问题为本——真实问题、伪问题与虚构性问题设置的学术边界OA
Questions as the Foundation:Academic Boundaries Among Genuine Questions,Pseudo-Questions,and Fabricated Problem Settings
本刊此前几篇按语,分别从学术根脉、智能工具、研究评价尺度、学科方向和强工具时代人的成长等角度,讨论了当代科研活动中的若干基础问题.沿着这些讨论,还需要追问一个更靠前的问题:一项研究凭什么开始,它所依据的问题怎样才能真正站得住? 本文的基本判断是,研究不能从答案、工具、热点或概念包装开始,而必须从真实问题开始.但国家需求、工程瓶颈、实验异常、对象失效和理论裂缝,只是问题的来源,还不是科学问题的完成形态.真实困难必须经过科学抽象:从复杂情境中识别主导变量、关键尺度和一般结构,界定对象、假设、边界与证据路径,形成可分析、可建模、可检验并具有一定迁移意义的科学命题.未经抽象的困难仍然只是困难;脱离对象和问题来源的抽象,也可能沦为形式空转.真实问题并不等同于一个未经抽象的现实物体.它可以来自工程实践和物理世界,也可以产生于方程、算子、变分结构、反问题、奇异极限和稳定性等知识结构内部.问题的"真",关键不在于是否直接对应一件看得见的装备,而在于其来源可以说明,概念可以定义,假设可以公开,推理或检验路径可以建立,结论有清楚的适用边界,并能够形成真实的知识增量或能力增量.判断问题时,还必须把三个不同层面分开.问题的轻与重,涉及价值和生长能力;问题的真与伪,涉及问题是否良构、是否成立;研究的诚与假,则涉及对象、数据、图像、证明和证据是否真实.一个问题可以轻而真实,伪问题也可能源于认识不足,并经过重新定义和压缩边界得到纠正;建立在虚构对象、失真数据或不可追溯证据之上的问题外观,则属于学术共同体必须拒绝的虚构性问题设置. 对力学而言,问题最终要回到对象、变量、载荷、边界、响应、机制、失效和验证;对应用数学而言,问题最终要回到定义、假设、数学结构、可解性、稳定性、收敛性、可辨识性和误差边界.人工智能可以扩大检索范围,加快计算、证明探索和表达,却不能替研究者判断问题是否值得问,不能代替科学抽象,也不能承担边界、证明与证据责任.研究从真实问题开始,期刊的学术判断和大学的科研训练,也应从真实问题开始.
Previous editorials of this journal have discussed academic lineage,intelligent tools,criteria for eval-uating research,disciplinary direction,and human development in an age of increasingly powerful tools.A more prior question must now be asked:what gives a study the right to begin,and under what conditions does its underlying question genuinely stand? Research should begin neither with an answer nor with a tool,a fashionable topic,or conceptual packa-ging.It should begin with a genuine question.Yet national needs,engineering bottlenecks,anomalous observa-tions,failures of real objects,and gaps in theory are sources of questions rather than their completed scientific form.A real difficulty must undergo scientific abstraction:dominant variables,controlling scales,and general structures must be identified;the object,assumptions,boundaries,and evidential pathway must be specified;and the difficulty must be transformed into a proposition that can be analysed,modelled,tested,and,where appropriate,transferred beyond its original setting.A difficulty without abstraction remains a difficulty,while abstraction detached from the original object may become an exercise in form alone.A genuine question need not correspond directly to a tangible physical object.It may arise from engineering practice and the physical world,but it may also emerge from equations,operators,variational structures,inverse problems,singular limits,and questions of stability.Its genuineness lies not in whether it refers directly to a visible object,but in whether its origin can be explained,its concepts defined,its assumptions disclosed,its reasoning or testing pathway examined,and the authority of its conclusions properly limited.Three different dimensions must be distinguished.The distinction between a minor and a major question concerns value and capacity for further growth.The distinction between a genuine and a pseudo-question concerns whether the question is well formu-lated and intellectually valid.The distinction between honest and fabricated research concerns whether the ob-ject,data,images,proof,and evidence are authentic.A minor question may be entirely genuine.A pseudo-question arising from incomplete understanding may sometimes be corrected by reformulation and boundary clarification.A fabricated problem setting supported by invented objects,distorted data,or untraceable evi-dence belongs to a different category and must be excluded from academic inquiry. For mechanics,a question must ultimately return to the object,variables,loads,boundaries,responses,mechanisms,failure,and validation.For applied mathematics,it must return to definitions,assumptions,mathematical structures,solvability,stability,convergence,identifiability,and error bounds.Artificial intelli-gence may broaden searches and accelerate computation,proof exploration,and expression,but it cannot as-sume responsibility for deciding whether a question is worth asking,carrying out the necessary scientific ab-straction,or defining the authority of proof and evidence.Research begins with genuine questions,and so should academic judgment in journals and research training in universities.
卢天健
南京航空航天大学,南京 210016
数理科学
问题意识真实问题科学抽象伪问题虚构性问题应用数学力学人工智能学术判断
problem awarenessgenuine questionsscientific abstractionpseudo-questionsfabricated prob-lem settingsapplied mathematicsmechanicsartificial intelligenceacademic judgment
《应用数学和力学》 2026 (8)
959-977,19
评论