【作者】 蒋柯；

【导师】 熊哲宏；

【作者基本信息】 华东师范大学 ， 基础心理学， 2009， 博士

【摘要】 本文一共包括12章。其中，第1章回顾了归纳问题的哲学和逻辑学研究历史。第2章对归纳推理的心理学研究作了综述，并参照康德的“先验哲学”、古德曼的“主观逻辑”以及施泰默的“进化主义”哲学、逻辑学研究思路，建构了关于归纳推理的心理学研究策略，推导出归纳推理应该具有领域特殊性特征，于是提出了本研究的基本假设：命题2．1．“归纳推理是进化而来的领域特殊性机制”。对命题2．1．的论证过程包括两个部分。第一个部分包括第3章，主要任务是将基本假设转化为可以进行经验验证的中级假设。为了实现这种转换，本研究依据了进化心理学范式的功能分析研究方法，分4个步骤进行论证：第1步：考察我们的祖先可能会面临什么样的适应性问题；第2步：考虑这个适应性问题是否可以通过一个独立的机制来完成；第3步：考察这个独立的机制将按照什么样的内在逻辑运行；第4步：检验现代人的认知过程，看是否存在这种独立的逻辑运算方式。这4个步骤中，前面两个属于理论推演，后两个属于经验验证，其中第3步提出经验研究的假设，第4步则验证假设。按照这4个步骤，本研究首先论证了“归纳推理是一个适应性问题”，然后继续推演出“归纳推理领域特殊性机制”。接下来，更进一步推论：“获得利益”和“避免伤害”是最重要的适应性领域，在这两个领域中，“存在领域特殊性的归纳逻辑”。于是，本研究的基本假设被转换成为两个中级假设：命题3．1：在“获得收益”和“避免伤害”两种条件下。人的归纳推理会表现出不同的特征。命题3．2：在“获得收益”和“避免伤害”两种条件下，人的归纳推理采用两种不同的论证逻辑。在“获得收益”条件下，人采用充分必要性论证策略；在“避免伤害”条件下，人采用充分性论证策略。论证的第二个部分包括第4～11章。第4章对本研究所作的实验的背景信息进行了介绍。包括：1)自变量和因变量的定义方法；2)无关变量的控制技术；3)被试的招募办法。第5章报告了实验1和实验2。两个实验的目的是分离归纳推理中的概念效应和相似性效应。结果发现人在进行归纳推理时，对概念信息的依赖胜于对相似性信息的依赖。第6章的实验3和实验4分离了归纳推理中的前提数量效应和前提多样性效应。结果发现，前提多样性并不是人归纳推理的固有特征。人在组织归纳推理时，倾向于依据前提的数量而忽略前提的多样性。第5章和第6章的4个实验确立了归纳推理作为一种先天“本能”的可能性。这是领域特殊性假设确立的基础。第7章的实验5首先检验了“利”、“害”条件下的归纳推理心理效应是否存在差异。结果发现，在“利”和“害”两种条件下，被试的归纳推理表现出显著的差异，验证了作为假设的命题3．1．。第7章还对特殊归纳和一般归纳的差异和类同作了分析，认为，特殊归纳和一般归纳的差异是由人对具体对象或抽象对象的表征差异导致的。总体上，两者只有量的差异而没有质性区别，所以在大多数研究中可以忽视两者的区别。从第8章到第11章，6个实验(实验6～10和实验10B)验证了命题3．2。这一部分运用了反应时、眼动以及信号检查理论等研究技术。所有的结果都支持：在利／害条件下，人们的归纳推理分别采用了两种不同的论证逻辑。最后，在第12章的总讨论中，本文提出一个整合性的理论假说：获利和避害是两种由自然选择塑造的先天的归纳推理机制：系统的科学教育能够培养出一套符合科学理性的推理习惯。因此，对于一个受过教育的人来说，可能拥有三套推理模式．它们在个体的日常思维中分别占不同的权重，因而形成了人的不同的思维风格。因此，只要测量人的三种推理模式的权重，就能够描述个体的思维风格、个性心理特征，以及预测其社会行为。此外，不同推理模式的假说能够解释许多社会冲突现象．更多还原

【Abstract】 This paper contained 12 chapters totally.In Chapter I,philosophical and logicalstudies about inductive problem were recounted.And in Chapter 2,psychological studiesof inductive inference were reviewed.At last,the conclusion being established was that wecould not set up a domain general model about inductive inference.So the article erected abasic hypothesis:Proposition 2.1.Inductive inference are domain specific mental machines shaped bynature selection.The demonstration of Proposition 2.1 was made up of two parts.The first part contained Chapter 3.In this chapter,the basic hypothesis wastransformed into two middle hypotheses.Guided by the paradigm of EvolutionaryPsychology,the transformation was achieved within 4 steps:First step:to explore what kinds of adaptive problems could our ancestors meet.Second step:to consider if it is possible that such adaptive problem could be dealt byspecific machine.Third step:to explore what kind of inner logic did the specific machine run follow.Fourth step:check-up the cognitive process of modern people,to examine if thespecific inner logic do exist.Among the fours,the first and second step belonged to theoretic demonstration,andthe third was to erect the empirical hypothesis,and the fourth was to validate thehypothesis by empirical methods.The Chapter 3 stopped at the third step.At first,“inductive inference was an adaptive problem”had been argued.From thisproposition,a new conclusion was deduced,that was:“inductive inference were domainspecific mental machines”.On the other side,“gain-benefit”and“avoid ing-injuring”weresuggested to be most important adaptive domain.Thus,the middle hypotheses wereestablished.They were expressed as two propositions:Proposition 3.1.There would be different inductive inference performance under the conditions of“gain-benefit”and“avoiding-injuring”.Proposition 3.2.There would be different inductive logic under the conditions of“gain-benefit”and“avoiding-injuring”.Under the condition of gain-benefit,thereasoning would be sufficient and hecessary:but with the condition of avoiding-injuring,the reasoning would just be sufficient.The second part of demonstration was composed by Chapter 4～Chapter 11.Chapter 4 introduced the background information of experimental study which wouldbe carried out.They included:1)the definition of independent variables and dependentvariables;2)the method to master the controlled variables;3)and how to recruit theparticipants.In Chapter 5,two experiments,Exp.1 and Exp.2,had compared the different effectsaroused by category and similarity when inductive inference were engaged.The resultsshowed that as to make inductive inference,people would like to depend on information ofcategory more than that of similarity.In Chapter 6,Exp.3 and Exp.4 had distinguished different effects of inductionbrought by amount of premise and variety of premise.The results showed that people weresensitive to premise-amount and tended to ignore premise-variety.The purpose of former studies were to define the horizon of the whole research.Therewere two important meanings with the conclusions:One:the hypothesis of domain specificity got a chance to be validated,because theadaptive domains were distinguished by category.Two:it was possible to be sure that inductive inference could be an innate faculty,forevidences coming from infant cognition studies had show that very young infant hadexhibited the ability to realize amount.In Chapter 7,Exp.5 had discovered that the performances of inductive inference wereentirely different as under conditions of“gain-benefit”and“avoid ing-injuring”.The resultvalidated one of two middle hypotheses,that was Proposition 3.1.From Chapter 8 to Chapter 11,there were 6 experiments to be carry out todemonstrate Proposition 3.2 In this section,reaction time,eye movement,and signaldetection were used.All the results revealed that there were two different information processing to match the conditions of gain-benefit and avoiding-injuring.And also,underthe two conditions,people’s reasoning followed two different logics.The conclusionsupported Proposition 3.2.When two middle hypotheses had been demonstrated,the basic hypothesis,that wasProposition 2.1.,could then be validated.At list,an integrative theory was put forward in Chapter 12.It suggested that gain-benefit and avoiding-injuring were two kinds of innate inductive inference machinesshaped by nature selection.But systemic science education could cultivate a scientificreasoning habit.Thus,there might be three kinds of reasoning fashions,and they wereendowed with different weight in people’s thinking.This theory could explain manyphenomena of conflict in social behavior.It also suggested that if we measured the weightof three reasoning fashions with inductive inference tasks,the result would be used as anindex about people’s thinking style.The index may be named as“inductive style”.It couldbe used to describe people’s personality,and also to predict the other tendency of socialbehavior.更多还原