Immanuel Kant · 伊曼努尔·康德
Critique of Pure Reason · 纯粹理性批判
Undated 未注日期
纯粹理性批判——散记
Reflections · Pages
关于存在是谓词的批驳(精简版)
首先,如果存在是谓词,那就说明这样的判断是有意义的:“A是存在的”。但在A被添加「存在」这个谓词之前,A本身已经以一种形式存在了,才能允许谓词的添加,因此这只会是循环定义。因为无论这个谓词是本质属性还是必然属性,都无法改变主词是其存在所依赖的源头的事实,而源头不能不存在。
那么,是否可以认为「存在」是一个绝对的普遍谓词?这里亚里士多德已经反驳过了,因为如果「存在」是一个绝对的普遍谓词,那没有外部的它将无法生成自身内部的差异。
此外,如果「存在」是谓词,那么特殊概念与特殊对象之间就会有内容上的差异,而将无法对应,知识就会是不可能的。而关于为什么一般概念完全可以差异于特殊的经验对象,这个问题很好解决:差异并不在于概念与对象的交互,而在于普遍概念与特殊概念的交互。在同样是概念的层级上,理智有抽象的功能。但在特殊经验对象与概念的交互上,两者必须严格在内容上相等同。
最后一个重要的反驳来自观察。“一个人类”或“存在的人类”与“人类”在意义上并没有区别。除非此处的“一”与“存在”各自被赋予特殊的意涵,比如“存在”意味着经验存在。而这个「存在」的意义刚刚已经被批驳过了。即使「存在」意味着「经验存在」,也无法改变其不能是一个谓词的事实。
康德的设想或许类似于,理性与现实之间是一种material implication的关系,只要是现实就必然是理性的,但不是所有理性的都是现实的。这样就会有一个问题,就是理性相对于现实是独立的,理性独立地奠基了现实。那理性与现实就是差异开的两个领域,哪怕现实是可以被理性包含的。假如理性与现实不是同一的而是差异的,那理性就不依靠现实;但理性作为一个确定的领域却必然否定性地与其之外的非-理性的领域处于关系中;或者可以说,理性作为一个确定的领域需要被非理性奠基。而这样的非理性为理性奠基,理性为现实奠基,可得非理性同样为现实奠基;现实因此并不完全被理性所包含。现实因此同样可以是非理性的。我们甚至可以这样说;因为理性与现实是相对独立的,尽管现实被理性所奠基,但理性并不与现实共延,因此现实在被理性包含的同时否定性地限制了理性的领域:“理性是现实与超出于现实的领域”。因此理性反过来同样被现实所限制;现实此时就作为理性的部分而超出了理性。
所以最后,先验为经验奠基,但经验又超出先验,那这就会陷入悖谬;除非,先验与经验被预先综合后直接统一在被给予性中;而这并不等于统觉,假如我们认为统觉是一个描述性的、间接被给予的“经验总已经被先验综合”的事态的话,此处的统一是直接性且主动生成的综合;不再是得到经验反推先验,而是强迫先验和经验的官能综合于其下;这就是理智直观,或者以康德的想法,是自由、理性、道德和行动,其最根本的意志或律令在于“运用你的理性”
先验的要进入经验成为实在,同时经验要永远允许一个超出于自身的可能性,这两者的统一或综合总是需要一个差异于两者的第三方做中介;就是时间。然而时间是被动的综合,总已经被给予的事态;这个被给予同时暗示了一种给予的行动;给予了时间的是超出于时间者,也就是自由。因此,同时外在于先验和经验的就是超验,就是对先验和经验进行主动综合、制造时间的动机。
而这样的话,超验的自由总是进入时间而在时间中制造时间;超验因此并不是完全与先验与经验相隔绝的,正如先验也并不与经验相隔绝,尽管它们确实相区隔。由此,康德最终走向了三批,超验、先验与经验的超混沌自由综合。这就是德勒兹的康德论……
纯粹理性批判内部有着某种不贴合,在开头康德提出统觉和先天综合判断的可能性时,他暗示了概念与其质料在根源处的同质性,也就是概念是可以自己生成自己的真实的;但他后面又将两个世界分离开来,认为真理必须由双重原则作为支撑,概念的真实必须在直观中被给予,而直观不能被还原为概念,而是两个分离的官能,概念因此不能自己生成自己的真实。这里的矛盾应该如何去解决?一种是认为统觉或是先天综合判断只是给出了经验知识的“底部”,本身只是一种“可能性”,也就是说,概念生成真实并施加规定于被外部给予的经验直观;尽管概念规定了直观如何被建构,但直观的被给予性并不是必然或必然如此的。换句话说,这种思路承认了,只要有直观,直观就会如此被规定;但如果没有“如此的”直观,这种规定也就不再有效力。因此,在这种理解下,先天综合判断其综合性并非是真正先天地从概念自身生成的,而是在直观被给予的后天中被回溯性地、范导性地前设出的。当一个范畴被认为是先天范畴,我们并不能从范畴本身推导出其确定内容的生成;相反,我们只能从它在经验中后天的被给予性中反推出它;如真是这样,这样的范畴便会遇到接下来的问题,也就是,它不被经验直接地给予,其存在便无意义,而当它的存在无意义时,它又如何为存在着的经验奠基?假如我们说,过去是在现在被回溯性地重构的,就像所有深谙拉康-齐泽克理论的理论家知道的那样,那半-先验式的过去又怎么反过来奠基如此真切地对时间产生了重构作用的现在呢?
另一种理解则是,统觉确实是真正先天综合的,可以自己生成自己的真实内容;那这里遇到的问题就会是,康德如何能把统觉与“上帝”“世界”等理念区分开呢?如何能在坚持统觉的同时反对对上帝的本体论证明呢?
也就是说,假如坚持第一种解读,思与在是相互独立的,概念并不揭示自身的存在,那概念的真实就要反过来被直观给予;假如诸范畴不能自身特殊化为自身的对象,那对象与诸范畴的差异和对应就要被直观给予;这结论恰恰抵触了前提,本来是要反对理智直观的(不能通过概念直观存在),却又回过来证明了理智直观的必然性(必须通过对存在的直观得到概念)
统觉这样只给出了“凡是现实的都是理性的”,而没有给出“凡是理性的都是现实的”;所以现实必然被理性规定,但是理性又依赖于现实的“被如此给予”而仍然只是一种偶然的必然性。如果要确保知识的确定性,只能在从超出理性的现实推导出理性的同时,还要坚持现实本身亦是理性的;也就是要坚持一个循环式的悖论,也就是要坚持理智直观的原则。
The Kantian Truth Model: the duality of principles
Every truth must be conditioned by two principles, one of form and one of content; the form must reflect into itself from the content so as to make correspondence. If the form is reflected into itself from form, then we have unobjective tautology; if it’s the content reflected into itself from the content, then it’s a misjudgment because they wouldn’t correspond at all(one particular cat is always different from the other). The content cannot be derived from form; if so, it’s unthinkable how should the self-sameness of form ever be different from itself. The form cannot be derived from content, either, since we should then question the transition from change to the change to the unchangeable. To put it into concrete terms, the rationalist method of deriving experience from pure thought is impossible since the “pure” thought must have already been something existent, hence experiential, while the empiricist thought of deriving all concepts from experience is equally impossible as the experience has always already been thought a priori, hence conceptual and abstract. So there’s an inescapable dualism in any truth.
所以康德也意识到这个问题了,如果每个真理都是dualistic的,那真理本身就总是已经综合了这两种原则了……那它就不会是绝对的clearcut……
AI translated the non-English passages; original English passages are preserved. The mixed-language original is available under Original.
A refutation of existence as a predicate (abridged)
First, if existence is a predicate, the judgment ‘A exists’ must be meaningful. Yet before the predicate ‘existence’ is added to A, A already exists in some form that permits the predicate to be added. This can therefore only be a circular definition. Whether the predicate is an essential or a necessary property cannot change the fact that the subject is the source on which its existence depends, and that source cannot fail to exist.
Could ‘existence’ then be regarded as an absolutely universal predicate? Aristotle has already refuted this: if ‘existence’ were an absolutely universal predicate, it would have no outside and thus could not generate differences within itself.
Furthermore, if ‘existence’ were a predicate, there would be a difference in content between a particular concept and a particular object. They could not correspond, and knowledge would be impossible. As for why a general concept can certainly differ from particular empirical objects, the problem is easily resolved: the difference lies not in the interaction between concept and object, but in that between universal and particular concepts. On the same conceptual level, the intellect has the function of abstraction. But in the interaction between a particular empirical object and a concept, the two must be strictly identical in content.
A final important objection comes from observation. ‘A human being’ or ‘an existent human being’ does not differ in meaning from ‘human being,’ unless ‘one’ and ‘existence’ are each given a special meaning here—for example, ‘existence’ means empirical existence. Yet this meaning of existence has just been refuted. Even if ‘existence’ means ‘empirical existence,’ this does not change the fact that it cannot be a predicate.
Kant’s conception may resemble a relation of material implication between reason and actuality: whatever is actual must be rational, but not everything rational is actual. This creates a problem: reason is independent of actuality and independently grounds it. Reason and actuality are therefore two distinct domains, even if actuality can be contained within reason. If reason and actuality are different rather than identical, reason does not depend on actuality. Yet reason, as a determinate domain, necessarily stands in a negative relation to the nonrational domain outside it. Or one could say that reason as a determinate domain needs to be grounded in the nonrational. Since the nonrational grounds reason and reason grounds actuality, it follows that the nonrational also grounds actuality. Actuality is therefore not wholly contained within reason and can also be nonrational. We can even say this: because reason and actuality are relatively independent, although actuality is grounded in reason, reason is not coextensive with actuality. While being contained within reason, actuality therefore negatively limits reason’s domain: ‘reason is actuality and the domain beyond actuality.’ Reason is thus also limited by actuality in return. Actuality now exceeds reason as a part of reason.
So ultimately the transcendental grounds the empirical, yet the empirical exceeds the transcendental, leading to paradox—unless the transcendental and the empirical are synthesized in advance and immediately unified in givenness. This is not equivalent to apperception, if we take apperception to be the descriptive, indirectly given state of affairs that ‘experience has always already been transcendentally synthesized.’ The unity here is an immediate and actively generative synthesis. It no longer works backward from experience to the transcendental, but compels the transcendental and empirical faculties to be synthesized beneath it. This is intellectual intuition, or, in Kant’s terms, freedom, reason, morality, and action, whose most fundamental will or imperative is ‘use your reason.’
The transcendental must enter experience and become actual, while experience must always allow a possibility exceeding itself. The unity or synthesis of these two always requires a third term, distinct from both, as mediator: time. Yet time is a passive synthesis, an always already given state of affairs. This givenness simultaneously implies an act of giving. What gives time exceeds time: freedom. Thus, what lies outside both the transcendental and the empirical is the transcendent, the motive that actively synthesizes the transcendental and the empirical and produces time.
In that case, transcendent freedom always enters time and produces time within time. The transcendent is therefore not wholly isolated from the transcendental and the empirical, just as the transcendental is not isolated from the empirical, although they are indeed distinct. Thus Kant finally moves toward the Third Critique: the hyperchaotic free synthesis of the transcendent, the transcendental, and the empirical. This is Deleuze’s account of Kant…
There is a certain lack of fit within the Critique of Pure Reason. When Kant introduces apperception and the possibility of synthetic a priori judgments at the beginning, he implies that concepts and their matter are homogeneous at their root, meaning that concepts can generate their own reality. Yet later he separates the two worlds, holding that truth requires the support of a duality of principles. The reality of a concept must be given in intuition, while intuition cannot be reduced to concepts. They are two separate faculties, so concepts cannot generate their own reality. How should this contradiction be resolved? One way is to regard apperception or synthetic a priori judgment as providing only the ‘bottom’ of empirical knowledge, itself merely a ‘possibility.’ That is, concepts generate reality and impose determinations upon empirical intuitions given from outside. Although concepts determine how intuitions are constructed, the givenness of intuition is neither necessary nor necessarily thus. In other words, this approach acknowledges that whenever there is intuition, it will be determined in this way; but without ‘such’ intuition, the determination ceases to have efficacy. On this understanding, the synthetic character of synthetic a priori judgment is not truly generated a priori from the concept itself. Rather, it is retrospectively and regulatively presupposed in the a posteriori givenness of intuition. When a category is regarded as a priori, we cannot deduce the generation of its determinate content from the category itself. On the contrary, we can infer it only backward from its a posteriori givenness in experience. If this is so, such a category faces the next problem: when it is not directly given by experience, its existence has no meaning. When its existence has no meaning, how can it ground existent experience? If we say that the past is retrospectively reconstructed in the present, as every theorist versed in Lacan and Žižek knows, how can that quasi-transcendental past in turn ground the present that so tangibly reconstructs time?
Another understanding is that apperception really is synthetic a priori and can generate its own real content. The problem then is: how can Kant distinguish apperception from Ideas such as ‘God’ and ‘world’? How can he uphold apperception while rejecting the ontological proof of God?
That is, if we maintain the first interpretation, thought and Being are mutually independent, and concepts do not reveal their own existence. The reality of a concept must then instead be given by intuition. If categories cannot particularize themselves into their own objects, the difference and correspondence between objects and categories must be given by intuition. This conclusion directly conflicts with the premise. What began as an attempt to oppose intellectual intuition—we cannot intuit existence through concepts—turns back to demonstrate the necessity of intellectual intuition: we must obtain concepts through an intuition of existence.
Apperception thus gives only ‘whatever is actual is rational,’ not ‘whatever is rational is actual.’ Actuality is necessarily determined by reason, but reason depends upon actuality’s ‘being given thus’ and remains merely a contingent necessity. To secure the certainty of knowledge, we must derive reason from an actuality exceeding reason while also insisting that actuality itself is rational. That is, we must uphold a circular paradox, the principle of intellectual intuition.
The Kantian Truth Model: the duality of principles
Every truth must be conditioned by two principles, one of form and one of content; the form must reflect into itself from the content so as to make correspondence. If the form is reflected into itself from form, then we have unobjective tautology; if it’s the content reflected into itself from the content, then it’s a misjudgment because they wouldn’t correspond at all(one particular cat is always different from the other). The content cannot be derived from form; if so, it’s unthinkable how should the self-sameness of form ever be different from itself. The form cannot be derived from content, either, since we should then question the transition from change to the change to the unchangeable. To put it into concrete terms, the rationalist method of deriving experience from pure thought is impossible since the “pure” thought must have already been something existent, hence experiential, while the empiricist thought of deriving all concepts from experience is equally impossible as the experience has always already been thought a priori, hence conceptual and abstract. So there’s an inescapable dualism in any truth.
So Kant also recognizes this problem: if every truth is dualistic, truth itself has always already synthesized these two principles… They therefore cannot be absolutely clear-cut…
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。
关于存在是谓词的批驳(精简版)
首先,如果存在是谓词,那就说明这样的判断是有意义的:“A是存在的”。但在A被添加「存在」这个谓词之前,A本身已经以一种形式存在了,才能允许谓词的添加,因此这只会是循环定义。因为无论这个谓词是本质属性还是必然属性,都无法改变主词是其存在所依赖的源头的事实,而源头不能不存在。
那么,是否可以认为「存在」是一个绝对的普遍谓词?这里亚里士多德已经反驳过了,因为如果「存在」是一个绝对的普遍谓词,那没有外部的它将无法生成自身内部的差异。
此外,如果「存在」是谓词,那么特殊概念与特殊对象之间就会有内容上的差异,而将无法对应,知识就会是不可能的。而关于为什么一般概念完全可以差异于特殊的经验对象,这个问题很好解决:差异并不在于概念与对象的交互,而在于普遍概念与特殊概念的交互。在同样是概念的层级上,理智有抽象的功能。但在特殊经验对象与概念的交互上,两者必须严格在内容上相等同。
最后一个重要的反驳来自观察。“一个人类”或“存在的人类”与“人类”在意义上并没有区别。除非此处的“一”与“存在”各自被赋予特殊的意涵,比如“存在”意味着经验存在。而这个「存在」的意义刚刚已经被批驳过了。即使「存在」意味着「经验存在」,也无法改变其不能是一个谓词的事实。
康德的设想或许类似于,理性与现实之间是一种material implication的关系,只要是现实就必然是理性的,但不是所有理性的都是现实的。这样就会有一个问题,就是理性相对于现实是独立的,理性独立地奠基了现实。那理性与现实就是差异开的两个领域,哪怕现实是可以被理性包含的。假如理性与现实不是同一的而是差异的,那理性就不依靠现实;但理性作为一个确定的领域却必然否定性地与其之外的非-理性的领域处于关系中;或者可以说,理性作为一个确定的领域需要被非理性奠基。而这样的非理性为理性奠基,理性为现实奠基,可得非理性同样为现实奠基;现实因此并不完全被理性所包含。现实因此同样可以是非理性的。我们甚至可以这样说;因为理性与现实是相对独立的,尽管现实被理性所奠基,但理性并不与现实共延,因此现实在被理性包含的同时否定性地限制了理性的领域:“理性是现实与超出于现实的领域”。因此理性反过来同样被现实所限制;现实此时就作为理性的部分而超出了理性。
所以最后,先验为经验奠基,但经验又超出先验,那这就会陷入悖谬;除非,先验与经验被预先综合后直接统一在被给予性中;而这并不等于统觉,假如我们认为统觉是一个描述性的、间接被给予的“经验总已经被先验综合”的事态的话,此处的统一是直接性且主动生成的综合;不再是得到经验反推先验,而是强迫先验和经验的官能综合于其下;这就是理智直观,或者以康德的想法,是自由、理性、道德和行动,其最根本的意志或律令在于“运用你的理性”
先验的要进入经验成为实在,同时经验要永远允许一个超出于自身的可能性,这两者的统一或综合总是需要一个差异于两者的第三方做中介;就是时间。然而时间是被动的综合,总已经被给予的事态;这个被给予同时暗示了一种给予的行动;给予了时间的是超出于时间者,也就是自由。因此,同时外在于先验和经验的就是超验,就是对先验和经验进行主动综合、制造时间的动机。
而这样的话,超验的自由总是进入时间而在时间中制造时间;超验因此并不是完全与先验与经验相隔绝的,正如先验也并不与经验相隔绝,尽管它们确实相区隔。由此,康德最终走向了三批,超验、先验与经验的超混沌自由综合。这就是德勒兹的康德论……
纯粹理性批判内部有着某种不贴合,在开头康德提出统觉和先天综合判断的可能性时,他暗示了概念与其质料在根源处的同质性,也就是概念是可以自己生成自己的真实的;但他后面又将两个世界分离开来,认为真理必须由双重原则作为支撑,概念的真实必须在直观中被给予,而直观不能被还原为概念,而是两个分离的官能,概念因此不能自己生成自己的真实。这里的矛盾应该如何去解决?一种是认为统觉或是先天综合判断只是给出了经验知识的“底部”,本身只是一种“可能性”,也就是说,概念生成真实并施加规定于被外部给予的经验直观;尽管概念规定了直观如何被建构,但直观的被给予性并不是必然或必然如此的。换句话说,这种思路承认了,只要有直观,直观就会如此被规定;但如果没有“如此的”直观,这种规定也就不再有效力。因此,在这种理解下,先天综合判断其综合性并非是真正先天地从概念自身生成的,而是在直观被给予的后天中被回溯性地、范导性地前设出的。当一个范畴被认为是先天范畴,我们并不能从范畴本身推导出其确定内容的生成;相反,我们只能从它在经验中后天的被给予性中反推出它;如真是这样,这样的范畴便会遇到接下来的问题,也就是,它不被经验直接地给予,其存在便无意义,而当它的存在无意义时,它又如何为存在着的经验奠基?假如我们说,过去是在现在被回溯性地重构的,就像所有深谙拉康-齐泽克理论的理论家知道的那样,那半-先验式的过去又怎么反过来奠基如此真切地对时间产生了重构作用的现在呢?
另一种理解则是,统觉确实是真正先天综合的,可以自己生成自己的真实内容;那这里遇到的问题就会是,康德如何能把统觉与“上帝”“世界”等理念区分开呢?如何能在坚持统觉的同时反对对上帝的本体论证明呢?
也就是说,假如坚持第一种解读,思与在是相互独立的,概念并不揭示自身的存在,那概念的真实就要反过来被直观给予;假如诸范畴不能自身特殊化为自身的对象,那对象与诸范畴的差异和对应就要被直观给予;这结论恰恰抵触了前提,本来是要反对理智直观的(不能通过概念直观存在),却又回过来证明了理智直观的必然性(必须通过对存在的直观得到概念)
统觉这样只给出了“凡是现实的都是理性的”,而没有给出“凡是理性的都是现实的”;所以现实必然被理性规定,但是理性又依赖于现实的“被如此给予”而仍然只是一种偶然的必然性。如果要确保知识的确定性,只能在从超出理性的现实推导出理性的同时,还要坚持现实本身亦是理性的;也就是要坚持一个循环式的悖论,也就是要坚持理智直观的原则。
康德式真理模型:原则的双重性
每一个真理都必须以两个原则为条件,一个属于形式,一个属于内容;形式必须从内容反思回自身,才能形成符合。如果形式从形式反思回自身,我们得到的就是不具客观性的同义反复;如果是内容从内容反思回自身,那就是误判,因为它们根本不会符合(一只特殊的猫总是不同于另一只猫)。内容不能从形式推导出来;否则,形式的自身同一性如何竟会不同于自身,就是不可思议的了。形式也不能从内容推导出来,因为那样我们就必须质疑,从变化过渡到不变者的过程。说得具体些,理性主义从纯粹思想中推导经验的方法是不可能的,因为“纯粹”思想必定已经是某种存在的东西,因此是经验性的;而经验主义从经验中推导一切概念的想法同样不可能,因为经验总是已经先天地被思过,因此是概念性的、抽象的。所以,任何真理之中都有一种无法逃避的二元性。
所以康德也意识到这个问题了,如果每个真理都是dualistic的,那真理本身就总是已经综合了这两种原则了……那它就不会是绝对的clearcut……