Seek and ye shall find: test results are what you hypothesize they are

Expert clinicians were given batteries of psychodiagnostic test results (Rorschach, TAT, Draw-A-Person, Bender-Gestalt, Wechsler) to analyze. For half, a battery came along with a suggestion that the person suffers from Borderline Personality disorder, and for half, that battery was accompanied by a suggestion that he suffers from Paranoid Personality disorder. In Study 1, the suggestion was made indirectly, through a background story that preceded the test results. In Study 2, the suggestion was made directly, by the instructions given. The experts saw in the tests what they hypothesized to be there. In particular, the target diagnoses were rated higher when they were hypothesized than when they were not. Copyright © 1998 John Wiley & Sons, Ltd.     

迪昂问题及其解决

迪昂问题是当代科学哲学争论的主题之一。波普尔、拉卡托斯和库恩都曾经对此问题给予不同的对待或解决。本文对此背景给以扼要的介绍,然后着重介绍和评价一种新的解决方案——基于贝叶斯方法的解决。相比之下,后一方案较为令人满意。     

Early Examples of Resource-Consciousness

As with the development of several logical notions, it is shown that the concept of resource-consciousness, i. e. the concern over the number of times that a given sentence is used in the proof of another sentence, has its origin in the foundations of geometry, pre-dating its appearence in logical circles as BCK-logic or affine logic.     

蔡元定对河图洛书的区分——兼论蔡氏父子的范数之学

蔡元定主张"河十洛九"、"河先洛后",与刘牧的"河九洛十"、"河洛同出"说相反。蔡元定强调河图与洛书的区分,与朱熹的"虚中为易、实中为范"说企图融合河图、洛书也不同。其目的是加强十数图与《周易》八卦的关联,九数图与《洪范》九畴的关联,进而将《洪范》纳入易学系统。其子蔡沈的《洪范皇极》以《周易》为象学,对应河图,以《洪范》为数学,对应洛书,正是这一目的的体现。这一区分是对刘牧河图、洛书说的发展,蔡氏父子由此构建了其范数之学。     

易学话语与伊斯兰教中国化——论《天方正学》对易学话语的利用问题

明清时期,伊斯兰教汉文著述大量出现,汉语伊斯兰教在很多非基本信仰问题的思考与表述上,表现出明显的中国特色。清代后期,以蒋湘南与蓝曦为代表的伊斯兰教学者,系统利用熟悉的易学资源阐释汉语伊斯兰教思想。其中蓝曦在《天方正学》中,发展了王岱舆、马注等人已经开始的以回释儒的传统,大量利用卦气说、伏羲画卦说、易象说与太极化生宇宙说一套易学话语系统,作为建构汉语伊斯兰教思想的重要资源,这可以看作易学在清代的一个发展。这种情况的出现和中国思想史上格义的传统、易学强大的阐释空间、伊斯兰教思想和中国传统思想本身的沟通之处、汉语伊斯兰教思想明显有苏非气息等因素有关。     

意见收敛定理与休谟问题

意见收敛定理是主观主义概率论的一条重要定理,它表明随着证据的增加,验前概率的主观性将被验后概率的客观性所代替。意见收敛定理被看作主观概率的动态合理性原则,因而被用来解决休谟问题,即归纳合理性问题。然而,哈金有说服力地表明,意见收敛定理证明的是条件概率Pr（h／e）的收敛,而不是验后概率Pre（h）的收敛。主观主义概率论暗中接受的一个等式是：Pre（h）=Pr（h／e）,通常称之为“条件化规则”。这样,归纳法的合理性问题变成条件化规则的合理性问题。为此,本文提出一个新的合理性原则,即“最少初始概率原则”,将它同“局部合理性”观念结合起来便可为条件化规则的合理性加以辩护。     

The Epsilon Calculus and Herbrand Complexity

Hilbert's ε-calculus is based on an extension of the language of predicate logic by a term-forming operator e _x . Two fundamental results about the ε-calculus, the first and second epsilon theorem, play a rôle similar to that which the cut-elimination theorem plays in sequent calculus. In particular, Herbrand's Theorem is a consequence of the epsilon theorems. The paper investigates the epsilon theorems and the complexity of the elimination procedure underlying their proof, as well as the length of Herbrand disjunctions of existential theorems obtained by this elimination procedure.     

动机的概率意义——概率归并定理与Monty Hall问题的非贝叶斯解

本文探讨Monty Hall问题认知困惑的形成因素。通过对该问题数学结构的分析,导出一个与该问题的解密切相关的定理——概率归并定理;并依据该定理给出了Monty Hall问题的一种更直观的数学解。     

An impossibility theorem for verisimilitude

In this paper, we show that Arrow’s well-known impossibility theorem is instrumental in bringing the ongoing discussion about verisimilitude to a more general level of abstraction. After some preparatory technical steps, we show that Arrow’s requirements for voting procedures in social choice are also natural desiderata for a general verisimilitude definition that places content and likeness considerations on the same footing. Our main result states that no qualitative unifying procedure of a functional form can simultaneously satisfy the requirements of Unanimity, Independence of irrelevant alternatives and Non-dictatorship at the level of sentence variables. By giving a formal account of the incompatibility of the considerations of content and likeness, our impossibility result makes it possible to systematize the discussion about verisimilitude, and to understand it in more general terms.     

拥有关系信息情境模型建构的影响因素

该研究修正了Radvansky1997年的实验设计 ,运用扇效应技术系统地探讨影响建构拥有关系情境模型的因素。共包括 4个实验 ,被试为 180名一年级本科生 ,研究材料是“1人 /多物”与“多人 / 1物”两类表示拥有关系的句型 ,每类句型都有 3种扇水平 ,用计算机逐个呈现句子 ,要求被试迅速再认作出按键反应 ,用MANOVA分析了被试对不同扇水平的两类句型的反应时。实验 1在中文条件下重复了Radvansky1997年的实验 ;实验 2探讨在明确的相同空间条件下 ,对“多人买 1物”拥有行为句式的整合是否表现出扇效应 ;实验 3探讨在信息采用顺向展开的情况下 ,“多人买 1物”拥有行为句式的整合是否表现出扇效应 ;实验 4进一步探讨在具有相同时空状语的条件下 ,“1人有多物”与“多人有 1物”两类拥有状态句式的整合是否表现出扇效应。结果表明 ,当具有明确的时空信息情况下 ,或者当信息顺向展开的情况下 ,“多人买 1物”的拥有行为句式的整合均没有表现出扇效应 ;而在有相同时空状语情况下 ,“1人有多物”的拥有状态句式的整合也没有表现出扇效应 ;说明这些句式可以整合为同一情境模型 ,与Radvansky的研究结果不同。研究结果初步表明 ,享有共同概念的若干事件信息是否整合为同一个情境模型 ,受到动词的动作性程度、是否有相