Notes on <Kant's Transcendental Idealism> - The Second Analogy
The argument of the Second Analogy is the culmination of the Transcendental Analytic. In the eyes of Kant himself, the whole project of establishing a “metaphysic of experience” stands or falls with the success of this argument. However, Kant formulates the Principle differently in First and Second Edition. In the First: “Everything that happens, that is, begins to be, presupposes something upon which it follows according to a rule” (A189). Second: “All alterations take place in conformity with the law of the connection of cause and effect” (B232). In reality, the difference between the two formulations is merely cosmetic. Since the “law of the connection of cause and effect” is equivalent to what is called the Principle of Production in the First Edition. We can safely say that both editions tries to establish the “whatever begins to exist must have a cause of its existence” principle.
This chapter is divided to three parts. The first deals with Kant’s general reflections on the conditions of the representation of an objective temporal order. THE second analyzes the main line of argument for the every-event-some-cause principle. The third considers the precise nature and scope of Kant’s claim, and analyzes one of the most significant objects.
The Transcendental Setting
In First Edition, Kant raises the general problem of explaining how knowledge of an objective temporal order is possible. The paragraph does make it clear that the argument for the causal principle cannot be separated from the transcendental perspective from which the problem is posed. The paragraph can be divided to four parts. The first raises the problem of explaining the possibility of knowledge of an objective temporal order. The second argues that transcendental realism cannot explain the possibility of such knowledge. The third formulates the problem in terms of the language and assumptions of transcendental idealism. The fourth sketches the “critical” or transcendentally idealistic solution.
The Problem of Knowledge of an Objective Temporal Order
“Objective temporal order” simply mans an order of occurrences in the world. The general problem with which all the Analogies are concerned is the possibility of the knowledge of such an order. Thus the general account which Kant prefaces to the argument of the Second Analogy really serves as an introduction to the Analogies as a whole. The specific problem is how the judgments of the form state A precedes state B in object x is possible. Such a succession is called “objective succession”.
The problem is that since apprehension is always successive, an an inspection of the “subjective order” in which representations occur in consciousness does not provide adequate evidence for forming reliable judgments about the quite different order of occurrences in the world.
If we are to understand the problem that concerns Kant in the Analogies we must first reject the assumption that the subjective order is a datum or bit of evidence from which we must somehow make inferences about an objective order. Subjective order corresponds to what is called in the Transcendental Deduction a “subjective unity of consciousness”. As such it can be made into a “subjective object” through introspection, but it is not itself given to the mind as such an object. The key point here is to speak transcendentally about this order is not to consider it as something introspected or actually represented; it is rather to regard it as the indeterminate pre-conceptualized material for sensible representation. What Kant is trying to say here is that if all we had were this indeterminate subjective order, we would not be able to represent any temporal order at all. The problem is to explain how time-consciousness, and with it the consciousness of objective succession, is possible. It is to provide the “formal conditions of empirical truth”. It is the very problem that was left unsolved by the Transcendental Deduction.
The Inadequacy of Transcendental Realism
Let’s first consider Kant’s own characterization of this problem in terms of his transcendental distinction:
If appearances were things in themselves, then since we have to deal solely with our representations, we could never determine from the succession of the representations how their manifold may be connected in the object. How things may be in themselves, apart from the representations through which they affect us, is entirely outside our sphere of knowledge. [A190/B235]
As we can see transcendental realism will not be able to account for the possibility of knowledge of an objective temporal order.
The Idealist Reformulation
While the transcendental realist has the problem of explaining the possibility of access to an objective temporal order of things in themselves, the transcendental idealist has the problem of explaining the possibility of distinguishing between an objective and a subjective temporal order within the realm of appearance.
In the house example we know that we do not identify the order of the representing of the parts with the order of successive states in the represented object. Kant raised the following question:
What am I to understand by the question: how the manifold may be connected in the appearance itself, which yet is nothing in itself? That which lies in the successive apprehension is here viewed as representation, while the appearance which is given to me, notwithstanding that it is nothing but the sum of these representations, is viewed as their object; and my concept, which I derive from the representation of apprehension, has to agree with it. [A191/B236]
The key to understanding the passage lies in recognizing the progressive, dialectical nature of Kant’s procedure in the entire discussion. He begins by raising the question of the conditions of the possibility of making judgments about an objective temporal order, which can also be expressed as a question regarding the grounds or legitimating conditions for such judgments. The transcendentally idealistic analysis of the general problem leads to a new question about the very concept of an objective temporal order of appearances.
Further reflection suggests that these two questions are really equivalent for the transcendental idealist. As we have already seen, it is the very essence of Kant’s “transcendental turn” that the meaning of “object” must be explicated in terms of the conditions of the representation of objects. This is the basic principle underlying the “Copernican” conception of an object as something = x that conforms to our mode of cognition. All Kant is doing here is applying this to a special object, that is “the objective temporal order”.
The “Critical” Solution
In the remainder of the paragraph Kant gives the gist of his solution to the problem. Kant’s answer is that we represent appearance as an object, that is, a temporal order of appearances as objective, by subjecting our representations to a rule. Kant explains this point in [A191/B236] and [A197/B242-43].
It is a rule for conceptualization or judgment, which expresses how he given representation are to be connected in the “objective unity of apperception.” The representation of an objective succession involves a determinate manner of conceiving the manifold of representation that are referred to the “object” in the judgment. For Kant, the necessity, and with it the objective validity of the thought, is produced by the imposition of an a priori rule. The argument of the Second Analogy consists largely in the application of this general principle to the special case of the succession of states of an object.
The Essential Argument
Commentators have found as many as six different “proofs” of the principle in the text of the Second Analogy. Five of them are really variants of the single line of argument. So we can first present the basic structure of Kantian defense of the causal principle without appealing to any of the unique features of the argument.
Before the argument, we need to consider how Kant construes the two key terms, “cause” and “event”. It is frequently noted that the essential feature of the Kantian conception is the element of necessity. Kant claims that “this concept makes strict demand that something, A, should be such that something else, B, follows from it necessarily and in accordance with an absolutely universal rule” (A91/B124). Thus the claim A is the cause of B is to claim that given A, together with certain presupposed “standing conditions” which Kant ignores, B will invariably follow.
The second key term is “event”. It refers to the coming to be of a state of determination of some object. For example, the freezing of water is an event because it involves the coming into existence of a new state of the water. An event thus involves an alteration.
The actual argument begins with an account of the essential features of event perception: “That something happens, some state which did not previously exist, comes to be, cannot be perceived unless it is preceded by an appearance that dos not contain in itself this state” (A191/B236-37). In other words, I cannot be aware that “something has happened” unless I can contrast the present state of some object or substance with its preceding state. This is only a necessary but not a sufficient condition of event perception. The problem is to determine the condition under which a succession of perceptions can be taken as the perception of a succession of states in the object.
The example of the perception of a ship sailing downstream is introduced to clarify the problem. He contrasts this with the perception of the house. Kant concludes from the comparison that in the case of ship, he is constrained to regard the order of my perceptions as determined or irreversible.
However this is commonly misunderstood by commentators that Kant is claiming a necessity of a subjective order of perception. If the irreversibility is understood in this way, then Kant must be taken as claiming that it somehow functions as an “inference ticket,” licensing judgments about the objective temporal order. But if this is the case then Kant is hopelessly confused; for this interpretation commits him to the very empirical idealism which as we have seen, he adamantly rejects. Thus this reading is clearly wrong. We cannot regard the succession ab in the perception of an objective succession AB as either property with these perceptions have in “empirical consciousness” or as a datum from which we can somehow infer that an objective succession has occurred.
How then are we to understand the irreversibility thesis? The answer is that it characterizes the way in which we connect perceptions in thought (the objective unity of apperception) if we are to experience through them an objective succession. Irreversibility does not refer to a given perceptual order, which we can inspect and then infer that it is somehow determined by the object; it refers rather to the conceptual ordering of the understanding through which the understanding determines the thought of an object. Prior to the conceptual determination, there is no thought of an object at all and no experience.
The task is therefore to determine the condition under which we think the order of perceptions as irreversible. Such a condition can only be supplied by an a priori rule. Moreover since the order in question is temporal, the rule must have the status of a transcendental schema. It is obvious that the it is the schema of the pure concept of causality, which is characterized as “the succession of the manifold insofar as it is subject to a rule”. Consequently, it is only by subjecting our perceptions to this rule that we can regard them as containing the representation of an event.
The move from the subjection of the perceptions to the rule to the subjection of the perceived event is clearly the key step in the argument. It might seem that all it establishes is the necessity of the subjecting the perceptions to the rule, but has no bearing on the rule-governedness of the object. Once again, the subjection of perceptions to a rule cannot be construed as the means for making the perceptions themselves into objects, but rather as the basis for conceiving of a distinct, objective temporal order in and through these perceptions. The principle here is fundamental to Kant’s whole philosophy: “the conditions of the possibility of experience in general are likewise conditions of the possibility of the objects of experience” (A158/B197).
I render my subjective synthesis of apprehension objective only by reference to a rule in accordance with which the appearances in their succession, that is, as they happen, are determined by the preceding state. [A195/B240]
In another passage, Kant contrasts the subjective synthesis of the imagination with the objectifying synthesis:
But if this synthesis of apprehension of the manifold of a given appearance, the order is determined in the object, or, to speak more correctly, is an order of successive synthesis that determines an object. [A201/B246]
This, in essence, is the central line of argument of the Second Analogy. The argument falls into following five steps:
(1). All event perception requires successive perceptions of an object.
(2) But this is merely a necessary and not sufficient condition of event-perception. The latter also requires the perception of successive states of the object and this can never be determined on the basis of the successiveness of the perceptions themselves.
(3) In order to consider a succession of perceptions as perceptions of successive states of an object, it is necessary to regard their order as irreversible (in conception).
(4) To regard perceptions in this way is just to subject them to an a priori rule, which in this case must be the schema of causality
(5) As a condition of the possibility of the experience of an objective succession, the schema is also a condition of the succession itself. The schema thus has “objective reality”, which is just what the argument is intended to prove.
It’s clear from the analysis that the argument is grounded in the transcendental analysis of the objectivity.
Finally, dos this argument provides an adequate answer to Hume’s skeptical challenge to the causal principle in the Treatise? Lewis pointed out that the response must be based upon a premise that Hume accepts, and that its conclusion expresses a necessary condition of this premise. I believe the argument meets this criteria, at least in part. The premise which Hume cannot doubt is that we do distinguish between a mere sequence of perceptions and the perception of an objective sequence. He cannot doubt this because the possibility of event awareness is presupposed by his own well-known account of how we come to form the belief that future sequences of events will resemble past sequences. The argument shows that it is only by the application of the schema of causality that we can experience such succession. The premise which Hume would not accept, the transcendental idealism.
Interpretation and Defense
Traditional interpretation of Kant’s real goal is to establish the lawfulness or uniformity of nature, which means the necessary conformity of nature to universal laws of the type found in Newtonian physics. The obvious failure of Kant’s argument to establish anything like this result is then taken as the basis for its complete rejection. The whole line of interpretation culminates in the non sequitur objection of Lovejoy and Strawson. I believe that Kant’s argument is intended merely to establish the modest thesis that every event has some cause, and that only such a reading of Kant’s intent is compatible with the argument which he provides.
The Nature of Kant’s Claim
The first point to note here is that Kant’s argument attempts to prove that the concept of schema of causality is a necessary condition of the experience of the succession of the states in an object, of an event, not that it is a condition for the ordering of distinct events.
In order to determine how much is built into the claim that every event has a cause, it is crucial to keep in mind the meaning of “event”; namely, the change of state of an object. It is necessary that every event must have a perceptual antecedent, given this, I believe that the central exegetical question is, what does the argument of the Second Analogy require us to assume about the connection between an event and its perceptual antecedent? Assuming “cause” means something like “initiating condition”, three possibilities suggest themselves: (1) the initial state A must itself be regarded as the cause of the change to state B; (2) the sequence of states A-B must be “lawlike” in the sense that, given some initiating condition, the transition from A to B is necessarily subsumable under a “covering law”; (3) the sequence can be “contingent” in the sense that it is not necessarily characterizable in either of the two previous manners, but the event is still subject to the principle of causality as a transcendental condition.
The first way of interpreting Kant’s argument leads directly to the famous reductio objection of Schopenhauer. He took Kant to be arguing that the only succession which we can regard as objective is that of cause and effect. Against this, Schopenhauer quite correctly pointed out that appearance can perfectly well follow after one another without following from one another. Kant claims only that without the appeal to the schema of causality we could not distinguish between the representation of successive states of a changing object and the successive representations of coexisting parts or properties of an enduring object. This does not entail that only succession we can experience is that of cause and effect.
Most versions of what can be called the “strong” interpretation of the Second Analogy construe Kant to be arguing for something like the second of the possibilities. It is obvious that many cases of objective succession do fit this model.
The point can be brought out by an example of a succession that is explicitly unlawlike. Let’s consider the case of Jones, who is perceived at t1 to be in a state of blissful inebriation, and immediately after, at t2, is observed to pass out. At first this would no doubt be regarded as a familiar instance of a “lawlike” succession: we simply subsume the successive states of Jones under an empirical law regarding the effects of alcohol. Suppose, however, that we learn subsequently that Jone’s passing out was really due to a slow-working drug ingested hours before the event. In that case the perceptual antecedent, drunkenness at t1 does not stand in any causal or “lawlike” connection with the event. Nevertheless it is obvious that the succession of states is itself objective and “necessary” in precisely the same sense as it would be if it did. The objectivity of the sequence is thus not a function of its lawlikeness.
Buchdahl supports a “weak” interpretation. THat judgments about objective temporal succession do not presuppose that the elements of the succession are connected by empirical laws. There are no additional assumption regarding the repeatability of the sequence and its relevance to other objects of x’s type that are either required or license by this presupposition.
In the example the ingestion of the drug makes possible the explanation of the event, and the search for such conditions is a requirement of reason in its regualtive capacity. This requires the assumption of the uniformity of nature or the affinity of appearances. It is not, however, in any sense part of the task of the Second Analogy to provide a justification for this requirement of reason.
The Non Sequitur Objection
Of all the objections against Second Analogy, the non sequitur objection is perhaps the most fundamental and potentially damaging. Nevertheless I believe the preceding considerations have put us in a position to dispose of it.
According to Lovejoy, Second Analogy contains nothing more than a restatement of Wolff’s proof of the principle of sufficient reasons. It is thus analytic and dogmatic rather than synthetic and critical. He maintains:
all this has no relation to the law of universal and uniform causation. for the manifest reason that a proof of the irreversibility of the sequence of my perception in a single instance of a phenomenon is not equivalent to a proof of the necessary uniformity of the sequence of my perceptions in repeated instances of a given kind of phenomenon. Yet, it is the latter alone that Hume denied and that Kant desires to establish.
Strawson makes much the same point, calling the move “a non sequitur” of numbing grossness.
to conceive this order of perception as necessary is equivalent to conceiving the transition or change from A to B as itself necessary, as falling under a rule of law of causal determination; it is equivalent to conceiving the event of change or transition as preceded by some condition such that an event of that type invariably and necessarily follows upon a condition of that type.
Here lies the non sequitur, which he traces to Kant’s illicit shift from a conceptual to a causal notion of necessity. It is conceptually necessary that in the perception of the sequence of states A-B the observer’s perception should follow the order A-B. Nevertheless he insists:
the necessity invoked in the conclusion of the argument is not a conceptual necessity at all: it is the causal necessity of the change occurring. given some antecedent state of affairs…
There are two points to be made about both objection. First is its complete neglect of the transcendentally idealistic trust of Kant’s argument. The objection presupposes a transcendentally realistic standpoint, and in effect, treats Kant as if he were an empirical idealist. More specifically, it assumes that Kant’s argument rests on an inference from a feature of our perceptions to a conclusion regarding the causal relations of ontological distinct entities or states of affairs that supposedly correspond to these perceptions. This would be a non sequitur of “numbing grossness”. but we have already seen that it is not Kant’s.
One should not, however, jump from this to a “subjectivistic” or “phnomenalistic” reading of the argument. according to which the successive states of affairs are either simply identified with the representations or construed as complexes of actual and possible representations. To do so is to ignore the transcendentally idealistic nature of Kant’s position. The argument of the Second Analogy maintains that the schema of causality is the condition to which our perceptions must be subjected if they are to yield the experience of an objective temporal order.
The second point about the non sequitur objection is that it rests upon a “strong” interpretation of the argument of the Second Analogy. The “weak” Interpretation of the argument requires that we assume only that in the succession of states A - B of some object x, there must be some antecedent condition that determines the x, which was in state A at t1, to enter into or assume state B at t2. Since this neither implies anything about the connection between the new state B and its perceptual antecedent A, nor requires that we assume anything about all objects of x’s type. there is no non sequitur. By the same token, the argument does not establish anything about the uniformity of nature or their conformity to necessary laws.