|Home Page Papers by Peirce Peirce-Related Papers
|This is an incompletely revised version of the original, which is being made available for purposes of critical feedback while the revision is still in process. It will in due time be submitted to the place of original publication as a revision to be substituted for the original whenever the occasion to update it arises. Paragraph numbers have been added to this version--bracketed, in the right margin--for purposes of scholarly reference to this version of it.
Charles Sanders Peirce was the second son of Sarah and Benjamin Peirce, the latter a professor of mathematics and astronomy at Harvard University, usually described as the leading American mathematician of his time. Unusually gifted by nature, trained rigorously in the development of his talents by his father, and and growing up in a household frequented by the scientific and literary intelligentsia who lived in or visited the college town of Cambridge, Massachusetts, Charles Peirce would seem to have been destined for an outstanding academic career at Harvard. But for reasons still not altogether clearinvolving such factors as an intemperate emotional disposition, a social climate within which his idiosyncrasies were found intolerable, the consequences of an unfortunate first marriage, and matters having to do with academic politics of the timehe was never able to acquire a secure academic position and worked chiefly as a physical scientist for the U. S. Coast and Geodetic Survey until 1891, after which he and his second wife, Juliette, lived in increasingly desperate poverty in a large house they called "Arisbe" near Milford, Pennsylvania which they had bought and moved into in better days (a few years before the termination of his position with the Coast Survey). He did, however, teach as a part-time Lecturer in Logic at Johns Hopkins University from 1879 through the spring of 1884, where he had a number of subsequently influential graduate students who regarded him highly, gave several series of lectures at Harvard University and elsewhere at various times across his career, maintained an extensive and wide-ranging correspondence with prominent figures in a variety of intellectual fields, and was better known and more widely published than is generally realized today.
Peirce was extraordinarily competent and accomplished in the sciences generally, in the manner of a true polymath--not a mere dilettante, he was solidly established as a working scientist--he nevertheless always regarded himself as primarily a logician, though in a far broader sense of the word "logic" than it commonly carried then or carries now. Indeed, it was his ambition from early on to ground and expand logic--deductive, inductive, and abductive (that is, the logic of hypothesis)--on a fundamentally new basis, namely, within the framework of a general theory of signs and representation (i.e. a semiotic), and he eventually came to the conclusion that logic in what he called "the broad sense" is that framework, the unified theory of deduction, induction, and abduction being best differentiated within it as "critical logic." Unfortunately, much of his best work on the general theory is still available only in manuscript form and copies thereof. This material is becoming increasingly available, though at a slow pace, as a new edition of his philosophical work (projected as some 30 volumes of 600 pages each) continues to be published across the coming years (five volumes of which have appeared thus far), and there is some hope that much of it can be made available more rapidly through the use of the new computer-based communications technologies. Peirce also left a substantial corpus of scientific work, including a book in astronomy, and a five volume collection of his work in mathematics, which contains much material of the first importance for semiotic, appeared in 1976.
In the following account of his semiotic, considerations pertaining to the chronological development of his thought will be ignored, with a few exceptions, and his ideas--regarded here as basically coherent--will be presented in a systematic though highly abbreviated way, the focus being on what is fundamental. No attempt will be made to reconcile here what others may take to be conflicting passages in his writings.
Peirce's semiotic can be regarded as a systematic development of what is implicit in his precise and highly abstract theoretical refinement of the vague pre-theoretical idea of a sign, as the colloquial word "sign" (considered as a noun) is defined in any comprehensive dictionary of the English language. A number of senses of the colloquial word can be distinguished, but one can nevertheless speak of the vague idea of a sign because in common to every sense is the idea of something which is capable of revealing something experientially--that is, as a phenomenon, fictional or real--to a being capable of responding to it (the sign) appropriately. In his theoretical refinement of this idea Peirce was guided, first, by his extensive knowledge of the antecedent theoretical usages of the word (and its rough equivalents in other Western languages, especially medieval Latin and classical Greek), and, second, by the aim of developing a highly abstract conception which would be basic in a generalized logic applicable to all processes and products which exhibit intelligence to any degree. Considered apart from its application to this or that special subject-matter or problem, then, the theoretical conception of a sign--or, more exactly, the conception of the triadic (that is, three-term) sign relation--is a highly abstract explication of the formal structure of intelligence, which Peirce himself regarded as co-extensive with life.
In application, this conception can become highly complex when, for example, distinctions are made between various types of signs; and it can function either descriptively or normatively, depending upon the specific situation and context of application and the aim of the inquiry for the purposes of which the conception is being used. (It is comparable in this respect to mathematics, considering the latter not as a calculus but as a formal science.) Within philosophy, semiotic would take the place of logic (as traditionally understood), epistemology, philosophy of mind, and so forth. In its extra-philosophical applications it would, he believed, be of use in any science which studies life processes or in any discipline which studies the products of intelligence as such (hence, in critical disciplines such as esthetic criticism, social criticism, etc.). Peirce even thought it might have some application within the physical sciences, namely, to processes which are non-conservative or irreversible (in the physicist's sense), though he realized that its actual scope of effective application was something to be discovered and demonstrated in practice, requiring interpretation by specialists into conceptions peculiar to a given field. It is best regarded as a theory, in the strict sense, only in respect to the question of what intelligence is; beyond that, it can be thought of as an analytic schema with such uses as attempts to apply it may or would prove it to have. Unlike European semiology, which it pre-dates by several decades (being datable from 1867), Peircean semiotic--which he sometimes spells "semeiotic"--is neither conceptually nor historically related to the special science of linguistics as a generalization or extrapolation therefrom.
In its theoretical refinement, this conception of the sign (the technical name for which is "representamen") retains the triadic relational structure implicit in the vague idea of a sign as described above, as that idea has been abstractively reduced in content to what Peirce regarded as minimally essential for the highly comprehensive range of application he envisioned for it. Thus a sign (sometimes called "representamen") is anything whatsoever which is capable of manifesting anything whatsoever in any respect (directly or indirectly), that which is manifested being called its "object"; and the response to the sign as such (that is, what we might ordinarily call an "interpretation" of its meaning or significance by an "interpreter") is thought of primarily in respect to its objective content rather than as an act, this possible objective content--the technical name for which is "interpretant"--being the third term of the basic sign relation. The technical name for the triadic sign-object-interpretant relation is "representation". The sign is said variously to "refer to," "stand for," "denote," or "represent" its object, depending upon which word is most appropriate in a given context of application; and it is said to "signify" or "mean" its interpretant.
There are also formulations of the relation in which the sign is said to be "determined by" its object and to "determine" its interpretant. It is important, though, that Peirce's conception of the generic sign relation not be understood primarily through one's prior understanding of such familiar words as "refers to", "represents", "denotes", "stands for", "signifies", "means", "determines", and so forth, but rather by noting, as we have, the pre-theoretical idea from which it was abstractively derived, and then by understanding that he regards the triadic sign relation as the most basic form or structural principle of a dynamic process, each moment of which he calls "semiosis."
The technical word "semiosis" refers to the action of a sign in generating or producing an interpretant of itself. An interpreter's act of interpreting can be regarded as a special case of an interpretant (though as so regarded that act is implicitly thought of as being the objective content of another interpreting act), but the causation is nevertheless primarily located in the sign-action itself, and in the sequential process of sign-action (that is, interpretant-production), in a way to be explained later. As he conceives it, any given sign is itself necessarily an interpretant of a prior sign of the same object (though we do not necessarily know what that prior sign is), which implies that the object of the sign is, in any given case, distanced from it by a logically intermediating sign. There is a sense, then, in which the object of a sign is always infinitely remote from it, though this apparently self-defeating conception of the semiotic object is not really such because--as we shall see--the object and a sign of it can also be identical, in a certain sense.
As regards the interpretant of a sign, it is always also a sign capable of generating a further interpretant, and so on ad infinitum, so that, paralleling the infinite logical remoteness of the object (which is the ultimate generating ancestor of all signs of itself), there is the infinite logical remoteness of the final interpretant. But here, too, an apparently self-defeating conception is not really such because of the special sense the word "final" bears in this context and because of the possibility of an identity in content of an actual interpretant with the final interpretant. (In passages in which Peirce speaks of the interpretational process as terminating in something which is not a sign he is to be understood as using the word "sign" in its colloquial sense, not as denying that all interpretants are themselves signs in the technical sense, that is, representamens.) There is, of course, such a thing as misinterpretation, but there can be no such thing as a misinterpretant. (In a misinterpretation an interpretant is being thought of in relation to the wrong sign and/or wrong object.)
Supposing it were ever realized, the complete composite final interpretant of all signs of a given object would be nothing other than the object itself, considered as fully manifested: that is, an object, all of its signs, and all interpretants of its signs would be materially indistinguishable, such that the semiosis process could be thought of as being the object itself in its protracted manifestation or self-actualization in time. However, since no object ever has or ever could fully manifest itself, or at least we can never be in the position, in practice, of supposing that it has done so, a working distinction must always be maintained between the object as such, its signs as such, and the interpretants of its sign as such. Even this, however, does not preclude the possibility of also recognizing, at the same time, a material identity (at least in part) of sign and/or object and/or interpretant at a given time. In other words, it is possible to recognize one and the same entity as playing two or even all three semiotic roles at once.
The clue to understanding the apparently puzzling infinities involved in Peirce's conception is that they pertain to the object-sign-interpretant relation considered in the abstract, whereas in the concrete application of semiotic conceptions there will always be legitimating reasons for regarding certain signs as the first signs of the object (for the purpose at hand) and certain interpretants as the last interpretants (again, for the purpose at hand, but not absolutely). The infinities involved in the theoretical conception function in practice as a theoretical reminder and a logical guarantee that there can never be an absolutely definitive and minute semiotic analysis or account of anything, but only an analysis or account which satisfies whatever intellectual motive instigates the concrete inquiry or specific line of inquiry. The key to understanding how something can be regarded simultaneously as one and many is that the conceptions of identity and difference cannot be applied to any subject-matter apart from some concrete analytic situation which implicitly provides working criteria of justification for their application and the manner of it. Just as in the case of arithmetical conceptions, there is no way of saying how many things there are unless one has first specified the type of thing to be counted, and, besides that, assumed (implicitly or explicitly) that what we might ordinarily count as one thing extending across a given period of time should be counted as one or many, as the case may be, for the purposes of the inquiry in progress. Moreover, the situationally implicit criteria for material identity are, in general, distinct from the analytical considerations which make it desirable or necessary to identify and distinguish that which is functioning semiotically as sign, as object, and as interpretant.
Consider, for example, any piece of paper upon which the first word of this sentence is inscribed. That piece of paper could be regarded as being one piece of paper during, say, the length of time it takes you to read this sentence, or it could be regarded as a succession of several numerically distinct pieces of paper during that time (if there were a concrete reason for so regarding the matter), nor would there be any contradiction involved in saying at once--in a semiotic analysis--that it is one and many in the same respect during that time; for this would simply mean that a double relativization was being taken for granted: a relativization to some material criterion of identity and a relativization to the needs motivating the semiotic analysis. In other words, for analytic purposes, we can conceptually divide any entity into as many different entities as necessary in order to maintain the object-sign-interpretant distinction without thereby denying that the object, the sign, and the interpretant are one and the same otherwise: in regarding it as a sign, and also as the object of that sign, and also as an interpretant of that sign our justification would be the need to do so in order to analyze it semiotically; in regarding it also as one and only one enduring thing our justification would be that there was no practical need to distinguish it into three apart from the purpose of analyzing it semiotically. Finally, in the case of one type of sign in particular (the iconic sign, which will be discussed later), the very nature of the sign-object relation is such that they can be regarded as at once one and many--and indeed must be in accounting for cases of self-representation.
Considered dynamically, the representation relation is the basic form--repetitively occurring and overlapping--of a process of semiosis without any determinable absolute beginning or end. This can be likened intuitively to a chain interminable in both directions, starting from any given point within it, though conceivable from without as originating in an ultimate ancestral object and as terminating in an ideal final interpretant. That is, one might devise a method of representation of a semiosis process by beginning with such a picture or--as Peirce would say--"icon" of the semiosis process. This would then have to be complicated by taking into account the fact that (a) any given "link" in the chain might represent a sign (which, bear in mind, is always also an interpretant) having two or more sign values, which would have to be represented by a corresponding number of branchings at that point, (b) any two or more given chains might merge into a single chain through an interpretant-sign which is a resultant of signs in different chains that have cooperated conjointly in producing that interpretant (much as premises conjoin in producing a conclusion), (c) any two or more given chains might originate ultimately from the same ancestral object (from which any number of chains may emanate), and (d) the individual "links" in the chain are always potentially analyzable into innumerably many possible sub-links and sub-sub-links (within sub-chains and sub-sub-chains) since a sign is not a "logical atom" but only something which it happens to be intellectually profitable to treat as a unit at a certain point in an analysis, though always analyzable in principle into sub-signs if or when the analytic project should require it, any of which may itself be a sub-sign within some more inclusive sign. Moreover, a sign may be inclusive both of what we would identify as the text and the context, or, on the other hand, it might be more profitable intellectually to retain the distinction between the two in a given analysis.
Peirce himself provides no general method of representation of semiosis processes of the sort I am suggesting, though in his work in formal logic he did develop an ingenious graphical technique for representing deductive semiosis in particular--called "existential graphs"--in which he experimented with three-dimensional techniques. Perhaps his graphical technique, or something like it, could be developed and modified in such a way as to provide a workable method of representation; perhaps it could not. If so, then the chain image would not be the appropriate one to work with. However that may be, though, the purpose in explaining Peirce's semiotic here in terms of a possible but presently non-existent method of representation is not to point out a necessity of developing such a notational device--which is not required for many uses of his ideas, at any rate--but only that of conveying in a brief space an intuitive way of understanding semiosis, as he conceived it, which may at least tend to discourage certain persistent misinterpretations of his conceptions.)
This chain-like image of the semiosis process must be complicated in still another way, though, which might require recourse to something analogous to a three-dimensional method of representation, and perhaps even to dimensions beyond that. The further complications arise from the fact that logical levels are implicit in the structure of the semiosis process in virtue of the way in which the basic representation relation is to be understood. These implicit logical level differences need not be recognized explicitly merely because they are present implicitly: in fact, they cannot be made completely explicit since there are an infinite number of such levels implicit in any semiosis chain, and they should not be made explicit except insofar as there is some concrete analytic need to do so since it is obviously desirable not to complicate one's analysis beyond necessity. There are, on the other hand, any number of possible reasons why it might be desirable or necessary to make some level differences explicit in a given analysis.
To see that and how these logical levels are involved, consider the following two characterizations of the representation relation, the first of which (from a 1902 publication) focuses on the relation primarily from the perspective of the sign-term and the second of which (from an 1867 publication) focuses on it primarily from the perspective of the interpretant-term.
The two characterizations differ verbally in several ways. For instance, the first characterizes representation in a more dynamical way than the second through its use of the word "determines", which carries a more obvious causal import than the word "represents". In fact, though, they say substantially the same thing (in a Peircean context) since "x determines y" (or "y is determined by x") is equivalent to "y refers to x" or "y represents x" or "y stands for x", given suitable accompanying changes in the grammatical structure of the sentences in which they occur. Thus to emphasize still more strongly that representation, construed dynamically, is the form of a semiosis process, characterization (1) can be reworded as a tautological statement as follows:
On the other hand, it can also be reworded in terms of reference, rather than determinism, in order to emphasize the logical aspect of representation:
Turning to characterization (2) above, it can also be reworded (as a tautological statement) in terms of reference:
(1R) and (2R) appear at first to differ substantially since the word "because" is used in the one and not in the other, but the difference is only an apparent one, given that we are understanding (2R) as tautological of what it is to be an interpretant. Also, given the context in which (2) was formulated by Peirce, we can regard the import of the phrase "in the same way", which occurs explicitly in (1), (1D), and (1R), as implicit in (2R).
One sign is at a (relatively) higher logical level than another if it refers to the other as a sign. (There is no absolutely lowest or highest logical level because there is no absolutely definitive semiotic analysis of anything.) Hence, as these rewordings of Peirce's characterizations of the representation relation should make clear, every interpretant-sign is at a higher logical level than the sign, S1, of which it is an interpretant. Moreover, the interpretant also always refers to the object of S1 independently of its reference to S1 as sign of that object, and in this second reference it is not at a higher logical level than S1. (There is no contradiction implicit in this since logical level differences are entirely relative to the respect in which something is being regarded semiotically.) However, in this second reference it necessarily refers to the object via reference to a second sign, S2, of the object (since there is no such thing as an unmediated reference to an object), and it is thus also an interpretant-sign relative to it. Now, S1 and S1's interpretant are either both interpretants of S2, or else both are interpretants of signs which have an identical antecedent at some point in their respective chains. This is a part of the import of the phrase "in the same way" as it occurs in characterization (1) and the reworded versions of it above. (Another part of its import is that when the semiosis chain is regarded dynamically, as in (1D), we are to understand that the interpretant inherits the generative power of the sign that generates it.)
Stated in the abstract, the meaning of this is no doubt obscure (though it does indicate, in a minimal way, how logical level differences are always implicit in the representation relation), so let us consider a concrete example. Take, for instance, the essay you are presently reading and regard the whole of it as a single sign. To simplify, assume that there is no doubt that what is being said here about Peirce's semiotic is true of it, and also assume that there is no doubt that you correctly understand what is being said here. This essay (that is, what is being said here) is, then, a sign (S1) whose object is Peirce's semiotic, and your understanding of this essay--not your act of understanding but rather its objective content--is an interpretant of this sign (and is therefore itself a further sign of Peirce's semiotic). Your understanding of this essay is at a higher logical level than the essay itself since it refers to this sign, and as its interpretant it refers to it as being a sign. Since this essay is understood as being a sign, a distinction is implicitly recognized between what is being said here (S1) and what is being talked about, namely, Peirce's semiotic, which implies some understanding of Peirce's semiotic--however little or great it may be--which is independent of what is being said about it here. (It could amount to nothing more than that there was someone called "Peirce" who had some sort of theory or idea called "semiotic.") For were there no independent understanding which could function as a second sign (S2) of what the subject-matter is, what is being said about it (S1) could not be referred to it as true of it: this essay would be like a (highly complex) predicate without a subject. In other words, it is by means of S2 that the subject-matter--the object--is identified. Given our initial suppositions, then, it follows that either this essay (S1) and your understanding of it (S1's interpretant) are both interpretants of S2 (which is to say that we both identify the subject-matter in exactly the same way, independently of what is said about it here), or else what is said here and your understanding of it are both interpretants of signs which have an identical antecedent at some point in their respective chains. To simplify, let us suppose that we would agree in identifying the object through S2.
Leaving aside the complications that would arise in the example if the simplifying assumptions were dropped (and ignoring certain other implicit complications even given these assumptions), the minimally essential factors which must be distinguished in order to understand semiotically the understanding of this text (that is, to understand what it means to regard this text as a sign) are as follows:
Thus we have (A) the sign of the object, (B) the object proper, and (C) the interpretant of the sign of the object. But we have isolated a fourth essential factor (D) in the semiosis process as well, namely, the subject-identifying sign, S2. Peirce calls this the "immediate object," as distinct from the object proper, which he calls the "dynamical object." (He sometimes calls the object proper the "real object," but as he realized himself, "dynamical object" is better inasmuch as the intended contrast is not between the real and the fictional--for the object proper can be either--but rather between the object as it really is in itself, as distinct from the object as it is represented to be--perhaps not altogether correctly, and certainly not completely--and taken for granted as being.) This fourth factor of the semiosis process--the immediate object, S2--is not, however, a fourth term of the representation relation, which is always only triadic, but is rather a sign of which a given sign as a sign is an interpretant, functioning logically to identify the object independently of how it is characterized by the given sign (or how it is understood as being characterized by that sign). In doing so it constitutes the point of logical overlap when the representation relation is recognized as structuring the chain of semiosis, the links of the chain being, as it were, "welded" together by it.
Note, though, that even in this highly simplified example, it is clear that a given interpretant is always generated or produced by at least two signs (each of which could be internally very complex)--the subject-identifying sign (S2) and the information-supplying sign (S1)--and is thus always a sort of synthesis, much as a simple subject-predicate sentence is a synthesis of what is already known or taken for granted about a certain subject-matter (provided by the subject-term) with what is added to that by what is said (the predicate term). Its conjoint generation by S1 and the immediate object (S2) is the result of a semiotic copulation, analogous to the logical copula which synthesizes the subject and predicate of a sentence. (The biological and erotic suggestions implicit in the word "copulation" are not to be ignored, by the way, when or if one is trying to apply the semiotic schema to subject-matters to which such ideas may be pertinent.) It is this synthesizing function of the interpretant which explains why Peirce speaks of it as a mediating representation in his 1867 formulation quoted earlier: in its occurrence it functions as a third thing synthesizing what he there called the "relate" (i.e. sign) and the "correlate" (i.e. the object as represented: the immediate object, the object in its immediacy). From a causal point of view, though, it would be the sign which would be mediating, namely, between the object and the interpretant (via the immediate object, which mediates between the dynamical object and the sign).
Since every sign is an interpretant and every interpretant a sign (the distinction between sign and interpretant being analytically relative), the diagrammatic representation of a relatively simple semiosis process obviously could be quite complex, even in cases where there was no need to represent logical level differences within it. In the example given, the simplifying assumptions were introduced in order to minimize that need (as well as other complications), since the representation of possible disagreement or error or misunderstanding would necessitate the representation of still further logical level differences. In general, there is no need to represent such differences insofar as one is only tracing a semiosis chain as a causal chain; but since the ideas of disagreement, error, misunderstanding, and the like, imply deviation (possible or actual) from some norm, any representation of semiosis considered normatively will necessarily involve the representation of level differences (since references of signs to signs as signs will have to be represented). As remarked above, the idea of a graphical mode of representation of semiosis is used here only as an expository aid, rather than because such a notational device is in fact needed, since verbal descriptions of matters using his technical terminology will often suffice. However, it could be argued--and Peirce would himself argue--that a verbal description of a semiosis process always does appeal implicitly to a graphical schematization of some sort (perhaps only "in the imagination") for its intelligibility, anyway, and to that we may add that not only would such a method of representation be a practical necessity in analyzing cases that are too complex to be understood with the use of words alone, it would also be in the attempt to develop such a method of representation that one would discover both powers and defects in Peirce's semiotic conceptions that might otherwise go undetected.
In addition to distinguishing between the dynamical and the immediate object, Peirce also distinguished between what he called the "immediate interpretant," the "dynamical interpretant," and the "final interpretant." The meaning of this distinction is in dispute, but it is perhaps best understood as follows. The dynamical interpretant is simply the actually occurring interpretant. The immediate interpretant is constituted by the range of possible interpretants of a given sign at a given time, and is thus its identity as the particular sign it is, insofar as that is (in principle) an establishable fact at that time. But this identity at a given time is, on Peirce's view, always more or less vague, which is to say that the limits of the generative power of the sign as such cannot be exactly determined since it is in itself--and not merely in relation to our possible knowledge--somewhat indeterminate. One might think of this in analogy with the way in which, say, a person might be indeterminate at a given time as regards a power or ability of doing something which he or she might definitely have or not have the ability to do at another time. For example, a novice billiard player might successfully make an extraordinarily difficult shot, such that one cannot say that it was done "by accident" since it was precisely the shot the player attempted to make, even though it might be many years later before one could say of him "He can make that sort of shot," because by then he had developed and demonstrated the ability to make such shots successfully with some regularity. During the questionable period of time when the player cannot definitely be said to have or not to have the ability, there is no fact about him or her which would contradict either the imputation or denial of possession of it: the being of the player is, as it were, vague or indeterminate in that respect during that time. A sign can be like that, and thus from the point of view of an interpreter it is not simply a matter of not knowing what the sign does or can signify, but of there being an objective indeterminacy in the sign itself in the relevant respect. The immediate interpretant is, then, the range--always vaguely circumscribed--of the interpretant-generating power of the sign at a given time.
The final interpretant, on the other hand, is the range of possible interpretants as that would be definitively established with the ultimate cessation of all growth in the powers of the sign as such. It is not the last interpretant that will occur, but rather the totality of all the sign-powers a given sign would have manifested when it had shown all that it could be--all that it could do--as a sign. For example, several decades ago a scholarly institute attempted to give a reasonably comprehensive account of the meaning of the word "freedom" (and its equivalents in other languages), particularly as it has occurred in philosophical discourse across the centuries. Some of these meanings may rarely occur in any contemporary discourse other than in this compendious volume itself; but then the very fact that they occur in the compendium shows that the sign still has such powers (though perhaps greatly enfeebled). Suppose that such an account were compiled and fully completed, without error and without restriction to philosophical discourse, at a time after which there never would in fact be any further changes in the sign's generative power. At that time--which would not, of course, be the end of time--the immediate and the final interpretant would be indistinguishable in content, though the conceptual distinction would not thereby collapse.
The application of the distinction between the immediate, dynamical, and final interpretant can be illustrated by the role of precedent cases in law (that is, in legal semiosis), and the reader can perhaps extrapolate by analogy from this to its applications in other fields. There are occasions when the dynamical interpretant--that is, the actually occurring interpretant--of a sign which is the law is not definitely identifiable because the law is too vague in the relevant respect: the facts of the case may be clear enough but the meaning of the law is not, and the judge must, as we say, exercise real judgment in the matter (which is to say that the judge must recognize something as being the relevant dynamical interpretant without benefit of recourse to any ascertainable basis in the immediate interpretant that would justify that recognition). The conscientious judge makes a guess, in effect, at what the final interpretant includes when he or she recognizes something as being a dynamical interpretant of that law at that time relative to that case. But it is the course of future legal interpretation of that law (in courts of appeal, in future juridical practice, and so on) which will determine whether the judge was or was not right in his or her attempt to anticipate the relevant content of the final interpretant--or, as we would ordinarily say, in the attempt to set a precedent that will be honored.
Peirce also distinguished the "emotional interpretant," the "energetic interpretant," and the "logical interpretant" and there is dispute on this as well, both as regards precisely what the distinction is and how it relates to the distinction just explained. Since it does not appear to be on a logical footing with the distinction just discussed nothing will be said about it until later.
Before proceeding to an account of Peirce's distinctions of types of signs, the nature of the causation involved in semiosis must be briefly explained. In some characterizations of the representation relation, Peirce speaks of the object "determining" the sign and the sign "determining" the interpretant. Now, the word "determines" and its cognates carry at once a logical and a causal sense, for Peirce. In its logical sense, it means that whatever the sign refers to must be referred to by its interpretant, which is a generalization of the idea that whatever the subject-term of a proposition refers to must be referred to by its predicate--which is, in turn, a way of defining the truth of a proposition. (This is why there can be no such thing as a misinterpretant, as distinct from a misinterpretation.) In its causal sense, it means that the sign causes or produces or generates its interpretant, such that the latter in turn causes or produces an interpretant, etc., and such that, in the course of such a causal chain, there is a real tendency for the object to manifest itself. (Bear in mind that what immediately determines a given sign is the immediate object, which is itself a sign of which the given sign is an interpretant. The dynamical object--the object itself--is always itself a sign, too, though this is not to say that it is nothing but a sign: indeed, no sign is ever nothing but a sign since, in Peirce's view, it could not stand in the triadic representation relation if it did not also stand in a dyadic relationship, which in turn presupposes that it bears a monadic property. Conversely, Peirce held that nothing whatever can so much as be thought of by us which does not stand in some triadic or representational relationship: these three types of properties are what he calls "the universal categories," following the Aristotelian and Kantian traditions.)
Now, something can tend toward something without ever actually attaining that towards which it tends, and wherever there is a real tendency there is what Peirce calls, indifferently, "final causation" or, "teleology". Semiosis processes are thus telic in overall form, and it will have to suffice here to say that Peirce's basic explication of teleology or tendency to an end, as he also and most commonly refers to it, is through the conception of probability: the laws of chance. He also works with conceptions of the sort identified sense his time as being "cybernetic" in character, involving feed-back correction loops, but he does not attempt to reduce teleology to this type of mechanism but instead treats such mechanisms as function to augment a primitive directionality or tendency to an end which is inherent in probability principles. A tendency, though, is a characteristic which is discernable only in an active relationship involving more than two terms, and it is presupposed that each generation of an interpretant by a sign is also a matter of what might best be called "brute-force" causation, but which he himself calls "efficient causation", which involves the sort of law which woujld be expressed by an unqualified universal proposition ("Whenever A then B") is regarded by Peirce as a limiting case of what could be expressed by something like "Given A, there is a tendency toward B", that is, a case where the tendency is completely rigid, as it were, and to say that A caused B in the brute-force sense does not involve implicit reference to any such law. Thus semiotic causation is itself telic or final causational, and it presupposes but cannot be explicated in terms of brute-force causation. [Still not coherent: keep rewriting.]
Peirce distinguished ten basic trichotomic sets of distinctions of types of signs by applying a pre-semiotical distinction of three modes of being to the generic representation relation in various respects. By cross-combining these one arrives at the mathematical possibility of 59,049 (= 310) kinds of signs; but most of these combinations would be conceptually incoherent, and the substantively possible combinations come to only 66, according to Peirce. Three of these basic distinction-sets are far more firmly established theoretically than the others since they are generated on the basis of the simple sign-object-interpretant conception only and are exemplified and discussed by Peirce extensively in his writings, whereas the other seven are based on differentiating the object into immediate and dynamical and the interpretant into immediate, dynamical, and final, and Peirce did not develop these distinctions or attempt to articulate the further distinction-sets they make possible until relatively late (and then only in an obscure, sketchy, and obviously tentative way). Although we cannot say in abstracto that any of the ten sets is more important than the others, we will restrict ourselves here to the three best-established trichotomies.
Peirce believed he had demonstrated that there are three basically distinct ways of regarding phenomenal entities: (1) only insofar as they bear monadic (that is, non-relational) properties; (2) insofar as they are terms in dyadic (two-term) relationships, which presupposes that they bear monadic properties, and (3) insofar as they are terms in triadic (three-term) relationships, which presupposes that they stand in dyadic relationships. All relationships of apparently greater complexity are supposedly analyzable into combinations of triadic and/or dyadic and/or monadic forms, though overt linguistic form is an unreliable guide to whether a given property is monadic, dyadic, or triadic. Thus "Jones is a father" is, linguistically, a monadic predication, but of course it implicitly involves a more complex predicate which can be analyzed one way and another depending upon whether it is biological or/and legal parenthood which is meant, and also depending upon the interests situationally motivating the analysis (since nothing can dictate a priori exactly what is to be taken into account and what ignored as unimportant in a concrete analysis). Terminologically, Peirce speaks of something "being" insofar as it has monadic properties, as "existing" and "occurring"--though not necessarily in what we commonsensically identify as the "real world"--insofar as it stands in dyadic relationships, and as "real" insofar as it stands in triadic relationships (though, again, not necessarily in the commonsensical "real world"). His justification cannot be explained here, but he regarded all irreducibly triadic relations as being or involving representation relations and all irreducibly dyadic relations as being or involving interactional occurrences: that is, this is what constitutes their irreducible triadicity and dyadicity, respectively. Hence, to regard an entity which is a sign as standing in a triadic relationship is to regard it as a sign; to regard it as standing in a dyadic relationship is to regard it as something occurring; and to regard it as bearing a monadic property is simply to regard it as being what it appears to be (that is, to regard it simply as an appearance: that which is manifest when something is manifested).
It will be helpful in what follows to understand that neither a monadic nor a relational property is regarded by Peirce as being independent of an entity or entities which bear(s) the property; that is, properties are always embodied. However, insofar as the analytic emphasis is on the property rather than its embodiment, the latter is left otherwise unidentified. Hence, the generality of a property consists in the fact of its being embodiable in an entity or entities capable of being otherwise identified with any number of entities, as distinct from its being independent of embodiment in such a way as to require linkage of itself to embodying entities. Moreover, a relation is an action (an active relating) of its terms, regarded with special reference to some one of its terms. This way of understanding relations is exemplified, by the way, in Peirce's characterizations of the representation relation in terms of what it is to be a sign (as in the one quoted version given earlier), and also in terms of what it is to be an interpretant (in the other quoted version). Hence, triadic relations in particular, which can be thought of as being of the nature of rules, laws, customs, habits, etc., are active properties of their embodying entities, so that one can think of the rule by thinking of an instance of its embodiment as such.
The first basic division of signs is into: (1) qualisigns; that is, signs considered as such but just insofar as it is the appearances they present--which he sometimes speaks of as "qualities"--which are pertinent to their value as signs; (2) sinsigns; that is, signs considered as such but just insofar as it is their occurrence or existence which is pertinent to their value as signs; and (3) legisigns; that is, signs considered as such and only as such, which is to say, as terms of a representation relation. In other words (and in reverse order), given something which is ex hypothesi a sign, and which is being regarded as such, it may be considered only as such, or it may be considered with special regard for the fact of its occurrence, or it may be considered with special regard for its appearance.
For example, the word "the", considered as a sign, occurs many times in this copy of this essay (and there are therefore a corresponding number of sinsigns of it); in its occurrences here it has only one appearance, in the sense of visual form (I am ignoring, for simplicity's sake, the difference between its capitalized and uncapitalized forms), and thus there is only one qualisign of it in this essay; but if we are thinking of it as the definite article--that is, as the sign whose object is the relation we have in mind when we think of how something functions grammatically as a definite article--then it is a legisign which is "replicated" by a number of occurrences outside of this paper as well, and with numerous different appearances (such as "Das", "Der", "Die", "la", "le", "el", "il", and so on, to mention only a few of its visual appearances, though, of course, replicas of it also occur acoustically, tactually, and so on). The word "replica" is a technical word of Peirce's used to refer to a sinsign considered as the actualization of a legisign, its replicas perhaps exhibiting many different qualisigns pertinent to them as its replicas. It should be noted, though, that the particular perceptible form or quality of a sinsign is never either a necessary or a sufficient condition of its being a replica of this or that legisign in particular. For we often have to disregard a sign's appearance in identifying what sign it is, as in the case of misprints, scribbles, mumbles, etc., and the functioning of the definite article, for example, may involve no qualisign, as in cases where this grammatical function is performed by the fact that it must be assumed in order to make sense of the discourse under analysis: this fact might, perhaps, be regardable as the sinsign since it is something that occurs, though the sinsign in that case could hardly be said to be such that it is regardable in its qualitative aspect, that is, as exhibiting a qualisign. (Peirce used several different terminologies for his technical conceptions, and the most used alternative for this trichotomy is "tone"/"token"/"type". However, with a few possible exceptions, the use of "token" and "type" by later logicians does not draw the same distinction he was drawing with the words, even though it has become customary to acknowledge Peirce as the originator of the distinction, nor would it be correct to say that the distinction between a legisign (or type) and its replicas is the relevant one either; for later logicians commonly have worked only with a dichotomic distinction, whereas Peirce's basic distinction is a trichotomy, and, moreover, presupposes importantly different systemic connections with other ideas such as cannot be attributed to these other logicians.
The second basic division of signs concerns the different sorts of facts about the sign, considered in relation to its object, which can ground it as a sign of its object: not what makes it a sign but rather what makes it the kind of sign it is, relative to its object. (If we distinguish between the object as immediate and as dynamical, it is the relation to the dynamical object which is intended here even though verification of the grounding would have to be via the immediate object.) Now, if the grounding is in virtue of the fact that the sign and its object have a property in common--that is, are alike, similar to, or resemble one another in some respect sufficiently for ignoring any differences between them in that respect--then the sign is an icon if it is a qualisign, or it is said to be "iconic" if it is a sinsign or a legisign. But if the grounding is in virtue of the fact that the sign and its object stand in a dyadic (that is, existential) relationship to one another, then it is an index if it is a sinsign, or it is said to be "indexical" if it is a legisign. (An icon proper can neither be an index nor be indexical.) Finally, if the grounding lies only in the fact that the sign will be represented in its interpretant as a sign of that object, regardless of any likeness or existential relation which it may have to that object, then the sign is a symbol and is necessarily a legisign. (The symbol could also be defined in terms of the idea of rule, habit, custom, law, convention, habitude, habitus, hexis, disposition, etc., but the present way of defining it is better, for reasons to be given below.)
Strictly speaking, there are no "symbolic" signs (as distinct from symbols proper) but rather only "symbol replicas," which are special cases of indices (indexes). However, no confusion is likely to result from using "symbolic sign" in place of "symbol replica"; and it would also be unobjectionable to call a qualisign which is embodied in a symbol replica a "symbolic sign," too, provided the qualisign was thought of as functioning in the identification of the sinsign as a replica of a certain symbol, were it not for a certain persisting confusion in semiotical writings between iconicity and symbolism which should not be encouraged (as, for example, when something obviously imitative of what it refers to is said to be a symbol of it on that account). Given this confusion--which arises in large measure simply from the fact that there are two distinct and well-established traditions of usage of the word "symbol", but which has tended to induce confusions that are more than merely verbal--it is probably best not to refer to qualisigns (which are always icons) as "symbolic signs."
In summary to this point, then, we have:
(1) qualisigns (which are always icons);
(2) sinsigns, which are either
(a) indexical, including symbol replicas, or
(b) iconic; and
(3) legisigns, which are either
(a) symbolic, or
(b) indexical, or
Several explanatory comments are necessary before proceeding. First, it should be noted that, since an icon and its object (and an iconic sign and its object) do not differ in the respect in virtue of which the one is an icon or is iconic of the other (or do not differ enough to make any difference), and since it is even possible for them not to differ in any respect, nothing precludes the possibility of a material identity (partial or total) of sign and object. This allows for the possibility of using the conception of self-representation if that is desirable or necessary, while at the same time distinguishing between the entity as sign and as object. Similar considerations apply to a sign and its interpretant, mutatis mutandis, and thus a material identity of sign, object, and interpretant can be recognized within a semiotic analysis. This is obviously a point of the first importance relative to the use of semiotic analysis, since it gives it great flexibility in application.
Second, the common property of sign and object can be monadic, dyadic, or triadic, which is why even legisigns can be iconic.
Third, the existential relation of sign and object in the case of the index is to be thought of as an interactional co-occurrence, including, of course, the sort of interaction we ordinarily have in mind when we speak of something as ("efficient") cause of the other, though not restricted to cases in which cause and effect occur successively (so that it sometimes make no difference which is taken as cause and which as effect). In any case, it is always only a singular instance of object-sign interaction which grounds the sign as index, with no implicit reference to a generalization about "constant conjunction" playing an essential role in the grounding, though, of course, such generalizations may be resorted to collaterally as an aid in ascertaining whether an object-sign interaction is involved.
Fourth, iconicity and/or indexicality, as conceived by Peirce, cannot be subsumed under symbolism as special cases thereof: the three are defined as coordinate. It is true that the presuppositional ordering of the triadic, dyadic,and monadic will introduce subordinative structuring in various ways whenever Peirce's schema is applied in practice, but there is no way of saying a priori and in general what will be subordinate to what (and in what way and sense), nor is there any reason to suppose that symbolism will in fact turn out to be the superordinating conception more often than others, given the intended scope of application of semiotic. This is said in view of current tendencies in semiotic theorizing to rely so heavily on the conception of a rule (or "code") that it may seem reasonable to regard iconicity and indexicality as special types of symbolism, given that symbolism is understood primarily in terms of rule or rule-following and given certain arguments to the effect that both likeness and "efficient" causation are implicitly constituted by conformity to rules. But we must distinguish between that which constitutes something as a sign and the sort of fact which constitutes the particular sign identity of something. The former is a matter of the overall telic form of the process of which it is a part, whereas the latter is a matter of what sort of thing would have to be appealed to in identifying a sign in its particularity, which is radically different in the three cases. Regardless of what constitutes a given sign, S, as being like something, O, be it conformity to some rule of likeness or not, it is the implicit appeal to the fact of likeness that differentiates it as an icon or as iconic; it is the implicit appeal to the quite different sort of fact--however constituted--of the existential interaction of S and O that differentiates it as an index or as indexical; and it is appeal to the still different sort of fact that S will generate a certain interpretant (which can be construed as a fact about its final interpretant, as this concept was explained earlier) which differentiates it as symbol of O. Now, neither an appeal to a likeness nor an appeal to an existential relationship--which would be appeals to non-semiotic properties--can be construed as a special case of an appeal to the interpretant of a sign: neither the likeness of S to O nor any existential relationship of the two implies or presupposes anything about S's sign-properties except that it has met necessary conditions for being an icon or an index, respectively.
Fifth, it is not only possible for the same sign to be at once a symbol, an index, and an icon (since it can be regarded as being at once a legisign, a sinsign, and a qualisign), it is even possible for it to refer to the same object in all three ways at once. Relative to human interests, though, one cannot say in general that it is either desirable or undesirable that it should do so. On the one hand, the respective and importantly different roles of the icon, index, and symbol might tend to be confused thereby. The role of the icon as such is that of presenting the properties of the object immediately by embodying them in itself, and therefore should be as complex as interest in the relevant properties of the object requires. The index as such functions in a contrary way, though, since its role is primarily that of directing attention away from itself and toward the object, but without reference to the latter in any respect other than as is required to identify it as the object (which already goes beyond its pure function as index). The role of the symbol, then, can only be that of synthesis: since the properties of its replicas do not, as such, have either an exhibitive or a deictic function, their role is that of bringing together signs that exhibit or show and signs that point out the things about which something is shown. (The symbol is thus what the ancient Greeks called a "logos," the primary meaning of which was that of a presentational gathering.) On the other hand, there are doubtless some important purposes for which it would be desirable that the different functions be as unobtrusive in their difference as possible.
Sixth, the formulation of the symbol in terms of how it will be interpreted is preferable (at least in a basic account of Peirce's ideas) to a formulation in terms of "rule", etc., for several reasons. (1) It discourages the tendency to confuse the idea of the symbol as being a rule (which is the correct idea, as far as it goes) with the idea of it as something which conforms to a rule: a confusion which persistently plagues semiotic theorizing. (2) Some philosophers--Peirce and Ludwig Wittgenstein most notably--have recognized that the conception of a rule, law, convention, etc., is itself so highly problematic that recourse to it really explains nothing, though it seems to do so. Thus the need for a deeper understanding is obscured at precisely the point where it should be made obvious. (3) The words "rule" and "convention" have become strongly associated in this century with ideas fundamentally antithetical to Peirce's way of regarding things. This is particularly true of the ideas now usually associated with the word "conventionalism", which are inappropriate in understanding Peirce's idea of the symbol. The formulation in terms of how something will be interpreted, properly understood, precludes construing even the symbol as being conventional in the sense of being a meaning establishable by any person or group of persons at any time simply in virtue of a will or decision to so at that time. For if it is borne in mind that the interpretant of a sign is itself a sign the significance of which is a matter of its interpretant, and so on, ad infinitum, it will be obvious that how a sign will be represented in its interpretant is never definitively established by any given dynamic (actual) interpretant, and the identity of a symbol, in particular, is thus essentially futuristic in somewhat the same sense in which it is true that "tomorrow never comes." Of course, there is also a sense in which tomorrow does come, and, analogously, a symbol's identity is not something that never is achieved nor is it such that we cannot reasonably surmise what its identity is. But symbols--like all signs--do have a life of their own, so to speak, and the juridical example given earlier, in connection with the distinction between the immediate, dynamical, and final interpretant, may be helpful in understanding the sense in which this is so.
The third basic division of signs concerns the sign in relation to its interpretant. (If the interpretant is thought of as differentiated into immediate, dynamical, and final then this distinction concerns its relation to the final interpretant.) The deciding question here concerns the extent to which the sign limits what its actual interpretant can be. (1) Although the sign always determines the interpretant to a reference to its own object, it may include within itself no sign of what that object is, independently of what it manifests about it, so that the interpretant can refer to anything to which the sign can refer: in this case it is a rheme. This is to say that the immediate object--the second means through which the interpretant refers to the sign's object--is left to be identified in the interpretant, though the later may also be only a rheme (in which case the independent identification of the object is deferred to another interpretant). (2) The sign may limit the reference of the interpretant by including in itself an identification of the object: in this case the sign is a dicisign or dicent. There is still some range of possible alternatives within which the actual interpretant can fall, however. (3) The sign may determine precisely what its interpretant is, leaving no latitude whatsoever, because it itself specifies its own interpretant: in this case the sign is an argument. Peirce uses the word "argument" here in a more generalized sense than is to be found in the writings of logicians, of course, though various traditional non-logical usages of the word--as, for example, when one speaks of the argument of a literary work--justify this extension to some extent.
This trichotomic distinction is a generalization of the logician's distinction between a propositional function, a proposition and an argument, as Peirce construed these distinctions. Loosely speaking, a propositional function is simply a proposition in which the subject term is left unspecified and is therefore identifiable only insofar as the predicate identifies it. For example, the present essay, taken as a whole and as a single sign, is a proposition--or, more generally, a dicisign--since there are frequent identifications of the subject-matter within it. But one could delete the title, ignore the context, and delete all other indications of what it is about, and it would then be a rheme. It cannot be said to be an argument, though arguments are to be found within it here and there, since it does not exhibit the overall form of a set of premises with a specified conclusion or anything reasonably regardable as analogous to that.
This distinction is applicable to anything regarded as a sign, and it may be far from clear in many cases whether or how the object of a sign is being identified and whether or in what way it could or should be regarded as an argument. The questions one would ask oneself, though, would be as follows. First, should it be regarded as revealing or making manifest or apparent something about something, such that any part of it can reasonably be construed as indicative of what it is about? If not then it is a rheme; if so then whatever it reveals anything about is its object and it is a dicisign. (In order to identify something as a dicisign one might have to include what we would ordinarily call the "context" of its occurrence: that is, the sign might have to be construed as including both text and context.) Second, should it be regarded as divisible into two parts or aspects, such that one part or aspect is in some sense compelled, or at least made acceptable, given the other? If not then it is only a dicisign at most; if so then it is an argument, that part which is compelled or made acceptable being the conclusion relative to the other as premise, in a suitably extended sense of "premise" and "conclusion". (The conclusion stands to the premise as interpretant to sign within the more comprehensive sign which is the argument.)
This raises the problem of invalid arguments; for it may be clear enough that a premise-conclusion analysis of the sign is pertinent, though the putative conclusion may not in fact be compelled or even made acceptable as a conclusion by the premise (or premises--but a conjunction of premises can be regarded as a single premise). It is obvious that this can occur in the case of explicit verbal arguments of the sort which logicians commonly treat, and analogues to this can doubtless be found easily in almost any subject-matter treated semiotically: for example, in literary works which lapse into inconsequentiality at a certain point, in musical pieces which exhibit a failure in development, or in natural telic processes when something goes wrong relative to what can be reasonably expected in view of what has preceded. Peirce's way of handling this is to say that, strictly speaking, there are no invalid arguments, by which he means that what we would ordinarily regard as such is to be regarded either as no argument at all or else as really being valid though seemingly invalid because of a misidentification of the kind of argument it is: for example, an apparently invalid deductive argument might be reconstruable as an inductive or retroductive argument (perhaps rather weak in its degree of strength).
The point is that in our representation of the sign, in a semiotic analysis, we are not obligated by the intentionality of the sign-producer to represent the sign as being an argument or as being of this or that kind or strength. It is only the sign itself to which we have to be faithful in analysis, which means that we are obligated to represent it as it really is, not as it may appear to be to this or that particular person, including the sign-producer. This matter cannot be discussed further here, nor can a number of other questions that naturally arise about this distinction, but it is worth remarking that an idea of the importance of it--largely ignored to date, unlike the other two trichotomies discussed above--can be acquired by reflecting upon the extent to which aesthetic criticism, in particular, is besieged with difficulties in connection with the question of the referentiality, assertionality, and inferentiality involved in works of art.
In tabular summary,then, there are ten possible cross-combinations of these three trichotomic distinctions:
For an account of the seven other trichotomies which Peirce rather tentatively distinguished see Savan (l980).
As noted earlier, Peirce made what appears to be a second three-way distinction of the interpretant into the "emotional," "energetic," and "logical" during the same period in which he distinguished the immediate, dynamical, and final interpretant. This second distinction is similar to the qualisign/sinsign/legisign distinction since it is based on regarding the interpretant in its monadic, dyadic, and triadic aspects, respectively, though we cannot simply say that it is the qualisign/sinsign/legisign distinction as applied to the interpretant since an interpretant regarded monadically or dyadically is not being regarded as a sign even though all interpretants necessarily are signs (that is, representamens). This distinction is almost always discussed by Peirce in contexts in which it is clear that he is concerned with interpreters and their interpretations, considered as psychical acts or responses (as distinct from the objective content of such acts or responses). Now, the sign relation is always only triadic and the use of the idea of an interpreter and its interpretational acts or responses is always supposedly eliminable in principle in favor of an analysis using only the conceptions of sign, object, and interpretant, though in concrete cases such a puristic analysis might be so prolix as to be highly undesirable, as a practical matter. (A puristic analysis replacing one which involved the use of such words as "interpreter" and "interpretation" as shorthand devices would usually require the representation of logical levels beyond those which are minimally implicit in all semiosis processes. The analytic strategies available for producing a puristic analysis for all cases have yet to be developed or demonstrated, but the integrity of Peirce's semiotic as a system clearly depends upon this possibility. From a philosophical point of view what is at stake is the elimination of implicit reliance on a Cartesian conception of mind in favor of understanding the mental--or, more generally, the psychical, i.e. anything which exhibits a telic form--in terms of semiosis processes.) Thus although it does not follow that the distinction in question is itself eliminable, inasmuch as it is a distinction of kinds of interpretants, its chief use would nevertheless appear to be in cases where reference to interpreters and interpretation is convenient or even a practical necessity. But the use of such terms of convenience is not a practical necessity in all cases, and it is therefore questionable whether the distinction should be regarded as a generic one on a footing with the immediate/dynamical/final interpretant distinction. That is, the distinction could always be drawn but it is not obvious that there would be any purpose in doing so in many cases. It does not, in any case, add anything to the principles involved in sign-typology. It should be understood, though, that the nature and relation of these two sets of interpretant-distinctions is a matter of great dispute in Peirce scholarship at this time. For other accounts of the interpretant distinctions see Fitzgerald (1964), Short (1981a), Buczynska-Garewicz (1981a), and Eco (1976a, l976b, and 1981).
Something which has been implicit throughout the foregoing account should be emphasized at this point, namely, that Peircean semiotic (as a general theory) is phenomenologically based, and the effective application of it will often--though not always--require the user to adopt the same basic point of view. Peirce's conception of phenomenology--which he also called "phaneroscopy"--was not derived from Husserl's (though it is similar to it in some respects), but a general account of it cannot be given here. For present purposes let it suffice to say that it means that semiotic analysis does not, in general, restrict itself to entities such we would ordinarily think of as being "in the real world." On the contrary, anything whatsoever which is referable to and describable is a possible subject-matter for semiotic analysis, regardless of whether we would regard it as being something sensed or remembered or dreamt or conceived or imagined or hallucinated, and so on. Or, to put it another way, to say that semiotic is phenomenological is to say that it presupposes no metaphysics, whether it be the hodge-podge metaphysics (largely Cartesian in type) endemic in modern "common sense" or a more explicit and sophisticated product of philosophical theorizing. (See Rosensohn (1974) on the topic of Peirce's phenomenology generally.) Also, it should be understood that the entities with which a semiotic analysis is concerned are not, in general, required to be of the nature of logical individuals, which is to say that they do not have to measure up to the so-called "laws" of non-contradiction or excluded middle: semiotic objects, signs, and interpretants can be highly indeterminate in the sense in which, say a fleeting object of the imagination or a dreamt object can be indeterminate relative to any number of possible characteristics. All that is necessary is that the entity be something to which reference can effectively be made, which requires in turn that it be capable of some identifying and differentiating description. (See Brock (l979) on the topic of the determinateness and indeterminateness of signs.)
General semiotic, as a whole, is divided by Peirce into three parts, the second presupposing the first, the third presupposing the second. "Speculative grammar", "critical logic", and "speculative rhetoric" are his most commonly used labels for these three parts, respectively. The word "speculative" is to be understood in the sense of "theoretical", as derived from the Greek word "theoretikos, and therefore in contradistinction from "applied". The word is used in commemoration of the Medieval attempt at a grammatica speculativa (Kloesel 1981). "Philosophical grammar" and "philosophical rhetoric" would doubtless be more apt labels at present, however. Philosophical grammar is the general theory of the nature of the representation relation and of the various types of signs, construed as structuring semiosis processes. (Hence the foregoing account of Peirce's semiotic is actually an account only of this, its first and foundational part.) The second part, critical logic, was developed by Peirce primarily as the unified theory of deduction induction, and retroduction (hypothetical inference), and it is clearly recognizable as an ambitious and innovative attempt at a comprehensive logic in the traditional sense. The question arises, though, as to whether critical logic can be still further broadened in conception such that it is applicable to semiosis of all sorts. This would involve an extension of the conceptions of deduction, induction, and retroduction so as to be applicable to art works of all sorts, for example, and to all natural processes which exhibit the requisite telic form. There are numerous indications that Peirce himself thought this possible, but it remains wholly undeveloped in any systematic way to date. The third part of semiotic, philosophical rhetoric, is concerned with semiosis as regards its effectiveness (efficacy). ("Methodeutic" is one of Peirce's many alternative labels for this part of semiotic.) Heuristics, the theory of inquiry, the logic of discovery (which is not to be confused with the logic of hypothesis), the technê or art of communication and exposition, and anything else concerned with the enhancement of the vitality of coherent semiosis processes in general would fall within the scope of philosophical rhetoric. Though Peirce did substantial work in this area, a systematic account of it has yet to be developed.
General semiotic is then distinguished by Peirce from what he calls "the psychical sciences" (or "psychognosy"), by which he means such special sciences as the various psychological and social sciences, linguistics, history, art/esthetic criticism of various sorts, and so forth. (The word "psychical" had not, of course, taken on the debased sense which it subsequently acquired in virtue of its unfortunate association with "psychic" phenomena so-called. The phrase "semiotical sciences" would be a felicitous substitute for "psychical sciences" now, provided one were convinced that these types of inquiry are indeed amenable to semiotical treatment.) These are those special sciences in which specialized semiotic conceptions would be the predominating ones, and they are contrasted by Peirce with the physical sciences, which include physics and astronomy, chemistry, and biology. However, in listing biology (including physiology) under the heading of the physical rather than the psychical sciences, contrary to what one would expect, Peirce did not mean that the phenomena studied by biology are non-psychical (non-semiotical) in character. The idea is rather that the biological sciences are importantly dependent upon chemical (hence physical) conceptions in a way that the psychical sciences indicated above are not. In fact, Peirce does not even hold that the phenomena studied by physics are all non-semiotical in character, for he speculates at times about the possibility of certain physical processes (those that are non-conservative or irreversible, for example) being psychical in character. His classification of the special sciences into these two main groups is not based on subject-matter differences but rather on what he believed to be importantly different basic orientations which a scientist may take, corresponding to different intellectual types among scientists. If the orientation is primarily toward efficient ("brute force") causation then he classifies the science as physical; if the orientation is primarily toward final or telic causation then he classifies the science as psychical. In any case, this neither precludes the use of semiotic conceptions in the physical sciences nor the use of non-semiotic conceptions in the psychical sciences. Moreover, as remarked earlier, dyadic conceptions of lawlike relations can be regarded as degenerate cases (in the mathematical sense of "degenerate") of conceptions of tendency or finality.
Everything known to have been published by Peirce himself or in his life-time is available in microfiche form in:
This is available with:
(1) and (2) are also available separately, the latter being in book form. The corpus of Peirce's voluminous unpublished work (i.e. his Nachlass is available in microfilm form in:
Microfilms of Peirce's professional correspondence is available from the same source. To accompany the foregoing there is:
To this should be added:
However, a new catalogue of the unpublished material, as radically reorganized (with previously fugitive manuscript fragments reunited, etc.), is in preparation at the Peirce Edition Project. The major collection of Peirce's writings which is currently available in book form is:
Also available in book form is:
There is very little overlap in content of items (6) through (9). However, a new edition of Peirce's philosophical works which will--when completed (some years hence)--make the Collected Papers redundant (though not items (7) through (9), which it will presuppose as supplements) is underway at the Peirce Edition Project at Indiana University-Purdue University at Indianapolis.
The microfiche of the totality of Peirce's published work in all fields (i.e. item (1) above) is also presupposed as supplementary to (10), and will remain of distinctive value since it contains much material that will probably never be available elsewhere. Although the corpus of unpublished works will be largely incorporated within the Chronological Edition, this will be some years hence, and even then there will probably be unincorporated material of scholarly interest which it was not editorially feasible to include. It should also be noted that there is scholarly material available both at the Institute for Studies in Pragmaticism (Texas Tech University) and at the Peirce Edition Project (referred to above) which is not available elsewhere.
Ransdell's home page is: