Version classiqueVersion mobile
OpenEdition Books

Ré-inventer le réel

Thomas Dutoit
Trevor Harris

Versions of the Iconic

John Pier

Texte intégral

1I shall be taking it for granted that the real, insofar as it can be counted among the constituents of whatever it is that we regard as human, is inextricably bound up with representation. There is no real, however ephemeral it is and whatever epistemological, ontological or ideological ramifications it may involve, without some form of representation even though, as Professor Miller has persuasively argued, the failure of representation is a prerequisite for representation. It can be asserted, I believe, that even at the dawn of the third millennium, Plato’s dialogues on mimesis and related matters remain discreetly relevant even though, in our postmodern world, the real, whether disseminated or oversaturated in a context of proliferating signifiers and multiple realities, seems to lie beyond the grasp of any verbal or other form of discourse. Traditionally, mimesis has been the reserve of esthetic and, more particularly, of literary matters, going from imitation in classical poetics to verisimilitude and the establishment of literary realism to the portrayal of the mind in modernist texts, etc., and it has also been the object of numerous writings in the history of poetics. At the same time, however, mimesis has played a role in social discourse, as witnessed by the theory and practice of rhetoric established in antiquity and, in our own century, metamorphosed by the growth of mass media and its accompanying refinements in the distribution of information, advertising and propaganda, further revolutionized by the emergence of global virtual reality. "The medium is the message," tells us Marshall McLuhan, and Jean Baudrillard, observing that we no longer live in an age where, for instance, it is possible to speak of a map that is ideally coextensive with the territory that it represents, maintains that signs of the real have come to replace the real itself, that the real is a projection of models, rather than vice versa, and that the processes of representation have been absorbed into a network of simulacra. Disneyland, he claims, stands as a simulacrum of America, a model aimed at lending a coherence to the American way of life for anyone who may have trouble grasping the truth unassisted by a guiding hand (Baudrillard, 1982, 24-28). And for my own part, I still remember watching CNN’s daily televised reports at the time of the Gulf War, informatively and reassuringly entitled "Showdown in the Desert".

  • 1 See in particular Gérard Genette, 1976, which traces the history of "secondary (or secondary mimolo (...)

2These matters extend to the outer perimeters of my present contribution, however, which are limited to a few observations on the iconic and, more specifically, on icons in the semiotic sense. My focus will not be the "verbal icon," introduced into literary theory and criticism by W. K. Wimsatt, nor will it be the "iconography" or "iconology" of art criticism, although semiotic iconicity certainly has a role to play in these notions. My concern stems, rather, from the problem debated in Plato’s Cratylus as to whether words result from convention between speakers (thései) or from a natural bond with the things they designate (phúsei), and although Cratylism has maintained an irrepressible place in the philosophy of language and in the history of poetics,1 it is the reemergence of these questions in the form of iconic signs in modern semiotic theory that will occupy us here. The departure point for any reflection on iconic signs is in the writings of Charles Sanders Peirce. In Peircean semiotics, icons stand alongside indices and symbols, this widely-known and influential triad being regarded by Peirce himself as "the most fundamental division of signs" (Collected Papers [abbreviated below to CP], 2.275). "An Index" he states, "is a sign which refers to the Object that it denotes by virtue of being really affected to that Object" (CP 2.248); such is the case of a pointing finger or of demonstrative and personal pronouns, which put the representamen in an existential relationship with its object. "A Symbol is a sign which refers to the Object that it denotes by virtue of a law, usually an association of general ideas" (CP 2.249), and is the category of arbitrary and conventional signs. "An Icon," however, "is a sign that denotes merely by virtue of characters of its own, and which it possesses, just the same, whether any such object actually exists or not" (CP 2.247); moreover, it is a sign which "represent[s] its object mainly by its similarity, no matter what its mode of being" (CP 2.276), or "a sign which stands for something merely because it resembles it" (CP 3.362). The icon (together with index and symbol, the other two members of the second trichotomy) thus places the accent on the relation between the representamen and its object, as opposed to signs of the first trichotomy, which emphasize the sign character of the representamen (qualisign, sinsign, legisign —"according as the sign in itself is a mere quality, is an actual existent, or is a general law" [CP 2.243]), and to signs of the third trichotomy, where the focus is on the interpretant (rheme: possibility; dicent: fact; argument: reason). According to one commentator: "[T]he most important thing to understand about iconic signs generally is that their distinctive function is that of making their objects immediately available as they are in themselves, in this and/or that respect (perhaps, even, in every respect)"; iconic signs are "immediately perceived as sensory objects in their own right prior to their use as representative of something else. This means that they are both self-representing and other-representing" (Ransdell, 1986, 68 and 70). A final point that bears mentioning is the fact that a pure icon is a strictly theoretical concept, for if an icon were to be pure, the representamen would be indistinguishable from its object. Consequently, actual icons are referred to by Peirce as "hypoicons" (CP 2.277).

  • 2 A well-known critique of the "six naïve notions" of iconism is that of Umberto Eco, for whom "modes (...)

3Now, there is a great deal more to be said about Peirce’s iconic signs than this, particularly if one wishes to place them in the larger context of his complete semiotic system and to take account of the various critiques and revisions that have succeeded Peirce’s original formulations.2 It is not my ambition in these few pages to address all of the issues at stake in these matters, but rather to draw attention to the three types of icons identified by Peirce — images, diagrams or metaphors — which, I shall be arguing, must be reckoned with when examining the processes by which human agency invents (or re-invents) the real. In order to illustrate these types of icon, I shall propose a few examples.


4The identity photo included in a passport enables the customs officer to determine the degree of similarity of qualities between signans and signatum. By comparing the features of each (angular or round features, color of hair, skin and eyes, knife slashes, etc.), the officer seeks to establish a degree of similarity between the photo and the individual before him adequate to identify the "real" John Smith for the purpose of allowing that individual to enter the country. Note that if the officer were to seek total identity between signans and signatum, his function would be radically undermined, either because he would allow no one to cross the border or because he would be unable to decide whether to allow entry to the passport or to the owner of the passport. Images, then, are icons that "partake of simple qualities" (CP 2.277).

5Examples that are significantly different from the one above, however, can be found in esthetic matters, where the merging of the real with its copy is frequently regarded as a mark of high artistic achievement (see CP 3.362 to this effect). Iconoclasts, as Baudrillard has shown, take much the opposite view, for they are motivated to destroy images of God that they fear will be worshipped in place of God (1982, 14-15). On the other hand, the semiotics of the imaginai icon can be extended to include the police line-up where, for instance, the detective seeks to establish a demonstrable degree of similarity between one of the suspects (signans) and the photo of the robber taken in the act by the bank’s security camera (signatum). Here again, the principle need not be restricted to the "real": during September 1997, the French press published expensively the photo of an Algerian women whose eight children had just been massacred, a photo that indisputably evoked the Pietà.


6The blueprint of a building (as opposed to a photo), a grid upon which the days of the month are visualized, graphs for the presentation of statistical data, the musical score of a symphony, the phonetic alphabet for the notation of phonemes — such schemata e xist in virtue, not of a similarity of qualities, but of an analogy of relations. For Peirce, they are examples of diagrams, a form of icon "which represent[s] the relations, mainly dyadic, or so regarded, of the parts of one thing by analogous relations in their own parts" (CP 2.277).


7Metaphor is a more widespread phenomenon than is sometimes thought. Take for example the metaphor "time is money," which points towards a whole series of expressions including "save time," "waste time," "spend time," "invest time," "budget time," etc. — a conceptualization of time that it would be pointless to reduce to "literal" meanings. Prepositions such as "up" and "down" become orientational metaphors in such expressions as "his spirits are up," "I’m feeling down," "market prices moved up/down," "she came down with the flu". Structural metaphors can organize entire stretches of discourse, as when an argument is presented as a journey ("set out to prove an hypothesis," "proceed step by step," "arrive at a conclusion"), or as a building ("lay the foundation of an argument," "construct/buttress/undermine an argument," "shaky/solid argument"). These examples, taken somewhat freely from George Lakoff’s and Mark Johnson’s Metaphors We Live By (1980), suggest that "metaphor is understanding and experiencing one kind of thing in terms of another" (5), so that metaphor, while culturally determined, is grounded in mental processes. They also serve to illustrate Peirce’s definition of metaphor as icons "which represent the representative character of a representamen by representing a parallelism in something else" (CP 2.277). Thus, the parallel established between time and money through the verb "to spend" results in a metaphor in which "to spend time" bears an iconic relation of similarity or resemblance to "to spend money," while it also stands as a metasign of "to spend money," whose status as a sign is thereby brought more fully into play.

8It must be remembered that actual icons are hypoicons, as an icon in a pure state is a possibility, a quality with no instantiation. Furthermore, like all other signs in Peirce’s system, icons are variously linked to other types of signs. The grid used to represent the days of a given month in a given year, for example, can be used to diagram the days of any other month of any other year. At issue here are signs of the second trichotomy in their relations with those of the first and the third: a rhematic iconic legisign ("a diagram, apart from its factual individuality") and its transformation into a rhematic iconic sin sign ("an individual diagram" [CP 2.254ff]). While the hybrid character of signs is an important aspect of Peircean semiotics and one that has been the occasion of much discussion, it raises questions that extend beyond the scope of my present argument; namely, to investigate some of the ways in which icons participate in the invention or the re-invention of the real, how they provide a sort of framework or model for the construction or elaboration of the real while they seem to merge with the real itself.

  • 3 See Dressier, 1987, and Landsberg, 1996, for similar views on diagrammatic iconicity. Also to be co (...)

9Icons are sometimes regarded as marginal and somewhat trivial phenomena with few manifestations beyond such things as identity photos, graphs, onomatopoeia and the like. There is growing evidence, however, to suggest that iconicity is a more pervasive feature of signifying systems than is commonly believed. This can best be seen in diagrammatic icons, a simple linguistic example being Caesar’s Veni, vidi, vici, where the order of the words follows the order of the reported actions: I came, I saw, I conquered. This principle has been seized on by a number of linguists who have produced studies in natural phonology, natural morphology, natural syntax, etc., such that there now exists a school of "natural linguistics". Research in this field involves reintroducing "motivation" into the study of linguistic phenomena in order to offset the linguistic orthodoxy of the "arbitrariness" of the sign. One of the first to observe the role of diagrams in grammatical structures was Roman Jakobson (1971a, 349ff. and 1971b, 707), and since that time natural linguists have often referred to Peirce’s diagrammatic iconicity in connection with the theoretical framework of their research. In his work on natural syntax, for example, John Haiman (1980, 1985) proceeds on the basis of two principles: (1) isomorphism (or the one form-one meaning principle) and its violations in the form of homonymy, synonymy, polysemy, etc., independent of whatever position the words may have relative to one another; (2) motivation, a diagrammatic property whereby words exhibit the same relations among themselves as do the relations among the referents of those words, as in the Veni, vidi, vici principle, although in fact language is replete with distortions of such diagrammatic parallelisms. It is significant that Haiman relates this motivation principle directly to the pages in Saussure’s Cours where, for instance, huit, neuf and dix are given as examples of arbitrary signs, while dix-huit and dix-neuf are described as "relatively motivated," since the relations between these words are dictated by the lexical system of numbers in the French language. He maintains, furthermore, that Saussure’s motivation, even though not clearly defined, is identical with Peirce’s diagrammatic iconicity (1985, 14).3

  • 4 Valuable background material on iconicity in literature can be found in Nanny, 1986, although this (...)

10The case can be plausibly argued, I believe, that not only language as such, but also the patterns and schema of all forms of discourse, literary and otherwise, can be described in terms of such a diagrammatic iconicity. Without attempting to provide here the detailed and systematic examination that the subject merits, I would like to suggest that poetic meter and prosody are prime instances of this principle, while narrative structures and categories come not far behind.4 It seems to me that, as far as the literary text is concerned, diagrammatic iconicity is manifested in a series of three stages that reflect Peirce’s distinction between the immediate object and the dynamical object. The immediate object is the object "as the Sign itself represents it, and whose Being is thus dependent upon the Representation of it in the Sign," as opposed to the dynamical object, "which is the Reality which by some means contrives to determine the Sign to its Representation" (CP 4.536), that is, the object that lies beyond the grasp of the sign, but that acts on the sign as such, resulting in imaginai, diagrammatic and metaphoric icons.

  • 5 The distinction between microtextual and macrotextual iconicity is made by Van Zoest, 1977,14ff.

11First of all are the various textual categories themselves, an example of which is the system of narrative time, as formulated by Gunther Müller and Eberhard Lämmert and later refined by Gérard Genette (1972, 77-182): 1) the chronological order of events in the story as opposed to their spatial distribution in the text together with the various "anachronies" or discordances between the two orders; 2) duration, that is, the greater or lesser degree of coincidence between story and text in the form of scene, ellipsis, pause and summary; 3) frequency, or the various relations of repetition between story and text. Equally concerned by this aspect of diagrammatic iconicity are other categories dealt with by narratologists including plot structure, focalization, discourse typology, voice, narrative level, etc., and it is useful to bear in mind that, just as these categories are manifested both microstructurally and macrostructurally, so too is diagrammatic iconicity.5 As such, categories of this sort come under the iconic legisigns referred to earlier; that is, they act as a sort of "grid" for the structuring of (in this case) the narrative text.

12The second aspect of diagrammatic iconicity in the literary text involves a broader framework for textual categories, and here I would like to suggest that the three forms of "mimesis" elaborated by Paul Ricoeur in his Temps et récit are of some relevance. Mimesis I, or prefiguration, involves the pre-textual dimension of human agency, while Mimesis II, or "configuration," includes the processes of mediation through the deployment of textual means or, in Ricœur’s terms, mise en intrigue (emplotment), and Mimesis III, or "refiguration," stands at the intersection between the world of the text and the world of the reader or listener. I take configuration to be the deployment of textual categories in individual texts, coming under the iconic sinsign.

13The third aspect of diagrammatic iconicity derives from vraisemblance and "naturalization" as discussed by Jonathan Culler in his Structuralist Poetics. In chapter 7 of this book, Culler focuses on the five interrelated ways that vraisemblance brings a text "in relation with another text which helps to make it intelligible": 1) the socially-given text of the "real world"; 2) the shared knowledge of the "cultural text"; 3) the conventions of genre; 4) "the natural attitude to the artificial"; 5) intertextual relations between specific texts. The interest of these categories in the present context is that, broadly speaking, they represent a corrective measure to the structuralist dogma of the arbitrariness of the sign through the re-introduction of the dialectics between what the Russian formalists called "motivation" and "defamiliarization," that is, the process of rendering the unknown intelligible by appealing to the known. It is within this dynamic process of motivation and defamiliarization that, as Culler argues, the naturalization of a text takes place. Naturalization, I would like to suggest, is a form of diagrammatic iconicity that is closely associated with the dynamical object: "the Real (no, perhaps because the Object is altogether fictive, I must choose a different term, therefore), say rather the Dynamical Object, which, from the nature of things, the Sign cannot express, which it can only indicate and leave the interpreter to find out by collateral experience" (CP 8.314).

14As an illustration of how diagrammatic iconicity actually functions in a text, we shall look briefly at Walter Abish’s Alphabetical Africa. This novel is composed in a peculiar manner in that chapter one is written using only words beginning with letter "a", chapter two only with words beginning with letters "a" and "b", etc. up until chapter twenty-six, where words beginning with all of the letters of the alphabet are used, while chapters twenty-seven through fifty-two progress in the opposite order. In addition to the novelistic content of the work being jolted out of the conventional patterns as a result of this alphabetically-motivated restriction, the alphabetical organization of the work produces numerous startling effects at the level of lexical choices and grammatical and stylistic features. Also, while the three main characters, Alex, Allen and Alva, appear right from the first page, it is only beginning (and ending) with the chapters where "I" is allowed to intervene that we have a clear idea of the fact that the novel as a whole is a first-person narration, and only in the chapters containing letter "y" that it becomes possible for the narrator to express his affection for the heroine, saying: "You were unaware of me at first" (Abish, 1974, 62). Reading Abish’s novel is something like assembling a puzzle in which the way to fit the pieces together is at odds with the scene represented: some of the green pieces of the meadow fit with the blue pieces of the sky, while to complete the trees or the mountain, pieces that are normally for the sky must be used, etc. It is thus a work that pits the building blocks of written language — the alphabet — against novelistic convention. Expressed another way, it throws into sharp relief a striking yet revealing discordance between two forms of diagrammatic iconicity whose complementarity is habitually considered to fall within the province of the natural: that of the alphabet and that of textual and narrative schemata.

15The situation is markedly different when a diagrammatic icon is taken for an object, as though the musical score of a symphony were to be taken for a performance of that symphony, or when an imaginai icon is taken for the original as, for instance, the photo of a building taken for the building itself. Examples such as these are comparable to what, in the verbal sphere, Roland Barthes (1968) calls l’effet de réel, lamentably translated, from a semiotic point of view, as "the reality effect". In his famous and influential essay on the semiotic status of the "concrete detail" in the literary text, it is interesting that Barthes refers to "the tripartite nature of the sign," that is, to the idea that the sign is constituted of a signifier, a signified and a referent, and not, as Saussure would have it, of a signifier and a signified. The concrete detail (or the perceiver’s impression thereof) involves "la collusion directe d’un référent et d’un signifiant" and consequently the evacuation of the signified, as though the unmediated details themselves were to function as signifiers instead of as referents, affirming: "nous sommes le réel." In such cases, the process of signification as it appears in the denotational language of a scientific discourse, for instance, is fundamentally modified: the real is no longer identical with the referent, but becomes the signified in a connotational or non-referential language — in other words, not the real itself, but its (illusionistic) effect. To put it in terms closer to the context of my present observations, it is as though "cock-a-doodle-doo" is in fact what the chicken says rather than an onomatopoeia, as though the imaginai icon were identical with the original rather than a copy of the original. Barthes himself, however, might differ from this analysis, not because French chickens say "cocorico," but because he focuses on the "direct collusion [secret agreement for nefarious purposes, trickery] between the referent and the signifier" rather than on a relation of similarity or resemblance. He speaks of the effect of the real, and not of an image, diagram or metaphor of the object, and for this reason it would seem that l’effet de réel, according to the case at hand, is located somewhere between the iconic and the indexical.

16Metaphor is an icon that functions in yet another way. According to Peircean theory, it begins with two symbols or arbitrary signs that establish a parallel between their respective objects, and this parallel then displays the "representative character" of the represen: amen or sign vehicle (signifier) with the result that the metaphor qualifies as a metasign. To illustrate this point, I have selected two short extracts from the first chapter of James Joyce’s A Portrait of the Artist as a Young Man where Stephen Dedalus, still a young boy, begins to become aware of and to understand the feminine:

And [Dante] did not like him to play with Eileen because Eileen was a protestant and when she was young she knew children that used to play with protestants and the protestants used to make fun of the litany of the Blessed Virgin. Tower of Ivory, they used to say, House of Gold! How could a woman be a tower of ivory or a house of gold?... Eileen had long white hands. One evening when playing tig she had put her hands over his eyes: long and white and thin and cold and soft. That was ivory: a cold white thing. That was the meaning of Tower of Ivory. (35-36)

17And a few pages later:

Eileen had long thin cool white hands too because she was a girl. They were like ivory; only soft. That was the meaning of Tower of Ivory but protestants could not understand it and made fun of it.... Her fair hair had streamed out behind her like gold in the sun. Tower of Ivory. House of Gold. By thinking of things you could understand them. (42-43)

  • 6 Note that when textual icons, be they topological, diagrammatic or metaphorical, involve interrupti (...)

18It is fascinating to see how in only a few sentences Stephen transforms the meaning of Tower of Ivory from a conventional symbol for the Blessed Virgin in the Roman Catholic Litany into a meaning that he finds personally intelligible. Unable to understand the religious symbolism of the Tower of Ivory, he literalizes the expression, wondering how a woman can be an object such as a tower of ivory. A partial answer is provided when he feels Eileen’s hands over his eyes: the form of her hands — long, white, thin and cold — reminds him of the form of a tower of ivory, but the meaning he gives to ivory remains incongruously literal, totally dissociated from both the Blessed Virgin and Eileen: "a cold white thing". However, the realization that Eileen’s hands are soft changes Stephen’s perception significantly: "They were like ivory; only soft. That was the meaning of Tower of Ivory." Eileen’s hands thus resemble ivory but are different because they are soft, which gives them a new meaning. In effect, they become a metonymy (index) for Eileen: soft hands mean Eileen. Stephen thus arrives at the answer to his question: "How could a woman be a tower of ivory?", and in doing so, he passes from the Tower of Ivory as an empty symbol for the Blessed Virgin to a meaningful metonymy for Eileen. But as the larger context of A Portrait will confirm, there is, beyond Eileen and beyond the Blessed Virgin, the extremely important question of the feminine, so that the Tower of Ivory is transformed into one of a number of metaphors for the feminine with extensive ramifications for the thematic and psychoanalytical dimensions of the work. Being a metaphor, then, the Tower of Ivory signifies its object indirectly and is thus a metasign. Moreover, as it is a metaphor that is not limited to a local manifestation, but informs a considerable stretch of discourse, it can be considered an example of macrostruc tural iconicity akin to the macrostructural features of diagrammatic iconicity.6

19The iconic, I have argued, is a function that underscores the self-representing quality of certain types of signs, exhibiting the sign as it is in itself, even while that sign is other-representing. The iconic sign seeks to incorporate the real into itself but can never become totally identified with the real, the reason why icons, once they leave the domain of pure possibility, are qualified as hypoicons. The examples of the iconic that I have discussed are drawn, not from the real, but from the fictional and the imaginary and, to various degrees, this is symptomatic of all of the papers that have been presented at this colloquium. The one nearexception to this observation — the one "real" object we have witnessed — is the shoe placed on the table by Jon Delogu during his intervention yesterday. The point was strikingly confirmed in the discussion following Anne Tomiche’s paper on Gertrude Stein, when it was pointed out by Stephen Romer that in Stein’s work the direct confrontation between word and thing paradoxically drives a wedge between word and reality. And in commenting on Lyotard’s affirmation that realism avoids the problem of reality, Kathie Birat noted that the real exists only in relation to something else. This is a singularly Peircean insight, and one that, in my view, puts into words what has proved to be one of the principal leitmotifs of our colloquium: we do not re-invent the real so much as we invent the "real".


Works Cited

Abish, Walter, 1974. Alphabetical Africa. New York: New Directions.

Bal, Mieke, 1978. "Mise en abyme et iconicité." Littérature 29: 116-128.

Barthes, Roland, 1968. "L’effet de réel." Communications 11: 84-89.

Baudrillard, Jean, 1982. Simulacres et Simulation (Collection Débats). Paris: Galilée.

Bense, Max, and Elisabeth Walther, eds., 1973. Wörterbuch der Semiotik. Cologne: Kiepenheuer & Witsch.

Bouissac, Paul, Michael Herzfeld and Roland Posner, eds., 1986. Iconicity. Essays on the Nature of Culture. Festschrift for Thomas Α. Sebeok on his 65th Birthday. (Probleme der Semiotik 4). Tübingen: Stauffenberg.

Culler, Jonathan, 1975. Structuralist Poetics. Structuralism, Linguistics, and the Study of Literature. Ithaca: Cornell UP.

Dressler, Wolfgang, Willi Mayerthaler, Oswald Panagl and Wolfgang U. Wurzel, eds., 1987. Leitmotifs in Natural Morphology. (Studies in Language Companion Series 10). Amsterdam: John Benjamins.

Eco, Umberto, 1976. A Theory of Semiotics. (Advances in Semiotics). Bloomington: Indiana UP.

Genette, Gérard, 1972. "Discours du récit, essai de méthode." In Figures III. (Collection Poétique). Paris: Seuil, 67-282.

—— 1976. Mimologiques. Voyage en Cratylie (collection Poétique). Paris: Seuil.

Haiman, John, 1980. " The iconicity of grammar: Isomorphism and motivation." Language 56.3: 515-540.

—— 1985. Natural Syntax. Iconicity and Erosion. (Studies in Linguistics 44). Cambridge: Cambridge UP.

Jakobson, Roman [1965], 1971a. "Quest for the Essence of Language." In Selected WritingsΠ. The Hague & Berlin: Mouton de Gruyter, 345-359.

—— [1968], 1971b. "Language in Relation to other Communication Systems." In Selected Writings II, 697-708.

Joyce, James [1916], 1964. A Portrait of the Artist as a Young Man. (The Viking Critical Library). Ed. Chester G. Anderson. New York: Penguin.

Lakoff, George, and Mark Johnson, 1980. Metaphors We Live By. Chicago & London: Univ. of Chicago.

Landsberg, Marge, ed., 1996. Iconicity in Syntax. Berlin: Mouton de Gruyter.

Nänny, Max, ed., 1986. Iconicity in Literature. Special issue of Word and Image 2.3.

Peirce, Charles Sanders, 1931-1958. Collected Papers. Ed. Charles Hartshorne and Paul Weiss (vols. 1-6); Arthur W. Burks (vols. 7-8). Cambridge: Harvard UP.

Ransdell, Joseph, 1986. "On Peirce’s Conception of the Iconic Sign." In Bouissac et al, eds., 51-74.

Ricœur, Paul, 1983-1985. Temps et récit. (L’ordre philosophique). 3 vols. Paris: Seuil.

Sebeok, Thomas Α., 1979. "Iconicity." In The Sign and Its Masters. Austin & London: University of Texas.

Van Zoest, Aart J. Α., 1977. "Le signe iconique dans les textes." Zagadnienia Rodzajów Literackich 20.2: 5-21.

Waugh, Linda R., 1993. "Against Arbitrariness: Imitation and Motivation Revived, with Consequences for Textual Meaning." Diacritics 23.2: 71-87.

Whiteside, Anna, 1988. "Verbal icons and self-reference." Semiotica 69.3-4: 315-329.


1 See in particular Gérard Genette, 1976, which traces the history of "secondary (or secondary mimologism)," showing, for instance, that the poetic function included in Roman Jakohson’s model of verbal communication in effect reintroduces mimeticism into the linguistic sign, considered by Saussure to be of a fundamentally arbitrary nature.

2 A well-known critique of the "six naïve notions" of iconism is that of Umberto Eco, for whom "modes of sign-producing functions" (1976, 217; author’s italics) replace types of signs, including iconic signs. As Eco’s reasoning follows lines that are distinctly different from those of the present argument, no attempt will be made here to take it into consideration, although Thomas A. Sebeok’s comment on Eco’s analysis appears to me incontrovertible: "The notion of an’icon’is very much impoverished when viewed, as it often is, in isolation rather than in the total context of a fully rounded doctrine of signs " (Sebeok, 1979, 117).

3 See Dressier, 1987, and Landsberg, 1996, for similar views on diagrammatic iconicity. Also to be consulted is Waugh, 1993.
Max Bense and Elisabeth Walther (1973), refer to Peirce’s diagrammatic icons as "structural icons" and to his imaginai icons as "topological icons". While the term topological icon seems to me an appropriate adaptation of Peirce’s images to textual features, that of structural icon lacks, in my opinion, the specificity of diagrammatic icon. For a study of Apollinaire’s calligrams that can be characterized as topological in this sense, see Whiteside, 1986.

4 Valuable background material on iconicity in literature can be found in Nanny, 1986, although this interesting collection of articles does not seek to make use of iconicity in the semiotic sense.

5 The distinction between microtextual and macrotextual iconicity is made by Van Zoest, 1977,14ff.

6 Note that when textual icons, be they topological, diagrammatic or metaphorical, involve interruption in diegetic level (which is not the case in the example taken from A Portrait), the result is a mise en abyme. For an insightful discussion on this point, see Bal, 1978.


University of Tours

© Presses universitaires François-Rabelais, 1999

Conditions d’utilisation :


Volume papier