The evolving theoretical framework of Peirce’s classification systems

Tony Jappy, University of Perpignan Via Domitia, Perpignan, France tony@univ-perp.fr Abstract: Establishing the correct order of trichotomies in Peirce’s projected ten-division typology of 1908 has been a preoccupation of Peirce scholars since at least 1945. Most seem to assume that the same phenomenological framework was adopted by Peirce in all his classification systems from 1903 to 1908 and that these ten divisions form a homogeneous set. The paper examines the status of typologies in Peirce’s semiotics and, by comparing two hexads of divisions from 1904 and 1908, shows how the theoretical framework of each was based upon entirely different principles. Keywords: typology, phenomenology, category, ontology, universe The evolving theoretical framework of Peirce’s classification systems 1. Introduction The problem facing any Peirce scholar attempting to give a general overview of his work in semiotics is that the theory is not homogeneous: not only does it’s development span almost fifty years, but, as many specialists have noted, it advances in a number of stages, usually seen as three, Atkin (2010) and Liszka (1996), or four Murphey (1993), for example. These stages correspond to the number of divisions Price derives at a given time – one in 1867, three in 1903, six in 1904, etc. – and also to the number of correlates participating in sign-action – three in 1903 and six in 1904. In an effort to synthesize the theory, most detailed presentations gravitate around the period of the Lowell Lectures in 1903 to the almost general neglect of the later statements on the classification of signs. Such a reduction, although understandable in view of the sheer volume of Peirce’s statements on semiotics, is misrepresentative for several reasons. First, his theory of signs, upon which the earlier and later typologies were defined, was not frozen in 1903, but continued to develop over the following five or six years, the typologies with them. Second, many of those Peirce scholars who have taken the trouble to investigate the later typologies have largely restricted their interest to the correct ordering of the ten divisions that Peirce hoped would yield sixty-six classes of signs, and, assuming them all to be based upon the same theoretical framework, have rarely bothered to review the foundations upon which they were based. Finally, such neglect tends to obscure the central role that the construction of typologies and the classification of signs enjoyed in Peirce’s conception of semiotics: ‘We may add now that logic also is a classificatory science….and that in his own lifetime as a whole, [Peirce] devoted more labor to the classification of signs than to any other single field of research.’ Thus wrote Max Fisch in his Introduction to the first volume of the Writings (1982: xxii). The paper seeks to compensate for this inattention by showing how an examination of the evolving typologies sheds light on the development of Peirce’s conception of signs and, above all, on an important shift in the theoretical framework which underwrote it. After having established the status of classifications in his logic, the paper sets out some of the major stages in their development and then describes the significant differences between the well-canvassed 1903 definition of the sign with its ten-class typology and the later conception by comparing two different hexadic classification systems. 2. The quest for sign-classes Should there be any doubt as to the priority which Peirce himself accorded the identification and classification of signs, we have his testimony as evidence. For example, he doesn’t hesitate to assimilate the tasks of the logician to that of the zoologist: ‘We are in the situation of a zoölogist who wants to know what ought to be the meaning of ‘fish’ in order to make fishes one of the great classes of vertebrates’ he wrote to Lady Welby in a letter dated 12 October 1904 (CP 8.332). In the series of variants composing the text referred to as ‘Pragmatism’ he further adds: Now how would you define a sign, Reader? I do not ask how the word is ordinarily used. I want such a definition as a zoologist would give of a fish, or a chemist of a fatty body, or of an aromatic body, – an analysis of the essential nature of a sign, if the word is to be used as applicable to everything which the most general science of sēmei’otic must regard as its business to study (R318 585, 1907) Moreover, for Peirce, the goal of the logician was not simply to classify signs, it was to identify and ‘collect’ as many as possible: Possibly a zoölogist or a botanist may have so definite a conception of what a species is that a single type-specimen may enable him to say whether a form of which he finds a specimen belongs to the same species or not. But it will be much safer to have a large number of individual specimens before him, from which he may get an idea of the amount and kind of individual or geographical variation to which the given species is subject. (CP 1.224, 1902) More importantly, there is Peirce’s testimony as to the extreme difficulty he found himself in late in his career, by now aware of the fact that he was unable to complete the task he had set himself in his logical researches. Consider, for example, the following statement: My excuse for not answering the question scientifically [that those signs that have a logical interpretant are either general or closely connected with generals] is that I am, as far as I know, a pioneer, or rather a backwoodsman, in the work of clearing and opening up what I call semiotic, that is, the doctrine of the essential nature and fundamental varieties of possible semiosis; and I find the field too vast, the labor too great, for a first-comer. (R318 119, 1907) There is no mistaking Peirce’s research project here – it was to identify all possible varieties of sign-classes in spite of the by now apparent immensity of the task. He doesn’t mention classifications as such at this point, nor does he give a single typology of consequence in the text ‘Pragmatism’ from which the quotation is drawn. It is nevertheless this complex process of definition and division and the subsequent classification of the results thus obtained that Peirce saw as one of the most important tasks of the logician. Any remaining doubt as to the significance of total classification in his conception of semiotics can surely be dispelled by the following statement from a letter to William James outlining his project for a ‘System of Logic’ which echoes the earlier extract from 1907, but more specifically: ‘My classification of signs, however, is intended to be a classification of possible signs and therefore observation of existing signs is only of use in suggesting and reminding one of varieties that one might otherwise overlook’ (EP2 500, 1909). From this we can gather that Peirce was more interested in establishing as exhaustive an inventory of signs as possible than in analyzing and classifying the signs around him as perhaps a present-day semiotician might. Moreover, as he suggests in the Logic Notebook, his search was not necessarily limited to ten divisions: ‘more than these ten I don’t enquire into, not because I don’t think they are in truth there, but simply because it will be all I possibly can do to define and to prove these ten’ (R339 360r, 1909). Cf., too, this remark in a draft letter to Lady Welby of 24 December 1908: ‘On these considerations I base a recognition of ten respects in which Signs may be divided. I do not say that these divisions are enough. But since every one of them turns out to be a trichotomy, it follows that in order to decide what classes of signs result from them, I have 310 or 59049, difficult questions to carefully consider; and therefore I will not undertake to carry my systematical division of signs any further, but will leave that for future explorers’ (CP 8.343). 3. Stages in the development of the classification systems We note, first, that in the four and a half years between August 1904 and December 1908 Peirce established in letters, drafts and the Logic Notebook no fewer than twelve different, mainly complete, typologies, whereas in the thirty-six years between 1866 and 1902 there was only one very simple classification system (Table 1) and none of significance after 1908, surely evidence of the considerable experimentation undertaken by Peirce during this short period. Table 1 displays the single-division typology of representamens of 1867, ‘representamen’ being at the time the term for the unit of representation, while ‘sign’, which was later to replace ‘representamen’, was another term for the index. At that time, too, the symbol, which Peirce considered the sole object of study of the logician, subdivided into the traditional tripartite distinction between term, proposition and argument. He later informed Lady Welby that he had modified this ‘old division’ in order to make it applicable to signs generally (SS 33, 1904). As far as the theoretical framework of this typology is concerned we note that at the time Peirce was attempting to derive his categories from the logic – it was logic therefore which provided the framework, not phenomenology, as this was not really fully defined until the turn of the century. Table 1: Peirce’s trichotomy of representamens, 1867 Second, in their very diverse ways, these typologies reflect how Peirce’s conception of sign-action developed between the two major statements on signs in the Lowell Lectures of 1903 and the letter to Lady Welby of 23 December 1908 (EP2 478–481). In 1903, for example, within Peirce’s theory of triadic relations the sign was defined thus: ‘A Sign, or Representamen, is a First which stands in such a genuine triadic relation to a Second, called its Object, as to be capable of determining a Third, called its Interpretant, to assume the same triadic relation to its Object in which it stands itself to the same Object’ (CP 2.274), a process in which the sign mediates between the determining object and the determined interpretant. By 1908 it had been explicitly defined as a medium as opposed to the unit of representation, and had effectively ceased to be one of three correlates of a triadic relation but, rather, one of a hexad of logically distributed correlates: the sign, two objects and three interpretants. Third, of those twelve typologies no fewer than eight were established in the two years between August 1904 and August 1906 in the Logic Notebook (R339 239v, 240r, 1904; 252r, 253r, 255r, 262r, 1905; 275r, 285r, 1906). Indeed, the years 1905 and 1906 proved to be a turning-point in the development of Peirce’s sign-systems in several ways: ‘From the summer of 1905 to the same time in 1906, I devoted much study to my ten trichotomies of signs’ (CP 8.363, 1908). To begin with, he had now begun working on ten divisions. Furthermore, in a draft letter to Lady Welby from 1906 the sign was redefined as a medium communicating to the three interpretants form emanating from the dynamic object: Peirce had already introduced the definition of the sign as a medium of communication in the 1905 paper ‘Issues of Pragmatism in the Normative Sciences’ (EP2 388–392), and in the manuscript R793 (1–2, 1906) I use the word “Sign” in the widest sense for any medium for the communication or extension of a Form (or feature). Being medium, it is determined by something, called its Object, and determines something, called its Interpretant or Interpretand... The Form, (and the Form is the Object of the Sign), as it really determines the former Subject, is quite independent of the sign; yet we may and indeed must say that the object of a sign can be nothing but what the sign represents it to be. Therefore, in order to reconcile these apparently conflicting Truths, it is indispensible to distinguish the immediate object from the dynamical object. The same form of distinction extends to the interpretant; but as it applies to the interpretant, it is complicated by the circumstance that the sign not only determines the interpretant to represent (or to take the form of) the object, but also determines the interpretant to represent the sign. (RL463 26–28) The page numbers in the references are those of Peirce himself and correspond to pages 30–32 in this portion of RL463. The relevant parts of the manuscript can also be found at EP2 477–478 and SS 195–201. As can be seen from the final sentence of the passage, this particular draft is also important for introducing, albeit implicitly, the order of semiosis into the presentation of the correlates in sign-action. After having introduced and the sign and the dynamic (Peirce’s term is ‘dynamical’, but for convenience the terminology has been standardized here) and immediate objects, Peirce goes on to complete the description of the process by describing the three interpretants in the following order: There is the Intentional Interpretant, which is a determination of the mind of the utterer; the Effectual Interpretant, which is a determination of the mind of the interpreter; and the Communicational Interpretant, or say the Cominterpretant, which is a determination of that mind into which the minds of utterer and interpreter have to be fused in order that any communication should take place. This mind may be called the commens. It consists of all that is, and must be, well understood between utterer and interpreter, at the outset, in order that the sign in question should fulfill its function. (RL463 29) An important aspect of these two passages is the fact that the intentional interpretant – at the time Peirce’s term for the immediate interpretant – is presented as both a determination of the sign and of the mind of the utterer. If it is the determination of the sign, then it must follow the sign in the process of semiosis. This being the case, it seems only logical that the sign should follow the immediate object and that this should follow from the dynamic. In this way the passages anticipate the later explicit statement of the order of determination in the 23 December 1908 letter to Lady Welby: Od, Oi, S, Ii, Id, If (SS 84, quoted below). In 1905, too, Peirce had established for the first time a typology involving not six divisions as before, but ten (R339 253r, 8 October 1905), after which all but one the typologies that he proposed involved ten divisions. 1906 was the year, finally, in which Peirce explicitly introduced a fundamental distinction between categories and universes in ‘Prolegomena to an Apology for Pragmaticism’, making explicit the universes to which the subjects mentioned in the extract (RL463 26–28) quoted earlier belonged: As noted by the editors of volume two of The Essential Peirce, (EP2 555n3), Peirce’s use of the concept of a universe occurs frequently at this time. Oh, I overhear what you are saying, O Reader: that a Universe and a Category are not at all the same thing; a Universe being a receptacle or class of Subjects, and a Category being a mode of Predication, or class of Predicates. I never said they were the same thing; but whether you describe the two correctly is a question for careful study. (CP 4.545, 1906) In short, the passage suggests that Peirce is turning his back on the logico-phenomenological framework within which he had established his theory of signs since the mid-1860s, and that he is evolving towards an ontological approach to classification, anticipating in this field, too, the definitions advanced in the 23 December 1908 letter. Table 2: The ten-class typology of 1903 Summarizing, then, Table 1 sets out the single, three-class division of 1867, Table 2 displays the three divisions yielding ten classes from his 1903 triadic definition of the sign, while the six-division typologies of 1904 and 1908 (respectively Tables 3 and 4) differ significantly in that there are no longer just three correlates in his new conception of the action of a sign, but six. Table 3, in which the symbol S stands for the sign, of course, sets out the six divisions in the order proposed by Peirce in the letter to Lady Welby of 12 October 1904 (SS 22–36), and is henceforth referred to as ‘hexada’. Table 3: Hexada, 1904 Finally, in the letter to Lady Welby dated 23 December 1908, Peirce posited a new hexad of divisions based upon the correlates themselves in the order defined below, this system generating twenty-eight classes of signs and henceforth referred to as ‘hexadb’. The process in which the sign is engaged in 1908 he describes as follows: It is evident that a Possible can determine nothing but a Possible; it is equally so that a Necessitant can be determined by nothing but a Necessitant. Hence, it follows from the Definition of a Sign that since the Dynamoid Object determines the Immediate Object, which determines the Sign itself, which determines the Destinate Interpretant, which determines the Effective Interpretant, which determines the Explicit Interpretant, the six trichotomies, instead of determining 729 classes of signs, as they would if they were independent, only yield 28 classes, and if, as I strongly opine (not to say almost prove) there are four other trichotomies of signs of the same order of importance, instead of making 59049 classes, these will only come to 66. (SS 84) This statement, apparently Peirce’s only reference to the 28-class system – although it necessarily participates in the construction of the sixty-six classes – not only establishes the logical order of determination holding between the six correlates but, by virtue of the reference to the modes of being (Possible, Existent and Necessitant) of the universes with respect to which each correlate can be referred for the purposes of classification, defines the dependency relation reducing the 729 possible classes to twenty-eight. Given his desire to identify all possible signs, the attraction of sixty-six far outweighed the twenty-eight offered by the six divisions based upon the correlates themselves, and probably explains why he never sought to develop what is in fact an ‘organically’ organized set of trichotomies based upon the process of sign-action itself. Moreover, as he suggests to William James in the letter outlining his ‘System of Logic’ mentioned earlier (EP2 500), he was not so much interested in classifying existing classes of signs as discovering new ones. Table 4: Hexadb, 1908 3.1. Completing Table 4 Whether the order of determination of the correlates advanced in the letter is the order of the trichotomies in the classification system yielding the twenty-eight classes of signs requires clarification. But, before continuing, there are two points to be noted regarding Table 4. First, Peirce never set out his typologies in rows in the manner of Table 4, preferring a vertical format as on Table 3, as in the typologies of the Logic Notebook, and as in the case of paragraph CP 8.344 below, for example, of which only the relevant six divisions have been retained: The ten respects according to which the chief divisions of signs are determined are as follows: 1st, According to the Mode of Apprehension of the Sign itself, 2nd, According to the Mode of Presentation of the Immediate Object, 3rd, According to the Mode of Being of the Dynamical Object, … 5th, According to the Mode of Presentation of the Immediate Interpretant, 6th, According to the Mode of Being of the Dynamical Interpretant, … 8th, According to the Nature of the Normal Interpretant…. (CP 8.344) In other words, the various ‘respects’ with reference to which the sign was to be classified were listed in an order beginning with the sign S, Oi, Od, Ii, Id, If (the normal interpretant has been standardized to If), an order which differs significantly from the twenty-eight class hexad on Table 3, where the order of the respects is S; S–Od, S–Oi (Oi); S–If, S–Id, S–Ii, and from the one on Table 4, where we have Od, Oi, S, Ii, Id, If. The advantage of the layout of the typology on Table 4 is that the classes of signs can be read quite simply along the rows. For example, a sign which is to produce self-control can be seen to be necessarily collective, and an abstractive sign to be necessarily gratific. Furthermore, if the hierarchy is respected, an action-producing copulant token as in the case of any verbal utterance can easily be seen to be a logically valid sign; conversely, the logical impossibility of having an abstractive, designative action-producing sign is also clearly displayed. The way Peirce set out his typologies was very different, as shown on Table 3, but this in no way invalidates the format adopted for reconstructed Table 4. Second, in his letter to Lady Welby Peirce only identified the first three divisions and the subdivisions they comprise. The final three trichotomies on Table 4 concerning the interpretants have been reconstructed from the typologies Peirce established in drafts in the days following the 23 December letter. In this way, the three subdivisions related to the Destinate interpretant (standardized to ‘immediate’, Ii, on Table 4), for example, are, in order of increasing complexity, hypothetical, categorical and relative. Note, too, that Peirce was not entirely happy with this particular division, as can be seen from paragraph CP 8.369: ‘V. As to the nature of the Immediate (or Felt?) Interpretant, a sign may be: Ejaculative, or merely giving utterance to feeling; Imperative, including, of course, Interrogatives; Significative. But later I made this the 7th Trichotomy and for the fifth substituted—with great hesitation—: Hypothetic, Categorical, Relative’. The seventh division as he re-worked it would be ineligible to figure in the hexad on Table 4 as S—Id is a relation, not an individual subject. Similarly, the Effective (dynamic) Interpretant trichotomy is based upon the sixth division in CP 8.370: ‘VI. As to the Nature of the Dynamical Interpretant: Sympathetic, or Congruentive; Shocking, or Percussive; Usual.’ The division containing the classes of signs identified by the nature of the final interpretant, referred to in the 23 December letter as the ‘Explicit’ interpretant, distinguishes between signs intended to produce, respectively in order of increasing complexity, feeling, action or self-control. The purposive, telic nature of the final interpretant, which is, after all, the defining characteristic of the signs which this division identifies, might be considered different from its ‘nature’, as Peirce’s definition has it, but by a process of elimination, the classes of signs thus defined can be assumed to correspond to this division. Table 4 has in this way been obtained from an incomplete description by Peirce but is easily reconstructed from other typologies, in spite of his misgivings concerning the viability of the classes forming the immediate interpretant division. 3.2. The ordering problem In view of remarks Peirce makes to Lady Welby in the letter of 23 December 1908 and in view, too, of the sometimes fierce debate concerning the correct ordering of the ten divisions among Peirce scholars, it is important to justify the order proposed in this paper. Two persuasive arguments plead in favour of the order retained on Table 4, which is, after all, the order of semiosis. First, if in semiosis a dynamic object and then an immediate object can ‘determine’ a sign, i.e. cause it to be such as it is, there is clearly a logical reason why the sign must appear after these two correlates in the table. If the sequence given on Table 4 is correct – the hierarchy ordering the universes applies to the classifying system as much as to the order of semiosis – it is difficult to see how in the process of semiosis the concept of beauty, for example, which is of a possible nature – beauty is an abstractive sign according to Peirce (SS 83–84) – could determine the type (a necessitant) that names it: such a determination would violate the universe hierarchy principle. Moreover, in CP 8.366 Peirce later identifies as abstractives such basic material qualities as colour, mass and whiteness: these can be found, for example, in a painting, a sculpture or a piece of architecture, and as such would be compatible not with a type, but, rather, with a simple mark. Note, too, that in the immediately following paragraph illustrating the way the universe hierarchy principle operates Peirce explicitly asserts the necessary compatibility of an abstractive and a mark: I was of the opinion that if the Dynamical Object be a mere Possible the Immediate Object could only be of the same nature, while if the Immediate Object were a Tendency or Habit then the Dynamical Object must be of the same nature. Consequently an Abstractive must be a Mark, while a Type must be a Collective, which shows how I conceived Abstractives and Collectives. (CP 8.367, 1908). Now, referring to Table 4, if the compatibility of a type is restricted to a (copulant) collective sign, the dynamic object must precede it in the typology: were the sign to occupy initial position on the table it would be compatible not only with collective signs but also with concretives and abstractives. Similarly, for the compatibility of an abstractive sign to be restricted to a mark, in this case, too, the dynamic object needs to precede the sign, as can clearly be seen from the table. The situation is complicated, however, by the fact that the order retained on Table 4 is in sharp contradiction with the order implied by remarks made by Peirce to Lady Welby concerning the noun beauty. In the letter he suggests to Lady Welby that ‘it is the ultimate reference, and not the grammatical form, that makes the sign [the word beauty] an Abstractive’ (SS 83). Although there are other criteria, the beauty example nevertheless offers a most compelling second argument in favor of the order retained on Table 4. Peirce’s remark concerning beauty is problematic as the word beauty is a type – all words in a dictionary are by definition necessitant and are therefore classified as types (SS 83). However, a type can only be classified as an abstractive sign whose dynamic object is a member of the universe of possibles if it precedes the dynamic object in the classification system. In this case, the order displayed on Table 4 would be incorrect, and would need to be replaced by a system displaying the order S, Oi, Od, Ii, Id, If, for example (Table 5). In this case the correlates are set out in an order favoured by Peirce in most of his typologies. See Jappy (in press: Ch. 3) for a discussion. Peirce’s disconcerting description of the common noun beauty as an abstractive sign therefore has significant consequences for the way the table is to be ordered. There is one conclusive way to approach the problem: namely, to analyze the term beauty with the sign in initial position as on Table 5. Table 5: Hexadb with the type beauty in initial position Table 5 displays the consequences of retaining the order suggested by Peirce’s discussion of the classification of beauty as an abstractive: the noun has been placed first on the table in the necessitant sign position as Peirce’s comments suggest – it is a type – followed by the immediate object division, with respect to which it is of no consequence whether the sign is copulant, designative, or descriptive, while with respect to its dynamic object it is classified as an abstractive. Now, according to the hierarchy principle, an abstractive sign, which is a sign with a possible object with respect to the Od division on Table 5, can only determine a sequence of interpretants from the universe of possibles, necessarily identifying the sign here as also hypothetical, sympathetic and gratific. This, in fact, is the real problem concerning the classification of the common noun beauty as an abstractive sign: namely, it is difficult to see how the effect produced by a verbal sign might be limited to feelings as when such a sign is classified as gratific – with the possible but theoretically dubious exception of onomatopoeia – if the interpretant sequence proposed on Table 5 were adopted. The very meaning of the word has first to be ‘processed’ cognitively, so to speak, at the immediate interpretant stage in semiosis as relative or categorical for it to be capable of producing any subsequent effect at all, even a feeling, and this is also the case if we classify beauty as a verbal token. Therefore Peirce’s remark to the effect that the type beauty is an abstractive sign seems to be yet another example of a ‘sop to Cerberus’, with the intention of simplifying matters for his British correspondent, and the order suggested on Table 4 must stand. 4. The two hexads compared Returning now to Table 4, we note two immensely significant consequences of its fundamentally different way of classifying signs from the single division of the period from 1867 to 1902 and from the three divisions yielding the ten classes of 1903. To begin with, the S—Od division which Peirce claimed to be the one he used most (CP 8.368) and the one yielding the universally known division into icon, index and symbol, has disappeared from the scheme on Table 4. The system here is based not on how a sign represents its object – is not based on the sign’s mode of representation, in other words – but, amongst other things, upon the sorts of objects it represents. The icon-index-symbol division is one of four divisions appended to the description of the hexad in the 23 December letter (SS 84–85), and becomes the fourth trichotomy in the ten Peirce describes in the draft of 25 December 1908 (EP2 489). Gone, too, is the important S—If division defining rheme, dicent and argument, the absence of the dicisign being particularly significant in view of the importance Peirce seems to have accorded it in the 9 March 1906 draft, and in view of the importance attributed to the proposition in traditional theories of logic it is no wonder that Peirce was at pains to retain these two divisions in particular in his projected 66-class system. Given the important logical status of these two divisions within earlier typologies the final sections of the paper examine the theoretical differences distinguishing hexada and hexadb. The descriptions come from two letters, and also have in common the fact that the sign-systems they describe are prefaced by two very different theoretical frameworks within which Peirce established their respective hexadic typologies. 4.1. The phenomenology of 1904: You know that I particularly approve of inventing new words for new ideas. I do not know that the study I call Ideoscopy can be called a new idea, but the word Phenomenology is used in a different sense. Ideoscopy consists in describing and classifying the ideas that belong to ordinary experience or that naturally arise in connection with ordinary life, without regard to their being valid or invalid or to their psychology. In pursuing this study I was long ago (1867) led, after only three or four years' study, to throw all ideas into the three classes of Firstness, of Secondness, and of Thirdness. This sort of notion is as distasteful to me as to anybody; and for years, I endeavored to pooh-pooh and refute it; but it long ago conquered me completely. Disagreeable as it is to attribute such meaning to numbers, and to a triad above all, it is as true as it is disagreeable. The ideas of Firstness, Secondness, and Thirdness are simple enough… I call these three ideas the cenopythagorean categories. (CP 8.328, 1904) This passage from the 12 October 1904 letter to Lady Welby is followed by a very detailed analysis of the three categories of Firstness, Secondness and Thirdness (CP 8.329–8.332) – the latter enabling Peirce to introduce the concept of sign – after which he defines the sign and its five correlates (CP 8.333). As in the manuscript R478, where he establishes his first two typologies, namely the icon-index-symbol division and the one replacing the traditional triad of term, proposition and argument respectively by sumisign, dicisign and suadisign (EP2 273-275, 1903), he prefaces his presentation of hexada in the letter to Lady Welby of 12 October 1904 by this very long introduction to the categories. These constitute the theoretical guarantee of the logical viability of the sign-systems of this period – both the definitions of the sign and the divisions which the sign and its correlates generated either singly or in combination – and the framework within which he was working at the time and had been for almost forty years was explicitly phenomenological. The typologies which this framework gave rise to are to be found typically on Tables 2 and 3. Noticeable on Table 3 are the nature of the divisions he sets up and the order in which they occur in the typology: this and the one for the August? 1904 entry in the Logic Notebook (R339 239v) The typology is to be found on the verso of a page dated 10 July 1903, but belongs clearly to the period when Peirce was working on six-element classifications that began in 1904. are the only two to employ this order, S, S—Od, S—Oi (Oi), S—If, S—Id, S—Ii. Hardwick’s note at this point raises an interesting problem. He suggests that the division concerning the immediate object is a relation: ‘In his letter dated 23 December 1908, Peirce provides a trichotomy for the immediate object’ (SS 33, n19), the highlighted preposition for suggesting that this was not the case in 1904. However, Peirce actually writes ‘In respect to its immediate object a sign… [Emphasis added]’, whereas in the case of the dynamic object and the signified interpretant, for example, Peirce writes ‘In respect/regard to their/its relation(s) to… [Emphasis added]’ (SS 33). In the first case the division appears to be established with respect to a correlate, in the second case it is established with respect to relations between sign and correlates, hence the hesitation at the S—Oi (Oi) position on Table 3. In view of the general introduction to the typology: ‘Now signs may be divided as to their own material nature, as to their relations to their objects, and as to their relations to their interpretants’ (SS 32), Table 3 is admittedly hypothetical. Whatever the case, it has no bearing on the general tenor of the paper. The majority of the other typologies in the Logic Notebook present an expanded version of the order of correlates in the original triadic relation that integrates the two objects and three interpretants in the following manner: S, Oi, Od, Ii, Id, If, with various others interleaved between them as in the 8 October 1905 entry (R339 252r), for example. We note, too, that there is only one independent correlate division, namely the sign, and that two divisions on Table 3, the relational respects S—Oi and S—Ii, were subsequently abandoned completely, with the immediate object and interpretant henceforth occupying independent positions in all the typologies to follow. Finally, as in the ten-class system of 1903 (Table 2), the majority of divisions are relational. 4.2. The ontology of 1908 Peirce begins the relevant passage in the 23 December letter to Lady Welby by defining the sign and its two major correlates, referring not to the categories as before, but to three universes distinguished by three modalities of being. In short, having defined a hexad of divisions in 1904 using the categories as distinguishing criteria he now defines another basing his distinctions on an entirely different set of criteria as the framework for hexadb: It is clearly indispensable to start with an accurate and broad analysis of the nature of a Sign. I define a Sign as anything which is so determined by something else, called its Object, and so determines an effect upon a person, which effect I call its Interpretant, that the latter is thereby mediately determined by the former... I recognize three Universes, which are distinguished by three Modalities of Being. One of these Universes embraces whatever has its Being in itself alone, except that whatever is in this Universe must be present to one consciousness, or be capable of being so present in its entire Being… I denominate the objects of this Universe Ideas, or Possibles, although the latter designation does not imply capability of actualization… Another Universe is that of, 1st, Objects whose Being consists in their Brute reactions, and of, 2nd, the Facts (reactions, events, qualities, etc.) concerning those Objects, all of which facts, in the last analysis, consist in their reactions. I call the Objects, Things, or more unambiguously, Existents, and the facts about them I call Facts… The third Universe consists of the co-being of whatever is in its Nature necessitant, that is, is a Habit, a law, or something expressible in a universal proposition. (SS 81–82) The theoretical framework within which Peirce is now working is ontological in the widest sense, involving the three universes defined above, three universes which are entirely different from the phenomenological categories of 1903–1904. A universe, says Peirce, is not the same as a category: ‘Let us begin with the question of Universes. It is rather a question of an advisable point of view than of the truth of a doctrine. A logical universe is, no doubt, a collection of logical subjects, but not necessarily of metaphysical Subjects, or ‘substances’; for it may be composed of characters, of elementary facts, etc.’ (CP 4.546, 1906). In this way, the correlates involved in semiosis figure on Table 4 as subjects susceptible of belonging to one or other of these universes. As seen earlier, this was not a new position on Peirce’s part, as he had already employed the concept of the subject in a passage from the 9 March 1906 draft to Lady Welby: In order that a Form may be extended or communicated, it is necessary that it should have been really embodied in a Subject independently of the communication; and it is necessary that there should be another subject in which the same form is embodied only as a consequence of the communication. The Form, (and the Form is the Object of the Sign), as it really determines the former Subject, is quite independent of the sign. (RL463 26–27, 1906) In this case, as in the letter two years later, the correlates thus described are not subdivided in any way by Firstness, Secondness or Thirdness but are subjects or members of a given universe: the dynamic object is one subject, the sign is another, etc. This does not mean, of course, that Peirce had now discarded his categories altogether, or that they were no longer of interest to him – indeed he continues to discuss them in his correspondence with Lady Welby and elsewhere (e.g. CP 1.288–292, c. 1908). Cf. this passage from a latter to James dated 25 December 1909: ‘I found Logic largely on a study which I call Phaneroscopy, which is the keen observation of and generalization from the direct Perception of what we are immediately aware of.’ (EP2 501) Nevertheless, after 1906 Peirce never again employed his categories as criteria in the classification of signs. 4.3. The evolving object Just why Peirce should have chosen to distinguish the different classes of his two late typologies by means of universes rather than by his earlier categories is difficult to establish, but the decision is surely not unrelated to his developing comprehension of the sign’s object. For it is not always realized that the object, no less than the introduction and development of the three interpretants, underwent a considerable evolution in the years following the Lowell Lectures. At that time, the basic triadic relation held between a single object, the representamen and a single interpretant. In addition to introducing a hexadic typology, the letter of 12 October 1904 also heralds a new stage in the development of the object: ‘I’m now prepared to give my division of signs, as soon as I’ve pointed out that a sign has two objects, its object as it is represented and its object in itself’ (CP 8.333). Now the very possibility that a single sign should be capable of representing an object quite distinct from the immediate one perceived in the sign thus stems from the now explicit distinction between the immediate and the dynamic objects. The fact that the two objects are two distinct subjects in a typology, as in hexadb, makes it possible for them to have different classificational values within it, a principle that Peirce had already advanced in the Logic Notebook: ‘The immediate object of a sign may be of quite a different nature from the real dynamical object’ (R339 277r, 1906). A further important aspect of Peirce’s evolving understanding of the object can be seen in the description that he gives of the (dynamic) object As can be clearly to be seen in the description to follow, these are objects which necessarily exist ‘outside’ the sign. in a text entitled ‘The Neglected Argument for the Reality of God’, which, like the letter describing hexadb, dates from 1908. In it Peirce details the sorts of entities – which include, of course, the sign, the two objects and the three interpretants, as can be seen on hexadb – that the three universes are receptacles of: Of the three Universes of Experience familiar to us all, the first comprises all mere Ideas, those airy nothings to which the mind of poet, pure mathematician, or another might give local habitation and a name within that mind. Their very airy-nothingness, the fact that their Being consists in mere capability of getting thought, not in anybody’s Actually thinking them, saves their reality. The second Universe is that of the Brute Actuality of things and facts. I am confident that their Being consists in reactions against Brute forces, notwithstanding objections redoubtable until they are closely and fairly examined. The third Universe comprises everything whose Being consists in active power to establish connections between different objects, especially between objects in different Universes. Such is everything which is essentially a Sign, —not the mere body of the sign, which is not essentially such, but, so to speak, the Sign’s Soul, which has its Being in its power of serving as intermediary between its Object and a Mind. Such, too, is a living consciousness, and such the life, the power of growth, of a plant. Such is a living institution, —a daily newspaper, a great fortune, a social ‘movement.’ (CP 6.455, 1908) Thus from the hexadb as represented on Table 4 we can infer that the universes determined by ‘a living consciousness, the life, the power of growth, of a plant….a living institution, —a daily newspaper, a great fortune, a social ‘movement’’ are far more complex from a logical point of view than the object as conceived in 1903: these are objects from the universe of necessitants, and therefore the signs representing them are to be classified, at that point in the hexad on Table 4 (Od), as collectives. However, as the table also shows, while a given sign can be classified as a collective in reference to its dynamic object, at the immediate object stage (Oi) that same sign can be classified as either copulant, designative or descriptive. Now if at this point the sign is designative or descriptive, the immediate object will necessarily belong to a different universe from that of the dynamic. Furthermore, even if in the immediate object division the sign is classified as copulant, to be physically perceivable at all it will have itself to be a token at (S): a given sign can therefore have both an immediate and a dynamic object more complex than itself. The possibility for a sign to be classified according to objects in different universes from itself, as in the case of a collective designative hypothetical token, clearly bears out Peirce’s earlier remark in the Logic Notebook to the effect that the immediate object may not be at all like the dynamic. Summarizing, then, we find that the two hexads are organized in entirely different ways. The first, from 1904, harks back to the principles expounded in the Lowell Lectures a year earlier, and uses the categories – predicates – as criteria with which to establish the subdivisions of each trichotomy. By 1908, however, Peirce had abandoned this phenomenological framework as a means of classifying signs, preferring the ontological nature of the three universes described above. 5. Conclusion While the purpose and methodology of Peirce’s typological project remained relatively constant throughout the period from the mid-1860s to 1908, the theoretical framework in which these operated evolved significantly, particularly once the expanded set of correlates was introduced explicitly in 1904 and the sign had been defined as a medium for the communication of form as opposed to a unit of representation in 1906. The three categories, which, irrespective of their origin, had accompanied all his work in the classification of signs from the earliest period until approximately 1904, was superseded in 1908 by a broad ontological vision embracing three universes, receptacles with respect to which the sign and its correlates could be referred in the course of the classification of a sign. The logical principles supporting this later typological approach to signs, the fruit of an evolution in Peirce’s conception of the object and of the rapid theoretical development that his conception of sign-action experienced in those years between 1904 and 1906, are, therefore, radically different from those of the earlier approach, and it is doubtful that the two will ever be combined in a satisfactory manner in the quest for the sixty-six classes that Peirce hoped to identify. In his search for possible signs, Peirce was clearly not interested in the novelty and semiotic potential of the 28-class system. Chapters 4 and 5 of Jappy (in press) suggest ways in which to exploit the semiotic potential of the 1908 hexad. The sixty-six classes proved to be a more interesting goal, and one can only surmise what might have happened if he had developed the full analytical power of the 1908 hexad as described above. 6. References Atkin, A., (2010), ‘Peirce's Theory of Signs’ http://plato.stanford.edu/entries/peirce-semiotics/ Accessed March 2016. Fisch, Max, (1982), Introduction to The Writings of Charles S. Peirce, Volume 1: 1857–1866, Fisch, M., C. Kloesel, E. Moore, and D. Roberts, (Eds), Bloomington: Indiana Univ. Press., pp. xv–xxxv. Jappy, Tony, (In press), Peirce’s Twenty-Eight Classes of Signs and the Philosophy of Representation, London: Bloomsbury Academic. Liszka, J., (1996), A General Introduction to the Semeiotic of Charles Sanders Peirce, Bloomington and Indianapolis: Indiana University Press. Murphey, M., (1993), The Development of Peirce’s Philosophy. Indianapolis and Cambridge: Hackett Publishing Co. Peirce, Charles S., (1931-1958), The Collected Papers of Charles Sanders Peirce, 8 vols. Charles Hartshorne, Paul Weiss and Arthur W. Burks, (Eds), Harvard University Press: Cambridge, Massachusetts. (CP) Peirce, Charles S., (1998), The Essential Peirce, vol. 2, Peirce Edition Project, (Eds), Bloomington: Indiana University Press. (EP2) Peirce, Charles. S. and Victoria Welby-Gregory, (1977), Semiotic and Significs: The Correspondence between C. S. Peirce and Victoria Lady Welby, Hardwick, C. S. (Ed.), Bloomington: Indiana University Press. (SS) Savan, David, (1988), An Introduction to C. S. Peirce’s Full System of Semeiotic, Toronto Semiotic Circle. Shapiro, Michael, (1983), The Sense of Grammar: Language as Semeiotic, Bloomington: Indiana University Press. Short, Thomas, (2007), Peirce's Theory of Signs, London and Cambridge: Cambridge University Press.