Junior annual essay ·
Kant at Colonus
In this essay 篇目
If we say Kant’s philosophy comprises a system, we do not miss our mark. Naturally, we want to understand what a system is and how systemicity is conceived. But there is an even more profound difficulty: what motivates Kant, and what ought to motivate a philosopher, to construct a system?
This is a question regarding the ground of Kant’s philosophy, the ground which supports and continually activates a system. This ground can be interpreted in two ways. First, as the theoretical foundation, the center about which a system revolves. Second, as the fluid earth, the problematic undercurrents the foundation intends to repress; whose return, however, is inevitable.
A riddle to which the hero offers a grand solution, only to be taken by surprise later by his own answer. Like Oedipus who cracks the Sphinx’s riddle.
Oedipus knows what ‘man’ is, and he knows ‘man’ for a temporal being: he crawls, he walks, till he finally is dragged along by a cane. But Oedipus is ignorant of his own parents. He does not know the ‘origin of man’, and is for that finally devoured by the raging Furies1. In the jaws of the Sphinx, Kant declares the same answer to the riddle (‘the subject!’), and in a similar vein only adumbrates a more profound difficulty concerning the origin of subjectivity......
But let me dispense with these dissimulating metaphors for now and speak more straightforwardly. The Kantian revolution of metaphysics is but his solution to a problem he identifies in traditional metaphysics. My task is to trace out this development. I shall first delineate the problem as Kant sees it in traditional metaphysics, then his central strategy to get around this problem. After this initial elucidation, I shall provide an analysis of his system while interpreting systemicity in its light. In my conclusion, I shall broach again the question of the Kantian ground.
Part I
If Aristotle assigns the fundamental task of first philosophy to be an inquiry into being, he does not forget to add: being is in a way oneness2. Throughout the history of metaphysics, the question of being has constantly been raised under its arithmetical expression: if there is one that simply is, how come many share in this oneness of being? if there are many that are, how come they all point to one being? In simpler words: how is one many, and how are many one?
This needs further clarification. How being and oneness become synonymous is not necessarily transparent. Aristotle’s response to this problem is tautological. He claims that being and oneness mean the same since it is no different for one to say that ‘something is’ or ‘something is one’3. But here he is already taking this synonymity between being and oneness to be self-evident, which I am not convinced is the case. I shall therefore delve deeper to inquire into the origin of this synonymity.
For ‘traditional metaphysics’ (I use the term to loosely refer to pre-Kantian metaphysics), the question of being is largely equivalent to the question of the logical affirmation of being. This is because an equivalence between the order of thought and the order of things is presupposed. If Aristotle’s philosophy is teleological, it is because his method of transposing logical determinations to ontological ones presupposes the transparency of thought to being, which he shall substantiate only at the very end in the notion of a divine nous. We shall see later how Kant challenges precisely this equivalence between thought and being when he breaks away from the tradition. As for now, I restrict myself only to drawing out the consequences of this presupposition for traditional metaphysics.
The difference between being qua oneness as opposed to non-being qua manyness, therefore, amounts to a difference between affirmation and negation as functions of judgment. This suggests to us to excavate the original connection between being and oneness via interpreting our act of logical affirmation. Now, I am presented with a general set of things, whose members are A, B, C, D, X, etc., of which I am to either affirm or deny the existence. An asymmetry between affirmation and negation gradually emerges as I continue to pass judgments. An affirmation of the existence of X takes the form of ‘X is’, which is simple and exclusive, i.e. it does not involve in its content anything other than X. A simple presence of X is sufficient for its affirmation, independent of anything else. X is, and only is X. A negative judgment takes the form of ‘X is not’, which opens up an indeterminate realm of possibility. For something must be in lieu of X, in whose presence I am allowed to make the judgment at all. For instance, a round square is not, but its constituents, the circle and the square, must exist, so that something can be present for me to judge. X is not, but something yet unknown must take its place. And here we see the fundamental asymmetry. The affirmation of X is simple and determinate. The negation is indeterminate. Merely by stating ‘X is not’, I do not obtain the positive content that makes this negation possible. ‘X is not’ implies ‘there is something non-X’, yet ‘non-X’ can potentially contain A, B, C, D, etc., all these being contents of something I know nothing about.
I now turn to one single object. My task is to determine its being. I first say, this object is not a chair. But this only rules out one possibility among infinitely many others. In this case, the object remains indeterminate for me. But what if I try a different approach? I say an object is either a body or a non-body; if a body, either cold or not cold, and if not cold, either hot or lukewarm. Through such an eliminative procedure, hopefully I can discover all the hidden content of an object. But before I proceed in this way, I must have already made an affirmation. I must first exhaust the whole terrain of possibilities to be able to obtain only a finite number of options. Before the negation of ‘cold’ can yield a determinate result, namely, ‘hot’ or ‘lukewarm’, I must first know that the object is a ‘body’. But in the immediacy of a negation, nothing like this happens. By negation of ‘cold’ in its pure immediacy, I know nothing more about the object, if I have not already put it under the mediating genus of ‘body’, and without knowing beforehand the object in question is a body. For if the object is not a body, a mere negation of ‘cold’ does not eliminate and narrow down the scope of possibilities for me, as everything that is not a body is technically ‘not cold’. Kant warns us exactly on this point when he says: ‘if somebody were to say that every body has either a good or a bad smell, a third case is possible, namely, that it has no smell at all, in which case both conflicting propositions would be false’4. Therefore, in order to negate determinately, a prior affirmation is necessary. This equally applies to the highest kinds, for they are also known to be of a finite number, thus depend on a prior affirmation.
The fundamental asymmetry between affirmation and negation is thus clarified. Affirmation by itself is always simple and determinate, negation obscure and indeterminate, for one never determines anything through negation without a prior mediation of affirmation. We can explain on this account what is implied in the problem of one and many. Manyness in its original meaning precisely does not mean the arithmetical ‘many’, but rather ‘indeterminateness’ or ‘infinitude’. The obscurity here is, if manyness originally means ‘infinitude’, this has nonetheless been translated to the arithmetical concept of ‘many’. And it is not clear why the infinitude of non-being should be more than the finitude of being, for they might not be quantitatively comparable concepts. It seems that only two finite things can be compared by their quantities, for an infinite thing has no determinate amount at all. Can we say a body is more than a spirit? And if we think the latter absurd, why do we not think that of the former? How does the qualitative difference between finitude and infinitude come to be taken as a quantitative one?
If we look at oneness and manyness only in abstraction as independent principles, this translation of the metaphysical to the arithmetical, or the qualitative to the quantitative, remains unintelligible. For the true thrust of the metaphysical question is to be elucidated only in a concrete problem. That is, the fundamental problem of metaphysics is formulated in an arithmetical way precisely because its most original concern is directed to the foundation of arithmetics: the questions ‘how is one many?’ and ‘how are many one?’ are only ramifications of the question, ‘How is number possible? What is the essence of number’? Again, the telos of the inquiry directs the inquiry itself; the task of metaphysics to ground arithmetics predetermines its arithmetical nature:
“By Zeus!” he said, “I seem to be far from thinking, I suppose, that I know the cause concerning any of these things, I who don’t even allow myself to assert that whenever anyone adds a one to a one, the one added to or the one that was added has become two, or that the one that was added and the one to which it was added became two by the addition of the one to the other. Here’s what I wonder about: When each of the two was separate from the other, then each was one and the pair were not two, but when they came close to each other, this then became the cause of their becoming two –– the concourse that comes from their being placed close to each other. Nor again can I yet be persuaded that if somebody splits a one apart, this –– the splitting –– has in turn become the cause of their having become two. For then this cause comes to be the contrary of the former cause of their becoming two. Then it was because they were led close to one another and were added, one to the other, but now it’s because they’re led away and separated one from the other. Nor do I any longer even persuade myself that I know why a one comes to be nor why, in a word, anything else comes to be or perishes or is by this way of proceeding. Instead, I’ve randomly smushed together another way myself, and that former one I don’t tolerate at all.”5
In the passage cited above, Socrates is asking none other than the question, ‘how are manyness (the principle of “the splitting”) and oneness (the principle of “bringing close”) combined to make a number’? The two causes, the ‘splitting’ and ‘bringing close’, seem contrary, only because it is not mentioned that the number ‘two’ is a combined product of these ‘two’ causes. Regarding this combination of causes, many likely stories are contrived. One common narrative goes like the following. First, ‘one’ is a simple affirmation. Yet the determination of ‘one’ soon reverts to its own un-determination, for its simplicity is without content6. With the bounds of oneness thus breached, the one nonetheless coheres with that which it is not, and what it is not is nevertheless itself, thus the one that is not and the not-one that is one is maintained in a ‘twofold-ness’. This ‘twofold-ness’ is unified in the one through the one, and is held together in the one, which, distinct from the twofold it encompasses, is then added to the twofold-ness as a third. Within the twofold-ness, however, the indeterminate remains, and continues to generate new pairs... This is a generative account of number qua evens and odds, and this same narrative runs through traditional metaphysics: the ‘one’ in its self-sameness generates a power of manyness, with which the unity of one combines to make ‘number’, thus making arithmetics possible. Therefore, from the very start, we see metaphysics tasked with the grounding of arithmetics, and this has remained one of its most lasting motives, as is obvious in the case of Leibniz, who strives to construct the mathesis universalis.
Since the end-goal predetermines the procedure, it is no wonder traditional metaphysics follows an arithmetical paradigm. We can briefly recall two samples of traditional metaphysics to convince ourselves that this is the case. For this purpose I choose to interpret Aristotle and Plotinus, since they represent, in relatively systematic ways, two major schools of traditional metaphysics. My interpretation is by necessity incomprehensive and cannot do justice to the interpreted, but I believe the arithmetical nature of traditional metaphysics, through my delineation, can be brought clearly to the eye.
The pinnacle of Aristotle’s metaphysics is the active nous. However, in order for nous to be active, it must in itself embrace a passive side to be acted upon. The pure activity of nous thus introduces in itself its own opposite. Nous in this sense cannot be a highest genus, since it is self-generative. It acts upon itself and self-differentiates, i.e. it is its own differentia. Below the divine nous are the highest sources (categories), which are the first products of the activity of nous. These highest sources, driven by the same divine activity, combine with each other in pairs, and are genera or differentiae to each other. These combinations bring forth many intellectual products. These new products in turn combine as genera or differentiae to each other and generate more complex products. As this process of generation proceeds unrelentingly, eventually some products are composite of so many antecedents that our finite intelligence loses sight of those ‘combinations’ completely, even as we cannot intuit a chiliagon or a thousand grains7. It is in this way our sensible world composite of form and material is generated and placed at the lower rung of the ontological ladder, with aether on the outermost sphere as the encompassing form and earth at the center as the encompassed material8, form being a power of unification that more closely indicates the original unity, material a passivity farther removed from that unity. Yet both form and material are far enough from the original unity to look obscured (hence un-intellectual) to us, for they each contain too great a number of antecedents, only more or less so.
It is true that Aristotle vehemently criticizes Pythagoreans’ doctrine of eidetic number, especially Plato’s supposedly ‘unwritten doctrine’ that grounds arithmetics through metaphysical speculations. Yet we must not fail to see that that criticism is only directed at the unreflective method the Pythagoreans have employed in such grounding. For Aristotle, the grounding of arithmetics through metaphysics cannot be a direct one, as if whenever two objects are present, ‘two-ness’ is itself present9. Rather, there is a fundamental gap between the mathematical operations of human beings and the ontological progression of real beings, for when we count, we are only abstracting mathematical units from the concrete determinations of all antecedents involved in the objects presented to us10. Our counting only reduces things to homogeneous units and does not see things in concreto. The grounding in question for Aristotle cannot therefore be as direct as how some specific eidetic number grounds its corresponding mathematical number, like ‘two-ness’ for two, but how the metaphysical account in general of the progression of beings should ground arithmetics in general. The Aristotelian grounding of arithmetics therefore proceeds as follows.
Since everything is involved in the original unity of nous, everything shares the same ‘oneness’ of being, so that all individuals can be reduced to homogeneity, though they remain mathematically distinct. In other words, each being is itself ‘one being’ as any other being. The possibility of number hence still hinges on the combination of the two metaphysical principles of oneness and manyness, so that things are distinct as many, yet limited by one common measure. But it is only in a realm where the hierarchy of beings is indiscernible that we are able to treat things as homogeneous. This realm is not inherent in the order of things, but only relative to our finite intelligence. For each thing in itself has its own unique composition of antecedents that is irreducible to any other, not even to its own immediate antecedent, as a particular cat is not the universal cat. It is only due to our abstraction that unique beings are reduced to the same. Now, abstraction is rooted in the sensible, for as generation proceeds to infinity, everything becomes equally obscured to us; only then can we indiscriminately treat all things as sharing in the same ‘form’ (the same ‘oneness’) and as homogeneous units. From there we even apply this abstraction derived from the sensible back onto the intelligible. Number for Aristotle therefore originates from an abstract measuring of a magnitude. As Aristotle himself says, ‘for a thing to be in number is for there to be some number of the thing’11. Yet even if arithmetics can only be thus indirectly grounded through metaphysics, as the order of things and the order of number do not completely coincide, (for instance, since there are ten highest kinds12, the twofold nous in its generative activity would make a jump from 2 to 12 by our human reckoning), the grounding still strongly entails a hierarchical order proceeding from finitude to infinitude.
Plotinus’ doctrine is no different on this front. Starting from the bottom of the ontological ladder, matter is relatively many. It is then unified through the unity of form, and thus unified with form, constitutes natural things. This unifying power of form, however, stems from a higher principle. Form is but an image of what Plotinus calls ‘The Highest Reason’, which then unifies, and is unified with, all natural things to constitute Nature. The Highest Reason in turn is a reflection of a higher principle, i.e. the higher part of The Soul. The higher part of The Soul unifies Nature and is unified with it to constitute The Soul proper. The higher part of The Soul again only reflects an activeness of The Intelligence, which similarly unifies The Soul to make The Intelligence. But the activeness of The Intelligence is no different from The One, which encompasses The Intelligence13.

Again, we see the same arithmetical paradigm at work. The One is the original unity which splits into a twofold-ness, i.e. the activity and passivity of The Intelligence. The Passive Intelligence, qua The Soul, is then divided into a higher and a lower part. The lower part of The Soul, qua Nature, further divides. This progression proceeds until finally the power of manyness is fully manifest through the lowest rung, i.e. matter. It is to be noted, however, that all beings, not just matter, generated in this progression are infinite, since they all contain the principle of manyness. The only difference here is how contained and suppressed the power of this principle is. The closer to The One, the more unified and concentrated is the infinitude inherent in the being; that is, the more is the infinitude ‘limited’.
If I seem to digress much from a simple exposition of Kantian philosophy, I hope this digression is not taken to be irrelevant to my task. Kant is born into a tradition, and his originality cannot be appropriately appreciated without first understanding what that tradition entails. Through this section, it is clear that traditional metaphysics operates in an arithmetical paradigm. This is not without consequence for metaphysics: first, it presumes the question of being and non-being to be equivalent to the question of oneness and manyness; next, it constructs an ontological hierarchy that roughly follows the progression of numbers, if not precisely coinciding with it, so that the higher on this ladder a being is placed, the simpler and the more unified this being is; thirdly, since all beings are involved in the arithmetical progression and encompassed by the same original unity, beings are primarily different by ‘degree’, and most of the distinctions are reducible to quantitative differences, namely, how ‘much’ a being partakes of the original unity; finally, since every being is conceived in the original unity as a ‘quantity’, and quantities are related to each other through ratio and proportion, logos becomes the predominant force in our understanding of being. The power of reason gathers what is many into what is one in a ratio, and many such ratios into one proportion (analogia), and in so doing places manyness under the command of a superior unity, judging and comparing beings according to ‘how much closer’ to the origin each of them is. Reason becomes not only the unifying and commanding power, but the ultimate judge. In the next section I shall argue that Kant’s Critique of Pure Reason is nothing less than a critique of precisely this arithmetical paradigm and the traditional metaphysics that follows it.
Part II
Kant’s opposition to the arithmetical paradigm is motivated by a difficulty he sees in traditional metaphysics. This difficulty comes with the theory of participation. On the one hand, this theory dictates a group of particulars to partake of a universal. On the other, these universals themselves must be particulars that partake of higher universals. This series is potentially infinite, and the difficulty here is how to move along this infinite progression. This difficulty is twofold.
First, starting from our sensible world, all perceived beings are presented as obscured. In order to ascend to the ground, I need to discover the antecedent contained in the composition of the thing. To ensure my success, I need a method. This method is usually termed ‘abstraction’, as commenced by Aristotle and inherited by Descartes, which consists in varying in imagination the modes a certain being can possess, thus sieving out the essence of that being from all its accidents. But this incurs the problem of infinite regress. Since a variation in our imagination can proceed in infinite ways, it is impossible to conclude the process without a leap of faith. When Descartes is meditating on that wax, he abstracts its essence from its color, its texture, its circumstance, its position, etc. Yet before he abstracts it from its color, he must first abstract its color from a particular red, blue, green... Before he abstracts its color from a particular red, he... Abstraction is interminable as a procedure dependent on the variations in imagination, and we are constantly stuck in the process. This induces Descartes to conclude that there must be some act more than simple variation ad infinitum that provides us with rational knowledge: viz., intellectual intuition14.
Then, suppose through intellectual intuition, I successfully sieve out the ultimate substance from everything else, this does not help my knowledge. For whatever can be abstracted away from a being is accidental to that being. And for the original unity, i.e. the substance, thus retained through abstraction, everything that is derivative of it would only be accidental to it. I then encounter a second problem, namely, the production of the world is now purely accidental and thus inexplicable.
These two problems, i.e. infinite regress from bottom to top and mysterious production from top to bottom, are problems because they fail to satisfy the principle of sufficient reason. With infinite regress, the ultimate ground for everything becomes unreachable, while with mysterious production, it is only due to God’s will, instead of necessity, that the world which is inferior to him is created.
Kant’s genius is that he sees the present difficulty not as accidental to traditional metaphysics, but rather inherent in its very beginning. Because traditional metaphysics follows an arithmetical paradigm, it blindly presumes that ‘oneness’ and ‘manyness’ are only quantitatively different, rather than qualitatively distinct. In truth, what Kant does is simply to supplement the arithmetical paradigm with a geometrical paradigm; ‘oneness’ and ‘manyness’ are not different amounts, but different kinds, as rectilinear and curvilinear figures are different.
It is likely that Kant owes this discovery of the geometrical approach to metaphysics to Spinoza, for Spinoza, prior to Kant, had already challenged the traditional arithmetical paradigm, which constantly subjects body to thought for its supposed inferiority and passivity. Contrary to this tradition, Spinoza claims that nobody as yet knows what bodies can do15, thus releasing body from the yoke of thought, and places the corporeal (Extension) on equal footing with the intellectual (Thought), each an attribute amongst the infinitely many attributes of God.
This is not to say Kant is secretly a Spinozist. Quite the opposite. As we shall see later, Kant has a different agenda in employing the geometrical approach. But even before that, we should understand Kant’s philosophy is only geometrical to an extent. The apparent geometricism of his philosophy is merely a supplement to an overall arithmetical structure.
Let us now delve deeper into the concrete of his new approach, so as to understand his solution to the present difficulty. As said above, geometricism in this context means Kant discerns between ‘oneness’ and ‘manyness’ not a quantitative difference but a qualitative distinction. Yet, why is this the solution? The Kantian solution cannot be properly understood without determining more precisely how the difficulty first comes about in traditional metaphysics.
For traditional metaphysics, ‘oneness’ and ‘manyness’ are only quantitatively different. Each being as a combination of both principles therefore stands more or less proximate to one or the other end, i.e. being or its privation, on a continuous spectrum. The closer a being is to ‘oneness’, the simpler, the more unified and the more perfect it is. As truth has to do with clarity and distinctness, which is proper only to unity, the pursuit of knowledge is to minimize the expression of ‘manyness’ and maximize that of ‘oneness’, so as to ascend from particulars to universals, from the sensible to the intellectual. And since all beings are only different by degrees, the knowledge one presumes to have of each being is not in principle different. The knowledge one has of the ultimate unity is the same kind of knowledge one has of any other being and is comparable to any other knowledge, only more eminent and perfect. This first allows the knowledge of the ultimate unity to be a ground for all knowledge, then necessitates it. For without the ground provided by the ultimate unity, all knowledge becomes indeterminable and void. One does not know where one is located in the infinite chain without knowing where the chain starts, even as one does not know the number without knowing the unit. Yet, since the chain is infinite, one can never traverse the chain step by step. And if, on the contrary, one starts from an intellectual intuition of the origin, it is difficult to see why the unity should place beings in an order of proximity as numbers are placed from less to more, when everything is equally encompassed by it. The former amounts to the problem of infinite regress, the latter the problem of mysterious production.
The geometrical approach enables Kant to achieve one simple thing: the redefinition of knowledge. If oneness is not ‘more perfect’ than manyness but is of a different kind, knowledge should not be a reduction of what is less perfect to what is more perfect, but a correspondence between two qualitatively distinct principles. For Kant, every being consists of a unity and its corresponding manifold, both of which contribute equally to the full knowledge of this being. This is Kant’s famous maxim: ‘thoughts without content are empty, intuitions without concepts are blind’16. The concept of an object, as Kant describes it, is the manifestation of a principle of unity: ‘knowledge consists in a determinate relation of given representations to an object; and an object is that in the concept of which the manifold of a given intuition is united’17. Concepts provide the unity that organizes the manifold of intuitions, not to reduce or abstract away from the manifold, but only to relate the unity to the manifold. Knowledge of the pure unity, which according to traditional metaphysics is the grounding of all knowledge, is thereby expelled from the foundation. For Kant now declares: if knowledge is only a composite of unity and manifold, the ‘knowledge’ of the pure unity, which is of a different kind from the composite, is no knowledge at all. The ground that traditional metaphysics seeks is thus illusory.
This redefinition of knowledge, achieved through the introduction of geometricism, solves the twofold difficulty Kant sees in traditional metaphysics. First, since knowledge is the correspondence between two qualitatively distinct principles, the true ground lies in this very correspondence, in lieu of the original unity. Therefore, one does not need to ascend from the sensible all the way up to the intellectual in order to ground knowledge, for knowledge occurs ‘here and now’, in a manner of speaking. This is the force of the Transcendental Analytic, for it establishes precisely this correspondence between concept and intuition. For instance, in order to know a causal relation, we need not go through everything in the causal chain to trace back to a first cause. Rather, what is required is only an irreversible succession of two representations:
When, therefore, I perceive that something happens, this representation contains that something precedes, because by reference to what precedes the appearance receives its relation of time, that is, to exist after a time in which it did not exist. Its determinate position in time within this relation can only be assigned to it, if in the antecedent state something is presupposed on which it follows always, that is, follows according to a rule. It thus follows, first of all, that I cannot reverse the series, and place that which happens before that upon which it follows; secondly, that if the preceding state is posited, the other event follows inevitably and necessarily.18
All other elements besides The Second Analogy in the Analytic should be interpreted in the same vein, for all these categories are what Kant calls formal conditions, i.e. rules applied by the subject only to singular and immediate representations. An analogy might help elucidate this notion. ‘Each’ and ‘all’, though both expressive of a universality, have different grammatical functions, for ‘each’ is followed by a singular noun, while ‘all’ by a plural. Following this distinction, formal conditions are rules applied to ‘each and every’ representation, since knowledge for Kant is always an immediate and singular application of the universal rule, whereas for traditional metaphysics, knowledge only resides in the ‘totality of all beings’ grounded in and through an original unity.
If the original unity is no longer the ground, there is no need to traverse the infinite chain in search of that unity, and no infinite regress can thus be incurred. This has both an ontological and an epistemological implication. First, universals are neither more real nor more perfect than particulars, as is established. Second, a knowledge no longer requires a higher knowledge to be the guarantee of its validity. This addresses a criticism constantly raised against Kant, i.e. in order for me to know my faculty of knowledge, do I not put myself in a vicious circle? In order to know, I have to know that I know, and have to know that I know I know, etc. In this view, the Critique of Pure Reason would be an impossible project, since Kant presupposes he knows what knowledge is. In traditional metaphysics, the solution to this problem is to divide knowledge into more or less universal categories. The highest knowledge is infinitely reflective and self-grounding, e.g. the absolute transparency of divine nous or the unity between Being and Thought in The Active Intelligence, so as to be a stoppage to the infinite regress of reflections. This solution, as we can deduce, is defective. Kant departs from the tradition and provides his own solution. The knowledge of knowledge, or the reflective knowledge, is not a higher knowledge, but only an elucidation or an analysis of a synthesis that has already been carried out:
This act we shall call by the general name of synthesis, in order to show that we cannot represent to ourselves anything as combined in the object without having previously combined it ourselves, and that of all representations combination is the only one which cannot be given through objects, but, as it is an act of the subject’s self-activity, can only be carried out by the subject itself. It can easily be perceived that this act of synthesis must originally be a single act and must be equally valid for all combination, and that its dissolution, that is analysis, which seems to be its opposite, always presupposes it. For where the understanding has not previously combined, there is nothing for it to dissolve, because only as having been combined through the understanding could it have been given to the faculty of representation.19
The Kantian epistemological project is therefore not a grounding of knowledge through a higher reflective knowledge, but an analysis of the latent syntheses, i.e. a transcendental analytic. The ‘knowledge of knowledge’ is nothing but a tool that clarifies what is by itself already knowledge, a ‘reflection in the mirror’ that shows what already is in its full right, i.e. the synthetic activity.
The problem of mysterious production is similarly resolved. This problem stems from the apparent arbitrariness of an arithmetical order. Why must all things be placed on an ontological ladder, some more perfect than others? Are these not all equally encompassed by ‘the one’? This incongruence in thought comes about only because the arithmetical paradigm of traditional metaphysics is itself arbitrary. Metaphysics is addressed to the opposition of ‘one’ and ‘many’ only because it concerns itself with the foundation of arithmetics. Now, Kant has not gone so far as to claim that the question of being and non-being has nothing to do with the question of one and many, which would be an equally nonsensical position. But he does see that arithmetical progression is not the only model via which one can conceive of one and many, and the ontological ladder is not the only possible order of things.
Since ‘one’ and ‘many’ have a qualitative distinction, they can be mixed not only in varying proportions, but also in varying relations. A concept is a mixture of ‘one’ and ‘many’ in a way qualitatively different from the way an intuition is a mixture of them. In the next section, I shall analyze in detail what these qualitatively distinct structures are and how they are interrelated. For now, it is sufficient for us to see that the problem of mysterious production is solved when the arithmetical order becomes just one structure in a larger system of corresponding structures. Its arbitrariness disappears as soon as it is grounded in relation to other structures in a harmonious totality.
Part III
A system is a harmonious whole that has all its parts so closely related to the totality, that each part is less of a constituent than a certain perspective on the totality. Each element of a system is reflective of the whole. This is what Kant understands a system to be:
When one is concerned to determine a particular power of the human soul in terms of its sources, contents, and bounds, then indeed, by the nature of human cognition, one cannot start except from the soul’s parts, their exact and (as far as is possible according to the current situation of what elements of the soul we have already acquired) complete exhibition. But there is also a second attentiveness that is more philosophical and architectonic: viz., to grasp correctly the idea of the whole and, on the basis of this idea in a pure power of reason, to fix one’s eyes upon all those parts in their reciprocal reference to one another by means of their derivation from the concept of that whole. This examination and warrant is possible only through the most intimate acquaintance with the system. Those who were irked by the first investigation and hence did not consider acquiring that acquaintance worth the trouble do not reach the second level, viz., that of the overview, which is a synthetic return to what had previously been given analytically; and it is no wonder if they find inconsistencies everywhere even though the gaps that suggest these inconsistencies are to be encountered not in the system itself but merely in their own incoherent progression of thought.20
It is in this way the Transcendental Aesthetic cannot be understood without the Transcendental Logic, which in turn remains unintelligible without us grasping the mutual dependence between the Analytic and the Dialectic. When I call Kant’s philosophy systematic, I understand its every part to be reflective of the whole, as a perspective on the whole.
I shall follow Kant’s suggestion and analyze the elements of his system with the whole in my view. As established in last section, ‘one’ and ‘many’ are mixed for Kant in qualitatively different ways that generate different structures. These structures are then elements that constitute the totality of Kant’s system. The analyses of each structure roughly correspond to the division of the book, viz., the Aesthetic, the Analytic, and the Dialectic.
Right off the bat Kant has made the distinction between intuitions and concepts clear. In the Metaphysical Expositions in the Aesthetic, Kant says of space, that
Space is not a discursive or, as we say, general concept of the relations of things in general, but a pure intuition. For, first of all, we can represent one space only; and when we speak of many spaces, we mean only parts of one and the same unique space. Nor, secondly, can these parts precede the one and all embracing space, as its constituents, out of which it can be assembled; rather, they are thought only as in it. Space is essentially one; the manifold in it, and therefore the general concept of spaces in general, arises entirely from limitations. Hence it follows that, with respect to space, an a priori intuition of it (one that is not empirical) must underlie all concepts of it. In the same manner, all geometrical principles, e.g. that in a triangle two sides together are greater than the third, are never derived from the general concepts of line and triangle, but from intuition, and derived moreover a priori with apodictic certainty.21
He says the same about time: ‘different times are only parts of one and the same time; and that representation which can be given only through a single object is an intuition’22. In other words, one and many in intuition has the structure of inherence. Many parts of an intuition inhere in one singular intuition, whereas the particular intuitions as limited parts of the whole inhere among many other parts, all homogeneous with each other. This structure is distinct from an arithmetical one, where an ontological hierarchy is the natural consequence. Concepts, on the contrary, do follow the arithmetical paradigm: ‘now it is quite true that every concept must be thought as a representation which is contained in an infinite number of different possible representations (as their common characteristic), and therefore the concept contains these under itself; but no concept, as such, can be thought as containing an infinite number of representations within itself’23. Through understanding, concepts are organized in a hierarchy, the particular subjected to the universal.
It is not difficult to see that both intuitions and concepts are mixtures of ‘one’ and ‘many’, as both require the unification of a manifold. Yet the structures by which ‘one’ and ‘many’ are mixed are distinct for them. For intuitions, each particular inheres in a larger one, and there is no distinction of rank between the particular and the universal; everything is organized on a homogeneous plane. This structure can be understood succinctly as one-among-many, as opposed to the structure of understanding, which is one-above-many. However, if knowledge requires the correspondence between the two structures, how is it possible that these two orders of things should ever work together?
Now, we can see how Kant has transformed metaphysics and relocated the ground. The ground is no longer ‘at the foundation’ of things, but rather ‘in between’ orders of things. The ground is no longer the immobile center, but becomes a transitive relation: this is the true significance of the Copernican Revolution. In order to address this ‘grounding’, we now move to the Analytic.
Knowledge for Kant ‘consists in a determinate relation of given representations to an object; and an object is that in the concept of which the manifold of a given intuition is united’24. This already answers the question for us: the two orders work together because these two orders constitute each other and are originally interdependent, so that one cannot be itself without the other.
The unity of space and time in intuitions is not possible without a higher unity in concepts. The concept, in its most originary determination, is not a ‘thing’, but an activity of combination that unifies the manifold of intuitions into a whole, and is the source of this unity of intuitions. As Kant makes explicit in his footnote to B161:
Space, represented as object (as required in geometry), contains more than mere form of intuition, namely, also the comprehension of the manifold given according to the form of sensibility into an intuitive representation, so that the form of intuition gives only a manifold, while the formal intuition give unity of representation. In the Aesthetic I had simply ascribed this unity to sensibility, in order to show that it precedes any concept, though it presupposes a synthesis not belonging to the senses, and through which all concepts of space and time become first of all possible. For as by this synthesis (the understanding determining sensibility) space and time are first given as intuitions, the unity of this a priori intuition belongs to space and time, and not to the concept of the understanding.25
In other words, the unity that constitutes space and time as such originates in our understanding, yet the unity per se is not separate from our plane of intuitions. The activity of unification proceeds from concepts and is projected onto the plane, thus generating an unified intuition as one-among-many. Though its source is a ‘higher unity’ in the understanding, the unity of intuitions remains on the plane of sensibility. The initial manifold that concepts subject and organize into a hierarchy is received as a unity through sensibility, yet it is a unity without any distinction of rank.
It is obvious that the concept of ‘right’ can never be sensible while the appearance of ‘body’ is not by itself a concept. Yet at the limit case, where the unity of intuitions (one-among-many) originates in the unity of concepts (one-above-many), the difficulty of keeping things apart is accentuated. This is where schematism and the faculty of imagination comes in. The role schematism plays is twofold: it unites sensibility and understanding in mutual affinity, yet also separates them, so as to rescue Kant’s philosophy from becoming yet another unfortunate example of traditional metaphysics.
Kant is clearly aware of the danger. After speaking of the mediating role of schemata, he does not forget to stress that ‘appearances must therefore be subsumed not directly under the categories, but only under their schemata’26. Schematism must be a bridge that separates two banks.
Kant lays out the structure of schematism in B180. He determines that
the image is a product of the empirical faculty of productive imagination; the schema of sensible concepts (such as of figures in space) is a product and, as it were, a monogram of the pure a priori imagination, through which and according to which images first become possible. These images, however, must always be connected with the concept only by means of the schema which they designate; in themselves they are never fully congruent with the concept.27
Two things are to be noticed here. First, schemata are not particular sensible images. They are rules ‘for the determination of our intuition, in accordance with a certain general concept’28. Yet, a schema is not in itself general. It only refers to a general concept, for it is a figure following the rule ‘according to which my imagination can generally register’29. Imagination is thus placed at a strange terrain of incipience, where a general figure is about to be raised above the particulars, yet without becoming a concept proper. The best way to understand this, I believe, is to think of the structure of schematism as an enfolding of many into one. Schemata are figural horizons that enfold appearances, without being above particulars. This is already implied in the Aesthetic, for space and time are there determined to be intuitions both singular and formal, as unities of intuition that encompass all possible intuitions. Opposed to either one-among-many or one-above-many, the faculty of imagination embraces an intermediate position, that of many-in-one.
Even without completing our full analysis of Kant’s system, it is not difficult to see that on the whole Kant has preserved the arithmetical paradigm. Concepts are still the higher unities to which intuitions are subjected. Compared to concepts, intuitions are ‘many’, for the structure of sensibility, i.e. one-among-many, entails that the unities of intuitions are those of flocks and folds, which unities must originate from a purer unity. Yet if the paradigm of traditional metaphysics is salvaged, it is not without sacrifice. First, intuitions are ‘many’ not in a quantitative sense. In his criticism of Leibniz, Kant points out that ‘the concept of “right”, as used by common sense, contains all that the subtlest speculation can draw out of it, except that in its ordinary and practical use we are not conscious of the manifold representations contained in these thoughts’30. Concepts therefore contain just as many representations as intuitions. The only difference is, a concept contains the manifold under itself, while an intuition contains the manifold within. Understanding is not inherently simpler or more perfect than sensibility. Rather, regarding clarity and distinctness, each should be judged by its own measure. Hence, even if understanding is the dominant faculty in our disposition for knowledge, this is not due to its superiority, but because sensibility and understanding can only work in this manner for knowledge to be possible. Concepts do not wield tyranny over intuitions, but only reign in harmony and freedom to which sensibility willingly submits. The horizon of our imagination is where this freedom is actualized.
Sensibility, imagination and understanding comprise the disposition of knowledge. Yet we still lack one last piece of the puzzle. In order to explain the unity of a concept, it is necessary that an ultimate unity be sought, for it is only from this unity that the power of unification originates.
On the face of it, it seems at this point traditional metaphysics has been fully resuscitated in Kant’s system, and his whole endeavor has miscarried. For is not this original unity rediscovered precisely the ultimate ground traditional metaphysics seeks? Does not this original unity, when posited, reclaim the problem of ground to itself? This is, however, not the case. It is true that the unity of a concept must have originated in a highest unity, but it is not necessary that the concept is ‘grounded’ in that unity. Rather, the ultimate unity is only projected to our understanding, while the ground is in the middle, even as imagination is the ground between sensibility and understanding.
Similar to the triad of sensibility-imagination-understanding, a mediating principle is to be found between the faculty of understanding and the faculty of reason. Just as traces of schematism are already present in the Aesthetic, we should expect the same with the Analytic.
The unity of reason is determined as the unconditional totality31. Its structure is one-for-many. It is crucial to note that this structure does not primarily entail a quantitative maximum of oneness, but rather an identification of the two principles, so that infinitude is grasped ‘all at once’ in the ultimate unity. We can recall the previous distinction between ‘each’ and ‘all’. In all other structures, there remains a distance between one and many, so that the infinitude is not given all at once, but always incompletely. For intuitions, I traverse them one by one in the plane; for schemata, I enfold each given intuition into the horizon; for concepts, I ascend step by step on the ladder. It is only with the unity of reason that the complete identity between one and many is reached.
Now, what could mediate the unity of understanding with that of reason? It has to be a principle that is unconditional, yet still works with ‘each’ singular representation. This principle is given in the Analytic as ‘the original synthetic unity of apperception’32. The unity of apperception is ‘that self-consciousness which, by producing the representation, I think (which must be capable of accompanying all other representations, and which is one and the same in all consciousness), cannot itself be accompanied by any further representations’33. Moreover, it refers directly to the act of understanding, for understanding is ‘nothing but the faculty of combining a priori and of bringing the manifold of given representations under the unity of apperception; and the principle of this unity is, in fact, the supreme principle of all human knowledge’34. That is to say, the unity of apperception has an absolute and unconditional representation, yet it is not separate from the act of understanding. This supreme principle only posits the unity in order to regulate our activity of understanding. In the Dialectic, we become aware that this unity is nothing but the transcendental idea par excellence.
A transcendental idea is transcendental on this account: it guides our faculty of knowledge, and in this sense is a ground for knowledge. As a ground, however, it is only remote, for knowledge occurs primarily in the correspondence between sensibility and understanding. In our disposition for knowledge, the correspondence between reason and understanding through self-consciousness only serves the other correspondence. The unconditional totality of reason is projected onto understanding in the cast of apperception, through which I retain the ideals of reason as the regulative grounds for my acts of understanding, as these ideas direct the rational subject only through a self-consciousness: the world is presented to me only as how far I am from knowing it; God is to me how I should serve him; soul is to me why I exist. This regulative use of reason therefore marks the birth of a moral subject.
Though this subject remains in his theoretical disposition, a turn to the practical disposition is already prepared for him. If the disposition for knowledge implies a subjection of the unity of reason to the unity of understanding, the disposition for morality is its reverse. This reversal would be impossible if Kant has not completely revised metaphysics and the meaning of ground. Only when ground comes to move and communicates one structure to the other in a harmonious correspondence can each structure be released from the arbitrary yoke of the ultimate oneness. Ground for Kant is never a ground of blind necessity, but of free possibility. It is only because each faculty is in free harmony with the other, a reversal such as from the theoretical to the practical can occur.
With this complete analysis of Kant’s system, we can finally understand the systemicity of his philosophy as such. In a system, each part is reflective of the whole. That is, each part is but a perspective on the whole. The faculty of sensibility is structured in the following way: each particular intuition is posited among many intuitions, while these many intuitions are encompassed by general projected figures according to rules dictated by the higher unities of concepts, and all these figures are encompassed by the totality of an ultimate singular horizon, the limitation of which constitutes each particular intuition. Thus in his analysis of sensibility, Kant has already introduced the functions of imagination (general figures), understanding (higher unities), and even reason (the ultimate totality), all of them projected onto the same plane of one-among-many. When we move to the perspective of understanding, we find all the intuitions enfolded by imagination into one unified manifold are organized in a hierarchical order directed by the supreme unity of apperception, in which all the ideas of reason stay regulative. Finally, we move to the perspective of reason, and see how everything we have come to know so far is organized according to a plan and unified into an architectonic system by our own doing, which makes us morally responsible for our knowledge.
Again, this systemicity where everything reflects the whole in harmonious correspondences is not possible without the rethinking of ground. It is no coincidence that both Leibniz and Kant, the two most penetrating thinkers on the question of ground against traditional metaphysics, have constructed systems. If Kant has borrowed from Spinoza geometricism as a method, he borrows from Leibniz the moving ground as a principle. Leibniz’s God is a God in transit who coordinates the totality of perspectives from each monad to make a perfect world, while the Kantian subject is a subject in transit who relates all his faculties into a harmonious whole. For Kant, as for Leibniz, there is ultimately no disorder in our world. Between the skeptics and the dogmatists there exists only a disagreement of dispositions. This disagreement is no real contradiction, but a realization of human freedom. When the skeptics and dogmatists each present their case in court, the judge is no longer reason, as it is itself on trial. The judge can be none but the whole of humanity, even as to mitigate the conflict between sensibility and reason, we need the whole disposition of man.
Conclusion
The main argument so far consists of three parts. First, I identified the paradigm of traditional metaphysics to be arithmetical and its main motivation to be providing a foundation for arithmetics. The consequences are, first, that every being is understood to be a ratio between oneness and manyness; secondly, that the ground is conceived to be of maximum oneness, below which everything is organized in a hierarchy; thirdly, logos becomes a predominant force as it relates beings qua quantities.
I then argued that Kant revolutionizes metaphysics by introducing a geometrical paradigm in which ‘one’ and ‘many’ are qualitatively distinct. This helps him redefine knowledge as a correspondence between one and many, instead of a reduction of many to one, thus relocating the ground of knowledge from the original unity to in-between two faculties, so as to avoid the paradoxes of infinite regress and mysterious production constantly incurred by traditional metaphysics.
Finally, I analyzed Kant’s philosophic system into corresponding structures. The system is divided into two triads. The first triad consists of sensibility, understanding and the mediating imagination. The second triad consists of understanding, reason and the mediating apperception. The first triad, i.e. the unity of understanding, subjects the second one, i.e. the unity of reason, since knowledge is primarily a product of the correspondence between intuitions and concepts. The second triad, however, regulates and directs the first in dictating to it the moral significance of our knowledge. Kant thus achieves a twofold grounding of knowledge: first substantially as a correspondence between the unity of concepts and the manifold of intuitions, then morally as motivated by a command from our reason. This answers the two questions Kant raises for himself in the Introduction, first, how is metaphysics as a science possible?, second, how is metaphysics as a natural disposition possible?35 Metaphysics is scientific as it is grounded in the correspondence between sensibility and understanding. It is natural as it obeys the command of reason.
In revolutionizing metaphysics, Kant has thoroughly reconceived the ground. First, he relocates the ground from center of the system to the in-between. Second, if the ground is now in the middle, it becomes likewise transitive and movable. Kant’s philosophy is a map, which registers and delineates the smallest vibrations between different structures as if they are tectonic plates. But if the ground is moving under our feet, an abyss threatens to open up any time. The ground becomes just as obscure as it is fluid: the common root whence our knowledge springs remains unknown to us36.
To the maw of the Sphinx, Kant throws out his solution: ‘the subject’. It is clearer now where the subject springs from. The subject comes to be through the ruins of traditional metaphysics, for as the ground relocates, Thought and Being come out of joint, and we move from the ultimate unity of Being to the immediate ‘here and now’ of Understanding. The ground is not remote, but near to us. It is in us.
But this only brings the crisis closer. Despite, or precisely because of this new grounding, the dark abyss of absolute freedom underlying harmony and order is finally exposed to the eye. Oedipus recognizes man from all his phases of life, and Kant reassembles man from all his faculties. Yet they cannot determine the obscure origin of man, the true ground he stands upon. Because of this, they soon vanish into the dark night of subjectivity, as the underworld, the earth’s unlit foundation, gapes open37.
Notes
- The Theban Plays: Oidipous At Colonus, by Sophocles, translated by Ruby Blondell. 40–45. ↩
- Aristotle’s Metaphysics, by Aristotle, translated by Joe Sachs. Book IV, Chapter 2, 1003b 30. ↩
- Ibid. ↩
- Critique Of Pure Reason, by Immanuel Kant, translated by Marcus Weigelt. Transcendental Doctrine of Elements, Transcendental Logic: Transcendental Dialectic, The Antinomy of Pure Reason, Section VII, B531. ↩
- Phaedo, by Plato, translated by Eva Brann, Peter Kalkavage and Eric Salem. 97A. ↩
- Parmenides, by Plato, translated by Mary Louise Gill and Paul Ryan. 137C–142A. ↩
- Aristotle’s Metaphysics. Book XII. ↩
- De Caelo, by Aristotle, translated by C. D. C. Reeve. Book IV, Chapter 3, 310b 10. ↩
- Aristotle’s Metaphysics. Book XIII, Chapter 6–10, 1080a 10–1087a 20. ↩
- Greek Mathematical Thought And The Origin Of Algebra, by Jacob Klein, translated by Eva Brann. Part I, Chapter 8. ↩
- Physics, by Aristotle, translated by C. D. C. Reeve. Book IV, Chapter 12, 221b 14. ↩
- Aristotle’s Categories And Propositions, by Aristotle, translated by Hippocrates G. Apostle. Chapter 4, 1b 25. ↩
- The Essential Plotinus, by Plotinus, translated by Elmer O’Brien. The Three Primal Hypostases (V, 1 [10]). Contemplation (III, 8 [30]). ↩
- Discourse On Method And Meditations On First Philosophy: Meditations On First Philosophy, by René Descartes, translated by Donald A. Cress. Meditation Two, 31. ↩
- Ethics, by Baruch Spinoza, translated by Samuel Shirley. Part III, Proposition 2, Scholium. ↩
- Critique Of Pure Reason. Transcendental Doctrine of Elements, Transcendental Logic: The Idea of a Transcendental Logic, I, B75. ↩
- Ibid. Transcendental Logic: Transcendental Analytic, 17, B137. ↩
- Ibid. Second Analogy, B243. ↩
- Ibid. 15, B130. ↩
- Critique Of Practical Reason, by Immanuel Kant, translated by Werner S. Pluhar. Preface, 10. ↩
- Critique Of Pure Reason. Transcendental Doctrine of Elements, Transcendental Aesthetic, 2, B39. ↩
- Ibid. 4, B47. ↩
- Ibid. 2, B39. ↩
- Ibid. Transcendental Logic: Transcendental Analytic, 17, B137. ↩
- Ibid. 26, B161, footnote. ↩
- Ibid. Analogies of Experience, B223. ↩
- Ibid. Transcendental Doctrine of the Power of Judgement, Chapter I, B181. ↩
- Ibid. B180. ↩
- Ibid. ↩
- Ibid. Transcendental Aesthetic, 8, B61. ↩
- Ibid. Transcendental Logic: Transcendental Dialectic, Book I, Section II, B379. ↩
- Ibid. Transcendental Analytic, 16, B132. ↩
- Ibid. ↩
- Ibid. B135 ↩
- Ibid. Introduction, VI, B22. ↩
- Ibid. VII, B29. ↩
- Oidipous At Colonus. 1660–1665. ↩
本文由 AI 从英文原文翻译。请以原文为准。
如果说康德的哲学构成一个体系,我们并没有说错。自然,我们想理解什么是体系,体系性又是如何被构想的。但还有一个更深刻的困难:是什么促使康德,又是什么应当促使一个哲学家,去构造体系?
这是关于康德哲学之根基的问题,关于那个支撑并不断激活体系的根基的问题。这一根基可以作两种解释。第一,作为理论基础,作为体系环绕的中心。第二,作为流动的大地,作为这一基础意图压制的问题暗流;然而,暗流的回归不可避免。
一个谜题,英雄为它提供宏大的解答,后来却被自己的答案打个措手不及。正如破解斯芬克斯之谜的俄狄浦斯。
俄狄浦斯知道“人”是什么,也知道“人”是时间性的在者:他爬行,他走路,直到最后被一根拐杖拖着前进。但俄狄浦斯不知道自己的父母。他不知道“人的起源”,并因此最终被暴怒的复仇女神吞噬1。在斯芬克斯的利齿之间,康德对谜题宣告了同样的答案——“主体!”——却也以类似的方式,只是勾出了关于主体性起源的一个更深刻的困难……
但让我暂时放下这些遮掩的隐喻,说得更直接些。康德对形而上学的革命,无非是他对自己在传统形而上学中发现的问题所作的解答。我的任务,就是追踪这一发展。我将首先刻画康德所看到的传统形而上学的问题,再刻画他绕开这一问题的核心策略。在这一初步阐明之后,我将分析他的体系,并据此解释体系性。在结论中,我将重新提出康德之根基的问题。
第一部分
亚里士多德把第一哲学的根本任务规定为对在的追问,却没有忘记补充:在,在某种意义上就是一2。贯穿形而上学的历史,在的问题不断以其算术表达被提出:如果有一个单纯在着的一,多又如何分享这一在的一?如果有许多在着的东西,它们又如何都指向一个在?说得简单些:一如何是多,多又如何是一?
这需要进一步澄清。在与一如何成为同义词,并不必然透明。亚里士多德对这一问题的回答是同义反复。他声称,在与一具有相同的意义,因为说“某物在”与说“某物是一”,并没有区别3。但在这里,他已经把在与一的这种同义性当作自明的,而我并不信服。因此,我将更深入地探究这种同义性的起源。
对“传统形而上学”而言——我宽泛地用这一术语指称前康德的形而上学——在的问题,在很大程度上等同于对在的逻辑肯定的问题。这是因为,它预设了思想秩序与事物秩序之间的等同性。如果亚里士多德的哲学具有目的论性质,那是因为他把逻辑规定转置为本体论规定的方法,预设了思想对于在的透明性;而他直到最后,才在神圣的努斯(nous)概念中为这一点提供根据。我们稍后将看到,康德脱离传统时所挑战的,恰恰就是思想与在的这种等同性。眼下,我只限于引出这一预设对传统形而上学的后果。
因此,作为一的在与作为多的非在之间的差异,就等于肯定与否定作为判断功能之间的差异。这提示我们,通过解释逻辑肯定的活动,发掘在与一之间的原初联系。现在,我面前有一个一般的事物集合,成员包括A、B、C、D、X等,我需要肯定或否定它们的存在。随着我不断作出判断,肯定与否定之间的不对称逐渐显现。对X之存在的肯定,采取“X在”的形式,这是简单且排他的,即其内容不涉及X以外的任何东西。X单纯地在场,就足以获得肯定,与其他一切无关。X在,而且只是X。否定判断采取“X不在”的形式,它开启了一个未规定的可能性领域。因为必定有什么代替X,在那东西的在场中,我才能作出这一判断。例如,圆的方形不在,但它的组成部分——圆与方形——必须存在,使我面前有某物可供判断。X不在,但某个尚未知晓的东西必定占据它的位置。这里,我们看见了根本的不对称。对X的肯定是简单而确定的,否定却是未规定的。仅仅陈述“X不在”,我并没有获得使这一否定成为可能的肯定内容。“X不在”蕴含着“有某个非X”,但“非X”可能包含A、B、C、D等等,而这些全是某种我一无所知的东西的内容。
现在,我转向一个单一对象。我的任务是规定它的在。我先说,这个对象不是椅子。但这只是在无限多的可能性中排除了一种。在这种情况下,对象对我而言仍未被规定。但如果换一种进路呢?我说,一个对象要么是物体,要么是非物体;如果是物体,要么冷,要么不冷;如果不冷,要么热,要么温。通过这样的排除程序,我希望能够发现对象的全部隐藏内容。但在以这种方式推进之前,我必定已经作出一个肯定。我必须先穷尽整个可能性领域,才能只获得有限数量的选项。在对“冷”的否定能够产生确定结果,即“热”或“温”之前,我必须先知道,对象是一个“物体”。但在否定的直接性中,根本不会发生这样的事。如果没有事先把对象置于“物体”这个中介的属之下,没有预先知道所讨论的对象是物体,那么在纯粹直接性中否定“冷”,并不会让我对对象知道得更多。因为如果对象不是物体,单纯否定“冷”就不会为我排除并缩小可能性的范围,因为严格说来,一切非物体都是“不冷的”。康德恰恰在这一点上告诫我们:“如果有人说,每个物体的气味不是香就是臭,那么还可能有第三种情况,即它根本没有气味;在这种情况下,两个相互冲突的命题都是假的。”4因此,为了进行确定的否定,一个在先的肯定是必要的。这同样适用于最高的类,因为它们也被认为数量有限,因而依赖一个在先的肯定。
由此,肯定与否定之间的根本不对称得到澄清。肯定自身总是简单而确定的,否定则晦暗而未规定,因为没有肯定的在先中介,就绝不可能通过否定规定任何东西。依据这一点,我们可以解释一与多的问题中蕴含着什么。多在其原初意义上,恰恰不是算术上的“许多”,而是“未规定性”或“无限性”。这里晦暗之处在于:即便多最初意味着“无限性”,它仍被转译为“多”这一算术概念。而为什么非在的无限性应当比在的有限性更多,却并不清楚,因为二者也许不是可以作数量比较的概念。似乎只有两个有限事物才能按其数量比较,因为无限事物根本没有确定的量。我们能说一个物体比一个精神更多吗?如果认为后一个说法荒谬,为什么不认为前一个也荒谬?有限与无限之间质的差异,如何被当成了量的差异?
如果只在抽象中把一与多看作独立原则,那么,这一从形而上学向算术、从质向量的转译,就仍不可理解。因为形而上学问题真正的推动力,只有在一个具体问题中才能被阐明。也就是说,形而上学的根本问题之所以以算术方式表述,恰恰是因为其最原初的关切指向算术的基础:“一如何是多?”与“多如何是一?”这两个问题,只是“数如何可能?数的本质是什么?”这一问题的分支。再一次,探究的目的引导着探究自身;形而上学为算术奠基的任务,预先规定了它的算术性质:
“凭宙斯起誓!”他说,“我想,我看来远不能认为自己知道这些事物中任何一个的原因;我甚至不容许自己断言,每当有人把一个一加到另一个一上时,被加的一或加上去的一就成为了二,或者加上去的一与被加的一因相加而成为二。我感到奇怪的是:两者彼此分开时,各自是一,而这一对却不是二;可一旦彼此靠近,使它们成为二的原因就成了这个——因被放在彼此近旁而产生的汇合。同样,我也还不能相信,如果有人把一个一劈开,这一劈开反过来又成了它们成为二的原因。因为这样,这个原因就与此前使它们成为二的原因相反了。先前是因为它们被引到彼此近旁,一个加到另一个上;现在却是因为它们被引开,彼此分离。我也不再相信,凭这种进路,我知道一个一为什么生成,或者总而言之,任何其他事物为什么生成、消亡或在。相反,我自己胡乱拼凑了另一条道路,而此前那条,我完全无法容忍。”5
在上面引述的段落中,苏格拉底所问的正是:“多——‘劈开’的原则——与一——‘引近’的原则——如何结合起来,构成一个数?”“劈开”与“引近”这两个原因之所以显得相反,只是因为没有提到:数“二”是这“两个”原因结合的产物。关于原因的这种结合,人们构想了许多或然的故事。一个常见的叙事如下:首先,“一”是简单的肯定。但“一”的规定很快又退回自身的未规定,因为它的简单性没有内容6。一的界限因此被突破,但一仍与它所不是的东西相凝聚,而它所不是的东西仍然是它自身;于是,不在的一与是一的非一,就在某种“二重性”中被维持。这种“二重性”通过一而在一中统一,被一维系;而一又区别于它所包容的二重,因此作为第三者加到二重性之上。然而,在二重性内部,未规定者仍然保留着,继续生成新的成对者……这是关于数作为偶与奇的生成性说明,同一叙事贯穿传统形而上学:“一”在其自同一性中生成多的力量,而一的统一性与这力量结合,形成“数”,从而使算术成为可能。因此,从一开始,我们就看见形而上学肩负着为算术奠基的任务;而这始终是它最持久的动机之一,致力于构造普全数学(mathesis universalis)的莱布尼茨,就是明显的例子。
既然最终目标预先规定程序,传统形而上学遵循算术范式也就不足为奇。我们可以简要回顾两个传统形而上学的样本,以确信确实如此。为此,我选择解释亚里士多德与普罗提诺,因为他们以相对系统的方式代表了传统形而上学的两个主要学派。我的解释必然不完整,不能充分还原被解释者;但我相信,通过这一刻画,传统形而上学的算术性质能够清楚地呈现出来。
亚里士多德形而上学的顶点,是主动的努斯。然而,努斯要成为主动的,就必须在自身中包含一个接受作用的被动面。因此,努斯的纯粹活性在自身内部引入了自己的对立者。在这个意义上,努斯不可能是最高的属,因为它是自我生成的。它作用于自身,并自我差异化,也就是说,它是自身的种差。神圣努斯之下是最高的源头——范畴——它们是努斯活动的最初产物。这些最高源头在同一神圣活动的驱动下,成对地彼此结合,互为属或种差。这些结合带来许多理智的产物。这些新产物又反过来作为彼此的属或种差相结合,生成更复杂的产物。随着这一生成过程不断推进,最终,一些产物由如此之多的在先要素复合而成,以至于我们有限的理智完全看不清那些“结合”,正如我们无法直观千边形或一千颗谷粒7。我们由形式与质料复合而成的可感世界,就是这样生成,并被安置在本体论阶梯的较低层:最外层天球的以太是包容的形式,中心的土是被包容的质料8;形式是更近地指示原初统一性的统一力量,质料则是更远离那一统一性的被动性。然而,形式与质料都已经足够远离原初统一,以至于对我们显得晦暗——因此不再是理智的——因为二者各自都包含过于庞大数量的在先要素,区别只是或多或少。
确实,亚里士多德猛烈批评毕达哥拉斯学派的理念数学说,尤其是柏拉图据说以形而上学思辨为算术奠基的“未成文学说”。但不能忽略,这一批评只针对毕达哥拉斯学派在这种奠基中采取的、未经反思的方法。对亚里士多德而言,通过形而上学为算术奠基不可能是直接的,仿佛每当两个对象在场时,“二性”本身就也在场9。相反,人的数学操作与实在在者的本体论演进之间,存在一道根本的裂隙;因为计数时,我们只是从呈现在面前的对象所包含的全部在先要素的具体规定中,抽象出数学单位10。我们的计数只是把事物还原为同质单位,并不具体地看待事物。因此,亚里士多德所追问的奠基,不可能直接到某个特定理念数为对应的数学数奠基——例如“二性”为二奠基——的程度,而是关于在者演进的一般形而上学说明,如何为一般的算术奠基。因此,亚里士多德对算术的奠基如下展开。
由于一切都被包含在努斯的原初统一中,一切便分享着在的同一个“一”,从而所有个体都可以被还原为同质者,尽管它们在数学上仍彼此有别。换句话说,每一个在者自身都像任何其他在者一样,是“一个在者”。因此,数的可能性仍取决于一与多这两条形而上学原则的结合,使事物作为多而有别,却又被一个共同尺度所限定。但只有在无法辨识在者层级的领域中,我们才能把事物当作同质者。这个领域并不内在于事物秩序,而只相对于我们有限的理智。因为每个事物自身都具有由在先要素组成的独特结构,不可还原为任何其他事物,甚至不可还原为其直接的在先者,正如某一只具体的猫不是普遍的猫。独特的在者只是由于我们的抽象,才被还原为相同者。现在,抽象扎根于可感者,因为当生成推进至无限时,一切对我们都同样变得晦暗;只有到那时,我们才能不加区分地把一切事物看作分享同一“形式”——同一个“一”——的同质单位。从那里,我们甚至把这一由可感者中得来的抽象,反过来应用于可理解者。因此,对亚里士多德而言,数起源于对某个量的抽象度量。正如他自己所说:“一个事物在数中,就是这个事物具有某个数。”11然而,即便算术只能这样间接地通过形而上学获得奠基,因为事物秩序与数的秩序并不完全重合——例如,既然最高的类有十个12,二重的努斯在其生成活动中,按人的计算就会从2跃至12——这一奠基仍然强烈地蕴含着一种从有限向无限推进的层级秩序。
普罗提诺的学说在这一点上并无不同。从本体论阶梯的底部开始,质料相对地是多。随后,它通过形式的统一性被统一,并由此与形式统一,构成自然事物。然而,形式的这种统一力量来自一个更高原则。形式只是普罗提诺所谓“最高理式”的图像;最高理式进而统一一切自然事物,并与它们统一,构成自然。最高理式又是一个更高原则,即灵魂较高部分的映照。灵魂的较高部分统一自然,并与自然统一,构成真正的灵魂。灵魂的较高部分,又只是映照理智的活性;理智以类似的方式统一灵魂,构成理智。但理智的活性与包容理智的太一并无差别13。

再一次,我们看见同一算术范式在运作。太一是原初统一,它分裂为二重性,即理智的活性与被动性。被动的理智作为灵魂,随后分为较高与较低的部分。灵魂的较低部分作为自然,又进一步分裂。这一演进不断推进,直至多的力量最终通过最低层,即质料,充分显现。不过,应当注意:在这一演进中生成的一切在者,而不只是质料,都是无限的,因为它们全都包含多的原则。区别只在于,这一原则的力量被容纳和压制的程度。越接近太一,在者内部的无限性就越统一、越集中;也就是说,无限性就越受到“限定”。
如果我似乎远远偏离了对康德哲学的简单阐释,我希望这一绕行不会被看作与任务无关。康德生于一个传统,只有先理解这一传统意味着什么,才能恰当地领会他的原创性。通过本节可以清楚看见,传统形而上学是在算术范式中运作的。这对形而上学并非没有后果:首先,它预设在与非在的问题等同于一与多的问题;其次,它构造了一种本体论层级,即便不与数的进程精确重合,也大体遵循这一进程,因此,一个在者被放在阶梯上越高的位置,它就越简单、越统一;第三,由于一切在者都卷入算术进程,并被同一个原初统一所包容,在者首先是以“程度”彼此区分的,而大多数区分都可以还原为量的差异,即某个在者在“多大程度上”分有原初统一;最后,由于每个在者都在原初统一中被构想为一种“量”,而量通过比与比例彼此关联,逻各斯便成为我们理解在时占主导地位的力量。理性的力量把多按某个比聚为一,又把许多个比聚为一个比例(analogia),由此把多置于更高统一性的支配之下,依据每个在者距离起源“更近多少”,来评判、比较在者。理性不仅成为统一与支配的力量,也成为最终的裁判。在下一节,我将论证,康德的《纯粹理性批判》恰恰就是对这一算术范式及遵循它的传统形而上学的批判。
第二部分
康德对算术范式的反对,源于他在传统形而上学中发现的一个困难。这个困难伴随着分有理论而来。一方面,这一理论规定一组特殊者分有某个普遍者;另一方面,这些普遍者自身又必须是分有更高普遍者的特殊者。这一系列在潜能上是无限的,而困难在于如何沿这一无限进程行进。这一困难具有两面。
首先,从我们的可感世界出发,一切被感知的在者都作为晦暗者呈现。为了上升至根据,我需要发现事物构成中包含的在先者。为了保证成功,我需要一种方法。这个方法通常被称为“抽象”,由亚里士多德开启,并被笛卡尔继承:在想象中改变某个在者能够具有的样式,从全部偶性中筛出其本质。但这招致了无限回退的问题。由于想象中的变更可以无限地进行,若没有信仰的一跃,就不可能结束这一过程。当笛卡尔沉思那块蜡时,他把蜡的本质从颜色、质地、环境、位置等之中抽象出来。但在从颜色中抽象之前,他必须先把颜色从某种具体的红、蓝、绿……之中抽象出来。在把颜色从某种具体的红中抽象出来之前,他又必须……作为依赖想象变更的程序,抽象是无法终结的,我们不断被困在这一过程中。这促使笛卡尔得出结论:必定有某种超出单纯无限变更的活动,为我们提供理性知识,即理智直观14。
其次,假设通过理智直观,我成功地从其他一切中筛出了最终的实体,这也无益于知识。因为,凡是能从某个在者中被抽象掉的东西,对这个在者都是偶然的。而对通过抽象保留下来的原初统一,即实体而言,一切由它派生之物,也都只对它具有偶然性。于是,我遭遇第二个问题:世界的产生现在成了纯粹偶然的,因而无法解释。
这两个问题,即自下而上的无限回退与自上而下的神秘产生,之所以是问题,在于它们不能满足充足理由原则。无限回退使一切事物最终的根据变得不可抵达;而神秘产生则意味着,低于神的世界之所以被创造,只是出于神的意志,而非必然性。
康德的天才在于,他看到这一困难并不是传统形而上学偶然遭遇的,而是内在于其开端本身。由于传统形而上学遵循算术范式,它盲目预设“一”与“多”只有量的差别,而非质的区分。实际上,康德所做的,只是以几何范式补充算术范式;“一”与“多”并不是不同的数量,而是不同的种类,正如直线图形与曲线图形彼此不同。
康德对形而上学几何进路的这一发现,很可能得益于斯宾诺莎,因为在康德之前,斯宾诺莎就已挑战传统的算术范式;后者由于身体被认为低等而被动,不断将身体隶属于思想。与这一传统相反,斯宾诺莎声称,迄今没有人知道身体能够做什么15,由此将身体从思想的轭下释放出来,把身体性的东西——广延——与理智性的东西——思想——置于平等地位,二者各自都是神的无限多属性之一。
这并不是说康德暗中是个斯宾诺莎主义者。恰恰相反。我们稍后会看到,康德采用几何进路时另有目的。但即便在那之前,我们也应当理解,康德的哲学只在一定程度上是几何性的。其哲学表面上的几何主义,仅仅补充了一种总体上的算术结构。
现在,让我们更深入这一新进路的具体内容,以理解康德如何解决眼前的困难。如前所述,这里的几何主义,意味着康德在“一”与“多”之间辨识到的不是量的差别,而是质的区分。然而,为什么这就构成了解答?如果不更精确地确定困难最初如何产生于传统形而上学之中,就无法恰当地理解康德的解答。
对传统形而上学而言,“一”与“多”只有量的差别。因此,每一个在者作为两条原则的结合,在一个连续谱上或多或少地接近其中一端,即在或在的缺失。一个在者越接近“一”,就越简单、越统一、越完善。由于真理关乎清楚与分明,而这只恰当地属于统一,知识的追求就是尽量减少“多”的表达,尽量增加“一”的表达,从而由特殊者上升至普遍者,由可感者上升至理智者。又由于一切在者只有程度差异,人自以为拥有的关于每个在者的知识,在原则上也没有区别。对最终统一性的知识,与对任何其他在者的知识属于同一种知识,可以与其他任何知识相比较,只是更卓越、更完善。这首先允许关于最终统一性的知识成为所有知识的根据,继而使之成为必然。因为没有最终统一性提供的根据,所有知识都会变得不可规定、空洞无物。不知道链条从何开始,就不知道自己在无限链条上的位置,正如不知道单位,就不知道数。但既然链条是无限的,我们就永远不能一步一步遍历它。反过来,如果从对起源的理智直观开始,那么,既然一切都同样被统一性包容,又很难看出,统一性为什么应当按远近安排在者,如同数从少到多被排列。前者就是无限回退的问题,后者就是神秘产生的问题。
几何进路使康德能够完成一件简单的事:重新定义知识。如果一并不比多“更完善”,而是属于不同的种类,那么,知识就不应是把较不完善者还原为较完善者,而应是两条在质上有别的原则之间的对应。对康德而言,每一个在者都由一个统一性及其对应的杂多构成,二者对这一在者的完整知识作出同等贡献。这就是康德那句著名格言:“没有内容的思想是空的,没有概念的直观是盲的。”16按照康德的描述,对象的概念是统一性原则的显现:“知识在于所给予的表征与某个对象之间的确定关系;而对象,就是在其概念中,一个被给予的直观的杂多得到统一的东西。”17概念提供组织直观杂多的统一性,不是为了还原或抽象掉杂多,而只是为了把统一性与杂多联系起来。传统形而上学据以奠基一切知识的纯粹统一性知识,由此被逐出基础。因为康德现在宣告:如果知识只是统一性与杂多的复合,那么,与复合物属于不同种类的纯粹统一性之“知识”,根本不是知识。传统形而上学寻求的根据,因此是幻相。
通过引入几何主义而完成的知识重定义,解决了康德在传统形而上学中发现的双重困难。首先,既然知识是两条质上有别的原则之间的对应,真正的根据就不在原初统一性,而在这一对应本身。因此,为了给知识奠基,人不必从可感者一路上升至理智者,因为可以说,知识发生在“此时此地”。这就是先验分析论的力量所在,因为它恰恰建立了概念与直观之间的这一对应。例如,为了认识某个因果关系,我们不必遍历因果链条中的一切,追溯至第一因。相反,所需的只是两个表征不可逆的相继:
因此,当我感知某件事发生时,这一表征就包含着某种在先之物,因为显象正是通过与在先者相关,才获得其时间关系,也就是在它尚不存在的一段时间之后存在。只有当我们在先前状态中预设某种东西,发生之事总是跟随其后,即依照规则跟随其后,才能在这一关系中赋予它确定的时间位置。由此首先可知,我不能倒转这一系列,把发生的东西置于它所跟随的东西之前;其次,如果先前状态被设定,另一事件就不可避免、必然地随之发生。18
分析论中除第二类比之外的所有其他要素,也应以同样的思路解释,因为这些范畴全都是康德所谓的形式条件,即主体只应用于单一而直接的表征的规则。一个类比或许有助于阐明这一概念。“每一个”与“所有”虽然都表达普遍性,却具有不同的语法功能,因为“每一个”后面跟着单数名词,“所有”后面则跟着复数名词。依照这一区分,形式条件是应用于“每一个”表征的规则,因为对康德而言,知识总是对普遍规则的直接而单一的应用;而对传统形而上学而言,知识只存在于奠基在原初统一性之中、并通过原初统一性获得奠基的“一切在者的总体”之中。
如果原初统一性不再是根据,就没有必要为了寻找它而遍历无限链条,因而也就不会招致无限回退。这同时具有本体论与认识论的意涵。第一,如已确立的,普遍者并不比特殊者更实在或更完善。第二,一项知识不再需要更高的知识来保证其有效性。这回应了一种经常针对康德提出的批评:为了认识我的认识能力,我难道不是把自己放进了恶性循环?为了知道,我必须知道自己知道,又必须知道自己知道自己知道,如此等等。从这一观点看,《纯粹理性批判》会是一个不可能的计划,因为康德预设自己知道什么是知识。传统形而上学解决这一问题的方式,是把知识分成普遍性或多或少的类别。最高知识无限地反思并自我奠基,例如神圣努斯的绝对透明,或主动理智中在与思想的统一,以此终止反思的无限回退。可以推知,这一解答存在缺陷。康德离开传统,提出自己的解答。知识的知识,或反思知识,并不是更高的知识,而只是对已经完成的综合的阐明或分析:
我们将以“综合”这一一般名称来称呼这个活动,以表明:除非我们先已自行结合,否则就不可能把任何东西表征为在对象中结合着;而在一切表征之中,结合是唯一不能通过对象被给予的表征,因为它是主体自身活动的行动,只能由主体自身实行。很容易看出,这一综合活动在本源上必定是单一的活动,并且对一切结合具有同等效力;而它的解体,即看起来与它相反的分析,总是预设综合。因为,知性若未曾先行结合,就没有什么可供解体;只有已经通过知性结合起来的东西,才能被给予表征能力。19
因此,康德的认识论计划,不是通过更高的反思知识给知识奠基,而是分析潜藏的综合,也就是一种先验分析论。“知识的知识”无非是一件工具,用来澄清自身已经是知识的东西;它是一种“镜中映照”,展示已经以其全部正当性在着的东西,即综合活动。
神秘产生的问题也以类似的方式得到解决。这个问题源于算术秩序表面上的任意性。为什么一切事物都必须被安置在一架本体论阶梯上,有些比另一些更完善?它们不都同样被“一”所包容吗?这种思想上的不协调之所以产生,只是因为传统形而上学的算术范式本身就是任意的。形而上学之所以处理“一”与“多”的对立,只是因为它关心算术的基础。康德并没有走到断言在与非在的问题同一与多的问题毫不相关的地步,那也同样是毫无意义的立场。但他确实看见,算术进程并不是构想一与多的唯一模型,本体论阶梯也不是事物唯一可能的秩序。
既然“一”与“多”具有质的区分,它们就不仅可以按不同比例混合,也可以按不同关系混合。概念是一与多的一种混合,其方式在质上不同于直观混合二者的方式。下一节,我将详细分析这些质上有别的结构是什么,以及它们如何相互关联。眼下,只要看见以下一点就够了:当算术秩序成为一个更大的、诸结构相互对应的体系中的某一种结构时,神秘产生的问题便得到解决。只要它在和谐总体中通过与其他结构的关系获得奠基,其任意性便消失了。
第三部分
体系是一个和谐的整体,其所有部分与总体的联系如此紧密,以至于每一部分与其说是组成要素,不如说是看待总体的某个视角。体系的每一个要素都映照着整体。康德正是这样理解体系的:
当我们关心的是从源泉、内容与界限上规定人类灵魂的某种特殊能力时,按照人类认识的本性,确实只能从灵魂的各部分开始,对它们作精确而完整的展示——在我们关于灵魂诸要素的当前认识状态所允许的限度内。但还有第二种、更富哲学性与建筑术性的注意:正确把握整体的理念,并根据纯粹理性能力中的这一理念,通过从整体概念中推导各部分,注视它们相互之间的关联。这种考察与确证,只有通过对体系最亲密的熟悉才可能完成。那些被第一番研究惹恼、因而不认为值得费力获得这种熟悉的人,便无法达到第二层次,即总览的层次;总览是以综合方式返回此前经分析而被给予的东西。他们若处处发现不一致,也就不足为奇,尽管暗示这些不一致的缺口并不在体系本身,而只在他们自己不连贯的思想进程之中。20
正是在这个意义上,不理解先验逻辑,就不能理解先验感性论;而如果不把握分析论与辩证论之间的相互依赖,先验逻辑又仍然不可理解。当我把康德哲学称为系统性的,我的意思是,它的每一部分都作为整体的一个视角,映照着整体。
我将遵循康德的建议,在观照整体的同时分析其体系的诸要素。如上一节已经确立的,对康德而言,“一”与“多”以质上不同的方式混合,生成不同的结构。这些结构继而成为构成康德体系总体的要素。对每一结构的分析,大致对应本书的划分,即感性论、分析论与辩证论。
一开始,康德就已明确区分直观与概念。在感性论的“形而上学阐明”中,康德谈到空间时说:
空间不是关于一般事物关系的推论性概念,或我们所谓的一般概念,而是纯粹直观。因为,首先,我们只能表征一个空间;当我们说许多空间时,所指的只是同一个独一空间的各部分。其次,这些部分也不能先于那一个包罗一切的空间,作为它得以组装的组成要素;相反,它们只能被思为在空间之内。空间本质上是一;其中的杂多,因而一般空间的一般概念,完全来自限定。因此,就空间而言,对空间的先天直观——一种非经验的直观——必须构成一切空间概念的基础。同样,一切几何原理,例如三角形两边之和大于第三边,绝不是从线与三角形的一般概念中推出,而是从直观中推出,并且是先天地、以确然的确定性推出。21
他对时间也说了同样的话:“不同的时间只是同一个时间的部分;而只能通过一个单一对象被给予的表征,就是直观。”22换句话说,直观中的一与多具有内属的结构。一个直观的许多部分内属于一个单一的直观,而特殊直观作为整体的受限部分,又内属于许多其他部分之间,所有部分彼此同质。这一结构不同于算术结构,后者的自然后果是本体论层级。相反,概念确实遵循算术范式:“每一个概念确实都必须被思为一种表征,包含于无限多个不同的可能表征之中,作为它们的共同特征,因而概念把这些表征包含在自身之下;但没有任何概念,作为概念,能够被思为在自身之内包含无限多个表征。”23通过知性,概念被组织为层级,特殊者隶属于普遍者。
不难看出,直观与概念都是“一”与“多”的混合,因为二者都要求杂多的统一。但二者混合“一”与“多”的结构不同。在直观中,每个特殊者内属于一个更大的直观,特殊与普遍之间没有等级区分;一切都被组织在一个同质平面上。这一结构可以简洁地理解为多中之一(one-among-many),区别于知性的结构,即多上之一(one-above-many)。然而,如果知识要求这两个结构相互对应,那么,这两种事物秩序究竟如何能够共同运作?
现在,我们可以看出康德如何改造形而上学、重新安置根据。根据不再“位于事物的基础”,而是“在诸事物秩序之间”。根据不再是不动的中心,而成为一种传递的关系:这就是哥白尼式革命的真正意义。为了处理这一“奠基”,我们现在转向分析论。
对康德而言,知识“在于所给予的表征与某个对象之间的确定关系;而对象,就是在其概念中,一个被给予的直观的杂多得到统一的东西”24。这已经回答了我们的问题:两种秩序之所以共同运作,是因为它们相互构成,在本源上彼此依赖,以至于缺少另一方,任何一方都无法成为自身。
没有概念中的更高统一性,直观中空间与时间的统一性就不可能。概念在其最本源的规定中,不是一个“东西”,而是把直观的杂多统一为整体的结合活动,是直观这一统一性的源泉。正如康德在B161的脚注中明确指出的:
被表征为对象的空间——如几何学所要求的——所包含的不只是直观的形式,还包括把依感性形式给予的杂多统摄为一个直观表征;因此,直观的形式只给予杂多,而形式直观则给予表征的统一性。在感性论中,我把这一统一性直接归于感性,以表明它先于任何概念,尽管它预设一种不属于感官的综合,空间与时间的一切概念首先通过这一综合才成为可能。因为,空间与时间首先通过这一综合——知性规定感性——作为直观被给予,所以,这一先天直观的统一性属于空间与时间,而不属于知性的概念。25
换句话说,把空间与时间构成为空间与时间的统一性,起源于知性,但这一统一性本身并不与直观的平面分离。统一活动从概念出发,被投射到平面之上,由此生成作为多中之一的统一直观。尽管源泉是知性中的“更高统一性”,直观的统一性却仍停留在感性平面上。概念使之隶属、并组织成层级的原初杂多,通过感性被接收为一个统一体,但这是一个没有任何等级区分的统一体。
显而易见,“权利”的概念永远不可能是可感的,而“物体”的显象自身也不是概念。但在极限情形中,直观的统一性——多中之一——起源于概念的统一性——多上之一——时,保持二者有别的困难就更加突出。这正是图型法与想象力进入的地方。图型法扮演双重角色:它使感性与知性在相互亲和中统一,同时又把二者分开,以免康德哲学成为传统形而上学又一个不幸的例子。
康德清楚地意识到这一危险。谈论图型的中介作用之后,他没有忘记强调:“因此,显象不能直接归摄于范畴之下,而只能归摄于范畴的图型之下。”26图型法必须是一座把两岸分开的桥。
康德在B180中展开图型法的结构。他规定:
图像是生产性想象力的经验能力的产物;感性概念——例如空间图形的概念——的图型,则是纯粹先天想象力的产物,可以说是它的简笔画;图像通过并按照图型才首先成为可能。然而,这些图像总是只能借助它们所标示的图型与概念联系;就其自身而言,它们绝不完全符合概念。27
这里有两点值得注意。第一,图型不是特殊的感性图像。它们是“按照某个一般概念,规定我们的直观”的规则28。然而,图型自身并不是一般的。它只是指向一个一般概念,因为它是一个遵循规则的图形,“按照这一规则,我的想象力能够一般地描画”29。因此,想象力被置于一种奇异的初生地带:一个一般图形正要升至特殊者之上,却还没有成为真正的概念。我认为,理解这一点的最佳方式,是把图型法的结构看作把多褶叠入一。图型是把显象褶叠于其中的形象视域,却并不高于特殊者。感性论已经暗示了这一点,因为空间与时间在那里被规定为既单一又形式性的直观,是包容一切可能直观的直观统一性。区别于多中之一或多上之一,想象力占据一个中间位置,即一内之多(many-in-one)。
即便尚未完成对康德体系的全部分析,也不难看出,康德在总体上保留了算术范式。概念仍是直观所隶属的更高统一性。相较于概念,直观是“多”,因为感性的结构,即多中之一,意味着直观的统一性是群聚与褶叠的统一性,而这些统一性必须起源于一个更纯粹的统一性。然而,即使传统形而上学的范式被挽救了,也并非没有代价。首先,直观之“多”不是量的意义上的。在批评莱布尼茨时,康德指出:“常识所使用的‘权利’概念,包含最精微的思辨能够从中引出的一切;只是在它日常而实践的使用中,我们并不意识到这些思想中所包含的杂多表征。”30因此,概念包含与直观同样多的表征。唯一区别是,概念把杂多包含在自身之下,直观则把杂多包含在自身之内。知性并非内在地比感性更简单或更完善。相反,就清楚与分明而言,二者都应依各自尺度接受评判。因此,即便知性是我们认识配置中的主导能力,也不是由于它更优越,而是因为感性与知性只能以这种方式运作,知识才可能。概念并不对直观施行暴政,而只是在和谐与自由中统治,感性也自愿服从这种和谐与自由。想象力的视域,就是这一自由得到实现的地方。
感性、想象力与知性构成了知识的配置。然而,拼图还缺少最后一块。为了说明概念的统一性,必须寻求一个最终的统一性,因为统一的力量只可能起源于它。
表面看来,传统形而上学似乎在这里已于康德体系中彻底复活,他的全部努力都失败了。因为,这个被重新发现的原初统一性,不正是传统形而上学所寻求的最终根据吗?这个原初统一性一经设定,不就把根据的问题重新收归自身了吗?然而,事实并非如此。概念的统一性确实必定起源于最高统一性,但概念并不必然“奠基”于那一统一性。相反,最终统一性只是被投射到知性之上,而根据则位于中间,正如想象力是感性与知性之间的根据。
类似于感性—想象力—知性这一三元组,在知性与理性之间,也应当找到一个中介原则。正如感性论中已经存在图型法的踪迹,我们也应期待分析论中出现类似情形。
理性的统一性被规定为无条件的总体性31。其结构是为多之一(one-for-many)。关键在于,这一结构首先并不意味着一在量上的最大值,而意味着两条原则的同一化,使无限性在最终统一性中“一下子”被把握。可以回想此前“每一个”与“所有”之间的区分。在其他所有结构中,一与多之间仍有距离,因此,无限性不是一下子被给予,而总是不完整地被给予。在直观中,我在平面上逐一遍历它们;在图型中,我把每一个被给予的直观褶叠入视域;在概念中,我沿阶梯逐级上升。只有在理性的统一性中,一与多的完全同一才得以达到。
现在,什么能够中介知性的统一性与理性的统一性?它必须是一条无条件、却仍对“每一个”单一表征起作用的原则。分析论把这一原则给出为“统觉的原初综合统一性”32。统觉的统一性是“这样一种自我意识:它产生我思这一表征——该表征必须能够伴随其他所有表征,并在一切意识中保持同一个——而它自身却不能被任何进一步的表征所伴随”33。此外,它直接指向知性的活动,因为知性“无非是先天地结合,并把被给予表征的杂多置于统觉统一性之下的能力;而这一统一性的原则,事实上就是一切人类知识的最高原则”34。也就是说,统觉的统一性具有一个绝对而无条件的表征,却不与知性活动分离。这一最高原则只为调节知性活动而设定统一性。在辩证论中,我们将意识到,这一统一性正是最严格意义上的先验理念。
先验理念之所以是先验的,在于它引导我们的认识能力,并在这个意义上成为知识的根据。然而,它只是遥远的根据,因为知识首先发生在感性与知性的对应之中。在我们的认识配置中,理性与知性通过自我意识形成的对应,只服务于另一重对应。理性无条件的总体性,以统觉的样式被投射到知性之上;通过统觉,我把理性的理想保留为我的知性活动的调节性根据,因为这些理念只通过自我意识引导理性主体:世界只作为我距离认识它还有多远而呈现给我;神对我而言是我应当怎样侍奉他;灵魂对我而言是我为何存在。因此,理性的这种调节性使用,标志着道德主体的诞生。
尽管这个主体仍处于理论的配置中,转向实践配置的准备却已经完成。如果认识配置意味着理性的统一性隶属于知性的统一性,那么道德配置就是其倒转。如果康德没有彻底修订形而上学及根据的意义,这一倒转就不可能。只有当根据开始运动,在和谐的对应中使一个结构与另一结构相通,每个结构才能从最终之“一”的任意束缚中解放出来。对康德而言,根据绝不是盲目必然性的根据,而是自由可能性的根据。只有因为各能力与其他能力处于自由的和谐之中,从理论向实践这样的倒转才可能发生。
通过这一对康德体系的完整分析,我们终于可以理解其哲学本身的体系性。在体系中,每一部分都映照整体。也就是说,每一部分不过是整体的一个视角。感性能力的结构如下:每个特殊直观被设定在许多直观之间,而这许多直观又被一般的投射图形所包容,遵循概念的更高统一性所规定的规则;所有这些图形,又被一个最终的单一视域的总体所包容,而对此视域的限定构成每个特殊直观。因此,在对感性的分析中,康德已经引入了想象力——一般图形——知性——更高统一性——乃至理性——最终总体——的功能,它们全都被投射到多中之一的同一平面。当我们移向知性的视角时,就会发现:所有被想象力褶叠为一个统一杂多的直观,都被组织为由统觉最高统一性引导的层级秩序,而理性的一切理念在其中保持调节性。最后,我们移向理性的视角,看见迄今所认识的一切,如何由我们自己的活动按照一个计划组织,并统一成一个建筑术体系;这也使我们在道德上为自己的知识负责。
再一次,这种一切都在和谐对应中映照整体的体系性,若没有对根据的重新思考,就不可能。莱布尼茨与康德这两位在反对传统形而上学时对根据问题思考得最深入的思想家,都构造了体系,这并非巧合。如果康德从斯宾诺莎那里借用了几何主义作为方法,那么,他就从莱布尼茨那里借用了运动的根据作为原则。莱布尼茨的神是一个行进中的神,协调每个单子的视角总体,造就一个完善的世界;康德的主体则是一个行进中的主体,把自己的全部能力联系为一个和谐整体。对康德,也对莱布尼茨而言,我们的世界最终没有无序。怀疑论者与独断论者之间,只存在配置的分歧。这一分歧并不是真实矛盾,而是人类自由的实现。当怀疑论者与独断论者各自在法庭上陈述案情时,裁判不再是理性,因为理性自身正在受审。裁判只能是整个人类,正如要调解感性与理性的冲突,就需要人的整个配置。
结论
迄今的主要论证包含三部分。首先,我指出传统形而上学的范式是算术性的,其主要动机是为算术提供基础。其后果是:第一,每个在者都被理解为一与多之间的一个比;第二,根据被构想为最大程度的一,其下一切都被组织为层级;第三,逻各斯通过把在者作为量联系起来,成为主导力量。
随后,我论证康德通过引入几何范式革新了形而上学;在这一范式中,“一”与“多”具有质的区分。这帮助他把知识重新定义为一与多的对应,而非把多还原为一,从而将知识的根据从原初统一性重新安置到两个能力之间,以避免传统形而上学不断招致的无限回退与神秘产生的悖论。
最后,我把康德的哲学体系分析为相互对应的结构。体系分成两个三元组。第一个三元组由感性、知性以及中介的想象力组成。第二个三元组由知性、理性以及中介的统觉组成。第一个三元组,即知性的统一性,使第二个三元组,即理性的统一性,隶属于自身,因为知识首先是直观与概念对应的产物。然而,第二个三元组又通过向第一个三元组规定知识的道德意义,调节并引导它。康德由此完成对知识的双重奠基:首先,在实质上,将知识奠基为概念统一性与直观杂多之间的对应;其次,在道德上,将知识奠基为由理性命令所激发的活动。这回答了康德在导论中向自己提出的两个问题:第一,形而上学作为科学如何可能?第二,形而上学作为自然禀赋如何可能?35形而上学之所以是科学,在于它奠基于感性与知性的对应;它之所以是自然的,在于它服从理性的命令。
在革新形而上学的过程中,康德彻底重新构想了根据。首先,他把根据从体系的中心重新安置到之间。其次,如果根据如今位于中间,它同样就成为传递的、可运动的。康德的哲学是一张地图,记录并刻画不同结构之间最微小的震动,仿佛这些结构是构造板块。但如果根基在脚下移动,深渊就随时可能裂开。根基越是流动,就越是晦暗:我们的知识从中生发的共同根源,对我们仍然未知36。
康德把自己的解答抛向斯芬克斯的巨口:“主体。”现在,主体从何生发变得更清楚了。主体从传统形而上学的废墟中生成,因为随着根据被重新安置,思想与在脱了节,我们也从在的最终统一性,转向知性直接的“此时此地”。根据并不遥远,而是贴近我们。它就在我们之内。
但这只使危机更加逼近。尽管有了这一新的奠基——或者恰恰因为有了它——和谐与秩序之下绝对自由的黑暗深渊,终于暴露在眼前。俄狄浦斯从人的一切生命阶段认出了人,康德则从人的全部能力重新拼合了人。然而,他们无法确定人的晦暗起源,无法确定人立足其上的真正根基。正因如此,当冥府——大地未被照亮的基础——张开时,他们很快便消失在主体性的黑夜之中37。
注释
- 忒拜三部曲:俄狄浦斯在科洛诺斯, 索福克勒斯著,英译者: Ruby Blondell. 40–45. ↩
- 亚里士多德的形而上学, 亚里士多德著,英译者: Joe Sachs. 卷 IV, 章 2, 1003b 30. ↩
- 同上。 ↩
- 纯粹理性批判, 伊曼努尔·康德著,英译者: Marcus Weigelt. 先验要素论, 先验逻辑: 先验辩证论, 纯粹理性的二律背反, 节 VII, B531. ↩
- 斐多, 柏拉图著,英译者: Eva Brann, Peter Kalkavage and Eric Salem. 97A. ↩
- 巴门尼德, 柏拉图著,英译者: Mary Louise Gill and Paul Ryan. 137C–142A. ↩
- 亚里士多德的形而上学. 卷 XII. ↩
- 论天, 亚里士多德著,英译者: C. D. C. Reeve. 卷 IV, 章 3, 310b 10. ↩
- 亚里士多德的形而上学. 卷 XIII, 章 6–10, 1080a 10–1087a 20. ↩
- 希腊数学思想与代数的起源, 雅各布·克莱因著,英译者: Eva Brann. 部分 I, 章 8. ↩
- 物理学, 亚里士多德著,英译者: C. D. C. Reeve. 卷 IV, 章 12, 221b 14. ↩
- 亚里士多德的范畴篇与命题篇, 亚里士多德著,英译者: Hippocrates G. Apostle. 章 4, 1b 25. ↩
- 普罗提诺选集, 普罗提诺著,英译者: Elmer O’Brien. 三个原初本体 (V, 1 [10]). 观照 (III, 8 [30]). ↩
- 谈谈方法与第一哲学沉思集:第一哲学沉思集, 勒内·笛卡尔著,英译者: Donald A. Cress. 第二沉思, 31. ↩
- 伦理学, 巴鲁赫·斯宾诺莎著,英译者: Samuel Shirley. 部分 III, 命题 2, 附释. ↩
- 纯粹理性批判. 先验要素论, 先验逻辑: The Idea of a 先验逻辑, I, B75. ↩
- 同上。 先验逻辑: 先验分析论, 17, B137. ↩
- 同上。 第二类比, B243. ↩
- 同上。 15, B130. ↩
- 实践理性批判, 伊曼努尔·康德著,英译者: Werner S. Pluhar. 序言, 10. ↩
- 纯粹理性批判. 先验要素论, 先验感性论, 2, B39. ↩
- 同上。 4, B47. ↩
- 同上。 2, B39. ↩
- 同上。 先验逻辑: 先验分析论, 17, B137. ↩
- 同上。 26, B161, 脚注. ↩
- 同上。 经验的类比, B223. ↩
- 同上。 先验判断力学说, 章 I, B181. ↩
- 同上。 B180. ↩
- 同上。 ↩
- 同上。 先验感性论, 8, B61. ↩
- 同上。 先验逻辑: 先验辩证论, 卷 I, 节 II, B379. ↩
- 同上。 先验分析论, 16, B132. ↩
- 同上。 ↩
- 同上。 B135 ↩
- 同上。 导论, VI, B22. ↩
- 同上。 VII, B29. ↩
- 俄狄浦斯在科洛诺斯. 1660–1665. ↩