How Does Doubt Become Knowledge? From Skepticism to the Scientific Method
An ancient stone observatory stands beneath an immense night sky. A weathered philosopher's chair faces a polished bronze mirror reflecting the stars. Scrolls, geometric diagrams, measuring instruments, and an unfinished astrolabe rest upon a stone table. The scene represents doubt not as denial, but as the disciplined beginning of inquiry—a place where questions become the foundation of knowledge.
How Does Doubt Become Knowledge? From Skepticism to the Scientific Method
How did philosophical doubt help shape modern science? Explore the intellectual journey from ancient skepticism to scientific inquiry.
Why Does Knowledge Begin by Questioning Itself?
Civilizations have developed many ways of establishing what should be believed. Myths situated events within sacred narratives, laws defined social obligations, and traditions preserved accumulated practices and judgments. Alongside these forms persisted a difficult question: how can people determine whether what they believe is true or justified?
Some early Greek thinkers began to seek natural principles and reasoned accounts of the cosmos alongside inherited mythic explanations. They disagreed about the fundamental constitution of reality, proposing water, the boundless, air, change, number, or being as possible principles. This was not a complete passage from myth to reason, because philosophical and mythic forms continued to coexist. It did, however, expand the demand that explanations be examined through argument rather than accepted solely through tradition.
Socrates is known primarily through the writings of Plato, Xenophon, Aristophanes, and later authors, and these sources do not present one entirely consistent portrait. In several of Plato’s early dialogues, Socrates asks speakers to define concepts they assume they understand and then tests their answers for contradiction. These exchanges often end in aporia, a condition of unresolved difficulty, rather than in a final doctrine.
The Socratic practice represented in these dialogues gives questioning a constructive function. Examination can reveal that confidence exceeds understanding. It does not establish that every belief is false, but it shows that a claim to knowledge requires more than assurance, reputation, or social authority.
Philosophy thereby developed a continuing responsibility to investigate the grounds of its claims. Doubt did not immediately become one unified method, and different schools assigned it different purposes. Nevertheless, disciplined questioning became one way of distinguishing an asserted belief from a belief supported by reasons.
When Did Doubt Become a Method?
Ancient skeptical traditions transformed questioning and the suspension of judgment into sustained philosophical practices. Academic skeptics investigated whether knowledge was possible and challenged claims to certainty. Pyrrhonian skeptics, as represented in Sextus Empiricus, set opposing appearances and arguments against one another and suspended judgment when neither side could be decisively preferred.
Pyrrhonism should not be reduced to the claim that nothing can be known, because that statement would itself become a dogmatic assertion. The Pyrrhonist continued to follow appearances, ordinary practices, customs, and practical skills without claiming knowledge of how things ultimately are. Suspension of judgment was also connected with the pursuit of tranquility, not with the construction of modern empirical science.
Ancient skepticism nevertheless placed lasting pressure on claims that relied on unsupported authority. It developed arguments concerning disagreement, perception, regress, relativity, and unproven assumptions. Its historical importance lies partly in showing how difficult it is to justify a claim without presupposing another claim that also requires justification.
Centuries later, René Descartes employed doubt for a different purpose. In ‘Meditations on First Philosophy’, he resolved to withhold assent from beliefs that admitted any reason for doubt. His aim was not to remain indefinitely uncertain but to discover a foundation secure enough to support the reconstruction of knowledge.
Cartesian doubt was therefore methodological and temporary within Descartes’s project. The act of doubting led him to the certainty that thinking was occurring and that he existed as a thinking being. His subsequent arguments concerning God, clear and distinct perception, and the external world remain disputed, but the initial procedure established doubt as an instrument for testing foundations.
Modern science did not emerge from Cartesian doubt alone. Its formation involved ancient natural philosophy, medieval scholarship, mathematical demonstration, Renaissance observation, new instruments, Baconian criticism of cognitive and cultural prejudices, experimental practices, and institutions that enabled results to be communicated and contested. Philosophical doubt contributed to this history by challenging inherited certainty, but it was one element within a larger transformation.
There is also no single procedure used identically by every science. Observation, measurement, hypothesis formation, modeling, experimentation, comparison, statistical inference, replication, criticism, and revision play different roles in different fields. What unites these practices is not one rigid sequence but the requirement that claims remain answerable to evidence and open to public examination.
Is Doubt the Opposite of Knowledge?
Modern culture often presents certainty and doubt as opposites. Knowledge appears to begin only after doubt has disappeared. The history of philosophy and science suggests a more complex relationship: doubt can weaken knowledge when it becomes indiscriminate, but it can strengthen inquiry when it identifies what still requires justification.
A belief protected from every possible question can become dogmatic. Doubt that rejects every possible reason can become indiscriminate suspension or endless hesitation. Inquiry operates between these conditions. It asks what degree of confidence the available evidence warrants and what evidence would require a judgment to change.
Scientific claims are not all supported in the same way or held with the same degree of confidence. Some findings are directly measurable and repeatedly reproduced; others depend on statistical models, historical traces, indirect observation, or converging lines of evidence. Scientific confidence is therefore graduated rather than simply present or absent.
Karl Popper argued that scientific theories should expose themselves to possible falsification and that surviving severe tests could corroborate, but never conclusively verify, a theory. This remains an influential account of scientific criticism. It is not, however, a complete description of all scientific practice, because evidence can bear on theories through confirmation, prediction, explanatory success, replication, and comparison with alternatives as well as through attempted refutation.
The authority of science therefore does not come from complete immunity to doubt. It comes from practices designed to expose claims to evidence, criticism, correction, and independent scrutiny. Particular institutions can fail to uphold these standards, but the standards make errors discoverable and revision possible.
Skepticism has endured because it repeatedly confronts philosophy with the difference between feeling certain and being justified. It does not prove that certainty is impossible. It requires confidence to remain proportionate to reasons and open to revision when those reasons change.
What Should We Question in the Age of Artificial Intelligence?
Generative artificial intelligence has changed how many people search for, summarize, and produce information. Tasks that once required extensive searching can sometimes be completed rapidly, and language models can generate fluent explanations across a wide range of subjects. Speed and linguistic coherence, however, do not establish factual accuracy.
Information is not identical with understanding, and fluency is not identical with truth. Generative systems can produce unsupported or false statements, combine sources incorrectly, or present uncertain material without adequate qualification. The quality of an answer therefore depends not only on how convincingly it is expressed but also on the evidence, provenance, reasoning, and verification supporting it.
This makes disciplined skepticism especially relevant. An answer should be examined by asking what kind of claim it contains, what source supports it, whether the cited source actually entails the conclusion, what assumptions connect the evidence to the claim, and what information could revise the result. These questions apply to human and machine-generated statements alike.
Artificial intelligence can generate questions and suggest lines of inquiry. It cannot thereby assume the whole human responsibility for deciding which purposes matter, which consequences are acceptable, or when a judgment should guide action. The central problem is therefore not whether machines can produce questions, but whether people will retain responsibility for selecting, testing, and acting upon them.
The challenge of the present age is not merely to obtain answers more quickly. It is to prevent speed, abundance, and fluency from being mistaken for justification. Doubt remains valuable when it slows judgment long enough for evidence and reasoning to be examined.
Perhaps an age is remembered not only for the answers it produced, but also for the claims it learned to test and the questions it refused to abandon.
References
Bacon, F. (2000). ‘The New Organon’ (L. Jardine & M. Silverthorne, Eds.). Cambridge University Press. Original work published 1620.
Descartes, R. (1985). ‘Meditations on First Philosophy’. In J. Cottingham, R. Stoothoff, & D. Murdoch (Trans.), ‘The Philosophical Writings of Descartes’ (Vol. 2). Cambridge University Press. Original work published 1641.
Godfrey-Smith, P. (2003). ‘Theory and Reality: An Introduction to the Philosophy of Science’. University of Chicago Press.
Kirk, G. S., Raven, J. E., & Schofield, M. (1983). ‘The Presocratic Philosophers’ (2nd ed.). Cambridge University Press.
National Institute of Standards and Technology. (2024). ‘Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile’. NIST AI 600-1. DOI: 10.6028/NIST.AI.600-1
Plato. (1997). ‘Apology’. In J. M. Cooper (Ed.), ‘Plato: Complete Works’. Hackett Publishing Company.
Popper, K. R. (2002). ‘The Logic of Scientific Discovery’. Routledge. Original English edition published 1959.
Sextus Empiricus. (2000). ‘Outlines of Scepticism’ (J. Annas & J. Barnes, Eds. & Trans.; 2nd ed.). Cambridge University Press.
지식은 왜 자기 자신을 의심하는 데서 시작할까?
의심은 언제 방법이 되었을까?
의심은 지식의 반대일까?
인공지능 시대에는 무엇을 의심해야 할까?
참고문헌
프랜시스 베이컨. ‘신기관’. 리사 자딘·마이클 실버손 편집. 케임브리지대학교출판부, 2000년. 1620년에 처음 출판된 저작입니다. 인간의 판단을 방해하는 ‘우상’을 비판하고 자연 탐구를 위한 체계적인 방법을 제안합니다.
르네 데카르트. ‘제일철학에 관한 성찰’. 존 코팅엄·로버트 스투트호프·두걸드 머독 옮김. ‘데카르트 철학 저작집’ 제2권. 케임브리지대학교출판부, 1985년. 1641년에 처음 출판된 원전입니다. 의심할 수 있는 믿음을 검토하여 확실한 지식의 토대를 찾으려는 데카르트의 방법적 회의를 확인할 수 있습니다.
피터 고드프리스미스. ‘이론과 실재: 과학철학 입문’. 시카고대학교출판부, 2003년. 포퍼의 반증주의를 포함한 주요 과학철학 이론과 실제 과학 연구의 관계를 설명하는 연구서입니다.
G. S. 커크·J. E. 레이븐·말컴 스코필드. ‘소크라테스 이전 철학자들’. 제2판. 케임브리지대학교출판부, 1983년. 초기 그리스 철학자들의 단편과 증언을 수록하고 그들의 자연철학과 논증을 해설한 학술서입니다.
미국 국립표준기술연구소. ‘인공지능 위험관리 프레임워크: 생성형 인공지능 프로파일’. NIST AI 600-1, 2024년. 생성형 인공지능의 사실과 다른 출력, 출처 검증, 신뢰성 문제를 다루는 공인 지침입니다. DOI: 10.6028/NIST.AI.600-1
플라톤. ‘변론’. 존 M. 쿠퍼 편집, ‘플라톤 전집’ 수록. 해킷출판사, 1997년. 소크라테스가 지혜를 자처하는 사람들의 주장을 검토하고 인간 지식의 한계를 논하는 모습을 보여주는 주요 원전입니다.
칼 R. 포퍼. ‘과학적 발견의 논리’. 라우틀리지, 2002년. 영어판은 1959년에 처음 출판되었습니다. 과학 이론의 반증 가능성과 엄격한 검사를 통한 잠정적 입증을 제안한 과학철학의 주요 저작입니다.
섹스투스 엠피리쿠스. ‘회의주의 개요’. 줄리아 아나스·조너선 반스 편집 및 번역. 제2판. 케임브리지대학교출판부, 2000년. 대립하는 논변과 판단 유보, 퓌론주의적 탐구의 목적을 확인할 수 있는 주요 원전입니다.

Comments
Post a Comment