Inventor of “stories in etchings” at the end of the 1820s, the Genevan Rodolphe Töpffer (1799–1846) initiated the theorization of this new form of storytelling. For the reader at the end of the twentieth century, the first “defense and illustration” of comics,1 his Essai de physiognomonie (1845), opens stimulating perspectives for a reflection on an art which, in the intervening period, has contributed in a decisive manner to the shaping of the modern imagination, thereby confirming the intuitions of the genial precursor.
Since this initial thunderclap, it is rarely noted that practice has become divorced from theory. The works that have contributed to the understanding of the comics phenomenon are extremely limited in number, and the relative legitimation of the “ninth art” in France has not actually led to their multiplication. Myopic scholarship, nostalgia, and idolatry have structured the discourses around comics for about three decades. All too often the history of the medium takes the form of an egalitarian chronicle where masterpieces and less glorious works are treated as equivalents, while, at the same time, the artists who “sell” are continually the object of fetishistic celebrations in which critical analysis has little place.
As rare as they have been, the milestones of thinking about comics nevertheless demonstrate an evolution in the approach to the subject. Pierre Fresnault-Deruelle—who, within the French universities, was, for a long time, alone in his interest—distinguishes four successive layers in the critical discourse:
the archeological age of the 1960s, where nostalgic authors exhumed readings from their childhoods (Lacassin 1971)
the sociohistorical and philosophic age of the 1970s, where the critics established the texts in their variants, reconstituted the relationships, etc. (Le Gallo 1967; Kunzle 1973)
the structuralist age (Fresnault-Deruelle 1972, 1977; Gubern 1972)
the semiotic and psychoanalytic age (Rey 1978; Apostolidès 1984; Tisseron 1985, 1987)2
I subscribe grosso modo to this periodization, but it is still necessary to qualify it. Of the four tendencies, none has been totally abandoned; they continue to coexist rather like divergent, or parallel, roads offered to the investigator, not exclusive from others (in particular, thematic criticism and genre studies: humor, fantasy, western, etc.). What interests me more is that Pierre Fresnault-Deruelle has marked the recent emergence of a “fifth stratum,” that of a “neo-semiotic criticism where the accent will be placed on the poetic dimension of comics.”3 It seems to me that this precisely recognizes the ambition of this book.
Comics will be considered here as a language, that is to say, not as a historical, sociological, or economic phenomena, which it is also, but as an original ensemble of productive mechanisms of meaning. This language will not be passed through the sieve of a grand constituted theory, such as structural analysis or narrative semiotics. Taking into account the given object, the perspective that I propose can no doubt be described as semiologic (or semiotic) in the broadest sense of that term. However, as there will be hardly any discussion of the sign in these pages—for reasons that will become clear in a moment—I situate myself, in regard to semiology, on the fringes of its disciplinary orthodoxy. I will not forego a few short detours through the realms of the semantic and the aesthetic, turning to my advantage everything that can contribute intelligibility to the medium. That is the reason that the term “neo-semiotic” appears to me completely adequate to qualify the point of view that The System of Comics demands.
Reading the researchers who have preceded me and, above all, of the vulgate spread by the media and by instruction manuals, has convinced me that a theory of comics must definitively renounce two current ideas that, even though inspired for the most part by the semiotic approaches produced up to this point, appear to me to be obstacles to real comprehension of the object. The first widespread idea is that the study of comics, like that of every other semiotic system, must pass through a decomposition into constitutive elementary units: the “smallest commutable elements that have a proper meaning,” to use Christian Metz’s phrase.4 I hold that this method cannot bring forth that which is truly specific about the language of comics.
The second idea is that comics are essentially a mixture of text and images, a specific combination of linguistic and visual codes, a meeting place between two “subjects of expression” (in the sense of the linguist Louis Hjelmslev). Against this conception, I intend to demonstrate the primacy of the image and, therefore, the necessity to accord a theoretical precedence to that which, provisionally, I designate under the generic term of “visual codes.”
I will begin by explaining myself on these two points.
For certain researchers, all drawing—and, singularly, the often willfully schematic linework of traditional comics—can be broken down into discreet units that can then be identified—points, line sections, spots—as equivalents (according to a precise system of homology or of analogy) to those of lexemes, morphemes, and phonemes in natural languages. Guy Gauthier, for example, defended this option in 1976: “We postulate therefore that, in every image, it is possible to isolate lines or groups of lines, spots or groups of spots, and to locate, for each signifier thus determined, a precise signified, itself corresponding to a part of the global signified.”5 The same author insisted: “The discrete units generated in the drawing style of Peanuts can be compared to the units of the first articulation of language, the image can be compared to one or more syntagms” (p. 126).
According to other researchers, the pertinent units are more highly elaborated and correspond to the illustrated message or to the figures—objects, characters, body parts. In an essay entitled “Comics lesen,” Ulrich Krafft distinguished four kinds of patterns, respectively: character in the foreground, object in the foreground, character in the background, object in the background. Then he broke up the “character” into smaller and smaller signs (Anzeichen), thus categorizing Donald Duck as the head within the body, the eye within the head, and the pupil within the eye.6
Following the terminology proposed by the Groupe Mu in their Traité du signe visual (terminology that I take to be essential), the elementary units distinguished by Krafft correspond to “sub-entities” of iconic signifiers, while those discussed by Gauthier are of an inferior standard, that of “marks.”7
As we know, the simultaneous existence of similar units within an image is controversial. If the Groupe Mu gives credence to this thesis in supplying a general and systematic description, no doubt the most convincing to this day, other eminent researchers have pleaded for the recognition of a semanticism specific to the image, which makes the economy of stable units analogous to those of language.8 This was already Émile Benveniste’s point of view:
The signifying relations of artistic “language” are revealed within a composition. Art is never more here than a particular work of art, where the artist freely establishes oppositions and values which he commands with total sovereignty, having neither an “answer” to wait upon, nor a contradiction to eliminate, but only a vision to express. . . . The significance of art never returns to a convention identically received between partners. Each time it must discover the terms, limitless in number and unpredictable in nature, so as to be reinvented for each work; in short, they are inapt to be fixed in an institution.9
The image provides the example of a semiotic system devoid of signs, or at least not reliant on a finished system of signs. It is in this sense that Benveniste maintained that “none of the plastic arts considered in their entirety can reproduce [the] model [of language],” the language in which he needs to resign himself to see “the only model of a system that can be semiotic at the same time in its formal structure and in its functioning.”10
Although I adhere without reservation to Benveniste’s affirmation, I am not trying to demonstrate the well-founded. I do not assume that the question of existence or nonexistence of visual signs is central in the analysis of the language of comics. I especially want to establish that the most important codes concern larger units, which are already highly elaborated. In this case, these codes govern the articulation, in time and space, of the units that we call “panels”; they obey criteria that are just as much visual as narrative—or, more precisely, discursive. These two orders of preoccupation sometimes superimpose themselves to the point of indistinction.
Entering inside the frame, in order to dissect the image by counting the iconic or plastic elements that compose the image, then studying the methods of articulation for these elements, supposes a profusion of concepts but does not lead to any significantly advanced theory. By this I mean that we touch upon only the most general mechanisms, none of which is particularly well suited to shed light on comics. I am convinced that we will not arrive at a coherent and thoughtful description of the language of comics by approaching them on this level of detail and incorporating a progressive enlargement. On the contrary, we need to approach from on high, from the level of grand articulations. (I don’t use the term “articulation” in the specific meaning that it has in linguistics but in the sense that it highlights the fact that every operation consists to “organize the collection of units functioning at the same level.”11)
In the concluding pages of his essay on Les dessous de la peinture, Hubert Damisch writes: “There, where semiology is vainly exhausted updating the ‘minimal units’ that would allow it to deal with painting as a ‘system of signs,’ painting demonstrates, in its very texture, that the problem demands to be taken upside down, at the level of relations between the terms, to the level, not of the ropes, but of the knots.”12 On the surface, this position is very close to mine but perhaps marred by a certain ambiguity: When describing the “relations between the terms” it is important to know with precision that which the terms bind. In so doing, the theory of painting will make the economy of a micro-semiotic approach much more difficult than the theory of comics. The reason for this methodical inequality is simple. The image in painting is unique and global; it cannot arouse delicate apprehension except at the price of decomposition (this was endlessly demonstrated by Alain Jaubert in Palettes, his remarkable television series about painting).13 On the contrary, the comics panel is fragmentary and caught in a system of proliferation; it never makes up the totality of the utterance but can and must be understood as a component in a larger apparatus.
Perhaps one objects that the fact of establishing the image as a base unit does not exempt an examination of the inferior elements that constitute it. It is true that these two approaches are not exclusive and that they can even complement each other. The Groupe Mu speaks of “this constant oscillation of the theory between the micro- and macro-semiotic, the first exhausted in the search for minimal stable units, the second challenging the existence of these in the name of originality each time renewed of complex utterances.”14 It is not important for me to challenge, alongside Benveniste, the existence of these units. It is only a question of knowing what, from the micro- or the macro-semiotic, is most useful for the elaboration of a complete model of the language of comics. I repeat: For the particular subject that is comics, the operativity of the micro-semiotic is revealed to be, in practice, extremely weak.
Guy Gauthier is elsewhere obliged to admit this. For one thing, he writes that, despite its “apparent complication,” the image “can always be reduced, sometimes, it is true, thanks to a work out of proportion with the results obtained,” on the other hand, in explaining his method allows at most the arrival of “the description of a code, or rather to a sub-code, since it characterizes a single artist while being accessible to millions of readers.”15 Despite his pretensions to scientificity, this method, when it distinguishes as many codes as there are artists, returns to stylistic analysis and not to the semiology of comics as such.
If the image is the base unit of the comics language it can be seen to confirm that the five “types of determinations” that characterize the “visual signs” according to Groupe Mu (the global properties, superordination, coordination, subordination, and preordination)16 all perfectly apply to this unit, and in a much clearer manner than to units of the inferior rank, such as, for example, the character.
It does not appear to me useful to fetishize a priori certain codes that are more specific to comics than others. This point merits a brief clarification. Christian Metz has insisted in several places that cinematographic language results from the combination of specific codes and nonspecific codes.17 In comics the codes that are truly specific to the form are perhaps less numerous than they are for film (if they even exist). Thus, the spatio-topical code, which organizes the co-presence of panels within space (and which I will establish later as a theoretical foundation), equally governs the framing relations of photo-novels. Further, this related medium has also adopted the speech balloon as a method of inserting writing into the heart of the image. At the end of the day, what makes comics a language that cannot be confused with any other is, on the one hand, the simultaneous mobilization of the entirety of codes (visual and discursive) that constitute it, and, at the same time, the fact that none of these codes probably belongs purely to it, consequently specifying themselves when they apply to particular “subjects of expression,” which is the drawing. Their “efficiency”18 finds itself notably singularized.
Comics are therefore an original combination of a (or two, with writing) subject(s) of expression, and of a collection of codes. This is the reason that it can only be described in the terms of a system. From then on, the problem posed to the analyst is not which code to privilege; it is to find an access road to the interior of the system that permits exploration in its totality so as to find coherence. Put another way, the objective must be to define the sufficiently encompassing categories for the majority, or the totality, of linguistic processes and the observable tropes in the field that can be explained by these concepts. In elaborating the concepts of spatio-topia, arthrology, and braiding, all three of which draw upon the macro-semiotic, I am obligated to realize this program.
If, at certain analytical moments, we move to the interior of the panel in order to concern ourselves with certain component elements, we will always do so with reference to the codes that, at a more elevated level of interrogation, determine these components. To give a simple example, one can see that a close-up has no value in itself, but as long as it is opposed to a general outline or is a part in a progression observable only if one takes into consideration the syntagm formed by a number of consecutive panels. In addition, this large framework can also “rhyme” with another large framework, and the two images thus bound are able to occupy opposing or symmetrical places on the page. The colors, and in a general way any units of an iconic or plastic nature, are simultaneously informed by the neighboring images and sometimes by the distant images. In short, the codes weave themselves inside a comics image in a specific fashion, which places the image in a narrative chain where the links are spread across space, in a situation of co-presence. The Québecois Yves Lacroix summed up the specificity of the medium very ably in speaking of “the soul of comics, its fundamental immobility, simultaneity and panopticism compels its units, otherwise known as the serial status.”19
Acting in its production or reading, the comics image is not that of painting. The meaning of this work will be to disengage and to analyze that which, between the unique fixed image (picture or illustration) and the animated image, is common to fixed sequential images.
“A thousand-year-old-logocentric tradition has trained us to conceive a relationship of the suzerainty of the verb to the image,” Michel Thévoz justifiably reminds us.20 This tradition has, in reality, produced two important consequences, which are not always as sufficiently distinguished from each other as they could be. One belongs to general semiology, the other to narratology:
The langue has been taken as the model of all language.
Fictional literature is considered almost everywhere and by nearly all as the model of all narrative forms. (This second consequence is partly a logical corollary of the first.)
Though it is historically based, this last conception is nonetheless theoretically untenable. The fact that written literature (itself preceded, and at one time accompanied, by oral literature) preceded by several millennia the quasi-simultaneous advent of cinema and modern comics confers on it no monopoly on the privilege of rights, merely a de facto anteriority. In other words, it is no longer possible to confuse narrative and literature, exposed as we have been to a range of media that have, more or less, recourse to the structures of the story.
The narrative genre, with all of its categories (intrigue, diegesis, situations, themes, dramatic conflicts, characters, etc.), exists in itself and can be analyzed as such, as a system of thought, as a manner of appropriating the world, or as an immemorial activity of the human soul. It cuts across different semiotic systems and can incarnate itself indifferently in each of them (or rather: differently, but without renouncing its particular technical nature, which is nothing other than the art of storytelling). I agree here, with Paul Ricoeur,21 that there exists a narrative genre and several narrative species: novel, film, stage play, but also comics, the photo-novel, and—why not?—also ballet and opera, without prejudging those that will be born tomorrow from technological progress (since comics and cinema owe their late birth—relative to literature—to technological evolutions, that is to say, for comics, the invention of lithography).22 Naturally, every narrative species proposes to the public another expository model of storytelling and is inclined to its particular competencies. Thus, as Ricoeur writes: “No mimetic art had gone as far in the representation of thought, feelings, and discourse as has the novel.”23 For its part, film has other assets, and comics have theirs also, a fact that is demonstrated by its continuing popularity after a century and a half of existence, despite the competition of cinema and of all the new images born of what Régis Debray calls the “videosphere.”
Töpffer saw in the text and in the image two equal components of comics, which he defined from their mixed character. This point of view, which was supportable at the time, is no longer today. Indeed, those who recognize in the verbal an equal status, in the economy of comics, to the image, begin from the principle that writing is the vehicle of storytelling in general. Yet the multiplicity of narrative forms has rendered this postulate obsolete.
To suppose that comics are essentially the site of a confrontation between the verbal and the iconic is, in my opinion, a theoretical counter-truth that leads to an impasse.24 Need I be specific? If I plead for the recognition of the image as preeminent in status, it is not for the reason that, except on rare occasions, in comics it occupies a more important space than that which is reserved for writing. Its predominance within the system attaches to what is essential to the production of the meaning that is made through it.
Some will surely meet this assertion with skepticism. Since Lessing, western thought has in effect clung to these two categories, “the story” and “the image,” taken as antinomic, beginning from the distinction between space and time. The cinematographic image was a time-image; it did not arouse the same theoretical embarrassment as the comics image. Of the two great forms of storytelling with images, it is undoubtedly comics that pose the most questions to the literary and plastic arts. Now, the apparent irreducibility of the image and the story is dialectically resolved through the play of successive images and through their coexistence, through their diegetic connections, and through their panoptic display, in which we have recognized the foundation of the medium. As we can see, it is through this collaboration between the arthrology and the spatio-topia that the sequential image is seen to be plainly narrative, without necessarily needing any verbal help.
The 1960s and 1970s, it is true, witnessed “a massive transfer of linguistic notions into the domain of the analysis of visual arts: we thus speak frequently of pictorial utterances, of filmic syntagms, etc.,” this application expanded linguistic concepts that relied on the idea that “all representation [can] not be anything but coded and that all contemplation of a figurative representation [is] a reading.”25 Yet this idea continues to be opposed by theorists who defend a more restrictive (dogmatic?) conception of the notion of narration and who refuse to extend it to the visual arts. Jean-Marie Schaeffer is one of the most convincing advocates of this linguistic orthodoxy. I am tempted primarily to oppose to this refusal the fact that it is manifestly counter-intuitive, that it goes against common experience: indeed, for the viewer of a film or the reader of a comic there is no doubt that one is being told a story! One also recalls that the generative process of all of these works usually begins through the creation of a scenario. But Schaeffer argues precisely that “narration is not given in or by the images (whereas in the case of a verbal structure, it is given in and by the connection of phrases): it is at once upstream from the work (as a narrative program) and downstream (as a reconstruction on the part of the spectator).”26
There is a certain kind of sophism in this position, which admits narration upstream and downstream but refuses to recognize it acting in the work itself! And one must ask by what miraculous cognitive alchemy the reader or spectator can reconstruct a story if that story wasn’t already contained in the work to which he is exposed. Schaeffer’s response is that the spectator extrapolates a story “beginning from what it (the image) represents thanks to what is shown.”27 He continues, “to tell a story in the first meaning of the term does not automatically imply that there is a narrative in the technical sense of term, that is to say an enunciative act assumed by the narrator.” Thus, he defines the two essential features that characterize narration as an enunciative act: “The specificity of logical links that carry out the reciprocal integration of elementary propositions thanks to the links of consecution (a and then b and then c and then . . .) and of causality (. . . c because b because a)” and the fact that “narrative assertions must refer to a speaker; it follows that all narration implies a narrator.”28
This demonstration, once again, is symptomatic of the linguistic hegemony in general semiotics29 and, therefore, of the too-frequent mechanical application of dogmas of literary narratology to every other form of storytelling. Linguistics always reduces the category of “story” to the authority of “narration” and does not recognize the presence of a narrator except insofar as certain markers belong to a particular verbal language. Consequently, it can only discredit image-based stories as narrative forms; the verdict is reached before the trial begins.
Instead of concluding that “it is appropriate to restrict the application of the technical notion of ‘narration’ to the verbal domain,”30 for my part I think: 1) that it is urgent to revise the technical notion that has ceased to be operative because it is in flagrant contradiction with the experience of the modern reader-viewer; and 2) that it is no less necessary to invent specific concepts to report on the extra-linguistic “logical links,” that “carry out the reciprocal integration of elementary propositions” in stories in images.
In Schaeffer’s argument, however, there is a point that may be retained and which is applicable to comics: This is the insistence on the active cooperation provided by the reader. Comics is a genre founded on reticence. Not only do the silent and immobile images lack the illusionist power of the filmic image, but their connections, far from producing a continuity that mimics reality, offer the reader a story that is full of holes, which appear as gaps in the meaning. If this double reticence recalls a “reconstruction on the part of the spectator,” the story “to be reconstructed” is no less set in the images, driven by the complex play of sequentiality. Moreover, if one believes François Dagognet, it is the role of art in general to manufacture “the surreal with the elliptic.”31 Every comics reader knows that, from the instant where he is projected into the fiction (the diegetic universe), he forgets, up to a certain point, the fragmented character and discontinuity of the enunciation. Allow me to recycle something that I wrote elsewhere on the particular illusionism of the narrative art of comics:
The panels return nothing but the fragments of the implied world in which the story unfolds, but this world is supposed to be continuous and homogenous, everything transpiring as if the reader, having entered into the world, will never again leave the image to which he has been offered access. The crossing of frames becomes a largely unconscious and mechanical operation, masked by an investment (absorption) in the virtual world postulated by the story. The diegesis, this fantastic virtual image, which comprises all of the panels, transcends them, and is where the reader can reside. If, according to Pierre Sterckx’s term, I can build a nest [nidifier] in a panel, it is because, in returning, each image comes to represent metonymically the totality of this world. . . . the multiplicity and spread of these images, the ubiquity of the characters, makes comics truly open to a consistent world, as I persuade myself all the more easily that I can live there that . . . I do not cease, in reading, to enter within and to exit.32
To sum up, the story is possibly full of holes, but it projects me into a world that is portrayed as consistent, and it is the continuity attributed to the fictional world that allows me to effortlessly fill in the gaps of the narration.
A number of comics lovers have said very similar things. Thus Pierre Fresnault-Deruelle: “the fascination that comics can carry out on the reader rests, among other elements, on its capacity to make us imagine everything other than what is actually shown to us: there is a rustling of voiceless signs (just like there is a swarming of the motionless) behind these docilely aligned frames”;33 or Federico Fellini: “Comics, more than film, benefits from the collaboration of the readers: one tells them a story that they tell to themselves; with their particular rhythm and imagination, in moving forward and backward.”34
In comics, as I have said before, narration passes first and principally (save for exceptions) by way of the images. André Gaudreault recalls that “for Plato mimesis is not, contrary to what is too frequently suggested, a category opposed to diegesis but well and truly one of the forms of it.”35 In truth, in an image-based story, as in film or comics, each element, whether it is visual, linguistic, or aural, participates fully in the narration. Christian Metz had the upper hand and one can’t say it more clearly: “in a narrative film, everything becomes narrative, even the grain of the film or the tone of voice.”36 The subjects that follow apply equally to comics as to cinema:
The terminology was principally fixed in reference to linguistic narration, in particular to novels. There, the narrative codings are superimposed to a first stage of major adjustments, those of language; it is because of them that we speak of enunciation, since the term is linguistic. By repercussion, if need be one can reserve “narration” to the lower level. But the narrative film is based on nothing, it doesn’t pile up on an equivalent of a language; it is itself, or rather it makes everything that it will be on the order of “language.” At the same time that enunciation is made narrative, narration takes charge of all enunciation.37
I want to contribute to clarifying in these pages the notion of stories in images, beginning with the particular case of comics, which I will postulate from here on as a predominantly visual narrative form. It seems to me in any case that Paul Ricoeur defined the proper perspective when he separated the authority of the story from its diverse concrete manifestations, and situated each on a plane of theoretical equality. This principled petition opens the field to comparative studies and to deepening the semiotic systems in their respective singularities.
The definitions of comics that can be found in dictionaries and encyclopedias, and also in the more specialized literature, are, as a general rule, unsatisfactory. It is easy to understand the reasons.
These definitions are of two sorts. The first, often concise, participates in an essentialist approach and looks to lock up some synthetic form of the “essence” of comics. This enterprise is no doubt doomed to failure if one considers that, far from verifying the long assumed poverty of expression and intrinsic infantilism, comics rest on a group of coordinating mechanisms that participate in the representation and the language, and that these mechanisms govern in their movements numerous and disparate parameters, of which the dynamic interaction takes on extremely varied forms from one comic to another. Whatever its successes on the plane of art, one must recognize that any comic:
is necessarily (constitutionally) a sophisticated structure
only actualizes certain potentialities of the medium, to the detriment of others that are reduced or excluded
Consequently, searching for the essence of comics is to be assured of finding not a shortage but a profusion of responses. In the brilliant essay by Alain Rey entitled Les Spectres de la bande,39 one thus reads on page 102 that “the essential” of comics is in “the organized space that cheats between the two dimensions of the format and the perceptive suggestion of the world”; page 104, that “the exchange between the textual and figural values creates the essence of comics”; page 200, that from now on the medium characterizes above all “a creative battle between figuration and narrativity, not between image and text, this last assumes nothing but the most superficial aspect of the story.” These are many different and fertile suggestions, and no doubt it would not be difficult to find in this single book half a dozen other analogous formulas that suggest some part of the truth.
But one also meets definitions of comics that are longer and more articulated, better conforming to the definition of a definition: “An enunciation of attributes that distinguish something, that belongs in particular to the exclusion of all others” (Littré). These differing definitions are retained as pertinent for the number and the identity of their attributes. Researchers have not failed to butt heads on this point, as one can see by looking at some clarifying examples.
The work of David Kunzle, The Early Comic Strip, launched a series intended to cover the entire history of comics. This first book examines the pre-Töpfferian period, from 1450 to 1825, grouping not only anonymous popular imagery but also painting and engraving cycles by artists such as Callot, Rubens, Greuze, and Hogarth, to name but a few. Kunzle formulates “four conditions” under which these stories in images can be considered proto-comics or, if one prefers, assimilated a posteriori:
I would propose a definition in which a “comic strip” of any period, in any country, fulfills the following conditions: 1/ There must be a sequence of separate images; 2/ There must be a preponderance of image over text; 3/ The medium in which the strip appears and for which it was originally intended must be reproductive, that is, in printed form, a mass medium; 4/ The sequence must tell a story which is both moral and topical.40
Bill Blackbeard, another, and no less eminent, American researcher, is violently opposed to this view. Challenging, and not without some bad faith, each of the conditions proposed by Kunzle, Blackbeard formulated the following definition:
A serially published, episodic, open-ended dramatic narrative or series of linked anecdotes about recurrent identified characters, told in successive drawings regularly enclosing ballooned dialogue or its equivalent and generally minimal narrative text.41
These two definitions are, to my understanding, both unacceptable. They are equally normative and self-interested, each made to measure in order to support an arbitrary slice of history. For example, the third of Kunzle’s conditions only serves to justify the fact that he chose the invention of printing as a starting point for The Early Comic Strip. While Blackbeard’s definition, which defends the thesis of the American origin for comics, applies only to printed comics and is destined to dismiss the entire field of comics that predates the appearance of the Yellow Kid in 1896.
In France, let us recall that Antoine Roux proposed a definition in six points in La Bande dessinée peut être éducative (Éd. de l’École 1970), a definition backhand-edly swept aside (and, here again, in part unjustly) by Yves Frémion in L’ABC de la BD, where one reads: “In ten years, none of these criteria, although a priori serious, has withstood history.”42
The difficulty of producing a valid definition of comics, a definition that permits discrimination in that which it is not but which excludes none of its historical manifestations, including its marginal or experimental visionaries (I am thinking, for example, of the works of Jean Teulé and of Martin Vaughn-James, where the reception can seem to be problematic), was indicated by Pierre Couperie in 1972:
Comics would be a story (but it is not necessarily a story . . .) constituted by handmade images from one or several artists (it must eliminate cinema and the photo-novel), fixed images (in difference from animation), multiple (contrary to the cartoon), and juxtaposed (in difference from illustration and engraved novels . . .). But this definition applies equally well to Trajan’s Column and the Bayeux tapestry.43
And Couperie adds that neither the framing of images, nor the use of the balloon, nor the mode of distribution are determining criteria.
So great is the diversity of what has been claimed as comics, or what is claimed today under diverse latitudes, that it has become almost impossible to retain any definitive criteria that is universally held to be true. I want to demonstrate this for two of the pertinent traits often erected as doctrinal elements:
the insertion, in the image, of verbal enunciations
the permanence, within the panels, of at least one identifiable character (a criterion notably insisted upon by Blackbeard)
Although used overwhelmingly, these elements must be seen to be contingent characteristics, suffering a number of exceptions. It follows that they can only produce reductive definitions.
Here, first of all, are some authors who have produced “mute” comics, that is to say, devoid of verbal enunciations, without dialogue or the narrational text (captions). Coming from Germany, this particular narrative form was widespread at the end of the nineteenth century with the pantomines of Caran d’Ache, K-Hito, or A. B. Frost, to mention not a single French, Spanish, or American artist. One later finds works “without words” in every category of comics: the daily comic strip and/or independent pages (Adamson by Oscar Jacobsson [1920]; The Little King by Otto Soglow [1931]; Vater und Sohn by e. o. plauen [1934]; Henry by Carl Anderson [1934]; Globi by J. K. Schiefe and R. Lips [1934]; Professeur Nimbus by André Daix [1934]; M. Subito by Robert Velter [1935]; Max l’explorateur by Guy Bara [1955]; etc.); complete stories published in the illustrated press (here the examples abound, recall only, among the successes, Allô! il est vivant by Raymond Poïvet [1964]; Sanguine by Philippe Caza [1976]; many episodes of Ken Parker by Milazzo and Berardi [mid-1980s]; Magic Glasses by Keko [1986]; or again the sketches of the German artist Sperzel, such as those that can be found in recent years in U-Comix and Kowalski); finally, in books, from Milt Gross (He Done Her Wrong [1930]) to Thierry Robin (La Teigne [1998]), passing through Moebius (Arzach [1975]), Crepax (La Lanterne Magique [1979]), Ana Juan (Requiem, with Gordillo [1985]), Avril and Petit-Roulet (Soirs de Paris [1989]), Hendrik Dorgathen (Space Dog [1993]), Alberto Breccia (Dracula, Dracul, Vlad?, bah . . . [1993]), Fabio (L’Oeil du Chat [1995]), Lewis Trondheim (La Mouche [1995]), Anna Sommer (Remueménage [1996]), and Peter Kuper (The System [1997]), and this list has no pretence to completeness.44
The permanence—and the present vitality—of this tradition does not prevent some researchers from asserting that “what distinguishes a comic from a cycle of frescoes is the fact that the written words are essential to the understanding of the story.”45 An amusing detail—and indicative of his blindness—the author next produced, in support of this observation, a Krazy Kat page in which the texts were masked, without seeming to notice that, unfortunately for him, the narration, developed in eleven images, remained perfectly intelligible despite the verbal amputation!
As for the presence of a recurrent character, there are diverse ways to bypass this. I will note six:
Thus, two dogmatic criteria, retained for the most part in current definitions of comics, must be dismissed. The difficulty encountered here is not particular to comics. It arises in almost identical terms for the most part, if not completely, in forms of modern art, like the cinema, and for forms where the evolution over the course of a century has smashed the traditional definition (novel, painting, music) into pieces. For example, Roger Odin shows clearly that it is almost impossible to express a definition of cinema that also applies to animated films and to all the forms of experimental or “widened” cinema. The aporia that the semiotician necessarily unblocks is thus described:
By what right do we exclude from cinema these productions when their authors present them explicitly as “films”? The fact that these productions do not enter into our definition of the “cinema,” is that a sufficient justification for this exclusion? If not, must we revise our definition of cinema in a more general-izable manner in order to integrate these counter-examples? But if so, where do we stop this generalization: at the absence of the film? At the absence of the screen? At the absence of the projector? Won’t we arrive at a sort of definition that tells us nothing about its object?46
Roger Odin suggests that it is necessary to surpass this immanent approach to cinema in order to take into account its social uses. No longer considering the “cinematic object” but the “cinematic field,” he concludes (p. 57) that “cinematic objects are definable objects, but variable objects in space and time.”
If one wishes to provide the basis of a reasonable definition for the totality of historical manifestations of the medium, and also for all of the other productions unrealized at this time but theoretically conceivable, one must recognize the relational play of a plurality of interdependent images as the unique ontological foundation of comics. The relationship established between these images admits several degrees and combines several operations, which I will distinguish later. But their common denominator and, therefore, the central element of comics, the first criteria in the foundational order, is iconic solidarity. I define this as interdependent images that, participating in a series, present the double characteristic of being separated—this specification dismisses unique enclosed images within a profusion of patterns or anecdotes—and which are plastically and semantically over-determined by the fact of their coexistence in praesentia.
No doubt giving the word “comics” such an extensive meaning is not without inconveniences. This is the danger noted by Pierre Couperie. From the steles, frescoes, and the ancient Egyptian books of the dead to the predellas of medieval painting, and from the Bayeux Tapestry to the polyptychs of every age, all the way to the pre-Colombian codex, the stations of the cross, the Emakimono (Japanese picture scrolls), storyboards for films and modern photo-novels, there are probably too many of these works of art that can find refuge in this potluck collection.47
Comics will encounter a problem similar to that which has long concerned the world of literature. Everyone admits that it is not sufficient to simply align words in order to make a literary work, for the reason that “of all the materials that humanity can utilize among others in the fine arts, language is perhaps the least specific, the least closely reserved to this end.”48 Resuming a debate begun in the time of Aristotle, Gérard Genette struggles to define the criteria of literarity, that is to say the conditions by which a text can be recognized as literary. I concede in the same way to the “essentialists” that it is not sufficient to simply align images, even interdependently, to produce a comic. Many other conditions can be legitimately debated, which would touch in priority, initially the “nature” of these images (their substance, their mode of production, their formal characteristics), followed by their mode(s) of articulation, eventually even the published form that they take, their distribution and the conditions of their reception—in short, everything that inscribes them in the specific process of communication.49 But it is improbable that unanimity will be reached on any of these conditions.
In reality, research on the essence of comics is not quite on the same order as that of a definition of literarity. The point is, in the second case, to separate the literary discourse from all the other forms of discourse, starting with day-to-day language. Literature is characterized by “a rupture with the ordinary regime of the language.” The clearly posed question from then on is to define “that which makes a verbal message a work of art,” according to the formulation of Roman Jakobson recalled by Genette. For the latter, the rupture can be analyzed in terms of fiction (in so far as a work of fiction develops in the reader an “aesthetic attitude” and a relative “disinterest” with regard to the real world), or perhaps in terms of diction, that is to say by the observation of formal traits that are “facts of style.” This opposition stretches to coincide with the division of the field of literature into “two great types: on the one hand fiction (dramatic or narrative), on the other lyric poetry, more and more often designated by the term poetry all told.”50
Comics rest on a device that is not known from familiar usage. It is not noted that everything can be expressed by this means—even if the practice of comics is, technically and financially speaking, available to everyone, as is confirmed by the aptitude of those children who devote themselves to it. One cannot help but compare it with other forms of creation (those, notably, that we have enumerated above) that participate with complete rights in the domains of art or fiction. Since comics are not based on a particular usage of a language, there is no place to define them in terms of diction. But neither are they bound exclusively with fictional forms, since there are examples of publicity or propagandistic comics, political and pedagogical comics, and, occasionally, comics journalism, where the concern is to inform or to testify. We can also add that the proliferation of autobiographical comics is a remarkable phenomenon of recent years, stemming from America, where the works of Robert Crumb, Art Spiegelman, and Harvey Pekar, notably, have opened the door. This plasticity of comics, which allows them to put in place messages of every order and narrations other than the fictional, demonstrates that before being an art, comics are well and truly a language.
But it is not necessary, at this stage of reflection, to push the concern for the delimitation of the medium further ahead. It will be enough for us that one cannot conceptualize comics without verifying the general rule, that of iconic solidarity. The necessary, if not sufficient, condition required to speak of comics is that the images will be multiple and correlated in some fashion.
This fact is empirically verified by whoever leafs through a comic book or comics magazine. What is put on view is always a space that has been divided up, compartmentalized, a collection of juxtaposed frames, where, to cite the fine formula of Henri Van Lier, a “multi-framed aircraft” sails in suspension, “in the white nothingness of the printed page.”51 A page of comics is offered at first to a synthetic global vision, but that cannot be satisfactory. It demands to be traversed, crossed, glanced at, and analytically deciphered. This moment-to-moment reading does not take a lesser account of the totality of the panoptic field that constitutes the page (or the double page), since the focal vision never ceases to be enriched by peripheral vision.
It is observable that the words for the French term “bandes dessinées” (drawn strips) itself implies a restrictive perception of the field that it is supposed to cover. The epithet, specifically assuming that the image will be the product of a drawing (dessin), seems to remove a priori all recourse to the photo, to typography, and even to painting. More seriously, the notion of the strip (bande) abusively privileges one of the components of the medium, the horizontal segment52 that sometimes constitutes a micro-story, sometimes nothing other than an ongoing continuing story, or only a portion of a page. If one believes Jean-Claude Glasser, the reign of this term is historically justified:
It is truly in the buildings of the Agence Opera Mundi that the expression “bande dessinée” was formed [in the 1930s], then progressively imposed itself. . . . It remained to designate the daily strips . . . which explains why it is not found in the illustrated magazines (illustrés) of the age where the Sunday pages predominated. . . . It is only in the 1950s that it ceased to apply only to daily strips.53
But what was formerly nothing but a lexical generalization has become a veritable impropriety. Now that the book [album] is, in Europe, the preponderant vehicle for comics, it follows that the page is the technical unit, market and aesthetic reference.54
Iconic solidarity is only the necessary condition so that visual messages can, in first approximation, be assimilated within a comic. As a physical object, every comic can be described as a collection of separate icons and interdependent images. If one considers any given production, one quickly notices that comics that satisfy this minimal condition are naturally longer, but also that they do not all obey the same intentions and do not mobilize the same mechanisms. All theoretical generalizations are cognizant of the trap of dogmatism. Far from wanting to defend a school of thought, an era or a standard against others, or again to prescribe any recipes, I want to force myself to note the diversity of all forms of comics and spare my reflections from any normative character.
That is why I have chosen the notion of the system, which defines an ideal, as emblematic of this reflection. The comics system will be a conceptual frame in which all of the actualizations of the “ninth art” can find their place and be thought of in relation to each other, taking into account their differences and their commonalities within the same medium. In this meaning, the notion of the system, “an ensemble of things that are held” (Littré), advances the fundamental concept of solidarity.
It is important now to define the exact nature of this iconic solidarity. Indeed, comics submit the images of which they are composed to different sorts of relations. To describe the entirety of these relations, I will use a generic term with a very broad meaning: namely, arthrology (from the Greek arthron: articulation).55
Every drawn image is incarnated and is displayed in a space. The fixed image, contrary to the moving image of cinema, which Gilles Deleuze has shown is at the same time a “movement-image” and a “time-image,”56 only exists in a single dimension. Comics panels, situated relationally, are, necessarily, placed in relation to space and operate on a share of space. These are the fundamental principles of this spatial distribution that will be examined at the sign of the spatio-topia, a term created by gathering, while maintaining distinct, the concept of space (espace) and that of place (lieu).57 The specific spaces of comics, like the word balloon (bulle), the panel and its frame, the strip (the horizontal band that is the first level of arrangement for the panels), and the page will be successively summoned, and their interactions analyzed.
The precedence accorded to the order of spatial and topological relations goes against most widespread opinion, which holds that, in comics, spatial organization will be totally pledged to the narrative strategies, and commanded by them. The story will create or dictate, relative to its development, the number, the dimension, and the disposition of panels. I believe on the contrary that, from the instant that an author begins the comics story that he undertakes, he thinks of this story, and his work still to be born, within a given mental form with which he must negotiate. This form is precisely the spatio-topical apparatus, one of the keys to the system of comics, a complex of units, parameters, and functions that it is up to us to describe. The taking into account of the form and the preconception of the mode of spatial organization that will be adopted are, as I hope to demonstrate, the preliminary conditions to every beginning, and the constraints that never cease to inform each phase of creation. From the moment of sketching the first panel of a comic, the author has always already taken, as for the behavior of engaging with the medium, some large strategic options (evidently modifiable by what follows), which concern the distribution of spaces and the occupation of places.58 It will belong to the page layout (mise-en-page) to specify these options and to provide each page with its definitive configuration.
But comics is not only an art of fragments, of scattering, of distribution; it is also an art of conjunction, of repetition, of linking together. Within the spatio-topical operation—that is, within the space that comics appropriates and develops—one can distinguish two degrees in the relations between the images. The elementary relations, of the linear type, compose what we will call the restricted arthrology. Governed by the operation of breaking down (decoupage), they put in place the sequential syntagms, which are most often subordinated to the narrative ends. It is at this level that writing takes priority, as a complementary function of narration. The other relations, translinear or distant, emerge from general arthrology and decline all of the modalities of braiding (tressage). They represent a more elaborated level of integration between the narrative flux (which can also be called the narrative energy or, again, to adopt an expression from Hubert Damisch, the “story shuttle” [navette du récit]) and the spatio-topical operation, in which the essential component, as Henri Van Lier has named it, is the “multiframe” (multicadre).
This is not, on the one side, a comparison of spaces that will adopt the spatiotopia, and on the other a comparison of content that comes out of arthrology. The articulations of the comics discourse are indistinguishable from the content-incarnated-in-space, or, if one prefers, the spaces-invested-with-content. Thus, the spatio-topia is a part of arthrology, an arbitrarily detached subgroup, with no other autonomy than that which it recognizes for itself, at a given moment, to the heuristic ends. Indeed, it is useful, in order to apprehend certain levels of the functionality of the comics language, to intellectually conceive of this reduction of the page as an assemblage of frames and empty bubbles. In reality, this assemblage is in no way observable as such, and does not preexist, in an already elaborated form, the final, complete version of the page.
Yet, it seems to me, the study of the system of comics must come to terms with the spatio-topia. This precedence is not justified, I wish to add, by the chronology of successive operations carried out in the course of a process of elaborating a comic. It holds to the preexistence of that which I have called a “mental form.” A scenario destined for comics (but what follows applies a fortiori to an improvised comic without a preliminary scenario—as is found, for example, in The Airtight Garage of Jerry Cornelius by Moebius) is not constructed in a purely abstract and speculative manner. It cannot be developed except in a dialogue with a certain preliminary idea of the medium, of its nature, of its competencies and its prescriptions. We must invent a scenario that can be incarnated in this medium (or, sometimes, adapt itself in function to a preexisting story), to make the best usage.59 This general and diffuse representation of comics, on which creation rests, withholds the spatio-topical components of the system, because they simultaneously constitute the framework and the base. The spatio-topia is the point of view that can be had on comics before thinking about any single comic, and starting from which it is possible to think about a new performance of the medium.
Very quickly, when elaborating the contents, when a discourse invests the multiframe, the question of linkages and of articulations will become preponderant. To articulate the frames is the process of page layout. Breakdown and page layout are the two fundamental operations of arthrology, on which the braiding eventually puts the finishing touches. The one and the other, however, help themselves to elements that stem at first from the spatio-topia. It is evident for the page layout, whose own role is to define a share of space, founded on two essential and complementary functions of the frame (that I will develop further along), the separative function and the readerly function.
Thus, one can define the mode of interaction between the authority of the spatio-topia and the arthrology as “dialogic” and “recursive.” Edgar Morin, from whom I borrowed these notions, defines them in the following way. The dialogic is “complex associations of necessary authority essential for the existence of a phenomenon.” The phenomena of “reciprocal feedback” can be qualified as recursive between the authorities that “are inter-regulated amongst themselves,” such as that “the effects and the products are at the same time causal and productive.”60 This is what I hope to establish with respect to the complex degree of interaction that underlies the comics system.
From this perspective, the privilege frequently accorded by critics and theoreticians to certain processes assumed to be specific to comics no doubt need to be revised. I point, for example, in Bande dessinée et figuration narrative (a book which can be seen as foundational), to a passage asserting that 80 percent of comics artists “neglect the techniques of page layout and breakdown that are specific to it.”61 (“It” refers to the language of the ninth art.) However, comics rarely mobilize truly specific processes and techniques. On the other hand, all comics, even those that provide the simplest appearance, are particular avatars of a system in which the components, and their interactions, draw a complex and unpublished totality. It is this system that we are now going to dissect.
I would like, at the moment of closing this introduction, to express my debt to Benoît Peeters and Thierry Smolderen, who have contributed in a decisive fashion to the maturation of certain ideas formulated here. I must also thank, for the precious commentaries that they have happily formulated on many of these pages, Gilles Ciment, Pascal Lefèvre, and Bernard Magné.