1 This paper grew out of my work as a Humboldt fellow in the philosophy department at the Georg-August Universität in Göttingen. I am grateful to the Humboldt Foundation and also to the FCAR of Quebec for their generous financial support. I am very grateful indeed to Wolfgang Carl, who most generously found time to read an earlier version of this and other papers. His comments and criticisms were invaluable. I should also like to extend my thanks to the philosophy department in Göttingen and Konrad Cramer in particular for their hospitality on a subsequent sabbatical. Finally, I would like to thank referees of this journal for very constructive and stimulating critical comments that have helped me rethink and sharpen ideas in this paper.

2 All citations to Kant's works — with the exception of the Critique of Pure Reason — refer to Kants gesammelte Schriften (the ‘Academy Edition’) in 29 volumes (Berlin: Walter de Gruyter [and predecessors] 1902). Where applicable, I give the volume, page and line number. Following custom, citations from the Critique of Pure Reason refer — where applicable — to the pagination of the first edition of 1781 (A) and that of the second edition of 1787 (B). I used the text edited by Erdmann, Benno Immanuel Kant's Kritik der reinen Vernunft, fifth edition (Berlin: Georg Reimer 1900)Google Scholar. All translations are my own.

3 Kant wrote up the reflections published as R4675 in volume 17 on a letter from a well-wisher of 20 May 1775. R4675 could not have been written before this date. Adickes says it was Kant's practice to use letters and other scraps of paper shortly after receiving them (18.269f). If that's right, R4675 could have been written as early as the summer of 1775. That the rest of the Duisburg Nachlaβ was written at the same time as R4675 is suggested by the uniformity of ink and handwriting throughout the bundle.

4 To be sure, Kant's correspondence from this time is a valuable source of information about his State of thinking during the ‘silent decade.’ But all too often, the philosophical ideas in these letters are highly abbreviated or too programmatic to tell us much. On the other hand, we also have transcripts of Kant's lectures. These too are valuable sources. However, they are not usually in Kant's own hand, but rather in that of his students. The Duisburg Nachlaβ is unique, because it shows Kant trying to work out at leisure and to his own satisfaction the best formulation for ideas that obviously strike him as having great philosophical significance. In short, the Duisburg Nachlaβ is what is left to us of a kind of philosophical laboratory Kant was running in the mid-1770s.

5 Let me briefly situate my work on the Duisburg Nachlaβ with respect to previous studies. There is little. The most important study is in Carl's, Wolfgang Der schweigende Kant (Göttingen: Vandenhoeck & Ruprecht 1989)Google Scholar. I owe a great deal to this book — starting with the optimism that sense can be made out of the Duisburg Nachlaβ and that it can teach us something important about Kant's philosophical development in the 1770s. My specific debts to Carl in this paper I will acknowledge along the way. What distinguishes my work from his is that I take into account the history of Kant's earlier metaphysical commitments: for the purposes of this paper, that means Kant's early cosmology. Thus I tend to read the Duisburg Nachlaβ in a more backward-looking way; he tends to read it in a more forward looking way. To the extent that I read the Duisburg Nachlaβ as pointing to the first Critique, it is to show that the relevant aspects of the later work should be understood as the result of efforts in the 1770s to give his early cosmology an epistemological turn and to correct underlying difficulties in it. Another full-length treatment of the Duisburg Nachlaβ in recent years is that of Guyer, Paul in Kant and the Claims of Knowledge (Cambridge: Cambridge University Press 1987)CrossRefGoogle Scholar. Like Guyer, I will be insisting in this paper on the significance of temporal determination in the Duisburg Nachlaβ. But Guyer does not take account take account of the significance of Kant's early cosmology, as I do; and, he seems to think that Kant is trying to show how temporal determination is possible through the so-called ‘functions of apperception’ in order to secure the unity of the self. I don't see how this can be possible since Kant seems to be taking the unity of the self for granted. In generai, Kant treats the thinking self in the Duisburg Nachlaβ as a rational substance: in short, he had not yet discovered the Paralogisms. Wolfgang Carl has pointed this out; I accept his conclusions. Cf. Der schweigende Kant, 88-93.

6 For all intents and purposes, the ‘sensible world’ of the Inaugural Dissertation and the ‘objective whole of appearances’ of the Duisburg Nachlaβ are one and the same. I will more usually speak of ‘objective wholes’ in what follows. Elsewhere I discuss Kant's efforts in the Duisburg Nachlaβ to specify the necessary condition for representing appearances as forming an ‘objective whole.’ I also discussed the sense in which this whole is ‘objective’ rather than subjective. See my ‘Kant's Metaphysical Reflections in the Duisburg Nachlaβ.’

7 Kant's Early Metaphysics, 132ff.

8 Cf. a footnote to the passage from the first Critique I just quoted (A418-19/B446-7) and the second-edition version of the Transcendental Deduction (B163-165).

9 A note of caution and qualification is in order here. Strictly speaking, the issue for Kant in the Inaugural Dissertation is under what condition we can represent sensibly given things as joined in one and the same phenomenally conditioned world-whole, whereas the issue he addresses in the Prolegomena is under what condition we can represent sensibly given things as joined in one and the same experience. Even in the Inaugural Dissertation itself, there is some question about whether we may reasonably regard the sensible world as a whole in the strict sense, i.e., as a completed whole. Nevertheless, Kant does sometimes speak of the sensible world in terms directly analogous to those he uses to speak of the intelligible world, and the intelligible world may be regarded as a completed whole. Thus sometimes he speaks of the sensible world as a whole in the general cosmological sense, without any qualification. But by the time of the Prolegomena, Kant denies unequivocally that we can convert experience into any kind of completed whole. That idea underlies his discussion of the antinomies. I do not believe, however, that any of this affects the main point for now. The main point for now is that both the sensible world, as conceived in the Inaugural Dissertation, and experience, as conceived in the later works, are one rather than many, exhibit internal unity, and are such that their parts may be thought by us to stand in universal, reciprocal relations to one another.

10 Hence Kant says that the law that determines the properties of a given geometrical figure does not lie in the figure itself as a determination of space, but rather in the understanding, inasmuch as this faculty is the source of the procedure for constructing the figure. Kant's example is the circle and the construction used in the proof of Proposition 35 in Book Three of the Elements: if in a circle two straight lines cut one another, the rectangle contained by the segments of the one is equal to the rectangle contained by the segments of the other. Kant asks whether ‘this law [se. the proposition from Euclid just stated] lies in the circle or in the understanding?’ He answers the question as follows: ‘If one goes through the proofs of this law, one soon notices that it [se. the law] can be derived only from that condition which the understanding lays at the foundation of the construction of this figure, namely the equality of the radii’ (4.320.34-321.3: the emphasis is mine). Thus the understanding — not the pure intuition of space — is the source of the relevant law inasmuch as it underlies the relevant constructions. The passage from the Prolegomena under consideration is highly suggestive of the important distinction in the B version of the Transcendental Deduction between the form of outer intuition and the formal intuition of space. This has been pointed out by Michael Friedman in Kant and the Exact Sciences, 197ff.

11 Friedman, ‘Space, the Understanding, and the Law of Gravitation: Prolegomena §38’ in Kant and the Exact Sciences.

12 I am indebted to one of this journal's referees for this example.

13 Notice that, in the passage we discussed earlier from the first Critique at A418- 419/B446-447 introducing the distinction between nature and world, Kant says explicitly that we may treat world as nature (i.e., as nature in the formal sense) insofar as it constitutes a dynamic whole with ‘unity in the existence of appearances.’ Hence it is not surprising for Kant to say that the principle of nature in the formal sense is a set of laws — just those laws that universally unite the existence of appearances with one another. But precisely because existence is at issue here, these laws will give us a unity of appearances in time as well as space. Space, as pure form of intuition, cannot all by itself handle either existence or time. So we should not expect it, on Kant's account, to do as principle of nature in the formal sense.

14 Cf., for example, Critique of Pure Reason, B162-3.

15 This leads Friedman to suggest that Kant's line of thinking in Prolegomena §38 is already operating with the distinction between space as form of intuition and the formal intuition of space. The former, to use the language of Prolegomena 38, is ‘uniform and indeterminate’ in that it cannot determine the temporal relations of objects of outer sense. The latter, on Friedman's suggestion, can. See Friedman, Kant and the Exact Sciences, 199Google Scholar.

16 The emphasis is mine.

17 One form such a coordinate System might take is described in the third section of the Inaugural Dissertation (2.401.28-38). I will discuss this passage below.

18 This is a fundamental assumption of the Analogies of Experience in the first Critique.

19 For a much fuller discussion of the principle of coexistence in the Nova dilucidatio, see Chapter Two of my Kant's Early Metaphysics and the Origins of the Critical Philosophy (Atascadero: Ridgeview 1995).

20 I will spell out what this is supposed to mean in due course. This much should be plain for now, however: the proto-principle of nature considered in the formal way (God's choice of law in the governance of the world) and the proto-principle of nature considered in the material way (God's being under the description indicated above) are two different aspects of the divine being.

21 Parallel passages can be found in other works, e.g., the Nova dilucidatio of 1755 (1.395.4-396.7) and the Preisschrifl of 1764 (2.296.32-297.30).

22 Another reason for thinking that Kant's early natural theological reflections might be regarded as the ancestor of his later treatment of the principles of nature considered in the material way is that the specific things Kant says in the Beweisgrund about God under the relevant description seem to correspond neatly to the specific things he says about the pure intuitions of Space and time in the Inaugural Dissertation. Indeed, we can imagine Kant trying initially to work out his account of the pure intuitions of space and time by wondering what result he would get by refitting his natural theological considerations to suit those conditions of human knowledge under which things can appear to us. I discuss the parallels in detall in my Kant's Early Metaphysics, 112-115.

23 The reason for looking at what follows in detall is that it will allow me to call attention to some very striking differences between Kant's treatment of the principles of nature considered in the formal and material ways in the Prolegomena and his treatment of the general cosmological ancestors of the same in the Beweisgrund.

24 On this account, Kant himself has to admit that the world is contingent: ‘All things of nature are contingent in their existence’ (2.106.5).

25 The emphasis is mine. See too 2.95.3-6; 2.133.12-13 where Kant speaks of the great unity in the manifold relations of space; and 2.94.21-23, where he speaks of Galileo's law of chords as being given by the figure of the circle itself.

26 Again, the emphasis is mine.

27 See my ‘Problems and Postulates: Kant on Reason and the Understanding’ in the Journal for the History of Philosophy 36 (1998).

28 In another passage along this line, Kant says this: ‘when I determine something Coming into being specifically in time, i.e., when I determine a reality in the sequence of time, time [sc. the pure intuition thereof ] is, to be sure, the condition in which, but the rule [whereby I effect this determination] is the condition through which’ (17.662.20-3).

29 This interpretation of what Kant is saying here is further suggested by the fact that at the end of these considerations Kant pretty explicitly distinguishes between the order of appearances as merely given in intuition and as established according to rules of perception (17.666.15-18).

30 See too R4675-17.648.9-17.

31 The implications of this are not clear.

32 A few lines earlier in R4682, Kant glosses ‘reality’ as that whereby something is an object of perception (17.668.28).

33 Note, however, that the idea was already announced in the Inaugural Dissertation. Kant writes: ‘But in sensitive things and phenomena, what comes before the logical use of the intellect is called appearance, while the reflective cognition that comes about, manifold appearances being compared through the intellect, is called experience. Thus there is no way from appearance to experience except through reflection according to the logical use of the intellect’ (2.394.2-7). The passage is interesting, because, as in parallel passages in the Duisburg Nachlaβ and even in later works, the transition from appearance to experience goes by way of the understanding or intellect and its use of concepts. One important, obvious difference is that the use of concepts Kant has in mind in the Inaugural Dissertation is merely logical, i.e., the use of concepts for the purpose of classification. Already in the Duisburg Nachlaβ, it seems that such use of concepts would be inadequate, since it is hard to see how any classifying could yield time determination. It is certainly plain in the later works that the logical use of concepts will not give us experience. Nevertheless, the passage from the Inaugural Dissertation is interesting, as one of the first landmarks in the path of a developing idea.

34 It should be noted that Kant has just characterized perception as ‘appearance of which one is conscious’ (R4679 17.664.1).

35 I discuss this relationship at much greater length elsewhere, in ‘Kant on the Self as Model of Experience’ (forthcoming in Kantian Review).

36 Carl, Der schweigende Kant (Göttingen: Vandenhoeck & Ruprecht 1989), 91-2Google Scholar

37 It should be noted that immediately preceding mention of the three functions of apperception, Kant had enumerated the three concepts that ‘concern objects as appearances’ [gehen auf Gegenstände als Erscheinungen] (17.646.23-24): that of substance, cause and effect, and whole. He says, of these concepts, that ‘they give us the threefold dimension of synthesis.’ The proximity of these remarks to the remarks about the three functions of apperception and so forth is suggestive, once again, of a very intimate relation between the three stated concepts of the understanding and the different expressions of the self's self-reflection.

38 See too R4674 -17.646.10-13.1 discuss this idea at greater length in ‘Kant on the Self as Model for Experience’ (forthcoming in Kantian Review).

39 Thus, given the presence of great bodies of water on the earths surface, the law governing the motion of the tides would be a natural consequence of the laws of motion, together with the law of universal gravitation, as Newton argued in Book Three of the Principia. Another example would be the law that governs the slowing of the earth's axial rotation, as Kant himself had argued in a paper of 1754, ‘Investigation of the Question whether the Earth's Dally Axial Rotation Has Undergone any Change.’

40 The paragraph that immediately precedes this passage is also relevant. See R4679 - 17.663.19-26.

41 A passage from R4679 is relevant here. It reads as follows: Every perception must be brought under a title of the understanding, because otherwise it [se. any perception] will give no concept and nothing will be thought thereby. By means of these concepts we avall ourselves of appearances, or rather the concepts specify the way in which we avall ourselves of appearances as of matter for thought…. We say: the stone has weight; the wood falls; (added the body moves); i.e., something acts, hence it is substance. The field has been readied; the meadow has been dried out; the glass has been brocken: these are effects that refer to a cause. The wall is sturdy; the wax is soft; the gold is dense: these are connections in the composite. Without such concepts [se. substance, cause and effect, and composite], appearances would be altogether separated and would not belong to one another’ (17.664.2-14). Kant cannot plausibly be taken to say that the understanding somehow derives thought or knowledge of stones, meadows, gold and their properties or behavior a priori, by reflecting on its own operations. The concepts used as subject or predicate in all of his example judgments are based on what we perceive, i.e., they are empirical. The only thing that the understanding contributes a priori are the concepts without which could not think of these things as substance, as standing in relations of cause and effect or as composites. As Kant says a line or two after the quoted passage, ‘Experience is a specification of the concepts of the understanding through given appearances. Appearances are the matter or the substrate’ (17.664.19-21).

42 It is curious to note that both here in this passage from the Critique and in the passage from R4679 I quoted in the previous passage, Kant uses the same word to characterize more precisely what the understanding or apperception secures for experience in general and empirical laws in particular: he calls it the ‘norm’ for appearances and the particular laws that govern them. The German in both passages is ‘Norm.’ It is — in context — a peculiar word. Its peculiarity heightens the sympathetic harmony of the two passages.

43 Compare R5552, which reads as follows: ‘The understanding lays down the law for nature, but a law reaching no further than that of the form of appearances, which grounds the possibility of experience as such. For nature, as the object of empirical knowledge, must conform to this [se. the form of appearances], because otherwise it would not be nature for us, as it would be impossible to find a connection [einen Zusammenhang] in it that would conform to our capacity to bring the manifold of appearances into a connected conscience [in ein Zusammenhangendes Bewußtsein], hence it would not be knowable’ (18.219.31-220.6).

44 I argued for this point in detall in Kant's Early Metaphysics, in the concluding chapter.