1. A PRELIMINARY NOTE ON SPELLING AND PRONUNCIATION

Most of the vocabulary of Peirce's doctrine of signs—for examples, representation, sign, object, and interpretant —is derived from Latin, and poses no difficulty of spelling or pronunciation. But for the science itself and for what it studies, he uses English forms of two Greek terms that are more troublesome in both these respects. Now Peirce was, among other things, linguist, philologist, lexicographer, and exponent of the ethics of terminology. So if we count him a founder of our science, we shall wish to know what these terms were, and how he spelled and pronounced their English forms.

For σημείωσιϛ—sign-action, the operation or functioning of a sign, sign-interpretation, or the act of inferring from signs—he uses two English forms, semiosis and semeiosy. The former he tells us to pronounce with the e and the first i long and with the accent on the о (5.484*).¹ He does not tell us where to place the accent in semeiosy (5.473), but I think he put it on the second syllable, pronouncing it "my."² For the plural of semiosis, he uses semioses (5.489).³

For σημειωτκή the art or science or doctrine or general theory of semioses—he uses semeiotic; much less often, semeiotics or semiotic; very rarely, seme otic; never semiotics. To tell us how to pronounce his preferred form, he marks it sēmeio'tic (Ms 318 p.15).⁴

His rationale for that spelling and pronunciation was probably twofold. (1) There is no more reason for semeiotics or semiotics than for logics or rhetorics. (2) Both the spelling and the pronunciation should (in this case, at least) be signs of etymology; that is, should make it evident that the derivation is from Greek σημԑιον, sign, not from Latin semi- ("half־"). There is nothing halfway about semeiotic; it is all about signs, and it is about all signs. And the о in semeiotic should be long because it has behind it a Greek omega, not an omicron.

In the remainder of this paper, I shall use in quotations whatever spellings Peirce there uses, but outside of quotations I shall use only semeiosis and semeiotic, and I invite the reader to pronounce them with me "See my o, sis" and "See my о tick." I cannot believe that Peirce ever pronounced the latter "semmy-AHT-ick."

* References in the form 5.484 are to the Collected Papers of Charles Sanders Peirce by volume and paragraph number. (Cambridge: Harvard University Press, vols. 1-6 edited by Hartshorne and Weiss, 1931-1935; vols. 7-8 edited by Burks, 1958.) References in the form "Ms 318 p. 15" are to the Charles S. Peirce Papers in the Houghton Library at Harvard University, quoted by permission of the Department of Philosophy. References in the form NE 3:886 are to The New Elements of Mathematics by Charles S. Peirce, edited by Carolyn Eisele (4 vols, in 5, The Hague: Mouton; Atlantic Highlands, N. J.: Humanities Press, 1976), by volume and page. References in the form W75 are to the pages of Semiotic and Signifies: The Correspondence between Charles S. Peirce and Victoria Lady Welby, edited by Charles S. Hardwick (Bloomington: Indiana University Press, 1977). Christian J. W. Kloesel has helped me by calling my attention to manuscript passages I might otherwise have missed, by criticizing drafts of this essay, and in numerous other ways.

2. THE FIRST FOUNDING (1865-1869)

Peirce's training was in chemistry. His career was in the service of the United States Coast Survey, 1859-1860, 1861-1891. His work for the Survey was primarily astronomical and geodetic, but it involved metrology, spectroscopy, optics, color theory, map projections, the four-color problem, and the history of astronomy and of science in general. His contributions to the annual reports of the Survey included one on the theory of errors of observations in the Report for 1873 and one on the economy of research in that for 1876. He deliberately diversified his researches beyond the requirements of his work for the Survey, not from ambition to contribute to as many sciences as possible, but with a view to advancing the logic of science; that is, of hypothesis and induction. His first professional publication was on the chemical theory of interpénétration; his second on the pronunciation of Shakespearian English. He was a mathematician also, but with a view to advancing the logic of mathematics, that is, of deduction.

In the spring of 1877, when he was being considered for election to the National Academy of Sciences, he submitted a list of four of his published papers in logic and asked that his eligibility be judged by these rather than by his contributions to the special sciences.⁵ He was elected, and in his letter of acceptance he expressed his "gratification at the recognition by the Academy of Logic as entitled to a place among the real sciences."⁶ Many of the papers he later presented to the Academy were in logic, and at least one in semeiotic.

For five years, 1879-1884, he was part-time Lecturer in Logic at the Johns Hopkins University, while continuing his work for the Coast Survey.⁷

From time to time he gave single courses of lectures at Harvard University, at the Lowell Institute in Boston, and elsewhere. These were usually in logic, in the history of logic, or in the history of science considered from the viewpoint of the logic of science.

His first such course was given at Harvard University in the spring of 1865, under the title "The Logic of Science." In the first half of the first lecture he reviewed various definitions and conceptions of logic, psychological and nonpsychological. In the second half he approached his own nonpsychological definition by way of Locke's identification of logic with semeiotic, "the doctrine of signs," in the last chapter of his Essay Concerning Human Understanding (1690). The resulting definition of logic, Peirce said, would serve as a first approximation; but it was too broad, since, of the three kinds of representations, logic treats only of symbols. (Locke had used "representation" as a synonym of "sign," and Peirce at this time was using "representations" as his technical term for signs in general.⁸)

On May 14,1865, Peirce began a book called Teleological Logic with a chapter of definitions, in which, like Locke, he makes semeiotic one of the three most general kinds of science. With no further help from Locke, he then makes symbolistic one of the three divisions of semeiotic, as he had done in his lecture; and he makes General Grammar, General Rhetoric, and General Logic the three divisions of Symbolistic (Ms 802).

In Boston in the fall of 1866 he gave a course of twelve Lowell Lectures on "The Logic of Science; or, Induction and Hypothesis," in which the doctrine of signs was carried into somewhat greater detail (Mss 351-59, esp. 357, 359).

The first published sketch of his semeiotic was in a paper "On a New List of Categories," which he presented to the American Academy of Arts and Sciences on May 14, 1867. Forty years later he described this paper as the outcome of "the hardest two years' mental work that I have ever done in my life" (1.561). He first establishes, in place of Aristotle's ten categories and Kant's twelve, a new list of three: Quality, Relation, Representation. He then uses these categories to distinguish: (1) three kinds of representations—likenesses (which he will later call icons), indices, and symbols; (2) a trivium of conceivable sciences—formal grammar, logic, and formal rhetoric; (3) a general division of symbols, common to all three of these sciences—terms, propositions, and arguments; and (4) three kinds of argument, distinguished by their three relations between premisses and conclusion—deduction (symbol), hypothesis (likeness), induction (index) (1.545-59).⁹

It is evident that Peirce is still using representation in the general sense in which he will later use sign. In effect, therefore, he is making of sign an ultimate and irreducible category. It would seem to follow, though he does not press the point, that we need an autonomous science or doctrine of signs. Other sciences—perhaps any other science—may supply indispensable data, but no synthesis of these will suffice to constitute the science.

Nevertheless, it might plausibly be objected, Peirce is a logician, and he concerns himself with semeiotic only so far as is necessary to place logic within the larger framework of that one of the three most general kinds of science that Locke, following the ancient Greeks, had distinguished. To that objection, however, it may fairly be replied that at no time of his life did Peirce set any limit to the intensity of cultivation of the larger field of semeiotic that would be advantageous for purposes of logic, even if the cultivating had to be done by logicians themselves because, for the time being, they were the only semeioticians.

In any case, it was not enough in Peirce's eyes for semeiotic to provide a pigeonhole for logic in the classification of the sciences. This became fully apparent in 1868-69 in a series of three articles in the Journal of Speculative Philosophy : "Questions Concerning Certain Faculties Claimed for Man," "Some Consequences of Four Incapacities," and "Grounds of Validity of the Laws of Logic: Further Consequences of Four Incapacities" (5.213-357).

The first two papers are there for the sake of the third. The upshot of the series is a theory of the validity of the laws of logic, including those of the logic of science (that is, of hypothesis and induction) as well as those of the logic of mathematics (that is, of deduction). Yet the first paper is in the form of a medieval quaestio, a disputed question, and the second begins with a four-point statement of "the spirit of Cartesianism," followed by an opposed four-point statement of the spirit of the scholasticism that it displaced. In respect of these four antitheses, "modern science and modern logic" are closer to the spirit of scholasticism. The first paper was "written in this spirit of opposition to Cartesianism." It was meant to illustrate as well as to commend the "multiform argumentation of the Middle Ages." It resulted in four denials.

These propositions cannot be regarded as certain, Peirce says; and the second paper puts them to the further test of tracing out some of their consequences. The third paper then constructs a theory of the validity of the laws of logic in the form of "further consequences" of these "four incapacities."

The central positive doctrine of the whole series is that "all thought is in signs" (5.253). Every thought continues another and is continued by still another. There are no uninferred premisses and no inference-terminating conclusions. Inferring is the sole act of cognitive mind. No cognition is adequately or accurately described as a two-term or dyadic relation between a knowing mind and an object known, whether that be an intuited first principle or a sense-datum, a "first impression of sense" (5.291). Cognition is a minimally three-termed or triadic relation (5.283). The sign-theory of cognition thus entails rejection not only of Cartesian rationalism but also of British empiricism.

The sign-theory of cognition leads into a semeiotic theory of the human self, "the man-sign" (5.313), and thence into a social theory of logic. "When we think, then, we ourselves, as we are at that moment, appear as a sign" (5.383) ; "the word or sign which man uses is the man himself" (5.314). "Finally, no present actual thought (which is a mere feeling) has any meaning, any intellectual value; for this lies not in what is actually thought, but in what this thought may be connected with in representation by subsequent thoughts; so that the meaning of a thought is altogether something virtual" (5.289). "Accordingly, just as we say that a body is in motion, and not that motion is in a body, we ought to say that we are in thought and not that thoughts are in us" (5.289nl).

"The real, then, is that which, sooner or later, information and reasoning would finally result in, and which is therefore independent of the vagaries of me and you. Thus, the very origin of the conception of reality shows that this conception essentially involves the notion of a COMMUNITY, without definite limits, and capable of an indefinite increase of knowledge" (5.311).¹⁰ "So the social principle is rooted intrinsically in logic" (5.354).

Along the way, with the help of his three categories, Peirce's doctrine of signs is worked out in greater detail in these three papers, and especially in the second of them.

As a first approximation, then, we may say that, if Peirce was a founder—perhaps the founder—of modern semeiotic, the first founding took place in the years 1865-1869. The most relevant publications were "On a New List of Categories" (1867) and the three papers developing the sign-theory of cognition (1868-1869). The chief occasions for the founding were that Peirce was invited to give lecture courses in "the logic of science" at Harvard in 1865 and at the Lowell Institute in 1866; that he presented five papers on logic to the American Academy of Arts and Sciences in 1867; and that the editor of the Journal of Speculative Philosophy challenged him in 1868 to show how, on his principles, the validity of the laws of logic could be "other than inexplicable" (5.318).

The semeiotic thus founded was semeiotic as viewed from the stand-point of logic and studied for the purposes of logic, and more particularly for those of the logic of science rather than for those of the logic of mathematics. But it was a semeiotic that included logic.

3. THE FIRST NON-PEIRCEAN ERECTION ON THIS FIRST FOUNDATION (1913)

So far as I am aware, nobody but Peirce himself deliberately built on this first foundation until forty-five years later. Then, in 1913, Josiah Royce, though acquainted with much of Peirce's later work, discovered in the doctrine of signs contained in these four early published papers just the foundation he needed for solving "the problem of Christianity." In a two-volume work under that title he moves toward the solution in the following four chapters:

The very first step toward the solution was to abandon the dyadic models of perception and conception and to adopt in their stead Peirce's triadic semeiotic model of interpretation.¹¹

4. PRAGMATISM A SECOND FOUNDING? (1877-1879)

As we shall see, when modern semeioticians began in the 1920s and 1930s to recognize Peirce as a founder of their science, the Peirce they had in mind was the founder of pragmatism. Pragmatism was, at least in the first place, a theory of meaning, and therefore a contribution to the doctrine of signs. Peirce's first published exposition of pragmatism was in a series of six "Illustrations of the Logic of Science" in the Popular Science Monthly in 1877-78.¹² A book under the same title was announced as in preparation for the International Scientific Series but never appeared. The "Illustrations" bore the following titles:

ILLUSTRATIONS OF THE LOGIC OF SCIENCEE

First Paper.—The Fixation of Belief.

Second Paper.—How to Make Our Ideas Clear.

Third Paper.—The Doctrine of Chances.

Fourth Paper.—The Probability of Induction.

Fifth Paper.—The Order of Nature.

Sixth Paper.—Deduction, Induction, and Hypothesis.¹³

A reader coming to these papers directly from that "On a New List of Categories" and those on the sign-theory of cognition and the validity of the laws of logic would soon make the following observations, (a) Peirce is having another go at the validity of the laws of logic, and more particularly those of the logic of science; that is, of hypothesis and induction. (b) The upshot is not radically different; we reach the social theory of logic at the same stage (2.654); but the pragmatism that is only implicit in the earlier papers, if present there at all, is now unfolded as the lesson in logic taught by Darwin's Origin of Species (5.364).¹⁴ (с) Though there is no mention of the categories or of the doctrine of signs, they are omnipresent, and the "Illustrations" become fully intelligible only in the light of the four papers of a decade earlier, (d) The categories are the key to the analysis of belief, doubt, and inquiry in the first paper, and to the distinction of the three grades of clarity in the second paper, (e) The sign-object-interpretant triad is the key to the maxim for attaining the third grade of clarity: "Consider what effects, which might conceivably have practical bearings, we conceive the object of our conception to have. Then, our conception of these effects is the whole of our conception of the object." (f) The whole series is thought out within the framework of the doctrine of signs, (g) Peirce has presumably suppressed the terminology and the technicalities of semeiotic so as not to put too great a strain on the readers of the Popular Science Monthly, (h) Perhaps the book never appeared because he decided that this suppression had been a mistake, but he did not find time for the rewriting that would have been needed to save the book from the same mistake, (i) Even so, the "Illustrations," just as they appeared in the Monthly, constitute an anti-Cartesian Discourse on the Method of Rightly Conducting the Reason and Searching for the Truth in the Sciences.

Take observation (e). In the second paper Peirce applies the maxim to the scientific conceptions of hardness, weight, and force, and to the logical and metaphysical conceptions of truth and reality; and in the third and following papers he applies it to the most difficult conception of the logic of science, that of probability. Take hardness, for example. The object is the physical property designated by the sign hard as used both by laymen and by mineralogists. The three grades of clarity are exemplified by three kinds of interprétants of this sign. The second presupposes the first, and the third presupposes the first and the second. The first is that of familiar feel, ready use, and easy recognition; the second is that of abstract genus-and-differentia or synonym-and-antonym definition. At the very least, what is hard is not soft, and what is harder than x is less soft than x. Suppose that the second kind of interpretant, and thereby the second grade of clarity, that of distinctness, is already attained; then the rule for reaching the third involves two further steps. In the first further step we specify, in this case, the sensible effects of one thing's being harder than another; say, of a diamond's being harder than glass. Sensible effects are not effects upon our senses, but perceivable public effects. For example, diamond will scratch glass but glass will not scratch diamond. In the second further step we specify practical bearings of these effects. Practical bearings are bearings on practice or conduct; that is, on habits of action. A sensible effect has a practical bearing if it is such that to conceive ourselves as being in a certain situation and having a certain desire is to be ready to act in a certain way if such a situation should ever arise. For example, we can conceive ourselves as desiring to divide a sheet of glass, and as having no regular glass-cutting tool available, but only a diamond ring. So to conceive is already to have formed the habit of using the diamond to cut the glass in such situations. On the other hand, we can conceive ourselves as having a sheet of glass we do not want scratched; say, a mirror. The habit of action determined by the belief that diamond is harder than glass will in that case be the habit of keeping the diamond ring away from the mirror. In each of these cases, the third and final interprétant, which marks the third level of clarity, consists of conceived sensible effect, conceived desire, and habit of action together. At that level of clarity, interprétants such as these constitute the whole of our conception of the object represented by the sign hard. The mineralogists' scale of hardness is arrived at by interpreting hard in this way, and the scale itself is so interpreted.

Much of this, however, would have escaped a reader unacquainted with Peirce's earlier papers. If the pragmatism of 1877-79 was indeed a second founding of semeiotic, this would have been evident at the time only to a reader who had the first founding very much in mind. In both foundings, the semeiotic is one that includes logic and that serves logic.

5. PHILODEMUS AND SEMEIOSIS (1879-1883)

In 1865, the first year of the first founding, Theodor Gomperz published an edition of the Herculaneum papyrus remains of a Greek treatise on inductive logic by the Epicurean philosopher Philodemus. The papyrus lacked the title, but the one most often given it is the Latin De signis ("On Signs").

Peirce seems not to have made the acquaintance of this work immediately, but at the Johns Hopkins University he had a student named Allan Marquand, with whom he made an intensive study of it in 1879-80. To meet the thesis requirement for his Ph.D. degree, Marquand translated the treatise under the title "On Inductive Signs and Inferences" and wrote an introduction to it. The introduction, or an abridgment of it, was published under the title "The Logic of the Epicureans" as the first essay in a volume of Studies in Logic edited by Peirce in 1883.¹⁵

One of the most striking features of the treatise is the frequency of the term semeiosis. The Greek suffix -sis means the act, action, activity, or process of. Peirce was prepared to understand semeiosis in either of two ways : (1) from the side of the sign, as sign-action, the functioning of a sign, or (2) from the side of the interprétant, as sign-interpreting or inferring from signs. Philodemus used it primarily in the latter sense, and even more narrowly as drawing inductive inferences from inductive signs. But for Peirce sign-action and sign-interpretation were not two different kinds of semeiosis but one and the same semeiosis considered from two points of view. To act as a sign is to determine an interpretant.

Furthermore, a sign is not a kind of thing. The world does not consist of two mutually exclusive kinds of things, signs and non-signs, each with its subdivisions, yet with no subdivision of the one overlapping any subdivision of the other. There is nothing that may not be a sign; perhaps, in a sufficiently generalized sense, everything is a sign: "all this universe is perfused with signs, if it is not composed exclusively of signs" (5.448nl). The fundamental distinction is not between things that are signs and things that are not, but between triadic or sign -action and dyadic or dynamical action (5.473). So the fundamental conception of semeiotic is not that of sign but that of semeiosis; and semeiotic should be defined in terms of semeiosis rather than of sign, unless sign has antecedently been defined in terms of semeiosis. A quarter of a century later, in 1907, Peirce could still describe himself as "a pioneer, or rather a backwoodsman, in the work of clearing and opening up what I call semeiotic, that is, the doctrine of the essential nature and fundamental varieties of possible semiosis" (5.488).

6. SEMEIOTIC AND THE LOGIC OF MATHEMATICS (1866-1911)

Peirce wrote in 1903 : "It has taken two generations to work out the explanation of mathematical reasoning" (NE 3:1119; cf. 1:256). What were the essential steps that he himself took or observed others taking? A list of some of them follows.

But first a prefatory note. It all started in 1854 with The Laws of Thought by George Boole, the Copernicus of modern logic (Ms 475 p.6).¹⁶ After an introductory chapter on the nature and design of the work, Boole began the work itself with a chapter entitled "Of Signs in General, and of the Signs Appropriate to the Science of Logic in Particular; also of the Laws to which that Class of Signs are Subject." Of Peirce's five papers on logic in 1867, the first was "On an Improvement in Boole's Calculus of Logic," the fourth took off from Boole, and Peirce later showed how study of Boole led him to the "natural classification of arguments" in the second (Ms 475 pp.2-28). Now for the steps:

(1) In a privately printed paper of 1866 (at 2.801-804) and in his second and third papers of 1867 (at 2.470, 474 and 1.559) Peirce showed, as he later put it, that "all logical thought" is "an operation upon symbols consisting in substitution" but did not claim or assume that such substitution is "an indecomposable operation."¹⁷

(2) In 1870 Peirce published his "Description of a Notation for the Logic of Relatives" (3.45-149), with sections on the various signs (for examples, of inclusion, equality, addition, multiplication, involution). The logic of relatives became the key to the inexhaustible richness of mathematical reasoning, its ability to draw indefinitely numerous necessary conclusions from a single hypothesis, a single premiss or conjunction of premisses (NE 4:58-59).

(3) In the same year his father, Benjamin Peirce, began his Linear Associative Algebra with the sentence: "Mathematics is the science which draws necessary conclusions." He went on to discuss "the language of algebra"—its letters and signs and rules of composition. The first principle he states is that of "the substitution of letters," which "is radically important, and is a leading element of originality in the present investigation."

(4) During the period in which son and father were working on (2) and (3), they had frequent conversations. The son later remembered two things: (a) The father at one point seemed inclined toward the view, later embraced by Dedekind, that mathematics is a branch of logic; but the son "argued strenuously against it," and thus the father "came to take the middle ground of his definition" (NE 3:526). (b) The father as mathematician and the son as logician were both struck by the contrary nature of their interests in the same propositions and in the systems of notation in which they were represented. Take the algebra of logic for example.

(5) In the 1870s, the British mathematicians Cayley, Sylvester, and Clifford made two-way connections between mathematics and chemistry. Cayley applied his mathematical theory of "trees" to a problem in chemistry. Sylvester and Clifford shortened to graph the "graphic formula" of the chemists, and, starting with the theory of invariants, they began adapting such graphs to mathematical uses.

(6) Sylvester became professor of mathematics at the Johns Hopkins University in 1876, founded the American Journal of Mathematics there in 1878, introduced the new term graph in the first issue, and said that Clifford had found "the universal pass key to the quantification of graphs."¹⁸

(7) Peirce joined the Hopkins faculty in 1879. As chemist, mathematician, friend of Clifford (who had died in the spring), and now younger colleague of Sylvester, he welcomed the adapting of chemical graphs to mathematical uses. To the Journal's first seven volumes (1878-1885) he contributed a review and four articles, as well as a new edition of his father's Linear Associative Algebra, with many notes and two addenda by himself.¹⁹

(8) Cayley was visiting lecturer at Hopkins from January to June 1882. Peirce, as usual, was attending meetings of the Mathematical Society, presenting papers to it, and taking part in discussions of papers presented by others. At its January meeting, for example, papers were presented by Cayley, Sylvester, and Peirce.²⁰ In the spring Peirce gave a short course of three lectures on the logic of relatives for students of mathematics.

(9) In 1883 George Chrystal gave an account of mathematics in the ninth edition of the Encyclopaedia Britannica, which Peirce took to be defining it as the science of making pure hypotheses, though Chrystal used the term conception rather than hypothesis (3.558). Chrystal, he said, "puts emphasis upon the definiteness of mathematical hypotheses. ... I incline to suspect that Prof. Chrystal has confounded definiteness with iconicity, or the capability of being represented in a diagram" (NE 2:595).

(10) In 1885 Peirce published the second of his two papers "On the Algebra of Logic," with the subtitle "A Contribution to the Philosophy of Notation" (3.359-403). It begins with a section on "Three Kinds of Signs"—icons, indices, and tokens—whose thesis is that "in a perfect system of logical notation signs of these several kinds must all be employed." He gives his student O. H. Mitchell credit for introducing indices, and thereby quantification, into the algebra of logic. He goes on to say that by means of tokens and indices alone "any proposition can be expressed; but it cannot be reasoned upon, for reasoning consists in the observation that where certain relations subsist certain others are found, and it accordingly requires the exhibition of the relations reasoned with in an icon." The theory of signs and the logic of relatives thus lead to the further conclusion that all deductive reasoning, including that of mathematics, involves experiment and observation (3.363). In the main body of the paper, Peirce presents in the form of twelve "icons" the algebraic foundations of a system of material implication, including truth-table analysis and quantification. One of these "icons," the fifth (3.384), has come to be called Peirce's Law.

Every one of the icons consists of symbols (here called tokens) and is a symbol. Some of the elementary symbols are indices. But what Peirce wants to emphasize is the iconicity of each formula as a whole. The logic of relatives has opened the way for him to extend the notion of iconicity from quasi-geometrical graphs, whose iconicity was already obvious, to algebraic formulations of the laws of logic, whose iconicity is rendered obvious by the logic of relatives.

It follows that, just as the world does not consist of two mutually exclusive kinds of things, signs and non-signs, so there are not three mutually exclusive kinds of signs: icons, indices, and symbols. These are rather elements or aspects of semeioses that vary greatly in relative prominence or importance from semeiosis to semeiosis. We may therefore call a sign, for short, by the name of that element or aspect which is most prominent in it, or to which we wish to direct attention, without thereby implying that it has no element or aspect of the other two kinds.

(11) In 1886 A. B. Kempe published in the Philosophical Transactions "A Memoir on the Theory of Mathematical Form" and sent an inscribed copy to Peirce, who annotated and indexed it. Kempe made an extensive use of graphs, and it was in part by critical study of this memoir that Peirce later arrived at his own two systems of graphs. As late as 1905, he called Kempe's "great memoir" "the most solid piece of work upon any branch of the stecheology of relations that has ever been done" (5.505).

(12) In 1889 Peirce contributed to The Century Dictionary the first dictionary definition of the new term graph:

(13) In 1894-95 Peirce drafted two textbooks: Elements of Mathematics (NE 2:1-232) and ՝New Elements of Geometry Based on Benjamin Pence's @@@@Work ե Teachings@@@ (NE 2:233-473). In the former he describes mathematics as "the exact study of ideal states of things" and says his father's definition "comes to much the same thing" (NE 2:10). "Two kinds of icons are chiefly used by mathematicians, namely, first, geometrical figures, drawn with lines, and, second, arrays of points or letters .... upon which experiments and observations can be made" (NE 2:24; cf. 2:12).

(14) In 1896, in a paper "On Quantity, with special reference to Collectional and Mathematical Infinity," Peirce finally concedes that his father's definition of mathematics is defective in that it omits the framing by the mathematician of the hypotheses from which he pro־ ceeds to draw necessary conclusions (NE 4:271); and he offers a definition of his own that makes good that defect (NE 3:40-41).

(15) In The Monist for January 1897, with references to Clifford and Kempe by name (3.468, 479nl) and to Sylvester by implication (3.470), Peirce presented the system of what he later called entitative graphs. While reading the proofs, he conceived another system, which he called existential graphs. Partial expositions of this second and more iconic system reached print in 1903 (4.394-417) and 1906 (4.530-72).²²

(16) In the Educational Review for 1898 Peirce published "The Logic of Mathematics in Relation to Education" (3.553-62).

Peirce describes the "stripping" as "skeletonization or diagrammatization"; that is, iconization.

(17) In 1901, in a draft of "On the Logic of drawing History from Ancient Documents, especially from Testimonies," Peirce divided deductions into two kinds, corollarial and theorematic, and gave a detailed example of each, both drawn from the doctrine of multitude (NE 4:1-12). He took this to be the most important division of deductions, and his own most important discovery in the logic of mathematics (NE 4:38, 56). He had already "opened up the subject of abstraction" (NE 4:1), distinguished its two kinds, prescission and subjectif action, and called the latter "the very nerve of mathematical thinking" (2.428). He now proceeded to divide theorematic reasoning into abstractional and non-abstractional (NE 4:49). Here again the theory of signs came into play. "Every subject partakes of the nature of an index. . . . The expressed subject of an ordinary proposition approaches most nearly to the nature of an index when it is a proper name. . . . Among, or along with, proper names we may put abstractions. . . ." (2.357). But this is matter for a separate long article or short book.

(18) In 1902, in the chapter of his Minute Logic on "The Simplest Mathematics," Peirce briefly restates the distinction between corollarial and theorematic deduction; speaks of the latter as "mathematical reasoning proper"; describes it as "reasoning with specially constructed schemata"; and says it "invariably depends upon experimentation with individual schemata," that is, with icons, whereas corollarial reasoning is largely "reasoning with words," that is, with symbols (4.233). In the same chapter, in an eleven-page passage omitted by the editors of the Collected Papers (at 4.261), he introduces two notations for the sixteen binary connectives of the two-valued propositional calculus. One of these may be called his box־X, the other his cursive notation. He says it was his Hopkins student Christine Ladd-Franklin "who first proposed to put the same character into four positions in order to represent the relationship between logical copulas, and ... it was a part of her proposal that when the relation signified was symmetrical, the sign should have a right and left symmetry." Peirce's own notations simply carrry out that proposal in a particular way (NE 3:272֊75n at 272).²³

(19) In his article on Symbolic Logic in Baldwin's Dictionary of Philosophy and Psychology in 1902, Peirce said the symbols should inelude graphical as well as algebraic ones, and that a system of symbols devised for the investigation of logic, as opposed to one intended as a calculus, "should be as analytical as possible, breaking up inferences into the greatest possible number of steps, and exhibiting them under the most general categories possible." "There must be operations of transformation. ... In order that these operations should be as analytically represented as possible, each elementary operation should be either an insertion or an omission" (4.372-74).²⁴

(20) In 1903, in his Syllabus of Certain Topics of Logic, there appeared the first published account of Peirce's existential graphs (4.394-417), including rules of transformation and code of permissions, from which it appears that in this system each elementary operation is an insertion or an omission. This is preceded by a section called "The Ethics of Terminology"—an ethics that applies to notations and other symbols as well as terms (2.219-26). And that is preceded by "An Outline Classification of the Sciences" (1.180-202). Logic is now a normative science, depending on ethics, as that does on esthetics. Above the normative sciences are mathematics and phenomenology.

(21) About 1904, in his καινά στοιχεία ("New Elements"), Peirce presents the best restatement so far of his general theory of signs (NE 4:238-63). Symbols are now genuine signs; indices are signs degenerate in the first degree; icons are signs degenerate in the second degree. A symbol sufficiently complete always involves an index; an index sufficiently complete always involves an icon (NE 4:256). But "the icon is very perfect in respect to signification, bringing its interpreter face to face with the very character signified. For this reason, it is the mathematical sign par excellence" (NE 4:242).

(22) About 1905 Peirce begins "The Rules of Existential Graphs" (Ms 1589) with a preface and an introductory section on "The Nomenclature," in which he confesses a violation of the ethics of terminology in his previous expositions. The preface reads:

(23) In 1906, in his "Prolegomena to an Apology for Pragmaticism," Peirce presents the fullest and most mature accounts both of his semeiotic (4.530-51) and of his existential graphs (4.552-72) that he sueceeded in publishing. A sample sentence:

(24) Up to this point, Peirce has concerned himself primarily with the classification of arguments. From the beginning he recognizes three kinds, which he calls at first deduction, induction, and hypothesis. The last he later calls abduction, and finally retroduction. He has set the logic of mathematics (that is, of analytic, deductive, or necessary arguments) over against the logic of science (that is, of ampliative or probable arguments, either retroductive or inductive). In 1908, however, in "A Neglected Argument for the Reality of God," he presents retroduction, deduction, and induction as successive stages of inquiry (8.468-73). To that extent, he absorbs the logic of mathematics into that of science. Deduction, he says, has two parts.

(25) The nearest thing to a retrospective summing up is in a long letter to J. H. Kehler in 1911 (NE 3:159-210), from which I quote two short passages.

He gives an exposition of the syntax of his existential graphs, in the course of which he remarks that

In concluding this section, I trust that its twenty-five selected steps in the working out of the explanation of mathematical reasoning have made it sufficiently evident that Peirce's lifelong study of the logic of mathematics was conducted throughout within the framework of the general theory of signs.

7. THE REBIRTH OF PRAGMATISM (1898-1911)—A THIRD FOUNDING?

In the United States, at least, it was in 1898 that the word pragmatism was first used in a public address and then in print as the name of a philosophic doctrine and method. The speaker was William James, addressing the Philosophical Union at the University of California at Berkeley on "Philosophical Conceptions and Practical Results." His address appeared as the leading article in the University Chronicle for September. It was widely circulated, and pragmatism soon became a movement, the liveliest so far in American philosophy.²⁵

Though James gave him full credit, Peirce soon felt the need of restating his own pragmatism, both to distinguish it from James's and Schiller's and to correct certain errors and omissions in his original statement of 1877-78; above all to make fully explicit the semeiotic framework within which it had been worked out. Peirce held that his own strictly limited form of pragmatism was provable, and it was only within the semeiotic framework that the proof could be made evident.²⁶ With this in view, he gave two series of lectures in 1903, one at Harvard University in the spring, the other at the Lowell Institute in the fall.

In 1905 he began a series of articles on pragmatism in The Monist. In the first, "What Pragmatism Is," his own form of it was renamed pragmaticism (5.411-37 at 414). In the second, "Issues of Pragmaticism," two doctrines that he had defended before he first formulated his pragmatism back in the 1870s—namely, critical common-sensism and scholastic realism—were now treated as consequences of it. The chief novelty in this article is the semeiotic of vagueness, one of the characters of critical common-sensism (5.438-63 at 446-50).

These two articles were meant only to prepare the way for the proof of pragmaticism in a third article. But after the second had appeared, Peirce decided that the best way to present the proof was by means of his existential graphs. So he devoted the third article to further "Prolegomena" to the proof. These, as we saw in step (23) of section 6, were a restatement of his general theory of signs—the last he succeeded in getting into print—and a much fuller exposition of his system of existential graphs.

But, alas! Though there are drafts of a fourth article and promises of a fifth and sixth, the third was the last to reach print. One of the unfinished tasks of Peirce scholarship is to construct the proof, largely from manuscripts not yet published, and to show how the graphs would have functioned in the exposition of it.

In sheer volume, his writings on the theory of signs in the nine years from 1903 through 1911—many of them still unpublished—exceed those of the preceding forty years. The most striking features of these later writings are the high frequency of focus on pragmaticism and the development of a semeiotic realism out of the type-token distinction.

In any case, the semeioticians who were soon to begin thinking of Peirce as founder of modern semeiotic had in mind chiefly his published writings of this last period, rather than those of what I have called the first and second foundings.

Meanwhile a relevant change had taken place in Peirce's view of the relation between logic and semeiotic. I report that change in the following section.

8. BACK TO LOCKE: FROM LOGIC-WITHIN-SEMEIOTIC TO LOGIC-AS-SEMEIOTIC (1865-1911)

We have seen in section 2 that Peirce at first refused to follow Locke in identifying logic with semeiotic, and defined it rather as one of the three parts of a symbolistic which in turn was one of the three parts of semeiotic. By the mid-1880s, however, as we saw in step (10) of section 6, he had come to realize that logic requires indices and icons; that it cannot do business with symbols that are neither indexicai nor iconic. About 1894, in the chapter on signs in his only finished treatise on logic (the so-called Grand Logic), he argued that in all reasoning we must use a "mixture" of icons, indices, and symbols. "We cannot dispense with any of them" (Ms 404 p.46). So the symbolistic trivium became the semeiotic trivium, with logic as its mid-science, and Peirce was halfway back to Locke.

But we have also seen in step (20) of section 6 that by 1903 he had gone the rest of the way. Logic was now semeiotic, as Locke held, and what Peirce had previously called logic he now called Critic. When and how did his conversion come about?

It was a gradual transition rather than a conversion. Even on the second half of the way back there was an intermediate stage, beginning about 1896, in which Peirce was saying such things as : "The term logic' is unscientifically by me employed in two different senses" (1.444). "The word logic is ambiguous. It is at once the name of a more general science and a specific branch of that science" (Ms 751 p.l). During this two-sense transitional stage, logic in its narrow sense was the mid-science of the semeiotic trivium; in its broad sense it was general semeiotic, embracing all three sub-sciences. But even the narrow sense was by no means as narrow as that which Peirce had given to logic in what I have called the first founding.

The journey back to Locke was completed when in 1902 he gave up the narrow sense altogether, identified logic unreservedly with semeiotic, and adopted Locke's term Cntic for what he had most recently been calling "logic in the narrow sense" (NE 4.20f.). Since Critic in this sense is the critic of arguments (4.9), and since this may need to be distinguished from the critic of morals or of works of art or of craftsmanship, Peirce sometimes calls it Critical Logic (2.93); more often, Logical Critic (6.475). To one occurrence of the latter phrase, however, he adds "or let us say 'critic' simply, as long as we have to do with no other than the logical kind" (Ms 852 p.2).

It is important to note, however, that though logic is now wholly semeiotic, it is still not the whole of semeiotic. It is semeiotic variously qualified as cenoscopic²⁷ (Ms 499 p.[15]), formal (NE 4:20f.), general (1.444), normative (2.111), speculative (Ms 693 p.188). It is "General Semeiotic, the a priori theory of signs" (Ms 634 p.14); "the quasi-necessary, or formal doctrine of signs" (2.227); "the pure theory of signs, in general" (Ms L 107 p.24). In addition to cenoscopic semeiotic, there are, or may be, idioscopic studies of signs as various as the idioscopic seiences themselves—physical, chemical, biological, geological, anthropological, psychological, medical, musical, economic, political, and so on. None of these is any part of logic, though the reasonings they employ may be made matter for logical study. Take psychology for example.

Those, namely, of cenoscopic semeiotic.

The explanation Peirce most often gives of his move from logic-within-semeiotic to logic-as-semeiotic is in terms of the classification of the sciences.²⁸ This was always a concern with him, but increasingly so after 1890, from dissatisfaction both with the definitions of science and with the classification of the sciences that he had contributed to the Century Dictionary. He came to think of science no longer as knowledge already possessed or acquired and systematized, but as ongoing investigation, as what research scientists do; and therefore to identify a given science not with a particular body of knowledge but with a social group, a subcommunity of the larger community of investigators. As he wrote Lady Welby in 1908, "the only natural lines of demarcation between nearly related sciences are the divisions between the social groups of devotees of those sciences" (8.342). But of course, in attempting to place a given science, the classifier would consider not only what the subcommunities are severally doing at present, but what changes are in progress, and how far they are likely to go in the near future.

"A great desideratum," he wrote in 1909, "is a general theory of all possible kinds of signs, their modes of signification, of denotation, and of information; and their whole behaviour and properties, so far as these are not accidental" (Ms 634 p.14). The task of supplying this need must be undertaken by some group of investigators. Nearly all that had hitherto been accomplished in that direction had been the work of logicians. No other group was so well prepared to take on the task, or could do so with less diversion from its previous concerns.

For examples, though "a piece of concerted music is a sign, and so is a word or signal of command," and "logic has no positive concern with either of these kinds of signs," it must nevertheless "concern itself with them negatively in defining the kind of signs it does deal with; and it is not likely that in our time there will be anybody to study the general physiology of the non-logical signs except the logician," who is in any case "obliged to do so, in some measure" (Ms 499 p.[15]).

So it came about that the last of Peirce's major unfinished works, which he hoped would in the twentieth century have some measure of the success that Mill's System of Logic had had in the nineteenth, was A System of Logic, considered as Semeiotic (Ms 640 p.10; NE 3:875); considered, that is, not as the whole of semeiotic, but as the whole of cenoscopic semeiotic.

9. THE LOGIC OF MATHEMATICS AGAIN (1903)

At this point we return briefly to section 6, step (14). How could Peirce defend his father's definition so long and then so abruptly change it? Because of the change in his conception of science that we have just been tracing. As he put it in 1903,

10. VICTORIA LADY WELBY AND SIGNIFICS (1903-1911)

In May 1903 Victoria Lady Welby published What is MeaningP, had a copy sent to Peirce, and wrote him asking for criticism. He replied, and he reviewed the book in The Nation along with Russell's Principles of Mathematics. The correspondence thus begun lasted eight years, until her final illness.²⁹

Along with the rebirth of pragmatism, his having at last a responsive correspondent was almost certainly a factor in Peirce's concentration on semeiotic in the last decade of his life, from his Harvard and Lowell lectures of 1903 onward. It may also have been a factor in the directions this concentration took, and in its characteristic emphases. Some of his best expositions are in letters to Lady Welby, and among his last ereative efforts were drafts of a paper for a Festschrift in her honor.

After first trying sensifics, Lady Welby had adopted significs as the name for the field to which she was devoting the latter part of her life. She had contributed a brief article under that title to Baldwin's Dictionary in 1902. She later contributed a much longer one to the Britannica, in which she distinguished "three main levels or classes" of "expressionvalue"—"those of Sense, Meaning, and Significance." Peirce wrote her that these nearly coincided with his own division of interprétants (Will). And in a letter to James about the same time, he referred to her distinction in an illuminating passage on the sign-object and sign-interprétant relations, and on the relations between the two relations (NE 3:844).

Lady Welby wrote on December 4, 1908: "You have always been kindly interested in the work to which my life is devoted" (W65). Peirce replied on the 23rd:

Or, as he put it in a postscript not mailed, "when I have myself been entirely absorbed in the very same subject since 1863, without meeting, before I made your acquaintance, a single mind to whom it did not seem very like bosh" (8.376).

11. THE SOP TO CERBERUS (1908)

Responding to questions about his work in logic, Peirce wrote to Philip E. B. Jourdain on December 5, 1908:

Less than three weeks later, in his letter of December 23 to Lady Welby, Peirce wrote:

What was that broader, that more generalized, conception? Negatively, it is apparent that it did not involve "the mind of an interpreter" or "an effect upon a person." Did it also not involve an utterer, a signgiver? In the last account of his theory of signs which Peirce had published, as a framework within which to introduce his existential graphs, the place of the sign-utterer or sign-giver had been taken by the Graphist.

What, then, was the sop to Cerberus? If we recall that the original motive of subsuming logic under semeiotic in 1865 was to avoid basing it on psychology, we can give a tentative and at least partial answer. The sop to Cerberus was lapsing from sign-talk into psych-talk—from semeiotic into psychology. Since Peirce was himself an experimental psychologist, perhaps the first on the American continent, and once thought of giving up logic for psychology,³⁰ no disparagement of psychology is implied. Certainly it was no disparagement of psychology to place it lower than semeiotic in the classification of the sciences, just as it was no disparagement of semeiotic to place that below mathematics.

If we were attempting to give a more positive and complete answer, we might well begin with Peirce's 1902 application to the Carnegie Institution for a grant to enable him to write a series of thirty-six memoirs on logic; and more particularly with his brief descriptions of Memoirs No. 11, "On the Logical Conception of Mind," and No. 12, "On the Definition of Logic."

A few sentences from one of the drafts of the application offer further hints.

If that does not answer our question, it sets us off on the right track. But we return from it to pursue the question how Peirce came to be recognized as a founder of semeiotic.

12. OGDEN AND RICHARDS: THE MEANING OF MEANING (1923)

Almost from the beginning of their correspondence in 1903, Lady Welby gave her visitors accounts of Peirce's letters, and frequently enclosed copies of extensive extracts from them in her letters to other correspondents. On May 2, 1911, she wrote Peirce that she thought she had found a disciple for him in C. K. Ogden, then still a student at Cambridge University (W138-39).

In Peirce's letters to Lady Welby, one of the most striking passages, is that concerning his early reading of Whately's Elements of Logic. Ogden was so impressed by it that in The Meaning of Meaning in 1923 he and Richards made Whately and Peirce the culminating figures in the movement "Towards a Science of Symbolism"—the nominalistic movement from Ockham through Hobbes, Locke, Leibniz, Berkeley, Condillac, Horne Tooke, and Taine. (In order to pass directly from Whately to Peirce, they depart from chronology by taking up Taine before Whately.) They quote a passage from Whately's introduction in which he professes to know nothing of any universals but signs. Signs, he says, are the instrument of thought, not merely the vehicle of expression and communication. In any case, the only logic he understands "is entirely conversant about language" and other signs. It knows nothing of "abstract ideas" or of non-semeiotic mental processes. Ogden and Richards then say:

After misquoting the Whately passage from Peirce's letter to Lady Welby, they continue:

They do not call Peirce a nominalist, but they suggest that his "scholastic realism" and his exclusion of psychological considerations may account for a lack of clarity at certain points in a semeiotic that was otherwise the final upshot of the nominalistic tradition they have been sketching. In an appendix they offer a thirteen-page digest of his theory of signs in the form of extracts from his published papers (chiefly from the "Prolegomena" of 1906) and from three of his longer letters to Lady Welby, one of which contains the Whately and the "sop to Cerberus" passages.

The Meaning of Meaning was the first book in any language from which it was possible to get a grasp of Peirce's semeiotic at first hand, in his own terms. F. P. Ramsey, reviewing the book in Mind, rightly said that its "excellent appendix on C. S. Peirce deserves especial mention."³² (Ludwig Wittgenstein may have known something of Peirce through Ramsey.)

The authors misquote three passages from Royce's The World and the Individual. Had they also looked into The Problem of Christianity, its chapters on interpretation would surely have led them to Peirce's cognition series of 1868-69, in which the doctrine that all thought is in signs was most fully argued and developed.³³ This is much more fundamental than anything they do quote. Had they known of it, they would surely have asked themselves where Peirce got that doctrine, and would have given what is almost certainly the right answer: He got it from Whately at the age of twelve. But at least they were on the right track in approaching Peirce from Whately and from Whately's nominalistic predecessors. It is unfortunate that no other writer on Peirce's theory of signs has taken the same approach.

13. CHARLES MORRIS: FOUNDATIONS OF THE THEORY OF SIGNS (1938)

The movement variously called logical positivism, logical empiricism, scientific empiricism, and the unity of science movement, began in German-speaking middle Europe in the 1920s, started a westward migration in the 1930s, and for a time found its main resting place, at least in English-speaking countries, at Chicago. Its chief single monument is the International Encyclopedia of Unified Science, edited by Otto Neurath, Rudolf Carnap, and Charles Morris, and published by the University of Chicago Press. After an introductory monograph called "Encyclopedia and Unified Science" (by six authors—Neurath, Bohr, Dewey, Russell, Carnap, and Morris), its first systematic monograph was Foundations of the Theory of Signs by Morris in 1938.³⁴

The position of Morris's monograph in the Encyclopedia was no accident. It was the outstanding feature of the very design of the Encyclopedia. The foundations of the theory of signs were the foundations for the unification of the sciences.

Morris had studied under George Herbert Mead and had written his dissertation on Symbolism and Reality in 1925. He had been "helped to identify the contours of a general theory of signs by The Meaning of Meaning." ³⁵

The first six volumes of Peirce's Collected Papers, edited by Charles Hartshorne and Paul Weiss, had come out in the earlier 1930s; the first in 1931, the sixth in 1935. (Hartshorne was a colleague of Morris's at Chicago.) Morris acquired each of the six volumes as it appeared, and annotated it extensively.³⁶ There were semeiotic materials in all six volumes, but especially in the second and fifth. By the time Morris wrote the Foundations, therefore, he had examined a much more nearly adequate body of evidence for Peirce's theory of signs than had been accessible to Ogden and Richards. But the same evidence was now in the hands of many other students, and interpretations or criticisms of Peirce no longer passed unchallenged.

Morris had a student named Estelle Allen De Lacy, who wrote her dissertation in 1935 on Meaning and Methodology in Hellenistic Philosophy, giving prominence to Philodemus. She assisted Morris for several years in collecting materials for a history of semeiotic. This was never written, but she and her husband, Phillip De Lacy, edited and translated Philodemus's De signis. ³⁷

Morris's later work, Signs, Language and Behavior (1946), has an appendix with a section on Peirce, which begins: "Peirce was the heir of the whole historical analysis of signs and has himself had a major influence upon contemporary discussion." In this book, as in the Foundations, Morris rightly took off from semeiosis, but about the same time Dewey challenged his earlier account of "the pragmatic dimension" of semeiosis as "the relation of signs to interpreters." Morris replied to Dewey and other critics in 1948 in "Signs About Signs About Signs," which brought Peirce to the center of semeiotic controversy,³⁸ as Morris's two books had brought Peirce to the center of fresh construction.

Another student of Morris's, Thomas A. Sebeok, has become the most productive and influential semeiotician of the present day. A special field of his, which he will forgive me for spelling zoösemeiotic, is one that farmer Peirce entered now and then with his horses and dogs, but found no time to cultivate systematically.

14. THE GESTATION PERIOD (1851-1865)

Whately's Elements of Logic was studied in the spring semester of the junior year in Harvard College. In September 1851, when about to enter upon his junior year, Peirce's older brother Jem (James Mills Peirce) bought his textbooks for the year, including Whately. Charles, who was turning twelve that month, came into Jem's room, glanced at the new textbooks, and asked what logic was. Jem's answer led Charles to stretch himself upon the carpet there in Jem's room, with Whately open before him. As Charles wrote F. A. Woods in 1913, in a few days he got all the good he could out of it, "so that 6 years later when I was, with the rest of my class, required to answer at recitations on the book, I needed no more than a slight rereading of the lessons" (Ms L 477).

There was no other episode of his boyhood that Peirce so often recounted. In other accounts he speaks of himself as having "in a few days mastered that illuminating work" (Ms 905, canceled p.5), as having been "intent" upon reading it "on several days" (Ms 842[s]), as having "buried" himself in it (W85), as having been "delighted" with it (Ms 1606 p.ll), as "poring over" it (NE 4:vi); and in at least four other accounts as "devouring" it.³⁹

The logicians of Peirce's youth, however critical they were of partieular points in Whately, ascribed to him the revival of logic at Oxford and elsewhere after a century or more of stagnation. As early as 1833, Sir William Hamilton wrote that by the publication of Whately's Elements in 1826 "a new life was suddenly communicated to the expiring study," and that the decade in which it appeared had "done more in Oxford for the cause of this science than the whole hundred and thirty years preceding."⁴⁰ In 1854 George Boole, in the preface to his Laws of Thought, said that for "a knowledge of the most important terms of the science, as usually treated, and of its general object .... there is no better guide than Archbishop Whately's Elements of Logic," to which "the present revival of attention to this class of studies seems in a great measure due." Augustus De Morgan in his article on logic in the English Cyclopaedia in 1860 wrote that Whately possessed "the talent of rendering a dry subject attractive in a sound form by style, illustration, and clearness combined. And to him is due the title of the restorer of logical study in England." Peirce's Harvard teacher, Francis Bowen, had written in the North American Review for October 1856:

Besides the passages in his introduction from which Ogden and Richards quoted, Whately had a chapter criticizing realism, and treating conceptualism as a variant of it.⁴² He made the same distinction between fact and arrangement⁴³ to which Peirce appealed in two of the most nominalistic passages of his Popular Science Monthly series: the application of the pragmatic maxim to the conception "hard" and the comment on Gray's "Elegy" (5.403 and 409 at end; cf. 7.340). In later stages of his long progress from nominalism into realism, Peirce corrected or rejected these (5.453, 457, 545; 1.27nl, 615; 8.216).

To keep from sliding into realism unawares, Whately prescribed the prophylactic measure of using "description" when tempted to say "kind" or "nature."⁴⁴ Peirce never quite lost the habit so formed, in spite of having gradually become more and more of a realist (1.27n, 204, 549n; 5.127, 483, 486; 8.251).

When Peirce recited on Whately's Logic in the spring of 1858, it had been the Harvard logic text, and nominalism had been "the Cambridge Metaphysics,"⁴⁵ for a quarter of a century. But the Logic was not the only book of Whately's on which Peirce had to recite. In the first term of his freshman year, he recited twice a week on Whately's Lessons on Morals and Christian Evidences. In both terms of his junior year and perhaps also in his senior year he recited on Whately's Elements of Rhetoric, which had a passage advocating nominalism more vigorously even than the one that Ogden and Richards quoted from the Logic. Here is the latter half of it:

The rhetoric text in Peirce's sophomore year, George Campbell's Philosophy of Rhetoric, inculcated similar views.

While still in college, Peirce had in his private library at least two other books of Whately's. One was A Selection of English Synonyms, by Whately's daughter Elizabeth Jane Whately, revised throughout by Whately himself, who said in his preface that it was "very much the best" work that had appeared on the subject, but that

It was doubtless this book that prompted Peirce in October 1857, early in his junior year, to begin writing "A Scientific Book of Synonyms in the English Language" (Mss 1140-42).

Also in Peirce's private library was Whately's Historic Doubts Relative to Napoleon Buonaparte, a parody of Hume's scepticism concerning miracles. This was almost certainly the germ from which Peirce's theory of historical method developed (2.625, 634, 642, 714; 5.589; 8.194f, 380, 382; Mss 1319-20).

Peirce also read the nominalists that Ogden and Richards later reviewed on their way to Whately and Peirce. Take Horne Tooke, for example. On January 1, 1861, Peirce's "Aunt Lizzie" (Charlotte Elizabeth Peirce) gave him a copy of the 1860 edition of The Diversions of Purley .⁴⁸ Though Horne Tooke was a follower of Locke, his thesis was that everything Locke had said in terms of ideas should rather have been said in terms of words. Though Peirce did not jettison the language of ideas, even in the article in which his pragmatism was first put forward —"How To Make Our Ideas Clear" (1878)—he could write as late as 1896: "What do we mean by an idea being clear? It is not needful to inquire first what an idea is. We can dispense with the word idea, and can ask what we mean by attaching a clear signification to a word" (Ms 953 p.8).

Peirce frequently said in later years that it was the extreme nominalists such as Ockham, Hobbes, Leibniz, and Berkeley who had especially urged the doctrine that "every thought is a sign" (5.470), that "thoughts are signs" (4.582), that "Any concept is a sign" (8.332), but that there is nothing inherently or peculiarly nominalistic about the doctrine, and that "the realists are, for the most part, content to let the proposition stand unchallenged, even when they have not decidedly affirmed its truth" (4.582).

Of Peirce himself it may be concluded that he committed himself in youth to a theory of cognition which he knew to be prima facie nominalistic, and that he at first conceived himself to be a nominalist in so doing; but that, step by step over a period of forty years or more, beginning in 1868, he transformed that nominalistic doctrine into a more and more realistic one.

In any case, he remained a nominalist throughout what I have called the gestation period of his semeiotic.⁴⁹

15. A FOURTH FOUNDING? (1976-)

Peirce's Letters to Lady Welby came out in 1953. Volumes 7-8 were added to the Collected Papers in 1958, both containing further materials on the general theory of signs, including a long draft of a letter not sent to Lady Welby and therefore not included in the 1953 edition (8.342-79). A microfilm edition of the Peirce manuscripts at Harvard University became available in 1964, and a Catalogue of them came out in 1967. The first of four volumes of his Nation reviews appeared in December 1975. Carolyn Eisele's The New Elements of Mathematics by Charles S. Peirce (1976) consists almost entirely of papers not previously published, and much more of this new material is relevant to semeiotic than would be guessed from the title, from the indexes, or from a casual glance through the four-volumes-in-five. A microfiche edition of the papers Peirce himself published appeared in 1977. An edition of the Peirce/Welby correspondence by Charles S. Hardwick, containing Lady Welby's letters to Peirce as well as his to her, appeared later in 1977. Several anthologies of Peirce's writings on semeiotic, both in English and in translation, are being prepared.

It remains the case, however, that Peirce's still unpublished writings on the theory of signs exceed in quantity those that have so far been published. A new and more comprehensive edition of his writings is now in preparation, to be arranged chronologically in fifteen or more volumes to appear over a period of ten or more years, plus a two-volume biography and a volume of bibliographies and indexes. The semeiotic materials appearing for the first time in this new edition will exceed in quantity those which first appeared in the eight volumes of Collected Papers.

There is already an extensive body of secondary literature, some of it purely expository, some of it critical; some of it continuing where Peirce left off; some of it inspired in part by Peirce but making no attempt to distinguish Peircean from non-Peircean elements in the new constructions in progress.

The continuing confusion of tongues in the semeiotic tower of Babel is such that, for some time to come, it will be worthwhile for semeioticians and Peirce scholars to study the new materials as they become available, and to attempt some of the unfinished tasks of Peircean semeiotic scholarship. Eight of these occur to me as worth mentioning here.

(1) Most needed, and perhaps even a prerequisite to the rest, is an annotated bibliography of Peirce's own relevant writings, published and unpublished, followed by a bibliography of the secondary literature and by a lexicon that quotes Peirce's best definitions or explanations of the terms he uses and that gives references to other relevant passages in his writings and in the secondary literature.

(2) The present paper has briefly shown how Peirce's lifelong study of the logic of mathematics was conducted throughout within the framework of semeiotic. This is worth showing in greater detail. But Peirce's work in the logic of mathematics was for the sake of his more extensive work in the logic of the positive sciences, and it remains to be shown how that also was conducted throughout within the same framework.

(3) Peirce said that the proof of pragmaticism on which he embarked in his Monist series of 1905 was "the one contribution of value" that he had still to make to philosophy, "For it would essentially involve the establishment of the truth of synechism" (5.415). What, in full, was the unfinished proof? How did the theory of signs and the system of existential graphs function in it? And how would it establish the truth of synechism?⁵⁰

(4) What, more exactly, was the "sop to Cerberus"? And what, more exactly, was that broader, that more generalized conception of sign that Peirce despaired of making understandable and understood?

(5) Suppose that Peirce had succeeded in writing A System of Logic, considered as Semeiotic, or rather suppose that he were writing it today, in full knowledge of developments in logic and in semeiotic since his time. What would be its distinguishing features? Imagine the System already published, and a competent critic writing a careful review article on it. How would the article go?

(6) As an approach to (5), consider that nearly everything that has so far been written about Peirce's general theory of signs belongs to the first of the three parts of the semeiotic trivium, leaving the second and third empty. But Peirce said his hardest and best work had been done on the third (NE 3:207).⁵¹ Interpreters and critics of his pragmatism and of his theory of the economy of research, for examples, have either detached them from semeiotic altogether or have failed to assign them properly, as he did, to its third part, Methodeutic, as presupposing the second, Critic. To what extent has our understanding of them been thereby vitiated? What were his other contributions to Methodeutic? And how about Critic?

(7) What were the steps by which Peirce passed from a nominalistic to a more and more realistic general theory of signs? "Everybody ought to be a nominalist at first, and to continue in that opinion until he is driven out of it by the force majeure of irreconcilable facts" (4.1). What was the force majeure at each step of the way?⁵²

(8) Among the recurring topics in Peirce's writings, early and late, are "first impressions of sense" and "immediate perception." It is perhaps obvious enough that the sign theory of cognition entails rejection of the former. It is less obvious that it entails acceptance of the latter. But as late as 1905 Peirce not only claimed to have adhered from the beginning to the doctrine of immediate perception, as held by Aristotle, by Reid and Hamilton, and by Kant in his refutation of Berkeley (8.261), but said that in his own case it was viewing logic as semeiotic that led "at once" to this doctrine.⁵³ These matters are worth arguing out in detail, and our understanding of Peirce will remain imperfect until that has been done.

If the new materials becoming available are as illuminating as the old, and if oncoming semeioticians and Peirce scholars carry out such tasks as these, and others not less fundamental, may we not look for a fourth founding before the end of the century?

It is my belief that such a fourth founding has already begun.

NOTES

1. Imagine a small boy for whom shaping his letters is still fun. One day he draws a big О that pleases him, and he proudly calls to his older sister, "See my O, sis!"

2. "See my O, see!"

3. The boy puts a pair of eyes in his big О and says, "See, my О sees!"

4. The boy draws another big О with a quivering or zigzag line and says, "See my О tick! Hear the clock tick, but see my О tick!" See further Thomas A. Sebeok, " 'Semiotics' and Its Congeners," in his Contributions to the Doctrine of Signs (Bloomington: Indiana University and Lisse: Peter de Ridder Press, 1976), pp.47-58; and Luigi Romeo, "The Derivation of 'Semiotics' through the History of the Discipline," Semiosis 6:37-50 (1977).

5. Peirce in a letter to his father without date but about April 15, 1877, in Ms L 333.

6. Letter to J. E. Hilgard, Secretary of the Academy, August 6, 1877, in the C. S. S. Peirce folder in the Archives of the National Academy of Sciences.

7. Max H. Fisch and Jackson I. Cope, "Peirce at The Johns Hopkins University," in Studies in the Philosophy of Charles Sanders Peirce, edited by Philip P. Wiener and Frederic H. Young (Cambridge: Harvard University Press, 1952), pp.277-311.

8. He uses the singular representamen once, in 1.557 (1867). The plurals representamina and representamens do not yet occur. He uses both later. By 1904 he has dropped this term, but he picks it up again at least once, in 1911 (Ms 675). He continues to use represent and representation, but seldom technically.

9. In 1.555 Peirce places these three categories between Being and Substance, making five in all; but he makes no use of the first and fifth as categories. By the time he wrote his 1870 paper on the logic of relatives, it was evident to him that, in any sense in which the central three are categories, the first and fifth are not; and they never reappear as such after 1867. They are hardly ever even mentioned in connection with the categories. In at least one account, however, Peirce explicitly says that "Being and Substance are of a different nature" (Ms L 75, Carnegie Application, "Statement," p.4 of longer draft with that heading). The best account of this matter is still that by Manley Thompson in The Pragmatic Philosophy of C. S. Peirce (Chicago: University of Chicago Press, 1953), pp.29-36. (Other questions concerning Peirce's theory of categories are dealt with in my essay "Hegel and Peirce," in Hegel and the History of Philosophy, edited by Joseph J. O'Malley, Keith W. Algozin, and Frederick G. Weiss [The Hague: Martinus Nijhoff, 1974], pp.171-93, at 173-78.)

10. The last clause is here corrected from an errata list not found by the editors of the Collected Papers.

11. Josiah Royce, The Problem of Christianity (New York: The Macmillan Co., 1913), vol. 2, pp. 107-325. See especially vol. 1, p.xi, and vol. 2, p.114. See also the last paragraph of section 12 below.

12. Popular Science Monthly 12:1-15 (November 1877), 286-302 (January 1878), 604-15 (March 1878), 705-18 (April 1878); 13:203-17 (June 1878), 470-82 (August 1878). Collected Papers (with later revisions and notes) 5.358-87, 388-410; 2.645-60, 669-93; 6.395-427; 2.619-44.

13. The first two papers appeared also in French in the Revue philosophique 6:553-69 (December 1878), 7:39-57 (January 1879), under the titles: La Logique de la Science, Première partie: Comment se fixe la croyance; Deuxième partie: Comment rendre nos idées claires.

14. To be sure, there are differences, but they might not strike a reader unacquainted with Bain's theory of belief. See Max H. Fisch, "Alexander Bain and the Genealogy of Pragmatism," Journal of the History of Ideas 15:413-44 (1954), at 438-40.

15. See Max H. Fisch, "Peirce's Arisbe: The Greek Influence in his Later Philosophy," Transactions of the Charles S. Peirce Society 7:187-210 (1971), at 190-91, 203.

16. Next in importance was Augustus De Morgan's fourth memoir on the syllogism, in 1860, which opened up the logic of relations (NE 4:125) and elaborated the syllogism of transposed quantity (4.103). A distant third was Sir William Hamilton's quantification of the predicate (1.29) and the controversy to which it gave rise (2.532-35).

17. "Substitution in Logic," The Monist 15:294-95 (1905), signed by Francis C. Russell but written by Peirce. See further steps (3), (19), (20), and (22) below. See also note 31 below.

18. American Journal of Mathematics l:126n (1878).

19. Ibid., 4:97-229 (1881) ; for errata see p.iv.

20. Johns Hopkins University Circulars 1:178-80.

21. For an explanation of this step, see section 9 below.

22. See Don D. Roberts, The Existential Graphs of Charles S. Peirce (The Hague: Mouton, 1973).

23. Cf. Ms 530, "A Proposed Logical Notation." (An American psychologist and logician, Shea Zellweger, is about to publish a new notation for the same connectives, which he calls "the logic alphabet." He accepts four of Peirce's criteria for a good notation, and follows Peirce in calling two of them iconicity and cursiveness. The other two, in substance contained in Peirce's box-X, he calls frame consistency and eusymmetry. Although his logic alphabet differs from Peirce's notations, he conceives his own notation as directly continuing Peirce's work.)

24. See step (1) above and steps (20) and (22) below. Peirce submitted a long article on "Mathematical Logic" for the same Dictionary, but Baldwin printed only the first five words of it and the appended bibliographical note. The article included a five-step analysis of the mathematician's procedure. The account of the peculiarities of mathematical reasoning ended: "Of still greater importance is the practice of making operations and relations of all kinds objects to be operated upon" (NE 3:742-50 at 749f.).

25. Max H. Fisch, "American Pragmatism Before and After 1898," in American Philosophy from Edwards to Quine, edited by Robert W. Shahan and Kenneth R. Merrill (Norman: University of Oklahoma Press, 1977), pp. 77-110.

26. Max H. Fisch, "The 'Proof' of Pragmatism," in Pragmatism and Purpose: Essays Presented to Thomas A. Goudge, edited by John G. Slater, Fred Wilson, and L. W. Sumner (Toronto: University of Toronto Press, forthcoming).

27. For Peirce's use of Bentham's cenoscopic-idioscopic distinction, see 1.241f., 8.199.

28. For Peirce's classification of the sciences as of 1902-1903, see 1.180-283. This is presented in tabular form between pages 48 and 49 of Thomas A. Goudge, The Thought of C. S. Peirce (Toronto: University of Toronto Press, 1950). See also the illuminating Ph.D. dissertation by Beverley E. Kent, Logic in the Context of Peirce's Classification of the Sciences (University of Waterloo, 1975); Dissertation Abstracts 36:2899-A, November 1975).

29. Charles S. Hardwick, ed., Semiotic and Signifies: The Correspondence between Charles S. Peirce and Victoria Lady Welby (Bloomington: Indiana University Press, 1977).

30. Fisch and Cope (note 7 above), p.292.

31. C. K. Ogden and I. A. Richards, The Meaning of Meaning (London: Kegan Paul, Trench, Trubner & Co.; New York: Harcourt, Brace & Co., 1923), p. 125, referring to the article cited in note 17 above.

32. Mind 33:109 (1924).

33. See sections 2 and 3 above.

34. Two hundred and sixty monographs were contemplated, to be collected in twenty-six volumes, but World War II and Neurath's death intervened. The ten monographs of Volume I were collected in 1955. When nine of the ten for Volume II were ready in 1969, it appeared along with a reprint of Volume I under the title Foundations of the Unity of Science: Toward an International Encyclopedia of Unified Science. The remainder of the project was indefinitely postponed. Meanwhile the monographs of both volumes had been appearing singly, as they became ready, beginning in 1938. Morris's monograph appeared in that year as Volume I, Number 2.

35. Charles Morris, Writings on the General Theory of Signs (The Hague: Mouton, 1971), p.7. (The Foundations and other writings mentioned below are reprinted in this volume.)

36. Morris has recently given a collection of his correspondence and other papers and a part of his library, including these volumes, to Indiana University-Purdue University at Indianapolis.

37. Philodemus: On Methods of Inference; A Study in Ancient Empiricism (Philological Monographs published by the American Philological Association, Number X, 1941). Estelle Allen De Lacy, "Meaning and Methodology in Hellenistic Philosophy," Philosophical Review 47:390-409, 1938.

38. Philosophy and Phenomenological Research 9:115-33 (1948). On the controversy between Morris and Dewey about Peirce see Max H. Fisch, "Dewey's Critical and Historical Studies," in Guide to the Works of John Dewey, edited by Jo Ann Boydston (Carbondale: Southern Illinois University Press, 1970), pp. 306-33 at 330-32.

39. Mss 842 p.7, 848 p.9; W77; letter to Samuel Barnett, December 20, 1909, in Emory University Library.

40. Sir William Hamilton, Discussions on Philosophy and Literature (New York: Harper & Brothers, 1860), p.126.

41. In 1864 Bowen began the preface to his own Treatise on Logic (Boston : John Allyn) : "The revival of the study of Logic, at least in England and America, as an important element of a University education, dates only from the publication of Dr. Whately's treatise on the subject, little over thirty years ago."

42. Elements of Logic (New York: Harper & Brothers, 1856), Book IV, ch. V, pp.294-303, at p.299*.

43. Ibid., pp. 296f.

44. Ibid., pp. 294*, 298, 302, 347.

45. Amos Bronson Alcott, writing to William Torrey Harris on April 2, 1868, and commenting on "Nominalism versus Realism" (Journal of Speculative Philosophy 2:57-61, 1868 [6.619-24]), says: "I take the author ... to be the son of the Cambridge Mathematical Professor, and speaking the best he has for the Cambridge Metaphysics. . . ." (quoted from a typewritten transcript sent to me by Harris's daughter, Edith Davidson Harris, in 1949. She gave her Alcott-Harris collection to the Concord Free Public Library in 1952, but the entire collection, including this letter, has been missing since 1960.) In a draft of this letter among the Alcott Papers in the Houghton Library of Harvard University, Alcott puts it as follows: "I take Peirce to be the son of the Cambridge mathematics Professor, and perhaps defending as he best can the Professor's metaphysics, if not of the College." So Benjamin Peirce, too, was understood to be a nominalist.

46. Elements of Rhetoric (London: John W. Parker, 1846), pp.20f.

47. Editor's preface to Elizabeth Jane Whately, A Selection of English Synonyms, 4th ed., rev. (London: John W. Parker and Son, 1858). Peirce's own copies of four of the five Whately books are listed in Ms 1555. All are of editions published in Cambridge, or Boston and Cambridge; the Napoleon in 1832, the Logic, Rhetoric, and Synonyms in the 1850s.

48. The call number of this copy in the Harvard University Library is 9265.11.

49. Max H. Fisch, "Peirce's Progress from Nominalism toward Realism," The Monist 51:159-78 (1967), shows that he was espousing nominalism under that name in 1867, well into the period of what I have called the first founding. (Among the defects of this article are that it pays too little attention to the theory of signs, fails to mention Whately and Harvard nominalism, and ignores the gestation period altogether. Moreover, its title was misleading. It was short for "Peirce's progress toward that degree—or that extremity—of realism which he eventually reached." But it was also meant to leave room for a subsequent paper, not yet written, on "Peirce's Lifelong Nominalism.")

50. See section 7 above and note 25.

51. Lines 9-11 of p.207 should read: "The third branch of logic is Methodeutic which shows how to conduct an inquiry. This is what the greater part of my life has been devoted to, though I base it upon Critic." The preceding sentence was left unfinished. From another draft of the same letter (Ms 231) : "In my own feeling, whatever I did in any other science than logic was only an exercize in methodeutic and as soon as I had the method of investigation thoroughly shown, my interest dropped off." From an earlier draft letter to William James (NE 3:874): "I have done a lot of work in Methodeutic that is valuable and very little of it is printed. This will be the most widely useful part of my Big Book."

52. See section 14 above and note 49.

53. Draft letters to James, July 22 and 26, 1905, in Ms L 224.