Graham Bird's interpretation of the Analogies of Experience fits them within his general account of the Critique of Pure Reason as providing a descriptive metaphysics.Footnote 1 What appears to be central to this mode of understanding the argument of the Analogies is a claim Bird made in an earlier essay when he stated that: ‘The Critique can be seen … as an extended essay on priority and dependence’ (Bird Reference Bird1999: 42). The dependence relations that are identified in Bird's account apply to what he terms ‘a nonanalytic conceptual relation’ (483). The conceptual relations identified are ones in which a dependence between two sets of notions is established as the basis of the possibility of our being able to utilize one of the sets of notions in question.
The advantages of this account become particularly clear in the detail of Bird's discussion of the Second Analogy. Rather than see the main point of the argument for this principle as resting on the description of causal relations between events, it is instead characterized in terms that enable us to articulate a notion of an event. So, to be able to locate events is to be able to describe a determinate order that has taken place, and this requires the conditions of such a determinate order to be specified. Since order is itself a relation in which two elements have to be placed, in a manner that demonstrates dependence, there has to be something that enables the dependence relation to be specified. Unlike Hume, who views the general causal principle as derived by induction from attention to observation of specific causal inferences being generalized, it follows on Bird's construal of Kant's argument that Kant states that no such inference can be made without prior recognition of the need for a general principle that enables the introduction of the very notion of determinate order.Footnote 2
This element of Bird's treatment is certainly worthy of being endorsed, though reversal of the priority in question is in itself insufficient to establish the apriority that Kant wishes to accord the general causal principle. In understanding how, on Bird's view, Kant does provide the further demonstration of apriority it will be necessary to address four specific issues. First, the manner in which the contrast between a ‘revolutionary’ and a ‘traditionalist’ reading affects the comprehension of the Analogies that Bird offers. Secondly, an articulation of how Bird's account of the Second Analogy in particular relates to the general interpretative options that the current secondary literature presents. Thirdly, a critical response to the way Bird relates the Second Analogy to the Third Antinomy, a relation based on his reading of the contrast between regulative and constitutive principles. Fourthly, a critical response to Bird's overall reading of the principles of the Analogies.
1. ‘Revolutionary’ and ‘Traditionalist’ Views of the Analogies
In arguing for a view of the Critique as providing a descriptive metaphysics, Bird consistently opposes the view that it is a prime aim of the work to argue against general or Cartesian scepticism. Whilst local scepticism is admitted to be often a target of Kant's arguments, it remains the case that even this is often defined in narrower terms than has been traditionally thought. So, the Second Analogy, as Kant's ‘reply’ to Hume, has often been evaluated as providing a response to a range of Humean doubts.Footnote 3
Hume is generally thought to have raised three separate questions: one concerns the rationality of inductive generalization, another concerns the ground of our belief in a general causal rule and a third the basis of attribution of specific causal claims. However, of these three distinct Humean questions, Bird understands the arguments of the Analogies to only address the belief in the general causal rule.Footnote 4 In taking it that the arguments of the Analogies neither provide nor are intended to provide any response to inductive scepticism or the question concerning the alleged necessity of specific causal connections, Bird's reading provides a minimalist view of the way in which it is a reply to Hume.Footnote 5
It is clearly part of what Bird views as ‘traditionalist’ commitments to think that Kant either does or should address the wider questions concerning causation raised in the Humean account, and it follows from Bird's view that Kant is closer to Hume than many would think either plausible or desirable.
2. The Principle of the Second Analogy: Interpretative Issues
Whilst Bird's reading of the Second Analogy offers an account of it that is, in important ways, more limited than has traditionally been suggested, there is a wide range of dispute concerning what the principle of the Second Analogy really asserts and how Kant argues for it. Bird's interpretation of the principle can be related quite clearly to a dominant reading of the Second Analogy, however, and in this respect is far from being ‘revolutionary’ in its implications. Whilst the proof strategy suggested for the principle by Bird is unusual, it is so in ways that highlight further the narrowness of his view of what the principle consists in.
In his general survey of the interpretations given of the Second Analogy, Steven M. Bayne identifies two central views of the principle at issue in it and five different strategies for proving the principle (Reference Bayne2004: 36–7). The two versions of the principle are ‘every event some cause’ (EESC) and ‘same cause same effect’ (SCSE).Footnote 6 The former interpretation is concerned only with particular events and makes no reference to event types, whilst the latter interpretation does involve reference to types and also to repeatability requirements in such a way as to address inductive scepticism in addition to problems with the general causal principle. Henry Allison formulates the distinction between them as that between ‘weak’ and ‘strong’ versions of the principle.Footnote 7
Bird's rendition of the proof of the Second Analogy is a form of the ‘weak’ reading but includes an anomalous moment for such a reading when he refers to ‘conceptual rules which discriminate types among appearances’ (457). This reference to ‘types’ is related to the ability to pick out objects by ‘distinguishing rules’, suggesting that to distinguish objects we require some sense of the causal rule that enables a reference to types of objects with respect to their law-governed character, not just to the notion that there is a general causal principle. However, this is not the direction Bird's analysis of the argument of the Second Analogy takes. Rather than flesh out this reference to ‘types’ Bird speaks of synthetic dependence of event on cause, ruling out in the process the view that, on the basis of the principle of the Second Analogy, we could be guaranteed to ‘establish the actual causes of any event’ (484).Footnote 8 These passages and the general drift of the argument hence commit Bird, despite his passing reference to ‘types’, to a reading of the principle of the Second Analogy as affirming EESC rather than SCSE.Footnote 9
The strategy by which Bird proves the principle of the Second Analogy is by means of an appeal to the notion of the pure category of cause and effect to determine the irreversible order taken to be at issue in the experience of an event.Footnote 10 It includes a claim for understanding this irreversibility in terms of a conceptual dependence that aligns rather strangely with aspects of Paul Guyer's view.Footnote 11 Whilst the description of individuation appears to mirror Guyer's, however, the manner in which Bird understands the contrast between subjective and objective succession is carried out in terms of a threefold distinction between content of perceptions, mode of perceiving and the event perceived. Whilst the mode of perceiving may either be of an event occurring between physical objects or purely mental occurrences, the content of the perceptions has to have the structure of irreversibility in order to convince us that what we are perceiving is actually an event.Footnote 12 For the content of perceptions to be recognized as having this structure is for the categories to be operative in them in a determination that is truth-valued. This description of the argument of the Second Analogy thus makes the reference to the category of causality the basis of the distinction between events and non-events, although we will, necessarily, never have assurance that we are actually experiencing an event. Bird's reconstruction of the argument of the Second Analogy thus emerges as self-consciously abstract and austere.Footnote 13
3. The Second Analogy and the Third Antinomy
There has traditionally been an assumption to the effect that one of the sources of the Third Antinomy is the dependence of the argument for the thesis on the Second Analogy.Footnote 14 By contrast, Bird argues that there is an important difference between the causal thesis of the Second Analogy and that at issue in the Third Antinomy on the grounds that the former does not assert the universal determinism under dispute in the latter. Bird argues for the difference between the Second Analogy and the Third Antinomy by appealing to the distinction between constitutive and regulative principles.
At this point it becomes possible to specify a number of difficulties with Bird's view of the Second Analogy, difficulties that will be elaborated further when we relate the Second Analogy to the First and Third Analogies. In articulating his account of the Second Analogy we sawFootnote 15 that Bird avoided a phenomenalist reading of it by focusing on the content of perceptions, but Kant himself states that the content of alterations is given only empirically whilst ‘the form of every alteration’ is what has to be considered a priori (A207/B252). When considering such form Kant refers not to a question about truth-valued determination but rather to ‘the law of continuity of all alteration’ which specifies: ‘All alteration is … only possible through a continuous action of the causality’ (A208/B254). If Kant's principle in the Second Analogy authorizes such a law then it follows from it that the unity of time, as connected to the unity of appearances, governs the universal and necessary possibility of continuous causal interaction.
If we relate the Second Analogy to the Third we can see further grounds for thinking of Kant's principles as requiring to be understood in a stronger sense than Bird suggests. The principle of the Third Analogy refers to ‘mutual interaction’ (A211) or ‘thoroughgoing reciprocity’ (B256), either of which seems to require that substances are viewed as constantly inter-related in space.Footnote 16 Now whether we regard the interaction of substances as direct or indirect is not so significant as recognition that what would be involved in admitting it as a universal condition – as is specified in the statement of the principle of the Third Analogy in both editions of the Critique – surely involves seeing this interaction as a determinate basis of constancy of appearances.Footnote 17
The relationship between the Second and Third Analogies is one in terms of which the universal laws expressed in each come together in a picture of a dynamical world-whole in which there is continuous action of causality between mutually dependent substances. The implications of connecting them together certainly appear to point in the direction of universal determinism. What is the basic reason then why Bird distinguishes the causality at issue in the Second and Third Analogies from that in the Third Antinomy?
4. Regulative and Constitutive Principles
When describing problems with a reading of the Second Analogy that views its principle as asserting that every phenomenal event is the product of a necessary law, Bird writes: ‘To understand it in that way would be to regard it as a “regulative” principle of reason rather than as a “constitutive” principle of the understanding’ (483). The understanding of the distinction at issue here is only referred to once, and that in a footnote, where Bird writes of it as a difference between ‘the expressive network, or affinity, of a priori concepts and intuitions which make certain objects of experience, such as events possible, and a doctrinal theory which asserts that the categories apply completely throughout the whole of experience’ (830).
In drawing the distinction between constitutive principles of understanding and regulative principles of reason in this way, Bird not only defines the rationale for his distinction between the causation under consideration in the Analogies and that described in the Antinomies, he also commits himself to a view of the regulative principles in particular that is hard to sustain. Whilst it is strictly correct to say that only the constitutive principles of understanding concern objects of experience, it is not only they which make experience itself possible and, furthermore, it is not on the grounds of reference to them that we can establish, as Bird's account suggests, a view of empirical truth. Kant quite emphatically states the opposite, writing that the search for unity by reason operating in accord with its regulative principles is ‘a necessary law’ as without it there would be no reason and ‘without reason no coherent employment of the understanding, and in the absence of this no sufficient criterion of empirical truth’ (A651/B679). So important does Kant think this point that he argues that to secure a criterion of empirical truth we have to presuppose the systematic unity of nature.
On the one hand, then, the regulative principles of reason appear rather more necessary than is granted on the construal of them Bird gives.Footnote 18 On the other hand, the suggestion that the reference of the constitutive principles is not to the whole of experience involves a diminution of their status in such a way that it becomes difficult to conceive how Kant could have described the law of the continuity of all alteration as emergent from the Second Analogy or have written that it was an indispensable law of empirical representation of time that appearances of past time determine all existences in the succeeding time: ‘and that these latter, as events, can take place only in so far as the appearances of past time determine their existence in time, that is, determine them according to a rule’ (A199/B244; my emphasis). Since what is here described is a universal condition of events that shows them in a deterministic order of time, it is difficult to see grounds for the reference to universal determinism in the Antinomies as only arising from regulative use of ideas of reason.
5. Events, Event-Types and the Laws of Nature
The basic problem underlying the account given in Bird of the principles of the Analogies is that there is no sustained reference to the notion of event-types or, put otherwise, to the way in which specific laws of nature relate to the transcendental principles of the Analogies. In one sense, the statement that the transcendental principles do not specifically describe the experience of specific causal connections is hardly news. In another, however, it is problematic to view specific causal connections as if they had no transcendental relation to the principles of the Analogies but were merely products of inductive determination. Amongst other problems, this ensures that Kant's account of causation is one that is a reply to Hume's doubts only in the narrowest sense. It also ensures a serious ‘gap’, however, between the transcendental principles and the provision of empirical laws, a ‘gap’ that threatens to undermine the contention that Kant has provided us with a metaphysics of experience at all.Footnote 19 If the Analogies provide principles for individuated events that do not describe actual experience of distinguishable events, then the ground of reference to continuity that Kant explicates as part of their principle is left unexplained and the basis of the claim to necessity in empirical laws left without any clear grounding in Kant's transcendental philosophy of nature.
The difference hence between us amounts to a distinction at the most general level between a ‘weak’ view that regards the nature that emerges from the Analogies as only generally formal and one that instead regards the possibility of the formulation of laws of nature such as we have in experience as grounded on and made possible by the general principles of the Analogies. In other terms, the transcendental philosophy of nature expressed in the Analogies is, on my view, one that shows not just how determinate truth-values are made possible by providing a reversal of the order of dependencies specified by empiricists but also by providing a basis for showing the ground of empirical natural laws in the provision of the concept of law-governed connections in the type of natural laws. This latter would, on my conception, include a necessary reference for example to substance in the possibility of natural law and with this a universal determination of substances in space such as is given in the Third Analogy.
Bird's reading cannot entirely justify the connection of the three Analogies to the thought of a dynamical world-whole since the principle of continuity that leads naturally to a conception of universal determinism is not included in his account. Without provision of the reference to the principle of continuity it follows quite clearly that the justification of the Analogies is not one that can answer the broadest empiricist problem, namely, the problem of the uniformity of nature. Bird's discussion of the Analogies is one that does not permit the principles which Kant specifically presents as the ground of the possibility of nature to lead to a view of that nature as uniform. By contrast, a stronger reading that views them as centrally connected to a conception of nature that relates these principles to the types of law expressed in natural laws enables the latter to be seen as part of a uniform nature. As Michael Friedman puts this: ‘The transcendental principles of the understanding are absolutely necessary and a priori: they are established entirely independent of all perception and experience. Empirical laws that somehow fall under these transcendental principles are then necessary and a priori in a derivative sense’ (Reference Friedman1992: 174). The way in which the ‘empirical laws’ fall under the transcendental principles is, I suggest, precisely through the conception that the Analogies, whilst formal transcendental principles, have to include more in their scope than is possible on Bird's account. They have, that is, to justify the conception not just of how events are made possible but also how it is possible to differentiate them into types.
