Jacob Klein · 雅各布·克莱因
Greek Mathematical Thought and the Origin of Algebra · 希腊数学思想与代数的起源
Undated 未注日期
希腊数学思想与代数的起源——散记
Reflections · Pages
Stevin做到了将数代数化,Descartes做到了将几何同样地代数化,而Wallis则做到了将代数形式重新表述为普遍比例。
Klein那边代数化大概的意思是,将数的总体概念(理念)本身化作一个可操作的数来看待
代数式的解因其在最开始就已经被决定为“确定数目的单子”,其与几何方法就有了区分。将未知项引入几何中,最后通过分析方法所能得到的量还需要再次经历一个综合的过程:比如求得的边能否与其它边搭成一个三角形。而代数则没有这个顾虑,因其解的“不可能性”并不来源于它无法与其它组成部分结合成一个整体,而来源于其性质本身。因此,无法形成三角形的边依旧是非几何的,但求得的负数解却是因其自身性质而不可能的(不可能有小于零的数) —— 也就是说,求得的边是在特殊的情况下并与其它同样特殊的单位经历了特殊的综合过程后才可以被认为无法组成一个“特殊的”三角形,其操作与操作对象远远没有代数具有普遍性。
在这个基础上,Vieta与Descartes的普全数学更进一步。但他们对代数的发展不能被简单理解为对数的抽象,因为“抽象”作为一种方法Aristotle同样使用过,而普全数学的“抽象”的特殊性需要与作为哲学方法的“抽象”区分开来。简单来说,普全数学的抽象所做的是将方法上的抽象个体化、对象化。换句话说,对于Aristotle,并不存在一个总体的数学对象,而数学对象的总体理解就意味着对存在本身的理解。而哪怕Plato与Vieta一样将数抬高到了理念的位置,这里依然需要注意两者根本上的差异。对于Plato来说,数只是存在的形式。换句话说,根本上,Plato是在借助数的特殊性质来讨论存在与非存在的问题。而存在对于Plato来说依然是特殊的诸理念的聚合物;对于Plato来说,一与多之间的张力表现为存在与数之间的张力;在数之上还有等待追问的存在之一。然而在Vieta的普全数学那里,数之上的层面被取消了,并被吸收进了数的内部,即,存在的普遍性本身成为了一个等待处理的“数”,而存在之问的遗忘与退隐就被凝聚在未知数“x”上。对于未知数“x”的象征就已经蕴含了这一层转折,也就是存在被无疑问地视作了一个特殊的存在者,而在古希腊时依然被视为问题的“存在/理念”与“存在者/对象”的关系在普全数学中就被掩盖了。因此,古希腊时具有相当张力的λόγος与διάνοια的关系,也就是“说”与“思”的关系,也被隐藏了。因为通过象征,“理念”被且只能被“说”出,但这个象征在Archimedes处还是一个等待思索的问题,因为“说”与“思”只在实践中被区分,Archimedes仅仅只是意识到在“思”达到一定程度后,“思”需要“说”作为自身的补充根据,但“说”尚未完全地抹去“思”的主权,也没有抹去对理念对象进行充实的要求;因此Archimedes甚至需要为“说”进行辩护。但是在普全数学的时代,“说”完全地替代了“思”,存在之问变动为对数进行操作的技艺。将“说”与“思”的关系的变化与对“数”的理解的变化联系起来,我们可以说,“思”是对理念对象进行具体呈现的过程,“说”则是用语词和符号对理念对象进行替代的过程。而我们在普全数学中目睹的转折是,“思”本身内部所蕴含的理念与理念对象的关系被扁平化,而“思”这个行动本身则成为了被“说”的对象:这恰恰是因为“思”这个行动本身的普遍性要求了符号进行表征。因此,在普全数学中,“思”成为了“说”的一个单元,并因此完全地失去了“思”的独立地位,也丧失了自己的本质。因此,普全数学根本上是“说”的科学,也就是λόγος的科学。只有在这个意义上,我们才能够理解为什么“对存在之思”是对λόγος的冲击,因为λόγος恰恰是不思的。
简洁一点说的话,可以说,在古希腊,“一”与“全”的张力是需要刻画的,但是在普全数学的时代,这个张力被沉淀在未知数“x”的表述中等待展开
Vieta将数学符号化,直接导致了存在的退隐,而数从此不再是外在被给予的确定的存在,而是可以被操作的、具有内在不确定性与随意性的概念符号,其可操作性源于其可变性
“一”与“多”在符号中被沉淀,是因为符号根本意义上是无对象的,而连对象本身都被隐匿了,自然也就无法去探问对象的存在模式了
克莱茵对胡赛尔的批评
克莱茵对胡赛尔的批评并非是我原本以为的技术细节或者内容方面的批评,而反而是在最本质立场上的批评。简单总结的话可以说,在克莱茵看来,胡赛尔太注重连续性了,而恰恰忽视了先验历史中的意义迁移,甚至是意义断裂。
这首先要从胡赛尔的“历史转向”开始谈起。胡赛尔的历史转向似乎是由于一种他所感受到的断裂:他意识到,科学正在陷入一种无法被(他个人)经验理解的危机,而他将这种现代科学的不可理解性归结于超出时间意识的历史性。在时间意识中,一切被给予者都经受着持续不断的变更却保持同一,因此在时间意识中,对事物的纯粹知识论把握是可能的。然而,任何“进入”了时间意识的被给予者必然是在此时间意识之外被给予,而数学与科学对象在这个意义上尤其特殊,因为它们具有一切对象中最高的客观性;这意味着它们的构成是不可能完全内在于(此)意识的。在这个意义上,数学与科学对象必须超出(此)意识而进入交互主体性的领域,作为传统和遗产被交付于(此)意识。而在主体的交互中,意义沉淀而被隐藏,最终导致了现代科学的不可理解;而对主体的意义交互的探问就是对其历史的探问。但是这里要注意的是,对于胡赛尔来说,此处并没有精神分析会强调的“断裂”,因为对于胡赛尔来说,历史问题其实依然是一个知识论问题。交互主体并非是主体与其他者,而是主体与其变更的连续整体;这个整体的意义网络由理念对象的同一性统合。所以,对胡赛尔来说,对历史事实的任意性和偶然性的强调没有意义:历史只是永恒的影子;他进入历史也仅仅是因为传统意义上的知识论被局限在单个意识的内部,而单个意识的时间变更无法独立成为根源罢了。
从这里已经可以看出,胡赛尔对连续性有着非同一般的执着。粗暴点说,我们在根本上可以将胡赛尔现象学的意志整个把握为一种将现代科学连接到根源的生活世界的冲动:当然要意识到生活世界并不是被预先给予的,而是被持续构建起来的。这也与胡赛尔将他从事的哲学工作视作为希腊哲学的目的论继承者有关。
然而,在克莱茵看来,正是这样一种执着使胡赛尔在对现代科学进行具体分析时犯下了根本的错误。对于克莱茵,现代科学是无法与生活世界直接连接的:因为现代科学在根本上与生活世界已经是两种造物。胡赛尔试图将现代科学还原为对基本经验与直观明见性的上升构造的努力之所以止步于对伽利略式的几何化自然的分析,是因为他忽视了现代科学另一个重要的组成部分:普全数学。而胡赛尔对普全数学重要性的忽视则恰恰来自于刚刚分析过的他对连续性的执着。这种执着使他无法脱离历史反思/回溯探问的方法论,而对于根源只能采用那种“之字形”的迂回探询;而“之字形”的探询依然以当下现代科学的事实为基本条件,将后见之明强加到了当时的历史境遇中,而无法注意到在现代科学的意义沉淀中已经包含了一个根本的意义迁移,即普全数学的诞生。克莱茵认为自己的数学研究彻底化了对普全数学起源的再激活,因其方法并非是回溯性的,而是前抛性(proactive)的,因此避免了像胡赛尔一样将现代科学的前设(即普全数学对自然世界的直接替代)忽略。正因为胡赛尔依然认为普全数学是对自然世界的直接替代,他的转向仅仅达成了对普全数学和自然世界的连接,并认为两者之间有直接指涉的连续关系;而克莱茵(通过具体的历史分析)则意识到,普全数学所替代的恰恰不是自然世界,而是希腊人的理念世界,而注意到了科学之历史演进中隐含的关键断裂。因此,与胡赛尔不同,尽管克莱茵(先于胡赛尔)实际上也只采用了“之字形”的迂回探问的方法,他却发掘出了另一种可能的探询方式,即对历史切面的普遍时段性探究,而这也是标准的历史学研究常用的方式。在这个意义上,克莱茵对胡赛尔似有超出。
以上都是Hopkins对两者的分析,不代表我的个人看法。根据Majolino的批评,Hopkins在根本上误解了胡赛尔的individuum概念,并将其偏狭地视作“个体经验对象”,而得出了“胡赛尔哲学是将现代科学连接到自然世界上的尝试”的结论。然而,根据Majolino的解读,individuum完全可以指涉具有独一性的理念对象,而这就意味着克莱茵将普全数学奠基于对希腊人理念世界的改造的努力并非是完全独立于胡赛尔影响的。
至于说为什么普全数学诞生时,奠基者如笛卡尔、Vieta和Stevin在认为自己构造的是唯一的科学的同时还将自己理解为古希腊传统的继承者,而后继者们则遗忘了后面这一点,转而只把自己的科学当作自然世界/实在的唯一真理,为什么后者沉淀下去了,那就不是克莱茵要操心的事情了。
Precision and exactitude: with Plato, the precise is itself the One or the Good. But with modern mathematical physics, the precise is good only as long as it produces an exact copy of the mathematical formula. Its oneness is only a component for the intrinsic moralism of science: “every single one conforms to the laws of nature”. The oneness is no longer a “what it is” but a neutral “something” in the phrase “how something comes to be”; as long as this something is differentiated, by magnitude or as a mere instance, it can be used. Yet what this something really is is somehow latent in the questioning.
克莱恩论数学数与理念数的分野
毕达哥拉斯学派的立场,简而言之,可以化为“一切事物都是可数的”。但这会遭遇两个问题。如果说一切事物都是可数的,那是什么在数呢?被数的存在者不能成为计数者本身。由此,数与被数的存在者有所间隔。第二个问题随之而来,假如被数的存在者差异于数本身,而数本身是「一个」数,属于数的这样的「一性」又从何而来呢?毕达哥拉斯学派的解决方式是提出数的形式的说法,通过奇偶性、平方立方等对数进行归类。数的归类学就是狭义上的数学(arithmetics),而对数与数之间的关系的探究则属于算学(logistics)的范畴。在这个意义上,对于毕达哥拉斯学派来说,算学或数学上理论和实践的分离还没有成为问题。
但这无法真正解决问题。因为三与五可以同享奇数的形式,但三与五之间的差别却无法体现。柏拉图因此提出了数之理念;并非是数的形式起了作用,而是数本身的理念。同时,由于数不能脱离于内容,作为内容物的纯粹单位同样需要得到保留。由此,柏拉图建立了这样的一个渐进式秩序:可感数→数学数→理念数。毕达哥拉斯学派将数学数的形式视作可感数的统一性根源,而柏拉图则更近一步,将理念数视作数学数的根源。而理念数不仅需要作为填充物的数学内容,还需要彰显其特异性的理念内容:比如一的理念内容就是善好(因为一意味着非偶然),而二的理念内容则是存在(因为只有当静止与运动的对子被聚集时,存在才在;存在必然要求不定之二)。理念数的提出同样解决了 斐多 中的问题;二永远是作为一个集合而在,而本身并非是一个独立于集合内容的理念。因此数之理念严格来说并非是单独的理念,而是理念之数。
这个意义上,亚里士多德对柏拉图的批评就显得尤其激进了。首先,亚里士多德不认为有纯粹的数学单位,或者他所认为的纯粹数学单位并不是“独立”的数学单位,而是“中立”的单位。亚里士多德处的数学单位是对可感数的抽象,也就意味着可感数单位与单位间的可转化性。在这个意义上,我之前的想法只对了一半;错的地方在于,亚里士多德会认为只有自己才完成了理论算学(因为数学单位意味着实体单位可以随时转换成另一种实体单位的部分,因此分数是有意义的),而不是相反;但柏拉图并不会认为亚里士多德所建立的能够超出实践算学的范畴。
另一个批评在于,亚里士多德否认数的内容来自于理念数,甚至否认理念数的存在。对于他来说,数的内容似乎仅仅局限于数学数与可感数。而他对理念数的拒绝则来自于更加根本的一种保留:他似乎很怀疑数具有任何的统一性,也不认为数必须要有「一个」根源。
所以柏拉图那里有的是不同层次的数的对应,而亚里士多德那里则是数与存在者的直接等同
“对存在的探问等价于对一的探问”,但一并不是数。柏拉图那里可以通过数学数的中介而达到理念数,但亚里士多德那里数学数到理念数这层阶梯被抽掉了,而只剩下形式与质料的混合关系。也就是反而是柏拉图那里可以在理念层面有复杂的复合关系,但亚里士多德却只是形式是一。
克莱恩论数学物理学
现代数学物理学以这项原则行动:将一切方法论上的普遍性都转变为对象的普遍性。在古希腊时期的数学科学隐含的普遍性很少会被开掘出来专题化,因此仅仅是作为隐藏的概念上的预设而具有方法论上的意义。
第一个明显的例子是理论算学。古希腊的理论算学是以纯粹单位为对象的计算,因此与纯粹单位具有观念上的直接性。尽管,理论算学的可能性在柏拉图和亚里士多德那里仍然有争议,因为亚里士多德并不认为存在对纯粹单位的“计数”;亚里士多德认为纯粹单位仅仅是对可被计数的现实单位的抽象,因此只有“理论数学”(theoretical arithmetics)而没有“理论算学”。对他来说,纯粹单位不具有确定性,因此不能被计算。但不论柏拉图和亚里士多德两者的分歧,他们都共同有这样的前设,即,只有确定的对象能被计算。但这被现代代数改变了;代数将算学方法论上的普遍性转化成了对象上的普遍性,因此,不具有确定性的数得以被字符表示,而原本具有确定性的数(比如“2”,意味着2个纯粹单位)则被视作普遍对象的一个实例。
第二个例子来自于分析几何。在笛卡尔和费马之前,几何虽然也设立了普遍关系,但几何形状之间的普遍关系却是以命题的形式被书写的。欧几里得或者是阿波罗尼乌斯发现的所有的普遍关系都是在确定的被给出的形状间建立的,特殊的形状作为特殊的图像对应着特殊的理念;阿波罗尼乌斯所表述的椭圆无法独立于特定的圆锥体,尽管是“每一个”圆锥体。而分析几何则将这图示化的“每一个”直接地转化成了一个包含了“所有”的图像(公式),而特定图像也相应地成为了“所有”的“每一个”实例。
而与这基本原则相违背的一个反例则来自于比例这个概念。在几何原本的第五卷,被讨论的并不是具有比例的特定物,而是具有比例的所有量。第五卷中的线段并非是一个对象,而是一个象征符号,作为一切量的承载者。需要注意的是,这里的线段不是图像,正因为其与线段本身没有严格对应关系;在关于三角形的命题中,“这个”被画出的三角形通过想象力显示了纯粹的数学三角形,但在第五卷中,被画出的“这条”线段并不显示纯粹的线段,而是显示纯粹的量。在这里,纯粹却具体的数学对象被抽离,图像被转化为符号,仅仅与最普遍之物吻合;而相应的,确定对象的消失使得原本具体的量也获得了解放。而第七卷关于数的内容则是在此基础上对第五卷内容的补充。
从这里出发,现代科学将自己视为古典时代科学的继承人;但其继承的概念的意涵发生了巨大的转变。首先,古典时代的宇宙秩序(比例)被转化为了宇宙的法则,原本的确定的数字比例被转化为符号公式;而符号就表示了一种随机性,哪怕是常数,这种常数也是无理由的。这样的可替代性,通过伽利略对柏拉图 蒂迈欧 的利用,最终以运动中的恒定的形式得到了表示。
意向性历史与事实性历史
在讨论这两种意义上的历史之前,先要澄清历史在此处意味着对起源的追问(ἱστορία);而对历史性起源的追问必然超出通常理解的历史本身,而达到了先验历史的维度。
先验历史这样的词是如何可能的?这不是一个矛盾吗?首先得从数学经验讨论。欧几里得将无法证明几何原本的任何命题,如果他不首先建立起相应的定义等;在这个意义上,几何命题有其起源。但这个起源并不是随机的,而是必然的。由此,我们得以讨论先验必然的历史。
先验意味着理念或知识。我们需要理解的因此就是知识的起源。知识的起源无法到朴素经验主义所定义的“经验数据库”中寻找,因此需要往先验主体性中挖掘。这个起源有两个规定:首先,起源必须与“过去”“现在”“未来”各有联系;另外,起源必须在根本上规定理念对象的统一性。
在这两个基本规定下,起源具有了两个“形式”:意向性历史与事实性历史。让我们接下来来考察两者的关系。
意向性历史在根本上被内在时间性(internal temporality,而不是内时间意识inner time consciousness)限定。这一种理念对象统一性的起源规定了在时间内对象的持存。因此,意向性历史的达成来自于对过去、现在和未来的前摄意识更改;在某种意义上,可以认为这是对理念的联结。
但是仅仅凭意向性历史无法解释这样一个疑问:如果仅仅只关注前摄意识对对象的无限修改,理念对象将永远显示为“未完成”的状态。也就是说,在意向性历史中,任何对象都无法成为一个业已完成的思维产物,因而无法成为真正的确切起源。因此,想要确定起源,仅仅凭意向性历史的追溯是不够的。
事实性历史则关涉于“原初给予”。在这个意义上,它并不是前摄意识,而恰恰是后摄意识。它关乎的是“第一个几何学家”的期望;在三角形第一次出现时,它的诸种性质就已经蕴含其中,等待未来的展开了。只有达到这样的原初给予的明见性,起源才真正获得了规定。在某种意义上,事实性历史是无时的,因为时间被凝缩在了一个确定产物中等待展开,而意向性历史则刻画了这样的展开过程。因此,可以说意向性历史是对时间内的各个相关联部分的刻画,而事实性历史则是对无时性与时间性的相关结构刻画。
从激活的方面说,意向性历史的激活是对过去到现在的重演,而事实性历史的激活则是将过去彻底重构为现时当下。
和福柯处在完全不同的维度中,福柯那里的知识型其实根本不是知识,只是条件化的认识模型……
起源重要是因为起源在根本意义上是先验主体的活动,而活动规定了产物,产物与活动的关系不是对称的。
但这并不意味着1.只有一个起源;2.起源与产物间不能相互作用。
用莱布尼茨那里的范畴理解的话,意向性历史是将过去和现在和未来把握在一个在时间当中持存的有机的整体当中,但事实性历史则是达成对理念单子(本体substance)的直观,并将其折叠起来的所有可能性都把握在这一个瞬间当中。一个是时间中的永恒,一个是时间之外的永恒。
AI translated the non-English passages; original English passages are preserved. The mixed-language original is available under Original.
Stevin algebraized number, Descartes likewise algebraized geometry, and Wallis reformulated algebraic forms as universal proportions.
Algebraization in Klein roughly means treating the total concept, or Idea, of number itself as an operable number.
An algebraic solution differs from a geometrical method because it is determined from the outset as ‘a determinate number of monads.’ When unknowns are introduced into geometry, the magnitudes finally obtained through analysis must undergo a further synthesis: can the resulting side, for example, form a triangle with the other sides? Algebra has no such concern, because a solution’s ‘impossibility’ comes not from its inability to combine with other components into a whole, but from its own nature. A side unable to form a triangle is thus still nongeometrical, whereas a negative solution obtained is impossible because of its own nature—there cannot be a number less than zero. That is, only under particular conditions and after a particular synthesis with other equally particular units can the resulting side be judged incapable of forming a ‘particular’ triangle. Neither the operation nor its objects are nearly as universal as in algebra.
On this basis, Viète’s and Descartes’ mathesis universalis goes further. Yet their development of algebra cannot simply be understood as an abstraction of number, since Aristotle also used abstraction as a method. The distinctiveness of abstraction in mathesis universalis must be distinguished from abstraction as a philosophical method. Simply put, mathesis universalis individuates and objectifies methodological abstraction. For Aristotle, there is no total mathematical object, and a total understanding of mathematical objects means understanding Being itself. Even though Plato, like Viète, raises number to the level of the Idea, their fundamental difference must be noted. For Plato, number is only the form of Being. Fundamentally, he uses number’s distinctive properties to discuss Being and nonbeing. Being for Plato remains an aggregate of particular Ideas. The tension between one and many appears as the tension between Being and number; above number there remains the one of Being to be questioned. In Viète’s mathesis universalis, however, the level above number is canceled and absorbed into number. The universality of Being itself becomes a ‘number’ awaiting treatment, while the forgetting and withdrawal of the question of Being are condensed into the unknown x. Symbolizing the unknown x already contains this turn: Being is unquestioningly regarded as a particular being, while the relation between Being/Idea and being/object, still a problem in ancient Greece, is concealed in mathesis universalis. The strongly tensioned Greek relation between λόγος and διάνοια, between saying and thinking, is thereby concealed as well. Through symbolization, the Idea is and can only be ‘said.’ Yet in Archimedes this symbolization remains a problem awaiting thought, because saying and thinking are distinguished only in practice. Archimedes recognizes merely that, after thinking reaches a certain degree, it needs saying as a supplementary ground. Saying has not wholly effaced thinking’s sovereignty or the requirement to fulfill ideal objects. Archimedes therefore even needs to defend saying. In the age of mathesis universalis, however, saying completely replaces thinking, and the question of Being turns into an art of operating on numbers. Connecting this change in the relation of saying and thinking with the changed understanding of number, we can say: thinking is the process of concretely presenting ideal objects, while saying substitutes words and signs for them. The turn we witness in mathesis universalis is that the relation between Idea and ideal object within thinking is flattened, while the act of thinking itself becomes an object of saying. This happens precisely because the universality of thinking as an act requires symbolic representation. Within mathesis universalis, thinking becomes a unit of saying, completely losing its independent standing and its essence. Mathesis universalis is therefore fundamentally a science of saying, a science of λόγος. Only in this sense can we understand why ‘thinking Being’ strikes against λόγος: λόγος is precisely unthinking.
More briefly, in ancient Greece the tension between one and all had to be characterized; in the age of mathesis universalis, this tension is sedimented in the expression of the unknown x, awaiting unfolding.
Viète’s symbolization of mathematics directly leads to the withdrawal of Being. Number is no longer an externally given, determinate being, but an operable conceptual sign with an intrinsic indeterminacy and arbitrariness. Its operability comes from its variability.
One and many are sedimented in signs because signs are fundamentally without objects. Once even the object itself is concealed, its mode of being naturally cannot be questioned.
Klein’s critique of Husserl
Klein’s critique of Husserl concerns neither technical details nor content, as I originally thought, but their most essential positions. Briefly, Klein thinks Husserl is too concerned with continuity and overlooks shifts of meaning, even ruptures of meaning, in transcendental history.
We must begin with Husserl’s ‘historical turn.’ This seems to arise from a rupture he experiences: he recognizes that science is entering a crisis unintelligible to his own experience, and attributes the unintelligibility of modern science to a historicity exceeding time-consciousness. Within time-consciousness, everything given undergoes continual modification while retaining identity, making a purely epistemological grasp of things possible. Yet anything that ‘enters’ time-consciousness must be given outside this time-consciousness. Mathematical and scientific objects are especially distinctive in this respect because they have the highest objectivity of all objects. Their constitution therefore cannot be wholly immanent to this consciousness. They must exceed it and enter the domain of intersubjectivity, handed down to this consciousness as tradition and inheritance. In the interaction of subjects, meaning sediments and is concealed, ultimately producing modern science’s unintelligibility. Questioning the interaction of meaning among subjects is therefore questioning its history. Note, however, that Husserl does not posit here the ‘rupture’ psychoanalysis would emphasize, because for him the historical problem remains epistemological. Intersubjectivity is not the subject and its other, but the continuous whole of the subject and its modifications, whose network of meaning is unified by the identity of ideal objects. For Husserl, emphasizing the arbitrariness and contingency of historical facts therefore has no significance: history is merely eternity’s shadow. He enters history only because traditional epistemology is confined within a single consciousness, whose temporal modifications cannot independently become the origin.
We can already see Husserl’s extraordinary commitment to continuity. Crudely put, the entire will of Husserlian phenomenology can fundamentally be grasped as an impulse to connect modern science to an originary lifeworld—while recognizing, of course, that this lifeworld is not pregiven but continually constituted. This also relates to Husserl’s view of his philosophical work as the teleological heir of Greek philosophy.
For Klein, however, precisely this commitment leads Husserl to a fundamental error in his concrete analysis of modern science. Modern science cannot be directly connected to the lifeworld, because the two are fundamentally different creations. Husserl’s attempt to reduce modern science to an ascending constitution from basic experience and intuitive evidence stops at the analysis of Galilean geometrized nature because he neglects another crucial component: mathesis universalis. His neglect of its importance comes precisely from the commitment to continuity just analyzed. This commitment prevents him from leaving the methodology of historical reflection or backward inquiry, permitting only a ‘zigzag’ approach to origins. Such inquiry still takes the present fact of modern science as its basic condition, imposing hindsight on the historical situation and failing to notice that modern science’s sedimentation of meaning already contains a fundamental shift: the birth of mathesis universalis. Klein believes his mathematical research radicalizes the reactivation of this origin because his method is proactive rather than retrospective. It thereby avoids Husserl’s neglect of a presupposition of modern science: the direct substitution of mathesis universalis for the natural world. Precisely because Husserl still regards mathesis universalis as directly substituting for the natural world, his turn achieves only their connection, supposing a continuous relation of direct reference between them. Through concrete historical analysis, Klein recognizes instead that what mathesis universalis replaces is not the natural world but the Greeks’ ideal world, discerning the crucial rupture within science’s historical development. Unlike Husserl, therefore, although Klein in practice also used only the ‘zigzag’ method of indirect inquiry—and did so before Husserl—he uncovered another possible approach: a general investigation of historical cross-sections as periods, also a common method of conventional historical research. In this sense, Klein seems to surpass Husserl.
The foregoing is Hopkins’ analysis of the two, not my own view. According to Majolino’s criticism, Hopkins fundamentally misunderstands Husserl’s concept of individuum, narrowly treating it as an ‘individual empirical object’ and thus concluding that Husserlian philosophy attempts to connect modern science to the natural world. On Majolino’s reading, however, individuum can perfectly well refer to a unique ideal object. This means that Klein’s attempt to ground mathesis universalis in a transformation of the Greeks’ ideal world is not wholly independent of Husserl’s influence.
Why, at the birth of mathesis universalis, founders such as Descartes, Viète, and Stevin regard themselves both as constructing the only science and as heirs to the ancient Greek tradition, while their successors forget the latter and regard their science only as the sole truth of the natural world or reality—why that latter aspect sediments away—is not something Klein needs to worry about.
Precision and exactitude: with Plato, the precise is itself the One or the Good. But with modern mathematical physics, the precise is good only as long as it produces an exact copy of the mathematical formula. Its oneness is only a component for the intrinsic moralism of science: “every single one conforms to the laws of nature”. The oneness is no longer a “what it is” but a neutral “something” in the phrase “how something comes to be”; as long as this something is differentiated, by magnitude or as a mere instance, it can be used. Yet what this something really is is somehow latent in the questioning.
Klein on the distinction between mathematical and eidetic numbers
The Pythagorean position can be reduced to ‘everything is countable.’ Yet this encounters two problems. If everything is countable, what does the counting? The beings counted cannot themselves be what counts. Number is therefore separated from the beings counted. A second problem follows: if the beings counted differ from number itself, while number itself is ‘one’ number, where does number’s oneness come from? The Pythagoreans propose forms of number, classifying numbers by oddness and evenness, squares and cubes, and so on. The classification of numbers is mathematics in the narrower sense, arithmetics, while investigating relations among numbers belongs to calculation, logistics. In this sense, the separation of theory and practice in calculation or mathematics has not yet become a problem for the Pythagoreans.
Yet this does not truly solve the problem. Three and five may share the form of odd number, but their difference cannot be expressed. Plato therefore proposes the Idea of number: what operates is not the form of number but number’s own Idea. At the same time, since number cannot be detached from content, pure units as its content must also be preserved. Plato thus establishes a progressive order: sensible number → mathematical number → eidetic number. The Pythagoreans regard the forms of mathematical numbers as the source unifying sensible numbers; Plato goes further, regarding eidetic numbers as the origin of mathematical numbers. Eidetic numbers require not only mathematical content as filling, but also ideal content manifesting their distinctiveness. The ideal content of one, for example, is goodness, because one means noncontingency; the ideal content of two is Being, because Being is only when the pair rest and movement is gathered, and necessarily requires the indefinite dyad. The proposal of eidetic numbers also solves the problem in the Phaedo: two always exists as a collection, not as an Idea independent of its contents. Strictly speaking, therefore, the Idea of number is not an isolated Idea but a number of Ideas.
Aristotle’s critique of Plato seems particularly radical in this sense. First, Aristotle does not accept pure mathematical units—or the pure mathematical units he accepts are not ‘independent’ but ‘neutral.’ His mathematical units are abstractions from sensible numbers, implying the convertibility of units of sensible number into one another. My earlier thought was therefore only half right. The mistaken half is that Aristotle would regard himself alone as completing theoretical calculation, not the reverse: mathematical units mean that substantial units can at any moment be converted into parts of another kind of substantial unit, so fractions are meaningful. Plato, however, would not think Aristotle’s construction exceeds the domain of practical calculation.
Another criticism is that Aristotle denies that number’s content comes from eidetic numbers, even denying their existence. For him, numerical content seems confined to mathematical and sensible numbers. His rejection of eidetic numbers comes from a more fundamental reservation: he seems deeply doubtful that number possesses any unity and does not think number must have ‘one’ origin.
So Plato has correspondences among numbers on different levels, whereas Aristotle directly identifies numbers with beings.
‘Questioning Being is equivalent to questioning the one,’ but the one is not number. In Plato, the mediation of mathematical numbers can lead to eidetic numbers. In Aristotle, the step from mathematical to eidetic numbers is removed, leaving only the mixed relation of form and matter. Thus, conversely, Plato can have complex composite relations on the ideal level, whereas in Aristotle only form is one.
Klein on mathematical physics
Modern mathematical physics acts on this principle: transform every methodological universality into the universality of an object. The universality implicit in ancient Greek mathematical science is rarely excavated and thematized, remaining a hidden conceptual presupposition with methodological significance.
The first clear example is theoretical calculation. Ancient Greek theoretical calculation takes pure units as its objects and thus has an ideal immediacy with pure units. Yet its possibility remains disputed between Plato and Aristotle, because Aristotle does not accept the existence of ‘counting’ pure units. For him, pure units are merely abstractions from actual countable units, so there is ‘theoretical arithmetics’ but no ‘theoretical logistics.’ Pure units lack determinacy and therefore cannot be calculated. Whatever their disagreement, however, Plato and Aristotle share the presupposition that only determinate objects can be calculated. Modern algebra changes this. It transforms calculation’s methodological universality into the universality of objects, allowing indeterminate numbers to be represented by characters, while formerly determinate numbers—for example, ‘2,’ meaning two pure units—become instances of universal objects.
The second example comes from analytic geometry. Before Descartes and Fermat, geometry establishes universal relations, but universal relations among geometrical figures are written as propositions. All the universal relations discovered by Euclid or Apollonius are established among determinate given figures, particular figures as particular images corresponding to particular Ideas. The ellipse described by Apollonius cannot be independent of a specific cone, even if it is ‘every’ cone. Analytic geometry directly transforms this diagrammatic ‘every’ into an image or formula encompassing ‘all,’ while the particular image correspondingly becomes ‘each’ instance of ‘all.’
A counterexample to this basic principle comes from proportion. Book V of the Elements discusses not particular things having proportions but all magnitudes having proportions. Its line segments are not objects but symbolic signs bearing all magnitudes. Note that these segments are not images, precisely because they do not strictly correspond to line segments themselves. In a proposition about triangles, ‘this’ drawn triangle exhibits a pure mathematical triangle through imagination. In Book V, however, ‘this’ drawn segment exhibits not a pure segment but pure magnitude. Here the pure yet concrete mathematical object is abstracted away, and the image becomes a sign corresponding only to what is most universal. Correspondingly, the disappearance of determinate objects liberates previously concrete magnitude. Book VII’s treatment of number supplements Book V on this basis.
From here modern science regards itself as the heir of classical science, yet the meaning of its inherited concepts changes enormously. First, the cosmic order of antiquity, proportion, is transformed into cosmic laws; determinate numerical proportions become symbolic formulas. Signs express a kind of randomness. Even constants are without reason. Through Galileo’s use of Plato’s Timaeus, this substitutability finally finds expression as constancy within movement.
Intentional history and factual history
Before discussing these two senses of history, we must clarify that history here means inquiry into origins, ἱστορία. Questioning historical origins necessarily exceeds history as ordinarily understood, reaching the dimension of transcendental history.
How is a term such as transcendental history possible? Is it not a contradiction? We must begin with mathematical experience. Euclid could not prove any proposition of the Elements without first establishing the corresponding definitions and so on. In this sense, geometrical propositions have an origin. Yet this origin is not contingent but necessary. We can therefore discuss a transcendentally necessary history.
Transcendental means Idea or knowledge. What we need to understand is therefore the origin of knowledge. This origin cannot be sought in the ‘database of experience’ defined by naïve empiricism; it must be excavated within transcendental subjectivity. The origin has two determinations: it must relate to past, present, and future, and it must fundamentally determine the unity of the ideal object.
Under these two basic determinations, origin has two ‘forms’: intentional history and factual history. Let us examine their relation.
Intentional history is fundamentally delimited by internal temporality, not inner time-consciousness. This origin of the ideal object’s unity determines the object’s persistence in time. Intentional history is thus accomplished through proactive modifications of consciousness concerning past, present, and future. In a sense, it can be regarded as a connection of Ideas.
Intentional history alone, however, cannot explain this difficulty: if we attend only to proactive consciousness’s infinite modifications of an object, the ideal object will always appear unfinished. In intentional history, no object can become an already completed product of thought and thus a genuinely definite origin. Tracing intentional history alone is therefore insufficient to establish an origin.
Factual history concerns ‘originary givenness.’ In this sense, it is not proactive but retroactive consciousness. It concerns the expectations of the ‘first geometer.’ When the triangle first appears, its various properties are already implied within it, awaiting future unfolding. Only by reaching the evidence of such originary givenness is the origin truly determined. Factual history is in a sense timeless, because time is condensed into a determinate product awaiting unfolding, whereas intentional history characterizes this unfolding process. We can thus say that intentional history characterizes the related parts within time, while factual history characterizes the correlational structure of timelessness and temporality.
In terms of activation, intentional history is activated by reenacting the passage from past to present, while factual history is activated by completely reconstructing the past as the current present.
This is on an entirely different dimension from Foucault. Foucault’s episteme is fundamentally not knowledge, but only a conditioned model of cognition…
Origin matters because, fundamentally, it is an activity of the transcendental subject. Activity determines its product; the relation between product and activity is not symmetrical.
Yet this does not mean 1. there is only one origin; or 2. origin and product cannot interact.
Understood through Leibnizian categories, intentional history grasps past, present, and future within an organic whole persisting in time. Factual history, however, attains an intuition of the ideal monad, or substance, grasping all its enfolded possibilities within a single instant. One is eternity within time; the other, eternity outside time.
非中文段落由 AI 翻译,中文原文保持不变。中英混合原稿可通过“原文”按钮查看。
Stevin做到了将数代数化,Descartes做到了将几何同样地代数化,而Wallis则做到了将代数形式重新表述为普遍比例。
Klein那边代数化大概的意思是,将数的总体概念(理念)本身化作一个可操作的数来看待
代数式的解因其在最开始就已经被决定为“确定数目的单子”,其与几何方法就有了区分。将未知项引入几何中,最后通过分析方法所能得到的量还需要再次经历一个综合的过程:比如求得的边能否与其它边搭成一个三角形。而代数则没有这个顾虑,因其解的“不可能性”并不来源于它无法与其它组成部分结合成一个整体,而来源于其性质本身。因此,无法形成三角形的边依旧是非几何的,但求得的负数解却是因其自身性质而不可能的(不可能有小于零的数) —— 也就是说,求得的边是在特殊的情况下并与其它同样特殊的单位经历了特殊的综合过程后才可以被认为无法组成一个“特殊的”三角形,其操作与操作对象远远没有代数具有普遍性。
在这个基础上,Vieta与Descartes的普全数学更进一步。但他们对代数的发展不能被简单理解为对数的抽象,因为“抽象”作为一种方法Aristotle同样使用过,而普全数学的“抽象”的特殊性需要与作为哲学方法的“抽象”区分开来。简单来说,普全数学的抽象所做的是将方法上的抽象个体化、对象化。换句话说,对于Aristotle,并不存在一个总体的数学对象,而数学对象的总体理解就意味着对存在本身的理解。而哪怕Plato与Vieta一样将数抬高到了理念的位置,这里依然需要注意两者根本上的差异。对于Plato来说,数只是存在的形式。换句话说,根本上,Plato是在借助数的特殊性质来讨论存在与非存在的问题。而存在对于Plato来说依然是特殊的诸理念的聚合物;对于Plato来说,一与多之间的张力表现为存在与数之间的张力;在数之上还有等待追问的存在之一。然而在Vieta的普全数学那里,数之上的层面被取消了,并被吸收进了数的内部,即,存在的普遍性本身成为了一个等待处理的“数”,而存在之问的遗忘与退隐就被凝聚在未知数“x”上。对于未知数“x”的象征就已经蕴含了这一层转折,也就是存在被无疑问地视作了一个特殊的存在者,而在古希腊时依然被视为问题的“存在/理念”与“存在者/对象”的关系在普全数学中就被掩盖了。因此,古希腊时具有相当张力的λόγος与διάνοια的关系,也就是“说”与“思”的关系,也被隐藏了。因为通过象征,“理念”被且只能被“说”出,但这个象征在Archimedes处还是一个等待思索的问题,因为“说”与“思”只在实践中被区分,Archimedes仅仅只是意识到在“思”达到一定程度后,“思”需要“说”作为自身的补充根据,但“说”尚未完全地抹去“思”的主权,也没有抹去对理念对象进行充实的要求;因此Archimedes甚至需要为“说”进行辩护。但是在普全数学的时代,“说”完全地替代了“思”,存在之问变动为对数进行操作的技艺。将“说”与“思”的关系的变化与对“数”的理解的变化联系起来,我们可以说,“思”是对理念对象进行具体呈现的过程,“说”则是用语词和符号对理念对象进行替代的过程。而我们在普全数学中目睹的转折是,“思”本身内部所蕴含的理念与理念对象的关系被扁平化,而“思”这个行动本身则成为了被“说”的对象:这恰恰是因为“思”这个行动本身的普遍性要求了符号进行表征。因此,在普全数学中,“思”成为了“说”的一个单元,并因此完全地失去了“思”的独立地位,也丧失了自己的本质。因此,普全数学根本上是“说”的科学,也就是λόγος的科学。只有在这个意义上,我们才能够理解为什么“对存在之思”是对λόγος的冲击,因为λόγος恰恰是不思的。
简洁一点说的话,可以说,在古希腊,“一”与“全”的张力是需要刻画的,但是在普全数学的时代,这个张力被沉淀在未知数“x”的表述中等待展开
Vieta将数学符号化,直接导致了存在的退隐,而数从此不再是外在被给予的确定的存在,而是可以被操作的、具有内在不确定性与随意性的概念符号,其可操作性源于其可变性
“一”与“多”在符号中被沉淀,是因为符号根本意义上是无对象的,而连对象本身都被隐匿了,自然也就无法去探问对象的存在模式了
克莱茵对胡赛尔的批评
克莱茵对胡赛尔的批评并非是我原本以为的技术细节或者内容方面的批评,而反而是在最本质立场上的批评。简单总结的话可以说,在克莱茵看来,胡赛尔太注重连续性了,而恰恰忽视了先验历史中的意义迁移,甚至是意义断裂。
这首先要从胡赛尔的“历史转向”开始谈起。胡赛尔的历史转向似乎是由于一种他所感受到的断裂:他意识到,科学正在陷入一种无法被(他个人)经验理解的危机,而他将这种现代科学的不可理解性归结于超出时间意识的历史性。在时间意识中,一切被给予者都经受着持续不断的变更却保持同一,因此在时间意识中,对事物的纯粹知识论把握是可能的。然而,任何“进入”了时间意识的被给予者必然是在此时间意识之外被给予,而数学与科学对象在这个意义上尤其特殊,因为它们具有一切对象中最高的客观性;这意味着它们的构成是不可能完全内在于(此)意识的。在这个意义上,数学与科学对象必须超出(此)意识而进入交互主体性的领域,作为传统和遗产被交付于(此)意识。而在主体的交互中,意义沉淀而被隐藏,最终导致了现代科学的不可理解;而对主体的意义交互的探问就是对其历史的探问。但是这里要注意的是,对于胡赛尔来说,此处并没有精神分析会强调的“断裂”,因为对于胡赛尔来说,历史问题其实依然是一个知识论问题。交互主体并非是主体与其他者,而是主体与其变更的连续整体;这个整体的意义网络由理念对象的同一性统合。所以,对胡赛尔来说,对历史事实的任意性和偶然性的强调没有意义:历史只是永恒的影子;他进入历史也仅仅是因为传统意义上的知识论被局限在单个意识的内部,而单个意识的时间变更无法独立成为根源罢了。
从这里已经可以看出,胡赛尔对连续性有着非同一般的执着。粗暴点说,我们在根本上可以将胡赛尔现象学的意志整个把握为一种将现代科学连接到根源的生活世界的冲动:当然要意识到生活世界并不是被预先给予的,而是被持续构建起来的。这也与胡赛尔将他从事的哲学工作视作为希腊哲学的目的论继承者有关。
然而,在克莱茵看来,正是这样一种执着使胡赛尔在对现代科学进行具体分析时犯下了根本的错误。对于克莱茵,现代科学是无法与生活世界直接连接的:因为现代科学在根本上与生活世界已经是两种造物。胡赛尔试图将现代科学还原为对基本经验与直观明见性的上升构造的努力之所以止步于对伽利略式的几何化自然的分析,是因为他忽视了现代科学另一个重要的组成部分:普全数学。而胡赛尔对普全数学重要性的忽视则恰恰来自于刚刚分析过的他对连续性的执着。这种执着使他无法脱离历史反思/回溯探问的方法论,而对于根源只能采用那种“之字形”的迂回探询;而“之字形”的探询依然以当下现代科学的事实为基本条件,将后见之明强加到了当时的历史境遇中,而无法注意到在现代科学的意义沉淀中已经包含了一个根本的意义迁移,即普全数学的诞生。克莱茵认为自己的数学研究彻底化了对普全数学起源的再激活,因其方法并非是回溯性的,而是前抛性(proactive)的,因此避免了像胡赛尔一样将现代科学的前设(即普全数学对自然世界的直接替代)忽略。正因为胡赛尔依然认为普全数学是对自然世界的直接替代,他的转向仅仅达成了对普全数学和自然世界的连接,并认为两者之间有直接指涉的连续关系;而克莱茵(通过具体的历史分析)则意识到,普全数学所替代的恰恰不是自然世界,而是希腊人的理念世界,而注意到了科学之历史演进中隐含的关键断裂。因此,与胡赛尔不同,尽管克莱茵(先于胡赛尔)实际上也只采用了“之字形”的迂回探问的方法,他却发掘出了另一种可能的探询方式,即对历史切面的普遍时段性探究,而这也是标准的历史学研究常用的方式。在这个意义上,克莱茵对胡赛尔似有超出。
以上都是Hopkins对两者的分析,不代表我的个人看法。根据Majolino的批评,Hopkins在根本上误解了胡赛尔的individuum概念,并将其偏狭地视作“个体经验对象”,而得出了“胡赛尔哲学是将现代科学连接到自然世界上的尝试”的结论。然而,根据Majolino的解读,individuum完全可以指涉具有独一性的理念对象,而这就意味着克莱茵将普全数学奠基于对希腊人理念世界的改造的努力并非是完全独立于胡赛尔影响的。
至于说为什么普全数学诞生时,奠基者如笛卡尔、Vieta和Stevin在认为自己构造的是唯一的科学的同时还将自己理解为古希腊传统的继承者,而后继者们则遗忘了后面这一点,转而只把自己的科学当作自然世界/实在的唯一真理,为什么后者沉淀下去了,那就不是克莱茵要操心的事情了。
精密(precision)与精确(exactitude):在柏拉图那里,精密者本身就是一或善。但在现代数学物理学那里,精密者只有在它生产出数学公式的精确副本时才是好的。它的一性只是科学内在道德主义的一个组成部分:“每一个个体都符合自然法则。”一性不再是“它是什么”,而是“某物如何生成”这一表述中中性的“某物”;只要这某物能够被区分,无论通过量值,还是仅仅作为一个实例,它就能被使用。然而,这某物究竟是什么,却以某种方式潜藏在发问之中。
克莱恩论数学数与理念数的分野
毕达哥拉斯学派的立场,简而言之,可以化为“一切事物都是可数的”。但这会遭遇两个问题。如果说一切事物都是可数的,那是什么在数呢?被数的存在者不能成为计数者本身。由此,数与被数的存在者有所间隔。第二个问题随之而来,假如被数的存在者差异于数本身,而数本身是「一个」数,属于数的这样的「一性」又从何而来呢?毕达哥拉斯学派的解决方式是提出数的形式的说法,通过奇偶性、平方立方等对数进行归类。数的归类学就是狭义上的数学(arithmetics),而对数与数之间的关系的探究则属于算学(logistics)的范畴。在这个意义上,对于毕达哥拉斯学派来说,算学或数学上理论和实践的分离还没有成为问题。
但这无法真正解决问题。因为三与五可以同享奇数的形式,但三与五之间的差别却无法体现。柏拉图因此提出了数之理念;并非是数的形式起了作用,而是数本身的理念。同时,由于数不能脱离于内容,作为内容物的纯粹单位同样需要得到保留。由此,柏拉图建立了这样的一个渐进式秩序:可感数→数学数→理念数。毕达哥拉斯学派将数学数的形式视作可感数的统一性根源,而柏拉图则更近一步,将理念数视作数学数的根源。而理念数不仅需要作为填充物的数学内容,还需要彰显其特异性的理念内容:比如一的理念内容就是善好(因为一意味着非偶然),而二的理念内容则是存在(因为只有当静止与运动的对子被聚集时,存在才在;存在必然要求不定之二)。理念数的提出同样解决了 斐多 中的问题;二永远是作为一个集合而在,而本身并非是一个独立于集合内容的理念。因此数之理念严格来说并非是单独的理念,而是理念之数。
这个意义上,亚里士多德对柏拉图的批评就显得尤其激进了。首先,亚里士多德不认为有纯粹的数学单位,或者他所认为的纯粹数学单位并不是“独立”的数学单位,而是“中立”的单位。亚里士多德处的数学单位是对可感数的抽象,也就意味着可感数单位与单位间的可转化性。在这个意义上,我之前的想法只对了一半;错的地方在于,亚里士多德会认为只有自己才完成了理论算学(因为数学单位意味着实体单位可以随时转换成另一种实体单位的部分,因此分数是有意义的),而不是相反;但柏拉图并不会认为亚里士多德所建立的能够超出实践算学的范畴。
另一个批评在于,亚里士多德否认数的内容来自于理念数,甚至否认理念数的存在。对于他来说,数的内容似乎仅仅局限于数学数与可感数。而他对理念数的拒绝则来自于更加根本的一种保留:他似乎很怀疑数具有任何的统一性,也不认为数必须要有「一个」根源。
所以柏拉图那里有的是不同层次的数的对应,而亚里士多德那里则是数与存在者的直接等同
“对存在的探问等价于对一的探问”,但一并不是数。柏拉图那里可以通过数学数的中介而达到理念数,但亚里士多德那里数学数到理念数这层阶梯被抽掉了,而只剩下形式与质料的混合关系。也就是反而是柏拉图那里可以在理念层面有复杂的复合关系,但亚里士多德却只是形式是一。
克莱恩论数学物理学
现代数学物理学以这项原则行动:将一切方法论上的普遍性都转变为对象的普遍性。在古希腊时期的数学科学隐含的普遍性很少会被开掘出来专题化,因此仅仅是作为隐藏的概念上的预设而具有方法论上的意义。
第一个明显的例子是理论算学。古希腊的理论算学是以纯粹单位为对象的计算,因此与纯粹单位具有观念上的直接性。尽管,理论算学的可能性在柏拉图和亚里士多德那里仍然有争议,因为亚里士多德并不认为存在对纯粹单位的“计数”;亚里士多德认为纯粹单位仅仅是对可被计数的现实单位的抽象,因此只有“理论数学”(theoretical arithmetics)而没有“理论算学”。对他来说,纯粹单位不具有确定性,因此不能被计算。但不论柏拉图和亚里士多德两者的分歧,他们都共同有这样的前设,即,只有确定的对象能被计算。但这被现代代数改变了;代数将算学方法论上的普遍性转化成了对象上的普遍性,因此,不具有确定性的数得以被字符表示,而原本具有确定性的数(比如“2”,意味着2个纯粹单位)则被视作普遍对象的一个实例。
第二个例子来自于分析几何。在笛卡尔和费马之前,几何虽然也设立了普遍关系,但几何形状之间的普遍关系却是以命题的形式被书写的。欧几里得或者是阿波罗尼乌斯发现的所有的普遍关系都是在确定的被给出的形状间建立的,特殊的形状作为特殊的图像对应着特殊的理念;阿波罗尼乌斯所表述的椭圆无法独立于特定的圆锥体,尽管是“每一个”圆锥体。而分析几何则将这图示化的“每一个”直接地转化成了一个包含了“所有”的图像(公式),而特定图像也相应地成为了“所有”的“每一个”实例。
而与这基本原则相违背的一个反例则来自于比例这个概念。在几何原本的第五卷,被讨论的并不是具有比例的特定物,而是具有比例的所有量。第五卷中的线段并非是一个对象,而是一个象征符号,作为一切量的承载者。需要注意的是,这里的线段不是图像,正因为其与线段本身没有严格对应关系;在关于三角形的命题中,“这个”被画出的三角形通过想象力显示了纯粹的数学三角形,但在第五卷中,被画出的“这条”线段并不显示纯粹的线段,而是显示纯粹的量。在这里,纯粹却具体的数学对象被抽离,图像被转化为符号,仅仅与最普遍之物吻合;而相应的,确定对象的消失使得原本具体的量也获得了解放。而第七卷关于数的内容则是在此基础上对第五卷内容的补充。
从这里出发,现代科学将自己视为古典时代科学的继承人;但其继承的概念的意涵发生了巨大的转变。首先,古典时代的宇宙秩序(比例)被转化为了宇宙的法则,原本的确定的数字比例被转化为符号公式;而符号就表示了一种随机性,哪怕是常数,这种常数也是无理由的。这样的可替代性,通过伽利略对柏拉图 蒂迈欧 的利用,最终以运动中的恒定的形式得到了表示。
意向性历史与事实性历史
在讨论这两种意义上的历史之前,先要澄清历史在此处意味着对起源的追问(ἱστορία);而对历史性起源的追问必然超出通常理解的历史本身,而达到了先验历史的维度。
先验历史这样的词是如何可能的?这不是一个矛盾吗?首先得从数学经验讨论。欧几里得将无法证明几何原本的任何命题,如果他不首先建立起相应的定义等;在这个意义上,几何命题有其起源。但这个起源并不是随机的,而是必然的。由此,我们得以讨论先验必然的历史。
先验意味着理念或知识。我们需要理解的因此就是知识的起源。知识的起源无法到朴素经验主义所定义的“经验数据库”中寻找,因此需要往先验主体性中挖掘。这个起源有两个规定:首先,起源必须与“过去”“现在”“未来”各有联系;另外,起源必须在根本上规定理念对象的统一性。
在这两个基本规定下,起源具有了两个“形式”:意向性历史与事实性历史。让我们接下来来考察两者的关系。
意向性历史在根本上被内在时间性(internal temporality,而不是内时间意识inner time consciousness)限定。这一种理念对象统一性的起源规定了在时间内对象的持存。因此,意向性历史的达成来自于对过去、现在和未来的前摄意识更改;在某种意义上,可以认为这是对理念的联结。
但是仅仅凭意向性历史无法解释这样一个疑问:如果仅仅只关注前摄意识对对象的无限修改,理念对象将永远显示为“未完成”的状态。也就是说,在意向性历史中,任何对象都无法成为一个业已完成的思维产物,因而无法成为真正的确切起源。因此,想要确定起源,仅仅凭意向性历史的追溯是不够的。
事实性历史则关涉于“原初给予”。在这个意义上,它并不是前摄意识,而恰恰是后摄意识。它关乎的是“第一个几何学家”的期望;在三角形第一次出现时,它的诸种性质就已经蕴含其中,等待未来的展开了。只有达到这样的原初给予的明见性,起源才真正获得了规定。在某种意义上,事实性历史是无时的,因为时间被凝缩在了一个确定产物中等待展开,而意向性历史则刻画了这样的展开过程。因此,可以说意向性历史是对时间内的各个相关联部分的刻画,而事实性历史则是对无时性与时间性的相关结构刻画。
从激活的方面说,意向性历史的激活是对过去到现在的重演,而事实性历史的激活则是将过去彻底重构为现时当下。
和福柯处在完全不同的维度中,福柯那里的知识型其实根本不是知识,只是条件化的认识模型……
起源重要是因为起源在根本意义上是先验主体的活动,而活动规定了产物,产物与活动的关系不是对称的。
但这并不意味着1.只有一个起源;2.起源与产物间不能相互作用。
用莱布尼茨那里的范畴理解的话,意向性历史是将过去和现在和未来把握在一个在时间当中持存的有机的整体当中,但事实性历史则是达成对理念单子(本体substance)的直观,并将其折叠起来的所有可能性都把握在这一个瞬间当中。一个是时间中的永恒,一个是时间之外的永恒。