Förster (1993, p. xxxif) points out in his introduction to the English version of Opus postumum that even as Kant was working on the final stages of his "Critique of Pure Reason", which was published in 1781, he had in fact begun to think about another way of presenting the results of the "completely new science" that had grown out of his investigations into what he had called the "idea of a transcendental philosophy"[A1, A13]. The necessity of developing this "other way" of discussing his work apparently became even clearer to Kant when he realised that the special sense he had developed for the term "transcendental" had not been widely understood. The first (anonymous) reviewer of the "Critique" in Göttingischen Gelehrten Anzeigen had in fact referred to Kant's work as representing "a higher idealism" and allied it with Berkeley's idealism about things, something which had distressed Kant considerably. Immediately after the "Critique" had been published, he therefore began working seriously on a brief account of what he had written on in the "Critique" based on a different kind of plan. It was, in his own words "a plan according to which even popularity might be gained for this study"[15]. The result was the "Prolegomena to Any Future Metaphysics" (1783), where Kant wrote in his preface that "... a mere plan preceding the Critique of Pure Reason would be unintelligible, unreliable and useless; it is all the more useful [however], as a sequel... With that work complete, I offer here a plan based on the analytic method, while the Critique itself had to be executed in the synthetic style."[16]
In his "Prolegomena" Kant sets out to answer the question of "whether such a thing as metaphysics is at all possible", and here, interestingly enough, he turns things on their head and starts out from the synthetic a priori propositions of mathematics and the natural sciences, which he alleges are uncontested. This being the case, it should then be possible to deduce from the principle that makes them possible the possibility of all other synthetic a priori propositions. In the "Prolegomena" it is thus the rational sciences of the objects of experience (i.e. mathematics and physics) which are seen as being able to provide the criterion that any science of non-empirical objects (i.e. a metaphysics) must meet. His idea was to make the "plan" of the "Prolegomena" the defining paradigm of transcendental knowledge by playing down the misunderstood idea of a priori reference to Gegenstände überhaupt as that which transcendental knowledge is concerned, and concentrating instead on the reference to possible experience. He writes that "The word `transcendental' ... does not signify something passing beyond all experience, but something that indeed precedes it a priori, but that is intended simply to make knowledge of experience possible." [AK 4:293], which Förster interprets as meaning that transcendental philosophy now has become for Kant exclusively a theory that discerns the a priori conditions of possible experience.
This shift in perspective led Kant to go on and write another paper entitled "The Metaphysical Foundations of Physical Science" (1786), which allowed him to challenge in a more systematic way the charges made by the previously mentioned anonymous review of "Critique of Pure Reason" in Göttingischen Gelehrten Anzeigen that his idealism could be compared with that of Berkeley. ("Metaphysical Foundations" was, incidentally, also a step on the way to Kant's subsequent refutation of idealism in the second edition of the "Critique".) The main project that Kant carried through apparently to his own satisfaction in "Metaphysical Foundations" was, following closely Förster's (1993, p. xxiv) more detailed discussion of this, the provision of a philosophical account for the apodictic certainty associated with the fundamental physical laws. Apodictic certainty was, however, only one of the two features which Kant believed any doctrine of nature must exhibit in order to qualify as a science; the other being systematic unity, and thus "Metaphysical Foundations" could not provide insight into the possibility of the systematicity of physics. Since he did not believe that collection of empirical data - an aggregate of perceptions - could provide the kind of a priori systematic unity one would expect to find in the laws and propositions of physics, Kant had then to find a way of making it possible for his philosophy to allow for the possibility of physics as a science, by providing principles for the investigation of nature - specific a priori topoi for classifying systematically forces of matter that can only be given empirically. It would also be necessary to demonstrate that we can a priori expect that nature would actually allow for such classification. The means by which he could do this latter task did not become available until he had written "Critique of Judgement" in 1790, and developed his principle of nature's appropriateness to our cognition, or as he himself put it: "...the principle of a purposiveness directed to our use of judgement as regards appearances" [AK 5:246, cited in Förster 1993, p. xxxv]. This principle provides the precondition under which a systematic empirical doctrine such as that of physics becomes a priori thinkable. When Kant's principle of the formal purposiveness of nature is seen in conjunction with his theory of matter, which had been explicated at some length in "Metaphysical Foundations", this makes a transition from the metaphysical foundations of natural science to physics not only possible, but also necessary. The task of the "Transition from the Metaphysical Foundations of Natural Science to Physics" - Kant's original working title for "Opus postumum" - was then to specify a method of bringing about the systematicity of physics through an outline of the systematicity of the objects of our outer senses, which is what physics as a science is concerned with.
Initially, Kant had believed that it would be sufficient for him to "follow the clue given by the categories and bring[ing] into play the moving forces of matter according to their quantity, quality, relation and modality in turn." [AK 21:311], but his numerous attempts to do so in the first part of the "Opus" repeatedly end up tailing off before he actually reaches the category of modality. This is due mainly to the fact that his discussion of the aggregate states of matter under the category of Quality raises the difficult problem of how to classify Calorific (Wärmestoff) which when present fluidifies matter, and when absent rigidifies it. Calorific can thus in itself neither be fluid nor rigid, and Kant finds it to be what he calls a "qualitas occulta". This impasse leads him to a further long discussion of Calorific (or ether), which he postulates must be some kind of "categorically given" "world-material" - the first cause of all moving forms of matter, which necessarily would required if we are to be able to have any kind of outer experiences at all. Since empty space cannot be an object of experience (space and time are, we remember, for Kant merely a priori unconditionally given wholes or intuitions), space, in order to be sensible, must be thought of as filled with a continuum of forces extended throughout the cosmos as a whole. Since experience must be a unity, this presupposes a constant motion of all matter on the sense organs of the human subject in order for perception to occur at all. The problem with this is that since Kant believed that we, as subjects, construct only indirect appearances, or appearances of appearances of the things in themselves, which cannot themselves be experienced, there would then be a problem to distinguish between the true object of physics, and these appearances. As Kant phrases it in the "Opus": "The objects of the senses, regarded metaphysically, are appearances; for physics, however, these objects are things [Sachen] in themselves, which affect sense, or as the subject affects itself (represent a priori)" . [AK 22:320]
His solution to this problem is to postulate that the subject constitutes itself as an empirical object by becoming an appearance of an object for itself. This must be done by shifting focus from the moving forces of matter that are in some way external to the subject, to the moving subject. This is possible since any physical body can be considered as a system of moving forces of matter, the concept of natural machines, or living organisms, cannot be excluded from the whole general system of moving forces of matter, something Kant had felt obliged to do previously. Whereas he now writes in this connection that:
"Organic bodies are natural machines, and, like other moving forces of matter, must be assessed according to their mechanical relationships, in the tendency of the metaphysical foundations of natural science; their appearances must be explained in this way, without crossing over into the system of the moving forces of matter according to final causes, which, being of empirical origin, belong to physics." [AK 21:186]
Because the subject has some consciousness of activating its own moving forces, it can thus anticipate the reacting moving forces of matter. "The subject which makes the sensible representation of space and time for itself is likewise an object to itself in this act. Self-intuition. For without this, there would be no self-consciousness of a substance" [AK 22:443] This allows the concepts of spatial and temporal relations become potentially sensible, something which would not have been possible before the positing of moving forces through which the subject is affected, these moving forces being necessarily prior. This leads him further into a discussion of the doctrine of self-positing (Selbstsetzungslehre), where he also begins to re-examine the notion of Ding an Sich, which in turn leads to a discussion of practical self-positing (i.e. the fact that the subject does not only constitute itself as an object of outer sense, but also as a person with rights and duties), and the idea of God. Finally he re-examines the whole question of what transcendental philosophy is, concluding that as well as being synthetic a priori knowledge from concepts, its is also the act of consciousness whereby the subject becomes the originator of itself as well as the whole system of technical-practical and moral-practical reason.
This particular later development of Kant's thought is especially interesting in this context to the extent that it shows how in his attempts to come to grips with the problem of qualifying philosophically the possibility of the existence of real external things; i.e. the object of study as it is conceived of by physics, rather than by metaphysics, he moved in a direction which brought him much closer to Peirce's concept of the real - that which "...is as it is, whether you or I or any group of men think it to be or not." [CP 8.118] - and of man as sign - "When we think, then, we ourselves, as we are at the moment, appear as a sign" [CP 5.283] - than one might have initially supposed possible. Further discussion of comparisons on these particular points would of course be of great interest, but since this all represents somewhat of a digression here, we shall merely leave it hanging for now and return to our prior discussion of Peirce and his classification of the sciences.
Murphey is of the opinion, and in this I would tend to agree with him, that the formulations by Peirce in the passage from the "Architectonic of Theories" from 1891 [CP 6.9] cited above need not necessarily be equated with the origins of his endeavour to create a classification of the sciences, but rather, as can be implied from what Peirce himself says, that his subsequent focus on the considerable task of creating a presuppositionally based classification of the sciences is something that would in any case naturally have developed as a consequence of his many intensive years of work prior to that on his Categories and Semeiotic. Beginning to write down a classification of the sciences would, in this kind of perspective, be for Peirce merely a logical next step towards a further development and refinement of his Semeiotic through the making of, as he put it, "a complete survey of human knowledge"; this, however, being the kind of monumental task which he had not felt he could begin with until after he had solved the most pressing and difficult problems that had occupied him during the years in which he was at work on the First and Second Systems and the subsequent complex problems that emerged out of this intensive period of work. With this work behind him providing his new starting point, however, the next step for Peirce in the direction of generalising the insights obtained thus far would be to, in his own words above "...proceed to the study of what the problem of philosophy consists in, and of the proper way of solving it." At this stage Peirce was becoming increasingly convinced that formal logic per se could not afford as solid a basis for knowledge as both Kant and he had assumed, and this is, in Murphey's view, the reason why his final classification of the sciences differs considerably from its predecessors in actually separating the theory of the categories from logic. It also seems clear that this is why Peirce's discussions of logic in general, and of its three subdivisions were given such a prominent position as they were - recall that in all twenty four memoirs (Memoirs 9-33) out of a total of thirty six were devoted to precisely these issues - in Peirce's Carnegie application of 1902. And indeed, the year 1902 is in fact a kind of turning point in this particular respect. Indeed, Murphey notes in this connection that Peirce's answer to the rhetorical question posed above: "If the categories are true of all thought, they are true of the special sciences by definition - why bother to prove in detail what is obvious a priori?" was very unclear in the Monist papers, which as we will recall, were all written between 1891 and 1893, and did not actually begin to become clear for him until around 1902.
In order to understand better why this may be the case it is necessary to understand Peirce's way of working. His belief in the architectonic theory of philosophy was probably the strongest and most long-lasting inheritance of his early commitment to Kant during what Murphey refers to as Peirce's "Kantian phase based on Kantian logic" (from 1857-1866), which Peirce later moved on from to develop his own system. Kant had previously defined the idea of "architectonic" in his "Critique of Pure Reason" as follows:
"By an architectonic I understand the art of constructing systems. As systemic unity is what first raises ordinary knowledge to the rank of science, that is, makes a system out of a mere aggregate of knowledge, architectonic is the doctrine of the scientific in our knowledge, and necessarily forms part of the doctrine of method.
... By a system I understand the unity of the manifold modes of knowledge under one idea. This idea is a concept provided by reason - of the form of the whole - in so far as the concept determines a priori not only the scope of its manifold content, but also the positions which the parts occupy relatively to one another. The scientific concept of reason contains, therefore, the end and the form of that whole which is congruent with this requirement. The unity of the end to which all the parts relate and the idea of which they all stand in relation to one another, makes it possible for us to determine from our knowledge of the other parts whether any part be missing, and to prevent any arbitrary addition, or in respect to its completeness any indeterminateness that does not conform to the limits which are thus determined a priori. The whole is thus an organised unity (articulatio), and not an aggregate (coacervatio). It may grow from within (per appositionem). It is thus like an animal body, the growth of which is not by addition of a new member, but by the rendering of each member, without change of proportion, stronger and more effective for its purposes." [A832f B860f][17]
Kant held that the basis of architectonic structure lay in formal logic. Peirce adopted initially both Kant's concept of system and his particular view of the role of logic in such systems, although he later became convinced, with regard to what he referred to as "the correspondences between the functions of judgement and the categories" in Kant's system, that "Its defect is that it affords no warrant for the correctness of the preliminary table, and does not display that direct reference to the unity of consistency which alone gives validity to the categories." [D2p1, cited in Murphey 1993, p. 56] Murphey goes on to maintain that although this important critical insight, which Peirce probably developed sometime in the period between 1862-1863, was the main motivating factor that started him upon his systematic studies of logic in general, it was nonetheless his basic belief in the two fundamental ideas of system as architectonic and the role of logic in such a system that constitutes the governing principle behind the development of his own philosophical system. Indeed, this appears almost certainly to be the case, as is witnessed by the following passage taken from the final version of Section 1 in his Carnegie application, where Peirce writes:
"Owing to my treating logic as a science, like the physical sciences in which I had been trained, and making my studies special, minute, exact, and checked by experience, and owing to the fact that logic had seldom before been so studied, discoveries poured in upon me in such a flood as to be embarrassing. This has been one reason why I have hitherto published but a few fragments of outlying parts of my work, or slight sketches of more important parts. For logic differs from the natural sciences and, in some measure, even from mathematics, in being more essentially systematic. Consequently, if new discoveries were made in the course of writing a paper, they would be apt to call for a remodelling of it, a work for mature reconsideration. Still, as far as I remember, no definitive conclusion of importance to which I have ever been led has required retraction, such were the advantages of the scientific methods of study. Modification in details and changes (very sparse) of the relative importance of principles are the greatest alterations I have ever been led to make. Even those have been due, not to the fault of the scientific method, but chiefly to my adherence to early teachings." [MS L75 FV 346-349]
Hence, since Peirce believed consistently in the basic idea of an architectonic system, he worked continuously to develop his system of philosophy and his logical method in such a way for the whole of his productive scientific life, and he seemed to have in his mind at all times a quite clear and definite conception of such an all-encompassing philosophical system, even though his understandings of what such a system might necessarily need to involve, for instance in terms of ontological systemics, came to change somewhat over time.