To Elements of Semiotics

    This book has six parts. Each part is proceeded by an introduction that sets its outline. Part I is itself introductory, and the plan of the whole book is sketched at its end.

    Perhaps it will serve the reader to know something of the history of this book, an on-and-off project of 15 years. I teach music theory in a university music department, all I was trained to do, but this book concerns general semiotics--not music, except by example. By telling how this came to be I will be able to thank some friends and colleagues for their ideas or support and also to indicate the orientation of my work.

    I entered Canada with my wife in 1965 to avoid the Vietnam War, and we spent some years quite isolated, without academic contacts. In 1968 a clipping from the New York Times sent me off to find Noam Chomsky's Aspects of Syntax. Its effect on me was explosive. Without having the simple good sense to hunt up background reading, I struggled to decode its jargon and master its logic. I nearly memorized it, attracted at the outset by what I mistook for an obvious parallel between Chomsky's description of sentences and Heinrich Schenker's descriptions of musical compositions. Shortly after my first academic appointment, I devised a grammar (programmed by James Gabura) exploiting those "parallels" to compose melodies with a computer.

    The output didnít include any hit tunes, but in 1972, after I read a paper on this work to an ethnomusicology conference, a big man in a blue suit tapped me on the shoulder and informed me that I was committing semiotics. It was not a policeman, but Jean-Jacques Nattiez, the leading animateur then and for further decades of the semiotics of music. Thus accused, I thought I should find out what semiotics might be. On his clue, I looked up the Toronto Semiotic Circle. Paul Bouissac was its leading spirit. His projects over the years provided a mentorship that remains a splendid impetus. Having attended a couple of exciting and confusing meetings of the circle, I asked Professor Bouissac how one could get the basics of this field. To my great fortune, he told me no one had figured that out yet. He suggested--as the best place to start--that I read Ferdinand de Saussure's Course in General Linguistics, which I did, and invited me read a paper on it for the circle with application to music (which I also did--on a song of Schubert). The host of fascinating and helpful people I met through the circle included David Savan, whose work soon had me thinking that the best place to start must be Charles Pierce, and Lubomir Dolezel, by whom I felt persuaded one must really start with the Prague Circle. President Carter's amnesty, in 1977, allowed me to attend (by coincidence in my hometown, Denver, Colorado) the second annual meeting of the Semiotics Society of America. There must only have been about 30 of us, but we spanned all generations and seemed to each other to be terribly bright. It was exciting to find art historians, film critics, philosophers doing it too, even if none of them knew beans about music. (Modestly enlarged, the society continues to attract a diversity of professions--even musicologists--and to inspire fascinating work. I am very indebted to its collegiality.)

    At that time and for some years my interest in semiotics was its bearing on music. By 1980 I was ready to set the world straight. The Toronto Semiotic Circle published my monograph, Structure and Significance in Music, billed too ambitiously as the first of two volumes, and then I got stuck.

    Added together, my bits of Peirce, bits of Saussure, bits of Prague semiotics, and bits of much else just made too much ad hocary. How could there be an intelligible semiotics of music if one did not start with a unified theory of semiotics in general, comprehensive, non-contradictory, and economical, as Louie Hjelmslev demanded? Umberto Eco had tried to pull it all together, but had produced, I then thought, more of a critical compilation than a synthesis with clear primary principles. How could it be otherwise? He was grounded in literature. Wasn't the job of a truly general semiotics--a theory for all signs and media--hopeless unless someone would try who was not from the start sufficiently distanced from linguistics and verbal art? Surely the job for a musician! Well, it happened that I needed to spend a summer in a New York suburb in 1983 while my daughter received medical treatment, and we were in circumstances that were impossible for doing music but great for doing words. Nearly a whole summer, just what it would take to write out my neat, clear, general theory. The draft didn't quite tally. This book, several interruptions and second thoughts later, is what I came up with.

    This is a book that attempts to draw the foundational problems of semiotics into a unified focus. Because its ideas emerge in part through dialogue with previous writings, it provides a cursory and biased introduction to semiotic literature, but it is not a history. It offers neither a comprehensive nor a balanced survey of that literature. As a book that formulates a basis for semiotics, I hope it will be useful to readers new to the subject, but it is not a textbook.. It pretends to no such anonymous certainty.

    It will seem oldfashioned to some readers that I look for unity or foundations. I cannot accept the grandly reductive idea that such ambitions ignore a great international dateline that puts us in a new "postmodern" age. When I started the work, I did not know what all my graduate students know today: that Hilary Putnam says foundations are out of the question, that Jacques Derrida shows us the foundations always weep, that Hans-Georg Gadamer warns we cannot build castles in the air and must shore up best we can the foundations of our ancestors. Yet the alternative of blind faith in a consensual discourse risks too much that our terms will be hostage to fashion. I would prefer that we understand construction and deconstruction as a dialectic and not a brand-new one. Deconstruction is liveliest when it targets formulated systems, and in this sense it depends on them. The salute my title directs to Euclid is no more old-fashioned than would be one in the opposite direction, say to Plato's Symposium, where Socrates and Alcibiades can protest too much that their matter is independent of their manner, only because the dependence of thought on figures has been so blatantly problematized. Those who find system is inherently totalitarian need to think more about the politics of exaggerated chaos. Those who argue that systematic thinking is inherently supportive of male dominance could have argued the same about voting 100 years ago. Better to share these resources than reject them.

    In the end, I donít think you will find my constructions excessively weighty. My ideal for universals is Snap-On tools, portable and adaptable, not the lever with which Alchemides offered to move the world. More often than not my elements do not arise from full definitions but from exploratory distinctions within categories already established by common speech. These are not rigorously formal but lend themselves, I believe, to the pragmatic test (in the pure Peircean sense of the term) of having intellectual consequences that discipline their continuing development.

    A recurring conceit in this book is that I avoid philosophy. Credit me please for knowing well that in reality I have no special dispensation. In the rest of my life, I accept Alfred North Whiteheadís dictum that there is no escaping metaphysical first principles. What not doing philosophy means, in the present context, is sometimes doing some philosophy shoddily, sometimes closing my eyes, sometimes taking evasive tactics, and then showing if I can that it doesnít matter. One thing Charles Morris got right: Every special study depends on pretending that a range of neighboring problems are unproblematic. My hope is to demonstrate that foundational semiotics has rich problems which do not require broad philosophical agreement among those who cooperate in investigating them. Our situation is parallel to that in sociology or psychology. Each field has substantial issues of its own. The issues of the philosophy of mind, the philosophy of society, and the philosophy of the sign are not entirely separate from their specialized offshoots, but they are enormously different. I hope I convince my readers that the tasks of semiotics are no more tasks of philosophy than they are tasks of linguistics, both of which have vital relations with it.

    By comparison to other theories of semiotics, what I might list on a patent application for this one would include its emphasis on a comparative perspective; its category of "elaborated signs" and some contributions to their structural analysis; its principle that semiotics emerges as a distinctive, unified, and insightful study when (and only when) its purview is strictly limited to consciousness; a definition of sign derived from Peirce's but radically revised to evade his metaphysical commitments; and, above all, an insistence that structure both develops and opposes reference.

    For most of the time of my professional career, theoretical semiotics has been largely a contest between structuralism and the broader philosophical tradition that we like to see as centering on Peirce. In his honor I use the label "pragmatism" for this alternative to structuralism, but I use it lightly, only as a label, without regard to its implications in other contexts. This received opposition sets a theory of significant structure against a theory of reference. The position I take in this book is that the dialectic does not properly belong to semiotic theory but to the use and development of signs themselves. In fewer words, the dialectic of structure and reference is constitutive of semiosis, not semiotics. "Wait there," my good readers will say "You canít patent that! The idea is ubiquitous." And you are right in some measure. The idea is everywhere, in Prague School aesthetics, in the notion of "constructs," in Julia Kristevaís opposition of "significance" and "signifiance", in the postmodern critique of semiotics, and elsewhere, but only, I must observe, by implication or in fragments. The principle that semiotic elaboration both develops and supplants reference, as a primary, even "universal," constitutive principle of semiosis, has not been squarely confronted. If I canít patent the idea, it is because I havenít confronted it well enough. There is much more to say on the matter than I do.

    Pragmatism provides tools to describe the sign as open. Structuralism provides the tools that describe closure. Far from exhausted, structuralism is a still new venture that begs further development in the face of many unsolved problems. My discussion of comparative articulation and my distinction between the syntax of structured objects (pattern) and the syntax of languages ( grammar) are intended to remove roadblocks and keep that ball in play. . . .

To Table of Contents



    Routinely, we hear that semiotics is the study of the sign and that a sign is something that stands for something else. These definitions are scant clues to the origins and motivations of semiotics as a characteristic intellectual movement of the twentieth century.

    Charles Peirce wanted to find out how the progress of knowledge was possible and why it was occurring. Ferdinand de Saussure wanted a standpoint for a disciplined analysis of language, as Charles Morris did for the social sciences. Roman Jakobson wanted to show that the logical patterns he found in phonology unified all manifestations of culture, inspiring Claude Levi-Strauss to explicate structures common to scientific and prescientific thought and Tom Sebeok to insist on the unity of intellectual and biological processes. Umberto Eco shows us the life of thought in avant-garde culture. My own original motivation had been to understand how music without words could seem to mean something and could seem to advance its meaning phrase by phrase, but unraveling the logic behind that puzzle became a motivation in itself.

    The premise each these authors seized on was that their investigations hinged on a fuller conceptual mastery of the notion of sign. That there are conceptual problems is almost immediately obvious. What realm of existence do signs inhabit? Are they like houses and roses? Like the facts of arithmetic? Like images in the imagination? Or something quite other than these? What does it mean "to mean" or "to refer" ? These are not questions that philosophy had neglected before the twentieth century nor that it neglected during the twentieth century, when, in fact, they emerged as a chief theme. The interaction of this philosophical concern with a host of specialized studies was the novelty.

    Part I is introductory but puts in play some notions and prejudices that are developed throughout the book. Chapter 1 regards our problems from the vantage of ordinary language. Chapter 2 traces the emergence of semiotics in our self-reflection. Chapter 3 is a cautionary sampler of historical sources. Chapter 4 situates the investigation to follow, closing with an overview of the book.



    Systematic structure is so nearly ubiquitous in signs that it may be more informative to begin with counterexamples: the heart figure, ©, and the nearly extinct barber pole. In a former time, that red-and-white spiral was part of a loose family of "logos," guild emblems, that identified trades, but each of these signs was utilized independently. They did not combine into larger "sentences" or depend on each other for identity or meaning. Some vocabularies, such as letters or words or notes, build combinations; the units in others function independently.

    Our main traffic light signals in Toronto are red, green, yellow, and flashing green, a simple vocabulary. No larger signifying units arise by combining two or more signals. The commands a dog knows might seem to make a system to the trainer, but to the dog, perhaps each one is an independent, self-contained unit, with a few exceptions perhaps for some complex sequences. Similar signs do not necessarily comprise a vocabulary. Beepers, bells, and buzzers, as for example in doorbells, telephone rings, clock alarms, and belt pagers, are similar but do not comprise a systematic vocabulary. In contrast, the meaning of a sentence is not just the sum of the meanings of its words. The meaning of a total position in chess is more than the sum of the meanings of the position of each piece.

    To begin the study of semiotics on the side of system rather than on the side of signification, as we do here, is partly arbitrary. To begin the study of semiotic system with linguistics, rather than with Vetruviusís analysis of architecture, axiomatics in math, harmony in music theory, or the instructions of the ten bamboo school of Chinese painting may be arbitrary a second time. Language is the most domineering species in the "semiosphere" (Lotman, 1976) much as Homo sapiens is in the biosphere, but biology does not need to start with humans.

    But the choice is only partly arbitrary. Before Saussure's systematic account of language we had no practical principles to compare the structures of the other systems with each other. Although semiotics draws on ancient philosophical currents, its twentieth century flowering was strongly characterized by initiatives in linguistics and literary studies. This is not the historical current of thought that provides the best conception of the sign. In that respect, philosophy offers more than linguistics. But it was ideas of system emerging from linguistics that most dynamically connected the philosophy of signs to the concrete data of culture. Chapter 5 reviews the notion of system that emerges from Saussure's Linguistics. Chapter 6, considering counterproposals aiming at a more comprehensive view of language than Saussure's, argues that these can not supplant his. Chapter 7 enlarges the idea of system to encompass grammar with its psychological entailments. Chapter 8, after reviewing a radical rebuttal of the postulate of systems, provides a framework for crediting their reality.




    We have so far acquired partial ideas of semiotic system and sign system. These ideas will prove essential to understanding signs but do not tell us what signs are. Saussure's conception of the sign as a signifier correlated with a signified is inadequate to give us a handle on the behavior of signs in mental life. What it leaves out is the angles. The sign is biased. Representation is of something as something. Turkey and pumpkin pie signify Thanksgiving as festive. The cross and the lamb both refer to Jesus but as a martyr and as a fount of gentle love respectively. When the lamb in the Psalms refers to the Jewish people, it refers to them as vulnerable. Chapter 3 took note of the sign's bias via brief references to Gottlob Frege and Charles Peirce. Even to call an object a "book" is to focus on its potential to be read rather than its weight or color or inflammability. We refer to it, in Peirce's phrase, "in some respect or capacity."

    Chapter 9 acknowledges Peirceís schematization of these relations in the context of a comprehensive philosophy. Chapter 10 develops a method of speaking about aspects of mind, a deliberately minimal method, intended to orient semiotics. Chapter 11 employs this method of speaking about mind to construct a schema of the sign derived in its form from Peirce but independent of his philosophical commitments. Chapter 12 considers the place of the sign in mental life, its boundaries, its correlation with consciousness, and its embodiment in the physical world, drawing our study of mental life out of psychology and into the world of signs.



    The elaboration of signs is fundamentally paradoxical. As signs become more elaborate, they tend to lose or loosen their hold on their objects. It is tempting to say of a particular fugue or a particular algebra, "This is not a sign; it does not refer to anything. It is only significant in itself for what it is." But that is not the whole truth of algebras or fugues. First, on study, we would find that these elaborate things arise out of simpler ones that were more clearly referential. Algebra arose from the same arithmetic that recorded commodities, and music appears in its oldest genres to be implicated in referential contexts of poetry, religion, or work. Second, the ties are not severed. When I say "arises," I don't have in mind either the history of those arts per se, nor their current forms of production per se, but something in the history of which we still find a living trace in current production. Third, and very important, the initial objection has no bounds. Can't we say that a thorough philosophy defines all its own terms, making a closed system that doesn't really refer to anything outside itself? Can't we say language is closed the same way? Don't we like to say "Nature follows art," to point out how we see the world in terms that artists have taught us, closing off the visual world as tightly as algebras and fugues? Is not every culture a closed system of representation, representing nothing outside itself except as a negation (an "other")?

    Closure approaches in all these cases because elaborated semiosis manifests a competition between structure and reference. This competition, or dialectic, has its moments of truce and synthesis, most obvious in language and perhaps also in visual depiction, allowing the elaboration of structure, if it is sufficiently constrained by rules, to subserve reference. competition is the general rule.

    Alert to this dialectic, we can return to structuralist semiotics not as an alternative to pragmatism but as its supplement. Sense is not the whole of meaning, but we do make sense of signs by grasping their structure.

    Frederick Jameson's study, The Prison House of Language, regards structuralism as any thought that takes as its model Saussure's linguistics or the new sciences of Sigmund Freud, Karl Marx, and Emile Durkheim. His deconstructions reveal problematic consequences of the premises of these authors.

    We have already sampled (in chapters 6 and 8) the dialogue that Jameson's study and the "poststructural" critique generally extend. To study structure is to examine an abstraction. When you make an abstraction, something is left behind. That something neglected is likely to turn out to be significant in another context. We are exploring worlds in which abstraction is of limited efficacy; yet abstraction remains an essential bootstrap. There can be no final verdict in the contest between studies of structure and studies of context. Structuralism is attractive when it makes progress in its own task. Structuralism has attracted exaggerated criticism in recent decades not because its task, like all good tasks, is limited but because structuralism got stuck.

    What is the task? Let us compare a mathematical and a semiotic perspective of what structure is.

    Whitehead and Russell defined "having the same structure" through the arithmetic of relational numbers in their Principia Mathematica. As Russell himself describes it in the retrospection of My Philosophical Development (1959), the key idea here is correspondence or mapping. Imagine a cloud of helium gas with its molecules randomly arranged, and imagine another cloud of hydrogen gas with the same number of molecules identically arranged or perhaps identically except upside down and with all the distances halved. The one-to-one correspondence, not just between the molecules of the two clouds but between all their relationships, is identity of structure. Russell's favorite examples of structure include the analogical relation between a sound and the phonograph record that plays it. The first is a dynamic, temporal arrangement--in air; the second, static and spatial--in vinyl. There is a sense in which they have the same structure.

    In this mathematical view, structure is a domain of similarity and difference, which, like color for the blind, has no substance of its own.

    Structure for semiotics is both more concrete and more elusive. It has to do with our consciousness of order and organization. We tend to reify structure, speaking not of "x and y having the same structure" but of "the structure of x." In structuralism we attempt to objectify an aspect in perception and thought that gains its salience for psychological, cultural, and/or logical reasons. No holds barred. We cannot expect a quest for the essence of structure to lead to a tidy, homogeneous formula. So far we have some clear ideas about certain types of structure and some vaguer ideas that there must be others. "Structuralism" has now become a battle cry in semiotics. To be "poststructuralist" is rather like having the head of a moose over your fireplace. Well and good, but structuralism should not be sealed and posted before its letter is finished.

    In my view, structuralism got stuck in my view in the hypnotic spell of some brilliant but excessively simple reductions, particularly a narrow construction of articulation and combinatory relations, a false lure that promised a closer approximation to the world of mathematics than we should hope to realize. Of necessity, what follows is still abstractive and reductive. I hope the pieces and angles in my Lego kit will not discourage extension or reconstruction. Structuralism is an unfinished project.

    Chapter 13 concerns variety in the way wholes are divided into parts. Chapter 14 concerns referential functions of syntactic hierarchy. Chapter 15 contrasts two types of textual coherence. Chapter 16 regards the coherent text in the constellation of other texts with which it interacts. Chapter 17 picks up a cue that was dropped in chapter 10, the possibility of taking a sign factor to be a process of consciousness rather than an item.



    We gradually shift our course here, attending less to what signs share and more to how they differ. The topics are merely examples, arbitrary ones from the standpoint of theory. I emphasize the arts with the conviction that they are the true laboratories of the great human project of sign making. The verbal arts have provided a field of analysis within semiotics so weighty and specialized that I wished to put my baggage at the other end of the boat. After chapter 18 regards the supportive relation between formalization and play via quick glances of several semiotic media, the discussion settles on music and visual art. This is an ancient and loaded pairing, which had much to do with the development of the concept of "absolute music," the chief locus of controversy in 19th-20th century musical aesthetics. I make no excursion into the history of this problem nor any effort to relate my results to the ancient quarrel. (See Dahlhaus 1978. ) Nevertheless, I hope the analyses offered will suggest how inadequate and uninformative a simple polarity of "musical" and "plastic" appears when we can acknowledge hierarchy of content, perspective, and suggestions of designation and modality in both.

    As our focus turns more and more intently to the representamen and object, with what may seem just a pedantic obligation of rather stiff and formal courtesy calls on the interpretant, the reader may be tempted to revisit the question whether a theory of signs is needed propose the analyses offered here. If so, note that our discussion of the representamen now turns on two questions inherited from our theory of the sign: How is the signifying structure equipped to entrain, specify, or generalize various objects? How does it constrain and elaborate their interpretation?



    In this book I have not considered the monumental literatureóthe writings of Michel Foucault, Jacques Lacan, Julia Kristeva, and others--that explores philosophical, social, and cultural issues in the light of semiotic consciousness; my concern has been to review foundational options in the construction of a semiotic theory. I do think it is the case, however, that our arrangement of distinctions regarding signs can facilitate thought on adjacent topics. I close with two speculative essays which intend to demonstrate that semiotic theory can suggest consequences. Both center on the notion of partially autonomous texts and systems. We cannot plan our cultures, but we do not need to think of culture as entirely hostage to blind forces. Semiotics provides a comparative perspective from which we can rethink issues in education, style, and governance.