《纯粹理性批判》中的第一经验类比（Analogies of Experience）要求在现象中有一个恒定且唯一之物来表象意识中的时间本身。恒定，或永恒，因为作为总体的时间本身是始终在那的；惟一，因为只有一个时间。这个惟一且永恒的东西，即现象中的实体（Substance）。康德明确把这个实体与洛克的托子（Substratum）区分开来，认为实体是以各种方式显现于人的，而不是像托子那样不可知，也不显现。
附录：On Understanding Substance as Mass
In the First Analogy of Experience, Kant argues that there must be some permanently persistent substance in the appearances which represents the persistence of time. Given Kant’s endorsement of Newtonian physics, commentators such as Eric Watkins suggest that such permanently persistent substance can be understood as Newtonian mass. In this paper, however, I argue that we face a dilemma when we try to cash out the notion of substance in terms of Newtonian mass.
The paper proceeds in three steps. In the first section, I present the reason why there needs to be a permanently persistent substance in the appearances, and discuss why it seems to be compelling to conceive of the permanently persistent substance as Newtonian Mass. Then, in the second section, I argue that there are (only) two ways of conceiving of the permanently persistent substance as Newtonian mass, namely, to conceive of substance as individuated mass and to conceive of substance as the sum total of mass in the world of appearances. I show that there are textual indications as well as philosophical reasons to support each option. In the third section, however, I argue that both ways suffer from inescapable problems. Thus, conceiving of the permanently persistent substance in terms of Newtonian mass is not viable.
Section I. The Permanently Persistent Substance
In this section, I shall first present the reason why Kant thinks that there must be a permanently persistent substance in the appearances. I then discuss why it is compelling to conceive of such substance as Newtonian mass.
In the chapter “System of all principles of pure understanding,” Kant discusses what makes possible the applications of the categories, i.e. the pure concepts of understanding, to objects, i.e. appearances that are given to sensible intuitions. That is, he discusses what it is that makes the categories have objective validity. Kant’s claim is that the applications of the categories are only possible under certain conditions, and these conditions are spelled out by the principles. For instance, the applications of the relational categories (substance-accidents, cause and effect, and mutual interactions) are possible if they are applied to objects according to the principles of Analogies of Experience. In addition to the three specific principles that correspond to each of the three relational categories, Kant also provides a general principle overarching all three Analogies. The general principle is stated in the second edition as follows: “Experience is possible only through the representations of a necessary connection of perceptions” (B 218). Watkins provides a helpful interpretation of this general principle:
“The general idea is that each of the three relational categories represents a necessary connection that is required for experience of a single time and of objects existing and being temporally related to each other within a single time to be possible.” (My emphasis)
Since this paper is focused on the notion of substance in the first Analogy, I shall ignore the second and third Analogies. So I now turn to a close examination of the first Analogy.
The first Analogy, i.e. the principle of the persistence of substance, is stated in the second edition as follows: “In all change of appearances substance persists, and its quantum is neither increased nor diminished in nature.” (B 224) Watkins summarizes Kant’s argument for the first Analogy as follows (which I take to be a correct interpretation):
Premise 1: Appearances, i.e. objects of experience, are made possible by time’s persistence.
Premise 2: We do not perceive time itself.
Therefore, In order to have experience of appearances, there must be some permanent substance in the appearances which can represent time or time’s persistence.
While the appearances, as the objects given to our intuitions, are changing, the substance in appearances always stays the same and is permanent. So, Kant calls the permanent substance “the substratum of everything real” (B 225). But, some clarifications about Kant’s use of the term “substratum” are needed to prevent potential confusions. Substratum in Kant’s text does not mean what Locke uses this term to mean, namely, the bearer of properties which is unchanging and about which we can have no knowledge. For, according to Locke, we can only know what is given to our senses, but since the underlying substratum cannot be given to our senses, we have no access to it and therefore cannot know it.
Kant, by constrast, does not think that there is any Lockean substratum in the world of appearances. For Kant, the fact that the states of the substance are changing and the substance stays the same does not mean the states are separable from the substance. Rather, the changing states of the substance are simply the ways in which the substance is given to us. Thus, we can know the substance, that is, we know the substance through its states. In order to avoid the Lockean implication of the term “substratum,” I shall only use “substance” to refer to the permanently persistent thing in the appearances despite Kant’s own use of “substratum” to talk about what is permanent in the appearances.
Since I have argued that Kant’s notion of substance is not the Lockean substratum, then what is the Kantian notion of substance? We need a positive account of what the substance is. It is obvious that such a permanently persistent thing cannot be captured by ordinary physical objects, no matter whether they are natural objects (say, rocks) or artifacts (say, ships), for neither artifacts nor natural objects always stay the same such that in principle they can never suffer changes. So, it seems no ordinarily construed physical things can be qualified as substance that is permanently persistent. On the other hand, it is very hard to imagine that anything non-physical could play the role the substance is supposed to play. For it is hard to imagine how a non-physical being could be given to our sensible intuition or could be spatiotemporally organized by our a priori intuitions. So, it is unlikely that Kant means something non-physical by “substance.” Thus, there are two constraints on spelling out what substance is. First, it is something physical. Second, as I have shown, the physical being that can be understood as substance cannot be ordinarily individuated physical things such as planet or rock.
In order to meet the above two conditions, Watkins suggests that, given Kant’s commitment to Newtonian science, it is likely that Kant has Newtonian mass in mind when he talks about the substance, since no matter how a physical object changes, its mass always stays the same. Since Newtonian mass is physical and is not an ordinarily individuated object, it seems quite compelling that the substance, which is permanently persistent, just is Newtonian mass. According to common sense, Newtonian mass is understood to be underlying objects such that we cannot directly perceive mass but can only perceive mass through the way it is given to our intuition, namely, through the perception of the objects that have mass. Thus, mass is neither unknowable nor directly perceivable, which seems to fit the description of the substance perfectly.
Moreover, there are many textual indications that suggest the identification of substance with mass. Let me note two examples. First, recall the general principle overarching the three specific Analogies, namely, “In all change of appearances substance persists, and its quantum is neither increased nor diminished in nature.” (B 224) It seems that “quantum” is most naturally to be understood as mass, for mass seems to be the only thing in nature that is neither increased nor diminished on Newtonian physics.
The other indication is Kant’s example to illustrate his claim that “he <a philosopher> thus assumed that as incontrovertible that even in fire the matter (substance) never disappears but rather only suffers an alteration in its form.” (B 288, my emphasis):
“A philosopher was asked: How much does the smoke weigh? He replied: If you take away from the weight of the wood that was burnt the weight of the ashes that are left over, you will have the weight of the smoke.” (B 288)
We can see that here Kant explicitly identifies substance with matter. And it is quite plausible to think that “matter” is just another way of saying “mass”. That is, “mass” seems to be the theoretical analog of the term “matter.” This hypothesis is supported by the example of the weight of smoke. For, in the example, the way to calculate the weight of smoke just is to calculate the mass (multiplies the gravitational constant).
However, despite the compelling reasons for the identification of substance with mass, in the next sections, I shall argue that the substance cannot be understood as Newtonian mass, for when we try to work out the details of understanding the substance as mass, we face an unavoidable dilemma.
Section II. Some Mass or the Sum Total of Mass
In this section, I shall argue that there are two ways of conceiving of substance as Newtonian mass, and then show that both ways have some support from the text and are to some extent philosophically plausible. So, both ways deserve detailed considerations. But, in the next section, I shall argue that both ways face insurmountable problems.
In identifying substance with mass, we need to settle an ambiguity: Is the mass meant to be some mass, say the mass of a rock which is 7 kilograms (a randomly chosen weight), or to be the sum total of mass in the world of appearances which is a very large but nonetheless definite amount? Since both some mass and the sum total of mass are permanently persistent, we cannot tell which way of identifying is more plausible with respect to the permanent persistence of substance. So, we must appeal to some other philosophically and/or textually interesting points to ground a preference in choosing one over the other.
Let us first consider identifying the substance with some or individuated mass. First, the first Analogy is the principle according to which the relational category substance-accident is to be applied. Kant defines accidents to be “the determinations of a substance that are nothing other than particular ways for it to exist.”(B 229) Many commentators interpret the relation to be between object and its properties or states. Thus it makes more sense to think that the mass, which is the underlying bearer of properties, is the individuated mass of some object, instead of the sum total of mass in the world of appearances. For instance, in the example of the weight of smoke, Kant seems to conceive of substance as the matter, i.e. mass, of an individual object. Moreover, if we conceive of substance as the sum total of mass in the world of appearances, it is very hard to imagine how substance can be the bearer of properties or what kind of properties of which substance is the bearer.
One might argue that, on the interpretation according to which substance is the sum total of mass, even though we could imagine no properties of which substance is the bearer, we can still conceive of substance as the bearer of (changing) states, i.e. the successive states of the world of appearances. I reply that Kant cannot accept such an idea because the states of the world are not objects of possible experience, for it is at least empirically true that no one could have the whole world of appearances as his object of experience. I will return to this point later on in the paper and use it to argue that conceiving of substance as the sum total of mass is untenable given Kant’s theoretic commitments.
The above discussion is about reasons to prefer the identification of substance with some mass. I now turn to the reasons to prefer the identifications of substance with the sum total of mass. There are some textual evidences in the first Analogy that suggest this latter identification. For instance, the following passage:
“…here the issue is only appearances in the field of experience, the unity of which would never be possible if we were to allow new things (as far as their substance is concerned) to arise. For then everything would disappear that alone can represent the unity of time, namely the identity of the substratum in which alone all change has its thoroughgoing unity. This persistence is therefore nothing more than the way in which we represent the existence of things (in appearances).” (B 229/A186, my emphasis)
In this passage, Kant seems to identify the permanent persistent substance that represents the persistence of time with the unity of appearances, which seems to be the sum total of mass in the whole world of appearances. Let me argue for my understanding of this passage that it indicates that Kant identifies substance with the sum total of mass. I shall argue by reductio: Suppose Kant identified substance with individuated mass in the above passage. Then, it would make no sense to think that the arising of new substance could make the representation of the unity of time impossible. For the arising of new substance in no sense affects the substance, i.e. the mass, of the original objects. Let me use an example to illustrate. Suppose there is a rock whose mass is 7 kilograms and there arises a new object out of nothing, whose mass is 5 kilograms. Insofar as the rock’s mass remains the same, whether or not there are new masses arising out of nothing does not affect the unity of the rock’s mass, which is 7 kilograms. Therefore, in this passage, Kant conceives of substance as the sum total of mass in the whole world of appearances.
So far I have shown that there are compelling reasons to identify substance with some mass or with the sum total of mass respectively. In the next section, I shall argue that there are also devastating reasons to each identification such that either way we go, we face unsolvable problems.
Section III. One Single Time and the Limit of Possible Experience
I now turn to the problems from which the each identification suffers. In this section, I shall argue that these problems make both identifications untenable. Let us first consider the identification of substance with individuated mass (i.e. some mass). I argue that the reason why individuated mass cannot be identified with substance is that individuated mass cannot represent the oneness of time. Recall Kant’s argument for the principle of the first Analogy: in order to have experiences of objects as temporal, we must identify a permanently persistent substance that can represent time in objects. While the states of the substance change, the substance persists so that the substance can represent time that persists. It is important to notice that time, which is supposed to be represented by substance in appearances, is one single time. But, individuated mass cannot represent one single time. For there are many individuated masses, for instance, the mass of a rock which is 7 kilograms, the mass of a cup which is 0.5 kilogram, and the mass of a table which is 3 kilograms, each of which is permanently persistent and undergoes changes. If one of them can represent time, any other also can. In that case, we do not have one single time. Rather, we have many times or time-series, each of which is persistent.
Let me explain in details why multiply individuated masses cannot represent on single time. If these individuated masses can represent one single time, there must be some one single thing that is shared by these individuated masses that serves to represent the singularity of time. Whatever this shared thing is, it is not any of these individuated masses. Therefore, individuated mass cannot present one single time. However, on the other hand, time has be to singular. Here is a passage in the first Analogy which explains why time has to be one single time rather than a plurality of times:
“Substances (in appearances) are the substrata of all time-determinations. The arising of some of them and the perishing of others would itself remove the sole condition of the empirical unity of time, and the appearances would then be related to two different times, in which existence flowed side by side, which is absurd. For there is only one time, in which all different times must not be placed simultaneously but only one after another.” (B 232/A189)
One might argue that it does not matter how many individuated masses can represent time, it only matters that there is an individuated mass that represents time. Insofar as there is such a substance, which is permanently persistent, it suffices to represent one single time. I reply that, in that case, we still do not know which individuated mass is suppose to be the representer of the one single time in appearances. For there is not reason to think that the mass of one object is more suitable to represent time than the mass of another object is, insofar as both of the individuated masses are permanently persistent. Any choice of one over the other is arbitrary. But the unity or singularity of time is not arbitrary, for there can only be one time-series which persists, and any other time-series or temporal relations are just temporal parts of this unique time-series. Thus, I conclude that individuated mass cannot be the representer of time in appearances.
I now turn to argue that the sum total of mass cannot represent time either. The idea of my argument is to make use of Kant’s solution to the Antinomies to show that the permanently persistent substance that represents time in the appearances cannot be the sum total of mass because the sum total of mass is not an object of possible experience. Let me lay out my argument in detail.
In “The Antinomy of Pure Reason” chapter, Kant presents four pairs of arguments concerning four cosmological ideas about the world-whole, namely, whether there is a beginning of time, whether there is indivisibly simple substance, whether there is a first cause, and whether there is a necessary existent. As Allen W. Wood argues, the four antinomies share a general form, namely, the thesis of each antinomy claims that there must be a first member of the conditioning-conditioned chain, while the antithesis of each antinomy claims that there is no first member of such a chain and that the chain goes back into infinity. Kant argues that there are valid arguments for each of the four theses as well as valid arguments for each of the four antitheses, so we need a solution to such contradictions.
Kant’s solution to the contradictions, as Wood argues, relies on his doctrine of transcendental idealism. As for the first two antinomies, Wood argues
The mathematical antinomies are generated by mathematical principles that apply to things only insofar as they are given in sensible intuition…But these [the first two] series of conditions are never given to intuition as a whole...The theses are false because the principles of possible experience make it impossible for objects corresponding to the cosmological ideas of a first event, a largest extent of the world or a simple substance, ever to be given to intuition.”
Thus, the reason why Kant thinks that the claims made by the theses of the first and second antinomies are false is that neither the beginning of time nor the spatial boundary of the world or an indivisible substance can ever be given to our sensible intuition. If something cannot be given to our sensible intuition, according to Kant, we cannot have experience of it. Let me call this principle the object-of-sensible-intuition principle, namely, if something cannot be given to our sensible intuitions, then it cannot be object of our possible experience. And we can apply this principle to an object to determine whether that object can be object of possible experience. That is, if the object in question can be given to our sensible intuition, then the object can be object of our possible experience, but if the object cannot be given to our sensible intuition, then it cannot be object of our possible experience.
Now, let me apply the object-of-sensible-intuition principle to the idea of the sum total of mass. We can see that the sum total of mass cannot be given to our sensible intuition, so, the sum total of mass cannot be object of our possible experience. For the world of appearances seems to mean the whole universe or cosmos (because everything in the universe stands in causal relations to each other), there is no way for all of the mass in the whole universe to be given to our sensible intuition. Actually, we do not even know whether there are spatial boundaries of the universe, so we do not even know whether the sum total of mass in the all universe is finite. Thus, the sum total of mass cannot be object of possible experience. So, the sum total of mass cannot be that which represents time in appearances. For the reason there must be a permanently persistent substance in appearances which represents time is to make our temporally connected representations of objects possible. But, if the sum total of mass cannot be object of experience, it cannot make our experience of object possible. Thus, the permanently persistent substance in appearances cannot be the sum total of mass.
One might object that in the antinomies, the cosmological ideas at issue are condition-condition series. (B 436/A410) But the sum total of mass is not a series. Rather, it is an aggregate about which the question of conditioning and conditioned does not arise at all. Thus, Kant’s remarks on the antinomies have no bearing on whether the idea of the sum total of mass has any objective validity or significance. Moreover, the first two antinomies concern whether the conditioning-conditioned series go on into infinities. And it seems that it is impossible for us to experience infinity, for no matter what we experience it is finite insofar as we have experienced it. But, the quantum of the sum total of mass seems to be a definite and finite amount. By virtues of what can we claim that the sum total of mass cannot be object of experience? Is this “cannot” an empirical cannot, or an In-Principle cannot? If the answer is the former, the empirical “cannot” does not seem to be strong enough to show that the sum total of mass cannot be experienced, because we cannot know or predict whether in the future empirical sciences and technologies will make the sum total of mass possible object of experience. If the answer is the latter, at least further explanations of why the sum total of mass, which is a finite and definite amount, cannot be object of possible experience in principle are needed.
To the first objection I have two replies. First, in the first antinomy, Kant also discusses whether there is boundary or the largest extent of space. It is not obvious that there is a spatial series in the sense that it is obvious that there is a temporal series in which one moment succeeds its previous moments. However, according to Kant, we can think of the space acquiring its quantum through repeatedly or successively adding spatial units to the previous spatial units. (A 428/B 456) That is, the way of conceiving of space as a spatial series depends on the way of conceiving of time as a temporal series, which is naturally serial. Then, by the same token, we can also think of the sum total of mass acquiring its quantum by successively adding massive units to previous massive units. Thus, if the object-of-sensible-intuition principle applies to the idea of the boundary of space, it should also apply to the idea of the sum total of mass of the whole world of appearances.
Second, the fact that Kant applies the object-of-sensible-intuition-principle to the first two (or three) cosmological ideas to solve the contradictions does not mean that the principle can only be employed to deal with the antinomies. If the principle is applicable to other ideas, we can also use the principle to deal with other ideas. Since the object-of-sensible-intuition principle is derived from transcendental idealism, which is an important element in the whole Critique, there is no reason why the principle cannot be applied to other ideas than cosmological ideas. Thus, it is legitimate to use the object-of-sensible-intuition principle to show that the sum total of mass of whole world of appearances cannot be object of possible experience. So, the sum total of mass cannot be what represents time in appearances.
My reply to the second objection has two steps. First, it needs to be clarified that, although the first two antinomies concern whether the conditioning-conditioned series are infinite, Kant’s solution by the object-of-sensible-intuition principle does not rely on the whether the series are infinite. The principle only concerns whether the things to which the cosmological ideas refer can be given to our sensible intuition. It does not concern whether the things are infinite. It seems true that infinity cannot be object of sensible intuition. But this does not mean that all finite things can be given to our sensible intuition. Actually Kant rejects the claim that all finite things can be given to our sensible intuition. For Kant thinks the thesis of the first antinomy is false, because the beginning of time or the boundary of space cannot be given to our sensible intuition so that it cannot be object of possible experience.
The second step of my reply is to spell out in which sense of “cannot,” the sum total of mass cannot be object of possible experience. It seems to me that the distinction between empirical “cannot” and In-Principle “cannot” is hard to cash out in the context of Critique. For, in the Critique, any legitimate claim to knowledge entails that the object of which the knowledge is can be experienced. Thus, it seems that the empiricality of the “cannot” entails the In-Principality of the “cannot”.
However, concerning the claim that we cannot predict whether in the future empirical sciences and technologies will make the sum total of mass possible object of experience, what would Kant say? Would Kant agree that future sciences and technologies might or could transform a transcendent idea into an idea which refers to object of possible experience? I do not think he would. For Kant thinks his Critique settles metaphysical questions once and for all by theoretical reason, which is static or a-historical. Future discoveries made by sciences and technologies should be able to do no damage to the doctrines in Critique. Moreover, it should be odd to Kant’s ear that progresses made by empirical sciences could have any bearings on the doctrines in the Critique, which he builds up from scratch employing only pure reason, which is absolutely a-historical.
Thus, I conclude that the above arguments show that identifying substance with the sum total of mass in the world of appearance is not tenable. Since I showed earlier in this section that identifying substance with individuated mass is not tenable either, I conclude that the general strategy of identifying substance with mass is untenable.
Section IV. Concluding Remarks
In this paper, I showed that a seemingly very promising way of understanding the permanently persistent substance discussed in the first Analogy, namely, conceiving of substance as Newtonian mass, is untenable. Then, I wonder whether there are other promising ways of providing a positive account of substance or actually it is the case that the notion of substance in the first Analogy is itself untenable. At this stage, maybe I could follow Kant’s stance on the things of themselves, namely, they exist, but we can have no knowledge about the way of their existence. But, at the same time, we need to have this minimal conviction that they exist. Similarly, concerning substance, we can have no knowledge about what the permanently persistent substance is, but we need to have the minimal conviction that it exists in the world of appearances and it serves to represent time.