【作者】 王卓娅；

【导师】 汪堂家；

【作者基本信息】 复旦大学 ， 外国哲学， 2010， 博士

【摘要】 自古希腊时期以来,探索思想的方法一直是哲人关心的问题。从柏拉图时期开始到19世纪,公理法—其中也不断受到质疑—被视为思考世界和理解世界的科学方法,如果没有公理,人们无法想象科学如何存在,因此包括理解在内的思想的方法被视为应该是逻辑的、有条理的。然而,对公理法的乐观估计被由施莱尔马赫引入的“解释学循环”概念打破,这个概念以“整体-部分”的循环关系揭示了潜藏于人类理解中的那种循环性,使得为思想在方法上找到一种类似公理的初始成为一种虚妄,与此同时,逻辑主义也在科学语言中发现了这种循环,使得科学语言的基础成了问题。解释学与逻辑主义在“循环性”上的相互认同使得许多科学哲学家们放弃了逻辑地、有条理地重构思想的计划,仅有为数不多的科学哲学家们仍然认为这个计划是可以实现的,由德国著名数学家、逻辑学家保罗·洛伦岑(Paul Lorenzen)和著名音乐学家、历史学家海德格尔的学生威尔海姆·卡姆拉(Wilhelm　Kamlah)以及他们的学生在20世纪60年代共同创立的“爱尔兰根建构主义学派”就是其中之一。与其它建构主义学派相似,它继承了康德批判“客观主义”的精神,认为在人类实践中已被给定的东西—例如,语言—是被建构的,而且这种建构必须是无循环的(without circles)、无间断的(without gaps)以及循序渐进的(step by step),它的特殊性就表现在它计划用分析语言的现代逻辑工具,以日常语言为基础,有条理地构造科学的语言,承诺解决科学的初始问题(Anfangsproblem)和根据问题(Begrundungsproblem)。虽然这一立场一经提出就招致诸多批判,例如,以波普尔为代表的批判主义和以哈贝马斯为代表的批判理论,前者更是提出了“明希豪森-三重困境(Munchenhaus-Trilemma)”证明放弃有条理地给出根据的正当性,但是迄今对爱尔兰根学派的计划还未能作出一种全面的批判,在此背景下,笔者尝试要让一种“全面的批判”可行,这种批判不是排斥爱尔兰根学派构造科学语言的建构主义方法,相反,笔者认为这种建构主义的方法是相当有效的,只是因为其中掺杂着诸多无用的内容而丧失了它的明晰性和普遍性。为了使爱尔兰根建构主义方法获得普遍性,笔者将采取的批判路径是：首先描述爱尔兰根学派的建构主义方法,然后考察这种方法是否能够实现它自己的方法承诺,最后将爱尔兰根建构主义方法整理为一种语义学的构想,让方法自身变得明晰、有序。除了第一章的引言和第五章的结语之外,文章的主体将分为三章。第二章描述对爱尔兰根学派建构科学语言的方法。根据作为具体构造理想语言(要根据建构主义方法才可被构造的语言)起点的那些要求：(1)无循环性和循序渐进地展开,在承诺诸构造步骤之间存在着某种关联和某种序的情况下使第一步成为“等语域下(没有元语言与对象语言之分)不可再给出根据的”；(2)通过指出每一步的适用范围和在此适用范围之内所力求达到的一些目标,为每一步的正当性进行辩护,这使得一种被成功构造的理想语言还可以证明与初始不同的其它的语言构造阶段是“可再给出根据的”。鉴于上述要求,构造理想语言的“第一步”是述谓行为,它体现了作区分的语言能力。在将述谓行为确定为构造理想语言的初始之后,为了能够合理地述谓并因此能够在语言上分辩出恰当的目标对象,必须首先弄清楚,所使用的述谓词是什么。爱尔兰根学派将述谓词规定为两部分：第一部分是不易混淆的声音形式和符号形式,由此得到作为信号的述谓词。第二部分是由此信号所传递的“含义”：根据无循环性要求,既不允许诉诸用理想语言表达的“含义”来规定含义,也不能使用在其它语言中的固定含义,因而认清述谓词含义的方法类似于在教小孩说话时所实际运用的方法,而不是通过翻译。在行为图示可被教授,并且可被习得的前提下,为了在等语域下获得作为语言初始的述谓词的含义,爱尔兰根学派模拟出一种老师-学生-情境,由此引入一个新概念“对话情境”。爱尔兰根学派用射箭—实例来表明在一种用于确定初始述谓词含义的对话情境下会发生什么,由此引入“典型确定的述谓词”,它们是由作为例证的各类情境与作为反例的各类情境而不是诉诸元语言来限定含义的述谓词在将以典型确定的述谓词为元素的有限集和证实一个述谓词φ的正例情境集与证伪它的反例情境集的并确立为“可靠的”因素之后,接下来要确定的是各种典型确定的述谓词之间的一种潜在关系,从而为一些已经与述谓词φ直接发生关联的个别情境增加一些可与之间接发生关联的个别情境。这种关系,虽然可用一种集合论的语言来表达,但是在老师-学生-情境下,老师是以非语言的方式准确强调某些实际的命令或禁令,从用一个述谓词φ述谓一个目标对象ψ转向用另一个述谓词φ*述谓ψ,这样做的意义在于使学生们不参照元语言的描述就能习得述谓词关系,由此引入了等语域下不可被表达但可以被实施的“述谓词规则”。根据建构主义者们的信念,与仅仅通过典型的规定相比,通过执行一个述谓词规则集,所意向的述谓词用法会得到更准确的规定,爱尔兰根学派称那些由述谓词规则规定用法的述谓词为“术语”,但是这种信念的正当性需要由“超验-担保”来辩护,即,超越那些有限多的情境,这正是爱尔兰根学派所未能澄清的地方。基于被确定为“可靠的”术语集,爱尔兰根学派的“概念”是从那些基于已确立的规则系统(正例,反例以及述谓词规则)在含义方面是等价的术语中抽象出来的。原子命题产生于述谓词与命名词项之间特定的关系,因为述谓词已被视为构造上可靠的,所以接下来的步骤要确保获得诸命名词项并且确保其与那些可用的述谓词能够联结在一起。为了从一开始就有方法保证,爱尔兰根学派选择放弃诉诸发展心理学的事实,仅明确把作区分的能力赋予所构造的“爱尔兰根的非凡者”,使“爱尔兰根的非凡者”有能力学习并且实现诸多行为图示。然而成问题的是,虽然可以充分“可靠地”假设老师和爱尔兰根的非凡者之间达到了诸多含义共性,但是很难确定为可靠的是,从指示行为转向这一指示行为的目标对象本身,因为要获得一个名称,必须从行为图示“指向一个目标对象”转向目标对象,为此很可能还要诉诸一种“超验-担保”,即,理想情况下可能就恰好存在一个目标对象,它能够充当证明某个属于“专名-称为”类的述谓词的肯定用法的实例。仅在确定了学生能够实现这种超验之后,才能把各个明确的限定性摹状词和专名称作“命名词项”。通过诸多命名词项,单个非语言行为在下一步得到命名,并且再度被理解为证实一个非语言行为图示的正例情境。从而可以在语言上将那个表示某种非语言行为图示的述谓词判归给由命名词项命名的某个非语言的行为,由此获得原子命题。下一步就是应用逻辑方法使诸相关规则得以贯彻从而规定,如何可以使某些命题相互关联,以便从原子命题过渡到复合命题。在对话情境中,学生要习得的不是另一些述谓词而是将命题联结在一起的逻辑小品词,保罗·洛伦岑(Paul Lorenzen)给出提示说“命题基本上用于对话”,于是,如何使命题发生关联的问题对学生而言就转变为如何对一些命题进行反驳和为一些命题作辩护,这在其之前习得的言语行为清单中找不到任何参考信息。为了使学生能够从判定和否定述谓词的行为转向为命题作辩护和反驳命题的行为,爱尔兰根学派假定学生以某一种方式获得为进行断定和反驳而必不可少的能力。笔者在文中尝试重构保罗·洛伦岑(Paul Lorenzen)提出的对话逻辑学,从而在确保明确表达其中概念、规则的同时避免繁冗陈述。由爱尔兰根学派建立的对话语义学称赞“直觉主义逻辑学”是真正的逻辑,由此介入到自布罗威尔以来持久存在于“经典派”与“直觉主义者”之间的争论。可证明正当的是,对话语义学通常在条理性上要胜于语义学,因为在等语域的应用训练中规定含义就可避免那种“含义转移”；不可证明正当的是,仅通过根据爱尔兰根学派的方法所构造的关于直觉主义逻辑的对话证明选择直觉主义逻辑是真正的逻辑,因为经典逻辑同样可以对话的方式被重构下一步要考虑的是如何为自然科学的语言奠基,以物理学为例,在此出现的复杂情况是虽然仅以对话的方式也能够判定某些物理学语言的命题真值,但是更可能发生的是,一旦有怀疑产生为了进行完全确定的测量会中断这一对话,此外,测量首先并不涉及表达对诸命题的反驳抑或为之作出的辩护,而是涉及到构造实验装置和某些仪器的读数,这将导致对诸多所谓“实质”科学的有条理的构造必须包括那些非言语的测量行为而且也必须确保获得这些行为的“初始”,爱尔兰根学派以“原物理学”来实现这些构造要求。这一章最后,基于在方法上确凿无疑地构造科学语言的目标,爱尔兰根计划要引入建构主义的伦理学并且确立文化释义的方法,而且这种取向可被证明是严谨的。第三章要解决的问题是允诺要解决科学中的初始问题和根据问题的整套爱尔兰根学派计划是否以及在何种程度上得以实现,并且由此可能会得出什么结论?要明确爱尔兰根计划如何解决初始问题,就需要明确如何对那些构造专有术语含义的表达与那些使用专有术语的表达作出区分。根据无循环性要求基本上不允许使用元语言来作出区分,但可设想通过将一个陈述句与表达其所处特定语境的语式相关联来表达命题,其中这种表达语式借助于语力行为就会识别出旨在构造含义的命题。然而,就标记命题是否旨在构造含义而言,爱尔兰根学派命题的语力行为并没有起到一种分类作用,因为在对话语义学范围内所涉及到的语力行为,即,断言和反驳,显然不适用于标明命题是旨在构成含义的。爱尔兰根学派恰恰是通过模拟老师和学生向他们的学生澄清,一个命题旨在构造含义,另一个命题则并非旨在构造含义。继而要澄清的问题是旨在构造含义的诸语境对理想语言的真值概念会产生哪些影响?在这个问题中所引入的分析性概念旨在对仅在“与经验的对话中”才可判定真值的那些命题与仅由于被使用的逻辑学和分别出现在其中的专有术语含义才可判定真值的那些命题之间作出区分。换句话说,这个概念带来了语言的相对化,将它引入一种科学语言可使人们能够对“语言眼镜”,由此人们看到“世界”,与透过这个“眼镜”可观察到的“实际发生的事情的经过”作出区分,尤其是在经验科学中这样一种可区分性使其语言中迄今未受到反驳的那些“假说”与言说这种语言的规范得到区分。最终,根据上述分析性概念,不需要描述那些构造含义的语境就可在根据爱尔兰根建构主义方法构造的语言中刻画此语言的“初始”：可在分析命题中找到“语言的初始”。爱尔兰根计划的根据问题要追问的是如何设想一种“非演绎”的根据概念,为描述事态的东西给出完全“充分”的根据。为此笔者要采取的研究策略是：首先鉴于建构主义方法论阐明本质上非演绎的爱尔兰根学派的根据概念,其次阐明借助这种根据概念如何能够为在爱尔兰根计划中提出的那些要求提供正当性证明,最后将说明爱尔兰根学派如何破解上述根据问题。在爱尔兰根建构科学语言的方法中出现的根据概念既有在建构主义伦理学中出现的有关行为、目的、目标的正当性证明的“三层根据(上层建筑)—需要(底基)概念,也有在对话逻辑中提到的、与导致三重困境的根据概念相对峙的、旨在判定内容性命题真值的辩护方法,不过更可取的是对两者的居中解释,即,一种根据现在可被理解为一种合理性的对话,这种对话可使所有参与者达成某种同意,使得在所有参与者那里那个受争议的观点在一种为此观点所模拟的不偏激的交往情境下可得到普遍认同。在下一步考察这种“合理性的对话”如何为判定命题真值给出根据。由于在科学语言中有待被给出根据的是初始命题,于是兴趣将集中在这种“合理性的对话”如何为初始命题的真值提供根据。从保罗·洛伦岑(Paul Lorenzen)的论述可知,为了在一种以建构主义的方式构造的语言的前提下说明所要求的初始命题的判定过程,需要模拟一种老师-学生-情境,这种旨在作为范式的初始命题的真值事实上是由教师的权威判定的。由此可设想的是,给出根据的基础隐藏在那些由老师实际上所遵循的意图之中。因此,有必要考察借助爱尔兰根学派提出的另一个根据概念“证明目的、行为、决定正当性的理性原则—道德原则—需要”如何为赋予老师权威的约定以及证明这种约定的决定给出根据,而且进一步要考察理性原则—道德原则—需要本身是否有充分根据?鉴于爱尔兰根建构科学语言的方法,考察作为“底基”的需要是否有充分根据就是对主观要求与真正的需要作出区分,这将诉诸由老师拟定的那些要求的清单,它们在爱尔兰根学派建构的语言中是“真正的需要”。考察作为“上层建筑”的理性原则和道德原则是否有充分根据在于“无条件服从的要求”,具体而言,这两种超规范在理想语言中被命名为“原则”,不会被视为要在对话情境下被讨论的建议,而是理想语言的初始,不可被给出根据以及不需要被给出根据是这些原则的特征,在老师-学生-情境下,学生承认这些原则被视为信仰行为,具体而言,相信老师权威的行为。在论述过建构主义给出根据的过程之后,可以对爱尔兰根学派如何解决根据问题作出一个初步的回答：给出根据总是局限于说明所属的那个初始,使得其“充分性”仅仅是一种相对于相关科学语言的充分性。然而需要注意的是爱尔兰根计划中的“给出根据”的术语还未与波普尔与阿尔伯特(Albert)在确立明希豪森—三重困境时所设想的“给出根据”相关联,尽管它与阿尔伯特(Albert)的三重困境在无循环地以及循序渐进地构造科学语言方面有论战发生,因而还不能把上述通过解决语言初始问题的方法得以解决的问题理解为哲学的根据问题,而是应该把这种问题理解为一种奠定基础的问题：在方法没有出现错误的情况下,可如何“为语言奠定一种基础”。其次,爱尔兰根学派为证明构造科学语言的各步正当性而设立的对话情境是非民主的,学生在给出根据时至多可诉诸他在构造含义的情境中正是被这样教导的,老师在给出根据时暗示自己知道为何他们给学生讲授这些知识,而不是其他知识,因而老师们的“原理性知识”与学生们的“程序性知识”间的基本差异可能会对达到含义共性、构建客观对称的言说者情境以及超越主体性等目标产生负面效果。除此之外,建构主义者们分别通过“重构意向”、“对话合理性”以及“参与者和受影响者的合理性”赋予其“根据概念”更多层次。第四章基于爱尔兰根计划可还原为对某类语义学的构想,引入第一语义学,并指出爱尔兰根的科学语言构造计划是第一语义学的模型。由于在爱尔兰根计划中还混杂着诸多无用内容,例如,各种世界观,因此还无法被视为一种精密的、普遍可行的为科学奠基的理论,此外,爱尔兰根计划旨在达到决定一种语言是否有意义的目标—含义共性,在通常是由语义学解决“含义构成”问题的前提下,可把爱尔兰根计划的合理性之核视为对某类语义学的构想。这种语义学有别于目前被视为语义学范式的解释语义学(第二语义学),而是在抛除了第二语义学的局限性(因为由元语言规定含义和为含义给出根据而引发的无穷倒退)之后对三个“语义学问题”作出有效回答的第一语义学。第一语义学的精妙之处在于它对语言与事实关系的构想,以“我们所看到的事实和语言向我们展现的一样”①为出发点,与通常为科学哲学所偏爱的那种符合论的信念相反,仅当人们考察了与“事实”的“含义”、“属性”等相关的语言游戏,才‘可解释什么是事实。因为,在第一语义学中,那些语言游戏首先规定了作为词项的“事实”是如何被使用的并因此对“事实”的含义作出规定。为了说明爱尔兰根的科学语言构造计划是第一语义学的模型,需要分析构成第一语义学的各个要素,并与爱尔兰根构造科学语言计划中的要素一一对应。在第五章,爱尔兰根计划所提出的要求,即,要以可靠的和有根据的方式构造科学语言,可担保这种科学语言的合理性以及科学本身的合理性。这个要求是否可能实现取决于如何理解“有根据的”与“可靠的”：鉴于要为相关语言奠定无循环的和没有无穷倒退的基础,“可靠的”旨在在方法上有一个准确的语言初始,“有根据的”旨在使科学在逻辑上得以成立,而用第二语义学构造科学语言是无法满足这两个要求的(总是诉诸元语言的翻译而导致无穷后退),因而爱尔兰根学派用第一语义学来构造科学语言,是因为它洞见到语言游戏中一系列有效信念与其说表现在某些孤立的命题中,不如说表现在相关言说者在他们的整个交往过程中赋予这些命题的状态之中。按照第一语义学重构科学语言既会要求标记出参与者们在语言游戏中所持有的信念,同时会要求参与者们认识到这种信念本身,即,他们的“语义的眼镜”。论文最后回到使爱尔兰根建构主义方法与批判理性主义的方法分道扬镳的那个问题：明希豪森-三重困境(Munchenhaus-Trilemma)，通过模拟两者的对话,揭示他们之间在此问题上的根本分歧。更多还原

【Abstract】 Since the ancient Greek times, what philosophers most concern is a method of thinking. From Plato’s time to the 19th century, the axiomatic method was considered as the scientific method to think about the world and to understand it. If there are no axioms, it’s hard to imagine how science exists. So the method of thinking, including understanding, should be logic and methodical. However, it is "the hermeneutic circle" which shatters such optimistic estimate of the axiomatic method. The hermeneutic circle reveals the circularity lurked in the human understanding by describing the circular relations between "the whole" and "the individual parts". As a result, it seems impossible to find a methodical beginning of thinking. Meanwhile, the Logicists find such circularity in the scientific language, which makes the foundation of the scientific language as a problem. Hermeneutic and Logicism coincide just at this point. Most of scientific philosophers renounce the attempt to reconstruct the thinking methodically, and only a few of them still try to fulfill this reconstruction, one of them is "Erlangen Constructivism". "Erlangen Constructivism" was found in the 1960s by Paul Lorenzen, who was an important mathematician physicist and logician, and Wilhelm Kamlah, who was a famous musician and historian, and their students. As the other constructivists, the Erlangen constructivists, who have inherited the critical spirit of Kants’, believe that what is given-for example, language- in praxis should be constructed and such construction should be step by step, without gaps, and without circles. It is the particularity of Erlangen Constructivism that it plans to construct the scientific language methodically on the basis of ordinary language by means of the modern logical analysis and then promises to solve the problems of Beginning and Grounding. Although it was criticized by other philosophers, for example, the critical rationalism, which put the Miinchenhaus-Trilemma forward as the argument for renouncing to give grounds for science, and the critical theory, very few philosophers, could successfully present a comprehensive critique of Erlangen Constructivism. What this essay wants to do is to make a feasible and comprehensive critique of Erlangen Constructivsm, which does not deny its constructive method but make it clear and generalized. For this purpose, the essay will be arranged in this way:It is the first step towards making a general description of the methods, field of application and technical details of Erlangen Constructivism. It is the second step towards examining, to what extent the concepts which are presented actually, are consistent with its promise. It is the final step towards systemizing the method of Erlangen Constructivism as an idea of a kind of semantics in order to make this method orderly and clear.The main dissertation consists of three chapters, in addition to the introduction and the Epilogue.Chapter 2 is the reconstruction of the essential part of the Erlangen philosophizing. On the contrary to most sciences nowadays, which have already abandoned efforts to find the grounds for themselves because of the Munchenhausen-Trilemma, the Erlangen Constructivists claim to trace the sciences again to a justified ground. They have attached great importance to the postulates of being free from circularity and developing step by step, which make them particularly interested in "the reliable beginning of all reasonable speech".Thereupon, to avoid this aporia, i.e. it is "always already" spoken and thus the beginning of language is uncatchable, the Erlangen Constructivists plan to construct the beginning of language "a primis fundamentis"(from the beginning) methodically. In these circumstances, it should be in an "isohyptical"(proceeding on the same level of language) way to establish the beginning of language and make further developments, or else people will refer to all kinds of ordinary languages, and as a result they will violate the postulate of being free from circularity.In order to guarantee the commonness of meaning isohyptically, the Erlangen Constructivists have suggested to train the participant, who feigns being speechless for the moment, to use the proper terms of the ideal language (orthosprache) in the correct way, which corresponds to the learning process of a first language explained by the Behaviorists. For some not very obvious reasons, the Erlangen School marked some of the learning contents preferable; however, it can’t become a good justification for such markedness by making a forced isohyptical impartation of them.Spoken in details, the first step is to presuppose that we can not further give reasons for the ability to differentiate, and that its derivatives-the ability to behavior and the ability to predicate-are the basis. In reference to the learnability and the teachability of the schemata of the speech acts and the non-speech acts it is possible to practise how to perform the elementary acts of predication, wherein the meanings of those predicators are first defined exemplarily by the production of the set of examples and the set of counter-examples. By the rules of predicators, the connection between those exemplarily defined predicators specifies the predicators as terms, and the acts of abstraction regarding synonymity make talking about the concepts possible. Thereafter, by means of the schemata of the acts of pointing, the targets are isolated for the acts of predication, and then are represented by the nominators in the language.Finally, the elementary propositions are constructed by using the nominators, predicators and the illocutionary acts of attributing and denying.After the construction of the linguistic elements it will begin to establish the illocutionary acts of assertion and thus create the prerequisites for teaching the constructivist logic. It is possible to construct the logical constants by connecting elementary propositions and complex propositions in some way, and the rules of dialogical arguments about the truth value of the assertions determine the usage of these constants.Following the introduction of modal logic, constructive arithmetic and constructive analysis, the next step is to guarantee the "empirical" components of science, especially the possibilities of measurement. Geometry marks the usage and productive conventions of the first measuring instrument (Alphagerat), which are identified as the first operationalisations methodically, as non-empirical theories, where according to the constructivist principles Geometry can also develop from the "bottom", that is to say, from the praxis of the Lebenswelt. It’s time to think of those theoretical postulations regarding the structural features of geometrical bodies, i.e. the principles of homogeneity, when people want to guarantee that the first basic geometrical bodies that are produced can behave constantly. But the Erlangen Constructivists haven’t clarified how to deduce these principles "from the bottom"Finally, Erlangen Constructivism focuses on those propositions which are implicitly based on "the last resort", i.e. the teacher’s unfalsifiable instructions, of grounding. The constructivist ethics, which is reconstructed in the above-mentioned way, takes the following basic idea as a starting point, that the suggestions for norms are based on the desire to make the aimed circumstances happen. As a result, such reasonable suggestion has to consider these two questions:why is a suggested aim worth pursuing? To what extent can a suggested norm be a means of achieving that aim? If there is a conflict between the aims, then the consultation has to be made according to the rational principle and the moral principle, where the first superprinciple should regulate the formal conditions of the argumentation, and the second one should rule on how to solve the conflicts between the supernorms. In this context Erlangen Constructivism suggests to choose the basic norm as a way out to guarantee the satisfaction of wants. However, due to the opacity of the concept of wants, to deliver on this way out "righteously", Erlangen Constructivism should make its way towards constructivist cultural studies.Chapter 3 probes into whether and to what extent the whole Erlangen program, which promises to solve the problems about Beginning and Grounding, is implemented. In the first place, it is necessary to distill the typical characteristics of the beginning, which is established according to the Erlangen prescription, from the research materials and thus we can examine whether those requirements, which are combined with these characteristics, are justifiable. In the second place, this part analyses how Erlangen Constructivism solves the problem about beginning. What the Erlangen constructivists want to construct is a public language, so the beginning, which they insist to inquiry into, is suitable to establish the commonness of meaning for a language. As a result, only "the constitution of meaning" can characterize this beginning, at which point the Erlangen Problem of beginning is converted into the problem about constructing the meaning of a language. The key point in identifying how to construct the meaning of a language lies in how to distinguish the context, which constructs the meaning of a language, and the context, in which a language is applied. Some contexts can be classified as constituting meaning by simulating a kind of teacher-student-situation, which can avoid using the metalanguage to accomplish such classification in order to satisfy the requirement of being free from circularity. The contexts that constitute the meaning of a language are:the logical rules, which can be enforced but can’t be expressed isohyptically; all kinds of operationalisations, which can not be substituted by the isohyptical formulations in principle; and the rules of meaning, which can be formulated isohyptically. In the third place, this part analyses how Erlangen Constructivism solves the problem of grounding. It is important to realize that the Erlanger concept of grounding is used as a technical term in the language involved. If a proper metalanguage exists, then we should read the predicator "is reasonable" as "is reasonable in the ideal language O". Similarly, the predicator "is justified"of a metalanguage can be read as "is justified by the system of norms N". This releases us from doing the specific research into the "deductivity" and "non-deductivity"of the procedure, which the Erlangen constructivists make use of, because at least in principle all of these grounds can be reconstructed as deductive relations; even if the deductivity is queered by appeal to the dialogue-game and the non-speech acts in the practical justification, there is no change in the deductivity of the concept of grounding per se. So the Erlangen concept of grounding is different from the one in the Munchenhausen-Trilemma, let alone the solution of the problem about ground in this Trilemma. However, to sum up, the first problem about grounding for the Erlangen Constructivists is how to ground the truth-decisions of a proposition. In light of the Erlangen prescription, it can be solved by combining the proposition with the beginning of the language involved in a simulated teacher-student-situation, which leads to the second problem of grounding in the sense of Erlangen Constructivism, i.e. how the conventions to which such combination resorts can be justified. According to the Erlangen prescription, still by means of a simulated teacher-student-situation, the teachers, who are able to speak the language involved, distinguish those conventions from the other suggestions, and make them into the beginning. From this, we can say that the term "grounding" in the sense of Erlangen Constructivism can be regarded as such a process, that by means of a rational dialogue those persons involved can reach agreement about some rules, which aim at constructing the scientific languages methodically to guarantee the rationality of science. In this sense, the Erlangen constructivists have indeed solved the problem of grounding, which is not a philosophical problem, but a problem of how to lay the foundations of the scientific language.Chapter 4 discusses what makes such a science, which has already been applied in practice, obligated to correct itself according to this reconstruction, which has been implemented in accordance with the regulations of Erlangen Constructivism. For this purpose, it is necessary to clarify what these regulations are directed toward. Because the postulation, i.e. to construct the science methodically by means of the scientific language, which has a prescriptive intention under the Erlangen programme, is combined with some restrictions when it is realized as a model, which are more "content-related" and thus can not be justified only by the maxims of Methodizitat. In the next step, this part demonstrates that those restrictions can be regarded as evidence to show that the Erlangen Constructivism has introduced a dogmatic foundation, which is not necessarily dogmatic in the end. It also points out the very reason why the Erlangen School fails to construct the universal and feasible theory which can lay the foundations for all kinds of sciences:the confusion between the postulation, i.e. to lay the foundation of scientific languages methodically, in content and the one in form. So the reconstruction of a "rational core" in the Erlangen program, which gets rid of those worthless contents, can give an adequate opinion on the normative requirement, which combined with that "rational core":first, it shows that a part of this requirement can be justified only on the basis of methodological considerations; then, it shows that that part of this requirement can be raised undogmatically, only when the other part of this requirement,which refers to those content-related restrictions, is relativised at the same time, to the effect that it is insignificant compared with those methodological requirements. In order to realize the above-mentioned program, it is necessary to explain why we need to lay the foundation of the basic principles of language:to guarantee that we can achieve this goal, i.e. the commonness of meaning, which decides whether a language makes sense. Because if a language can not guarantee that the laymen, who are interested in it, can speak it after a certain time, it is not only useless in science, but also not a language any longer. As a rule, it is the semantics with reference to these languages that takes charge of this problem of the rule-based constitution of meaning. So, more precisely, the thesis of this chapter is that the reduction of those unnecessarily restrictive requirements in content of the Erlangen program to its "rational core" shows this program as the conception of a certain type of semantics, and those steps of constructing a language which are performed by the Erlanger constructivists constitute a model of this type of semantics. For this purpose, the discussion will proceed in this way:the first step is to assert several facts about secondary semantics (Secondarsemantik) and primary semantics (Primarsemantik) in reference to the professorial dissertation of Peter Hinst. The second step is to show a way to construct a language by means of the primary semantics and explain to what extent the "rational core" of Erlangen Constructivism, which lays the foundation for all kinds of scientific languages, can be interpreted as a model of the construction of language by means of primary semantics.Chapter 5 is the conclusion of this dissertation, which reviews the Erlangen postulation, i.e. to construct scientific languages in a reliable and reasonable way and thus justifies the reconstruction of science by means of primary semantics. In the end, it concentrates on the problem, which separates the Erlangen School and Critical Rationalism regarding the ways in which the philosophy of science develops, i.e. the Munchenhausen-Trilemma, and simulates a dialogue between them to reveal the fundamental disagreements between them about this problem.更多还原