress. Prejudice and racism continue to exist in her small town, and cultural conflicts remain within her family. Yet Ellen enters adulthood with greater confidence, a clearer understanding of her identity, and the courage to define herself on her own terms. Her growth suggests that finding one’s voice does not mean simply learning how to speak louder or resist those who cause pain. It means developing the confidence to acknowledge all the parts of oneself without allowing others to determine which parts are acceptable.

About the Author

C. Clark Triplett is Emeritus Dean of Graduate Studies and Professor of Psychology at Missouri Baptist University. He served as co-editor of The Final Crossing: Death and Dying in Literature (Peter Lang, 2015), a co-editor of Worlds Gone Awry: Essays on Dystopian Fiction (McFarland, 2018), a co-editor of Certainty and Ambiguity in Global Mystery Fiction: Essays on the Moral Imagination (Bloomsbury Academic, 2024), and co-editor of Echoes from the Hills: Critical Essays on Ozarks Literature (forthcoming, University of Arkansas Press). Triplett’s poems have appeared in Cantos, Fireflies’ Light, and the Asahi Haikuist Network. He earned a BA from Southwest Baptist University, an MDiv from Covenant Theological Seminary, an MSEd from Southern Illinois University at Edwardsville, and a PhD from Saint Louis University.

© K-Global Scholars and Professionals Forum. All rights reserved. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

This article examines David’s leadership transformation journey from kingdom builder to a man after God’s own heart, exploring how success, failure, repentance, and divine grace shaped his development as Israel’s most influential king. Extending the Legacy Trap Framework, it considers how a leader may move beyond self-protection, personal achievement, and the misuse of power through repentance, character formation, institutional stewardship, and responsible succession. Using the integrated 5Is–5Ps framework, the study analyzes David’s progression through kingdom building, leadership crises, personal restoration, covenant stewardship, and legacy formation.

Keywords: David; Leadership Transformation; Legacy Trap; Character Formation; Repentance and Restoration; Leadership Stewardship

1. Introduction

Success reveals what a leader can build; failure reveals who that leader has truly become. This article extends the Legacy Trap Framework (Hong, 2026) by shifting the focus from how successful leaders drift toward misuse of power and institutional decline to how leaders may respond to failure through repentance, character transformation, stewardship, and responsible succession. The first half of life is often devoted to discovering identity, acquiring skills, building relationships, pursuing opportunities, and preparing for greater responsibility, while the second half increasingly tests whether achievement can mature into transformation, stewardship, and legacy. The central questions therefore shift from “How do I succeed?” to “How do I respond when success exposes my weaknesses?” and from “How do I gain influence?” to “What will remain after failure, repentance, and restoration?” This developmental progression is consistent with Erikson’s (1982) understanding of identity formation and later-life integration, Bennis’s (2009) view of leadership as a lifelong process of becoming, and Collins’s (2026) emphasis on self-knowledge, meaning, and the responsible use of accumulated influence.

A previous discussion of the Legacy Trap (Hong, 2026) examined how leaders may gradually shift from institutional stewardship toward personal recognition, symbolic achievement, concentrated power, and concern for how history will remember them. David’s life provides a complementary perspective because it demonstrates both the danger of this trap and the possibility of transformation after failure. At the height of his success, David misused authority, attempted to conceal wrongdoing, and experienced consequences that damaged his family and kingdom; yet his willingness to accept correction, repent, endure discipline, rebuild stewardship, and prepare a successor prevented failure from becoming the final definition of his leadership. His story therefore extends the Legacy Trap discussion by asking not only how leaders decline, but also how they may recover moral direction and leave a legacy that continues serving future generations.

Few figures embody this journey more powerfully than David, Israel’s greatest king and the biblical figure described as “a man after God’s own heart” (1 Samuel 13:14; Acts 13:22). This designation does not imply moral perfection, because David’s life included adultery, deception, abuse of power, family dysfunction, and painful political consequences; instead, it points to the orientation of his heart as he sought God’s will, responded to correction, repented when confronted, accepted discipline, and repeatedly returned to God through humility, faith, and obedience. If 1 Samuel shows how God developed David before leadership, 2 Samuel and 1 Kings show how God transformed him through leadership itself. His reign included military victory, national unification, Jerusalem’s establishment as the capital, the arrival of the Ark, covenant promise, family crisis, rebellion, restoration, and succession, demonstrating how achievement, failure, repentance, and grace together shaped his character and legacy.

This article examines David’s transformation from kingdom builder to a man after God’s own heart through the integrated 5Is–5Ps framework: Identity–Purpose, Interest–Priority, Insight–Perspective, Integrity–Principles, and Influence–Power. Drawing on major events in 2 Samuel and 1 Kings 1–2, it shows how leadership maturity develops through achievement, failure, discipline, restoration, stewardship, and the responsible transfer of influence to future generations. This emphasis on transformation beyond positional authority is consistent with Burns’s (1978) distinction between transactional and transformational leadership and Heifetz’s (1994) understanding of adaptive leadership as learning through difficult challenges. The five dimensions operate throughout David’s life, but their emphasis shifts over time: Identity–Purpose and Interest–Priority are most visible during kingdom building, Insight–Perspective becomes central through failure and restoration, and Integrity–Principles and Influence–Power become increasingly prominent in later stewardship and succession. David’s life therefore offers a constructive response to the Legacy Trap by showing that enduring leadership emerges when achievement is disciplined by repentance, power is redirected toward stewardship, and influence is responsibly entrusted to future generations.

2. Stage One: Kingdom Building and Covenant Leadership (2 Samuel 1–10)

Although all five 5Is–5Ps dimensions were present throughout David’s reign, the first stage is most clearly characterized by Identity–Purpose and Interest–Priority, as his calling as God’s anointed king was translated into national unity, political stability, military security, religious renewal, covenant faithfulness, and service to future generations.

2.1 Saul’s Death and the Unification of Israel

David’s kingship began not through political ambition but through patient waiting for God’s timing, even though Samuel had anointed him years earlier and opportunities arose to seize the throne by force. After Saul’s death, David lamented Saul and Jonathan, accepted kingship over Judah at Hebron, and refrained from immediately attacking the northern tribes that remained loyal to Ish-bosheth. His patience, diplomacy, and restraint demonstrated that leadership legitimacy is built through trust, service, and divine calling rather than formal authority or personal ambition alone (Bennis, 2009). Over time, David’s military competence, political wisdom, and growing public trust enabled him to unite all twelve tribes under one kingdom. The elders of Israel recognized his long-standing service and invited him to become king over the entire nation, transforming a loose tribal confederation into a more cohesive political community. David’s authority endured because faithful service, proven leadership, and God’s providential timing prepared both the leader and the nation for unification.

2.2 Bringing the Ark of God to Jerusalem

After establishing Jerusalem as the capital of the united kingdom, David made its spiritual identity a central priority by bringing the Ark of God into the city. Because the Ark represented God’s covenant presence, David understood that lasting national unity required more than military power, administrative order, or political authority. By relocating the Ark, he sought to unite Israel geographically, politically, and spiritually around a shared covenant relationship with God. The procession into Jerusalem revealed a leader who placed God’s authority above his own royal prestige. David’s joyful dancing before the Ark expressed humility, gratitude, and wholehearted worship, even at the risk of appearing undignified, reflecting Nouwen’s (1989) view of leadership as a movement from power toward servanthood. The event demonstrated that David’s legitimacy rested not only on political achievement, but also on worship, stewardship, and submission to God’s sovereignty.

Figure 1
David’s Leadership Transformation Journey

Note. Source: Author-developed framework based on 2 Samuel, 1 Kings, and leadership literature. AI-assisted visualization was used for graphic presentation only; the conceptual framework and interpretation are the author’s.

2.3 The Davidic Covenant: Messianic Promises

The defining event of David’s reign occurred in 2 Samuel 7, when his desire to build a permanent temple was met by God’s promise to establish his house, kingdom, and throne forever. This covenant moved David’s leadership beyond military success and national governance by placing his reign within God’s long-term redemptive purpose. It also expanded his horizon from serving one generation to participating in a promise associated with the future Messiah, reflecting the kind of long-range leadership vision described by Greenstein (2009).

At the same time, David faced difficult leadership tensions within his own administration, especially through Joab, whose military ability and loyalty were repeatedly compromised by impulsive violence and political recklessness. Joab’s killings of Abner, Absalom, and Amasa created crises that conflicted with David’s moral preferences and threatened national unity, yet removing such a powerful and effective commander carried serious political costs. David’s experience illustrates Niebuhr’s (1932) observation that leaders often pursue moral purposes within institutions shaped by competing interests, power struggles, and imperfect choices. He was required to preserve national stability and organizational cohesion while bearing the consequences of Joab’s actions and his own limited political leverage. This tension demonstrates that leadership requires not only vision and achievement but also patience, judgment, and the wisdom to navigate morally difficult relationships within complex systems.

3. Stage Two: Leadership Failure, Discipline, and Restoration (2 Samuel 11–20)

Although identity, purpose, interest, and priority remained active, David’s failures and restoration brought Insight–Perspective to the foreground as success, power, temptation, and reduced accountability drew him toward the Legacy Trap. The resulting family crisis and public suffering exposed the consequences of distorted priorities, while discipline, repentance, and restoration deepened his moral perspective, spiritual maturity, and understanding of covenant stewardship.

3.1 Bathsheba Incident and Nathan’s Rebuke

David’s greatest leadership failure occurred at the height of national success, when military victories, political stability, and public recognition had greatly expanded his influence. During this period of prosperity, he committed adultery with Bathsheba and arranged the death of her husband, Uriah, to conceal his wrongdoing. These actions revealed profound failures of integrity, stewardship, and the responsible use of power. Ludwig and Longenecker (1993) describe this pattern as the “Bathsheba Syndrome,” in which achievement, authority, and reduced accountability create conditions for self-deception and ethical collapse. David’s actions also illustrate the central dynamic of the Legacy Trap: accumulated success and authority weakened accountability, while the protection of position and reputation began to outweigh truth, justice, and responsible stewardship.

God confronted David through the prophet Nathan, whose parable exposed both the seriousness of his sin and the danger of leaders becoming insulated from correction. David did not deny responsibility or shift blame but confessed, “I have sinned against the Lord,” (NIV, 2011) while Psalms 32 and 51 reflect his brokenness, repentance, and spiritual renewal. Although the consequences of his actions remained, the episode taught him that restoration begins not with preserving image or authority, but with humility, confession, and renewed dependence on God’s grace. His response illustrates the difference between remaining trapped in self-protection and beginning ethical recovery, showing that genuine repentance can open the way toward restored stewardship and renewed leadership legitimacy.

3.2 Disorder within David’s House

Nathan’s warning that “the sword will never depart from your house” became painfully visible as turmoil spread through David’s family. Amnon’s assault of Tamar exposed serious failures of justice and protection within the royal household, while David’s inability to respond decisively deepened resentment and led Absalom to kill Amnon in revenge. What began as David’s personal moral failure gradually expanded into a wider family and political crisis, revealing the sharp contrast between stability in the kingdom and disorder within his own household.

The conflict eventually escalated into Absalom’s rebellion, forcing David to flee Jerusalem and endure humiliation, uncertainty, and loss. Even after Absalom’s death, succession tensions continued through Adonijah’s attempt to seize the throne. These events taught David that achievements in governance, military affairs, or organizational leadership cannot compensate for neglected relationships and unresolved conflict within one’s closest circle of responsibility. His experience supports Schein and Schein’s (2021) argument that enduring leadership depends on trust, openness, and healthy relationships rather than formal authority alone, while also demonstrating that leadership failure rarely remains private: unresolved misconduct at the center of authority can spread through families, institutions, and succession systems.

3.3 Rebellion, Discipline, and Restoration

Absalom’s rebellion became the greatest political crisis of David’s reign and one of the most painful chapters of his leadership journey. Years earlier, David had fled from Saul as an innocent servant, but he now fled Jerusalem as a disciplined king confronting the consequences of his own failures. Absalom’s popularity and political maneuvering threatened both David’s authority and the stability of the nation, while David’s love for his son conflicted with his duty as king. The crisis reflects Niebuhr’s (1932) observation that political leaders often face competing moral obligations and Heifetz’s (1994) view that adaptive leadership develops through painful confrontation with difficult realities.

Throughout the rebellion, David demonstrated humility, patience, and submission to God’s sovereignty. He left the Ark in Jerusalem, refused to retaliate against Shimei, and resisted efforts to manipulate circumstances for his own advantage. Although the rebellion ended with his restoration to the throne, David did not escape the consequences of his failure; instead, the experience deepened his understanding of authority, suffering, dependence, and stewardship. He gradually moved beyond the Legacy Trap by relinquishing self-justification, accepting discipline, and submitting his authority to purposes greater than personal survival. His experience shows that enduring leadership rests not merely on recovering power, but on allowing failure to produce humility, accountability, and renewed faithfulness.

4. Stage Three: Leadership Legacy and Succession (2 Samuel 21–24; 1 Kings 1–2)

As David entered the final stage of his reign, all five dimensions continued to interact, but Integrity–Principles and Influence–Power became especially prominent through repentance, leader development, institutional stewardship, and succession. His later years illustrate a movement from personal accomplishment toward stewardship across generations, consistent with the Legacy Trap argument that enduring leaders should be evaluated by the institutions, relationships, people, and succession systems they leave behind (Hong, 2026; Bennis, 2009; Burns, 1978).

4.1 David’s Mighty Men and Organizational Leadership

David’s kingdom was built not by one leader alone but through the courage and loyalty of the group known as David’s Mighty Men, many of whom shared hardship, danger, and uncertainty with him during his years as a fugitive. Their long-term commitment reflected David’s ability to build trust, inspire sacrifice, and unite diverse people around a common purpose. They also demonstrate a central principle of organizational leadership: great leaders multiply leadership by developing followers who can exercise initiative, make decisions, and contribute independently to collective success. In the language of institutional stewardship, David’s enduring influence came not from concentrating power in himself but from preparing capable people to carry responsibility beyond his personal presence, resembling Kearns Goodwin’s (2005) account of Abraham Lincoln’s leadership of talented individuals toward a shared national purpose.

Table 1
David’s Leadership Transformation (2 Samuel–1 Kings 2)

Note. Source: Author’s framework, integrating the 5Is (Identity, Interest, Insight, Integrity, Influence) and 5Ps (Purpose, Priority, Perspective, Principles, Power) dimensions of David’s leadership journey.

Table 1 presents David’s leadership transformation through the integrated 5Is–5Ps framework, showing how personal formation and leadership stewardship developed together across his reign. His Influence–Power extended beyond military and political achievement to leader development, institutional continuity, Solomon’s succession, and the enduring Messianic promise, revealing transformation as a lifelong integration of achievement, adversity, character, and legacy.

4.2 Military Successes and the Census Failure

Throughout his reign, David achieved major military victories over the Ammonites, Arameans, Philistines, Moabites, and Edomites, expanding Israel’s territory and strengthening its security. These accomplishments transformed Israel from a vulnerable tribal confederation into a unified and respected regional kingdom. His leadership, together with the abilities of his commanders and soldiers, brought stability, influence, and prosperity that earlier generations had not achieved. Yet Willard (1998) warns that spiritual formation is endangered when confidence shifts from dependence on God to trust in visible accomplishments.

One of David’s final failures therefore emerged at the height of national success, when he ordered a census of Israel’s fighting men. The census represented a late-life return to the logic of the Legacy Trap because visible strength, numerical measurement, and national capacity began to displace dependence, restraint, and covenant responsibility. Although Joab questioned the command and the resulting judgment exposed the limits of military and political power, David repented, accepted responsibility, and submitted himself to divine correction. His successes and failures reveal that leaders never become permanently immune to the Legacy Trap; leadership transformation requires continuing self-examination, inward character formation, accountability, and the responsible use of authority within imperfect political and organizational systems.

4.3 Final Words and Succession to Solomon

David’s final leadership responsibility was succession planning, an essential but often neglected dimension of effective leadership. As his life neared its end, he recognized that the kingdom’s future depended not only on past achievements but also on a stable transfer of authority to the next generation. Despite political tensions and competing claims, he secured Solomon’s succession and prepared him to continue the mission, including the future construction of the Temple. By relinquishing control and preparing a successor, David demonstrated that mature leadership preserves institutional continuity beyond the tenure of one individual (Bennis, 2009).

Succession provided the clearest evidence that David’s final legacy was no longer defined solely by what he personally controlled or completed. He accepted that another leader would build the Temple and carry the mission forward, demonstrating that faithful stewardship sometimes requires leaving important work unfinished and entrusting it to the next generation. His final words emphasized obedience to God above military success, political power, or personal accomplishment, as he instructed Solomon to uphold the covenant and govern with wisdom and justice. David’s influence therefore extended beyond his lifetime, showing that leadership escapes the Legacy Trap when power is no longer used to preserve personal prominence but is faithfully entrusted to future generations.

4.4 From the Legacy Trap to Enduring Stewardship

David’s life illustrates both the danger of the Legacy Trap and the possibility of transformation beyond it. At the height of success, he misused power, concealed wrongdoing, neglected accountability, and experienced consequences that spread through his family and kingdom. Yet his story did not end with personal decline because he accepted prophetic correction, confessed his failure, endured discipline, rebuilt his understanding of stewardship, developed other leaders, and prepared Solomon for succession. David’s enduring legacy therefore rests not on the absence of failure or the preservation of personal reputation, but on the redirection of power from self-protection toward repentance, covenant faithfulness, institutional continuity, and service to future generations.

5. Conclusion

David’s life reveals that success can build a kingdom, but only repentance, discipline, and faithful stewardship can create a legacy that endures across generations. His misuse of power demonstrates the danger of the Legacy Trap, while his willingness to accept correction and return to God shows that failure need not become the final definition of leadership. His transformation through family crisis, political rebellion, restoration, leader development, and succession demonstrates that identity, purpose, priorities, integrity, and influence are tested most deeply after power has been gained. The final measure of leadership is therefore not how much authority one accumulates or how prominently one is remembered, but what kind of person one becomes, what institutions and people one strengthens, and what future one prepares for others. David’s greatest legacy was not simply the kingdom he ruled, but the heart he cultivated before God, the responsibility he entrusted to the next generation, and the Messianic promise fulfilled in Jesus Christ.

References

Bennis, W. G. (2009). On becoming a leader (4th ed.). Basic Books.

Burns, J. M. (1978). Leadership. Harper & Row.

Collins, J. (2026). What to make of a life: Cliffs, fog, fire and the self-knowledge imperative. Harper Edge.

Erikson, E. H. (1982). The life cycle completed. W. W. Norton.

Greenstein, F. I. (2009). The presidential difference: Leadership style from FDR to Barack Obama (3rd ed.). Princeton University Press.

Heifetz, R. A. (1994). Leadership without easy answers. Belknap Press of Harvard University Press.

Hong, P. C. (2026, August 11). Legacy traps in politics: Lessons on purpose, character, and legacy. K-GSP Forum, 1–12.

The Holy Bible, New International Version. (2011). Zondervan.

Kearns Goodwin, D. (2005). Team of rivals: The political genius of Abraham Lincoln. Simon & Schuster.

Ludwig, D. C., & Longenecker, C. O. (1993). The Bathsheba syndrome: The ethical failure of successful leaders. Journal of Business Ethics, 12(4), 265–273.

Niebuhr, R. (1932). Moral man and immoral society: A study in ethics and politics. Charles Scribner’s Sons.

Nouwen, H. J. M. (1989). In the name of Jesus: Reflections on Christian leadership. Crossroad.

Schein, E. H., & Schein, P. A. (2021). Humble leadership: The power of relationships, openness, and trust. Berrett-Koehler.

Willard, D. (1998). The divine conspiracy: Rediscovering our hidden life in God. Harper San Francisco.

About the Author

Distinguished Professor, Dr. Paul Hong (Editor in Chief) — University of Toledo

Paul C. Hong is Distinguished University Professor at the University of Toledo. Drawing on more than four decades of leadership and missional service and three decades of scholarship in leadership and supply chain management, his research examines how individuals, organizations, and societies develop the capabilities necessary for resilience, adaptation, and long-term impact. His recent work integrates history, management, philosophy, and faith-based narratives to explore leadership formation through the development of identity and purpose, interests and priorities, insight and perspective, integrity and principles, and influence through stewardship—qualities that remain essential in the age of artificial intelligence.

K-GSP Forum · Leadership & Governance Research · August 2026 · All rights reserved

Original Article:

8 13 26 David's Leadership Transformation Journey Paulhongr

2.14MB ∙ PDF file

Download

Download

Suggested Citation: Paul C. Hong (2026, August 13). David’s Leadership Transformation Journey. K-GSP Forum, 1–10.

다윗의 리더십 변혁 여정

폴 C. 홍(Paul C. Hong)은 미국 톨레도대학교의 석좌교수로, 리더십과 공급망 관리에 대한 오랜 연구와 선교적 봉사 경험을 바탕으로 개인과 조직이 어떻게 정체성, 목적, 통찰, 성품, 영향력을 발전시키는지를 탐구해 왔다. 이 글은 다윗의 생애를 단순한 성공담이나 영웅 이야기로 보지 않고, 성공과 실패, 회개와 징계, 회복과 계승을 통하여 한 지도자가 어떻게 변화되는지를 분석한다. 다윗이 “하나님의 마음에 합한 사람”으로 불린 것은 도덕적으로 완전했기 때문이 아니라, 실패한 뒤에도 변명하지 않고 하나님께 돌아가며 자신의 삶을 다시 하나님의 목적에 맞추려 했기 때문이다. 따라서 다윗의 진정한 위대함은 왕국의 크기나 군사적 업적보다, 권력을 소유한 이후에도 회개하고 배우며 자신의 영향력을 다음 세대를 위해 책임 있게 사용한 데 있다.

다윗의 리더십 여정은 세 단계로 전개된다. 첫 단계는 사울의 죽음 이후 헤브론에서 왕이 되고, 이스라엘 열두 지파를 통합하며, 예루살렘을 수도로 세우고, 언약궤를 옮겨 온 왕국 건설의 시기이다. 그는 왕위를 강제로 빼앗지 않고 하나님의 때를 기다렸으며, 정치적 통합뿐 아니라 하나님의 임재와 언약을 중심으로 국가의 영적 기초를 세우려 했다. 또한 다윗 언약을 통하여 자신의 통치가 한 세대의 정치적 성공을 넘어 장차 메시아적 약속과 연결된다는 사실을 깨달았다. 이 시기의 핵심은 정체성에서 목적이 나오고, 관심이 우선순위를 결정한다는 점이며, 다윗은 목자에서 언약의 왕으로 성장하면서 개인적 생존보다 국가의 통합, 예배, 안보와 미래 세대의 준비를 우선시하였다.

두 번째 단계는 성공의 절정에서 찾아온 리더십 위기이다. 다윗은 밧세바와의 간음과 우리아의 죽음을 통하여 권력, 욕망, 자기기만이 결합될 때 성공한 지도자도 얼마나 심각하게 무너질 수 있는지를 보여 주었다. 그러나 나단의 책망을 받은 뒤 책임을 회피하지 않고 죄를 인정했으며, 시편 32편과 51편에서 드러나듯 회개와 은혜를 통하여 회복의 길로 들어갔다. 그럼에도 그의 죄는 암논과 다말의 비극, 압살롬의 반란, 가족의 분열과 정치적 혼란이라는 오랜 결과를 남겼다. 이 과정을 통해 다윗은 공적 성공이 가정과 관계의 건강을 보장하지 않으며, 진정한 지도력은 실패가 없는 상태가 아니라 실패에 어떻게 반응하고, 징계를 받아들이며, 다시 올바른 방향으로 돌아가는가에 의해 평가된다는 사실을 배웠다.

마지막 단계에서 다윗의 관심은 정복과 확장에서 성품, 유산, 후계자 준비로 이동했다. 그는 용사들과 지도자들을 세워 자신의 영향력을 조직 전체로 확장했고, 말년의 인구조사 실패를 통해 여전히 교만과 자기 의존의 위험에 노출되어 있음을 드러냈지만 다시 회개하고 하나님의 주권을 인정했다. 무엇보다 솔로몬에게 왕권을 넘기고 성전 건축을 준비함으로써, 성숙한 지도자는 모든 일을 자신이 완성하려 하지 않고 사명과 권한을 다음 세대에 책임 있게 이양해야 함을 보여 주었다. 글의 5Is–5Ps 틀은 이러한 변화를 정체성–목적, 관심–우선순위, 통찰–관점, 성실성–원칙, 영향력–권력으로 설명하며, 리더십은 단순히 권력을 얻는 과정이 아니라 그 권력을 하나님의 뜻, 공동체의 유익, 그리고 미래 세대를 위해 바르게 사용하는 과정임을 강조한다. 결국 다윗의 삶은 위대한 지도자는 자신이 세운 왕국보다 하나님 앞에서 형성된 마음과 다음 세대에 남긴 신실한 유산으로 기억된다는 사실을 가르친다.

Dasan Jeong Yak-yong (1762-1836), a renowned Confucian scholar of the Joseon Dynasty, left a lasting literary legacy in Gangjin, making the county a destination for those who admire his literary, philosophical, and social contributions. A century after Dasan’s years of exile in the county, it continued to nurture a vibrant literary spirit. In the early twentieth century, Gangjin produced two prominent poets: Kim Yun-sik (1903-1950), better known by his pen name Yeongnang (영랑 김윤식, 永郞 金允植), and Kim Hyeon-gu (김현구, 金玄鳩, 1904-1950). Kim Yun-sik’s “모란이 피기까지는” (“Until the Peonies Bloom,” 1934) and “누이의 마음아 나를 보아라” (“My Sister, Look at Me,” 1935) remain widely studied poems in Korean language textbooks. Although less widely recognized than Kim Yun-sik, Kim Hyeon-gu was an important member of the Simunhakpa (시문학파, 詩文学派), a 1930s literary movement that promoted pure poetry, alongside Kim Yun-sik.

The literary legacies of the two Kims are well preserved in Gangjin’s county seat. Visitors can explore Yeongnang’s birthplace, where peonies bloom in spring; Kim Hyeon-gu’s forest trail; the Simunhakpa Memorial Hall (시문학파기념관) near Yeongnang’s home; and a nearby alley adorned with representative poems by the two poets. Since these sites are located close to one another, visitors can explore them at a leisurely pace. I visited these places on June 2, 2018, when the weather was especially pleasant. The following brief biographies introduce the poets whose literary footprints continue to shape Gangjin’s cultural landscape.

Yeongnang was one of Korea’s foremost lyric poets and a leading member of the Simunhakpa (Poetic Literature School). He adopted the pen name “Yeongnang” when he began publishing in the journal Simunhak in 1930. After studying at Hwimun School in Seoul and later at Aoyama Gakuin in Japan, his education was interrupted by the 1923 Great Kantō Earthquake. During the March First Independence Movement in 1919, he was imprisoned for six months after attempting to organize demonstrations in his hometown. Following Korea’s liberation in 1945, he became active in public affairs, serving as director of the Publication Bureau in the Ministry of Public Information and participating in local politics. He died in 1950 after being struck by stray gunfire during the Korean War.

Kim published more than eighty poems, translations, essays, and literary reviews, but he is remembered above all for his lyrical poetry. His early poems, including “The Heart Shining on Camellia Leaves” (Dongbaekip-e Bitnaneun Maeum), “Lying on the Hillside” (Eondeoge Baro Nuwŏ), and the beloved “Until the Peonies Bloom?” (Moran-i Pigi-kkajineun), are celebrated for their musical language, emotional restraint, and delicate evocation of nature. Around 1940, however, his poetry took a darker turn, reflecting the despair and oppression of the late colonial period in works such as “Carrying Poison” (Dogeul Chago) and “Geomungo.” After liberation, his poems expressed renewed hope and a commitment to rebuilding the nation. Today, Kim is regarded as one of the greatest modern Korean poets, and his poetry remains widely taught in Korean schools and cherished for its lyrical beauty and artistic refinement.

Kim Hyeon-gu was educated at Paichai High School and later in Japan. He made his literary debut in 1930 by publishing four poems—including “My Love, the River Is So Blue” (Imiyeo Gangmuri Mopsido Pureotseumnida), “A Seagull Floating on the Water,” “Holy Spring and Sorrowful Spring” (Georukan Bomgwa Seulpeun Bom), and “Nirvana” (Jeokmyeol)—in the second issue of the influential literary journal Simunhak. He later published notable poems such as “Lying on the Grass” (Pul Ue Nuwo), “Where My Heart Dwells” (Nae Maeum Saneun Got), and “The Road” (Gil).

Although less prolific and less widely known than his contemporary Yeongnang, Kim Hyeon-gu was an important member of the Simunhakpa. He wrote sparingly, carefully crafting poems distinguished by their gentle musicality, emotional restraint, and sensitive reflections on nature and human experience. A posthumous collection of his poems, Hyeon-gu Sijip (Collected Poems of Hyeon-gu), was published in 1970, helping preserve the legacy of a poet whose quiet voice remains an integral part of modern Korean literary history.

Gangjin County, South Korea, lies along the southern coast of the Korean Peninsula. Across Gangjin Bay to the south is Wando County, which adjoins Dadohaehaesang National Park (the National Park of the Sea of Many Islands). Composite image based on map data © OpenStreetMap contributors.

The information board introduces three landmarks in Gangjin County associated with Yeongnang, Kim Hyeon-gu, and Dasan Jeong Yak-yong.

NOTE: In translating 감성 강진, “Gangjin of Sensibility” or “Gangjin, a Place of Sensibility” would sound more evocative than “Sensitivity Gangjin.” 감성 is usually translated as emotional sensitivity or aesthetic awareness, whereas sensitivity can suggest being easily offended or emotionally delicate.

An alley in the county seat of Gangjin features murals of the poets Kim Yeongnang and Kim Hyeon-gu. The panel in this photo asks, “Who was the poet Kim Hyeon-gu?”

This panel offers a brief biography of the poet Kim Hyeon-gu. Below is a full translation. One thing worth noting is that his repeated efforts to publish a collection of his poems during his lifetime never came to fruition for one reason or another. His insistence on publishing the book as a noncommercial edition created yet another obstacle.

Hyeon-gu (玄鳩), 1903-1950

Kim Hyeon-gu (김현구, 金玄鳩; November 30, 1903–October 3, 1950), pen name Hyeon-gu (玄鳩), was born at 179 Seoseong-ri, Gangjin-eup, Gangjin County, South Jeolla Province.

In May 1930, he published four poems in the second issue of Simunhak (Poetic Literature), including “My Love, the River Is So Blue,” “A Seagull Floating on the Water,” “Holy Spring and Sad Spring,” and “Nirvana” (“Extinction”). After making his literary debut, Hyeon-gu published eight more poems through the journals Munye Wolgan (Literary Monthly) and Munhak (Literature) through April 1934.

In 1935, Simunhaksa planned to publish Hyeon-gu’s Poems, following the publication of Jeong Ji-yong’s Poems and Yeongnang’s Poems. However, the plan fell through because of Park Yong-cheol’s illness, which led to his death in 1938. In 1941, Hyeon-gu planned to publish the collection under the title Musang (Impermanence) and attempted to have it reissued by Gwangmyeong Publishing and Printing Company. This attempt also failed because of a dispute with the publisher, as Hyeon-gu insisted that the book be published as a noncommercial edition.

In 1949, he asked Yeongnang, who was then director of publications at the Ministry of Public Information, to arrange for the publication of his poetry collection, but the project was again thwarted by the outbreak of the Korean War in 1950. Finally, in 1970, his son Won-bae and others published Hyeon-gu’s Poems (玄鳩詩集) as a noncommercial edition. The volume contains 82 poems, including 70 posthumously discovered poems and 12 poems that had been published during his lifetime.

Seven representative poems by Kim Hyeon-gu are displayed on a wall along the alley: “Forest Path,” “Pavilion of Falling Blossoms,” “All Through the Night,” “Black Dove,” “The Young Bride Gathering Greens,” “Life,” and “Nostalgia.”

Featured on an alley wall is the complete text of Kim Hyeon-gu’s poem “Nostalgia,” translated below:

Nostalgia
Kim Hyeon-gu

When a breeze brushes the clear night sky,
I see, far away, the blue road to the land of stars.
My heart longs for something—
is it the distant homeland of a soul
that departed long ago?
While the whole world sleeps
in endless stillness,
my heart, gazing upward,
longs for the land of stars
that drifts across the night, waving and whispering.
What lonely soul has it captured?

A wall along the alley features the complete text of Kim Hyeon-gu’s poem “The Young Bride Gathering Greens,” translated below:

The Young Bride Gathering Greens
Kim Hyeon-gu

Barefoot, hoe in hand, she comes to the fields again today.
The white kerchief on the head of the young bride gathering greens
flutters in the wind as she hums to herself.
The tears of this painful world—how they are the sorrow of longing for her beloved.
Humming as she gathers greens, she gazes down the distant road and sighs:
Will the one she parted from return with this spring?
Seeing a traveler’s footsteps slow along the road,
she wonders: has the one she longs for not come, even this spring?

Kim Hyeon-gu’s poem “My Love, the River Is So Blue” is inscribed on a rock. Below is a translation:

My Love, the River Is So Blue
Kim Hyeon-gu

As though a sigh might carry them away,
silver haze floats in the distance,
rippling over the far-off mountain ridges.
The winding old road shines white.
My love, the river is so blue.
The sunlight that trembled
on the broken castle stones has disappeared,
and the dim moonlight
spreads slowly across the fields.
There is not even a breeze
to cool this burning face.
My love, do you know
how boundless my heartache is?

Tapgol Well, located at the entrance to Yeongnang’s birthplace, is an old village well that inspired the poet’s imagination and serves as the symbolic starting point of the Gangjin-eup Literary Walking Trail.

The thatched-roof Gangjin Sensibility Travel Information Center on the Gangjin-eup Literary Walking Trail.

This rock bears the text of Yeongnang’s poem “Upon the Spring Path Where My Heart Flows Quietly.” Below is a translation of the poem:

Upon the Spring Path Where My Heart Flows Quietly
Kim Yeongnang

Like sunlight whispering along a stone wall,
like spring water smiling beneath the grass,
upon the beautiful spring path
where my heart flows quietly,
today, just for a day,
I wish to gaze up at the sky.

Like the blush rising on a young maiden’s cheek,
like a gentle wave softly moistening the heart,
I wish to gaze upon
a tender emerald sky
flowing lightly like silk.

This information board introduces the birthplace of Yeongnang, providing a biography and chronology of the poet. The final paragraph states:

After the poet moved to Seoul in 1948, his birthplace changed hands several times. In December 1985, Gangjin County purchased the property, and in February 1986, it was designated a provincial cultural heritage site by South Jeolla Province. In October 2007, it was elevated to a nationally designated cultural heritage site. In 2008, the Korean government posthumously awarded him the Order of Cultural Merit, Geumgwan (Gold Crown).

This stone marker stands at the entrance to the birthplace of Yeongnang. The inscription in Chinese characters reads, “Nationally Designated Important Folklore Cultural Heritage No. 252: Birthplace of Yeongnang Kim Yun-sik.”

The open gate to the birthplace of Yeongnang. This is a typical view of a thatched-roof house, once common in rural South Korea until the 1970s and 1980s. Today, such houses are rare and are preserved mainly as cultural heritage at the Korean Folk Village and other heritage sites.

This is Yeongnang’s neatly preserved birthplace in a quiet neighborhood of Gangjin-eup.

Yeongnang’s poem “A Heart Shining on Camellia Leaves” is inscribed on this rock. The poem reads in translation:

A Heart Shining on Camellia Leaves
Kim Yeongnang

Somewhere within my heart,
an endless river flows.
The rising morning light
glints upon its silvery ripples.
As though in my breast, my eyes,
or even in my veins,
my heart lies quietly hidden,
trembling softly in its secret place.
Somewhere within my heart,
an endless river flows.

This rock bears Yeongnang’s “Until the Peonies Bloom,” one of the most celebrated poems of modern Korean literature. Below is a translation:

Until the Peonies Bloom
Kim Yeongnang

Until the peonies bloom,
I shall continue to wait
for my spring.
On the day the peonies
fall, petal by petal,
I shall finally sink
into the sorrow of having lost my spring.
One day in May,
on that sweltering day,
even the fallen petals wither away,
and the peonies disappear
without a trace from the world.
Then my rising hope
collapses in disappointment.
Once the peonies have withered,
that is all—my year is over,
and for all three hundred and sixty days
I shall grieve in sorrow.
Until the peonies bloom,
I shall continue to wait
for the spring of radiant sorrow.

Yeongnang’s poem “My Sister, Look at Me” is inscribed on this rock. The poem captures a sister’s quiet anxiety as autumn approaches. Below is a translation:

My Sister, Look at Me
Kim Yeongnang

“Oh my, the autumn leaves are turning.”
Red persimmon leaves,
burning in the courtyard,
come fluttering down.
My sister looks up, startled.
“Oh my, the autumn leaves are turning.”

With Chuseok just a day or two away,
she must be worried
because the wind has finally gone still.
My sister, look at me.
“Oh my, the autumn leaves are turning.”

Yeongnang’s poem “I Gaze into the Clear Well Before the Courtyard” is inscribed here. Below is a translation:

I Gaze into the Clear Well Before the Courtyard
Kim Yeongnang

I gaze into the clear well
before the courtyard.

Deep beneath the earth,
a captive soul
seems always to be gazing up
at the distant sky.

I gaze into the clear well,
clear and filled with twinkling stars.

Deep beneath the earth,
a soul lies peacefully at rest.
Tonight, its eyes are sparkling,
as though calling out
to its body above.

The clear well
before the courtyard
is the face of my soul.

A portrait of Yeongnang leans against a wall in one of the rooms at his birthplace.

Near Yeongnang’s birthplace stands the Simunhakpa Memorial Hall. The three statues at the entrance commemorate Yeongnang Kim Yun-sik, Jeong Ji-yong, and Yonga Park Yong-cheol, the leading poets who spearheaded the founding of Simunhak (Poetic Literature), a literary journal launched in 1930 that played an important role in the development of modern Korean pure poetry.

An interpretive sign on the ground identifies the statues, from left to right, as Yeongnang, Jeong Ji-yong, and Yonga Park Yong-cheol, respectively.

The left side features an enlarged, inscribed reproduction of the cover of the first issue of Poetic Literature. The right side explains the journal’s importance as the “umbilical cord” of modern Korean poetry and displays photographs of six of its prominent members.

The Simunhakpa Memorial Hall hosts various literary events. The banner in the photo announces “A Wellspring of Humanities: Yeongnang Sensibility School,” a program held at both the memorial hall and Yeongnang’s birthplace from March 5 to December 28, 2018. All the photos in this photo essay were taken on June 2, 2018. After visiting sites related to Yeongnang and Kim Hyeon-gu, I headed to places connected with Dasan Jeong Yak-yong, which are featured in a separate photo essay.

About the Author

Professor of English and Creative Writing and Associate Dean of the School of Humanities and Theology, Missouri Baptist University

John J. Han, PhD, is the author, editor, co-editor, or translator of 36 books. His forthcoming book, Echoes from the Hills: Critical Essays on Ozarks Literature (co-edited with C. Clark Triplett), will be published by the University of Arkansas Press. He has also published more than 3,000 poems, primarily haiku, in a wide range of journals and anthologies. In addition to his long tenure at Missouri Baptist University, he has taught at Kansas State University, the University of Nebraska–Lincoln, Nebraska Wesleyan University, and Washington University in St. Louis. He has also served as a visiting scholar at Georgia College & State University and the University of Ljubljana in Slovenia.

© K-Global Scholars and Professionals Forum. All rights reserved. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

From Smarter AI to Safer AI: The New Global Competition for Trustworthy Artificial Intelligence

Why AI Safety Evaluation May Become the Next Strategic Infrastructure

The Question Has Changed

For much of the past decade, the global artificial intelligence competition was defined by a simple question:

Who can build the most powerful AI?

The answer was measured in model size, computing power, benchmark scores, reasoning ability, multimodality, and the speed with which new capabilities could be commercialized.

But that question is rapidly becoming inadequate.

As artificial intelligence moves from a conversational tool into an autonomous decision-maker, software engineer, scientific assistant, cybersecurity system, medical assistant, and potentially an agent capable of interacting with the physical and digital world, another question is becoming unavoidable:

Can we trust the AI we are building?

This is why the emergence of national AI safety organizations and systematic AI evaluation programs represents something much larger than another branch of technology policy. It signals the beginning of a new institutional architecture for the AI era.

The United Kingdom’s AI safety research and evaluation efforts are particularly important because they illustrate the emergence of an independent capability for testing frontier AI systems rather than simply accepting developers’ own claims about safety. The UK’s AI Security Institute has been conducting evaluations of frontier AI systems since 2023 and has increasingly emphasized systematic, evidence-based measurement of emerging capabilities and risks.

The larger historical significance is clear.

The AI race is gradually becoming a race not only for intelligence, but for trustworthy intelligence.

The Industrial Revolution Had Safety Standards. The AI Revolution Needs Them Too

Every major technological revolution eventually discovers the same lesson: powerful technology cannot become reliable infrastructure without independent mechanisms for testing, certification, monitoring, and accountability.

The automobile industry developed crash testing.

The aviation industry developed rigorous certification and airworthiness standards.

The pharmaceutical industry developed clinical trials.

The financial system developed auditing and regulatory supervision.

AI now needs an equivalent institutional infrastructure.

This does not mean that every AI system should be subjected to the same regulatory burden. A chatbot helping someone write a birthday card is fundamentally different from an autonomous AI system controlling critical infrastructure, conducting cybersecurity operations, or assisting with high-consequence scientific research.

The central principle should therefore be risk proportionality.

The more powerful, autonomous, connected, and consequential an AI system becomes, the stronger its evaluation and assurance requirements should become.

Recent scholarship reinforces this broader conception of AI safety. Gyevnár and Kasirzadeh (2025), for example, argue that AI safety should not be reduced to existential-risk scenarios. AI safety encompasses a much broader spectrum of technical, social, and governance problems.

This is an important conceptual correction.

AI safety is not only about preventing a hypothetical superintelligence from destroying humanity.

It is also about preventing today’s AI systems from making predictable mistakes that can harm individuals, organizations, and societies.

AI Safety Must Move from Principles to Evidence

For years, organizations have published AI principles emphasizing fairness, transparency, accountability, privacy, and safety.

These principles are important.

But principles alone do not make an AI system safe.

The critical transition is from:

“We promise that our AI is safe.”

to:

“Here is independent evidence demonstrating how our AI performs under defined safety conditions.”

This is the difference between AI ethics and AI assurance.

AI assurance asks much more difficult questions:

Can the system resist adversarial manipulation?

Does it remain safe when given unexpected instructions?

Can it be induced to bypass safeguards?

Does it behave differently when deployed with external tools?

Can its outputs be independently verified?

Can failures be reproduced?

Can humans intervene effectively?

What happens when the system is connected to the Internet?

And, perhaps most importantly:

How do we know that the system remains safe after deployment?

A recent AAAI study on AI assurance illustrates this transition from broad principles toward operational mechanisms for assessing transparency, reliability, consistency, and auditability (Cross et al., 2025).

The implication is profound.

The future of trustworthy AI will require not only AI developers, but also AI evaluators, auditors, certifiers, red teams, standards organizations, and independent testing laboratories.

In other words, AI will need an ecosystem of trust.

The New Frontier: Testing AI Under Stress

Traditional software testing generally assumes that a program should behave deterministically under a specified set of inputs.

Generative AI is different.

Large language models are probabilistic, context-sensitive, and capable of producing different outputs under apparently similar conditions. Their behavior can also change dramatically when connected to tools, external databases, agents, or other AI systems.

Therefore, conventional software testing is not enough.

AI safety testing increasingly needs to resemble stress testing.

A model should not merely be tested under normal circumstances.

It should be challenged.

It should be exposed to adversarial prompts.

It should encounter ambiguous instructions.

It should face conflicting objectives.

It should be tested against prompt injection.

It should be placed in unfamiliar contexts.

And, where appropriate, it should be evaluated in realistic environments.

The importance of this approach is demonstrated by Zhou et al. (2026), whose study in Nature Machine Intelligence benchmarked large language models against safety risks in scientific laboratory settings. The study illustrates why AI safety cannot be evaluated solely through abstract language benchmarks; models must also be examined in realistic, domain-specific environments where errors can have consequential effects.

This represents an important change in philosophy:

AI safety should be measured where AI is actually going to be used.

The Expanding AI Attack Surface

There is another reason why AI safety cannot be separated from cybersecurity.

As AI systems become more autonomous, their attack surface expands.

An isolated language model is one thing.

An AI agent with access to email, cloud infrastructure, databases, source-code repositories, financial systems, enterprise APIs, and the Internet is something entirely different.

The agent may be highly capable—but so is the potential attacker who manipulates it.

Recent research on the security of LLM agents highlights this duality: agentic AI can simultaneously strengthen cybersecurity defenses while expanding the attack surface through autonomy, tool use, and system integration (Xu et al., 2026).

This means that AI safety must include cybersecurity by design.

A trustworthy AI system should therefore incorporate:

identity and access control,

zero-trust architecture,

secure tool invocation,

sandboxing,

continuous monitoring,

runtime policy enforcement,

audit logging,

data integrity protection,

and human intervention mechanisms.

The distinction between AI safety and cybersecurity is consequently becoming increasingly artificial.

In the future, AI safety will be impossible without AI security.

From Benchmarking to Continuous Assurance

One of the most important developments in AI governance is the recognition that a single pre-deployment test is not sufficient.

An AI model may pass an evaluation today and behave differently tomorrow.

Why?

Because the surrounding system may change.

The model may be updated.

The prompt architecture may change.

New tools may be connected.

The retrieval database may change.

The operating environment may change.

Users may discover previously unknown failure modes.

Attackers may develop new techniques.

Therefore, AI assurance must become continuous.

This suggests a lifecycle model:

Design → Test → Red Team → Deploy → Monitor → Re-evaluate → Update → Re-certify

Such a model represents a fundamental shift from traditional product certification.

AI safety should not be treated as a certificate obtained once.

It should be treated as a continuous process of accumulating confidence.

This is particularly important because recent research has begun to map AI safety across multiple layers rather than treating “safety” as a single property. Chen et al. (2026), in a recent Artificial Intelligence Review survey, identify concerns spanning functional correctness, robustness, safeguard bypass, training-data integrity, privacy, transparency, content authenticity, misuse resistance, and ecosystem-level control.

The message is clear:

There is no single AI safety test.

There is an entire safety system.

The Independent Auditor May Become as Important as the AI Developer

The financial world provides an instructive analogy.

A corporation does not simply announce that its financial statements are accurate and expect society to accept that claim.

Independent auditors examine them.

Why?

Because trust requires independence.

The same principle will increasingly apply to frontier AI.

If an AI company develops a model, evaluates its own model, defines its own safety criteria, publishes its own results, and certifies its own compliance, a fundamental conflict of interest may emerge.

Independent evaluation therefore becomes strategically important.

The emerging AI assurance ecosystem could eventually include:

AI safety laboratories

independent AI auditors

government testing institutes

academic evaluation centers

industry certification bodies

red-team organizations

and international standards organizations.

Such institutions would not replace AI developers.

They would make AI development more trustworthy.

This is particularly important for frontier AI because many safety-relevant details may not be publicly observable. Rigorous external assessment may require secure access to information that cannot simply be disclosed publicly.

Thus, the future may require a carefully designed balance between transparency and confidentiality.

Trust Is Not a Public-Relations Problem

The ultimate objective of AI safety is not simply to make companies appear responsible.

It is to make AI systems genuinely more trustworthy.

Research on public trust in generative AI reinforces this point. Huynh and Aichner (2025) examine the determinants and outcomes of cognitive trust in generative AI, highlighting the importance of understanding why users trust AI and what consequences that trust produces.

This creates an important paradox.

AI systems must become trustworthy.

But humans must also learn when not to trust them.

Blind trust is not trustworthy AI.

The goal should be calibrated trust.

A user should trust an AI system when evidence supports that trust and question it when uncertainty is high.

This requires AI systems that communicate uncertainty, disclose limitations, preserve provenance, support verification, and make escalation to humans possible.

In this sense, the future of AI is not simply about creating more intelligent machines.

It is about creating intelligent systems that know the boundaries of their reliability—and help humans recognize those boundaries.

The Human Must Remain in the Safety Loop

There is a temptation to believe that increasingly sophisticated AI will eventually eliminate the need for human oversight.

That would be a dangerous assumption.

Human oversight is not merely an emergency brake.

It is part of the architecture of responsible AI.

Human beings provide contextual understanding, ethical judgment, institutional accountability, and the ability to redefine objectives when circumstances change.

This does not mean that humans should manually approve every AI action.

That would be impractical.

Instead, AI systems should be designed according to risk-sensitive human oversight.

Low-risk decisions can be highly automated.

Medium-risk decisions can require human confirmation or sampling.

High-risk decisions should require meaningful human review and the ability to intervene.

This principle becomes even more important as AI moves from generating text to taking actions.

The more an AI system can act, the more important it becomes to define who remains accountable.

What Korea Should Do Now

Korea has an unusual opportunity.

It possesses strong capabilities in semiconductors, telecommunications, cybersecurity, manufacturing, robotics, and digital infrastructure.

These strengths can be combined to create a distinctive national strategy:

not merely becoming an AI power but becoming an AI trust power.

Korea should establish a national AI safety evaluation infrastructure capable of independently testing frontier and high-risk AI systems.

Such an institution should not simply duplicate existing regulatory functions.

It should specialize in technical evaluation.

Its mission could include:

model red teaming,

AI cybersecurity testing,

agentic-AI evaluation,

privacy and data-integrity testing,

adversarial robustness,

hallucination and misinformation assessment,

autonomous-agent safety,

and continuous post-deployment monitoring.

Korea should also connect this infrastructure with universities, government laboratories, semiconductor companies, telecommunications operators, hospitals, financial institutions, and cybersecurity organizations.

The goal should be to create a Korean AI Assurance Ecosystem.

Such an ecosystem could eventually become an exportable national capability.

Korea has already demonstrated that standards and infrastructure can become strategic assets.

AI assurance could be the next one.

From AI Competition to AI Civilization

The deeper significance of AI safety extends beyond technology policy.

Human civilization has repeatedly learned that technological power without institutional responsibility creates instability.

The printing press transformed knowledge.

The steam engine transformed production.

Electricity transformed society.

The Internet transformed communication.

AI is transforming intelligence itself.

That makes AI fundamentally different.

Previous technologies amplified human physical capabilities.

AI increasingly amplifies human cognitive capabilities.

And when cognitive power becomes autonomous, the consequences are potentially much larger.

Therefore, the central question of the AI era is not simply:

How intelligent can machines become?

It is:

How wisely can human beings govern increasingly intelligent machines?

That is a civilizational question.

The New Meaning of AI Leadership

For years, AI leadership was associated with bigger models, larger datasets, faster chips, and greater computing capacity.

Those factors will remain important.

But they will no longer be sufficient.

The next generation of AI leaders will need another capability:

the ability to demonstrate that their AI systems can be trusted.

This will require a new form of infrastructure consisting of:

evaluation,

verification,

auditing,

security,

monitoring,

governance,

and human accountability.

The countries that build this infrastructure early may gain an unexpected competitive advantage.

They will not merely produce AI.

They will make AI deployable.

And in a world where AI enters healthcare, finance, education, defense, transportation, scientific research, and critical infrastructure, deployability will increasingly depend on trust.

Conclusion: Intelligence Must Earn Trust

The greatest mistake of the AI era would be to assume that technological capability automatically creates social legitimacy.

It does not.

A powerful AI system can still be unreliable.

A highly intelligent AI system can still be insecure.

A technically impressive AI system can still be unsafe.

And an AI system that cannot demonstrate why it should be trusted will ultimately face limits on where society is willing to deploy it.

The emerging global movement toward AI safety evaluation therefore represents more than a regulatory trend.

It is the beginning of a new technological institution.

The world is moving from:

AI development → AI evaluation → AI assurance → AI trust.

The question of the next decade will not simply be:

“Who has the smartest AI?”

It will increasingly be:

“Whose AI can we safely trust?”

That is where the next great competition will be fought.

And perhaps the most important lesson is this:

The future will not belong simply to those who build the most intelligent machines. It will belong to those who can make intelligence worthy of trust.

Smart AI may win the race for capability.
Safe AI will win the race for civilization.

References

1. Chen, C., Gong, X., Liu, Z., Jiang, W., Goh, S. Q., & Lam, K.-Y. (2026). AI safety landscape for large language models: Taxonomy, state-of-the-art, and future directions. Artificial Intelligence Review. https://doi.org/10.1007/s10462-026-11590-x

2. Cross, M., Gutmann, A., Psychoula, I., & Friere, P. (2025). From AI principles to AI assurance: An online safety case study. Proceedings of the AAAI Symposium Series, 7(1), 2–10. https://doi.org/10.1609/aaaiss.v7i1.36860

3. Gyevnár, B., & Kasirzadeh, A. (2025). AI safety for everyone. Nature Machine Intelligence, 7, 531–542. https://doi.org/10.1038/s42256-025-01020-y

4. Huynh, M.-T., & Aichner, T. (2025). In generative artificial intelligence we trust: Unpacking determinants and outcomes for cognitive trust. AI & Society, 40, 5849–5869. https://doi.org/10.1007/s00146-025-02378-8

5. Zhou, Y., Yang, J., Huang, Y., Guo, K., Emory, Z., Ghosh, B., Bedar, A., Shekar, S., Liang, Z., Chen, P.-Y., Gao, T., Geyer, W., Moniz, N., Chawla, N. V., & Zhang, X. (2026). Benchmarking large language models on safety risks in scientific laboratories. Nature Machine Intelligence, 8, 20–31. https://doi.org/10.1038/s42256-025-01152-1

더 똑똑한 AI에서 더 안전한 AI로: 신뢰할 수 있는 인공지능을 둘러싼 새로운 국제 경쟁

왜 AI 안전성 평가가 차세대 전략 인프라가 될 것인가

질문이 달라지고 있다

지난 10여 년 동안 세계 인공지능(AI) 경쟁은 매우 단순한 질문으로 정의되어 왔다.

“누가 가장 강력한 AI를 만들 수 있는가?”

그 답은 모델의 규모, 컴퓨팅 능력, 벤치마크 성능, 추론 능력, 멀티모달 능력, 그리고 새로운 기능을 얼마나 빠르게 상용화할 수 있는가에 의해 측정되었다.

그러나 이제 이러한 질문만으로는 충분하지 않다.

AI가 단순한 대화 도구를 넘어 자율적인 의사결정자, 소프트웨어 엔지니어, 과학 연구 보조자, 사이버보안 시스템, 의료 보조자, 그리고 디지털 세계와 물리적 세계에 직접 상호작용할 수 있는 에이전트로 발전하면서 새로운 질문이 피할 수 없는 과제가 되고 있다.

“우리가 만들고 있는 AI를 과연 신뢰할 수 있는가?”

이 때문에 국가 차원의 AI 안전 연구기관과 체계적인 AI 평가 프로그램의 등장은 단순히 기술정책의 한 분야가 새롭게 생겨났다는 의미를 넘어선다. 그것은 AI 시대를 위한 새로운 제도적 안전 인프라가 형성되기 시작했음을 의미한다.

영국의 AI 안전 연구 및 평가 노력은 특히 중요하다. 영국은 2023년부터 첨단 AI 시스템에 대한 독립적인 평가 역량을 발전시켜 왔으며, 개발자가 주장하는 안전성을 그대로 받아들이는 것이 아니라 실제 증거에 기반하여 새로운 능력과 위험을 체계적으로 측정하는 방향으로 나아가고 있다.

결국 역사적 의미는 분명하다.

AI 경쟁은 이제 지능만을 겨루는 경쟁에서 신뢰할 수 있는 지능을 만드는 경쟁으로 바뀌고 있다.

산업혁명에는 안전기준이 있었다. AI 혁명에도 필요하다

모든 거대한 기술혁명은 결국 하나의 공통된 교훈을 발견한다.

강력한 기술은 독립적인 시험, 인증, 감시, 책임 체계 없이는 신뢰받는 사회 인프라가 될 수 없다.

자동차 산업은 충돌시험을 발전시켰다.

항공산업은 엄격한 감항성 및 안전 인증체계를 발전시켰다.

제약산업은 임상시험 제도를 발전시켰다.

금융시스템은 감사와 규제감독 체계를 발전시켰다.

이제 AI에도 이에 상응하는 제도적 인프라가 필요하다.

물론 모든 AI 시스템에 동일한 수준의 규제를 적용해야 한다는 의미는 아니다. 누군가의 생일 축하 메시지를 작성하는 AI와 핵심 인프라를 통제하거나 사이버보안 작업을 수행하거나 고위험 과학 연구를 지원하는 자율 AI 시스템은 본질적으로 다르다.

따라서 핵심 원칙은 위험에 비례하는 안전관리(risk proportionality)가 되어야 한다.

AI 시스템이 더욱 강력하고, 자율적이며, 외부 시스템과 연결되고, 사회적 영향력이 커질수록 더욱 엄격한 평가와 보증 절차가 필요하다.

최근 연구 역시 AI 안전을 이러한 보다 넓은 관점에서 바라볼 필요성을 강조한다. Gyevnár와 Kasirzadeh(2025)는 AI 안전을 단순히 존재론적 위험이나 초지능 문제로 한정해서는 안 되며, 기술적·사회적·거버넌스 차원의 다양한 문제를 포함하는 개념으로 이해해야 한다고 주장한다.

이는 중요한 개념적 전환이다.

AI 안전은 가상의 초지능이 인류를 파괴하는 것을 막는 문제만이 아니다.

오늘날 사용되는 AI가 예측 가능한 실수를 통해 개인과 조직, 사회에 피해를 주는 것을 방지하는 것도 AI 안전이다.

AI 안전은 원칙에서 증거로 이동해야 한다

지난 몇 년 동안 많은 기업과 정부기관은 공정성, 투명성, 책임성, 개인정보 보호, 안전을 강조하는 AI 원칙을 발표해 왔다.

이러한 원칙은 중요하다.

그러나 원칙만으로 AI가 안전해지는 것은 아니다.

가장 중요한 변화는 다음과 같은 전환이다.

“우리의 AI는 안전하다고 약속합니다.”

에서

“우리 AI가 정의된 안전조건에서 어떻게 작동하는지를 보여주는 독립적인 증거가 있습니다.”

로 이동하는 것이다.

이것이 바로 AI 윤리(AI ethics)와 AI 보증(AI assurance)의 중요한 차이다.

AI 보증은 훨씬 더 어려운 질문을 던진다.

AI가 적대적인 조작을 견딜 수 있는가?

예상하지 못한 지시가 주어졌을 때도 안전하게 작동하는가?

안전장치를 우회하도록 유도할 수 있는가?

외부 도구와 연결되었을 때 행동이 달라지는가?

AI의 결과를 독립적으로 검증할 수 있는가?

오류를 재현할 수 있는가?

문제가 발생했을 때 인간이 효과적으로 개입할 수 있는가?

그리고 가장 중요한 질문은 이것이다.

배포된 이후에도 AI가 계속 안전하다는 것을 어떻게 확인할 것인가?

Cross et al.(2025)은 AI 원칙을 실제적인 AI 보증 체계로 전환하는 문제를 다루면서 투명성, 신뢰성, 일관성, 감사 가능성 등의 요소를 실제 안전사례(safety case)와 연결할 필요성을 제시했다.

이러한 변화가 의미하는 것은 분명하다.

미래의 신뢰할 수 있는 AI 생태계에는 AI 개발자뿐만 아니라 AI 평가자, 감사자, 인증기관, 레드팀, 표준화기관, 독립적인 시험기관이 필요하다.

다시 말해 AI에도 하나의 신뢰 생태계(trust ecosystem)가 필요하다.

새로운 최전선: 스트레스 상황에서 AI를 시험하라

전통적인 소프트웨어 테스트는 일반적으로 프로그램이 특정 입력에 대해 예측 가능한 방식으로 작동해야 한다는 전제에서 출발한다.

생성형 AI는 다르다.

대규모 언어모델은 확률적이고 문맥에 민감하며, 유사한 입력에서도 서로 다른 결과를 만들어낼 수 있다. 또한 외부 도구, 데이터베이스, 에이전트 또는 다른 AI 시스템과 연결되면 행동이 크게 달라질 수 있다.

따라서 전통적인 소프트웨어 테스트만으로는 충분하지 않다.

AI 안전성 평가 역시 스트레스 테스트(stress testing)와 같은 방향으로 발전해야 한다.

AI는 정상적인 상황에서만 시험해서는 안 된다.

AI를 도전적인 상황에 노출해야 한다.

적대적 프롬트를 제공해야 한다.

모호한 지시를 내려야 한다.

서로 충돌하는 목표를 제시해야 한다.

프롬트 인젝션 공격에 노출시켜야 한다.

익숙하지 않은 환경에서 시험해야 한다.

그리고 필요한 경우 실제 사용환경과 유사한 조건에서 평가해야 한다.

Zhou et al.(2026)은 Nature Machine Intelligence에 발표한 연구에서 대규모 언어모델을 실제 과학 실험실 환경에서 발생할 수 있는 안전 위험에 대해 벤치마킹했다. 이 연구는 AI 안전을 추상적인 언어 벤치마크만으로 평가해서는 안 되며, 실제 사용환경에서 발생할 수 있는 구체적인 위험을 중심으로 평가해야 한다는 점을 보여준다.

따라서 중요한 원칙은 다음과 같다.

AI 안전은 AI가 실제로 사용될 환경에서 측정되어야 한다.

확대되는 AI 공격 표면

AI 안전을 사이버보안과 분리할 수 없는 또 하나의 이유가 있다.

AI 시스템이 더욱 자율화될수록 공격 표면(attack surface)이 확대된다.

독립적으로 작동하는 언어모델과, 이메일, 클라우드 인프라, 데이터베이스, 소스코드 저장소, 금융시스템, 기업 API, 인터넷 등에 접근할 수 있는 AI 에이전트는 전혀 다른 존재다.

AI 에이전트가 강력해지는 만큼 이를 조작하려는 공격자 역시 강력해질 수 있다.

최근 LLM 에이전트 보안 연구는 자율성, 도구 사용, 시스템 통합이 AI의 방어능력을 향상시키는 동시에 새로운 공격 표면을 만들어낸다는 이중성을 지적하고 있다.

따라서 AI 안전은 처음부터 사이버보안을 포함해야 한다.

신뢰할 수 있는 AI 시스템은 다음과 같은 요소를 갖추어야 한다.

신원 및 접근통제

제로트러스트 아키텍처

안전한 도구 호출(tool invocation)

샌드박싱

지속적 모니터링

실행단계 정책 적용(runtime policy enforcement)

감사 로그

그리고 인간의 개입 메커니즘

결국 AI 안전과 AI 보안의 경계는 점점 의미가 없어지고 있다.

미래의 AI 안전은 AI 보안 없이는 달성될 수 없다.

벤치마킹에서 지속적 보증으로

AI 거버넌스에서 가장 중요한 변화 가운데 하나는 배포 전에 한 번 안전성 검사를 하는 것만으로는 충분하지 않다는 인식이다.

오늘 안전성 평가를 통과한 AI가 내일도 동일하게 행동한다는 보장은 없다.

왜 그런가?

AI를 둘러싼 환경이 계속 변화하기 때문이다.

모델이 업데이트될 수 있다.

프롬트 구조가 변경될 수 있다.

새로운 도구가 연결될 수 있다.

검색 데이터베이스가 변경될 수 있다.

운영환경이 변화할 수 있다.

사용자가 이전에 발견하지 못했던 새로운 오류를 발견할 수 있다.

공격자는 새로운 공격기법을 개발할 수 있다.

따라서 AI 보증은 지속적인 과정이 되어야 한다.

이를 하나의 생애주기 모델로 표현하면 다음과 같다.

설계 → 테스트 → 레드팀 → 배포 → 모니터링 → 재평가 → 업데이트 → 재인증

이것은 전통적인 제품 인증과 근본적으로 다르다.

AI 안전은 한 번 획득하고 끝나는 인증서가 되어서는 안 된다.

AI 안전은 신뢰도를 지속적으로 축적해 나가는 과정이어야 한다.

Chen et al.(2026)의 최근 연구 역시 LLM 안전을 단일한 속성으로 보지 않고 기능적 정확성, 강건성, 안전장치 우회, 학습데이터 무결성, 개인정보 보호, 투명성, 콘텐츠 진위성, 오용 방지, 생태계 수준의 통제 등 여러 층위로 분석한다.

메시지는 분명하다.

하나의 AI 안전성 테스트만으로는 충분하지 않다.

필요한 것은 하나의 AI 안전 시스템이다.

독립적인 AI 감사자는 개발자만큼 중요해질 것이다

금융 분야는 중요한 역사적 교훈을 제공한다.

기업이 재무제표가 정확하다고 스스로 선언한다고 해서 사회가 그것을 그대로 받아들이지는 않는다.

독립적인 감사자가 검증한다.

왜 그런가?

신뢰에는 독립성이 필요하기 때문이다.

앞으로 이러한 원칙은 첨단 AI에도 적용될 가능성이 높다.

AI 기업이 모델을 개발하고, 스스로 안전성을 평가하고, 자체적인 안전기준을 만들고, 결과를 발표하고, 스스로 안전성을 인증한다면 이해상충 문제가 발생할 수 있다.

따라서 독립적인 AI 평가가 점점 중요해질 것이다.

미래의 AI 보증 생태계는 다음과 같은 기관들로 구성될 수 있다.

AI 안전 연구소

독립적인 AI 감사기관

정부 AI 평가기관

대학 AI 평가센터

산업계 인증기관

레드팀 조직

그리고 국제표준화기관

이러한 기관들은 AI 개발자를 대체하는 것이 아니다.

오히려 AI 개발을 더욱 신뢰할 수 있도록 만들어주는 역할을 한다.

특히 첨단 AI의 경우 안전성과 관련된 모든 정보를 공개하기 어려울 수 있다. 따라서 독립적인 평가를 위해서는 공개와 기밀성 사이의 정교한 균형이 필요하다.

신뢰는 홍보의 문제가 아니다

AI 안전의 궁극적인 목적은 기업이 책임 있는 것처럼 보이게 만드는 것이 아니다.

목적은 실제로 AI 시스템을 더욱 신뢰할 수 있게 만드는 것이다.

Huynh와 Aichner(2025)는 생성형 AI에 대한 인지적 신뢰의 결정요인과 결과를 분석하면서 사람들이 왜 AI를 신뢰하는지, 그리고 그러한 신뢰가 어떤 결과를 가져오는지를 연구했다.

여기에는 중요한 역설이 존재한다.

AI는 신뢰할 수 있어야 한다.

그러나 인간 역시 언제 AI를 신뢰해서는 안 되는지를 알아야 한다.

맹목적인 신뢰는 신뢰할 수 있는 AI가 아니다.

우리가 지향해야 하는 것은 보정된 신뢰(calibrated trust)다.

사용자는 충분한 증거가 있을 때 AI를 신뢰해야 하며, 불확실성이 높은 경우에는 AI의 판단에 의문을 제기해야 한다.

이를 위해 AI 시스템은 불확실성을 표현하고, 한계를 공개하며, 정보의 출처와 근거를 제공하고, 검증을 지원하고, 필요한 경우 인간에게 판단을 넘길 수 있어야 한다.

따라서 미래의 AI는 단순히 더 똑똑한 시스템이 되어서는 안 된다.

자신의 신뢰성의 경계를 인식하고, 인간이 그 경계를 이해하도록 도와주는 지능형 시스템이 되어야 한다.

인간은 안전의 고리 안에 남아 있어야 한다

점점 더 발전하는 AI가 결국 인간의 감독을 필요 없게 만들 것이라고 생각하기 쉽다.

그러나 이것은 위험한 가정이다.

인간의 감독은 단순한 비상 브레이크가 아니다.

인간의 감독 자체가 책임 있는 AI의 핵심 아키텍처다.

인간은 상황에 대한 맥락적 이해, 윤리적 판단, 제도적 책임, 그리고 상황 변화에 따른 목표 재정의 능력을 갖는다.

물론 모든 AI의 모든 행동을 인간이 수동으로 승인해야 한다는 의미는 아니다.

그것은 현실적으로 불가능하다.

대신 위험에 민감한 인간 감독(risk-sensitive human oversight)이 필요하다.

위험이 낮은 결정은 높은 수준의 자동화를 허용할 수 있다.

중간 수준의 위험을 가진 결정은 인간의 확인이나 표본검사를 요구할 수 있다.

고위험 결정은 의미 있는 인간의 검토와 개입 가능성을 요구해야 한다.

특히 AI가 단순히 콘텐츠를 생성하는 것을 넘어 직접 행동하기 시작할수록 이러한 원칙은 더욱 중요해진다.

AI가 더 많은 행동을 할 수 있게 될수록,

누가 최종적인 책임을 지는가?

라는 질문이 더욱 중요해진다.

한국은 지금 무엇을 해야 하는가

한국은 특별한 기회를 가지고 있다.

한국은 반도체, 통신, 사이버보안, 제조업, 로봇, 디지털 인프라 분야에서 강력한 역량을 보유하고 있다.

이러한 역량을 결합한다면 한국은 단순한 AI 강국을 넘어,

AI 신뢰 강국(AI Trust Power)

으로 발전할 수 있다.

한국은 첨단 및 고위험 AI 시스템을 독립적으로 시험할 수 있는 국가 차원의 AI 안전 평가 인프라를 구축해야 한다.

이 기관은 기존의 규제기관 역할을 단순히 반복해서는 안 된다.

핵심 임무는 기술적 평가와 검증이어야 한다.

예를 들어 다음과 같은 분야가 포함될 수 있다.

AI 모델 레드팀

AI 사이버보안 테스트

AI 에이전트 평가

개인정보 및 데이터 무결성 평가

적대적 공격에 대한 강건성 평가

환각 및 허위정보 평가

자율 AI 에이전트 안전성 평가

그리고 배포 이후 지속적인 모니터링

한국은 이러한 인프라를 대학, 정부 연구기관, 반도체 기업, 통신사업자, 병원, 금융기관, 사이버보안 기업과 연결해야 한다.

목표는 한국형 AI Assurance Ecosystem, 즉 한국형 AI 보증 생태계를 구축하는 것이다.

이러한 생태계는 장기적으로 한국이 해외에 수출할 수 있는 새로운 국가적 역량이 될 수도 있다.

한국은 이미 표준과 인프라가 전략적 자산이 될 수 있다는 사실을 보여주었다.

AI 보증은 그 다음 전략적 자산이 될 수 있다.

AI 경쟁에서 AI 문명으로

AI 안전의 더욱 깊은 의미는 단순한 기술정책을 넘어선다.

인류의 역사는 기술적 힘과 제도적 책임 사이의 균형이 얼마나 중요한지를 반복해서 보여주었다.

인쇄술은 지식을 혁명적으로 확산시켰다.

증기기관은 생산을 혁신했다.

전기는 인간사회를 변화시켰다.

인터넷은 인간의 소통방식을 근본적으로 바꾸었다.

그리고 AI는 이제 지능 자체를 변화시키고 있다.

바로 이 점에서 AI는 이전의 기술과 근본적으로 다르다.

이전의 기술들은 인간의 물리적 능력을 확대했다.

AI는 인간의 인지적 능력을 확대하기 시작했다.

그리고 인지적 능력이 자율적으로 작동하기 시작하면 그 결과는 훨씬 더 커질 수 있다.

따라서 AI 시대의 핵심 질문은 단순히,

“기계가 얼마나 똑똑해질 수 있는가?”

가 아니다.

더 중요한 질문은,

“인간은 점점 더 똑똑해지는 기계를 얼마나 현명하게 통제할 수 있는가?”

이다.

이것은 기술적 질문이 아니라 문명적 질문이다.

AI 리더십의 새로운 의미

지난 수년 동안 AI 리더십은 더 큰 모델, 더 많은 데이터, 더 빠른 칩, 더 강력한 컴퓨팅 능력과 연결되어 있었다.

이러한 요소들은 앞으로도 중요하다.

그러나 더 이상 그것만으로는 충분하지 않다.

차세대 AI 리더에게는 또 하나의 능력이 필요하다.

자신의 AI 시스템이 신뢰할 수 있다는 것을 증명할 수 있는 능력이다.

이를 위해서는 다음과 같은 새로운 인프라가 필요하다.

평가

검증

감사

보안

모니터링

거버넌스

그리고 인간의 책임성

이러한 인프라를 먼저 구축하는 국가가 예상하지 못한 경쟁우위를 확보할 수 있다.

그들은 단순히 AI를 생산하는 국가가 아니다.

AI를 실제 사회에 안전하게 배치할 수 있는 국가가 될 것이다.

그리고 AI가 의료, 금융, 교육, 국방, 교통, 과학 연구, 핵심 인프라로 들어가는 세계에서 AI의 실제 활용 가능성은 점점 더 신뢰에 의해 결정될 것이다.

결론: 지능은 신뢰를 얻어야 한다

AI 시대의 가장 큰 실수는 기술적 능력이 자동적으로 사회적 정당성을 만들어준다고 생각하는 것이다.

그렇지 않다.

강력한 AI도 신뢰할 수 없을 수 있다.

매우 지능적인 AI도 안전하지 않을 수 있다.

기술적으로 뛰어난 AI도 보안에 취약할 수 있다.

그리고 왜 신뢰해야 하는지를 증명하지 못하는 AI는 결국 사회가 허용하는 활용 범위에 한계를 가질 수밖에 없다.

따라서 전 세계적으로 등장하고 있는 AI 안전 평가 움직임은 단순한 규제 흐름이 아니다.

그것은 새로운 기술제도의 탄생이다.

세계는 이제

AI 개발 → AI 평가 → AI 보증 → AI 신뢰

의 방향으로 움직이고 있다.

앞으로의 질문은 단순히

“누가 가장 똑똑한 AI를 가지고 있는가?”

가 아닐 것이다.

점점 더 중요한 질문은

“누구의 AI를 우리가 안전하게 신뢰할 수 있는가?”

가 될 것이다.

바로 이것이 다음 세대 AI 경쟁의 핵심 전장이 될 것이다.

그리고 가장 중요한 교훈은 이것이다.

미래는 단순히 가장 지능적인 기계를 만드는 사람의 것이 아니다.
지능을 신뢰할 수 있는 것으로 만드는 사람의 것이다.

Smart AI가 능력의 경쟁에서 앞설 수 있다면,
Safe AI는 문명의 경쟁에서 승리할 것이다.

참고문헌

Chen, C., Gong, X., Liu, Z., Jiang, W., Goh, S. Q., & Lam, K.-Y. (2026). AI safety landscape for large language models: Taxonomy, state-of-the-art, and future directions. Artificial Intelligence Review. https://doi.org/10.1007/s10462-026-11590-x

Cross, M., Gutmann, A., Psychoula, I., & Friere, P. (2025). From AI principles to AI assurance: An online safety case study. Proceedings of the AAAI Symposium Series, 7(1), 2–10. https://doi.org/10.1609/aaaiss.v7i1.36860

Gyevnár, B., & Kasirzadeh, A. (2025). AI safety for everyone. Nature Machine Intelligence, 7, 531–542. https://doi.org/10.1038/s42256-025-01020-y

Huynh, M.-T., & Aichner, T. (2025). In generative artificial intelligence we trust: Unpacking determinants and outcomes for cognitive trust. AI & Society, 40, 5849–5869. https://doi.org/10.1007/s00146-025-02378-8

Zhou, Y., Yang, J., Huang, Y., Guo, K., Emory, Z., Ghosh, B., Bedar, A., Shekar, S., Liang, Z., Chen, P.-Y., Gao, T., Geyer, W., Moniz, N., Chawla, N. V., & Zhang, X. (2026). Benchmarking large language models on safety risks in scientific laboratories. Nature Machine Intelligence, 8, 20–31. https://doi.org/10.1038/s42256-025-01152-1

2026년 8월 12일

{솔티}

Prof. Dr. Young Choi (Editor in Chief) — Regent University

Young B. Choi is a Professor in the Department of Engineering & Computer Science at Regent University. He published 38 books with ‘Selected Readings in Cybersecurity’ (2018) (over 800 copies archived globally at university/college libraries around the world) and ‘Cybersecurity Applications and Artificial Intelligence’ (2023) available in seven major world languages. He proposed the world’s first global and universal telecommunications “Service Order Handling (SOH)” Model (T-SOH Model) (1995) with Dr. Adrian Tang. With this innovative research work, he received the IEEE NOMS ’96 Best Paper Award and became the first recipient of the Outstanding Contribution Award of the TeleManagement Forum in 1998. His research areas include Natural Language Processing-focused AI, AI-applied cybersecurity, network and telecom service management, and Korean studies on Gani Choi Rip’s Jeonggwan (靜觀: Quiet Contemplation) philosophy and Shilhak ( 實學: Practical Learning).

© K-GSP (K-Global Scholars and Professionals) Forum. All rights reserved. August 2026. Content published in the K-GSP Forum may not be reproduced, distributed, or transmitted in any form without prior written permission from the K-GSP Forum, except for brief quotations with full attribution.

Choi, Y. B. (2026, August 11). From Smarter AI to Safer AI: The New Global Competition for Trustworthy Artificial Intelligence - Why AI Safety Evaluation May Become the Next Strategic Infrastructure. K-GSP.

The image depicts the Buddha and his followers receiving offerings, as illustrated in an eighteenth-century Burmese watercolour manuscript.

Image credit: Life of the Buddha, pages 15-18, Burmese watercolour parabeik, 18th century, Bodleian Libraries, University of Oxford, MS. Burm. a. 12. Digital image via Wikimedia Commons, licensed under CC BY 4.0. The image has been cropped and resized for this publication.

INTRODUCTION VIA HAIBUN

Precepts for Buddhist Monks

There are 250 rules for Buddhist monks to follow. The number varies somewhat by region and tradition, but 250 is the standard in the Buddhist tradition followed in China and Korea. Some of these rules are traditionally attributed to the time of the Buddha. When he began his new religious movement, there were no established rules governing the monastic community. However, soon after ordained believers began living together, the need arose to regulate communal life through various rules. Many of the rules are presented as responses to specific incidents that arose in the monastic community. Over time, they were compiled and organized into increasingly elaborate systems governing the lives of ordained monks.

The 250 precepts are organized within the Vinaya, the Buddhist tradition of monastic discipline, into eight categories that cover both serious offenses and the everyday conduct of monastic life. At the most serious end are the four Pārājika (Disqualification) rules. Breaking any of these results in expulsion from the monastic community: sexual intercourse, theft, killing a human being, and falsely claiming a spiritual attainment one has not achieved. These four cardinal rules are followed by rules concerning serious misconduct, rules dealing with uncertain cases, rules involving improper possessions, rules requiring confession and repentance, rules requiring confession in specific circumstances, rules governing proper conduct and daily behavior, and rules for resolving disputes within the monastic community. Taken together, the 250 rules range from grave offenses that can end a monk’s membership in the community to practical guidelines for dealing with possessions, speech, bodily conduct, etiquette, and communal life.

Some of the rules seem anachronistic (“Do not make a fire outdoors”), while others sound frankly amusing (“Do not tickle another person” and “Do not dig into the middle of the rice and eat from there”). Most, however, reflect a deep insight into human psychology. For example, monks are admonished not to sit alone with a woman in a secluded place. Rules like this are designed to protect the integrity of the monastic community and its members. Some rules even sound as though they were taken from a book of Western etiquette: “Do not speak while food is in your mouth” and “Do not slurp your food into your mouth.” All these rules demonstrate a desire to maintain the purity of thought and conduct in everyday life.

250 rules for monks—
we need one about
smartphones

Precepts for Buddhist Nuns

Whereas Buddhist monks are required to follow 250 rules, Buddhist nuns are subject to 348. That is, they have 98 more precepts to follow. It may seem unfair that their monastic life is governed by almost 100 more rules than that of their male counterparts. However, this disparity should be considered in a socio-historical context. In the Buddha’s time, admitting women into a monastic community represented a significant departure from prevailing social conventions. As members of a religious community, they were understood to need protection from society while also safeguarding the community’s integrity and reputation.

In the Four-Part Vinaya tradition followed in Korean Buddhism, the 348 rules are divided into seven categories that cover everything from serious offenses to the details of everyday monastic life. At the most serious end are the eight Pārājika rules, violations of which result in expulsion from the monastic community. These include sexual intercourse, theft, killing a human being, and falsely claiming a spiritual attainment one has not achieved. The seven categories also include rules concerning serious misconduct, rules involving improper possessions, rules requiring confession and repentance, rules requiring confession in specific circumstances, rules governing proper conduct and etiquette, and rules for resolving disputes within the monastic community. Taken together, the 348 rules cover a wide range of behavior, from major offenses to seemingly small matters involving possessions, speech, clothing, bodily conduct, etiquette, and relationships with others.

While the rules for monks emphasize sexual purity, those for their female counterparts are even more stringent. Among the eight cardinal rules for nuns, three (1, 5, and 6) specifically regulate interactions with men: avoid engaging in sexual intercourse; do not caress, with lustful intent, the parts of a lustful man below the armpits or above the knees; and do not meet the gaze of a lustful man, hold his hand or clothing, stand or talk with him in a secluded place, walk together with him, lean against each other, or make an appointment to meet again. There are many more rules concerning personal conduct, including one governing how nuns should clean themselves in the bathroom. The precepts for women offer both a fascinating glimpse into historical concerns about women’s conduct and a reminder of the need to reconsider such rules in a new age.

348 rules for nuns—
protection from the eyes
sunglasses hide

The Buddhist Precepts for the Modern Age

The rules for ordained Buddhist monks and nuns serve as valuable guides for upright and harmonious living, and most of them are reasonable and sound deeply insightful. Still, some of them seem outdated. For example, the rules related to begging for food make sense in the context of early Indian monasticism, but traditional alms-rounds have not generally been the dominant practice in Mahayana Buddhist regions such as China, Korea, and Japan. The rule restricting bathing to no more than once every fifteen days also sounds both anachronistic and unhygienic. Obviously, the rule “Do not keep medicine for more than seven days” reflects a very different medical context from our own. Modern medicines often need to be taken over much longer periods.

remote Korean temple—
monks watch U.S. news
by satellite

TRANSLATION

Note: The Korean text for the 250 precepts for monks can be found at https://whoami555.tistory.com/13742867 (출가 비구 250계 사분율). The Korean text for the 348 precepts for nuns comes from https://cafe.daum.net/mobuddhism/NNzs/608 (비구니 348계율). These texts contain some missing words, vague wording, and even missing rules. When a rule seemed too vague or unclear to translate reliably, I noted this and provided a tentative translation. When a rule was missing, I noted that as well.

The 250 Precepts for Ordained Buddhist Monks

[1] Pārājika (Disqualification): 1-4 (4 rules)

1) Do not engage in sexual intercourse.

2) Do not steal.

3) Do not kill a human being.

4) Do not tell a major lie.

[2] Saṅghādisesa (Serious Misconduct): 5-17 (13 rules)

5) Do not intentionally discharge semen.

6) Do not touch a woman’s body with sexual intent.

7) Do not speak to a woman about sexual matters.

8) Do not praise your own religious practice to a woman to solicit sexual favors.

9) Do not act as a matchmaker.

10) When building a dwelling without a donor, do not exceed the prescribed dimensions without the community’s approval.

11) When building a dwelling with a donor, do not build without the community’s approval.

12) Do not falsely accuse someone, out of personal animosity, of having committed a serious offense.

13) Do not falsely accuse another person of an offense by fabricating details out of personal animosity.

14) Do not reject admonition intended to prevent the disruption of the community’s harmony.

15) Do not resist admonition while supporting a group that seeks to disrupt the harmony of the community.

16) If you are admonished to leave a village because you have defiled a house where you have received alms, do not disparage the admonition; obey it properly.

17) Do not resist admonition while maliciously opposing the community.

[3] Aniyata (Undetermined Cases): 18-19 (2 rules)

18) [The Korean text is unclear.] Do not render a judgment concerning an offense when the person was mentally or physically incapacitated.

19) [The Korean text is unclear.] Do not render a judgment when circumstances prevent the person’s words from being heard or understood.

[4] Nissaggiya Pācittiya (Improper Possessions): 20-49 (30 rules)

20) Do not keep a robe beyond the prescribed period.

21) Do not sleep separated from your three robes.

22) Do not seek a robe after the prescribed robe-making period has passed.

23) Do not accept robes from a nun who is not a relative.

24) Do not have a nun who is not a relative wash your undergarment.

25) Do not ask for a robe from a layperson who is not a relative.

26) Do not accept an excessive number of robes.

27) Do not persuade donors to provide more money to make a finer robe.

28) Do not combine money from two donors to have one fine robe made.

29) Do not demand a robe urgently after the prescribed time has passed.

30) Do not make robes from silk produced by silkworms.

31) Do not make robes from black wool.

32) Do not make robes from white wool.

33) Do not make a new robe before six years have passed.

34) Add pieces of an old robe to a new robe.

35) Do not carry wool yourself for more than three yojanas. [Yojana: a traditional Indian unit of distance, the exact length of which varied by period and tradition.]

36) Do not have a nun who is not a relative wash, dye, or card your wool.

37) Do not accept gold, silver, copper, or other valuables.

38) Do not exchange or trade valuables.

39) Do not sell goods.

40) Do not keep an extra alms bowl beyond the permitted time limit.

41) When your alms bowl is still usable, do not obtain another new one for yourself.

42) Do not obtain thread yourself and have someone who is not a relative weave it for you.

43) Do not ask donors to give excessive thread to the weaver making your robe.

44) Do not reclaim a robe that you have given away.

45) If you keep medicine, do not keep it for more than seven days.

46) Do not obtain or wear a bathing robe before the proper time.

47) Do not accept a robe for the sick before the proper time, nor keep it beyond the permitted period.

48) In a dangerous forest hermitage, do not sleep without your three robes beyond the prescribed period.

49) Do not take goods intended for the community.

[5] Pācittiya (Confession and Repentance): 50-139 (90 rules)

50) Do not tell a minor lie.

51) Do not insult or scold another person.

52) Do not sow discord between people.

53) Do not sleep in the same room as a woman.

54) Do not sleep with a person who has not received full ordination for more than two nights and three days.

55) Do not recite the Dharma together with a person who has not received full ordination.

56) Do not reveal to a person who has not received full ordination that another monk has committed a serious offense.

57) Do not disclose your own spiritual attainments to a person who has not received full ordination.

58) When teaching the Dharma to a woman, do not speak more than five or six sentences unless a wise man is present.

59) Do not dig the ground yourself or have someone else dig it for you. [This precept comes from the Vinaya’s concern about harming living beings in the soil. Digging the ground could injure or kill small creatures—worms, insects, and other organisms believed to inhabit the earth.]

60) Do not cut down an old tree known as a “spirit tree.”

61) Do not cause the community distress through your actions or speech.

62) Do not suspect or reprimand a monk in charge of duties.

63) Do not leave behind a sleeping platform or meditation cushion after using it.

64) Do not leave bedding spread out in a room when you depart.

65) Do not forcibly take a dwelling already occupied by another person.

66) Do not become angry and drive another person out.

67) Do not sit on a couch with missing legs on the second floor.

68) Do not instruct a nun unless she first provides filtered water. [The nun must provide water that has been filtered to remove insects. The underlying concern is the Buddhist prohibition against harming living beings in water.]

69) Do not use more than three layers of thatch when roofing.

70) Do not instruct a nun without the community’s authorization.

71) Do not continue teaching a nun until after sunset.

72) Do not slander a monk who has instructed a nun.

73) Do not give robes to a nun who is not a relative, except in exchange.

74) Do not have a robe made for a non-relative nun.

75) Do not sit together with a nun in a secluded place.

76) Do not arrange to travel together with a nun.

77) [Number 77 is missing from the Korean text.]

78) Do not eat food praised by a nun.

79) Do not arrange to travel together with a woman.

80) Do not sleep a second night in a lodging intended for only one night’s stay unless you are ill.

81) Do not wander from place to place eating meals.

82) Do not gather separately from the community to eat.

83) Do not accept excessive amounts of food meant for a daughter sent to her husband’s home or food from merchants.

84) Do not violate the rules concerning leftover food.

85) Do not cause another person to violate the rules concerning leftover food.

86) Do not eat at an improper time.

87) Do not eat stale food.

88) Do not eat food that has not been properly offered.

89) Do not ask for choice food.

90) Do not personally give food to non-Buddhist ascetics.

91) Do not enter a village without informing another monk.

92) Do not sit shamelessly in a couple’s home when the husband shows signs of displeasure.

93) Do not sit alone with a woman in a secluded place.

94) Do not sit alone with a woman in an open but isolated place.

95) Do not drive away a fellow traveling monk from the village.

96) Do not accept medicine after the prescribed period has passed.

97) Do not mingle with crowds to watch spectacles.

98) If you must stay overnight among a crowd for business, do not stay longer than two nights and three days.

99) Do not watch a battle in progress. [This prohibition concerns watching warfare as a spectacle, rather than simply being present in a military area.]

100) Do not drink alcohol.

101) Do not play in the water.

102) Do not tickle another person.

103) Do not resist admonition.

104) Do not frighten another monk.

105) Do not bathe more than once every fifteen days.

106) Do not make a fire outdoors.

107) Do not take another person’s robes and bowl.

108) Do not take and wear robes left in another’s care without informing the custodian.

109) Dye new robes before wearing them.

110) Do not intentionally kill an animal.

111) Do not drink water containing living beings.

112) Do not cause another monk anxiety.

113) Do not conceal another person’s serious offense.

114) Do not give full ordination to a person who is under twenty years of age.

115) Do not revive a dispute after it has been settled.

116) Do not travel together with smugglers.

117) Do not reject counsel because of your wrong view.

118) Do not shield an offender.

119) Do not shelter a novice who has been expelled.

120) Do not resist an admonition to study the precepts.

121) Do not disparage the precepts.

122) Do not conceal your offense and then confess it merely because you fear someone else will expose it.

123) Do not agree to a resolution and later regret it.

124) Do not leave a meeting without delegating your vote.

125) Do not raise objections after having delegated your vote.

126) Do not eavesdrop on the disputes of other monks.

127) Do not strike anyone out of anger.

128) Do not make a gesture of striking someone out of anger.

129) Do not maliciously accuse another person of a serious offense without evidence.

130) Do not maliciously accuse another person of a serious offense without evidence. [This appears to duplicate no. 129 in the Korean text.]

131) Do not handle gems or jeweled ornaments.

132) Do not go into a village at an improper time.

133) [Number 133 is missing from the Korean text.]

134) Do not stuff a sitting mat with cotton.

135) Do not make a needle case out of bone, teeth, or horn.

136) Make the sitting mat three cheok two ch’i long and two cheok four ch’i wide; if necessary, each dimension may be increased by eight ch’i. [Cheok (尺) and ch’i (寸) are traditional East Asian units of length. One cheok is divided into ten ch’i. Their exact modern equivalent varied by period and region.]

137) Do not make a robe for covering sores larger than six cheok four ch’i long and three cheok two ch’i wide.

138) A bathing garment should be no more than nine cheok six ch’i long and four cheok wide.

139) Do not make a robe the same size as the Buddha’s robe.

[6] Pāṭidesanīya (Confession in Specific Circumstances): 140-143 (4 rules)

140) While staying in a village dwelling, do not accept food from a nun who is not a relative.

141) While staying in a village dwelling, do not accept food obtained through a nun’s improper solicitation of a donor.

142) Do not go to a house where the community has agreed not to receive offerings, and do not accept excessive offerings there unless you are ill.

143) Do not accept a meal while sitting in a monastery located in a dangerous wilderness area unless you are ill.

[7] Sekhiya (Proper Conduct and Etiquette): 144-243 (100 rules)

144) Wear your undergarment properly.

145) Wear your three robes properly.

146) Do not go to a donor’s house with your robes gathered up.

147) Do not sit in a donor’s house with your robes gathered up.

148) Do not enter a donor’s house with your robe wrapped around your neck.

149) Do not sit in a donor’s house with your robe wrapped around your neck.

150) Do not enter a donor’s house with your head covered.

151) Do not sit in a donor’s house with your head covered.

152) Do not enter a donor’s house while running or jumping.

153) Do not run into a donor’s house and sit down.

154) Do not enter a donor’s house while squatting or crouching.

155) Do not enter a donor’s house with your hands clasped behind your back.

156) Do not sit in a donor’s house with your hands clasped behind your back.

157) Do not enter a donor’s house while swaying or moving your body in an undignified manner.

158) Do not sit in a donor’s house while swaying or moving your body in an undignified manner.

159) Do not enter a donor’s house while swinging your arms.

160) Do not sit in a donor’s house while swinging your arms. [The Korean text says “enter” again; the context strongly suggests “sit.”]

161) Enter a donor’s house with your body properly covered.

162) Sit in a donor’s house with your body properly covered.

163) Do not enter a donor’s house while looking restlessly from side to side.

164) Do not sit in a donor’s house while looking restlessly from side to side.

165) Enter a donor’s house quietly.

166) Sit quietly in a donor’s house.

167) Do not enter a donor’s house laughing and joking frivolously.

168) Do not sit in a donor’s house laughing and joking frivolously.

169) Eat your meal with the Five Contemplations.

170) Receive rice neatly into your alms bowl.

171) Receive soup neatly into your alms bowl.